ChangeLog revision 446f62d6
1commit 36703f66c3b06875651606a6280d5dc9d9dad51e 2Author: Michel Dänzer <michel.daenzer@amd.com> 3Date: Tue Mar 19 18:01:02 2019 +0100 4 5 Bump version for 19.0.1 release 6 7commit 4407c78bd86da4460ee07a15a365e07d99e0dd27 8Author: Dave Airlie <airlied@redhat.com> 9Date: Thu Mar 14 11:17:32 2019 +0100 10 11 modesetting: add tile property support 12 13 This adds tiling support to the driver, it retrieves the tile info from 14 the kernel and translates it into the server format and exposes the 15 property. 16 17 (Ported from xserver commits 8fb8bbb3062f1a06621ab7030a9e89d5e8367b35 18 and 6abdb54a11dac4e8854ff94ecdcb90a14321ab31) 19 (Ported from amdgpu commit 6ee857726166f495abcd68e4ff60e3a09593d079) 20 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 21 22commit 79bc0e054f37026377d54cac6cd8127d4aa9baca 23Author: Michel Dänzer <michel.daenzer@amd.com> 24Date: Thu Mar 14 11:10:51 2019 +0100 25 26 Use radeon_finish in drmmode_crtc_scanout_update 27 28 radeon_glamor_finish only works if we're using glamor, otherwise it'll 29 crash. 30 31 Fixes: ce7db51020d3 "Cancel pending scanout update in drmmode_crtc_scanout_update" 32 Bug: https://bugs.debian.org/924540 33 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 34 35commit f6cd4a67d7de5d3ff1a6e58a8c83749fc8ffc310 36Author: Michel Dänzer <michel.daenzer@amd.com> 37Date: Fri Mar 8 11:48:10 2019 +0100 38 39 Revert "glamor: Avoid glamor_create_pixmap for pixmaps backing windows" 40 41 This reverts commit 274703087f80342f51fa69c935bb9a1cb0c4ae47. 42 43 Reports of visual corruption were bisected to this, e.g. 44 https://bugs.archlinux.org/task/61941 . I can reproduce this with Turks, 45 but not with Bonaire. I assume it's a Mesa/glamor bug, but let's revert 46 for now. 47 48 Acked-by: Alex Deucher <alexander.deucher@amd.com> 49 50commit 0d132d99e0b750896a78f47d73a8639680495d8c 51Author: Michel Dänzer <michel.daenzer@amd.com> 52Date: Wed Mar 6 17:48:03 2019 +0100 53 54 Bump version for 19.0.0 release 55 56commit c301b8af25d2c2cd49035a4395ebe6c3612df366 57Author: Michel Dänzer <michel.daenzer@amd.com> 58Date: Fri Mar 1 18:28:11 2019 +0100 59 60 dri2: Call drm_queue_handle_deferred in dri2_deferred_event 61 62 drm_queue_handler just puts the event on the signalled list; without 63 calling drm_queue_handle_deferred, actual processing of the event may be 64 delayed indefinitely, e.g. until another event arrives from the kernel. 65 66 This could result in DRI2 clients hanging during DPMS off. 67 68 Fixes: ba83a866af5a "Add radeon_drm_handle_event wrapper for 69 drmHandleEvent" 70 (Ported from amdgpu commit 09be74a3d1dd9604336d9a27f98d132b262dcbaf) 71 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 72 73commit 705020b6247eaa062edc9c88e6ad52f8c5468051 74Author: Michel Dänzer <michel.daenzer@amd.com> 75Date: Fri Mar 1 18:23:30 2019 +0100 76 77 present: Check that flip and screen pixmap pitches match 78 79 If they don't, flipping will result in corrupted display. 80 81 Test case: 82 83 * Run Xorg at 1920x1080 with no window manager 84 * glxgears -geometry 2048x1080 85 86 The Present extension code in xserver 1.21 will check for this. 87 88 (Ported from amdgpu commit a636f42b496b0604ca00a144690ece61d1a88a27) 89 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 90 91commit 15697ee242c30b9ea6775624e8282e0171a113a7 92Author: Michel Dänzer <michel.daenzer@amd.com> 93Date: Mon Jan 28 18:27:10 2019 +0100 94 95 Keep waiting for a pending flip if drm_handle_event returns 0 96 97 drm_wait_pending_flip stopped waiting if drm_handle_event returned 0, 98 but that might have processed only some unrelated DRM events. As long as 99 the flip is pending, we have to keep waiting for its completion event. 100 101 Noticed while working on the previous fix. 102 103 (Ported from amdgpu commit 9045fb310f88780e250e60b80431ca153330e61b) 104 105commit 227123de3d862e691131708b7f55260bee17f2b7 106Author: Michel Dänzer <michel.daenzer@amd.com> 107Date: Mon Jan 28 18:24:41 2019 +0100 108 109 Call drmHandleEvent again if it was interrupted by a signal 110 111 drmHandleEvent can be interrupted by a signal in read(), in which case 112 it doesn't process any events but returns -1, which 113 drm_handle_event propagated to its callers. This could cause the 114 following failure cascade: 115 116 1. drm_wait_pending_flip stopped waiting for a pending flip. 117 2. Its caller cleared drmmode_crtc->flip_pending before the flip 118 completed. 119 3. Another flip was attempted but got an unexpected EBUSY error because 120 the previous flip was still pending. 121 4. TearFree was disabled due to the error. 122 123 The solution is to call drmHandleEvent if it was interrupted by a 124 signal. We can do that in drm_handle_event, because when that is called, 125 either it is known that there are events ready to be processed, or the 126 caller has to wait for events to arrive anyway. 127 128 Bugzilla: https://bugs.freedesktop.org/109364 129 (Ported from amdgpu commit 3ff2cc225f6bc08364ee007fa54e9d0150adaf11) 130 131commit 1bfdccf7639ee2f655dc659cafa63830ba28be85 132Author: Michel Dänzer <michel.daenzer@amd.com> 133Date: Mon Jan 28 18:11:10 2019 +0100 134 135 Only update drmmode_crtc->flip_pending after actually submitting a flip 136 137 And only clear it if it matches the framebuffer of the completed flip 138 being processed. 139 140 Fixes 141 142 (WW) RADEON(0): flip queue failed: Device or resource busy 143 (WW) RADEON(0): Page flip failed: Device or resource busy 144 (EE) RADEON(0): present flip failed 145 146 due to clobbering drmmode_crtc->flip_pending. 147 148 Reproducer: Enable TearFree, run warzone2100 fullscreen, toggle 149 Vertical sync on/off under Video Options. Discovered while investigating 150 https://bugs.freedesktop.org/109364 . 151 152 (Ported from amdgpu commit e72a02ba1d35743fefd939458b9d8cddce86e7f5) 153 154commit dcd3527299c1f6d6faa401c565fa884f4d8f3287 155Author: Michel Dänzer <michel.daenzer@amd.com> 156Date: Mon Jan 28 18:08:35 2019 +0100 157 158 Don't allow TearFree scanout flips to complete in the same vblank period 159 160 We were using a relative target of 0, meaning "complete the flip ASAP". 161 This could result in the flip sometimes, but not always completing in 162 the same vertical blank period where the corresponding drawing occurred, 163 potentially causing judder artifacts with applications updating their 164 window contents synchronized to the display refresh. A good way to test 165 this is the vsynctester.com site in a windowed browser, where the judder 166 results in the large "VSYNC" text intermittently appearing red or cyan 167 instead of the expected gray. 168 169 To avoid this, use a relative target MSC of 1, meaning that if a 170 vertical blank period is in progress, the flip will only complete in the 171 next one. 172 173 Reported by Julian Tempel and Brandon Wright in 174 https://bugs.freedesktop.org/106175 . 175 176 (Ported from amdgpu commit a1b479c7d0066c481af920f297d6af9009dda11e) 177 178commit 274703087f80342f51fa69c935bb9a1cb0c4ae47 179Author: Michel Dänzer <michel.daenzer@amd.com> 180Date: Mon Jan 28 18:06:50 2019 +0100 181 182 glamor: Avoid glamor_create_pixmap for pixmaps backing windows 183 184 If the compositing manager uses direct rendering (as is usually the case 185 these days), the storage of a pixmap allocated by glamor_create_pixmap 186 needs to be reallocated for sharing it with the compositing manager. 187 Instead, allocate pixmap storage which can be shared directly. 188 189 (Ported from amdgpu commit bf326f2ea19daa6c8da23d6788ff301ae70b8e69) 190 191commit 6d1dfe2523e900517bd1e8743c87d6990a82c800 192Author: Michel Dänzer <michel.daenzer@amd.com> 193Date: Mon Jan 28 18:04:41 2019 +0100 194 195 dri2: Flush in dri2_create_buffer2 after calling glamor_set_pixmap_bo 196 197 To make sure the client can't use the shared pixmap storage for direct 198 rendering first, which could produce garbage. 199 200 Bugzilla: https://bugs.freedesktop.org/109235 201 (Ported from amdgpu commit ebd32b1c07208f8dbe853e089f5e4b7c6a7a658a) 202 203commit 77d7abf46446522e686c6b6f1e4857458589ef37 204Author: Michel Dänzer <michel.daenzer@amd.com> 205Date: Mon Jan 28 18:00:20 2019 +0100 206 207 dri3: Flush if necessary in dri3_fd_from_pixmap 208 209 To make sure the client can't use the shared pixmap storage for direct 210 rendering first, which could produce garbage. 211 212 Bugzilla: https://bugs.freedesktop.org/109235 213 (Ported from amdgpu commit d168532ee739f7e33a2798051e64ba445dd3859f) 214 215commit b1c01698f577577e4a88bad0ae08fb5d998e7ebb 216Author: Michel Dänzer <michel.daenzer@amd.com> 217Date: Wed Jan 9 11:26:14 2019 +0100 218 219 Only call drmmode_validate_leases if RandR is enabled 220 221 It would crash if RandR is disabled, e.g. because Xinerama is enabled. 222 223 Bugzilla: https://bugs.freedesktop.org/109230 224 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 225 226commit 38db1bbcfc019c92884c7819a6630c70e543f6b2 227Author: Michel Dänzer <michel.daenzer@amd.com> 228Date: Wed Jan 9 11:33:18 2019 +0100 229 230 Only call drmmode_uevent_init if RandR is enabled 231 232 There's no point in listening for hotplug events if RandR is disabled, 233 as there's no other mechanism for them to be propagated. We were already 234 mostly ignoring them in that case. 235 236 Inspired by 237 https://gitlab.freedesktop.org/xorg/driver/xf86-video-intel/commit/1a489142c8e6a4828348cc9afbd0f430d3b1e2d8 238 (via https://bugs.freedesktop.org/109230#c11). 239 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 240 241commit 803f872f7d4b2d80be434bb42ce64dfd295b122c 242Author: Michel Dänzer <michel.daenzer@amd.com> 243Date: Fri Dec 21 18:04:21 2018 +0100 244 245 Use two HW cursor buffers per CRTC 246 247 Switch to the other buffer when xf86_config->cursor changes. Avoids 248 these issues possible when re-using the same buffer: 249 250 * The HW may intermittently display a mix of the old and new cursor 251 images. 252 * If the hotspot changes, the HW may intermittently display the new 253 cursor image at the location corresponding to the old image's hotspot. 254 255 Bugzilla: https://bugs.freedesktop.org/108832 256 (Ported from amdgpu commit 0d60233d26ec70d4e1faa343b438e33829c6d5e4) 257 258commit 91e557f78ad261e76a1829f54722c2c0781742d2 259Author: Michel Dänzer <michel.daenzer@amd.com> 260Date: Fri Dec 21 18:00:09 2018 +0100 261 262 Update cursor position in drmmode_show_cursor if hotspot changed 263 264 The cursor position is updated to be consistent with the new hotspot in 265 the same ioctl call. 266 267 (Ported from amdgpu commit b04697de5270e8e45744a7025c24df1f454a4cf0) 268 269commit 92df709786830d4e30a106dd49d8e0355c50c8f0 270Author: Michel Dänzer <michel.daenzer@amd.com> 271Date: Fri Dec 21 17:54:36 2018 +0100 272 273 Use drmIoctl in drmmode_show_cursor 274 275 This should be functionally equivalent to what drmModeSetCursor(2) do 276 behind the scenes, but allows for new tricks in following changes. 277 278 (Ported from amdgpu commit b344e1559e936046ef02c777fc4f6bcefa3830bc) 279 280commit e14c3d2f86c7be2b5c3d06a47bf0abe954207d0b 281Author: Michel Dänzer <michel.daenzer@amd.com> 282Date: Fri Dec 21 17:38:51 2018 +0100 283 284 Drop RADEONInfoRec::cursor_bo array 285 286 Not needed or even useful for anything. 287 288 (Ported from amdgpu commit e95044e45350870fa7e237860e89ade91ac03550) 289 290commit f66254c171f5a3b052a2a9e0339f17dfb5a60dc2 291Author: Michel Dänzer <michel.daenzer@amd.com> 292Date: Fri Dec 21 12:49:27 2018 +0100 293 294 Automatically try re-enabling TearFree after a flip failed 295 296 Specifically, after both the page flip and vblank ioctls failed, but 297 then the vblank ioctl started working again. This can happen 298 intermittently e.g. when hotplugging a DP display. Previously, TearFree 299 would stay disabled in that case until a modeset was triggered somehow. 300 301 Bugzilla: https://bugs.freedesktop.org/103791 302 (Ported from amdgpu commit bcfa6c258fdf41a9928f8a3c78fc528d0fafee25) 303 304commit ce7db51020d32f17e442338bfd305220feb51630 305Author: Michel Dänzer <michel.daenzer@amd.com> 306Date: Fri Dec 21 12:47:18 2018 +0100 307 308 Cancel pending scanout update in drmmode_crtc_scanout_update 309 310 drmmode_crtc_scanout_update does the equivalent of a scanout update, 311 so no need to do it again. This might also avoid issues if there's a 312 pending scanout update at this point. 313 314 (Ported from amdgpu commit 4e7a24ac5a64e402146953ec5850d13c05742116) 315 316commit 21d65e5b78c8889e363aee8596cd0b0f942fee46 317Author: Michel Dänzer <michel.daenzer@amd.com> 318Date: Fri Dec 21 12:38:04 2018 +0100 319 320 Perform scanout buffer update immediately if drmmode_wait_vblank fails 321 322 Otherwise the damaged screen contents may never be displayed in that 323 case. 324 325 (Ported from amdgpu commit 500fadb16285146e91f62fce3a0ce1360ca684ba) 326 327commit 64942d2c49f9fa1afcc42c07943d076a40963e51 328Author: Michel Dänzer <michel.daenzer@amd.com> 329Date: Thu Dec 20 19:00:08 2018 +0100 330 331 Move deferred vblank events to separate drm_vblank_deferred list 332 333 It was still possible for nested xorg_list_for_each_entry_safe loops 334 to occur over the drm_vblank_signalled list, which could mess up that 335 list. Moving deferred events to a separate list allows processing the 336 drm_vblank_signalled list without xorg_list_for_each_entry_safe. 337 338 Bugzilla: https://bugs.freedesktop.org/108600 339 (Ported from amdgpu commit 51ba6dddee40c3688d4c7b12eabeab516ed153b7) 340 341commit f450632077843a95a6ef269febbfb64a605045ed 342Author: Michel Dänzer <michel.daenzer@amd.com> 343Date: Thu Dec 20 18:48:19 2018 +0100 344 345 Explicitly keep track of whether a DRM event is for a flip or not 346 347 When an async flip is performed, and TearFree is enabled on the CRTC 348 used for timing, we schedule a vblank event for completing the page 349 flip. The DRM event queuing code treated this event like a vblank event, 350 but it needs to be treated like a page flip event. 351 352 (Ported from amdgpu commit e2c7369cae65069aa93eed1c0b678f975ce5c274) 353 354commit 189b6facb3988c00c96d970f8c13ed8d58fa3998 355Author: Michel Dänzer <michel.daenzer@amd.com> 356Date: Thu Dec 20 18:44:24 2018 +0100 357 358 Use drm_abort_one in drm_queue_handler 359 360 At this point, we've already established that e->handler is NULL, no 361 need to check again in drm_queue_handle_one. This also makes it clearer 362 what's happening. 363 364 (Ported from amdgpu commit eda571222f5a6be47f8897e82d85199bb9d95251) 365 366commit 06957e3067b9eb38d30b32d98a07895399f4cb36 367Author: Michel Dänzer <michel.daenzer@amd.com> 368Date: Thu Dec 20 18:40:19 2018 +0100 369 370 glamor: Can work at depth >= 15 with current xserver Git master 371 372 (Ported from amdgpu commit 0734cdf544ffd3f2ac8749ad0e4bf43f8a5cea50) 373 374commit 99ac121770da53196124d80375a5c8edbcf827fa 375Author: Michel Dänzer <michel.daenzer@amd.com> 376Date: Thu Dec 20 18:38:33 2018 +0100 377 378 Skip gamma correction of cursor data if premultiplied R/G/B > alpha 379 380 The un-premultiplied R/G/B values would overflow the gamma LUT, so just 381 pass through the data unchanged, and leave it up to the HW how to 382 interpret such weird premultiplied alpha pixels. 383 384 Bugzilla: https://bugs.freedesktop.org/108355 385 (Ported from amdgpu commit 13c94a373b4858a2d2aa14c22b5f98d53c84c0d9) 386 387commit 0c40a76d1c050d018e6d59bebb5efc9c62be308c 388Author: Michel Dänzer <michel.daenzer@amd.com> 389Date: Thu Dec 20 18:25:21 2018 +0100 390 391 Detect and fix up non-premultiplied cursor data 392 393 X server >= 1.18 already has code for this, but handle it with older X 394 servers as well. 395 396 (Ported from amdgpu commits ad6dfb0124860cf67730bde85867f81d9258c84d & 397 426f9a49655f01863cf4d898f525e5f95984e0c4) 398 399commit 0058fd2ebf4c900b12f129984e98886a7ac84b2f 400Author: Mario Kleiner <mario.kleiner.de@gmail.com> 401Date: Tue Dec 18 11:40:35 2018 +0100 402 403 Fix crash when page flipping in multi-X-Screen/Zaphod mode 404 405 radeon_do_pageflip() indexed the flipdata->fb[] array 406 indexing over config->num_crtc, but the flip completion 407 routines, e.g., drmmode_flip_handler(), index that array 408 via the crtc hw id from drmmode_get_crtc_id(crtc). 409 410 This is mismatched and causes indexing into the wrong 411 array slot at flip completion -> Server crash. 412 413 Always use drmmode_get_crtc_id(crtc) for indexing into 414 the array to fix this. 415 416 Tested on a dual-X-Screen setup with one video output 417 assigned to each X-Screen, page-flipping an OpenGL app 418 on either of both X-Screens. This used to crash when 419 flipping on X-Screen 1, now it doesn't anymore. 420 421 Fixes: 740f0850f1e4 "Store FB for each CRTC in drmmode_flipdata_rec" 422 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 423 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 424 425commit b559fd83b2b026b50fe39a2b69bdc0901fd3330c 426Author: Michel Dänzer <michel.daenzer@amd.com> 427Date: Mon Nov 19 11:38:22 2018 +0100 428 429 Post-release version bump 430 431commit 33cbd449b5305b8e8c0319af8becc1ab1832c101 432Author: Alan Coopersmith <alan.coopersmith@oracle.com> 433Date: Sun Nov 18 12:22:16 2018 -0800 434 435 Update README for gitlab migration 436 437 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 438 439commit 5e6fa5c17a810127f0049816f20db6b871ca77e0 440Author: Michel Dänzer <michel.daenzer@amd.com> 441Date: Wed Oct 24 18:22:05 2018 +0200 442 443 Allow up to six instances in Zaphod mode 444 445 Corresponding to up to six CRTCs being available in the hardware. 446 447 (Ported from amdgpu commit c9d43c1deb9a9cfc41a8d6439caf46d12d220853) 448 449commit c480fd066fe129fa6561fca8c09f037613b753e8 450Author: Michel Dänzer <michel.daenzer@amd.com> 451Date: Wed Oct 24 18:19:42 2018 +0200 452 453 Make wait_pending_flip / handle_deferred symmetric in set_mode_major 454 455 We were always calling the latter, but not always the former, which 456 could result in handling deferred DRM events prematurely. 457 458 (Ported from amdgpu commit 955373a3e69baa241a1f267e96d04ddb902f689f) 459 460commit 86fe8d27b9a3f043e5288ce50eaf1f5fffd24516 461Author: Michel Dänzer <michel.daenzer@amd.com> 462Date: Wed Oct 24 18:14:02 2018 +0200 463 464 Handle pending scanout update in drmmode_crtc_scanout_free 465 466 We have to wait for a pending scanout flip or abort a pending scanout 467 update, otherwise the corresponding event handler will likely crash 468 after drmmode_crtc_scanout_free cleaned up the data structures. 469 470 Fixes crash after VT switch while dedicated scanout pixmaps are enabled 471 for any CRTC. 472 473 (Ported from amdgpu commit 0cd2c337d2c02b8ec2bd994d6124b4aaaad10741) 474 475commit f892d3791219d1041e0cbb1b866e15774004aa18 476Author: Michel Dänzer <michel.daenzer@amd.com> 477Date: Mon Oct 15 17:14:41 2018 +0200 478 479 dri3: Handle radeon_get_pixmap_bo returning NULL 480 481 We were trying to already, but testing the wrong pointer. 482 483 Fixes: b85b7b11f5b5 "Add struct radeon_buffer" 484 Bug: https://bugs.debian.org/910846 485 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 486 487commit 80514b7b1b88f46f3965ff023035dea83b57e652 488Author: Michel Dänzer <michel.daenzer@amd.com> 489Date: Fri Sep 14 17:48:49 2018 +0200 490 491 Bump version for 18.1.0 release 492 493commit 2d58830c3feafc54dccc0b7bf761a466437d4a09 494Author: Michel Dänzer <michel.daenzer@amd.com> 495Date: Wed Sep 5 18:23:29 2018 +0200 496 497 Fix uninitialized use of local variable pitch in radeon_setup_kernel_mem 498 499 Fixes server reset. 500 501 Pointed out by clang: 502 503 ../../src/radeon_kms.c:2721:9: warning: variable 'pitch' is used uninitialized whenever 'if' condition is false [-Wsometimes-uninitialized] 504 if (!info->front_buffer) { 505 ^~~~~~~~~~~~~~~~~~~ 506 ../../src/radeon_kms.c:2765:27: note: uninitialized use occurs here 507 pScrn->displayWidth = pitch / cpp; 508 ^~~~~ 509 510commit 5d5d883496842da84d9418e91cb13454751da625 511Author: Michel Dänzer <michel.daenzer@amd.com> 512Date: Fri Sep 7 18:16:22 2018 +0200 513 514 Bail early from drm_wait_pending_flip if there's no pending flip 515 516 No need to process any events in that case. 517 518 (Ported from amdgpu commit ca5eb9894fff153c0a1df7bdc4a4745713309e27) 519 520commit 4c7d5e50a5e469a541bc463cecb505fe850c0824 521Author: Michel Dänzer <michel.daenzer@amd.com> 522Date: Wed Sep 5 11:29:43 2018 +0200 523 524 Don't use xorg_list_for_each_entry_safe for signalled flips 525 526 drm_wait_pending_flip can get called from drm_handle_event, in which 527 case xorg_list_for_each_entry_safe can end up processing the same entry 528 in both. To avoid this, just process the first list entry until the list 529 is empty. 530 531 (Ported from amdgpu commit 26770be44b89b83bf39c28f2fe284c8cb92ed0c0) 532 533commit 87b9a3e516d19dd1b89a64f6cac990fae53fc1b3 534Author: Michel Dänzer <michel.daenzer@amd.com> 535Date: Wed Sep 5 11:27:25 2018 +0200 536 537 Always delete entry from list in drm_queue_handler 538 539 We left entries without a handler hook in the list, so the list could 540 keep taking longer to process and use up more memory. 541 542 (Ported from amdgpu commit 7eea3e2cd74eed22e982319144e18ae5b1087b78) 543 544commit de88ea2755611bdcb18d91d8234d2ab5be8ff2e9 545Author: Michel Dänzer <michel.daenzer@amd.com> 546Date: Fri Aug 24 11:00:16 2018 +0200 547 548 glamor: Handle ihandle == -1 in radeon_glamor_set_shared_pixmap_backing 549 550commit 84bde2df5b453f8aab35aa18b0cf1f20b8d20488 551Author: Michel Dänzer <michel.daenzer@amd.com> 552Date: Fri Aug 24 10:58:51 2018 +0200 553 554 EXA: Handle ihandle == -1 in RADEONEXASharePixmapBacking 555 556commit 1799680f7bd84e0618f34f4c7486799521ddaf83 557Author: Michel Dänzer <michel.daenzer@amd.com> 558Date: Fri Aug 24 10:56:58 2018 +0200 559 560 Handle ihandle == -1 in radeon_set_shared_pixmap_backing 561 562 It means to stop using the shared pixmap backing. 563 564 Fixes crash when changing PRIME slave output configuration. 565 566commit 3c42bd04800211f64a17640c7ce7c4111ffd4978 567Author: Michel Dänzer <michel.daenzer@amd.com> 568Date: Fri Aug 24 10:55:49 2018 +0200 569 570 EXA: Handle NULL BO pointer in radeon_set_pixmap_bo 571 572commit cba8fe4d64819aaa8ba516aa68dbe6d2aa153046 573Author: Michel Dänzer <michel.daenzer@amd.com> 574Date: Thu Aug 23 18:59:18 2018 +0200 575 576 Use AC_CONFIG_MACRO_DIR instead of AC_CONFIG_MACRO_DIRS 577 578 Older versions of autoconf only supported the former. 579 580 Acked-by: Alex Deucher <alexander.deucher@amd.com> 581 582commit 7b01c10137aba24c8f61dd9b2a19ea257ad24371 583Author: Michel Dänzer <michel.daenzer@amd.com> 584Date: Thu Aug 23 18:42:28 2018 +0200 585 586 Add m4 directory 587 588 Although normally it only warns about it, under some circumstances, 589 aclocal can error out if this directory doesn't exist. 590 591 Reported-by: John Lumby <johnlumby@hotmail.com> 592 Acked-by: Alex Deucher <alexander.deucher@amd.com> 593 594commit 60cd28bbbe4f9db391b4f27c5ff9289d5c2bef40 595Author: Michel Dänzer <michel.daenzer@amd.com> 596Date: Thu Aug 16 18:06:31 2018 +0200 597 598 Remove drmmode_crtc_private_rec::present_vblank_* related code 599 600 Not needed anymore with the more robust mechanisms for preventing nested 601 drmHandleEvent calls introduced in the previous changes. 602 603 (Ported from amdgpu commit 85cd8eef0cbed7b409b07f58d76dacd34aa3ddea) 604 Acked-by: Alex Deucher <alexander.deucher@amd.com> 605 606commit c42f6e2e61d166c8d3ef3fcad175d7050a00288b 607Author: Michel Dänzer <michel.daenzer@amd.com> 608Date: Fri Jul 20 16:56:22 2018 +0200 609 610 Defer vblank event handling while waiting for a pending flip 611 612 This is to avoid submitting more flips while we are waiting for pending 613 ones to complete. 614 615 (Ported from amdgpu commit e52872da69ecc84dafb3355839e35b0383f0d228) 616 Acked-by: Alex Deucher <alexander.deucher@amd.com> 617 618commit ba83a866af5a3784fc4822614375cc081e93197c 619Author: Michel Dänzer <michel.daenzer@amd.com> 620Date: Thu Aug 16 17:44:45 2018 +0200 621 622 Add radeon_drm_handle_event wrapper for drmHandleEvent 623 624 Instead of processing DRM events directly from drmHandleEvent's 625 callbacks, there are three phases: 626 627 1. drmHandleEvent is called, and signalled events are re-queued to 628 _signalled lists from its callbacks. 629 2. Signalled page flip completion events are processed. 630 3. Signalled vblank events are processed. 631 632 This should make sure that we never call drmHandleEvent from one of its 633 callbacks, which would usually result in blocking forever. 634 635 (Ported from amdgpu commit 739181c8d3334ff14b5a607895dfdeb29b0d9020) 636 Acked-by: Alex Deucher <alexander.deucher@amd.com> 637 638commit 93621e408c17dd9e082236c17f051c06558d7f4d 639Author: Michel Dänzer <michel.daenzer@amd.com> 640Date: Thu Aug 16 17:27:06 2018 +0200 641 642 Add radeon_drm_wait_pending_flip function 643 644 Replacing the drmmode_crtc_wait_pending_event macro. 645 646 (Ported from amdgpu commit 6029794e8a35417faf825491a89b85f713c77fc1) 647 Acked-by: Alex Deucher <alexander.deucher@amd.com> 648 649commit ecdf0b7ec9378bc386ce8276f16fb16d09f72812 650Author: Michel Dänzer <michel.daenzer@amd.com> 651Date: Thu Aug 16 17:14:42 2018 +0200 652 653 Move DRM event queue related initialization to radeon_drm_queue_init 654 655 And make radeon_drm_queue_handler not directly accessible outside of 656 radeon_drm_queue.c. 657 658 (Ported from amdgpu commit 0148283984c77f7a6e97026edc3093497547e0a4) 659 Acked-by: Alex Deucher <alexander.deucher@amd.com> 660 661commit 824189b3da9edc33e1a4f5c6130a043da73c1a4c 662Author: Michel Dänzer <michel.daenzer@amd.com> 663Date: Thu Aug 16 18:22:27 2018 +0200 664 665 Use correct FB handle in radeon_do_pageflip 666 667 We were always using the handle of the client provided FB, which 668 prevented RandR transforms from working, and could result in a black 669 screen. 670 671 Bugzilla: https://bugs.freedesktop.org/107519 672 Fixes: 740f0850f1e4 "Store FB for each CRTC in drmmode_flipdata_rec" 673 (Ported from amdgpu commit f6cd72e64e85896b6d155bee0930e59771dcb701) 674 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 675 676commit 740f0850f1e40403c8dd727e074eae36caeb1f63 677Author: Michel Dänzer <michel.daenzer@amd.com> 678Date: Thu Aug 2 18:49:48 2018 +0200 679 680 Store FB for each CRTC in drmmode_flipdata_rec 681 682 We were only storing the FB provided by the client, but on CRTCs with 683 TearFree enabled, we use a separate FB. This could cause 684 drmmode_flip_handler to fail to clear drmmode_crtc->flip_pending, which 685 could result in a hang when waiting for the pending flip to complete. We 686 were trying to avoid that by always clearing drmmode_crtc->flip_pending 687 when TearFree is enabled, but that wasn't reliable, because 688 drmmode_crtc->tear_free can already be FALSE at this point when 689 disabling TearFree. 690 691 Now that we're keeping track of each CRTC's flip FB separately, 692 drmmode_flip_handler can reliably clear flip_pending, and we no longer 693 need the TearFree hack. 694 695 (Ported from amdgpu commit 9b6782c821e0bdc53336d98f87ddde752faf7902) 696 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 697 698commit ef2a6b818fa47ad571bb0bc105aa8193d51a2190 699Author: Michel Dänzer <michel.daenzer@amd.com> 700Date: Thu Aug 2 18:48:04 2018 +0200 701 702 glamor: Set RADEON_CREATE_PIXMAP_DRI2 for DRI3 pixmaps 703 704 Not doing this resulted in falling back to software for DRI3 client 705 presentation operations with ShadowPrimary. 706 707 (Ported from amdgpu commit 2989d40ef74d9966e8e8df2ef7727b2cc48d4960) 708 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 709 710commit db28d35ce9fd07a2a4703f3df0633d4c8291ff9b 711Author: Michel Dänzer <michel.daenzer@amd.com> 712Date: Fri Jul 27 11:30:57 2018 +0200 713 714 glamor: Use glamor_egl_create_textured_pixmap_from_gbm_bo when possible 715 716 Inspired by the modesetting driver. 717 718 Fixes screen pixmap corruption with Xorg < 1.20, and as a bonus, 719 simplifies the code slightly. 720 721 Bugzilla: https://bugs.freedesktop.org/107385 722 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 723 724commit 7de4ea15880565c6536ed6859fd32ec89ce0b4c0 725Author: Michel Dänzer <michel.daenzer@amd.com> 726Date: Fri Jul 27 09:13:12 2018 +0200 727 728 Bump version to 18.0.99 729 730 Forgot to do this after the 18.0.1 release. 731 732commit fed411290b696ecf9f0e799ec4bad054e4a95017 733Author: Michel Dänzer <michel.daenzer@amd.com> 734Date: Tue Jul 24 17:28:46 2018 +0200 735 736 Use strcpy for RandR output property names 737 738 Instead of strncpy with the string length. Avoids new warnings with GCC 739 8: 740 741 ../../src/drmmode_display.c: In function ‘drmmode_output_create_resources’: 742 ../../src/drmmode_display.c:2240:2: warning: ‘strncpy’ output truncated before terminating nul copying 8 bytes from a string of the same length [-Wstringop-truncation] 743 strncpy(tearfree_prop->name, "TearFree", 8); 744 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 745 ../../src/drmmode_display.c:2244:2: warning: ‘strncpy’ output truncated before terminating nul copying 3 bytes from a string of the same length [-Wstringop-truncation] 746 strncpy(tearfree_prop->enums[0].name, "off", 3); 747 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 748 ../../src/drmmode_display.c:2245:2: warning: ‘strncpy’ output truncated before terminating nul copying 2 bytes from a string of the same length [-Wstringop-truncation] 749 strncpy(tearfree_prop->enums[1].name, "on", 2); 750 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 751 ../../src/drmmode_display.c:2247:2: warning: ‘strncpy’ output truncated before terminating nul copying 4 bytes from a string of the same length [-Wstringop-truncation] 752 strncpy(tearfree_prop->enums[2].name, "auto", 4); 753 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 754 755 (Ported from amdgpu commit f3b2ed37d683f8616a0a31ff63133ddb8fe1a4a3) 756 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 757 758commit d258413c956cd46a237d4b5b94a35817ca7ce8f4 759Author: Michel Dänzer <michel.daenzer@amd.com> 760Date: Mon Jul 23 18:54:40 2018 +0200 761 762 Remove drmmode_terminate_leases 763 764 The RandR screen private is already freed when our CloseScreen runs, so 765 this can't do anything useful. This cleanup has to be done by the X 766 server itself. 767 768 (Ported from amdgpu commit 5f06d6b8ba570b500956ad26fee711d5ac427818) 769 Acked-by: Alex Deucher <alexander.deucher@amd.com> 770 771commit ffdbd46b0abcf8cbe73eb8930ddcd7fd2485d9b5 772Author: Michel Dänzer <michel.daenzer@amd.com> 773Date: Mon Jul 23 11:02:32 2018 +0200 774 775 Hardcode "non-desktop" RandR property name 776 777 It's a bit silly to require current randrproto just for this definition, 778 which can't really change anyway. 779 780 Suggested-by: Qiang Yu <qiang.yu@amd.com> 781 (Ported from amdgpu commit ae8e02c6fc4ef5d5340b8cd4739e66b19b9e3386) 782 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 783 784commit 64bd009d17986c71726724859f100ad40a6e3096 785Author: Michel Dänzer <michel.daenzer@amd.com> 786Date: Thu Jul 19 11:33:42 2018 +0200 787 788 Don't allocate drmmode_output->props twice 789 790 This was accidentally duplicated when porting RandR lease support from 791 amdgpu. 792 793 Trivial. 794 795commit 499d2f9d5d301ef1efd4ffc2952677609ef05122 796Author: Michel Dänzer <michel.daenzer@amd.com> 797Date: Tue Jul 17 11:43:12 2018 +0200 798 799 glamor: Invalidate cached GEM handle in radeon_set_pixmap_bo 800 801 We continued using the stale cached handle, causing issues e.g. when 802 resizing the screen via RandR. 803 804 Reported-by: iive on IRC 805 Acked-by: Alex Deucher <alexander.deucher@amd.com> 806 807commit d9a139bc6d453d33593685de67d1425611f3ab6f 808Author: Michel Dänzer <michel.daenzer@amd.com> 809Date: Mon Jul 16 12:51:26 2018 +0200 810 811 glamor: Fix glamor_block_handler argument in radeon_glamor_finish 812 813 I missed this before because this code isn't compiled with current 814 xserver. 815 816 Trivial. 817 818commit f533b1f654952cee794de49d28d01947a6571daf 819Author: Keith Packard <keithp@keithp.com> 820Date: Wed Jul 11 19:13:28 2018 +0200 821 822 Add RandR leases support 823 824 Signed-off-by: Keith Packard <keithp@keithp.com> 825 (Ported from xserver commit e4e3447603b5fd3a38a92c3f972396d1f81168ad) 826 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 827 (Ported from amdgpu commit 61040bdfa360975614fb47aa7ea1b3a1abac3427) 828 Acked-by: Alex Deucher <alexander.deucher@amd.com> 829 830commit b4f0f44a9f67f2eafd4a2b5ab919e6ea7fa2acf7 831Author: Keith Packard <keithp@keithp.com> 832Date: Wed Jul 11 19:10:20 2018 +0200 833 834 modesetting: Create CONNECTOR_ID properties for outputs [v2] 835 836 This lets a DRM client map between X outputs and kernel connectors. 837 838 v2: 839 Change CONNECTOR_ID to enum -- Adam Jackson <ajax@nwnk.net> 840 841 Signed-off-by: Keith Packard <keithp@keithp.com> 842 (Ported from xserver commit 023d4aba8d45e9e3630b944ecfb650c081799b96) 843 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 844 (Ported from amdgpu commit ab7e39c5a03e24c3ce3ee2f22ada7572bc2d9aa7) 845 Acked-by: Alex Deucher <alexander.deucher@amd.com> 846 847commit 612bda0a5e769f23478b364cf89620222cfac349 848Author: Keith Packard <keithp@keithp.com> 849Date: Wed Jul 11 19:07:59 2018 +0200 850 851 modesetting: Record non-desktop kernel property at PreInit time 852 853 Save any value of the kernel non-desktop property in the xf86Output 854 structure to avoid non-desktop outputs in the default configuration. 855 856 [Also bump randrproto requirement to a version that defines 857 RR_PROPERTY_NON_DESKTOP - ajax] 858 859 Signed-off-by: Keith Packard <keithp@keithp.com> 860 (Ported from xserver commit b91c787c4cd2d20685db69426c539938c556128a) 861 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 862 (Ported from amdgpu commit 14db71a606128c4a207f43298809af279b77e2a8) 863 Acked-by: Alex Deucher <alexander.deucher@amd.com> 864 865commit 4b3e5f81c6032003237bb05c3ab96423c39524fc 866Author: Michel Dänzer <michel.daenzer@amd.com> 867Date: Wed Jul 11 19:04:26 2018 +0200 868 869 Call drmmode_crtc_gamma_do_set from drmmode_setup_colormap 870 871 Instead of from drmmode_set_mode_major. There's no need to re-set the 872 gamma LUT on every modeset, the kernel should preserve it. 873 874 (Ported from amdgpu commit baea4fa492f635cdfe746a84be2e337d9aeae8a9) 875 Acked-by: Alex Deucher <alexander.deucher@amd.com> 876 877commit 16ddc109660dbe813c457765eda66422fb745a3e 878Author: Michel Dänzer <michel.daenzer@amd.com> 879Date: Wed Jul 11 19:03:12 2018 +0200 880 881 Remove #if 0'd code 882 883 This has always been disabled, no need to keep it. 884 885 (Ported from amdgpu commit 19a40758be04e1d451a030f452efb49e8aaad541) 886 Acked-by: Alex Deucher <alexander.deucher@amd.com> 887 888commit 447ef0458fac047919d021f2ba1753a647f5c503 889Author: Michel Dänzer <michel.daenzer@amd.com> 890Date: Wed Jul 11 19:01:31 2018 +0200 891 892 Check dimensions passed to drmmode_xf86crtc_resize 893 894 When enabling a secondary GPU output, Xorg can try resizing the screen 895 beyond the limit advertised by the driver, leading to drmModeAddFB 896 failing and primary GPU outputs turning off. Check for this and bail 897 instead. 898 899 (Ported from amdgpu commit 940c8b39f79789d4d5ddb8ab8d25a8ae05932756) 900 Acked-by: Alex Deucher <alexander.deucher@amd.com> 901 902commit 6a0c01bbd4ed48c696c38952ee33ce21afec9f91 903Author: Michel Dänzer <michel.daenzer@amd.com> 904Date: Wed Jul 11 18:59:58 2018 +0200 905 906 Use drmmode_crtc_dpms in drmmode_set_desired_modes 907 908 Simplifies the latter slightly. 909 910 (Ported from amdgpu commit 74124f2c17dbb4b752707bb7eee398ae099e8a2c) 911 Acked-by: Alex Deucher <alexander.deucher@amd.com> 912 913commit 59441ee3dae15e0e81ed20688b0ba6dba12d7917 914Author: Michel Dänzer <michel.daenzer@amd.com> 915Date: Wed Jul 11 18:58:20 2018 +0200 916 917 Call drmmode_do_crtc_dpms from drmmode_crtc_dpms as well 918 919 Leo pointed out that drmmode_do_crtc_dpms wasn't getting called when 920 turning off an output with 921 922 xrandr --output <output> --off 923 924 This meant that the vblank sequence number and timestamp wouldn't be 925 saved before turning off the CRTC in this case. 926 927 Reported-by: Leo (Sunpeng) Li <sunpeng.li@amd.com> 928 (Ported from amdgpu commit ceeacb455cd058492a493aac954deab8455804b5) 929 Acked-by: Alex Deucher <alexander.deucher@amd.com> 930 931commit 4050b0ad51b1c65945c6474981d1228888738cd4 932Author: Michel Dänzer <michel.daenzer@amd.com> 933Date: Wed Jul 11 18:42:58 2018 +0200 934 935 Replace 'foo == NULL' with '!foo' 936 937 Shorter and sweeter. :) 938 939 (Ported from amdgpu commit e8e688f3852fb06b0c34ed5bce47c9493bcd1613) 940 Acked-by: Alex Deucher <alexander.deucher@amd.com> 941 942commit cf8bc72e3473cef2b511e2c938eb00aca82de909 943Author: Michel Dänzer <michel.daenzer@amd.com> 944Date: Tue Jul 10 18:57:49 2018 +0200 945 946 Wait for pending flips in drmmode_output_set_tear_free 947 948 This prevents a nested call to drmHandleEvent, which would hang. 949 950 Fixes hangs when disabling TearFree on a CRTC while a DRI3 client is 951 page flipping. 952 953 (Ported from amdgpu commit 04947b83cce3a7782e59dece2c7797cc396c1e05) 954 Acked-by: Alex Deucher <alexander.deucher@amd.com> 955 956commit f01d8cf2bd9681b8f5f0e2eddec0a79614389771 957Author: Michel Dänzer <michel.daenzer@amd.com> 958Date: Tue Jul 10 18:53:54 2018 +0200 959 960 Refactor drmmode_output_set_tear_free helper 961 962 Preparation for the following fix, no functional change intended. 963 964 (Ported from amdgpu commit fa30f4601de7a44edfb4a95873bd648946fd4292) 965 Acked-by: Alex Deucher <alexander.deucher@amd.com> 966 967commit ee7e15746148122abf6728fb0f59d6c8ae329e4e 968Author: Michel Dänzer <michel.daenzer@amd.com> 969Date: Tue Jul 10 18:53:06 2018 +0200 970 971 Set drmmode_crtc->scanout_id = 0 when TearFree is disabled 972 973 When disabling TearFree, drmmode_crtc->scanout_id could remain as 1, 974 but drmmode_set_mode_major would destroy drmmode_crtc->scanout[1], so 975 scanout_do_update() would keep bailing, and the scanout buffer would 976 stop being updated. 977 978 Fixes freeze after disabling TearFree on a CRTC with active RandR 979 rotation or other transform. 980 981 (Ported from amdgpu commit 7db0c8e9d7586cff4312d4b93684d35de3e6376f) 982 Acked-by: Alex Deucher <alexander.deucher@amd.com> 983 984commit e9d2d149481e2a9c7cba50d43c6a5146124be3f2 985Author: Michel Dänzer <michel.daenzer@amd.com> 986Date: Tue Jul 10 18:52:08 2018 +0200 987 988 Simplify drmmode_handle_transform 989 990 Set crtc->driverIsPerformingTransform for any case we can handle before 991 calling xf86CrtcRotate. We already clear it afterwards when the latter 992 clears crtc->transform_in_use. 993 994 This should allow our separate scanout buffer mechanism to be used in 995 more cases. 996 997 (Cherry picked from amdgpu commit 8e544b4a0de6717feb4abf00052d57c5b726b5ce) 998 Acked-by: Alex Deucher <alexander.deucher@amd.com> 999 1000commit 89d38e976fd6ba6e026cda7c5d05971f1f177f13 1001Author: Michel Dänzer <michel.daenzer@amd.com> 1002Date: Tue Jul 10 18:51:00 2018 +0200 1003 1004 Don't call scanout_flip/update with a legacy RandR scanout buffer 1005 1006 It means we are not using our own scanout buffers. 1007 1008 Fixes crash when TearFree is supposed to be enabled, but 1009 drmmode_handle_transform doesn't set crtc->driverIsPerformingTransform. 1010 1011 Bugzilla: https://bugs.freedesktop.org/105736 1012 (Ported from amdgpu commit 463477661c88cab3a87746499e5838c5b9f9a13b) 1013 Acked-by: Alex Deucher <alexander.deucher@amd.com> 1014 1015commit 9f2409c4d3a9ecad8e769783e307e57a75b1255e 1016Author: Michel Dänzer <michel.daenzer@amd.com> 1017Date: Tue Jul 10 18:49:05 2018 +0200 1018 1019 Simplify drmmode_crtc_scanout_update 1020 1021 Use our own BoxRec for the extents, and RegionEmpty for clearing the 1022 scanout damage region. 1023 1024 (Ported from amdgpu commit 72c3e9c7308fbcdf85708b72f9be14a5f2f8e7b5) 1025 Acked-by: Alex Deucher <alexander.deucher@amd.com> 1026 1027commit e07c38649280b3f4361005bc4c256f2145a72537 1028Author: Michel Dänzer <michel.daenzer@amd.com> 1029Date: Tue Jul 10 18:48:16 2018 +0200 1030 1031 Update RandR CRTC state if set_mode_major fails in set_desired_modes 1032 1033 Without this, RandR would report the CRTC and its outputs as enabled, 1034 even though they were actually off due to the failure. 1035 1036 (Cherry picked from amdgpu commit 4dcda0b48d62944c841cd9540f4ad4c7ac8dee47) 1037 Acked-by: Alex Deucher <alexander.deucher@amd.com> 1038 1039commit 05390ae36a0abb19f1d533ff95a3fceaafdf79c8 1040Author: Michel Dänzer <michel.daenzer@amd.com> 1041Date: Tue Jul 10 18:45:51 2018 +0200 1042 1043 Abort scanout_update_pending event when possible 1044 1045 We don't need to wait for a non-TearFree scanout update before scanning 1046 out from the screen pixmap or before flipping, as the scanout update 1047 won't be visible anyway. Instead, just abort it. 1048 1049 (Ported from amdgpu commit 36d01989cd842588f12fdae5b2cba5fdcf9c91dd) 1050 Acked-by: Alex Deucher <alexander.deucher@amd.com> 1051 1052commit 290291a11598ba9aa594417998502b0e0ac96970 1053Author: Michel Dänzer <michel.daenzer@amd.com> 1054Date: Tue Jul 10 18:43:28 2018 +0200 1055 1056 Track DRM event queue sequence number in scanout_update_pending 1057 1058 Preparation for next change, no behaviour change intended. 1059 (Ported from amdgpu commit 04a5c5f7cfacad8d9ccffe81e388cc3da2036cb5) 1060 Acked-by: Alex Deucher <alexander.deucher@amd.com> 1061 1062commit d14ff6b77824d0e32728869907e88d6866243101 1063Author: Michel Dänzer <michel.daenzer@amd.com> 1064Date: Tue Jul 10 18:41:43 2018 +0200 1065 1066 Ignore RADEON_DRM_QUEUE_ERROR (0) in radeon_drm_abort_entry 1067 1068 This allows a following change to be slightly simpler. 1069 (Ported from amdgpu commit 8fcc3a9b43d3907052a83a96e5a2423afab5ad3f) 1070 Acked-by: Alex Deucher <alexander.deucher@amd.com> 1071 1072commit 731d4b386a55cdb468e37b69d41284150952cf8c 1073Author: Emil Velikov <emil.velikov@collabora.com> 1074Date: Tue Jul 10 18:39:31 2018 +0200 1075 1076 Do not export the DriverRec RADEON 1077 1078 Unused externally and should not be exported. 1079 1080 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 1081 (Ported from amdgpu commit 7fb8b49895e225b3908c8bd186539de23afe91d1) 1082 Acked-by: Alex Deucher <alexander.deucher@amd.com> 1083 1084commit cc1d0824a8a7ef29f8911b95695f7cb1b4abe9b8 1085Author: Jim Qu <Jim.Qu@amd.com> 1086Date: Tue Jul 10 18:36:42 2018 +0200 1087 1088 Wait for pending scanout update before calling drmmode_crtc_scanout_free 1089 1090 There is a case that when set screen from reverse to normal, the old 1091 scanout damage is freed in modesetting before scanout update handler, 1092 so it causes segment fault issue. 1093 1094 Signed-off-by: Jim Qu <Jim.Qu@amd.com> 1095 1096 [ Michel Dänzer: Only call drmmode_crtc_wait_pending_event before 1097 drmmode_crtc_scanout_free is actually called, slightly tweak commit 1098 message ] 1099 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 1100 (Ported from amdgpu commit 9f6a8905611b5b1d8fcd31bebbc9af7ca1355cc3) 1101 Acked-by: Alex Deucher <alexander.deucher@amd.com> 1102 1103commit 413622bcccc9d36a9c87679f379dc84b21a3c655 1104Author: Michel Dänzer <michel.daenzer@amd.com> 1105Date: Tue Jul 10 17:23:24 2018 +0200 1106 1107 Swap pixmap privates in radeon_dri2_exchange_buffers 1108 1109 Instead of only the BOs. 1110 1111 This matches what amdgpu does, and fixes issues with DRI2 page flipping. 1112 1113 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1114 1115commit 3c4c0213c11d623cba7adbc28dde652694f2f758 1116Author: Michel Dänzer <michel.daenzer@amd.com> 1117Date: Fri Jun 29 17:57:03 2018 +0200 1118 1119 glamor: Use GBM for BO allocation when possible 1120 1121 Inspired by amdgpu. This avoids various issues due to a GEM handle 1122 lifetime conflict between us and Mesa with current glamor. 1123 1124 Bugzilla: https://bugs.freedesktop.org/105381 1125 Tested-by: Konstantin Kharlamov <hi-angel@yandex.ru> 1126 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1127 1128commit b85b7b11f5b5e792f21951b881bd8433d3d70858 1129Author: Michel Dänzer <michel.daenzer@amd.com> 1130Date: Fri May 18 18:49:35 2018 +0200 1131 1132 Add struct radeon_buffer 1133 1134 Inspired by amdgpu, preparation for the following change. For now, this 1135 is mostly a wrapper around struct radeon_bo, no functional change 1136 intended. 1137 1138 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1139 1140commit 37ba075b34130c41bb7a2261bd666af5b29ffaf3 1141Author: Michel Dänzer <michel.daenzer@amd.com> 1142Date: Fri Jun 15 18:23:53 2018 +0200 1143 1144 Refactor radeon_finish helper 1145 1146 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1147 1148commit 6c986e997159ad0086f940294b244fc4c30b61fc 1149Author: Jammy Zhou <Jammy.Zhou@amd.com> 1150Date: Fri Jun 15 18:41:34 2018 +0200 1151 1152 Remove throttling from radeon_dri2_copy_region2 1153 1154 Throttling should be handled by the client-side drivers. 1155 1156 Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> 1157 (Ported from amdgpu commit 8a34a8149860ac15e83ccdbd8d9a527d8d3e5997) 1158 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1159 1160commit 90b94d40449f665f2d12874598062a5e5e5b64cd 1161Author: Michel Dänzer <michel.daenzer@amd.com> 1162Date: Fri Jun 15 18:27:46 2018 +0200 1163 1164 Move flush from radeon_scanout_do_update to its callers 1165 1166 No functional change intended. 1167 1168 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1169 1170commit 5da2bf43e64f2d702fe05016263f617c10f8f134 1171Author: Michel Dänzer <michel.daenzer@amd.com> 1172Date: Mon May 28 18:35:32 2018 +0200 1173 1174 Factor out radeon_surface_initialize helper 1175 1176 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1177 1178commit ba5d5402b3e2e807d7e47205ac83f930b6c8caf5 1179Author: Michel Dänzer <michel.daenzer@amd.com> 1180Date: Thu May 17 12:36:29 2018 +0200 1181 1182 glamor: Don't store radeon_surfaces in pixmaps 1183 1184 Only EXA needs them. 1185 1186 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1187 1188commit eec4a41925127ae490f0a5156a881a08d521e28e 1189Author: Michel Dänzer <michel.daenzer@amd.com> 1190Date: Thu May 17 18:25:37 2018 +0200 1191 1192 Only initialize libdrm_radeon surface manager for >= R600 1193 1194 Not used with older GPUs. 1195 1196 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1197 1198commit ac18a993a3a68629805b8f272a339e25444c0897 1199Author: Michel Dänzer <michel.daenzer@amd.com> 1200Date: Wed May 23 12:18:36 2018 +0200 1201 1202 EXA: Remove old RADEONEXACreatePixmap hook 1203 1204 Not used by any supported version of xserver. 1205 1206 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1207 1208commit 25836375e0aab17a2e6bd91acb1c19b837ae2109 1209Author: Michel Dänzer <michel.daenzer@amd.com> 1210Date: Thu May 17 18:18:42 2018 +0200 1211 1212 Drop unused drmmode_create_bo_pixmap surface parameter 1213 1214 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1215 1216commit 65c9dfea4e841b7d6f795c7489fede58c5e9631f 1217Author: Michel Dänzer <michel.daenzer@amd.com> 1218Date: Tue May 15 18:59:30 2018 +0200 1219 1220 glamor: Bail CreatePixmap on unsupported pixmap depth 1221 1222 Fixes crash in that case. 1223 1224 Bugzilla: https://bugs.freedesktop.org/106293 1225 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1226 1227commit 3dcfce8d0f495d09d7836caf98ef30d625b78a13 1228Author: Michel Dänzer <michel.daenzer@amd.com> 1229Date: Thu May 17 09:50:50 2018 +0200 1230 1231 Bail from dri2_create_buffer2 if we can't get a pixmap 1232 1233 We would store the NULL pointer and continue, which would lead to a 1234 crash down the road. 1235 1236 Bugzilla: https://bugs.freedesktop.org/106293 1237 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1238 1239commit fdba5373522c008f43e9ca5a9e76e35c6e7ad73d 1240Author: Michel Dänzer <michel.daenzer@amd.com> 1241Date: Thu Mar 15 17:19:59 2018 +0100 1242 1243 Bump version for 18.0.1 release 1244 1245commit e0f5020117d0a6340f4208ec52bd3e3b3aa99a1b 1246Author: Michel Dänzer <michel.daenzer@amd.com> 1247Date: Wed Mar 14 16:54:52 2018 +0100 1248 1249 Pass extents to radeon_scanout_do_update by value 1250 1251 radeon_scanout_extents_intersect could leave the scanout damage region 1252 in an invalid state, triggering debugging checks in pixman: 1253 1254 *** BUG *** 1255 In pixman_region_append_non_o: The expression r->x1 < r->x2 was false 1256 Set a breakpoint on '_pixman_log_error' to debug 1257 1258 (Ported from amdgpu commit 8af989546907ad9fb491d940e1936d3bfc89276b) 1259 1260commit b915e8e6fb956e983c2ce2a5565c20838c85707c 1261Author: Michel Dänzer <michel.daenzer@amd.com> 1262Date: Wed Mar 7 17:28:55 2018 +0100 1263 1264 Wrap the whole miPointerScreenFuncRec, instead of only Set/MoveCursor 1265 1266 We were clobbering entries in mi's global miSpritePointerFuncs struct, 1267 which cannot work correctly with multiple primary screens. Instead, 1268 assign a pointer to our own wrapper struct to PointPriv->spriteFuncs. 1269 1270 Fixes crashes with multiple primary screens. 1271 1272 Fixes: 1fe8ca75974c ("Keep track of how many SW cursors are visible on 1273 each screen") 1274 Reported-by: Mario Kleiner <mario.kleiner.de@gmail.com> 1275 1276commit 7e18aea984e186dd2ab1144470b0c0e832562ad1 1277Author: Michel Dänzer <michel.daenzer@amd.com> 1278Date: Tue Mar 6 18:15:45 2018 +0100 1279 1280 Bump version for 18.0.0 release 1281 1282commit 504b8721b17a672caf1ed3eab087027c02458cab 1283Author: Michel Dänzer <michel.daenzer@amd.com> 1284Date: Mon Mar 5 18:41:44 2018 +0100 1285 1286 Only change Set/MoveCursor hooks from what we expect 1287 1288 Since xf86CursorCloseScreen runs after RADEONCloseScreen_KMS, 1289 PointPriv->spriteFuncs doesn't point to the same struct in the latter as 1290 in RADEONCursorInit_KMS. So we were restoring info->Set/MoveCursor to 1291 the wrong struct. Then in the next server generation, 1292 info->Set/MoveCursor would end up pointing to 1293 drmmode_sprite_set/move_cursor, resulting in an infinite loop if one of 1294 them was called. 1295 1296 To avoid this, only change the Set/MoveCursor hooks if their values 1297 match our expectations, otherwise leave them as is. This is kind of a 1298 hack, but the alternative would be invasive and thus risky changes to 1299 the way we're wrapping CloseScreen, and it's not even clear that can 1300 work without changing xserver code. 1301 1302 Fixes: 1fe8ca75974c ("Keep track of how many SW cursors are visible on 1303 each screen") 1304 Acked-by: Alex Deucher <alexander.deucher@amd.com> 1305 1306commit e0d23092337efe4c1e406c9124ca8d8856863e65 1307Author: Keith Packard <keithp@keithp.com> 1308Date: Fri Mar 2 12:13:43 2018 +0100 1309 1310 modesetting: Update property values at detect and uevent time 1311 1312 We were updating the link-status property when a uevent came in, but 1313 we also want to update the non-desktop property, and potentially 1314 others as well. We also want to check at detect time in case we don't 1315 get a hotplug event. 1316 1317 This patch updates every property provided by the kernel, sending 1318 changes to DIX so it can track things as well. 1319 1320 Signed-off-by: Keith Packard <keithp@keithp.com> 1321 1322 (Ported from amdgpu commit 374cb8fef4fdbb648af089ee80803ec78321f1b2) 1323 1324commit aef8fe6e9e94ebecbbb6adf940160d3a7ab4d5ac 1325Author: Daniel Martin <consume.noise@gmail.com> 1326Date: Fri Mar 2 12:12:09 2018 +0100 1327 1328 modesetting: Reset output_id if drmModeGetConnector failed 1329 1330 If drmModeGetConnector() fails in drmmode_output_detect(), we have to 1331 reset the output_id to -1 too. 1332 1333 Yet another spot leading to a potential NULL dereference when handling 1334 the mode_output member as output_id was != -1. Though, this case should 1335 be very hard to hit. 1336 1337 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 1338 1339 (Ported from amdgpu commit 10054b6c3d9a755b30abb43020121b9631fa296d) 1340 1341commit 1ef12a92ca53cd6fbfed3ef8628616445b007c9a 1342Author: Daniel Martin <consume.noise@gmail.com> 1343Date: Fri Mar 2 12:08:09 2018 +0100 1344 1345 modesetting: Use helper to fetch drmModeProperty(Blob)s 1346 1347 Replace the various loops to lookup drmModeProperty(Blob)s by 1348 introducing helper functions. 1349 1350 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 1351 1352 (Ported from amdgpu commit fb58e06acd6c6bd59de2dbdadbca27eb1dd0025b) 1353 1354commit ace1e3871c39a86becdf473a47fdc096e8ef39f5 1355Author: Michel Dänzer <michel.daenzer@amd.com> 1356Date: Wed Feb 21 17:53:59 2018 +0100 1357 1358 Disable all unused CRTCs before setting desired modes 1359 1360 This might avoid modeset failures in some cases where a CRTC which isn't 1361 used by Xorg was enabled before. 1362 1363 (Ported from amdgpu commit e3aae7a24296f640c0153d1459f3e0820485468a) 1364 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1365 1366commit f2915eb7a4beb6140ebec183aeac02111fc6ba77 1367Author: Christoph Haag <haagch@frickel.club> 1368Date: Thu Mar 1 15:07:00 2018 +0100 1369 1370 fix include order for present.h configure test 1371 1372 xorg-server.h defines _XSERVER64 which is used in X.h to choose the 1373 correct definition of XID 1374 1375 this prevents a failure in the present.h configure test that disables 1376 DRI3 on X.Org 1.20 1377 1378 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 1379 1380commit f777e909fd23e063729ecb038997db445f82ef5e 1381Author: Michel Dänzer <michel.daenzer@amd.com> 1382Date: Fri Feb 16 16:55:03 2018 +0100 1383 1384 Don't bail from drmmode_set_desired_modes immediately 1385 1386 If we fail to find or set the mode for a CRTC, keep trying for the 1387 remaining CRTCs, and only return FALSE if we failed for all CRTCs that 1388 should be on. 1389 1390 (Ported from amdgpu commit f5ac5f385f41d1547cfd7ccc8bb35a537a8fffeb) 1391 1392commit 2d63748667d0bcfcd9999e4e41f775eba87897a0 1393Author: Michel Dänzer <michel.daenzer@amd.com> 1394Date: Thu Feb 15 17:47:14 2018 +0100 1395 1396 Simplify depth 30 "kernel too old" error message 1397 1398 We know it's depth 30, no need to have xf86DrvMsg fill it in. 1399 1400 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1401 1402commit 680b4ab43f00b1f635f9359281e13a031d09ad24 1403Author: Michel Dänzer <michel.daenzer@amd.com> 1404Date: Wed Jan 24 12:29:17 2018 +0100 1405 1406 If glamor is too old for depth 30, fall back to EXA or ShadowFB 1407 1408 Instead of not starting up at all. 1409 1410 Corresponding to amdgpu commit 37c7260bdef3a53b0f0295a531f33938e9aad8cf. 1411 1412 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1413 1414commit 16954fab59f95bdca4540c4496b2ec0759c447e9 1415Author: Michel Dänzer <michel.daenzer@amd.com> 1416Date: Wed Feb 14 18:20:04 2018 +0100 1417 1418 Call RADEONFreeRec from RADEONFreeScreen_KMS even if info == NULL 1419 1420 It's safe now. 1421 1422 (Ported from amdgpu commit c9bd1399a13cea2e1331af2c826ca054b88db071) 1423 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1424 1425commit 6b44b8c584f3b63712858be11277842c97d03426 1426Author: Michel Dänzer <michel.daenzer@amd.com> 1427Date: Wed Feb 14 18:06:18 2018 +0100 1428 1429 Don't call RADEONFreeRec from RADEONPreInit_KMS 1430 1431 If the latter fails, Xorg will call RADEONFreeScreen_KMS, which calls 1432 the former. 1433 1434 (Ported from amdgpu commit 103b7285845b786929fb509083c57e074c48f9be) 1435 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1436 1437commit 8a55e66bfd77dcb2553a43e86215b9d3c87b03f0 1438Author: Michel Dänzer <michel.daenzer@amd.com> 1439Date: Wed Feb 14 18:13:16 2018 +0100 1440 1441 Don't use RADEONEntPriv in RADEONFreeRec 1442 1443 It crashes if info == NULL. 1444 1445 (Ported from amdgpu commits fb8444e731765588c0ff1e9053c1c7b73f5f0907 & 1446 cfccf4c4e7e5c73fe4040fabeb1b43283cf29b33) 1447 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1448 1449commit 703cbb7177c6f69264836e975c1921e552fd60a4 1450Author: Michel Dänzer <michel.daenzer@amd.com> 1451Date: Wed Feb 14 17:28:10 2018 +0100 1452 1453 glamor: Allow depth 30 with Xorg >= 1.19.99.1 1454 1455 Corresponding to amdgpu commit 6aee5770fb913713bb1b9a1af8f0d0892a66f21a. 1456 1457 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1458 1459commit 733f606dd6ca8350e6e7f0858bfff5454ddc98ed 1460Author: Michel Dänzer <michel.daenzer@amd.com> 1461Date: Thu Jan 25 16:59:30 2018 +0100 1462 1463 Always use screen depth/bpp for KMS framebuffers 1464 1465 DRI clients can use depth 32 pixmaps while the screen is depth 24, in 1466 which case page flipping would fail. 1467 1468 Reported-by: Mario Kleiner <mario.kleiner.de@gmail.com> 1469 Reviewed-and-Tested-by: Mario Kleiner <mario.kleiner.de@gmail.com> 1470 1471commit 83124426040bc1f2dba74c8eeb5a5a9c7c64e664 1472Author: Mario Kleiner <mario.kleiner.de@gmail.com> 1473Date: Wed Jan 24 18:19:38 2018 +0100 1474 1475 Make XvMC extension initialize at depth 30. 1476 1477 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 1478 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 1479 1480commit 2d171c892d373b45ef9f1ee068648437e2529a78 1481Author: Mario Kleiner <mario.kleiner.de@gmail.com> 1482Date: Wed Jan 24 18:19:37 2018 +0100 1483 1484 Add missing depth 30 -> cpp=4 handling for DRI2. 1485 1486 Missing this apparently didn't cause any ill side effects, only a bit of 1487 confusion when looking at xtrace's of DRI2 clients. 1488 1489 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 1490 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 1491 1492commit 574bfab4bf1fcd95163a8f33cea2889189429d30 1493Author: Mario Kleiner <mario.kleiner.de@gmail.com> 1494Date: Mon Jan 22 03:14:37 2018 +0100 1495 1496 Support exa screen color depth 30 on Linux 3.16 and later. (v2) 1497 1498 Linux 3.16's radeon-kms introduced 10 bpc fb support. 1499 Currently, as of X-Server 1.19, this works with exa accel, 1500 but not yet when using glamor acceleration. 1501 1502 v2: Style fixes, check for glamor first, as suggested 1503 by Michel. 1504 1505 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 1506 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 1507 1508commit ee7e6c25989e4faf1adfa4754b7edd5a9a4722ce 1509Author: Fredrik Höglund <fredrik@kde.org> 1510Date: Mon Jan 22 03:14:36 2018 +0100 1511 1512 exa: Accelerate ARGB2101010 pictures 1513 1514 Tested by Mario on a Radeon HD 4000 series for the r600 1515 exa path, and on Radeon HD 5770 for the evergreen exa 1516 path. 1517 1518 Signed-off-by: Fredrik Höglund <fredrik@kde.org> 1519 Reviewed-and-Tested-by: Mario Kleiner <mario.kleiner.de@gmail.com> 1520 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 1521 1522commit 1f1d4b1fa7d4b22dd8553f7e71251bf17ca7a7b1 1523Author: Mario Kleiner <mario.kleiner.de@gmail.com> 1524Date: Mon Jan 22 03:14:35 2018 +0100 1525 1526 Skip xf86HandleColormaps() at color depth 30. (v2) 1527 1528 The hardware gamma luts get bypassed at color 1529 depth 30 anyway, so skip their setup. 1530 1531 v2: Also don't hook up drmmode_crtc_gamma_set() 1532 to .gamma_set for a screen with depth 30, 1533 so the server can spare itself the effort of 1534 updating and setting unused cluts. 1535 Suggested by Michel. 1536 1537 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 1538 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 1539 1540commit 21f6753462464acfd3c452393328c977a375ce26 1541Author: Mario Kleiner <mario.kleiner.de@gmail.com> 1542Date: Mon Jan 22 03:14:34 2018 +0100 1543 1544 Define per x-screen individual drmmode_crtc_funcs 1545 1546 This allows to en-/disable some functions depending on individual screen 1547 settings. 1548 1549 Prep work for more efficient depth 30 support. 1550 1551 Suggested-by: Michel Dänzer <michel.daenzer@amd.com> 1552 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 1553 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 1554 1555commit 1fe8ca75974c5241c3b556b757f9c966c9ce5002 1556Author: Michel Dänzer <michel.daenzer@amd.com> 1557Date: Wed Dec 27 17:13:30 2017 +0100 1558 1559 Keep track of how many SW cursors are visible on each screen 1560 1561 And use this to determine when we cannot use page flipping for DRI 1562 clients. We previously did this based on whether the HW cursor cannot 1563 be used on at least one CRTC, which had at least two issues: 1564 1565 * Even while the HW cursor cannot be used, no SW cursor may actually be 1566 visible (e.g. because all cursors are disabled), in which case we can 1567 use page flipping for DRI clients anyway 1568 * Even while the HW cursor can be used, there may be SW cursors visible 1569 from non-core pointer devices, in which case we cannot use page 1570 flipping for DRI clients anyway 1571 1572 (Ported from amdgpu commit 69e20839bfeb3ee0b0a732d72de0a32d6c5435fc) 1573 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1574 1575commit b4ce8913ac392df339081f6f3e28bf79a239ee75 1576Author: Michel Dänzer <michel.daenzer@amd.com> 1577Date: Wed Dec 27 17:05:26 2017 +0100 1578 1579 Move cursor related ScreenInit calls into RADEONCursorInit_KMS 1580 1581 And bail if xf86_cursors_init fails. 1582 1583 (Ported from amdgpu commit dfccaa7043ccb157a1f8be7313123792bb7e7001) 1584 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1585 1586commit ee907e9b9563609cb4a4817e17215b50bb1c7177 1587Author: Michel Dänzer <michel.daenzer@amd.com> 1588Date: Fri Dec 1 15:49:11 2017 +0100 1589 1590 Add radeon_dirty_src_drawable helper 1591 1592 Allows tidying up radeon_dirty_src_equals and redisplay_dirty slightly. 1593 1594 (Cherry picked from amdgpu commit 1d65ac395971571094df21ca0408d5972c6b56ec) 1595 Acked-by: Alex Deucher <alexander.deucher@amd.com> 1596 1597commit fccfd75ecf9ccfa628a8ff9c1fbb7f60c0f10ea5 1598Author: Joakim Tjernlund <joakim.tjernlund@infinera.com> 1599Date: Fri Nov 24 17:27:09 2017 +0100 1600 1601 Fix non GLAMOR build 1602 1603 Non GLAMOR builds needs pRADEONEnt so move the USE_GLAMOR define 1604 down to make pRADEONEnt defined in this case. 1605 1606 Signed-off-by: Joakim Tjernlund <joakim.tjernlund@infinera.com> 1607 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 1608 1609commit 15350bd5538c21cd4c6b406d565ca765b7e0218a 1610Author: Michel Dänzer <michel.daenzer@amd.com> 1611Date: Wed Nov 15 18:01:29 2017 +0100 1612 1613 Use correct ScrnInfoPtr in redisplay_dirty 1614 1615 We used the destination pixmap's screen for flushing drawing commands. 1616 But when we are the master screen, the destination pixmap is from the 1617 slave screen. 1618 1619 Fixes crash when the slave screen isn't using the same acceleration 1620 architecture as us. 1621 1622 Bugzilla: https://bugs.freedesktop.org/103613 1623 Fixes: 01b040b4a807 ("Adapt to PixmapDirtyUpdateRec::src being a 1624 DrawablePtr") 1625 (Ported from amdgpu commit 3a4f7422913093ed9e26b73ecd7f9e773478cb1e) 1626 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1627 1628commit 6441210bc53978a2c46c572e9174215b14f9e813 1629Author: Daniel Martin <consume.noise@gmail.com> 1630Date: Fri Oct 20 10:05:35 2017 +0200 1631 1632 modesetting: Check crtc before searching link-status property 1633 1634 No need to lookup the link-status property if we don't have a crtc. 1635 1636 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 1637 (Ported from xserver commit 8d7f7e24261e68459e6f0a865e243473f65fe7ad) 1638 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1639 1640commit cafcb55768b58dc63a8c5682000eba21ad904689 1641Author: Keith Packard <keithp@keithp.com> 1642Date: Mon Sep 25 16:18:22 2017 -0700 1643 1644 modesetting: Skip no-longer-present connectors when resetting BAD links 1645 1646 Outputs may have NULL mode_output (connector) pointers if the 1647 connector disappears while the server is running. Skip these when 1648 resetting outputs with BAD link status. 1649 1650 (Ported from xserver commit 37f4e7651a2fd51efa613a08a1e705553be33e76) 1651 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1652 1653commit 0cd14777b7e39619f231b841a2d73f3122458030 1654Author: Michel Dänzer <michel.daenzer@amd.com> 1655Date: Fri Oct 20 17:20:30 2017 +0200 1656 1657 Always call drmModeFreeProperty after drmModeGetProperty 1658 1659 We were not doing so in all cases, leaking memory allocated by the 1660 latter. 1661 1662 (Cherry picked from amdgpu commit f6b39bcd45cb06976ba8a3600df77fc471c63995) 1663 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1664 1665commit 84c7d9c1c71d9cee9f50f96671032529b606cf04 1666Author: Michel Dänzer <michel.daenzer@amd.com> 1667Date: Fri Oct 20 17:19:09 2017 +0200 1668 1669 Call TimerFree for timer created in LeaveVT 1670 1671 We were leaking the memory allocated by TimerSet. 1672 1673 (Ported from amdgpu commit 84aad09f18fed6b52b0c073f0bbd675a6de07807) 1674 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1675 1676commit d588015f4f60ce270bcfd8433215fc981a84f143 1677Author: Michel Dänzer <michel.daenzer@amd.com> 1678Date: Fri Oct 20 17:13:36 2017 +0200 1679 1680 Free memory returned by xf86GetEntityInfo 1681 1682 We were leaking it. 1683 1684 (Ported from amdgpu commit cfccf4c4e7e5c73fe4040fabeb1b43283cf29b33) 1685 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1686 1687commit ec90c94881239791d0a66a02fc6d003c73a6c1f3 1688Author: Michel Dänzer <michel.daenzer@amd.com> 1689Date: Fri Oct 20 16:51:16 2017 +0200 1690 1691 Free pRADEONEnt memory in RADEONFreeRec 1692 1693 We were leaking it. 1694 1695 (Inspired by amdgpu commit 9d84934309e4ccd9a43c73d958b8ff10ef2fc990) 1696 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1697 1698commit b93ae9b159790e8a979e9d5d567a0e6fccd7d08a 1699Author: Michel Dänzer <michel.daenzer@amd.com> 1700Date: Thu Oct 19 18:57:45 2017 +0200 1701 1702 Unreference pixmap's FB with EXA as well in radeon_set_pixmap_bo 1703 1704 Not doing so resulted in DRI2 page flips not actually changing the FB 1705 being scanned out, showing intermittent flicker of the "back" buffer 1706 rendering. 1707 1708 Bugzilla: https://bugs.freedesktop.org/102643 1709 Fixes: 55e513b978b2 "Use reference counting for tracking KMS framebuffer 1710 lifetimes" 1711 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1712 1713commit 61113768bd79ec5c2164835e2e048e8f2a42ce9a 1714Author: Darren Salt <devspam@moreofthesa.me.uk> 1715Date: Thu Oct 19 18:17:07 2017 +0200 1716 1717 Clarify when TearFree is automatically enabled. 1718 1719 (Ported from amdgpu commit 2f72be038d22c54620e436af30121dd89f79a003) 1720 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1721 1722commit 7d435354099119234d443b07e2df1c7b9f97cf3c 1723Author: Michel Dänzer <michel.daenzer@amd.com> 1724Date: Wed Oct 4 11:11:33 2017 +0200 1725 1726 Bail if there's a problem with ShadowFB 1727 1728 If we hit a problem while setting up ShadowFB, just carrying on trying 1729 to set up HW acceleration instead is unlikely to work. 1730 1731 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1732 1733commit b71836c9962f8f2eadf0c2188e03faed737c18da 1734Author: Michel Dänzer <michel.daenzer@amd.com> 1735Date: Wed Oct 4 11:05:49 2017 +0200 1736 1737 Post-release version bump 1738 1739commit 34da04daec82077571558ac3fe1ec0c1203a01ad 1740Author: Michel Dänzer <michel.daenzer@amd.com> 1741Date: Tue Oct 3 12:47:14 2017 +0200 1742 1743 Fix VT switching with ShadowFB 1744 1745 We were trying to call acceleration specific functions from LeaveVT. 1746 Instead, memset the scanout buffer to all 0 in LeaveVT and allocate a 1747 new one in EnterVT. 1748 1749 Bugzilla: https://bugs.freedesktop.org/102948 1750 Fixes: 06a465484101 ("Make all active CRTCs scan out an all-black 1751 framebuffer in LeaveVT") 1752 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 1753 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1754 1755commit f9aca8c1a7a423b9e738c88400ae821bdfbcdb1c 1756Author: Michel Dänzer <michel.daenzer@amd.com> 1757Date: Fri Sep 8 17:02:59 2017 +0900 1758 1759 Bump version for 7.10.0 release 1760 1761commit 5cdd334b3402c2431deb3a87a8d04ef590da53ee 1762Author: Michel Dänzer <michel.daenzer@amd.com> 1763Date: Wed Aug 30 18:48:19 2017 +0900 1764 1765 Require xserver >= 1.13 1766 1767 xserver 1.13.0 was released on September 6th, 2012, almost 5 years ago. 1768 1769 This allows cleaning up a bunch of backwards compatibility code. 1770 1771 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1772 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 1773 1774commit 9d9c565c84601f4c6c73ad769f86491088683f7a 1775Author: Michel Dänzer <michel.daenzer@amd.com> 1776Date: Tue Aug 29 18:07:51 2017 +0900 1777 1778 Use a timer for unreferencing the all-black FB 1779 1780 The timer fires 1 second after LeaveVT. This gives the next DRM master 1781 enough time to set up scanout of its own buffers. 1782 1783 Fixes prolonged intermittent black screen when switching from Xorg to 1784 e.g. the GDM Wayland mode login VT. 1785 1786 Fixes: 06a465484101 ("Make all active CRTCs scan out an all-black 1787 framebuffer in LeaveVT") 1788 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1789 1790commit e4a3df19d588a4310fcb889ef34e205d0e92e4d7 1791Author: Michel Dänzer <michel.daenzer@amd.com> 1792Date: Mon Aug 28 17:54:23 2017 +0900 1793 1794 Remove drmmode_scanout_free 1795 1796 Not used anymore. 1797 1798 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1799 1800commit 06a465484101f21e99d3a0a62fb03440bcaff93e 1801Author: Michel Dänzer <michel.daenzer@amd.com> 1802Date: Wed Aug 23 18:00:15 2017 +0900 1803 1804 Make all active CRTCs scan out an all-black framebuffer in LeaveVT 1805 1806 And destroy all other FBs. This is so that other DRM masters can only 1807 get access to this all-black FB, not to any other FB we created, while 1808 we're switched away and not DRM master. 1809 1810 Fixes: 55e513b978b2 ("Use reference counting for tracking KMS 1811 framebuffer lifetimes") 1812 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1813 1814commit 7f0cd68d1b0c132e32ae736371bce3e12ed33c7a 1815Author: Michel Dänzer <michel.daenzer@amd.com> 1816Date: Mon Aug 28 17:53:19 2017 +0900 1817 1818 Create radeon_master_screen helper 1819 1820 Preparatory, no functional change intended yet. 1821 1822 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1823 1824commit 20f6b56fdb74d88086e8e094013fedbb14e50a24 1825Author: Michel Dänzer <michel.daenzer@amd.com> 1826Date: Thu Aug 24 17:10:29 2017 +0900 1827 1828 Create radeon_pixmap_get_fb_ptr helper 1829 1830 Preparatory, no functional change intended yet. 1831 1832 Also inline radeon_pixmap_create_fb into radeon_pixmap_get_fb, since 1833 there's only one call-site anymore. 1834 1835 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1836 1837commit 4bc992c31059eb50e22df4ebf5b92d08411f41ef 1838Author: Michel Dänzer <michel.daenzer@amd.com> 1839Date: Wed Aug 23 17:24:53 2017 +0900 1840 1841 Create drmmode_set_mode helper 1842 1843 Preparatory, no functional change intended yet. 1844 1845 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1846 1847commit 3f6210ca2c8ef60d59efc8139151d3b9838bb875 1848Author: Michel Dänzer <michel.daenzer@amd.com> 1849Date: Thu Jun 22 18:37:26 2017 +0900 1850 1851 Create radeon_pixmap_clear helper 1852 1853 Preparatory, no functional change intended yet. 1854 1855 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1856 1857commit 0472a605e0ec8fec1892bbc3a84698b7ef9c5296 1858Author: Martin Peres <martin.peres@linux.intel.com> 1859Date: Mon Apr 10 16:48:21 2017 +0300 1860 1861 modesetting: re-set the crtc's mode when link-status goes BAD 1862 1863 Despite all the careful planning of the kernel, a link may become 1864 insufficient to handle the currently-set mode. At this point, the 1865 kernel should mark this particular configuration as being broken 1866 and potentially prune the mode before setting the offending connector's 1867 link-status to BAD and send the userspace a hotplug event. This may 1868 happen right after a modeset or later on. 1869 1870 Upon receiving a hot-plug event, we iterate through the connectors to 1871 re-apply the currently-set mode on all the connectors that have a 1872 link-status property set to BAD. The kernel may be able to get the 1873 link to work by dropping to using a lower link bpp (with the same 1874 display bpp). However, the modeset may fail if the kernel has pruned 1875 the mode, so to make users aware of this problem a warning is outputed 1876 in the logs to warn about having a potentially-black display. 1877 1878 This patch does not modify the current behaviour of always propagating 1879 the events to the randr clients. This allows desktop environments to 1880 re-probe the connectors and select a new resolution based on the new 1881 (currated) mode list if a mode disapeared. This behaviour is expected in 1882 order to pass the Display Port compliance tests. 1883 1884 (Ported from xserver commit bcee1b76aa0db8525b491485e90b8740763d7de6) 1885 1886 [ Michel: Bump libdrm dependency to >= 2.4.78 for 1887 DRM_MODE_LINK_STATUS_BAD ] 1888 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1889 1890commit d822a0f47070374ad0c1a97b559bae27724dc52a 1891Author: Michel Dänzer <michel.daenzer@amd.com> 1892Date: Thu Aug 17 16:20:52 2017 +0900 1893 1894 Make radeon_scanout_do_update take a PixmapPtr instead of a DrawablePtr 1895 1896 All callers were already passing in a pixmap. 1897 1898 This allows simplifying the rotated scanout case slightly. 1899 1900 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1901 1902commit 3e24770b1b472fc15df56d06f5f04778c9db63dd 1903Author: Michel Dänzer <michel.daenzer@amd.com> 1904Date: Wed Aug 9 13:02:34 2017 +0900 1905 1906 Use xorg_list_append for the DRM event list 1907 1908 We were adding entries at the start of the list, i.e. the list was 1909 ordered from most recently added to least recently added. However, the 1910 corresponding DRM events are generally expected to arrive in the same 1911 order as they are queued, which means that radeon_drm_queue_alloc would 1912 generally have to traverse the whole list to find the entry 1913 corresponding to an arrived event. Fix this by adding entries at the end 1914 of the list. 1915 1916 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1917 1918commit 36ce7920136c0d723c9397a84e7dd5926a9c7943 1919Author: Michel Dänzer <michel.daenzer@amd.com> 1920Date: Wed Aug 9 19:01:15 2017 +0900 1921 1922 Consolidate radeon_scanout_flip_abort/handler helpers 1923 1924 While at it, make them use crtc->driver_private. 1925 1926 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1927 1928commit d314cbfb228bb4b8762714f98d0c114a8ee3f061 1929Author: Michel Dänzer <michel.daenzer@amd.com> 1930Date: Mon Jul 31 17:12:59 2017 +0900 1931 1932 Always allow DRI2 page flipping with TearFree 1933 1934 Even if TearFree is enabled for the CRTC we're synchronizing to. 1935 1936 v2: 1937 * Move manpage hunk here from previous change. 1938 1939 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> # v1 1940 1941commit 4445765af5b97d0cfd10889fe6d6f58f2ce85659 1942Author: Michel Dänzer <michel.daenzer@amd.com> 1943Date: Thu Feb 9 13:01:03 2017 +0900 1944 1945 Always allow Present page flipping with TearFree 1946 1947 Even if TearFree is active for the the CRTC we're synchronizing to. In 1948 that case, for Present flips synchronized to vertical blank, the other 1949 scanout buffer is immediately synchronized and flipped to during the 1950 target vertical blank period. For Present flips not synchronized to 1951 vertical blank, we simply use the MSC and timestamp values of the last 1952 vertical blank period for timing purposes, and let the normal TearFree 1953 mechanism handle display updates. 1954 1955 v2: 1956 * Move manpage hunk to next change, since TearFree can still prevent 1957 DRI2 page flipping with this change. 1958 1959 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> # v1 1960 1961commit 65e0c5ea1b4adff21d673dbf54af99704c429627 1962Author: Michel Dänzer <michel.daenzer@amd.com> 1963Date: Wed Feb 8 17:52:28 2017 +0900 1964 1965 Pass extents to radeon_scanout_do_update 1966 1967 Preparation for following change, no functional change intended yet. 1968 1969 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1970 1971commit 1443270e52e8562bd8dc3603f301963bd4027cef 1972Author: Michel Dänzer <michel.daenzer@amd.com> 1973Date: Tue Feb 7 18:58:23 2017 +0900 1974 1975 Add source drawable parameter to radeon_scanout_do_update 1976 1977 Preparation for following changes, no functional change intended yet. 1978 1979 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1980 1981commit 7c10ee9c88378d773c0bcf651fdc5d9f2c6dc5e5 1982Author: Michel Dänzer <michel.daenzer@amd.com> 1983Date: Thu Aug 3 18:16:36 2017 +0900 1984 1985 Handle multiple "pending" Present flips 1986 1987 The xserver Present code can submit a flip in response to notifying it 1988 that a vblank event arrived. This can happen before the completion event 1989 of the previous flip is processed. In that case, we were clearing the 1990 drmmode_crtc->flip_pending field prematurely. 1991 1992 Prevent this by only clearing drmmode_crtc->flip_pending when it matches 1993 the framebuffer being scanned out since the flip whose completion event 1994 we're processing. 1995 1996 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 1997 1998commit e6d7dc2070f4d21a6900916bb70a31839112882c 1999Author: Michel Dänzer <michel.daenzer@amd.com> 2000Date: Thu Aug 3 17:51:20 2017 +0900 2001 2002 Wait for pending flips synchronously before turning off a CRTC 2003 2004 Allows removing drmmode_clear_pending_flip and the pending_dpms_mode 2005 field and cleaning up the code considerably. 2006 2007 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2008 2009commit f87acdbfb1b0b6d2769764772a52ea8b81675e20 2010Author: Michel Dänzer <michel.daenzer@amd.com> 2011Date: Fri Jul 28 12:28:53 2017 +0900 2012 2013 Create drmmode_crtc_wait_pending_event helper macro 2014 2015 Preparation for following change, no functional change intended yet. 2016 2017 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2018 2019commit 99f1d7a474af3683fe1a66f50c0bb8935478ff0a 2020Author: Michel Dänzer <michel.daenzer@amd.com> 2021Date: Mon Aug 14 12:23:04 2017 +0900 2022 2023 Create drmmode_wait_vblank helper 2024 2025 Allows cleaning up the code considerably. 2026 2027 v2: 2028 * Fix "drmWaiVBlank" typo, add blank line for readability (Slava Abramov) 2029 * Rename in/out sequence parameters to "target_seq" and "result_seq", 2030 hopefully that will be clearer. 2031 2032 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> # v1 2033 2034commit 49cc61ab970ee28d4509b4e2dd0a57165136889f 2035Author: Michel Dänzer <michel.daenzer@amd.com> 2036Date: Tue Aug 1 18:45:53 2017 +0900 2037 2038 Pass reference CRTC to radeon_do_pageflip directly 2039 2040 Simplifies the code slightly. 2041 2042 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2043 2044commit 3f120fa1d5d921656a367751bc079e020e9ab105 2045Author: Michel Dänzer <michel.daenzer@amd.com> 2046Date: Thu Aug 3 17:42:51 2017 +0900 2047 2048 Remove drmmode_crtc->scanout_destroy[] array 2049 2050 No longer necessary since we're reference counting framebuffers. 2051 2052 Reviewed-by: Alex Deucher <alexander.deucher@amd.com 2053 2054commit 01b040b4a80754b8c8a305cbe968416182b54246 2055Author: Michel Dänzer <michel.daenzer@amd.com> 2056Date: Tue Aug 15 17:15:27 2017 +0900 2057 2058 Adapt to PixmapDirtyUpdateRec::src being a DrawablePtr 2059 2060commit 5309bde0c4e28adf2b167191c6d7011a19e31eed 2061Author: Michel Dänzer <michel.daenzer@amd.com> 2062Date: Thu Jul 27 16:11:19 2017 +0900 2063 2064 Allow DRI page flipping when some CRTCs use separate scanout buffers 2065 2066 As long as the CRTC we're synchronizing to doesn't. 2067 2068 v2: 2069 * Remove redundant checks from can_exchange which still prevented DRI2 2070 page flipping 2071 2072 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2073 2074commit 9bc3eef74452d924f9101c024f66ad9b14c404c8 2075Author: Michel Dänzer <michel.daenzer@amd.com> 2076Date: Thu Jul 27 15:46:41 2017 +0900 2077 2078 Add drmmode_crtc_can_flip helper 2079 2080 To reduce code duplication between DRI2 and Present. No functional 2081 change intended yet. 2082 2083 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2084 2085commit c2d26890691ec105858f086b63170ad94c6f7f05 2086Author: Michel Dänzer <michel.daenzer@amd.com> 2087Date: Wed Jul 12 12:13:03 2017 +0900 2088 2089 Use root window (pixmap) instead of screen pixmap for scanout updates 2090 2091 Preparation for following changes, no functional change intended yet. 2092 2093 v2: 2094 * Add drmmode_screen_damage_destroy callback to prevent use-after-free 2095 on server shutdown 2096 2097 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2098 2099commit 377c9dc7e51cc8ebd06816b08682a8098a49508a 2100Author: Michel Dänzer <michel.daenzer@amd.com> 2101Date: Thu Jul 27 15:27:54 2017 +0900 2102 2103 autogen.sh: Pass -f to autoreconf 2104 2105 To ensure that any existing copies of autotools files will be replaced 2106 with the current versions. 2107 2108 (Cherry picked from amdgpu commit a47c0093338d80d84e7033ad15d051925d542ca0) 2109 Acked-by: Alex Deucher <alexander.deucher@amd.com> 2110 2111commit bf0a759e3bced3a93bead033a26871cc340bd309 2112Author: Michel Dänzer <michel.daenzer@amd.com> 2113Date: Thu Jul 27 15:26:23 2017 +0900 2114 2115 Makefile.am: Set ACLOCAL_AMFLAGS = -I m4 2116 2117 Suggested by one of the tools called by autoreconf. 2118 2119 (Cherry picked from amdgpu commit 842bad4b951296ca25f47b50cb358e502bf30ebb) 2120 Acked-by: Alex Deucher <alexander.deucher@amd.com> 2121 2122commit e628f604d032d32971a1ca8670469e88939e3de2 2123Author: Michel Dänzer <michel.daenzer@amd.com> 2124Date: Thu Jul 27 15:24:37 2017 +0900 2125 2126 Add AC_CONFIG_MACRO_DIRS([m4]) to configure.ac 2127 2128 Suggested by one of the tools called by autoreconf. 2129 2130 (Cherry picked from amdgpu commit 227b399badaad9bbef0be5a776ce008d0d243449) 2131 Acked-by: Alex Deucher <alexander.deucher@amd.com> 2132 2133commit 7d7abf99b5441ddb04dbee99bc8fa7abc30d4c46 2134Author: Michel Dänzer <michel.daenzer@amd.com> 2135Date: Wed Jul 26 18:13:35 2017 +0900 2136 2137 Only handle reflection in the driver with Xorg < 1.16 2138 2139 Xorg doesn't handle the hardware cursor correctly in that case for 2140 rotation and general transforms, and we can't force the SW cursor. 2141 2142 Fixes: f2bc882f1c10 ("Handle rotation in the driver also with Xorg 2143 1.12-1.18") 2144 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2145 2146commit 94dc2b80f3ef0b2c17c20501d824fb0447d52e7a 2147Author: Michel Dänzer <michel.daenzer@amd.com> 2148Date: Tue Mar 14 16:57:17 2017 +0900 2149 2150 If a TearFree flip fails, fall back to non-TearFree operation 2151 2152 In order to avoid possible freeze / log file spam in that case. 2153 2154 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99769 2155 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2156 2157commit aff267ee36cc6a703a532f91f82adc1ba1425ff3 2158Author: Michel Dänzer <michel.daenzer@amd.com> 2159Date: Tue Mar 14 16:23:45 2017 +0900 2160 2161 Use drmmode_crtc->scanout_id instead of 0 to check for scanout buffer 2162 2163 Preparation for following change, no functional change intended. 2164 2165 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2166 2167commit df88d51c5005476e6fbfca3f31b54be079ec5aaa 2168Author: Michel Dänzer <michel.daenzer@amd.com> 2169Date: Tue Jun 27 19:15:18 2017 +0900 2170 2171 Use pRADEONEnt->fd exclusively for the DRM file descriptor 2172 2173 This brings us closer to amdgpu. 2174 2175 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2176 2177commit 19626bce4e5e31c863eedb503ea3884ac3f60bea 2178Author: Michel Dänzer <michel.daenzer@amd.com> 2179Date: Tue Jun 27 18:32:28 2017 +0900 2180 2181 Improve drmmode_fb_reference debugging code 2182 2183 If a reference count is <= 0, call FatalError with the call location 2184 (in case it doesn't get resolved in the backtrace printed by 2185 FatalError). 2186 2187 (Ported from amdgpu commit 1b6ff5fd9933c00ec1ec90dfc62e0b531927749b) 2188 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2189 2190commit bc46ffdf71ab3dfa0f95572529e818f2b619d380 2191Author: Michel Dänzer <michel.daenzer@amd.com> 2192Date: Tue Jun 27 18:19:31 2017 +0900 2193 2194 Increase reference count of FB assigned to drmmode_crtc->flip_pending 2195 2196 Otherwise, it could happen that we destroy the FB before the flip 2197 completes, resulting in use-after-free and most likely a crash. 2198 2199 (Ported from amdgpu commit af7221e1c4d2dbdfd488eb0976a835584ea8441c) 2200 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2201 2202commit c9dd28cb0c9c3de676eadac61e727732510f6b9b 2203Author: Michel Dänzer <michel.daenzer@amd.com> 2204Date: Thu Jun 22 12:38:55 2017 +0900 2205 2206 Only call drmmode_scanout_free for non-GPU screens in LeaveVT 2207 2208 Destroying the scanout buffers of GPU screens resulted in a crash when 2209 switching back to the Xorg VT. 2210 2211 Fixes: 4cfa4615f79f ("Use drmmode_crtc_scanout_* helpers for RandR 1.4 2212 scanout pixmaps") 2213 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2214 2215commit 78fad9ca1635cca04b89807e3db029477cf46681 2216Author: Michel Dänzer <michel.daenzer@amd.com> 2217Date: Mon Jun 26 18:58:30 2017 +0900 2218 2219 Include xf86Pci.h for DRICreatePCIBusID with xserver Git master 2220 2221 The declaration has been moved there from dri.h. 2222 2223 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2224 2225commit ef95e3a3cc9189c8906fe5b4d70bfa9f1055d7ce 2226Author: Adam Jackson <ajax@redhat.com> 2227Date: Tue Jun 13 09:32:44 2017 -0400 2228 2229 modesetting: Validate the atom for enum properties 2230 2231 The client could have said anything here, and if what they said doesn't 2232 actually name an atom NameForAtom() will return NULL, and strcmp() will 2233 be unhappy about that. 2234 2235 [copied from xserver d4995a3936ae283b9080fdaa0905daa669ebacfc] 2236 2237 Signed-off-by: Adam Jackson <ajax@redhat.com> 2238 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 2239 2240commit d80d01a73c2eaba2e3649b7bc0a3541b3ff782f6 2241Author: Michel Dänzer <michel.daenzer@amd.com> 2242Date: Thu May 18 11:12:34 2017 +0900 2243 2244 Update URLs 2245 2246 * Point to the amd-gfx mailing list 2247 * Specify the component in all bugzilla URLs 2248 * Use https:// for all HTML URLs 2249 2250 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2251 2252commit 2b7d77b90108911777a11ecaa63435552000c958 2253Author: Eric Anholt <eric@anholt.net> 2254Date: Tue May 16 11:04:29 2017 -0700 2255 2256 Use plain glamor_egl_create_textured_screen(). 2257 2258 Since 5064ffab631 (2014), glamor's implementation of _ext just drops the 2259 back_pixmap arg, which we were passing NULL (the default) to anyway. 2260 2261 Signed-off-by: Eric Anholt <eric@anholt.net> 2262 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 2263 2264commit 3a9f7e87c30b9f6334d38dfdd1cb15b28fe37fcb 2265Author: Michel Dänzer <michel.daenzer@amd.com> 2266Date: Thu May 11 18:57:25 2017 +0900 2267 2268 Remove unused struct members from drmmode_display.h 2269 2270 (Ported from amdgpu commit 462ac3341e5bfbded9086d3d9043821d19352b3e) 2271 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2272 2273commit 7dc68e26755466f9056f8c72195ab8690660693d 2274Author: Michel Dänzer <michel.daenzer@amd.com> 2275Date: Tue May 2 11:47:58 2017 +0900 2276 2277 Simplify tracking of PRIME scanout pixmap 2278 2279 Remember the shared pixmap passed to drmmode_set_scanout_pixmap for each 2280 CRTC, and just compare against that. 2281 2282 Fixes leaving stale entries in ScreenRec::pixmap_dirty_list under some 2283 circumstances, which would usually result in use-after-free and a crash 2284 down the line. 2285 2286 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2287 2288commit 55e513b978b2afc52b7cafc5bfcb0d1dc78d75f6 2289Author: Michel Dänzer <michel.daenzer@amd.com> 2290Date: Sat Apr 29 17:05:58 2017 +0900 2291 2292 Use reference counting for tracking KMS framebuffer lifetimes 2293 2294 References are held by the pixmaps corresponding to the FBs (so 2295 the same KMS FB can be reused as long as the pixmap exists) and by the 2296 CRTCs scanning out from them (so a KMS FB is only destroyed once it's 2297 not being scanned out anymore, preventing intermittent black screens and 2298 worse issues due to a CRTC turning off when it should be on). 2299 2300 v2: 2301 * Only increase reference count in drmmode_fb_reference if it was sane 2302 before 2303 * Make drmmode_fb_reference's indentation match the rest of 2304 drmmode_display.h 2305 2306 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2307 2308commit 944391b0052466b71bf9919b56139dc197a7e072 2309Author: Michel Dänzer <michel.daenzer@amd.com> 2310Date: Wed Feb 8 18:14:30 2017 +0900 2311 2312 Pass pixmap instead of handle to radeon_do_pageflip 2313 2314 This brings us in line with amdgpu and prepares for the following 2315 change, no functional change intended. 2316 2317 (Ported from amdgpu commit e463b849f3e9d7b69e64a65619a22e00e78d297b) 2318 2319 v2: 2320 * Be more consistent with the amdgpu code, which should make porting 2321 the following change to amdgpu easier 2322 2323 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2324 2325commit f32c45194ac6f82cbe42d255ed72f857018778e0 2326Author: Michel Dänzer <michel.daenzer@amd.com> 2327Date: Tue May 9 11:18:00 2017 +0900 2328 2329 Apply gamma correction to HW cursor 2330 2331 The display hardware CLUT we're currently using for gamma correction 2332 doesn't affect the HW cursor, so we have to apply it manually when 2333 uploading the HW cursor data. 2334 2335 This currently only works in depth 24/32. 2336 2337 (Ported from amdgpu commit 82fa615f38137add75f9cd4bb49c48dd88de916f) 2338 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2339 2340commit 4c91f36d3058180b5a2d6a23e9b82f5c933d8716 2341Author: Michel Dänzer <michel.daenzer@amd.com> 2342Date: Mon Apr 17 18:32:31 2017 +0900 2343 2344 Don't set modes before RADEONWindowExposures_oneshot is called 2345 2346 The root window contents may be undefined before that, so we don't want 2347 to show anything yet. 2348 2349 Fixes a crash on startup with rotation and virtual resolution set in 2350 xorg.conf. 2351 2352 Bugzilla: https://bugs.freedesktop.org/100276 2353 Fixes: cc9d6b7db9c2 ("Move DPMS check from radeon_scanout_do_update to 2354 radeon_scanout_flip") 2355 (Ported from amdgpu commit 981bac185cfd74ae50dffc28f57cf34623a9595f) 2356 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2357 2358commit d6c29250dbc293c932ca14997eb23399610b106d 2359Author: Daniel Stone <daniels@collabora.com> 2360Date: Fri Apr 7 14:27:58 2017 +0100 2361 2362 Set correct DRM event context version 2363 2364 DRM_EVENT_CONTEXT_VERSION is the latest context version supported by 2365 whatever version of libdrm is present. We were blindly asserting we 2366 supported whatever version that may be, even if we actually didn't. 2367 2368 Set the version as 2, which should be bumped only with the appropriate 2369 version checks. 2370 2371 Signed-off-by: Daniel Stone <daniels@collabora.com> 2372 (Ported from xserver commit 0c8e6ed85810e96d84173a52d628863802a78d82) 2373 v2: Remove second paragraph of commit log, we always initialize 2374 page_flip_handler2 = NULL (Emil Velikov) 2375 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 2376 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> # v1 2377 2378commit 23985d3ef3150ef8cad30b92ef0cd8ba6cfa1f4d 2379Author: Michel Dänzer <michel.daenzer@amd.com> 2380Date: Thu Mar 23 18:06:45 2017 +0900 2381 2382 manpage: Don't put "'" at the beginning of a line 2383 2384 It caused the whole line to be dropped. 2385 2386 Fixes: 58cd1600057e ("Allow toggling TearFree at runtime via output 2387 property") 2388 Reported-by: Andy Furniss <adf.lists@gmail.com> 2389 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (amdgpu commit) 2390 (Ported from amdgpu commit 165b51447643ce37f391f25ca6aecb8d76fabaa3) 2391 2392commit dc30f5e27ef8a8e980adb21c1fea7c4f60c7e896 2393Author: Michel Dänzer <michel.daenzer@amd.com> 2394Date: Thu Mar 23 18:06:23 2017 +0900 2395 2396 Post-release version bump 2397 2398commit 66227060872219c1a0dc84fa4cea6264f7548446 2399Author: Michel Dänzer <michel.daenzer@amd.com> 2400Date: Thu Mar 16 16:38:15 2017 +0900 2401 2402 Bump version for 7.9.0 release 2403 2404commit 9a71445094b728f3d78db8f6808b4782ee19a453 2405Author: Michel Dänzer <michel.daenzer@amd.com> 2406Date: Wed Mar 8 17:42:56 2017 +0900 2407 2408 Pass TRUE to drmmode_set_desired_modes the first time for GPU screens 2409 2410 This is the only place we call drmmode_set_desired_modes for GPU screens 2411 during server startup. Without this change, the display outputs of 2412 secondary GPUs may stay on even while Xorg isn't using them. 2413 2414 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2415 2416commit 67ae5e00a748ad52cf92738d401afff2947b1891 2417Author: Michel Dänzer <michel.daenzer@amd.com> 2418Date: Wed Mar 8 11:20:30 2017 +0900 2419 2420 Skip some initialization steps for GPU screens 2421 2422 Xorg doesn't use the following functionality of GPU screens, so don't 2423 bother initializing it: 2424 2425 * DRI page flipping 2426 * DRI3 / Present / SYNC fences 2427 * XVideo / XvMC 2428 * Root window with background None 2429 2430 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2431 2432commit f0e7948e1c0e984fc27f235f365639e9cf628291 2433Author: Jochen Rollwagen <joro-2013@t-online.de> 2434Date: Sun Mar 5 11:32:38 2017 +0100 2435 2436 Fix build for XServer 1.13 2437 2438 Latest git build stops with the error message 2439 2440 radeon_kms.c: In function 'RADEONWindowExposures_oneshot': 2441 radeon_kms.c:1644:45: error: expected expression before 'RegionPtr' 2442 pScreen->WindowExposures(pWin, pRegion, RegionPtr pBSRegion); 2443 2444 This patch fixes the build. 2445 2446 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 2447 2448commit d63881623f0686a66a2e3e3c1f84e496aa52ec6b 2449Author: Michel Dänzer <michel.daenzer@amd.com> 2450Date: Thu Mar 2 16:22:51 2017 +0900 2451 2452 Don't call radeon_cs_flush_indirect & radeon_bo_wait in drmmode_copy_fb 2453 2454 RADEONWindowExposures_oneshot takes care of it. 2455 2456 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2457 2458commit cc9d6b7db9c2078be1e530a64af6d517c6a42024 2459Author: Michel Dänzer <michel.daenzer@amd.com> 2460Date: Wed Mar 1 17:35:59 2017 +0900 2461 2462 Move DPMS check from radeon_scanout_do_update to radeon_scanout_flip 2463 2464 When radeon_scanout_do_update is called from 2465 drmmode_crtc_scanout_update, drmmode_crtc->pending_dpms_mode may still 2466 be != DPMSModeOn, e.g. during server startup. 2467 2468 Fixes intermittently showing garbage with TearFree enabled. 2469 2470commit 0a12bf1085505017068dfdfd31d23133e51b45b9 2471Author: Michel Dänzer <michel.daenzer@amd.com> 2472Date: Wed Mar 1 18:00:40 2017 +0900 2473 2474 Call drmmode_set_desired_modes from a WindowExposures hook 2475 2476 This is the earliest opportunity where the root window contents are 2477 guaranteed to be initialized, and prevents drmmode_set_mode_major from 2478 getting called before drmmode_set_desired_modes via RADEONUnblank -> 2479 drmmode_crtc_dpms. Also, in contrast to the BlockHandler hook, this is 2480 called when running Xorg with -pogo. 2481 2482 Fixes intermittently showing garbage on server startup or after server 2483 reset. 2484 2485 As a bonus, this avoids trouble due to higher layers (e.g. the tigervnc 2486 Xorg module) calling RADEONBlockHandler_oneshot repeatedly even after 2487 we set pScreen->BlockHandler = RADEONBlockHandler_KMS. 2488 2489 v2: 2490 * Drop spaces between XORG_VERSION_NUMERIC arguments 2491 * Call radeon_bo_wait after radeon_cs_flush_indirect 2492 2493 Bugzilla: https://bugs.freedesktop.org/99457 2494 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (v1) 2495 2496commit df2d749a4db33298c8ce9f2cfb77c20c5538c9cc 2497Author: Michel Dänzer <michel.daenzer@amd.com> 2498Date: Thu Mar 2 16:05:42 2017 +0900 2499 2500 Fix bogus indentation 2501 2502 Trivial. 2503 2504 Fixes: 58cd1600057e ("Allow toggling TearFree at runtime via output property") 2505 2506commit 9035b6abea557828e672ee455f0c84e43da0906f 2507Author: Michel Dänzer <michel.daenzer@amd.com> 2508Date: Tue Feb 28 18:13:42 2017 +0900 2509 2510 present: Flush before flipping 2511 2512 This isn't necessary for DRI clients, but the Present extension can also 2513 be used for presenting normal pixmaps rendered to via the X11 protocol. 2514 2515 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2516 2517commit 0a4eb0e12f0c9c653cf4cea6fd62e1a507eb261c 2518Author: Michel Dänzer <michel.daenzer@amd.com> 2519Date: Fri Feb 10 12:52:02 2017 +0900 2520 2521 present: Use async flip for unflip if possible 2522 2523 In that case, unflip operations should finish faster in general. 2524 2525 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2526 2527commit 9a951a3e551db58ba50e7a594521ceac54d90615 2528Author: Michel Dänzer <michel.daenzer@amd.com> 2529Date: Tue Feb 28 18:06:55 2017 +0900 2530 2531 present: Also flush before using a flip to unflip 2532 2533 Not doing so might result in intermittently scanning out stale contents 2534 of the screen pixmap. 2535 2536 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2537 2538commit f2bc882f1c1082bed9f496cfab6c8f07a76bc122 2539Author: Michel Dänzer <michel.daenzer@amd.com> 2540Date: Tue Feb 28 17:31:14 2017 +0900 2541 2542 Handle rotation in the driver also with Xorg 1.12-1.18 2543 2544 We cannot use the HW cursor in that case, but in turn we get more 2545 efficient and less teary updates of rotated outputs. 2546 2547 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2548 2549commit ae921a3150f69c38b5b3c88a9e37d54fdf0d5093 2550Author: Michel Dänzer <michel.daenzer@amd.com> 2551Date: Tue Feb 28 17:47:09 2017 +0900 2552 2553 Fold drmmode_crtc_scanout_allocate into drmmode_crtc_scanout_create 2554 2555 Not used anywhere else anymore. 2556 2557 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2558 2559commit 987a34adb319923ad36e2b47a26837248f187c3e 2560Author: Michel Dänzer <michel.daenzer@amd.com> 2561Date: Tue Feb 28 17:17:20 2017 +0900 2562 2563 Call drmmode_crtc_scanout_create in drmmode_crtc_shadow_allocate as well 2564 2565 Calling drmmode_crtc_scanout_allocate in drmmode_crtc_shadow_allocate 2566 resulted in drmmode_crtc_scanout_create called from 2567 drmmode_crtc_shadow_create passing an uninitialized pitch value to 2568 drmmode_create_bo_pixmap. 2569 2570 Fixes issues such as failure to allocate the scanout pixmap or visual 2571 corruption and GPUVM faults when attempting to use rotation with Xorg 2572 <1.19. 2573 2574 Bugzilla: https://bugs.freedesktop.org/99916 2575 Fixes: ea30d856ba5e ("Pass pitch from drmmode_crtc_scanout_allocate to drmmode_create_bo_pixmap") 2576 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2577 2578commit 244d4bc7f8c8f6bc90f49556c0b9344c8aa40295 2579Author: Michel Dänzer <michel.daenzer@amd.com> 2580Date: Thu Feb 9 18:55:04 2017 +0900 2581 2582 present: Wait for screen pixmap BO idle before setting modes for unflip 2583 2584 To make sure the screen pixmap contents are up to date when it starts 2585 being scanned out. 2586 2587 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2588 2589commit 3ff29e5a14451916bc66b4e0028e9a317f0723f8 2590Author: Michel Dänzer <michel.daenzer@amd.com> 2591Date: Thu Feb 9 18:50:40 2017 +0900 2592 2593 present: Only call drmModeRmFB after setting modes for unflip 2594 2595 Fixes display intermittently blanking when a modeset is used for unflip. 2596 2597 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2598 2599commit 0c29deb5a97d9a57e994cc0053c49ddf7aca6ecb 2600Author: Michel Dänzer <michel.daenzer@amd.com> 2601Date: Tue Feb 21 16:56:27 2017 +0900 2602 2603 Use drmmode_crtc_scanout_free in drmmode_fini 2604 2605 We were leaking drmmode_crtc->scanout_damage, which caused trouble on 2606 server reset. Fixes server reset with active separate scanout pixmaps. 2607 2608 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2609 2610commit 58cd1600057e41aade0106d4acf78e23eac6e44f 2611Author: Michel Dänzer <michel.daenzer@amd.com> 2612Date: Thu Dec 1 18:37:57 2016 +0900 2613 2614 Allow toggling TearFree at runtime via output property 2615 2616 Option "TearFree" now sets the default value of the output property. 2617 See the manpage update for details. 2618 2619 TearFree is now enabled by default for outputs using rotation or other 2620 RandR transforms, and for RandR 1.4 slave outputs. 2621 2622 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2623 2624commit 305e2cbf335837a2ab6a24e9ff65815afe038296 2625Author: Michel Dänzer <michel.daenzer@amd.com> 2626Date: Tue Feb 21 15:55:13 2017 +0900 2627 2628 Factor out drmmode_crtc_scanout_update helper 2629 2630 Cleanup in preparation for following change, no functional change 2631 intended. 2632 2633 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2634 2635commit 649644a88347a6d03de68f8c41db03a82deeb23b 2636Author: Michel Dänzer <michel.daenzer@amd.com> 2637Date: Tue Feb 14 17:36:31 2017 +0900 2638 2639 Factor out radeon_prime_dirty_to_crtc helper 2640 2641 Cleanup in preparation for the following change, no functional change 2642 intended. 2643 2644 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2645 2646commit 1351e48efe7a2c28eab447e16f36a00fbd02ae48 2647Author: Michel Dänzer <michel.daenzer@amd.com> 2648Date: Tue Feb 7 18:43:56 2017 +0900 2649 2650 Don't destroy current FB if drmModeAddFB fails 2651 2652 It would probably result in a black screen. 2653 2654 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2655 2656commit 481394e3c9f9f7d88bb66fe9ae8834c87952a8ab 2657Author: Michel Dänzer <michel.daenzer@amd.com> 2658Date: Tue Feb 7 18:39:05 2017 +0900 2659 2660 Fix flip event data leak if calloc or drmModeAddFB fails 2661 2662 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2663 2664commit 80fc78633962c1ec996b1721bfb757b98dc85bfc 2665Author: Michel Dänzer <michel.daenzer@amd.com> 2666Date: Tue Jan 31 16:25:37 2017 +0900 2667 2668 Don't handle Option "SwapbuffersWait" at all with glamor 2669 2670 It never had any effect with glamor. 2671 2672 v2: Better formatting, no functional change from v1. 2673 2674 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (v1) 2675 2676commit d611d21d9bf9aaa100ff7a5762e5980634846207 2677Author: Michel Dänzer <michel.daenzer@amd.com> 2678Date: Tue Jan 24 18:05:45 2017 +0900 2679 2680 Enable tiling by default with glamor on PALM 2681 2682 The DFS check is only relevant for EXA. 2683 2684 The lack of tiling prevented DRI3 clients from using page flipping, 2685 resulting in tearing. 2686 2687 Bugzilla: https://bugs.freedesktop.org/99491 2688 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2689 2690commit 55a33205cb9bd8091b4bcd478c2d1d191248681b 2691Author: Michel Dänzer <michel.daenzer@amd.com> 2692Date: Tue Jan 24 18:22:49 2017 +0900 2693 2694 Add 10-radeon.conf xorg.conf.d snippet 2695 2696 This instructs Xorg >= 1.16 to try loading the radeon driver for devices 2697 managed by the radeon kernel driver, even if the ati wrapper driver 2698 isn't available. 2699 2700 Copied from the amdgpu driver. 2701 2702 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2703 2704commit 8e6a4e96b7b27559e186f71b5547abb0a80b96dd 2705Author: Mihail Konev <k.mvc@ya.ru> 2706Date: Thu Jan 26 14:00:22 2017 +1000 2707 2708 autogen: add default patch prefix 2709 2710 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 2711 2712commit e062933b5489705bdcba7a266d14ba6fd3e767ba 2713Author: Emil Velikov <emil.l.velikov@gmail.com> 2714Date: Mon Mar 9 12:00:52 2015 +0000 2715 2716 autogen.sh: use quoted string variables 2717 2718 Place quotes around the $srcdir, $ORIGDIR and $0 variables to prevent 2719 fall-outs, when they contain space. 2720 2721 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 2722 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 2723 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 2724 2725commit 4fe183111f82d412760b175b3c9c8583acba804a 2726Author: Peter Hutterer <peter.hutterer@who-t.net> 2727Date: Tue Jan 24 10:32:07 2017 +1000 2728 2729 autogen.sh: use exec instead of waiting for configure to finish 2730 2731 Syncs the invocation of configure with the one from the server. 2732 2733 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 2734 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 2735 2736commit a1787e4615a76ab0bb12498be66591c86a6b07ae 2737Author: Michel Dänzer <michel.daenzer@amd.com> 2738Date: Wed Jan 18 19:07:11 2017 +0900 2739 2740 ati: Support loading the amdgpu driver from the ati wrapper 2741 2742 If .../share/X11/xorg.conf.d/10-amdgpu.conf doesn't exist, but the ati 2743 wrapper is loaded, it will otherwise try to use the radeon driver even 2744 for GPUs driven by the amdgpu kernel driver. This can only fail, 2745 potentially in bad ways. 2746 2747 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2748 2749commit 41b82c776b571e1556f300d77c715aebdd87d8e6 2750Author: Michel Dänzer <michel.daenzer@amd.com> 2751Date: Tue Jan 24 18:14:23 2017 +0900 2752 2753 .editorconfig: src/ati.c only uses spaces for indentation 2754 2755 Trivial. 2756 2757commit ea30d856ba5e7274c8ea499293b8b0e721b8e082 2758Author: Michel Dänzer <michel.daenzer@amd.com> 2759Date: Fri Jan 6 17:36:05 2017 +0900 2760 2761 Pass pitch from drmmode_crtc_scanout_allocate to drmmode_create_bo_pixmap 2762 2763 radeon_bo_get_tiling can fail, e.g. for linear BOs. 2764 2765 Fixes TearFree not working in those cases. 2766 2767 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97873 2768 Fixes: 040a7b80e1fc ("Explicitly set the fbcon pixmap pitch again") 2769 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2770 2771commit 720dd07e79b5f5cbbf3ef8bd92300578c485d583 2772Author: Jammy Zhou <Jammy.Zhou@amd.com> 2773Date: Mon Dec 19 17:27:19 2016 +0900 2774 2775 Use render node for DRI3 if available 2776 2777 Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> 2778 (ported from amdgpu commit ea558e645786b08d75307716036045170e97b43e) 2779 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2780 [ Second attempt, let's see if there's any fallout this time... ] 2781 2782commit 92fc387a6256ce3c95d62142735783ea07162c54 2783Author: Michel Dänzer <michel.daenzer@amd.com> 2784Date: Thu Dec 15 12:44:47 2016 +0900 2785 2786 Simplify drmmode_handle_uevents 2787 2788 No functional change intended. 2789 2790 Reviewed-by: Jim Qu <Jim.Qu@amd.com> 2791 (Cherry picked from amdgpu commit edd276185d42962a13faf9ec9eeebc754ef284e7.) 2792 2793commit 404c64752774d23a2db9aa3134f6ada3155708c9 2794Author: jimqu <Jim.Qu@amd.com> 2795Date: Wed Dec 14 11:46:16 2016 +0900 2796 2797 udev_monitor_receive_device() will block when hotplug monitor 2798 2799 udev_monitor_receive_device() will block and wait for the event of udev 2800 use select() to ensure that this will not block. 2801 2802 Signed-off-by: JimQu <Jim.Qu@amd.com> 2803 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 2804 (Cherry picked from amdgpu commit 732cf4d3a248b288532ad0f3443da49e08dc7507) 2805 2806commit a89c0bced423ce3e09f66edafc35642a7cc03b2f 2807Author: Jochen Rollwagen <joro-2013@t-online.de> 2808Date: Fri Dec 2 10:56:39 2016 +0100 2809 2810 Calculate log base 2 in radeon.h based on clz for all platforms 2811 2812 This commit replaces the inline assembler code (for x86 platforms) and 2813 loop (for non-x86 platforms) in RADEONLog2 with a one-liner version 2814 based on clz (count leading zeroes). 2815 2816 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 2817 2818commit e2942449171fe628a7726e59bcaab65e27d88563 2819Author: Michel Dänzer <michel.daenzer@amd.com> 2820Date: Mon Nov 21 18:47:53 2016 +0900 2821 2822 Call radeon_drm_abort_entry on failure to flip to a scanout pixmap 2823 2824 Fixes leaking the corresponding struct radeon_drm_queue list entry in 2825 that case. 2826 2827 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2828 2829commit 14c3f59f5157885ad8f941f0bad6c0c5e3db12f8 2830Author: Michel Dänzer <michel.daenzer@amd.com> 2831Date: Mon Nov 21 18:45:30 2016 +0900 2832 2833 Call ValidateGC after ChangeClip in radeon_sync_scanout_pixmaps 2834 2835 The wrong order meant that the clipping region wasn't actually applied, 2836 so it always copied the full contents from the other scanout pixmap. 2837 2838 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2839 2840commit a995f5830916a0fee5126263d1bfe48632be3a15 2841Author: Michel Dänzer <michel.daenzer@amd.com> 2842Date: Mon Nov 21 18:39:00 2016 +0900 2843 2844 Fix radeon_scanout_extents_intersect for GPU screens 2845 2846 Fixes incorrect screen updates with TearFree enabled on PRIME slave 2847 outputs which are not located at (0, 0). 2848 2849 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2850 2851commit e543ef3a2fb304cbe3a965fb780632af2e4186f4 2852Author: Michel Dänzer <michel.daenzer@amd.com> 2853Date: Mon Nov 21 18:33:25 2016 +0900 2854 2855 Take current scanout_id into account everywhere involved with TearFree 2856 2857 Fixes various potential issues with TearFree enabled, e.g. outputs 2858 freezing after display configuration changes. 2859 2860 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2861 2862commit f130b10e63f7526360b41aa0918b4940f63f662a 2863Author: Michel Dänzer <michel.daenzer@amd.com> 2864Date: Fri Nov 25 18:12:40 2016 +0900 2865 2866 Add radeon_is_gpu_screen helper 2867 2868 This will hopefully decrease the chance of accidentally breaking the 2869 build against xserver < 1.13 in the future. 2870 2871 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2872 2873commit 487aa62a2a23b86e4ea4714fdfd465c9e513141f 2874Author: Jochen Rollwagen <joro-2013@t-online.de> 2875Date: Thu Nov 24 18:44:01 2016 +0100 2876 2877 fix build for xserver < 1.13 2878 2879 same procedure every few patches..... 2880 2881 Fixes: 13c6bc5e3827 ("Don't install Flush/EventCallback for GPU screens") 2882 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 2883 2884commit 5fea5ef2f07eee4a0f94baab427010b936f1d4b4 2885Author: Michel Dänzer <michel.daenzer@amd.com> 2886Date: Thu Nov 24 18:45:22 2016 +0900 2887 2888 Pass fb_id into drmmode_page_flip_target_absolute/relative 2889 2890 drmmode->fb_id isn't what we need in the TearFree case. Fixes TearFree 2891 freezing with 2892 2893 (WW) RADEON(0): flip queue failed in radeon_scanout_flip: No such file or directory 2894 2895 in the log file. 2896 2897 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98834 2898 Fixes: 1106b2f773ad ("Use DRM_MODE_PAGE_FLIP_TARGET_ABSOLUTE/RELATIVE flags when available") 2899 2900commit 13c6bc5e382765fe567091e1c616c0a26eec04ca 2901Author: Michel Dänzer <michel.daenzer@amd.com> 2902Date: Mon Aug 1 18:11:57 2016 +0900 2903 2904 Don't install Flush/EventCallback for GPU screens 2905 2906 Their purpose is to flush GPU rendering commands corresponding to damage 2907 events, but there can be no damage events corresponding to GPU screen 2908 rendering operations. 2909 2910 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2911 2912commit f11531c99fcd6473f58b4d10efaf3efd84304d8e 2913Author: Michel Dänzer <michel.daenzer@amd.com> 2914Date: Mon Nov 21 18:15:05 2016 +0900 2915 2916 Enable glamor by default with >= R600 and Xorg >= 1.18.3 2917 2918 glamor should now perform at least as well as EXA in general, and this 2919 allows DRI3 to be enabled by default for >= R600. 2920 2921 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2922 2923commit 1106b2f773ad0611c729b27f4c192a26b43ef1e7 2924Author: Michel Dänzer <michel.daenzer@amd.com> 2925Date: Wed Jul 6 15:25:33 2016 +0900 2926 2927 Use DRM_MODE_PAGE_FLIP_TARGET_ABSOLUTE/RELATIVE flags when available 2928 2929 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2930 2931commit aab80d2203efb8569b26dd02a1c651c792e1ef0c 2932Author: Michel Dänzer <michel.daenzer@amd.com> 2933Date: Fri Nov 18 12:11:38 2016 +0900 2934 2935 Post-release version bump 2936 2937commit 5cbe1ee8e499e1b6b2646e341946292721d07e69 2938Author: Michel Dänzer <michel.daenzer@amd.com> 2939Date: Thu Nov 17 11:27:16 2016 +0900 2940 2941 Bump version for 7.8.0 release 2942 2943commit 68023e01842ed7e5c798be77c25993bbe6a5a966 2944Author: Michel Dänzer <michel.daenzer@amd.com> 2945Date: Thu Nov 17 11:32:37 2016 +0900 2946 2947 Update manpage entry for Option "TearFree" 2948 2949 It's now effective for arbitrary transforms as well. 2950 2951commit 9760ef33cba5795eddeda4d5c2fcbe2dcce21689 2952Author: Michel Dänzer <michel.daenzer@amd.com> 2953Date: Tue Nov 8 13:02:43 2016 +0900 2954 2955 Use pRADEONEnt to find both screens of a GPU in radeon_mode_hotplug 2956 2957 Fixes misbehaviour when hotplugging DisplayPort connectors on secondary 2958 GPUs. 2959 2960 Fixes: c801f9f10a5d ("Handle Zaphod mode correctly in radeon_mode_hotplug") 2961 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98626 2962 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2963 2964commit 35bec4937d89b48a79acfcb4f814b7370cb631b2 2965Author: Michel Dänzer <michel.daenzer@amd.com> 2966Date: Tue Nov 8 13:01:58 2016 +0900 2967 2968 Refactor radeon_mode_hotplug 2969 2970 Preparation for the next change, no functional change intended. 2971 2972 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2973 2974commit be9f67339a19a6649eab52aa2e0c8971cd9b4727 2975Author: Michel Dänzer <michel.daenzer@amd.com> 2976Date: Mon Nov 7 16:34:37 2016 +0900 2977 2978 Reindent code in radeon_dri2_create_buffer2 2979 2980 Fixes warning about misleading indentation from recent versions of gcc: 2981 2982 ../../src/radeon_dri2.c: In function ‘radeon_dri2_create_buffer2’: 2983 ../../src/radeon_dri2.c:224:2: warning: this ‘if’ clause does not guard... [-Wmisleading-indentation] 2984 if (flags & RADEON_CREATE_PIXMAP_TILING_MACRO) 2985 ^~ 2986 ../../src/radeon_dri2.c:227:6: note: ...this statement, but the latter is misleadingly indented as if it is guarded by the ‘if’ 2987 if (aligned_width == front_width) 2988 ^~ 2989 2990 No functional change intended. 2991 2992 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 2993 2994commit 350a2645a1b127227ff294c0b62d20000d0fd48a 2995Author: Michel Dänzer <michel.daenzer@amd.com> 2996Date: Tue Nov 1 16:30:42 2016 +0900 2997 2998 Check Xorg version at runtime instead of build time in two places 2999 3000 This means that all possible paths can be handled as intended, no matter 3001 which Xorg version the driver happened to be compiled against. 3002 3003 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3004 3005commit 5df36de39952c3a26cb2fbc125f298139a9dd5bc 3006Author: Michel Dänzer <michel.daenzer@amd.com> 3007Date: Fri Aug 26 18:24:08 2016 +0900 3008 3009 Require xserver 1.10 or newer 3010 3011 1.10.0 was released in February 2011. 3012 3013 We've been accidentally requiring 1.10 or newer since 121a6de72da5 ("Keep 3014 track of damage event related flushes per-client v2"). 3015 3016 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3017 3018commit 22b5ce9548393ba2ff73ee234ecd82eeaf0ef6c4 3019Author: Michel Dänzer <michel.daenzer@amd.com> 3020Date: Tue Oct 25 17:28:03 2016 +0900 3021 3022 Consume all available udev events at once 3023 3024 We get multiple udev events for actions like docking a laptop into its 3025 station or plugging a monitor to the station. By consuming as many 3026 events as we can, we reduce the number of output re-evalutions. 3027 3028 It depends on the timing how many events can be consumed at once. 3029 3030 (Inspired by xserver commit 363f4273dd4aec3e26cc57ecb6c20f27e6c813d8) 3031 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3032 3033commit 82d3c8f5500d2a6fb1495e217a0b79c396f1534c 3034Author: Hans de Goede <hdegoede@redhat.com> 3035Date: Tue Oct 25 16:56:40 2016 +0900 3036 3037 PRIME: Fix swapping of provider sink / source capabilities 3038 3039 When a card has import capability it can be an offload _sink_, not a 3040 source and vice versa for export capability. 3041 3042 This went unnoticed sofar because most gpus have both import and export 3043 capability. 3044 3045 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 3046 (Ported from xserver commit 94a1c77259ce39ba59ad87615df39b570ffab435) 3047 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3048 3049commit 6c940446ddadf418ee4959e46fa552b6c1cf6704 3050Author: Michel Dänzer <michel.daenzer@amd.com> 3051Date: Tue Oct 25 16:42:03 2016 +0900 3052 3053 Always call PixmapStopDirtyTracking in drmmode_set_scanout_pixmap 3054 3055 Otherwise, we may leak screen->pixmap_dirty_list entries if 3056 drmmode_set_scanout_pixmap is called repatedly with ppix != NULL, which 3057 can happen from RRReplaceScanoutPixmap. 3058 3059 (Inspired by xserver commit b773a9c8126222e5fed2904d012fbf917a9f22fd) 3060 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3061 3062commit 61df12e2377cbb19a19ca9d5624df8959822da9f 3063Author: Michel Dänzer <michel.daenzer@amd.com> 3064Date: Wed Oct 19 18:55:33 2016 +0900 3065 3066 Don't rely on randr_crtc->scanout_pixmap in drmmode_set_scanout_pixmap 3067 3068 RRReplaceScanoutPixmap may set randr_crtc->scanout_pixmap = NULL before 3069 we get here. 3070 3071 (Inspired by xserver commit f4c37eeee7953df1fe0e3196eda452acf0078e61) 3072 v2: Always return TRUE in the if (!ppix) block. 3073 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (v1) 3074 3075commit 49cf3b5032a7ce40afe514b7092440e3e19e05aa 3076Author: Michel Dänzer <michel.daenzer@amd.com> 3077Date: Thu Oct 20 15:33:09 2016 +0900 3078 3079 Sayōnara, AM_MAINTAINER_MODE! 3080 3081 If --enable-maintainer-mode got lost from config.status for any reason, 3082 builds would fail in mysterious ways after changing between different 3083 Git commits. 3084 3085 There are more reasons for dropping it in the automake manual: 3086 3087 https://www.gnu.org/software/automake/manual/html_node/maintainer_002dmode.html 3088 3089 I'm not aware of any reason why --disable-maintainer-mode would ever be 3090 useful with this project. 3091 3092 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3093 3094commit 4b17de1cad12e7be12f2c71e5149bfc04564702d 3095Author: Michel Dänzer <michel.daenzer@amd.com> 3096Date: Wed Oct 19 18:09:14 2016 +0900 3097 3098 Remove PCI IDs and bus type from ati_pciids.csv 3099 3100 This cleans up the list of unique chipsets a little further. 3101 3102 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3103 3104commit 1ce1b1656acc6211deb2091ff7f28d51b6daf86b 3105Author: Michel Dänzer <michel.daenzer@amd.com> 3106Date: Wed Oct 19 18:08:36 2016 +0900 3107 3108 Order unique chipsets according to first appearance in ati_pciids.csv 3109 3110 Instead of lexically. This makes it more likely for similar generations 3111 to be close to each other in the list of unique chipsets. 3112 3113 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3114 3115commit 7c66be961de3477fc685eb71b46265adbae923f0 3116Author: Michel Dänzer <michel.daenzer@amd.com> 3117Date: Mon Oct 17 18:38:55 2016 +0900 3118 3119 Enable HW cursor support with PRIME slave output & Xorg > 1.18.99.901 3120 3121 Supported since Xorg 1.18.99.2, but buggy until 1.18.99.901. 3122 3123 (Ported from amdgpu commit 7cc04035c55788261cda89a915c433c2add6cad9) 3124 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3125 3126commit d016f20b14ffe4f90f7457375b2d9f5cf0f5650f 3127Author: Michel Dänzer <michel.daenzer@amd.com> 3128Date: Thu Oct 13 16:33:45 2016 +0900 3129 3130 Restore location of radeon_set_pixmap_bo call in drmmode_create_bo_pixmap 3131 3132 radeon_set_pixmap_bo sets the tiling flags, so it needs to be called 3133 before radeon_get_pixmap_tiling_flags. Fixes a regression with EXA and 3134 TearFree. 3135 3136 Fixes: 8523a733b6a5 ("Propagate failure from radeon_set_pixmap_bo") 3137 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98097 3138 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3139 3140commit 937398f6cd50db105d5e53ab553eb284d44121f4 3141Author: Michel Dänzer <michel.daenzer@amd.com> 3142Date: Thu Oct 6 18:52:37 2016 +0900 3143 3144 Rotate and reflect cursor hotspot position for drmModeSetCursor2 3145 3146 We were always passing the hotspot position in the X screen coordinate 3147 space, but drmModeSetCursor2 needs it in the CRTC coordinate space. The 3148 wrong hotspot position would cause the kernel driver to adjust the 3149 HW cursor position incorrectly when the hotspot position changed. 3150 3151 (Ported from amdgpu commit d42773eb45baff5933730e26878a0b45fcf07b65) 3152 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3153 3154commit 3fc839ff49f01c24eb94d5e1f0ab4065de47bc17 3155Author: Michel Dänzer <michel.daenzer@amd.com> 3156Date: Wed Sep 21 17:50:36 2016 +0900 3157 3158 Only list each unique chipset family once in the log file 3159 3160 (Ported from amdgpu commit 6a1ba044c2b71081e6060d0c096917d6238f2145) 3161 3162commit 9a5eb1915a240e3021ac38a6dacafef81e7bb743 3163Author: Michel Dänzer <michel.daenzer@amd.com> 3164Date: Wed Sep 21 17:48:38 2016 +0900 3165 3166 Clear damage in radeon_scanout_update if it doesn't intersect the CRTC 3167 3168 There's no need to test that same damage again. 3169 3170 (Ported from amdgpu commit a576430526cbc404de64b30e1377a356644e8024) 3171 3172commit 792a30dbe5f7d7815ea14032dbab55b5de8eabca 3173Author: Michel Dänzer <michel.daenzer@amd.com> 3174Date: Wed Sep 21 17:43:57 2016 +0900 3175 3176 Remove w/h parameters from radeon_scanout_extents_intersect 3177 3178 We can use the dimensions of the CRTC's mode instead. 3179 3180 (Ported from amdgpu commit ede7f2bcae63be65e05e3029bfe7c742e5978932) 3181 3182commit 80d794f11a5f047663897e4349da75d89ceff7c7 3183Author: Michel Dänzer <michel.daenzer@amd.com> 3184Date: Wed Sep 21 17:38:57 2016 +0900 3185 3186 Make the dedicated scanout mechanism work with arbitrary transforms 3187 3188 This makes TearFree work with arbitrary transforms, and makes transforms 3189 work better even without TearFree, with xserver >= 1.12. 3190 3191 (Ported from amdgpu commit bf000ea7ef91f5ecb59fc3c1ab8ed9eddcc0841d) 3192 3193commit 09ba7b3b7b1b724aa386dbf2d33f567d7f26326b 3194Author: Michel Dänzer <michel.daenzer@amd.com> 3195Date: Wed Sep 21 17:00:22 2016 +0900 3196 3197 Remove drmmode_load_palette 3198 3199 Not used by any supported version of xserver. 3200 3201 (Ported from amdgpu commits 1091f28e1fa239ee1a973d84a8376fa4a95d7247 3202 and 5a4d3267ac3823fe58b51b0b9075b82375d7180c) 3203 3204commit 730f443955cf2bcedd760c9fef5b11c81eacdd84 3205Author: Michel Dänzer <michel.daenzer@amd.com> 3206Date: Wed Sep 21 16:48:49 2016 +0900 3207 3208 present: Separate checks for flips vs unflips 3209 3210 All unflip checks apply to flips as well, but not vice versa. 3211 3212 (Ported from amdgpu commit 4d506c23c9a628204fa23607931557b07ada3e31) 3213 3214commit 343d28672411a510de21fa57986324ef8bcdd1b6 3215Author: Michel Dänzer <michel.daenzer@amd.com> 3216Date: Wed Sep 21 16:47:17 2016 +0900 3217 3218 dri2: Don't allow flipping when using a dedicated scanout buffer 3219 3220 Fixes issues when mixing rotation and page flipping with current xserver 3221 Git master. 3222 3223 (Ported from amdgpu commit decabd574f90d3df397c80ec931b3fde8a4afb49) 3224 3225commit 83e4781d15d66d6fa0e153eb4cd0e3d9e2ebed9f 3226Author: Michel Dänzer <michel.daenzer@amd.com> 3227Date: Wed Sep 21 16:44:50 2016 +0900 3228 3229 present: Don't allow flipping when using a dedicated scanout buffer 3230 3231 Fixes issues when mixing rotation and page flipping with current xserver 3232 Git master. 3233 3234 (Ported from amdgpu commit 3ed28ce7cd26f89969617ba901ff253091d0d469) 3235 3236commit 8419db3de6157875f9a840773350ecd29cd42d83 3237Author: Michel Dänzer <michel.daenzer@amd.com> 3238Date: Wed Sep 21 16:40:42 2016 +0900 3239 3240 Make sure drmmode_crtc->scanout[] are destroyed when not needed 3241 3242 We failed to do this when going back to scanning out directly from the 3243 screen pixmap. 3244 3245 As a bonus, since we now destroy drmmode_crtc->scanout[] after setting 3246 the new scanout buffer, we may avoid the CRTC turning off intermittently 3247 in this case. 3248 3249 (Ported from amdgpu commit 9c3324715fd395fd486ea341654d78f4f298b97f) 3250 3251commit 956e03d6a6b9478dd8e77c81f426c0d371c5d7a1 3252Author: Michel Dänzer <michel.daenzer@amd.com> 3253Date: Wed Sep 21 16:37:53 2016 +0900 3254 3255 Simplify drmmode_set_mode_major error handling 3256 3257 Initialize ret = FALSE and only set it to TRUE when we've succeeded. 3258 3259 (Ported from amdgpu commit 3bce0519a4008cf87c0e31a7a579e10f5dcdd2f3) 3260 3261commit ace455d1db4db3f9b8b434a3747acfa7e1273db4 3262Author: Hans de Goede <hdegoede@redhat.com> 3263Date: Wed Sep 21 16:31:51 2016 +0900 3264 3265 Only add main fb if necessary 3266 3267 If we're doing reverse-prime; or doing rotation the main fb is not used, 3268 and there is no reason to add it in this case. 3269 3270 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 3271 (Ported from xserver commit 4313122dea0df9affc280ee698e929489061ccc6) 3272 (Ported from amdgpu commit a3ca1500703837cbb8d49c554199a25dea7d5e1e) 3273 3274commit eea370dc4f96e680cc29b416fe62d6e2ea09bc91 3275Author: Hans de Goede <hdegoede@redhat.com> 3276Date: Wed Sep 21 16:27:11 2016 +0900 3277 3278 Remove unnecessary fb addition from drmmode_xf86crtc_resize 3279 3280 drmmode_set_mode_major() is the only user of drmmode->fb_id and will 3281 create it if necessary. 3282 3283 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 3284 (Ported from xserver commit 877453212166fdc912e0d687cdecee11aba563b5) 3285 (Ported from amdgpu commit 9ca1c24235ff5ab2e028333fc326e2eff008c574) 3286 3287commit cf52c76484a4adfa3e0b2271ea0c00f46d32ca48 3288Author: Michel Dänzer <michel.daenzer@amd.com> 3289Date: Tue Sep 13 18:21:37 2016 +0900 3290 3291 Free priv in amdgpu_set_pixmap_bo also if priv->bo == NULL 3292 3293 Fixes memory leak when destroying pixmaps with priv->bo == NULL. 3294 3295 Reported-by: Qiang Yu <qiang.yu@amd.com> 3296 (Ported from amdgpu commit 7f7f9825caf3983902491da27c16d14cd8bf9b7d) 3297 3298commit 50abcc19e29bb50451d031b2c9e99d691d4faf70 3299Author: Michel Dänzer <michel.daenzer@amd.com> 3300Date: Tue Sep 13 18:18:37 2016 +0900 3301 3302 glamor: Fix leak of pixmap private when replacing BO 3303 3304 Reported-by: Qiang Yu <qiang.yu@amd.com> 3305 (Ported from amdgpu commit 397aedafee437c125b8ac1feafb1c3b466842aeb) 3306 3307commit 8523a733b6a5de6116a6332fefc871d4c32652d8 3308Author: Michel Dänzer <michel.daenzer@amd.com> 3309Date: Mon Sep 12 18:54:33 2016 +0900 3310 3311 Propagate failure from radeon_set_pixmap_bo 3312 3313 (Ported from amdgpu commits c315c00e44afc91a7c8e2eab5af836d9643ebb88 3314 and 0d42082108c264568e2aadd15ace70e72388bc65) 3315 3316commit 53be26b00e83f871f0afd39caa5a7a1d6ec4aea1 3317Author: Michel Dänzer <michel.daenzer@amd.com> 3318Date: Wed Aug 31 16:46:56 2016 +0900 3319 3320 Add support for ScreenPtr::SyncSharedPixmap 3321 3322 This allows deferring shared pixmap updates between different drivers. 3323 3324 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3325 3326commit 156b5bc00da2eecb71cad517136f8cd74cf5d2c9 3327Author: Michel Dänzer <michel.daenzer@amd.com> 3328Date: Tue Sep 20 17:50:20 2016 +0900 3329 3330 Only define transform_region function for XF86_CRTC_VERSION >= 4 3331 3332 We're not using it with older xserver. 3333 3334 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3335 3336commit 80cc892ee1ce54fad3cb7dd11bd9df18c359136f 3337Author: Michel Dänzer <michel.daenzer@amd.com> 3338Date: Tue Sep 20 17:45:16 2016 +0900 3339 3340 Use local implementation of RegionDuplicate for older xserver 3341 3342 It was only added in xserver 1.15. Fixes build against older xserver. 3343 3344 Reported-by: Pali Rohár <pali.rohar@gmail.com> 3345 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3346 3347commit 12d30eeb9711bd2b1609d6bbb74c4a1760596f72 3348Author: Qiang Yu <Qiang.Yu@amd.com> 3349Date: Wed Sep 14 17:03:03 2016 +0900 3350 3351 DRI2: Fix radeon_dri2_exchange_buffers width/height copy'n'paste error 3352 3353 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 3354 (Ported from amdgpu commit 73c8dc000ad6b2b53ba3aa7155f5e8f6b55623b7) 3355 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3356 3357commit 11cc6843aa3b745aa1361f1a65e465e16696b914 3358Author: Michel Dänzer <michel.daenzer@amd.com> 3359Date: Wed Sep 14 17:01:13 2016 +0900 3360 3361 DRI2: Add interpolated_vblanks in radeon_dri2_get_crtc_msc 3362 3363 We need that in radeon_dri2_drawable_crtc as well for priv->vblank_delta 3364 to work as intended. 3365 3366 radeon_dri2_get_msc was already doing this. 3367 3368 Fixes hangs in some cases when using VDPAU via DRI2 and moving the 3369 window between CRTCs. 3370 3371 (Ported from amdgpu commit abd1a7901c95e4bc78415cf1b7923623b9177152) 3372 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3373 3374commit 380daff43cfefe2a8e4d496eaf65673f2815c878 3375Author: Qiang Yu <Qiang.Yu@amd.com> 3376Date: Wed Sep 14 16:59:38 2016 +0900 3377 3378 Fix radeon_mode_hotplug crash on multi GPU platform. 3379 3380 On multi GPU platform, some screen is created by other GPU DDX. 3381 3382 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 3383 (Ported from amdgpu commit 978242977e5dc905e1d5a46b1b0d34b356c7af26) 3384 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3385 3386commit 4a5fa37f74c233c6b9c6a08306688628a8e216e8 3387Author: Michel Dänzer <michel.daenzer@amd.com> 3388Date: Wed Sep 14 16:56:22 2016 +0900 3389 3390 Destroy all dedicated scanout buffers during CloseScreen 3391 3392 Fixes leaking active scanout buffers across a server reset, which also 3393 fixes server reset with glamor and active scanout buffers. 3394 3395 (Ported from amdgpu commit d96dabc71b1b32dc4b422a9633cdd4e0e95da052) 3396 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3397 3398commit cc2555835cd4c5fd2ae4f999a4bf7c18cdb1dda4 3399Author: Michel Dänzer <michel.daenzer@amd.com> 3400Date: Mon Sep 12 18:26:57 2016 +0900 3401 3402 glamor: Reallocate linear pixmap BO if necessary for DRI2 PRIME 3403 3404 Fixes corruption when using DRI2 PRIME render offloading with the master 3405 screen using this driver. 3406 3407 (Ported from amdgpu commit 0007c2f018ba663303d91d847e7c085269a23062) 3408 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3409 3410commit b3e5259e60157fdbdf46ee59b1b78995c2b15f72 3411Author: Michel Dänzer <michel.daenzer@amd.com> 3412Date: Mon Sep 12 18:12:06 2016 +0900 3413 3414 Move DRI2's local fixup_glamor helper to radeon_glamor_set_pixmap_bo 3415 3416 So it can be used outside of the DRI2 code. 3417 3418 (Ported from amdgpu commit 5518bf5d793439b5bab369e5fc18de9a4a3b9dd6) 3419 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3420 3421commit 38632bbd5ff80a9cf8ce584b2bc499d17d15befe 3422Author: Michel Dänzer <michel.daenzer@amd.com> 3423Date: Mon Sep 12 18:05:51 2016 +0900 3424 3425 Consolidate get_drawable_pixmap helper 3426 3427 There were two static helpers for the same purpose. Consolidate them 3428 into a single inline helper which can be used anywhere. 3429 3430 (Ported from amdgpu commit 641f4647b7f51dfd2da330376cd10fa9702b6423) 3431 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3432 3433commit e91858e435672c32f9c4a854b3dec048199d6f7f 3434Author: Michel Dänzer <michel.daenzer@amd.com> 3435Date: Mon Sep 12 18:02:07 2016 +0900 3436 3437 glamor: Fix radeon_glamor_share_pixmap_backing for priv->bo == NULL 3438 3439 Fixes crash when running a compositor and DRI_PRIME client via DRI2. 3440 3441 Reported-by: Qiang Yu <qiang.yu@amd.com> 3442 (Ported from amdgpu commit b36c77695ba77b59a0ccd868454e3af4fc04d5ff) 3443 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3444 3445commit 38797a33117222dadbc89e5f21ed8cd5deef9bea 3446Author: Michel Dänzer <michel.daenzer@amd.com> 3447Date: Fri Sep 2 11:08:28 2016 +0900 3448 3449 Make TearFree effective with PRIME slave scanout 3450 3451 TearFree can now prevent tearing with any possible display 3452 configuration. 3453 3454 Note that there may still be inter-GPU tearing if the primary GPU uses 3455 a different driver. 3456 3457 v2: 3458 * Also test dirty->slave_dst in radeon_prime_scanout_do_update 3459 3460 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> [v1] 3461 3462commit eda1f3df6aaed683036369fe8820da4dac3c2ae2 3463Author: Michel Dänzer <michel.daenzer@amd.com> 3464Date: Thu Sep 1 12:54:13 2016 +0900 3465 3466 Synchronize scanout pixmaps for TearFree 3467 3468 Copy the damaged areas which are still valid in the other scanout pixmap 3469 from there, then only copy the remaining damaged area from the screen 3470 pixmap. 3471 3472 This is slightly more efficient (only needs one Damage record instead of 3473 two, and only needs to copy each screen update across PCIe once with 3474 ShadowPrimary and a discrete GPU), and will be significantly more 3475 efficient for PRIME with the following change. 3476 3477 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3478 3479commit 2f6e5fb15f1a9ce523c85550e493f8bda9d0c00f 3480Author: Michel Dänzer <michel.daenzer@amd.com> 3481Date: Fri Sep 2 17:23:16 2016 +0900 3482 3483 Move up radeon_scanout_extents_intersect 3484 3485 Will be needed higher up by the following changes. No functional change. 3486 3487 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3488 3489commit 5a57005178fc13b6f7e513458ca6dae72a3e5783 3490Author: Michel Dänzer <michel.daenzer@amd.com> 3491Date: Thu Sep 1 17:19:27 2016 +0900 3492 3493 Factor out transform_region helper 3494 3495 While we're at it, fix leaking the memory allocated for xRectangles. 3496 3497 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3498 3499commit 99232f64db52812a843cd616d263d3a6b90eef3d 3500Author: Michel Dänzer <michel.daenzer@amd.com> 3501Date: Tue Dec 1 17:58:13 2015 +0900 3502 3503 Only copy from screen pixmap to shared pixmap on demand for slave scanout 3504 3505 Only copy once for each time we update the corresponding scanout pixmap. 3506 This can significantly reduce the bandwidth usage when there are 3507 frequent updates to the screen pixmap. 3508 3509 This initial implementation only works when both the master and slave 3510 screens use this driver. 3511 3512 v2: 3513 * Reduce churn in radeon_prime_scanout_update_handler 3514 * Clear the correct damage in radeon_dirty_update 3515 3516 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3517 3518commit a92c27484703abc7c410b6ae0e4b8d1efbbb8e6f 3519Author: Michel Dänzer <michel.daenzer@amd.com> 3520Date: Fri Aug 26 18:09:03 2016 +0900 3521 3522 Fix build against xserver < 1.13 3523 3524 pScreen->isGPU was only introduced in 1.13. 3525 3526 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97490 3527 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3528 3529commit 4bd2d01552f18153afa03a8947b22eebf3d67c6b 3530Author: Michel Dänzer <michel.daenzer@amd.com> 3531Date: Fri Aug 26 12:02:38 2016 +0900 3532 3533 Also handle disabled CRTCs in drmmode_clear_pending_flip 3534 3535 If disabling a CRTC had to be deferred due to a pending flip in 3536 drmmode_crtc_dpms, there may no longer be any outputs associated with 3537 the CRTC when we get here. So we have to check for !crtc->enabled and 3538 call drmmode_crtc_dpms in that case as well. 3539 3540 Fixes: 9090309e057d ("Wait for pending flips to complete before turning 3541 off an output or CRTC") 3542 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3543 3544commit a36fdaff40d5b4795a1400c348a80eee94892212 3545Author: Michel Dänzer <michel.daenzer@amd.com> 3546Date: Wed Aug 24 22:52:11 2016 +0900 3547 3548 Don't override crtc parameter value in drmmode_flip_handler/abort 3549 3550 When overriding the crtc parameter value of the last pending CRTC, 3551 drmmode_clear_pending_flip would work on the wrong CRTC, and the last 3552 pending CRTC's flip_pending flag might never get cleared. This would 3553 prevent that CRTC from properly turning off and back on again. 3554 3555 Fixes: 9090309e057d ("Wait for pending flips to complete before turning 3556 off an output or CRTC") 3557 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97392 3558 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3559 3560commit e520ce0ec0adf91ddce5c932d4b3f9477fd49304 3561Author: Michel Dänzer <michel.daenzer@amd.com> 3562Date: Thu Aug 25 18:22:34 2016 +0900 3563 3564 Also call drmmode_clear_pending_flip from radeon_scanout_flip_abort 3565 3566 Not doing so could break DPMS with TearFree. 3567 3568 Reported-and-Tested-by: furkan on IRC 3569 Fixes: 9090309e057d ("Wait for pending flips to complete before turning 3570 off an output or CRTC") 3571 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3572 3573commit b0867063abb197b9134166706d99fcbe5f204bb5 3574Author: Michel Dänzer <michel.daenzer@amd.com> 3575Date: Mon Nov 30 18:54:12 2015 +0900 3576 3577 Track damage accurately for RandR 1.4 slave scanout 3578 3579 This further reduces the PCIe bandwidth usage. 3580 3581 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3582 3583commit ad0a0656dd0e74683e6d7789decba827aa29c221 3584Author: Michel Dänzer <michel.daenzer@amd.com> 3585Date: Fri Nov 27 16:31:21 2015 +0900 3586 3587 Handle RandR 1.4 slave dirty updates via radeon_drm_queue 3588 3589 This reduces PCIe bandwidth usage and tearing. 3590 3591 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3592 3593commit 4cfa4615f79f64062e5e771cd45dd7048f48b4f6 3594Author: Michel Dänzer <michel.daenzer@amd.com> 3595Date: Fri Nov 27 16:53:30 2015 +0900 3596 3597 Use drmmode_crtc_scanout_* helpers for RandR 1.4 scanout pixmaps 3598 3599 This should allow using multiple CRTCs via RandR 1.4 even with xserver 3600 < 1.17. It also simplifies the code a little, and paves the way for 3601 following changes. 3602 3603 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3604 3605commit 121a6de72da5fcf9a32408eff36b2235f3dfbcfe 3606Author: Michel Dänzer <michel.daenzer@amd.com> 3607Date: Mon Jul 11 12:51:46 2016 +0900 3608 3609 Keep track of damage event related flushes per-client v2 3610 3611 This further reduces the compositing slowdown due to flushing overhead, 3612 by only flushing when the X server actually sends XDamageNotify events 3613 to a client, and there hasn't been a flush yet in the meantime. 3614 3615 v2: Use ScreenPrivateKey, fixes invalid memory access with GPU screens 3616 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3617 3618commit 9090309e057dc703d1a5bffd88e6cae14108cfc3 3619Author: Michel Dänzer <michel.daenzer@amd.com> 3620Date: Tue May 10 18:45:30 2016 +0900 3621 3622 Wait for pending flips to complete before turning off an output or CRTC 3623 3624 At least with older kernels, the flip may never complete otherwise, 3625 which can result in us hanging in drmmode_set_mode_major. 3626 3627 Fixes: https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-ati/+bug/1577170 3628 3629 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3630 3631commit 9a1afbf61fbb2827c86bd86d295fa0848980d60b 3632Author: Michel Dänzer <michel.daenzer@amd.com> 3633Date: Mon Jul 11 12:22:09 2016 +0900 3634 3635 Use EventCallback to avoid flushing every time in the FlushCallback 3636 3637 We only need to flush for XDamageNotify events. 3638 3639 Significantly reduces compositing slowdown due to flushing overhead, in 3640 particular with glamor. 3641 3642 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3643 3644commit 94fe42f29e0b00a26e810581d6c438ac6d8ecd8a 3645Author: Michel Dänzer <michel.daenzer@amd.com> 3646Date: Tue Jul 12 17:36:27 2016 +0900 3647 3648 Don't enable DRI3 by default with EXA 3649 3650 It doesn't work correctly in all cases, see e.g. 3651 https://bugs.freedesktop.org/show_bug.cgi?id=95475 . I'm not sure this 3652 is fixable, given EXA's architecture. 3653 3654 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3655 3656commit 6d91fb4fc701895473ff675f440a8eef655e80ca 3657Author: Michel Dänzer <michel.daenzer@amd.com> 3658Date: Tue Jul 12 15:57:28 2016 +0900 3659 3660 Don't enable micro-tiling for scanout buffers on pre-R600 3661 3662 The display engine didn't support it. 3663 3664 Fixes display corruption with options "TearFree" and "ShadowPrimary" 3665 (and rotation or transforms with current xserver) on pre-R600. 3666 3667 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3668 3669commit a37af701768b12d86868a831a79f1e02ee4968cf 3670Author: Michel Dänzer <michel.daenzer@amd.com> 3671Date: Wed Jul 6 12:46:01 2016 +0900 3672 3673 Add explicit RADEON_DRM_QUEUE_ERROR define 3674 3675 Should make the radeon_drm_queue_alloc error handling clearer, and gets 3676 rid of a compile warning about it returning NULL. 3677 3678 Reviewed-by: Alexandre Demers <alexandre.f.demers@gmail.com> 3679 3680commit 024afff27eb9f4c60043214099ddb42bbfb3d289 3681Author: Keith Packard <keithp@keithp.com> 3682Date: Mon Jul 18 21:24:23 2016 -0700 3683 3684 Adapt to video API 22 3685 3686 Deal with drm fd wakeup 3687 3688 Signed-off-by: Keith Packard <keithp@keithp.com> 3689 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 3690 3691commit eac4db4f0cb22a23d6e69161dcaacc7d53978aae 3692Author: Adam Jackson <ajax@redhat.com> 3693Date: Tue Jul 19 10:03:56 2016 -0400 3694 3695 Adapt Block/WakeupHandler signature for ABI 23 3696 3697 Signed-off-by: Adam Jackson <ajax@redhat.com> 3698 3699commit 3be841d0ae7d505cef325993205b12d15e98dba9 3700Author: Michel Dänzer <michel.daenzer@amd.com> 3701Date: Fri Jun 24 11:12:49 2016 +0900 3702 3703 Only use RandR APIs if RandR is enabled 3704 3705 Fixes crash with Xinerama enabled, which disables RandR. 3706 3707 Fixes: https://bugs.debian.org/827984 3708 3709 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3710 3711commit 7835558acdce318130ba4a09ef936fd675e3197d 3712Author: Michel Dänzer <michel.daenzer@amd.com> 3713Date: Mon Mar 28 18:49:15 2016 +0900 3714 3715 Adapt to XF86_CRTC_VERSION 7 3716 3717 Now the HW cursor can be used with TearFree rotation. 3718 3719 This also allows always using the separate scanout pixmap mechanism for 3720 rotation, so that should be much smoother even without TearFree enabled. 3721 3722 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3723 3724commit 0945db4d902056bda3d9ad4a4de2dfa685d10a70 3725Author: Tan Hu <tan.hu@zte.com.cn> 3726Date: Fri May 27 17:05:15 2016 +0800 3727 3728 EXA/6xx/7xx: accelerate PictOpOver with component alpha 3729 3730 Subpixel text rendering is typically done with a solid src and 3731 a pixmap mask. Traditionally, this cannot be accelerated in a single 3732 pass and requires two passes [1]. However, we can cheat a little 3733 with a constant blend color. 3734 3735 We can use: 3736 const.A = src.A / src.A 3737 const.R = src.R / src.A 3738 const.G = src.G / src.A 3739 const.B = src.B / src.A 3740 3741 dst.A = const.A * (src.A * mask.A) + (1 - (src.A * mask.A)) * dst.A 3742 dst.R = const.R * (src.A * mask.R) + (1 - (src.A * mask.R)) * dst.R 3743 dst.G = const.G * (src.A * mask.G) + (1 - (src.A * mask.G)) * dst.G 3744 dst.B = const.B * (src.A * mask.B) + (1 - (src.A * mask.B)) * dst.B 3745 3746 This only needs a single source value. src.A is cancelled down in 3747 the right places. 3748 3749 [1] http://anholt.livejournal.com/32058.html 3750 3751 r6xx still be used on some machine, 3752 Ported from commit 4375a6e75e5d41139be7031a0dee58c057ecbd07. 3753 3754 Signed-off-by: Tan Hu <tan.hu@zte.com.cn> 3755 Reviewed-by: Grigori Goronzy <greg@chown.ath.cx> 3756 3757commit 9b9ad669c748f53247e53fa3f3b03a77da5e5cb3 3758Author: Tan Hu <tan.hu@zte.com.cn> 3759Date: Fri May 27 17:05:14 2016 +0800 3760 3761 EXA/6xx/7xx: fast solid pixmap support 3762 3763 Solid pixmaps are currently implemented with scratch pixmaps, which 3764 is slow. This replaces the hack with a proper implementation. The 3765 Composite shader can now either sample a src/mask or use a constant 3766 value. 3767 3768 r6xx still be used on some machine, 3769 Ported from commit 94d0d14914a025525a0766669b556eaa6681def7. 3770 3771 Signed-off-by: Tan Hu <tan.hu@zte.com.cn> 3772 Reviewed-by: Grigori Goronzy <greg@chown.ath.cx> 3773 3774commit aa07b365d7b0610411e118f105e49daff5f5a5cf 3775Author: Michel Dänzer <michel.daenzer@amd.com> 3776Date: Tue May 31 18:16:26 2016 +0900 3777 3778 Add .editorconfig file 3779 3780 Basically a conversion from .dir-locals.el, but also correctly handles 3781 files which predominantly use tabs for indentation. 3782 3783 Also, EditorConfig supports many more editors and IDEs. 3784 3785 Acked-by: Alex Deucher <alexander.deucher@amd.com> 3786 3787commit 040a7b80e1fcbaa93ac17f7113d696d9b853cf8a 3788Author: Michel Dänzer <michel.daenzer@amd.com> 3789Date: Tue Apr 12 18:18:43 2016 +0900 3790 3791 Explicitly set the fbcon pixmap pitch again 3792 3793 The kernel driver returns 0 as the pitch of the fbcon BO via the 3794 DRM_RADEON_GEM_GET_TILING ioctl, so we ended up using an incorrect 3795 pitch in some cases. 3796 3797 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94901 3798 3799 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3800 3801commit 1181b9c582f10b6c523e4b2988e2ce87ecf3d367 3802Author: Michel Dänzer <michel.daenzer@amd.com> 3803Date: Tue May 10 19:02:20 2016 +0900 3804 3805 Enable DRI3 by default when building for Xorg >= 1.18.3 3806 3807 Seems to work well enough in general now. 3808 3809 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3810 3811commit c801f9f10a5d72d935faf21e72f7e7808fb4f05f 3812Author: Michel Dänzer <michel.daenzer@amd.com> 3813Date: Wed Apr 13 11:12:38 2016 +0900 3814 3815 Handle Zaphod mode correctly in radeon_mode_hotplug 3816 3817 We need to scan both screens of the entity for existing connectors, and 3818 enumerate DVI & HDMI connectors consistently regardless of which screen 3819 they're assigned to. 3820 3821 Fixes crash when hot-(un)plugging connectors in Zaphod mode. 3822 3823 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93415 3824 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3825 3826commit 1ca677309720e2f6c953c9e76f5b34c22a4416c6 3827Author: Michel Dänzer <michel.daenzer@amd.com> 3828Date: Tue Mar 15 16:47:35 2016 +0900 3829 3830 present: Support async flips 3831 3832 The xserver Present code only calls radeon_present_flip with 3833 sync_flip=FALSE if radeon_present_screen_init sets 3834 PresentCapabilityAsync, and the latter only sets it if the kernel driver 3835 advertises support for async flips. 3836 3837 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3838 3839commit 90a915c62d012e99193833aecc93974e68880c60 3840Author: Michel Dänzer <michel.daenzer@amd.com> 3841Date: Tue Mar 15 16:42:16 2016 +0900 3842 3843 Add support for async flips to radeon_do_pageflip 3844 3845 Will be used by the next change. No functional change here. 3846 3847 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3848 3849commit ba8b6288c8e6fc4be5d7144ecbe9a1f241881674 3850Author: Qiang Yu <Qiang.Yu@amd.com> 3851Date: Mon Apr 11 16:35:37 2016 +0900 3852 3853 Remove RR_Capability_SinkOutput for GPU without CRTC 3854 3855 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 3856 (Ported from amdgpu commit a0bbb373f902e0ffc14570c85faec7e44134f62e) 3857 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3858 3859commit bd4c72c8625996d842824ce4963f2d759fe2954a 3860Author: Tom St Denis <tom.stdenis@amd.com> 3861Date: Fri Apr 8 10:22:11 2016 -0400 3862 3863 dri3: Return NULL from radeon_dri3_pixmap_from_fd if calloc fails. 3864 3865 Signed-off-by: Tom St Denis <tom.stdenis@amd.com> 3866 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 3867 3868commit 3300ea01fbdf770dc084eebbf2854fba35144220 3869Author: Michel Dänzer <michel.daenzer@amd.com> 3870Date: Mon Apr 11 16:11:57 2016 +0900 3871 3872 Post 7.7.0 release version bump 3873 3874commit df6662864b56d1aeeea4c67d9b668e64197afb20 3875Author: Michel Dänzer <michel.daenzer@amd.com> 3876Date: Thu Apr 7 11:44:05 2016 +0900 3877 3878 Bump version for 7.7.0 release 3879 3880commit c12614cfdc5729041c204e3ead1dbfde38011f0c 3881Author: Michel Dänzer <michel.daenzer@amd.com> 3882Date: Mon Apr 4 14:59:41 2016 +0900 3883 3884 Update manpage entry for Option "TearFree" 3885 3886 It's now effective for rotation as well. 3887 3888 (Ported from amdgpu commit faf9d720b7d650f5f1ea657a874d08eac3972e60) 3889 3890 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3891 3892commit 4693b1bd5b5c381e8b7b68a6f7f0c6696d6a68df 3893Author: Michel Dänzer <michel.daenzer@amd.com> 3894Date: Thu Mar 31 17:02:55 2016 +0900 3895 3896 Identify DRM event queue entries by sequence number instead of by pointer 3897 3898 If the memory for an entry was allocated at the same address as that for 3899 a previously cancelled entry, the handler could theoretically be called 3900 prematurely, triggered by the DRM event which was submitted for the 3901 cancelled entry. 3902 3903 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3904 3905commit 83734317e6bdaeebb4462a63f541e73a1d7c2f77 3906Author: Michel Dänzer <michel.daenzer@amd.com> 3907Date: Wed Mar 30 11:44:09 2016 +0900 3908 3909 Update pixmap pitch in radeon_set_pixmap_bo 3910 3911 Stop second guessing it in drmmode_crtc_scanout_create. 3912 3913 Fixes display corruption in some cases with TearFree enabled. 3914 3915 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94751 3916 3917 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3918 3919commit 0b3aac1de9db42bfca545fa331e4985836682ec7 3920Author: Michel Dänzer <michel.daenzer@amd.com> 3921Date: Mon Mar 28 18:43:36 2016 +0900 3922 3923 DRI3: Refuse to open DRM file descriptor for ssh clients (v2) 3924 3925 Fixes hangs when attempting to use DRI3 on display connections forwarded 3926 via SSH. 3927 3928 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93261 3929 3930 v2: Don't do this for Xorg > 1.18.99.1 since the corresponding xserver 3931 change has landed in Git master. 3932 3933 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (v1) 3934 3935commit 8a6cd4bda05b9569b3dd0a5a75b2cc385b9ecba9 3936Author: Michel Dänzer <michel.daenzer@amd.com> 3937Date: Fri Mar 25 11:50:15 2016 +0900 3938 3939 Revert "Use render node for DRI3 if available" 3940 3941 This reverts commit cd94248ffa7d8fe0b57476f79e7e860dee66d1b0. 3942 3943 It broke VDPAU<->GL interop with DRI3 enabled, because the Gallium VDPAU 3944 code doesn't support DRI3 yet. We can consider re-enabling this once 3945 there is a Mesa release where the Gallium VDPAU code supports DRI3. 3946 3947 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94675 3948 3949 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3950 3951commit ed31cffba0d1bd4b14e5348a1456e4377277059c 3952Author: Michel Dänzer <michel.daenzer@amd.com> 3953Date: Thu Mar 24 18:08:22 2016 +0900 3954 3955 Call RADEONBlockHandler_KMS before setting initial modes 3956 3957 Doing it the other way around meant that there was still a possibility 3958 for the front buffer contents to be uninitialized when they start being 3959 scanned out. 3960 3961 (Ported from amdgpu commit 4a60b4b1851a3cbc2d8ad9048d68eeb6947cf132) 3962 3963 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3964 3965commit e592f32f8b5f5873fcc18b10a69dd5e4ccf11073 3966Author: Michel Dänzer <michel.daenzer@amd.com> 3967Date: Wed Mar 23 18:25:48 2016 +0900 3968 3969 Require xserver 1.9 or newer 3970 3971 1.9.0 was released in August 2010. 3972 3973 We were already unintentionally relying on things not available in 1.8 3974 for at least a year, and nobody has complained. 3975 3976 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3977 3978commit bde466e5d44cad64b4e4eceaa5de80fdbf86356e 3979Author: Michel Dänzer <michel.daenzer@amd.com> 3980Date: Wed Mar 23 18:16:42 2016 +0900 3981 3982 Fix build against older versions of xserver 3983 3984 Also slightly clean up the error handling in radeon_scanout_do_update. 3985 3986 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94614 3987 3988 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 3989 3990commit b2b079be0e7ef2704c3df9ee25fe46166e450e83 3991Author: Michel Dänzer <michel.daenzer@amd.com> 3992Date: Wed Mar 23 17:29:47 2016 +0900 3993 3994 Use radeon_alloc_pixmap_bo for allocating scanout BOs 3995 3996 This enables tiling for scanout BOs used for TearFree / ShadowPrimary, 3997 and simplifies the code a little. 3998 3999 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4000 4001commit d21ac4669a8b2cdd4eec5e5a94d1950b7423b8b5 4002Author: Michel Dänzer <michel.daenzer@amd.com> 4003Date: Wed Mar 23 17:27:27 2016 +0900 4004 4005 DRI3 only works with acceleration 4006 4007 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94214 4008 4009 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4010 4011commit eb1a2c57e6c45d7ca57b2e2be643194243b7aae1 4012Author: Michel Dänzer <michel.daenzer@amd.com> 4013Date: Wed Mar 23 17:23:31 2016 +0900 4014 4015 TearFree can only work with acceleration 4016 4017 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4018 4019commit d670c5c9851b4eff21c845d26c7d7e4eb5ee0fa9 4020Author: Michel Dänzer <michel.daenzer@amd.com> 4021Date: Fri Mar 18 16:46:08 2016 +0900 4022 4023 Check for xf86CursorResetCursor 4024 4025 If it's available, Xorg calls it on each mode configuration change. It 4026 does what xf86_reload_cursors does (and more), so we don't need to call 4027 the latter anymore. 4028 4029 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4030 4031commit 7f3d0780ca65a90117c2a61362dbc0899bd9c0b0 4032Author: Michel Dänzer <michel.daenzer@amd.com> 4033Date: Wed Mar 9 18:28:47 2016 +0900 4034 4035 Don't try DRI2/Present flipping while the HW cursor can't be used 4036 4037 Flipping doesn't interact correctly with SW cursor: A flip makes the SW 4038 cursor disappear. It will only appear again when the cursor is moved, 4039 but it will be surrounded by corruption, because the SW cursor code 4040 will restore stale screen contents at the old cursor location before 4041 drawing the cursor at the new location. 4042 4043 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4044 4045commit 3de480e83c0a1824838d662d6d67c9fe85277298 4046Author: Michel Dänzer <michel.daenzer@amd.com> 4047Date: Wed Mar 9 16:44:32 2016 +0900 4048 4049 Factor out HW cursor checking code into drmmode_can_use_hw_cursor 4050 4051 And add a check for RandR 1.4 multihead. 4052 4053 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4054 4055commit 2e0b458e747c7df81dd5f9579a762262af3350e1 4056Author: Ryan Kennedy <Ryan.Kennedy@amd.com> 4057Date: Sun Mar 20 17:57:56 2016 -0400 4058 4059 Use correct tiling in drmmode_create_bo_pixmap 4060 4061 The callers were using 0 for the tiling flags, causing surfaces to be 4062 incorrectly treated as linear. This issue was causing corruption on Cayman 4063 just after login with GNOME3. The kernel was rejecting the cs in 4064 drmmode_copy_fb due to invalid texture pitch. 4065 4066 Signed-off-by: Ryan Kennedy <Ryan.Kennedy@amd.com> 4067 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 4068 4069commit a03271de5ecdaa7790d1316e993c4450b91fe936 4070Author: Michel Dänzer <michel.daenzer@amd.com> 4071Date: Wed Mar 9 18:23:10 2016 +0900 4072 4073 present: Return rotated CRTCs from radeon_present_get_crtc 4074 4075 Sync-to-vblank works fine with rotation. We're still checking for 4076 rotation in radeon_present_check_flip. 4077 4078 Returning NULL from here resulted in the xserver present code falling 4079 back to the fake CRTC running at 1 fps. 4080 4081 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4082 4083commit 40191d82370eb7e58bd34c44966cbf44c3703229 4084Author: Michel Dänzer <michel.daenzer@amd.com> 4085Date: Wed Mar 9 18:15:57 2016 +0900 4086 4087 present: Clear drmmode->fb_id before calling set_mode_major for unflip 4088 4089 Without this, drmmode_set_mode_major may just re-set the FB for the 4090 last flipped BO, in which case the display will probably freeze. 4091 4092 Reproduction recipe: Enable rotation while a fullscreen client is 4093 flipping. 4094 4095 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4096 4097commit 798c4fd16d339b1ad5fd729cc884be084c60e38b 4098Author: Michel Dänzer <michel.daenzer@amd.com> 4099Date: Fri Dec 25 18:57:42 2015 +0900 4100 4101 Make Option "TearFree" effective for rotated/reflected outputs as well (v2) 4102 4103 Support varies by xserver version: 4104 4105 < 1.12: No support for the driver handling rotation/reflection 4106 1.12-1.15: Support for driver handling rotation/reflection, but there's 4107 a bug preventing the HW cursor from being visible everywhere 4108 it should be on rotated outputs, so we can only support 4109 TearFree for reflection. 4110 >= 1.16: While the bug above is still there (fixes pending review), 4111 the driver can force SW cursor for rotated outputs, so we 4112 can support TearFree for rotation as well. 4113 4114 v2: Don't set crtc->driverIsPerformingTransform after xf86CrtcRotate if 4115 it wasn't set before. Fixes breaking rotation with TearFree disabled. 4116 4117 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (v1) 4118 4119commit eb611a2e4ecce7a1ab85fd72b9b78e3269311dd5 4120Author: Michel Dänzer <michel.daenzer@amd.com> 4121Date: Tue Feb 2 18:04:11 2016 +0900 4122 4123 Consolidate pScreen usage in drmmode_set_mode_major 4124 4125 We were already relying on pScrn->pScreen being non-NULL in some cases, 4126 which is supposedly always true ever since this function is no longer 4127 getting called from ScreenInit. 4128 4129 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4130 4131commit 06602171386e538081c298645fb7ca1a70fe80cc 4132Author: Michel Dänzer <michel.daenzer@amd.com> 4133Date: Thu Feb 4 17:13:57 2016 +0900 4134 4135 Remove check for XF86_CRTC_VERSION 3 4136 4137 We require xserver >= 1.8, which was already at version 3. 4138 4139 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4140 4141commit a88985f5d1e39caca49ceb65678aaa9cb622a0d2 4142Author: Michel Dänzer <michel.daenzer@amd.com> 4143Date: Mon Feb 29 18:12:43 2016 +0900 4144 4145 Deal with modesets and page flips crossing on a CRTC 4146 4147 If we set a mode while a flip is pending, the kernel driver may program 4148 the flip to the hardware after the modeset. If that happens, the hardware 4149 will display the BO from the flip, whereas we will assume it displays the 4150 BO from the modeset. In other words, the display will most likely freeze, 4151 at least until another modeset. 4152 4153 Prevent this condition by waiting for a pending flip to finish before 4154 setting a mode. 4155 4156 Fixes display freezing when setting rotation or a transform with 4157 TearFree enabled. 4158 4159 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4160 4161commit f5d968cbba3c9b7ec202161f2157d8d64778c817 4162Author: Michel Dänzer <michel.daenzer@amd.com> 4163Date: Thu May 21 12:54:31 2015 +0900 4164 4165 Make DRM event queue xf86CrtcPtr based instead of ScrnInfoPtr based 4166 4167 This allows for a minor simplification of the code. 4168 4169 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4170 4171commit e87365117acbd80b7d80fbb5eb30890ef7153291 4172Author: Michel Dänzer <michel.daenzer@amd.com> 4173Date: Thu Feb 25 16:49:16 2016 +0900 4174 4175 DRI2: Also clear dri2_flipping when client disconnects before event 4176 4177 Fixes the following problem: 4178 4179 With DRI3 enabled, run glxgears with LIBGL_DRI3_DISABLE=1, make it 4180 fullscreen and press Escape while it's still fullscreen. This could 4181 result in dri2_flipping not getting cleared, spuriously preventing apps 4182 using DRI3 from flipping. 4183 4184 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4185 4186commit d5dbb07db22d5420c81dfebc060f0dd86e7b8a20 4187Author: Michel Dänzer <michel.daenzer@amd.com> 4188Date: Thu Feb 25 18:02:12 2016 +0900 4189 4190 Remove radeon_scanout_flip_handler 4191 4192 No longer necessary now that radeon_drm_queue_handler can handle 4193 e->handler == NULL. 4194 4195 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4196 4197commit 3989766edde85d1abe7024577b98fc9b007bc02a 4198Author: Michel Dänzer <michel.daenzer@amd.com> 4199Date: Thu Feb 25 16:43:48 2016 +0900 4200 4201 drm_queue: Don't abort events immediately from radeon_drm_abort_client 4202 4203 Keep them around until the DRM event arrives, but then call the abort 4204 functions instead of the handler functions. 4205 4206 This is a prerequisite for the following fix. 4207 4208 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4209 4210commit 7bc089ee9740a3ef91f16e36786d8ed64e165acd 4211Author: jimqu <Jim.Qu@amd.com> 4212Date: Tue Feb 16 17:05:48 2016 +0900 4213 4214 glamor: Return NullPixmap on failure to create shareable pixmap 4215 4216 If we were asked to create a shareable pixmap, it doesn't make sense 4217 to return a pixmap which isn't shareable. Doing so caused trouble down 4218 the line such as a crash with older versions of glamor when trying to 4219 use GLX pixmaps of bpp < 32 via DRI2. 4220 4221 Signed-off-by: JimQu <jim.qu@amd.com> 4222 (ported from amdgpu commit 2fcb7dadd3c71cd405cbbaafc777697538ca9c29) 4223 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4224 4225commit 4cc32031467157ab8788f5c684fb1ac67ae96ff9 4226Author: jimqu <Jim.Qu@amd.com> 4227Date: Tue Feb 16 17:03:24 2016 +0900 4228 4229 Move radeon_glamor_destroy_pixmap before radeon_glamor_create_pixmap 4230 4231 The next commit will call the former from the latter. No functional 4232 change. 4233 4234 Signed-off-by: JimQu <jim.qu@amd.com> 4235 (ported from amdgpu commit 5269a2228bff6023c1a7f3e8534027e1d7addc25) 4236 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4237 4238commit a9b181083d56b5de9b6822d0ed89fcc86de777a6 4239Author: Michel Dänzer <michel.daenzer@amd.com> 4240Date: Tue Feb 16 16:59:40 2016 +0900 4241 4242 Remove pci_dev test from radeon_get_scrninfo 4243 4244 The pci_dev parameter can never be NULL since we only support KMS. 4245 4246 Reported-by: Tom St Denis <tom.stdenis@amd.com> 4247 (ported from amdgpu commit 6e42c58375a4c3229da93c27bbd104af145c6163) 4248 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4249 4250commit 2c7f068b0a0493daae86e5eacdf8d9812fcd28dc 4251Author: Mykola Lysenko <Mykola.Lysenko@amd.com> 4252Date: Mon Feb 15 18:21:39 2016 +0900 4253 4254 Initialize drmmode_crtc dpms_mode to DPMSModeOff 4255 4256 This disables query of disabled pipes for drmWaitVBlank on X start 4257 4258 Signed-off-by: Mykola Lysenko <Mykola.Lysenko@amd.com> 4259 (ported from amdgpu commit 4eb9cedca080b30c57ded349a397620ee7d0cd46) 4260 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4261 4262commit 9358e44f03c95396b7f7a4f7398f5b24d2559bb2 4263Author: Michel Dänzer <michel.daenzer@amd.com> 4264Date: Mon Feb 15 18:20:31 2016 +0900 4265 4266 sync: Check if miSyncShmScreenInit symbol is resolved at runtime 4267 4268 It may be disabled in the Xorg build, either explicitly or because the 4269 xshmfence library isn't available. 4270 4271 (ported from amdgpu commit 1d0b0c1794e65e581a48aa9fb19679d928d82a5d) 4272 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4273 4274commit cd94248ffa7d8fe0b57476f79e7e860dee66d1b0 4275Author: Jammy Zhou <Jammy.Zhou@amd.com> 4276Date: Tue Feb 16 16:36:33 2016 +0900 4277 4278 Use render node for DRI3 if available 4279 4280 Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> 4281 (ported from amdgpu commit ea558e645786b08d75307716036045170e97b43e) 4282 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4283 4284commit d41fcccf2f11d0393d252e76bbbe6d233c4ac443 4285Author: Michel Dänzer <michel.daenzer@amd.com> 4286Date: Wed Jan 20 11:20:30 2016 +0900 4287 4288 Set RADEON_GEM_NO_CPU_ACCESS flag for BOs which don't need CPU access 4289 4290 Failing to do this was resulting in the kernel driver unnecessarily 4291 leaving open the possibility of CPU access to those BOs. 4292 4293 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4294 4295commit 9483a3d777919b224f70c3b4d01e4b320a57db31 4296Author: Michel Dänzer <michel.daenzer@amd.com> 4297Date: Fri Jan 15 16:48:46 2016 +0900 4298 4299 Fix RandR CRTC transforms 4300 4301 Currently, Xorg will only transform the cursor as of the first time the 4302 cursor image changes after a transform is set. 4303 4304 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80678 4305 4306 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4307 4308commit b16856b25086ffb27365ac2249b8da921066ce62 4309Author: Michel Dänzer <michel.daenzer@amd.com> 4310Date: Thu Jan 14 17:32:54 2016 +0900 4311 4312 Set the RandR primary output on startup if Xorg hasn't 4313 4314 Fixes xrandr (XRRGetOutputPrimary) not reporting any output as primary 4315 after startup. 4316 4317 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4318 4319commit c7cf00487cd6d4a5d0f39d5b92ff04f6420d6a32 4320Author: Michel Dänzer <michel.daenzer@amd.com> 4321Date: Fri Nov 27 12:52:28 2015 +0900 4322 4323 Build RandR 1.4 provider name from chipset name and bus ID 4324 4325 Instead of just "radeon", it's now e.g. "KAVERI @ pci:0000:00:01.0". 4326 4327 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4328 4329commit ff9a6b6f079a8419f4e6fadfee778060618bf735 4330Author: Michel Dänzer <michel.daenzer@amd.com> 4331Date: Fri Dec 11 17:21:27 2015 +0900 4332 4333 HAS_DIRTYTRACKING_ROTATION also supports multiple CRTCs 4334 4335 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4336 4337commit 78fbca095ae9887a2d3de48bb07975e2d1126e68 4338Author: Michel Dänzer <michel.daenzer@amd.com> 4339Date: Tue Dec 1 15:38:13 2015 +0900 4340 4341 Load fb module before glamoregl/shadow modules 4342 4343 Fixes unresolved symbols on some systems. 4344 4345 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93105 4346 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4347 4348commit b19417e2fddf4df725951aea5ad5e9558338f59e 4349Author: Michel Dänzer <michel.daenzer@amd.com> 4350Date: Fri Nov 20 18:48:56 2015 +0900 4351 4352 Don't advertise any PRIME offloading capabilities without acceleration 4353 4354 Acceleration is required even for display offloading. Trying to enable 4355 display offloading without acceleration resulted in a crash. 4356 4357 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4358 4359commit 10b7c3def58bb34acc38f076bc230e25b454ab79 4360Author: Michel Dänzer <michel.daenzer@amd.com> 4361Date: Tue Nov 10 16:31:09 2015 +0900 4362 4363 glamor: Deal with glamor_egl_destroy_textured_pixmap being removed 4364 4365 When it's not available, it's safe to call down to the glamor 4366 DestroyPixmap hook instead. 4367 4368 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4369 4370commit 535e5438b2c32f774b9c8c27ee0289b4749548ef 4371Author: Michel Dänzer <michel.daenzer@amd.com> 4372Date: Tue Nov 10 17:31:52 2015 +0900 4373 4374 glamor: Restore all ScreenRec hooks during CloseScreen 4375 4376 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4377 4378commit dfb52777e43f3855476cff812d64de687a199674 4379Author: Michel Dänzer <michel.daenzer@amd.com> 4380Date: Thu Nov 12 13:49:04 2015 +0900 4381 4382 Post 7.6.1 release version bump 4383 4384commit 99cb8c3faf1a4ce368b7500f17a2a7868c15e8e8 4385Author: Michel Dänzer <michel.daenzer@amd.com> 4386Date: Thu Nov 12 13:38:14 2015 +0900 4387 4388 Set version for 7.6.1 release 4389 4390commit 4a2114a717ec8d78cacd8fe158e132d62013faac 4391Author: Michel Dänzer <michel.daenzer@amd.com> 4392Date: Thu Nov 12 13:32:25 2015 +0900 4393 4394 Remove duplicate free(output_ids) call 4395 4396 Fixes double-free regression introduced by commit 875ad48e ("Simplify 4397 drmmode_set_mode_major() and avoid leaking memory."). 4398 4399commit 960a9f37096b53ec9603a99c7b1f3679f915e03f 4400Author: Michel Dänzer <michel.daenzer@amd.com> 4401Date: Thu Nov 12 13:20:18 2015 +0900 4402 4403 Post 7.6.0 release version bump 4404 4405commit 9aae4b71fb402841439e3267fcb7f410d8429c3a 4406Author: Michel Dänzer <michel.daenzer@amd.com> 4407Date: Thu Nov 12 11:56:48 2015 +0900 4408 4409 Bump version for 7.6.0 release 4410 4411commit 98291869ac4a542a0b478920586407ff9d2c8ef0 4412Author: Tom St Denis <tom.stdenis@amd.com> 4413Date: Wed Nov 11 16:01:41 2015 +0900 4414 4415 Clean up radeon_dri2_create_buffer2() 4416 4417 Remove the depth_pixmap variable from the function and clear 4418 out any dead/odd behaviour that results. 4419 4420 Signed-off-by: Tom St Denis <tom.stdenis@amd.com> 4421 (ported from amdgpu commit 6000aef4e2f0a121b94023484406fb6f04688f74) 4422 4423 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4424 4425commit c6fc7e309a8a922f94a1f5f3e8bfb9058cff7ad1 4426Author: Michel Dänzer <michel.daenzer@amd.com> 4427Date: Wed Nov 11 15:59:06 2015 +0900 4428 4429 Properly handle drmModeAddFB failure in drmmode_crtc_scanout_allocate 4430 4431 We were printing an error message, but not propagating the failure. That 4432 would probably lead to trouble down the road. 4433 4434 (ported from amdgpu commit 21e72fb2418b5cc7fc849a9cf951186e209036b0) 4435 4436 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4437 4438commit 4e4f4d53da0539ef9feb8766230a6e9927ae005b 4439Author: Michel Dänzer <michel.daenzer@amd.com> 4440Date: Wed Nov 11 15:57:21 2015 +0900 4441 4442 Eliminate redundant data parameter from drmmode_crtc_scanout_create 4443 4444 drmmode_crtc_scanout_create just needs to call 4445 drmmode_crtc_scanout_allocate when scanout->bo is NULL. 4446 4447 This makes it clearer to the reader / compiler that 4448 drmmode_crtc_scanout_create doesn't dereference scanout->bo when it's 4449 NULL. 4450 4451 (ported from amdgpu commit 8da1d0c870e1081d77925807d6e3bbc61a23f54f) 4452 4453 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4454 4455commit 421a7e797bdd58d83e81af7a6512cc715a3df514 4456Author: Michel Dänzer <michel.daenzer@amd.com> 4457Date: Wed Nov 11 15:51:58 2015 +0900 4458 4459 Don't advertise rotation support without hardware acceleration v2 4460 4461 Rotation currently doesn't work without acceleration (doesn't actually 4462 rotate with Option "NoAccel", crashes with Option "AccelMethod" "none" 4463 or when glamor fails to initialize) and would probably be too slow 4464 anyway. 4465 4466 v2: Also remove now dead code checking for ShadowFB from 4467 drmmode_crtc_scanout_allocate(). 4468 4469 (ported from amdgpu commit dc40582d5ff94d812cbc08f95cf14b80cd0f410d) 4470 4471 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4472 4473commit 875ad48e7b5cdb7beefbf18dddcbee3ed22b5446 4474Author: Tom St Denis <tom.stdenis@amd.com> 4475Date: Wed Nov 11 15:51:19 2015 +0900 4476 4477 Simplify drmmode_set_mode_major() and avoid leaking memory. 4478 4479 The function would leak the memory allocated for output_ids. This 4480 patch addresses that as well as simplifies the logic somewhat. 4481 4482 Signed-off-by: Tom St Denis <tom.stdenis@amd.com> 4483 (ported from amdgpu commit 460560502a1bdf26d06f3c30df46fa9f28ffb9e5) 4484 4485 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4486 4487commit 789d7d6a04cca6b36fb088a074027807ccb8dd61 4488Author: Tom St Denis <tom.stdenis@amd.com> 4489Date: Wed Nov 11 15:48:51 2015 +0900 4490 4491 Clean up allocation in RADEONInitVideo() 4492 4493 The allocation of the adapters should use the correct sizeof (even if 4494 allocating an array of pointers). 4495 4496 Signed-off-by: Tom St Denis <tom.stdenis@amd.com> 4497 (ported from amdgpu commit db3bb2061b9ac16b0922d9afae99874820356a04) 4498 4499 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4500 4501commit d88fa0dd5d37604de8efb05853738cfaca6a3166 4502Author: Tom St Denis <tom.stdenis@amd.com> 4503Date: Wed Nov 11 15:46:50 2015 +0900 4504 4505 Simplify pick best crtc to fold two loops into one 4506 4507 This patch folds the two for loops from radeon_pick_best_crtc() into 4508 one to reduce the LOC and make the routine easier to read. 4509 4510 Signed-off-by: Tom St Denis <tom.stdenis@amd.com> 4511 (ported from amdgpu commit 3055724aef76a624718f26d5f0f9e9d567ffbcfb) 4512 4513 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4514 4515commit dbbcd75b3c80aba77673904d46bca97779fd8a8d 4516Author: Tom St Denis <tom.stdenis@amd.com> 4517Date: Wed Nov 11 12:54:54 2015 +0900 4518 4519 dri2: Avoid calculation with undefined msc value 4520 4521 If the get_msc() call fails for any reason we should avoid updating the 4522 vblank counter delta with undefined data. 4523 4524 Signed-off-by: Tom St Denis <tom.stdenis@amd.com> 4525 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> (minor fixups) 4526 (ported from amdgpu commit 8823c3d4c6db70cff7699b31088f2d92db8faaf4) 4527 4528 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4529 4530commit 7186a8713ba004de4991f21c1a9fc4abc62aeff4 4531Author: Stephen Chandler Paul <cpaul@redhat.com> 4532Date: Fri Oct 23 09:59:36 2015 -0400 4533 4534 Handle failures in setting a CRTC to a DRM mode properly 4535 4536 This fixes a bug where running the card out of PPLL's when hotplugging 4537 another monitor would result in all of the displays going blank and 4538 failing to work properly until X was restarted or the user switched to 4539 another VT. 4540 4541 [Michel Dänzer: Pass errno instead of -ret to strerror()] 4542 4543 Signed-off-by: Stephen Chandler Paul <cpaul@redhat.com> 4544 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 4545 4546commit 548e97b3b7d1e94075a54ca2bb4eb683025098a7 4547Author: Michel Dänzer <michel.daenzer@amd.com> 4548Date: Fri Oct 16 16:26:58 2015 +0900 4549 4550 Call xf86CrtcRotate from initial drmmode_set_desired_modes call 4551 4552 Fixes various problems when rotation is specified in xorg.conf. 4553 4554 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92475 4555 4556 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4557 4558commit ce9914af8d8d5243977023ec7b09c605f9cca8b4 4559Author: Michel Dänzer <michel.daenzer@amd.com> 4560Date: Fri Oct 2 18:07:18 2015 +0900 4561 4562 Only align screen / scanout pixmap height where necessary 4563 4564 When using glamor acceleration, the pixmap's header has to have a height 4565 that matches exactly what the actual height is minus the GPU memory 4566 alignment. Otherwise CRTCs scanning out from the main scanout buffer 4567 (e.g. every CRTC that isn't rotated or transformed in some way) won't 4568 always work. This results in a bug where rotating one monitor in a 4569 multi-monitor setup won't always work properly. Easiest way to reproduce 4570 this: 4571 4572 - Have two monitors (I've gotten this working with a 1920x1080 and 4573 1280x1024, along with two 1920x1080s) 4574 - Rotate one of them from 0° to 90°, then rotate the same monitor from 4575 90° to 180°. The monitor that hasn't been rotated won't properly 4576 update, and will stay on a blank screen 4577 4578 This doesn't seem to make any difference when using EXA for 4579 acceleration. 4580 4581 Compared to Stephen Chandler's patch, this drops the height alignment 4582 in most places and only keeps it where it's really necessary. 4583 4584 Reported-and-Tested-by: Stephen Chandler Paul <cpaul@redhat.com> 4585 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4586 4587commit 0288a4b87b65ba54f37fbeeea3cb32238deee92e 4588Author: Michel Dänzer <michel.daenzer@amd.com> 4589Date: Wed Aug 19 16:54:14 2015 +0900 4590 4591 DRI2: Keep MSC monotonic when moving window between CRTCs 4592 4593 This mirrors the DRI3 implementation in xserver. Fixes VDPAU video 4594 playback hanging when moving the window between CRTCs. 4595 4596 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=66384 4597 4598 (Ported from amdgpu commit 63948ea091a9b324327ade7ec4fc5d67ca7e6f6f) 4599 4600 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4601 4602commit 1d886b526dc49f32dc6744b7a882894bdac4e846 4603Author: Michel Dänzer <michel.daenzer@amd.com> 4604Date: Wed Aug 19 18:11:45 2015 +0900 4605 4606 DRI2: Don't ignore rotated CRTCs in amdgpu_dri2_drawable_crtc 4607 4608 Waiting for vblank interrupts works fine with rotated CRTCs. The only 4609 case we can't handle with rotation is page flipping, which is handled 4610 in can_exchange(). 4611 4612 This fixes gnome-shell hanging on rotation, probably because 4613 amdgpu_dri2_get_msc returned MSC/UST 0 for rotated CRTCs. 4614 4615 (Ported from amdgpu commit 7b3212e33cd36fb6f122774df27b56ec4e1a22b8) 4616 4617 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4618 4619commit 270da55340766074cabff8af4258e29fe2f0fc81 4620Author: Alex Deucher <alexander.deucher@amd.com> 4621Date: Mon Aug 10 15:32:34 2015 -0400 4622 4623 add new OLAND pci id 4624 4625 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 4626 4627commit f123096877f0a0044f3d0315285441f2b5790c29 4628Author: Mario Kleiner <mario.kleiner.de@gmail.com> 4629Date: Mon Aug 10 23:34:39 2015 +0200 4630 4631 Make selection between DRI2 and DRI3 consistent with other drivers. (v2) 4632 4633 Add Option "DRI" to allow selection of maximum DRI level. 4634 4635 This allows the user to select the maximum level of DRI 4636 implementation to use, DRI2 or DRI3. It replaces the old 4637 option "DRI3" which had exactly the same purpose, but 4638 differs from the method used in both intel ddx and nouveau ddx. 4639 Make this consistent before a new stable driver is released. 4640 4641 v2: Retain handling of old Option "DRI3" for backwards 4642 compatibility, but Option "DRI" will take precedence 4643 over "DRI3" if both are provided. 4644 4645 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 4646 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 4647 4648commit 3791fceabf2cb037467dc41c15364e9f9ec1e47e 4649Author: Michel Dänzer <michel.daenzer@amd.com> 4650Date: Thu Aug 6 16:27:01 2015 +0900 4651 4652 Wait for scanout BO initialization to finish before setting mode 4653 4654 This should avoid intermittent artifacts which could sometimes be visible 4655 when setting a new scanout pixmap, e.g. on server startup or when 4656 changing resolutions. 4657 4658 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4659 4660commit 4e3dfa69e4630df2e0ec0f5b81d61159757c4664 4661Author: Michel Dänzer <michel.daenzer@amd.com> 4662Date: Thu Aug 6 16:16:38 2015 +0900 4663 4664 Only call drmmode_copy_fb (at most) once on server startup 4665 4666 It doesn't make sense to copy the screen contents from console when VT 4667 switching back to Xorg or when Xorg resets. 4668 4669 Fixes intermittent artifacts when VT switching back from console to the 4670 gdm login screen. 4671 4672 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4673 4674commit 09c7cdb923965f9a1ea11d2f449bc02114408938 4675Author: Michel Dänzer <michel.daenzer@amd.com> 4676Date: Thu Jul 30 15:58:54 2015 +0900 4677 4678 glamor: Move declaration of struct radeon_pixmap out of #if/#else blocks 4679 4680 Reviewed-by: Dave Airlie <airlied@redhat.com> 4681 4682commit 936582fde0db461c5c3d78ce4f5f4c93a88a489d 4683Author: Michel Dänzer <michel.daenzer@amd.com> 4684Date: Thu Jul 30 15:56:36 2015 +0900 4685 4686 Remove unused local variable pRADEONEnt 4687 4688 ../../src/radeon_probe.c: In function 'radeon_get_scrninfo': 4689 ../../src/radeon_probe.c:157:22: warning: variable 'pRADEONEnt' set but not used [-Wunused-but-set-variable] 4690 RADEONEntPtr pRADEONEnt; 4691 ^ 4692 4693 Reviewed-by: Dave Airlie <airlied@redhat.com> 4694 4695commit 5510cd6027d2387efdf33575e3bfc424cb11bfd8 4696Author: Dave Airlie <airlied@redhat.com> 4697Date: Mon Jul 27 09:22:57 2015 +1000 4698 4699 radeon: move radeon_pixmap forward declaration into other block 4700 4701 There is already a radeon_pixmap forward decl here, the #else 4702 block is missing one. 4703 4704 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4705 Signed-off-by: Dave Airlie <airlied@redhat.com> 4706 4707commit b32a0a3de84a44b9af4f1ca8be19f10d7fa31b12 4708Author: Dave Airlie <airlied@redhat.com> 4709Date: Wed Jul 22 13:35:07 2015 +1000 4710 4711 radeon: cleanup the entity rec 4712 4713 Some of these were set, some of them were 4714 always opposites, so clean things up. 4715 4716 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4717 Signed-off-by: Dave Airlie <airlied@redhat.com> 4718 4719commit fcb32231a38f9461d12720cbf72f63502197a711 4720Author: Emil Velikov <emil.l.velikov@gmail.com> 4721Date: Tue Jul 21 23:55:20 2015 +0100 4722 4723 Do not link radeon_drv.so against libpciaccess 4724 4725 It's only used/needed by ati_drv.so 4726 4727 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 4728 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 4729 4730commit b6d871bf299c7d0f106c07ee4d8bd3b2337f53cc 4731Author: Dave Airlie <airlied@gmail.com> 4732Date: Thu Jul 9 19:25:36 2015 +1000 4733 4734 radeon: adopt for new X server dirty tracking APIs. 4735 4736 Signed-off-by: Dave Airlie <airlied@redhat.com> 4737 4738commit 95f5d09e3667ded027ae648c97eb4737d8bf67c5 4739Author: Michel Dänzer <michel.daenzer@amd.com> 4740Date: Thu May 28 11:24:42 2015 +0900 4741 4742 present: Handle DPMS off in radeon_present_get_ust_msc 4743 4744 The DRM_IOCTL_WAIT_VBLANK ioctl may return an error during DPMS off, 4745 which would trigger an error message in drmmode_crtc_get_ust_msc. 4746 4747 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4748 4749commit 211862b777d0be251a4662f5dd24f2d400544c09 4750Author: Michel Dänzer <michel.daenzer@amd.com> 4751Date: Fri May 1 18:20:01 2015 +0900 4752 4753 present: Look at all CRTCs to determine if we can flip 4754 4755 Inspired by modesetting driver change by Kenneth Graunke. 4756 4757 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4758 4759commit 802d33e474a82262d9cdf11b03568b0c4929cd0d 4760Author: Michel Dänzer <michel.daenzer@amd.com> 4761Date: Fri May 1 18:54:16 2015 +0900 4762 4763 present: Fall back to modeset for unflip operation 4764 4765 It's not always possible to use the page flip ioctl for this, e.g. 4766 during DPMS off. We were previously just skipping the unflip in that 4767 case, which could result in hangs when setting DPMS off while a 4768 fullscreen Present app is running, e.g. at the GNOME3 lock screen. 4769 4770 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4771 4772commit a8ed62010d5012dfb27773595c446b217f3c00c5 4773Author: Piotr Redlewski <predlewski@gmail.com> 4774Date: Mon Jul 6 19:57:07 2015 +0200 4775 4776 Do not try to enable already enabled CRTCs in DPMS hook 4777 4778 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91227 4779 4780 Signed-off-by: Piotr Redlewski <predlewski@gmail.com> 4781 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 4782 4783commit fc07c370dac4649511eb7322d05a61605a539f39 4784Author: Mario Kleiner <mario.kleiner.de@gmail.com> 4785Date: Sun Jun 28 02:45:58 2015 +0200 4786 4787 Fix inconsistent default eg_tile_split in evergreen_accel.c 4788 4789 eg_tile_split() maps the default: switch-case to tile split 4790 flag 6, like for a tile split size of 4096 Bytes. All other 4791 instances of tile split mappings or reverse mappings in both 4792 the ddx and in Mesa's radeon gallium drivers assign the default: 4793 case to tile split flag 4, consistent with a default of 1024 4794 Bytes. 4795 4796 Make this one outlier consistent with other instances of 4797 mappings in the code to avoid surprises. 4798 4799 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 4800 4801commit ddaba449e8d6fe9fc0d97085e4045843fd8d7af9 4802Author: Piotr Redlewski <predlewski@gmail.com> 4803Date: Wed Jun 24 18:44:39 2015 +0200 4804 4805 Disable CRTCs when disabling the outputs 4806 4807 When turning the outputs off (DPMSModeOff), CRTCs stayed enabled. This led 4808 to higher gpu temperatures than with fglrx driver. 4809 4810 v2: when entering DPMS also disable active CRTCs 4811 v3: use drmmode_set_mode_major() for enabling CRTCs when leaving DPMS 4812 4813 Signed-off-by: Piotr Redlewski <predlewski@gmail.com> 4814 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 4815 4816commit acc11877423ecd81a6e0a7f38466f80e43efee20 4817Author: Michel Dänzer <michel.daenzer@amd.com> 4818Date: Tue Jun 16 17:34:57 2015 +0900 4819 4820 Skip disabled CRTCs in radeon_scanout_(do_)update 4821 4822 The vblank / page flip ioctls don't work as expected for a disabled CRTC. 4823 4824 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4825 4826commit fc9fadaebbc8aead6e030d93a9ccd84561f8f59e 4827Author: Mario Kleiner <mario.kleiner.de@gmail.com> 4828Date: Mon Jun 22 00:09:17 2015 +0200 4829 4830 Don't set TILE_SPLIT flags if surface.tile_split == 0. 4831 4832 On pre-Evergreen hw, libdrm's r6_surface_best() helper 4833 for the surface managers radeon_surface_best() routine 4834 is a no-op and therefore doesn't assign any tile_split 4835 settings to created surfaces, so it leaves 4836 surface.tile_split on its "undefined" value of 0. 4837 4838 Mesa's DRI3/Present backend creates DRI3 Pixmaps via 4839 the DRIImage extension and the radeon gallium driver 4840 implementation of that extension uses the libdrm 4841 surface manager for backing bo creation and treats 4842 an undefined surface.tile_split==0, as returned by 4843 the surface manager for pre-evergreen, as a signal 4844 to not assign any tile_split flags to the DRI3 Pixmaps 4845 bo. 4846 4847 The ddx also uses libdrm surface manager to create the 4848 x-screen pixmap, but so far treated the returned undefined 4849 surface.tile_split==0 by mapping it to eg_tile_split()'s 4850 default tile_split flags, which are different from Mesa's 4851 tiling flags for DRI3 pixmaps. Under DRI3/Present this 4852 causes a mismatch of src pixmap and destination root 4853 pixmaps tiling flags and thereby prevents page flipping 4854 for pixmap presents. 4855 4856 Change the ddx code to treat surface.tile_split==0 the 4857 same way as the radeon gallium driver to avoid mismatched 4858 tiling flags and thereby allow DRI3/Present page-flip to 4859 work on pre-Evergreen hw. 4860 4861 Tested on RV730 and Evergreen "Juniper". 4862 4863 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 4864 4865commit 49f5b0bc301414df049e00d226034e3d6e56421b 4866Author: Michel Dänzer <michel.daenzer@amd.com> 4867Date: Fri May 1 19:02:36 2015 +0900 4868 4869 Don't attempt a DRI2/Present page flip while the other one is flipping 4870 4871 Fixes corrupted display and hangs when switching between DRI2 and DRI3 4872 fullscreen apps, e.g. a compositor using DRI3 and a fullscreen app using 4873 DRI2 or vice versa. 4874 4875 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4876 4877commit afab7839fc15722dbaa7203d00fe7f6ce5336b9d 4878Author: Mario Kleiner <mario.kleiner.de@gmail.com> 4879Date: Wed Jun 10 20:55:08 2015 +0200 4880 4881 Allow/Fix use of multiple ZaphodHead outputs per x-screen. (v2) 4882 4883 Defining multiple ZaphodHead outputs per x-screen in a 4884 multiple x-screen's per gpu configuration caused all 4885 outputs except one per x-screen to go dark, because 4886 there was a fixed mapping x-screen number -> crtc number, 4887 limiting the number of crtc's per x-screen to one. 4888 4889 On a ZaphodHead's setup, be more clever and assign 4890 as many crtc's to a given x-screen as there are 4891 ZaphodHeads defined for that screen, assuming 4892 there are enough unused crtc's available. 4893 4894 Tested on a triple display setup with different combos 4895 of one, two or three ZaphodHeads per one, two or three 4896 x-screens. 4897 4898 This is a port of similar code from xf86-video-nouveau. 4899 4900 v2: Implement suggestions by Michel Dänzer: Less verbose 4901 debug output, more clear warning message on crtc allocation 4902 failure. Move clearing of per gpu assigned_crtc mask to 4903 CloseScreen, indeed testing shows no need for the more 4904 complex new server generation check from v1. 4905 4906 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 4907 (v1) Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4908 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 4909 4910commit d7c82731a8bf3d381bc571b94d80d9bb2dd6e40d 4911Author: Michel Dänzer <michel.daenzer@amd.com> 4912Date: Thu May 21 16:01:16 2015 +0900 4913 4914 DRI2: Fix handling of drmmode_crtc_get_ust_msc return code 4915 4916 Fixes regression introduced in commit 4917 76c2923ac5c7230a8b2f9f8329c308d28b44d9c0 ("DRI2: Split out helper for 4918 getting UST and MSC of a specific CRTC"). 4919 4920 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4921 4922commit e58fc380ccf2a581d28f041fd74b963626ca5404 4923Author: Michel Dänzer <michel.daenzer@amd.com> 4924Date: Thu May 28 10:59:22 2015 +0900 4925 4926 glamor: Only wait for GPU writes in radeon_glamor_prepare_access_cpu_ro 4927 4928 We don't need to wait for GPU reads to finish before reading with the 4929 CPU. 4930 4931 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4932 4933commit 5f82a720374c9c1caebb42bfbeea1f0cf8847d28 4934Author: Michel Dänzer <michel.daenzer@amd.com> 4935Date: Thu May 21 15:46:32 2015 +0900 4936 4937 present: Remove get_drmmode_crtc helper 4938 4939 It was getting the drmmode_crtc_private_ptr in a roundabout way. 4940 4941 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4942 4943commit d64a13ebe0ecd241ee3260dbffd8f4a01e254183 4944Author: Michel Dänzer <michel.daenzer@amd.com> 4945Date: Wed May 20 09:58:38 2015 +0900 4946 4947 Replace SyncFence typedef with declaration of struct _SyncFence 4948 4949 Fixes build failure in some environments: 4950 4951 CC radeon_sync.lo 4952 In file included from radeon_sync.c:28: 4953 /home/tinderbox/xorg-build/include/xorg/misync.h:31: error: redefinition of typedef 'SyncFence' 4954 radeon.h:93: note: previous declaration of 'SyncFence' was here 4955 make[2]: *** [radeon_sync.lo] Error 1 4956 4957 Reported-Tested-and-Acked-by: David Airlie <airlied@redhat.com> 4958 4959commit 818c180c8932233b214a35ba0647af82f7bcec3d 4960Author: Michel Dänzer <michel.daenzer@amd.com> 4961Date: Tue May 12 17:43:57 2015 +0900 4962 4963 glamor: Deal with glamor_glyphs_init being removed from xserver 4964 4965 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 4966 4967commit 7c4b78ab10b82c6dba9f72034ff7583859cca63d 4968Author: Alex Deucher <alexander.deucher@amd.com> 4969Date: Tue May 12 13:21:24 2015 -0400 4970 4971 add new bonaire pci id 4972 4973 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 4974 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 4975 4976commit 80eede245d1eda27eaba108b0761a24bfd69aff6 4977Author: Mario Kleiner <mario.kleiner.de@gmail.com> 4978Date: Sat May 9 07:24:36 2015 +0200 4979 4980 present: Fix present notify timestamps and counts. 4981 4982 Invalid (msc,ust)=(0,0) were returned to the server 4983 because a wrong crtc_id was used to specify which 4984 crtc should deliver pageflip completion data. Fix 4985 it in accordance with the dri2 implementation. 4986 4987 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 4988 4989commit e3be8b0a8cf484ff16597413a6172788178e80c8 4990Author: Mario Kleiner <mario.kleiner.de@gmail.com> 4991Date: Sat May 9 07:24:35 2015 +0200 4992 4993 present: Move check for async flips 4994 4995 Check for and reject currently unsupported async_flip 4996 inside radeon_present_check_flip() instead of inside 4997 radeon_present_flip(). 4998 4999 This way the server can detect early that async flips 5000 aren't supported by the ddx and can deal with this 5001 correctly by using its non-vsync'ed CopyRegion fallback. 5002 5003 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 5004 5005commit 1584dc545c78e0bce8d4b4b9f26b568e2c211453 5006Author: Michel Dänzer <michel.daenzer@amd.com> 5007Date: Tue Apr 28 17:14:11 2015 +0900 5008 5009 Skip disabled CRTCs in drmmode_set_desired_modes() even if set_hw == FALSE 5010 5011 Not skipping a disabled CRTC results in a crash. 5012 5013 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90187 5014 5015 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5016 5017commit 340baf3804b7d9de8082badfc715a9bee7856e8f 5018Author: Michel Dänzer <michel.daenzer@amd.com> 5019Date: Mon Apr 27 15:07:06 2015 +0900 5020 5021 Don't call radeon_get_pixmap_private() when using EXA 5022 5023 It only works when using glamor. 5024 5025 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90169 5026 5027 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5028 5029commit 80f3d727f93cb6efedd2b39338d2301035965fe2 5030Author: Michel Dänzer <michel.daenzer@amd.com> 5031Date: Wed Apr 22 13:33:15 2015 +0900 5032 5033 On screen resize, clear the new buffer before displaying it 5034 5035 Fixes garbage being intermittently visible during a screen resize. 5036 5037 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=27757#c7 5038 5039 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5040 5041commit 3999bf88cdb192fe2f30b03bd2ed6f6a3f9f9057 5042Author: Michel Dänzer <michel.daenzer@amd.com> 5043Date: Thu Apr 2 18:29:38 2015 +0900 5044 5045 Make drmmode_copy_fb() work with glamor as well 5046 5047 Needed for Xorg -background none. 5048 5049 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5050 5051commit a4a8cdbcc10c1c5f07485a2af9e9e81e490c3e1d 5052Author: Michel Dänzer <michel.daenzer@amd.com> 5053Date: Thu Apr 2 17:46:34 2015 +0900 5054 5055 Update scanout pixmap contents before setting a mode with it 5056 5057 This ensures the scanout pixmaps used for Option "TearFree" and Option 5058 "ShadowPrimary" have been initialized when their initial mode is set. 5059 5060 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5061 5062commit 673e1c7637687c74fc9bdeeeffb7ace0d04b734f 5063Author: Michel Dänzer <michel.daenzer@amd.com> 5064Date: Thu Apr 2 17:54:33 2015 +0900 5065 5066 Defer initial modeset until the first BlockHandler invocation 5067 5068 This ensures that the screen pixmap contents have been initialized when 5069 the initial modes are set. 5070 5071 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=27757 5072 5073 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5074 5075commit 37874a4eeace5df04b02c8fc28f67b824e3f0f5f 5076Author: Michel Dänzer <michel.daenzer@amd.com> 5077Date: Tue Apr 21 17:19:15 2015 +0900 5078 5079 Defer initial drmmode_copy_fb call until root window creation 5080 5081 That's late enough for acceleration to be fully initialized, but still 5082 early enough to set pScreen->canDoBGNoneRoot. 5083 5084 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5085 5086commit 39c497f3efca5ca08343b884f44c93215dcdef31 5087Author: Michel Dänzer <michel.daenzer@amd.com> 5088Date: Thu Apr 2 18:10:42 2015 +0900 5089 5090 Only copy fbcon BO contents if bgNoneRoot is TRUE 5091 5092 Otherwise, the X server will initialize the screen pixmap contents 5093 anyway. 5094 5095 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5096 5097commit 1af044d7eee211fd4b248c236280274a68334da5 5098Author: Michel Dänzer <michel.daenzer@amd.com> 5099Date: Wed Apr 22 18:43:02 2015 +0900 5100 5101 DRI2: Clear old->devPrivate.ptr in fixup_glamor 5102 5103 It doesn't point to the memory of the newly allocated BO. Fixes crash 5104 running piglit with Option "ShadowPrimary" enabled. 5105 5106 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5107 5108commit 43159ef400c3b18b9f4d3e6fa1c4aef2d60d38fe 5109Author: Michel Dänzer <michel.daenzer@amd.com> 5110Date: Mon Apr 20 18:44:36 2015 +0900 5111 5112 Add Option "TearFree" v4 5113 5114 Avoids tearing by flipping between two scanout BOs per (non-rotated) CRTC 5115 5116 v2: 5117 * Fix condition for TearFree log message (Richard Wilbur) 5118 * Log warning message about DRI page flipping being enabled because of 5119 TearFree (or ShadowPrimary) also when building without glamor support 5120 5121 v3: 5122 * Only override fb_id/x/y if all scanout pixmaps have been successfully 5123 allocated 5124 5125 v4: 5126 * Make log warning clearer if drmModePageFlip returns an error 5127 5128 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (v1) 5129 5130commit ed401f5b4f07375db17ff05e294907ec95fc946d 5131Author: Michel Dänzer <michel.daenzer@amd.com> 5132Date: Fri Apr 3 12:11:00 2015 +0900 5133 5134 glamor: Remove the stride member of struct radeon_pixmap 5135 5136 Its value was always the same as that of the PixmapRec devKind member. 5137 5138 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5139 5140commit ae92d1765fa370a8d94c2856ad6c45d273ec3c69 5141Author: Michel Dänzer <michel.daenzer@amd.com> 5142Date: Wed Mar 18 16:23:24 2015 +0900 5143 5144 glamor: Add Option "ShadowPrimary" v2 5145 5146 When this option is enabled, most pixmaps (including the screen pixmap) 5147 are allocated in system RAM and mostly accessed by the CPU. Changed areas 5148 of the screen pixmap are copied to dedicated per-CRTC scanout pixmaps 5149 regularly, triggered by the vblank interrupt. 5150 5151 v2: 5152 * Set region data pointer to NULL for keeping only the extents 5153 * Move pRegion and pBox local variable declarations closer to their uses 5154 in drmmode_set_mode_major() 5155 5156commit eea79472a84672ee4dc7adc4487cec6a4037048a 5157Author: Michel Dänzer <michel.daenzer@amd.com> 5158Date: Wed Apr 1 15:51:52 2015 +0900 5159 5160 glamor: Add wrappers for the X server rendering hooks 5161 5162 They can choose between using the GPU or CPU for the operation. 5163 5164commit 051d46382656ffc3e6cac1aab3aee7efdf5b623a 5165Author: Michel Dänzer <michel.daenzer@amd.com> 5166Date: Thu Mar 19 17:34:27 2015 +0900 5167 5168 glamor: Add radeon_pixmap parameter to radeon_glamor_create_textured_pixmap 5169 5170commit 2fa021f77372ca93375a3d13a0c43a9089674899 5171Author: Michel Dänzer <michel.daenzer@amd.com> 5172Date: Thu Mar 19 17:38:47 2015 +0900 5173 5174 glamor: Remove unused function radeon_glamor_pixmap_is_offscreen 5175 5176commit e96349ba6281fd18b8bf9c76629128276b065e6c 5177Author: Michel Dänzer <michel.daenzer@amd.com> 5178Date: Wed Apr 1 16:04:13 2015 +0900 5179 5180 Add RADEON_CREATE_PIXMAP_SCANOUT flag 5181 5182 It means that the pixmap is used for scanout exclusively. 5183 5184commit 9be7dd382e86d2b804de81d4e2af7431b2e16843 5185Author: Michel Dänzer <michel.daenzer@amd.com> 5186Date: Wed Apr 1 12:40:16 2015 +0900 5187 5188 Split out struct drmmode_scanout for rotation shadow buffer information 5189 5190 Will be used for other kinds of dedicated scanout buffers as well. 5191 5192commit c32b0530302739f6512755bccf281c2300617376 5193Author: Michel Dänzer <michel.daenzer@amd.com> 5194Date: Thu Mar 19 17:46:48 2015 +0900 5195 5196 Rename scanout_pixmap_x field to prime_pixmap_x 5197 5198 To avoid confusion with upcoming changes. 5199 5200commit 5921ba4ca705a0d919515626088f3948cc4848c1 5201Author: Michel Dänzer <michel.daenzer@amd.com> 5202Date: Tue Mar 31 15:14:52 2015 +0900 5203 5204 present: Don't flip between BOs with different tiling parameters 5205 5206 The kernel driver doesn't handle that correctly yet. 5207 5208 Fixes or at least avoids issues with OpenGL fullscreen apps with DRI3 5209 enabled and using PRIME or with (2D) tiling disabled. 5210 5211 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89720 5212 5213 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5214 5215commit 428e416e7cb04a1e0527da39cfebf70218879a77 5216Author: Michel Dänzer <michel.daenzer@amd.com> 5217Date: Thu Apr 2 10:34:03 2015 +0900 5218 5219 Add radeon_get_pixmap_tiling_flags helper 5220 5221 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5222 5223commit 98fb4199e63fedd4607cddee64bf602d6398df81 5224Author: Michel Dänzer <michel.daenzer@amd.com> 5225Date: Tue Mar 31 12:25:18 2015 +0900 5226 5227 Only enable SYNC extension fences and the Present extension along with DRI3 5228 5229 This avoids some trouble with the Gallium nine state tracker, which uses 5230 the Present extension even when DRI3 is disabled. 5231 5232 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89772 5233 5234 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5235 5236commit f8b0f23e9f4af9f9097ee5e72d53b45173163c41 5237Author: Michel Dänzer <michel.daenzer@amd.com> 5238Date: Fri Mar 27 12:34:55 2015 +0900 5239 5240 DRI2: Use radeon_get_pixmap_handle 5241 5242 Now we can share pixmaps with no struct radeon_bo via DRI2. 5243 5244 Fixes VDPAU video playback freezing when using an OpenGL compositor with 5245 DRI3 enabled. 5246 5247 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89755 5248 5249 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5250 Tested-by: Nick Sarnie <commendsarnex@gmail.com> 5251 5252commit ccbda955ebae1d457d35293833f12791e0f9fb0b 5253Author: Michel Dänzer <michel.daenzer@amd.com> 5254Date: Fri Mar 27 12:16:44 2015 +0900 5255 5256 Move get_pixmap_handle helper to radeon_bo_helper.c 5257 5258 No functional change. 5259 5260 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5261 Tested-by: Nick Sarnie <commendsarnex@gmail.com> 5262 5263commit de5ddd09db82141b263338dcf0c28e01f58268ee 5264Author: Michel Dänzer <michel.daenzer@amd.com> 5265Date: Thu Mar 26 16:33:02 2015 +0900 5266 5267 Move radeon_drm_handler/abort_proc fields to drmmode_flipdata_rec 5268 5269 Their values are the same for all DRM flip ioctl calls within a single 5270 radeon_do_pageflip() call. 5271 5272 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5273 5274commit e8c0f6319fbf4c3ea11e22ab1a68837031bdec8c 5275Author: Michel Dänzer <michel.daenzer@amd.com> 5276Date: Thu Mar 26 16:27:35 2015 +0900 5277 5278 Simplify radeon_do_pageflip() error handling slightly more 5279 5280 We don't need the local variable old_fb_id. 5281 5282 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5283 5284commit 8fc22360d5520469c82092ccb0fcf2af330c573f 5285Author: Michel Dänzer <michel.daenzer@amd.com> 5286Date: Thu Mar 26 15:58:01 2015 +0900 5287 5288 Increase robustness against DRM page flip ioctl failures v3 5289 5290 Centralize cleanup, only clean up things that have been allocated for 5291 the failed ioctl call. 5292 5293 Fixes double-free after a flip ioctl failure. 5294 5295 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89681 5296 5297 v2: Only call drmModeRmFB for flipdata->old_fb_id on receipt of last DRM 5298 page flip event. Fixes Black screen on making glxgears fullscreen with 5299 DRI3 enabled. 5300 v3: Avoid double-free of flipdata in the unlikely case that calloc fails 5301 for flipcarrier, but only for the second or later CRTC. 5302 5303 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (v2) 5304 5305commit b16609b453bb1a181198cf27778f205dc23fb642 5306Author: Michel Dänzer <michel.daenzer@amd.com> 5307Date: Thu Mar 26 17:15:21 2015 +0900 5308 5309 glamor: Handle GLAMOR_* flags removed from xserver 5310 5311 The behaviour is the same as when the removed flags were passed in. 5312 5313 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5314 5315commit 391900a670addec39515f924265bfa9f8bfa9ec0 5316Author: Michel Dänzer <michel.daenzer@amd.com> 5317Date: Wed Mar 18 12:56:07 2015 +0900 5318 5319 glamor: Avoid generating GEM flink names for BOs shared via DRI3 v2 5320 5321 We can't create our own struct radeon_bo representation in this case 5322 because destroying that would make the GEM handle inaccessible to glamor 5323 as well. So just get the handle directly via dma-buf. 5324 5325 v2: Close dma-buf file descriptor, pointed out by Axel Davy. 5326 5327 Reviewed-by: Axel Davy <axel.davy@ens.fr> 5328 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5329 5330commit dfad91fffb5bd013785223b42d78886df839eacf 5331Author: Michel Dänzer <michel.daenzer@amd.com> 5332Date: Thu Mar 19 12:37:01 2015 +0900 5333 5334 Present: Add radeon_present_get_pixmap_handle helper 5335 5336 Reviewed-by: Axel Davy <axel.davy@ens.fr> 5337 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5338 5339commit 7b4fc4a677d252d01c2bf80d162bc35814059eaa 5340Author: Michel Dänzer <michel.daenzer@amd.com> 5341Date: Thu Mar 19 12:15:52 2015 +0900 5342 5343 Make radeon_do_pageflip take a BO handle directly 5344 5345 Reviewed-by: Axel Davy <axel.davy@ens.fr> 5346 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5347 5348commit af6076241c0d322b295a4e898407ae2472bd8eb4 5349Author: Michel Dänzer <michel.daenzer@amd.com> 5350Date: Tue Mar 24 13:37:01 2015 +0900 5351 5352 Adapt radeon_sync.c for misyncshm.h vs misync.h as well 5353 5354 This should have been part of the previous commit. Without this, SYNC 5355 extension fences (and by extension DRI3) were disabled because it checked 5356 for HAVE_MISYNC_H, but configure now only defines HAVE_MISYNCSHM_H. 5357 5358commit aa7825eb29cdf6ac9d7b28ad18186807ff384687 5359Author: Michel Dänzer <michel.daenzer@amd.com> 5360Date: Mon Mar 23 10:12:17 2015 +0900 5361 5362 configure.ac: Check for misyncshm.h again 5363 5364 We need this header, which was added after the other misync*.h headers, 5365 along with DRI3 support. 5366 5367commit 6291baaed261e36a63dc001307427fe00ba82259 5368Author: Michel Dänzer <michel.daenzer@amd.com> 5369Date: Thu Mar 19 00:12:06 2015 +0900 5370 5371 EXA: Return NULL from radeon_get_pixmap_bo if there is no driver private 5372 5373 This mirrors what the glamor part of the function does. Fixes a crash 5374 running glxgears_pixmap with DRI3 enabled, reported by "marvin24" on IRC. 5375 5376commit f68d9b5ba0c91a725b5eec9386c61bea8824c299 5377Author: Michel Dänzer <michel.daenzer@amd.com> 5378Date: Wed Mar 18 18:29:27 2015 +0900 5379 5380 dri3: Use screen->CreatePixmap instead of fbCreatePixmap directly 5381 5382 Fixes crash with EXA reported by "marvin24" on IRC. 5383 5384commit 4b0997e56dec0053cb2cb793e0f4ae35055ff7e6 5385Author: Michel Dänzer <michel.daenzer@amd.com> 5386Date: Wed Mar 18 12:30:09 2015 +0900 5387 5388 glamor: Add glamor_fd_from_pixmap define for standalone glamor tree 5389 5390commit af1862a37570fa512a525ab47d72b30400d2e2d6 5391Author: Michel Dänzer <michel.daenzer@amd.com> 5392Date: Wed Mar 18 11:05:40 2015 +0900 5393 5394 Always include misync.h before other misync headers 5395 5396 Older versions of xserver didn't include misync.h from other misync 5397 headers as needed. 5398 5399commit fcd37f65f485291084c174666bd605e215bf1398 5400Author: Michel Dänzer <michel.daenzer@amd.com> 5401Date: Wed Mar 18 10:39:03 2015 +0900 5402 5403 DRI3: Use open hook instead of open_client 5404 5405 We don't need the client pointer, and this allows it to work with older 5406 versions of xserver as well. 5407 5408commit f940fd741b15f03393037c5bb904cd74f012de9d 5409Author: Michel Dänzer <michel.daenzer@amd.com> 5410Date: Wed Mar 18 10:06:47 2015 +0900 5411 5412 Initialize boolean variable before calling xf86GetOptValBool 5413 5414 We were just lucky that it contained 0 in most cases. 5415 5416commit 4a98f60117c387a228d5cbaadb6e298fb4e865df 5417Author: Michel Dänzer <michel.daenzer@amd.com> 5418Date: Tue Mar 17 16:47:11 2015 +0900 5419 5420 Add xorg_list_for_each_entry_safe fallback in radeon_list.h 5421 5422 Fixes another build failure against older versions of xserver reported 5423 by "Pali" on IRC. 5424 5425commit 694e04720b886060fe3eefdce59741f218c8269f 5426Author: Michel Dänzer <michel.daenzer@amd.com> 5427Date: Tue Mar 17 16:32:09 2015 +0900 5428 5429 Simplify includes in radeon_{dri3,present}.c 5430 5431 Just include radeon.h first. Fixes build failures against older versions 5432 of xserver because compat-api.h wasn't getting picked up. Reported by 5433 "Pali" on IRC. 5434 5435commit 64e1e4dbdd3caee6f5d8f6b6c094b4533fa94953 5436Author: Michel Dänzer <michel.daenzer@amd.com> 5437Date: Tue Mar 10 16:53:11 2015 +0900 5438 5439 Add DRI3 support v2 5440 5441 Must be enabled with 5442 5443 Option "DRI3" 5444 5445 in xorg.conf. 5446 5447 v2: Adapt to v2 of patches 11/12. 5448 5449 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5450 5451commit 69ae0194778fe4276895839db92383f63f0b5de4 5452Author: David Heidelberger <david.heidelberger@ixit.cz> 5453Date: Fri Mar 6 17:57:22 2015 +0900 5454 5455 Handle tiling in radeon_set_shared_pixmap_backing 5456 5457 [ Michel Dänzer: Fixups for glamor ] 5458 Signed-off-by: David Heidelberger <david.heidelberger@ixit.cz> 5459 Signed-off-by: Axel Davy <axel.davy@ens.fr> 5460 5461 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5462 5463commit 3c65fb849e1ba9fb6454bcaa55b696548902f3fc 5464Author: Michel Dänzer <michel.daenzer@amd.com> 5465Date: Tue Mar 10 16:52:54 2015 +0900 5466 5467 Add support for the Present extension v2 5468 5469 v2: Fix up for struct radeon_drm_queue -> radeon_drm_queue_entry. 5470 Swapped order of patches 11 & 12 because the Present extension uses 5471 SYNC fences. 5472 5473 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5474 5475commit 8fc9a241ab59ffbcdc178d6415332c88a54e85fe 5476Author: Michel Dänzer <michel.daenzer@amd.com> 5477Date: Thu Mar 5 18:18:56 2015 +0900 5478 5479 Add support for SYNC extension fences v2 5480 5481 v2: Swapped order of patches 11 & 12 because the Present extension uses 5482 SYNC fences. 5483 5484 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5485 5486commit 4a35e2f33d9cdfb608423046391311109f96fb6b 5487Author: Michel Dänzer <michel.daenzer@amd.com> 5488Date: Thu Mar 5 18:34:07 2015 +0900 5489 5490 Fold radeon_glamor_flush into radeon_cs_flush_indirect 5491 5492 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5493 5494commit 4b8adebb80158bcf81ada83bb88517febe931b12 5495Author: Michel Dänzer <michel.daenzer@amd.com> 5496Date: Thu Mar 5 18:40:23 2015 +0900 5497 5498 Move #include "radeon_glamor.h" from radeon.h to where it's needed 5499 5500 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5501 5502commit 76c2923ac5c7230a8b2f9f8329c308d28b44d9c0 5503Author: Michel Dänzer <michel.daenzer@amd.com> 5504Date: Wed Mar 4 10:30:19 2015 +0900 5505 5506 DRI2: Split out helper for getting UST and MSC of a specific CRTC 5507 5508 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5509 5510commit 6c3a721cde9317233072b573f9502348dcd21b16 5511Author: Michel Dänzer <michel.daenzer@amd.com> 5512Date: Tue Mar 10 16:52:18 2015 +0900 5513 5514 DRI2: Use helper functions for DRM event queue management v3 5515 5516 This is mostly in preparation for Present support, but it also simplifies 5517 the DRI2 specific code a little. 5518 5519 v2: Fix up for struct radeon_drm_queue -> radeon_drm_queue_entry. 5520 v3: Removed excess 0s from conversion from microseconds to seconds, 5521 thanks to Richard Wilbur <richard.wilbur@gmail.com> for the catch! 5522 5523 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5524 5525commit c3fa22a479e61d1899fa9d327d9c4e2a7f64b0c1 5526Author: Michel Dänzer <michel.daenzer@amd.com> 5527Date: Wed Mar 11 17:47:59 2015 +0900 5528 5529 DRI2: Move radeon_dri2_flip_event_handler 5530 5531 In preparation for the next change, which will modify it to a static 5532 function which needs to be in the new place. No functional change. 5533 5534 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5535 5536commit 65045112fdc8a9fa36e0e00f46739a6152b775ff 5537Author: Michel Dänzer <michel.daenzer@amd.com> 5538Date: Wed Mar 11 17:33:54 2015 +0900 5539 5540 DRI2: Remove superfluous assignments to *_info->frame 5541 5542 That field is only used for page flipping. 5543 5544 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5545 5546commit ad27f16f308079d06a2b1c788b3cb0947531253a 5547Author: Michel Dänzer <michel.daenzer@amd.com> 5548Date: Wed Mar 11 17:30:11 2015 +0900 5549 5550 DRI2: Simplify blit fallback handling for scheduled swaps 5551 5552 Also use radeon_dri2_schedule_event when possible. 5553 5554 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5555 5556commit b4af8a327ed8420f0ff4ea0f113f4a59406ed4d3 5557Author: Michel Dänzer <michel.daenzer@amd.com> 5558Date: Mon Mar 2 18:59:54 2015 +0900 5559 5560 Add DRM event queue helpers v2 5561 5562 v2: Rename struct radeon_drm_queue to struct radeon_drm_queue_event, 5563 thanks to Richard Wilbur <richard.wilbur@gmail.com> for the suggestion. 5564 Also changed the corresponding parameter and local variable names from 5565 'q' to 'e'. 5566 5567 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5568 5569commit 7c3470f4b659206ed23f761948936ede3a2dba3d 5570Author: Michel Dänzer <michel.daenzer@amd.com> 5571Date: Fri Mar 6 18:51:29 2015 +0900 5572 5573 Move xorg_list backwards compatibility to new radeon_list.h header 5574 5575 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5576 5577commit 7388d0b6c54b9d536fdb161e3aa61b326627b939 5578Author: Michel Dänzer <michel.daenzer@amd.com> 5579Date: Thu Mar 5 17:35:06 2015 +0900 5580 5581 Require at least xserver 1.8 5582 5583 So we can rely on the list.h header. 5584 5585 xserver 1.8 was released in April 2010. 5586 5587 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5588 5589commit b8ec9ed4fe86952763b963c86f0af0dcae69aa6c 5590Author: Jerome Glisse <jglisse@redhat.com> 5591Date: Thu Feb 19 14:47:41 2015 -0500 5592 5593 Proper leak fix, previous leak fix was bogus. 5594 5595 Signed-off-by: Jérôme Glisse <jglisse@redhat.com> 5596 5597commit 63dc36dc49f93cb00111b497ab6805194bc9d240 5598Author: Jerome Glisse <jglisse@redhat.com> 5599Date: Thu Feb 19 11:43:08 2015 -0500 5600 5601 Avoid leaking memory on output. 5602 5603 Signed-off-by: Jérôme Glisse <jglisse@redhat.com> 5604 5605commit c80ea1e3e8c5f155cedadbe4588870fb1f410ab4 5606Author: Rolf Eike Beer <eb@emlix.com> 5607Date: Tue Jan 13 12:11:26 2015 +0100 5608 5609 radeon: remove unneeded inclusion of sarea.h 5610 5611 None of the structs or defines from that header is used. 5612 5613 Signed-off-by: Rolf Eike Beer <eb@emlix.com> 5614 5615commit 04da199231bb3f11cf17f94574a8df05855a7b82 5616Author: Adam Jackson <ajax@redhat.com> 5617Date: Wed Dec 17 14:03:58 2014 -0500 5618 5619 Remove dead accelDFS flag 5620 5621 Signed-off-by: Adam Jackson <ajax@redhat.com> 5622 5623commit c9f8f642fd495937400618a4fc25ecae3f8888fc 5624Author: Michel Dänzer <michel.daenzer@amd.com> 5625Date: Mon Nov 17 12:32:29 2014 +0900 5626 5627 Prefer drmModeSetCursor2 over drmModeSetCursor 5628 5629 The former includes information about the position of the hotspot within 5630 the cursor image. 5631 5632 Copied from xf86-video-modesetting. 5633 5634 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5635 5636commit 2f11dcd43966cf2ee26e61960fd72e6644f5e037 5637Author: Dave Airlie <airlied@redhat.com> 5638Date: Mon Nov 10 13:49:29 2014 +1000 5639 5640 radeon: add support for DP 1.2 display hotplug (v2) 5641 5642 This allows for dynamic creation of conneectors when the 5643 kernel tells us. 5644 5645 v2: fix dpms off crash 5646 5647 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5648 Signed-off-by: Dave Airlie <airlied@redhat.com> 5649 5650commit c88424d1f4aaa78b569e5d44f0b4a47de2f422f4 5651Author: Dave Airlie <airlied@redhat.com> 5652Date: Mon Nov 10 14:17:54 2014 +1000 5653 5654 radeon: move output name creation to its own function 5655 5656 The secondary indent is deliberate to make the next patch more 5657 parseable for mst support. 5658 5659 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5660 Signed-off-by: Dave Airlie <airlied@redhat.com> 5661 5662commit 32b003cb7657e07d5af6338ad44d768eda87fd33 5663Author: Dave Airlie <airlied@redhat.com> 5664Date: Mon Nov 10 14:12:34 2014 +1000 5665 5666 radeon: stop caching mode resources 5667 5668 This is step one towards MST connector hotplug support, 5669 it stop caching the mode resources structure, and 5670 just passes a pointer to it around. 5671 5672 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5673 Signed-off-by: Dave Airlie <airlied@redhat.com> 5674 5675commit 7c7b38e0b375b6e8853ad2d1092302ea83f6f570 5676Author: David Heidelberger <david.heidelberger@ixit.cz> 5677Date: Sun Oct 12 16:34:21 2014 +0200 5678 5679 radeon/vdpau: don't report VDPAU for < r300 5680 5681 Signed-off-by: David Heidelberger <david.heidelberger@ixit.cz> 5682 5683commit c74de9fec13fac2c836bb2a07ae6f90e1d61e667 5684Author: Michel Dänzer <michel.daenzer@amd.com> 5685Date: Wed Aug 6 11:08:00 2014 +0900 5686 5687 PRIME: Don't advertise offload capabilities when acceleration is disabled 5688 5689 Xorg tends to crash if the user tries to actually use the offload 5690 capabilities with acceleration disabled. 5691 5692 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=57200 5693 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5694 5695commit 636a6382f80412ae4e28abe0cd9f98f84cd52291 5696Author: Michel Dänzer <michel.daenzer@amd.com> 5697Date: Wed Oct 15 12:02:17 2014 +0900 5698 5699 Remove duplicate OPTION_PAGE_FLIP entry 5700 5701 Reported by 'amanual' on IRC 5702 5703commit c854b4479ec5122390ede81b655b259efd9b1800 5704Author: Alex Deucher <alexdeucher@gmail.com> 5705Date: Thu Oct 2 10:22:22 2014 -0400 5706 5707 radeon: bump version post release 5708 5709 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 5710 5711commit 068a59e010ce6bfcd54f5a18cc08c55c54b8618d 5712Author: Alex Deucher <alexander.deucher@amd.com> 5713Date: Wed Oct 1 23:38:52 2014 -0400 5714 5715 radeon: bump version for release 5716 5717 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 5718 5719commit 5fee565e420d3efbf72fdf4f63c3d5d93d8ceddb 5720Author: Alex Deucher <alexander.deucher@amd.com> 5721Date: Wed Oct 1 23:36:31 2014 -0400 5722 5723 radeon: require libdrm_radeon 2.4.58 for latest SI/CI pci ids 5724 5725 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 5726 5727commit 56c7fb8ada4cd9cf096c6b06a8d368d286d74e68 5728Author: Thomas DeRensis <tderensis@gmail.com> 5729Date: Wed Sep 10 21:20:32 2014 -0400 5730 5731 radeon: fix build warnings regarding const qualifier 5732 5733 Signed-off-by: Thomas DeRensis <tderensis@gmail.com> 5734 5735commit b9a9b244336ed50d7df4f36135f875a3bb6ca948 5736Author: Thomas DeRensis <tderensis@gmail.com> 5737Date: Thu Sep 4 21:45:19 2014 -0400 5738 5739 radeon: fix a leak in radeon_vbo_get_bo() 5740 5741 The dma_bo struct was never freed in the error path. Reported by cppcheck. 5742 5743 Signed-off-by: Thomas DeRensis <tderensis@gmail.com> 5744 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 5745 5746commit ba5fcc4d6b6eae3f2c55c6229a53d24d160789c1 5747Author: Alex Deucher <alexander.deucher@amd.com> 5748Date: Thu Aug 21 11:32:45 2014 -0400 5749 5750 radeon/kms: fix mullins pci id 5751 5752 strange errant change from: 5753 39fef269f521c92a2a31c80447e9401bacb3797a 5754 5755 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 5756 5757commit efef34930650d6d80f7b527f4cee76d9e5954ace 5758Author: Alex Deucher <alexander.deucher@amd.com> 5759Date: Thu Aug 21 11:30:17 2014 -0400 5760 5761 radeon/kms: add new SI pci ids 5762 5763 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 5764 5765commit 39fef269f521c92a2a31c80447e9401bacb3797a 5766Author: Alex Deucher <alexander.deucher@amd.com> 5767Date: Thu Aug 21 11:27:49 2014 -0400 5768 5769 radeon/kms: add new CIK pci ids 5770 5771 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 5772 5773commit 7dab6b2fe9cf690b0dae1c127d0374322bd8278e 5774Author: Michel Dänzer <michel.daenzer@amd.com> 5775Date: Wed Aug 20 17:51:35 2014 +0900 5776 5777 Revert "glamor: Set environment variable RADEON_THREAD=0" 5778 5779 This reverts commit 4b5060f357a3cb248c9359c92c1e9c42ef6434c8. 5780 5781 Further testing shows that disabling the thread is only a moderate win in 5782 some cases, but a much bigger loss in some other cases. 5783 5784 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5785 5786commit 1ff383360886c5cee1bb75abcc20675aca0336a6 5787Author: Alex Deucher <alexander.deucher@amd.com> 5788Date: Tue Aug 12 12:27:12 2014 -0400 5789 5790 radeon: fix warnings when building against older xservers 5791 5792 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 5793 5794commit 94202cbfbca05a503acdc1cca2f8409d141173af 5795Author: Alex Deucher <alexdeucher@gmail.com> 5796Date: Fri Aug 1 21:55:40 2014 +0200 5797 5798 radeon: enable hawaii accel conditionally (v3) 5799 5800 Only if the kernel has the new CP firmware. 5801 5802 v2: check value of ACCEL_WORKING2 5803 v3 (Andreas Boll): 5804 - check for value 2 or 3 of ACCEL_WORKING2 5805 - update man page 5806 5807 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> (v2) 5808 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 5809 Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com> 5810 5811commit 03930edd49f6b8a8d79910c7be5408b47db9649b 5812Author: Andreas Boll <andreas.boll.dev@gmail.com> 5813Date: Mon Aug 4 16:23:13 2014 +0200 5814 5815 radeon: remove definitions already present in radeon_drm.h 5816 5817 Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com> 5818 5819commit 91849fba0742ef61ba327e71fc3ce8f754af0a6f 5820Author: Andreas Boll <andreas.boll.dev@gmail.com> 5821Date: Mon Aug 4 16:23:12 2014 +0200 5822 5823 radeon: drop radeon_drm.h 5824 5825 Now we use libdrm's radeon_drm.h. 5826 5827 Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com> 5828 5829commit b4658901bf1d619f8ff9b5f94344894f935ee6c8 5830Author: Andreas Boll <andreas.boll.dev@gmail.com> 5831Date: Mon Aug 4 16:23:11 2014 +0200 5832 5833 radeon: move RADEON_TILING_{MASK, LINEAR} from radeon_drm.h to radeon.h 5834 5835 This allows us to drop radeon_drm.h from xf86-video-ati and use instead 5836 radeon_drm.h from libdrm. 5837 5838 Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com> 5839 5840commit 656bae7361c1e018553ef6d6d8c9efad616a4513 5841Author: Andreas Boll <andreas.boll.dev@gmail.com> 5842Date: Mon Aug 4 16:23:10 2014 +0200 5843 5844 radeon: drop redundant radeon_drm.h includes 5845 5846 Already included via radeon.h. 5847 5848 Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com> 5849 5850commit fbf575cb010e558a87fad0de45660738c8180896 5851Author: Michel Dänzer <michel.daenzer@amd.com> 5852Date: Wed Jul 30 17:03:24 2014 +0900 5853 5854 Add Emacs .dir-local.el file 5855 5856 Based on the one from the Gallium radeon winsys, but enabling tabs for 5857 indentation. 5858 5859commit 4b5060f357a3cb248c9359c92c1e9c42ef6434c8 5860Author: Michel Dänzer <michel.daenzer@amd.com> 5861Date: Thu Jun 19 18:27:59 2014 +0900 5862 5863 glamor: Set environment variable RADEON_THREAD=0 5864 5865 Reviewed-by: Marek Olšák <marek.olsak@amd.com> 5866 5867commit 9b54caf6509a9c02dd17c9c43d6be8f7ddc98054 5868Author: Hans de Goede <hdegoede@redhat.com> 5869Date: Mon Jul 28 15:55:12 2014 +0200 5870 5871 configure: Include xorg-server.h before 5872 glamor.h 5873 5874 glamor.h cannot be included without first including xorg-server.h, this also 5875 applies to including it from configure snippets. 5876 5877 Without this the configure glamor checks fail on systems with the latest 5878 glibc, throwing this error: 5879 5880 In file included from /usr/include/xorg/misc.h:115:0, 5881 from /usr/include/xorg/screenint.h:50, 5882 from /usr/include/xorg/scrnintstr.h:50, 5883 from /usr/include/xorg/glamor.h:32, 5884 from conftest.c:61: 5885 /usr/include/xorg/os.h:579:2: error: expected identifier or '(' before 5886 '__exten 5887 strndup(const char *str, size_t n); 5888 ^ 5889 5890 This is caused by HAVE_STRNDUP not being set (it is set from xorg-server.h), 5891 causing os.h to redefine it. 5892 5893 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 5894 5895commit c4ae0e2cbcc0e2ebf9f13ee92d59b5120254a1dc 5896Author: Michel Dänzer <michel.daenzer@amd.com> 5897Date: Mon Jun 30 10:20:12 2014 +0900 5898 5899 Handle CRTC DPMS from output DPMS hooks 5900 5901 This fixes at least two issues: 5902 5903 The CRTC DPMS hook isn't called after a modeset, so the vertical blank 5904 interrupt emulation code considered the CRTC disabled after a modeset. As 5905 a side effect, page flipping was no longer used after a modeset. 5906 5907 This change also makes sure the vertical blank interrupt emulation code 5908 runs before the hardware CRTC is disabled and after it's enabled from the 5909 output DPMS hook. The wrong order could cause gnome-shell to hang after 5910 a suspend/resume and/or DPMS off/on cycle. 5911 5912 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 5913 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5914 5915commit c0c3cac8613e31e310242695d0955b452f116e25 5916Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 5917Date: Wed Jun 25 16:21:57 2014 +0200 5918 5919 bump version post release 5920 5921commit 906a0ec9224146098bb4581486129d2934d36495 5922Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 5923Date: Wed Jun 25 16:10:21 2014 +0200 5924 5925 bump version for release 5926 5927commit cc615d06db0332fc6e673b55632bcc7bf957b44b 5928Author: Michel Dänzer <michel.daenzer@amd.com> 5929Date: Wed Jun 4 16:58:53 2014 +0900 5930 5931 Rename Option "NoAccel" to "Accel" 5932 5933 Removes the need for a double negation when forcing acceleration on. 5934 5935 Note that this change is backwards compatible, as the option parser 5936 automagically handles the 'No' prefix. 5937 5938 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5939 5940commit 851b2cf8714618843725f6d067915375485ade9d 5941Author: Adam Jackson <ajax@redhat.com> 5942Date: Wed May 21 09:34:32 2014 -0400 5943 5944 kms: Use own thunk function instead of shadowUpdatePackedWeak 5945 5946 I plan to delete the Weak functions from a future server. 5947 5948 Signed-off-by: Adam Jackson <ajax@redhat.com> 5949 5950commit b2dba2906f0b2284f17f53fd5251ba0f03d52a8b 5951Author: Michel Dänzer <michel.daenzer@amd.com> 5952Date: Thu May 15 16:07:53 2014 +0900 5953 5954 Don't disable acceleration on >= SI on attempts to force EXA 5955 5956 Also make this case clear in the log file: 5957 5958 (WW) RADEON(0): EXA not supported, using glamor 5959 5960 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 5961 5962commit be1469cc23aba46daf3293b3d09c5f2e792e7f42 5963Author: Alex Deucher <alexander.deucher@amd.com> 5964Date: Fri May 2 17:48:07 2014 -0400 5965 5966 radeon: enable tiling for mullins 5967 5968 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 5969 5970commit 2ae8e4b8d12f5c9bab6655eb8cd3c5c1d5cfb10e 5971Author: Samuel Li <samuel.li@amd.com> 5972Date: Tue Nov 12 15:30:42 2013 -0500 5973 5974 radeon: add Mullins pci ids. 5975 5976 Signed-off-by: Samuel Li <samuel.li@amd.com> 5977 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 5978 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 5979 5980commit 5c86a3461597647224c33d5190b4531aeeb2655f 5981Author: Samuel Li <samuel.li@amd.com> 5982Date: Thu Apr 17 15:17:28 2014 -0400 5983 5984 radeon: add support for Mullins. 5985 5986 Signed-off-by: Samuel Li <samuel.li@amd.com> 5987 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 5988 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 5989 5990commit fdb90ffc50acbb7d5ba0598470f9feeac6ce55fc 5991Author: Alex Deucher <alexander.deucher@amd.com> 5992Date: Fri May 2 17:45:45 2014 -0400 5993 5994 radeon: require libdrm_radeon 2.4.54 for mullins support 5995 5996 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 5997 5998commit 06e3c8c53ef4bd159f5864eabf726438d008b49a 5999Author: Dave Airlie <airlied@redhat.com> 6000Date: Wed Apr 23 13:39:42 2014 +1000 6001 6002 radeon: fix use-after-free in modesetting cleanup 6003 6004 noticed while looking at something else. 6005 6006 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 6007 Signed-off-by: Dave Airlie <airlied@redhat.com> 6008 6009commit dbac18c361f9e514ecb40d0617f9d68b65a542e0 6010Author: Michel Dänzer <michel.daenzer@amd.com> 6011Date: Mon Apr 28 17:51:56 2014 +0900 6012 6013 Revert "Adapt to load_cursor_argb signature change in xserver 1.15.99.902" 6014 6015 This reverts commit 48d3dbc8a0d3bfde88f46e402e530438f9317715. 6016 6017 xserver Git has been updated to be backwards compatible with the 6018 previous API. 6019 6020commit c84230d686c078aac1dc98d82153f8b02521b2e1 6021Author: Michel Dänzer <michel.daenzer@amd.com> 6022Date: Fri Apr 25 09:17:51 2014 +0900 6023 6024 dri2: Handle PRIME for source buffer as well in radeon_dri2_copy_region2 6025 6026 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=77810 6027 6028 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6029 6030commit 48d3dbc8a0d3bfde88f46e402e530438f9317715 6031Author: Michel Dänzer <michel.daenzer@amd.com> 6032Date: Tue Apr 15 17:45:35 2014 +0900 6033 6034 Adapt to load_cursor_argb signature change in xserver 1.15.99.902 6035 6036 Apart from the compiler warning below, not doing this may result in 6037 accidentally using software cursors. 6038 6039 ../../src/drmmode_display.c:808:5: warning: initialization from incompatible pointer type [enabled by default] 6040 .load_cursor_argb = drmmode_load_cursor_argb, 6041 ^ 6042 ../../src/drmmode_display.c:808:5: warning: (near initialization for 'drmmode_crtc_funcs.load_cursor_argb') [enabled by default] 6043 6044 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6045 6046commit 409786a8f780d78a83bf0bddea5d37117ff6fa39 6047Author: Michel Dänzer <michel.daenzer@amd.com> 6048Date: Thu Apr 10 11:43:04 2014 +0900 6049 6050 glamor: Fix test for creating shared pixmaps 6051 6052 The pixmap usage hint is not a bitmask in general. The test for 6053 CREATE_PIXMAP_USAGE_SHARED was incorrectly triggering for a glamor internal 6054 usage hint being added in the xserver tree. 6055 6056 Tested-by: Ed Tomlinson <edtoml@gmail.com> 6057 6058commit aecf1c4e5f4718adcfb85836830d065d3f4f97a5 6059Author: Michel Dänzer <michel.daenzer@amd.com> 6060Date: Thu Apr 10 15:10:56 2014 +0900 6061 6062 dri2: Fix conflicting CreatePixmap usage flag definitions 6063 6064 RADEON_CREATE_PIXMAP_TILING_MICRO_SQUARE was the same as 6065 RADEON_CREATE_PIXMAP_DRI2. 6066 6067 Disambiguate the definitions and rearrange them to try and prevent this 6068 from happening again. 6069 6070 Tested-by: Ed Tomlinson <edtoml@gmail.com> 6071 6072commit b50da3b96c212086cb58501dbe988d64f1f35b6d 6073Author: Hans de Goede <hdegoede@redhat.com> 6074Date: Fri Apr 11 09:44:37 2014 +0200 6075 6076 Fix building on older servers without xf86platformBus.h 6077 6078 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 6079 6080commit ed0cfbb4fe77146b0b38f777bc28f3a4ea6da07f 6081Author: Hans de Goede <hdegoede@redhat.com> 6082Date: Fri Mar 7 13:27:30 2014 +0100 6083 6084 Add support for server managed fds 6085 6086 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 6087 6088commit 3d7861fe112f25874319d4cdc12b745fbcd359cf 6089Author: Hans de Goede <hdegoede@redhat.com> 6090Date: Mon Mar 17 10:38:13 2014 +0100 6091 6092 Add radeon_get_drm_master_fd helper function 6093 6094 This is a preparation patch for adding server-managed-fd support without it 6095 turning into a goto fest. 6096 6097 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 6098 6099commit a63342ad15408071437c80b411d14196f3288aed 6100Author: Hans de Goede <hdegoede@redhat.com> 6101Date: Mon Mar 17 10:36:55 2014 +0100 6102 6103 radeon_open_drm_master get rid of unnecessary goto 6104 6105 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 6106 6107commit bdc412044f6ced056cd57320d1b2ee0d967c2191 6108Author: Michel Dänzer <michel.daenzer@amd.com> 6109Date: Thu Mar 13 16:40:19 2014 +0900 6110 6111 Build against glamor in the xserver tree if available 6112 6113commit 921a153f9964ca452e1241f76c7f7d653f42ceaf 6114Author: Alex Deucher <alexander.deucher@amd.com> 6115Date: Tue Mar 11 12:24:13 2014 -0400 6116 6117 update man page to reflect tiling changes for CI parts 6118 6119 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6120 6121commit 5eee8a4d5c86bb1cc34d8caf2f2b64b53c241fa5 6122Author: Alex Deucher <alexander.deucher@amd.com> 6123Date: Mon Mar 10 16:20:09 2014 -0400 6124 6125 radeon: enable tiling by default on CIK 6126 6127 Now that mesa 10.1 is released, we can enable this by 6128 default for CIK parts. Tiling improves memory bandwidth 6129 utilization. 6130 6131 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6132 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 6133 6134commit 691ec3d99c30111a4789830dfccb6eb5d3c40187 6135Author: Alex Deucher <alexander.deucher@amd.com> 6136Date: Mon Mar 10 16:17:34 2014 -0400 6137 6138 radeon: require libdrm 2.4.51 6139 6140 Required for proper tiling support on CIK parts. 6141 6142 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6143 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 6144 6145commit 515bcf14d514f9dcaaf30fd0bf1ef6dd6ba9a0cd 6146Author: Michel Dänzer <michel.daenzer@amd.com> 6147Date: Tue Mar 4 12:34:26 2014 +0900 6148 6149 Allow enabling glamor on R500 (and R300) class 3D engines as well. 6150 6151 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=75709 6152 6153 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6154 6155commit ea6d0affe52d82becadab6fb1c87f9261b0605a2 6156Author: Michel Dänzer <michel.daenzer@amd.com> 6157Date: Mon Feb 24 13:12:21 2014 +0900 6158 6159 Only log debugging output about initializing colormaps when we're doing so 6160 6161 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6162 6163commit 8da17f30c70f4494ce22ad781a1cee17041812f3 6164Author: Jérôme Glisse <jglisse@redhat.com> 6165Date: Mon Feb 24 14:50:25 2014 -0500 6166 6167 evergreen: fix shader constant upload on ppc 6168 6169 The number of dword we have to swap is (16*4) ie 16 vectors of 6170 4 floats each not 16 floats. Never hit this issue before because 6171 we never had more than 4 constant vector. 6172 6173 Signed-off-by: Jérôme Glisse <jglisse@redhat.com> 6174 6175commit cadb6b493942a84bfeb298751dce0dee39257a06 6176Author: Alex Deucher <alexander.deucher@amd.com> 6177Date: Fri Feb 21 08:33:21 2014 -0500 6178 6179 radeon: don't install colormap handling if there are no crtcs 6180 6181 Fixes a crash on cards with 0 crtcs. 6182 6183 Discussion: 6184 http://lists.freedesktop.org/archives/dri-devel/2014-February/054186.html 6185 6186 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6187 6188commit 7cd972a85705341dd8306eefc558ed9e5def05d7 6189Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 6190Date: Thu Feb 20 11:14:03 2014 +0100 6191 6192 return immediately in preinit when called with PROBE_DETECT 6193 6194 This fixes a crash with Xorg -configure. 6195 6196 Bug: 6197 https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/1278046 6198 6199 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 6200 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6201 6202commit 8de6f7b2f476d3baa9c7e2bb3544e4bafaad46b7 6203Author: Alex Deucher <alexander.deucher@amd.com> 6204Date: Fri Jan 24 11:04:30 2014 -0500 6205 6206 bump version post release 6207 6208 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6209 6210commit 0333f5bda27dc0ec2edc180c7a4dc9a432f13f97 6211Author: Alex Deucher <alexander.deucher@amd.com> 6212Date: Fri Jan 24 10:19:49 2014 -0500 6213 6214 radeon: bump version for release 6215 6216 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6217 6218commit fc4167f2a85d9cba65078d8fc6f08c7a619ad66e 6219Author: Alex Deucher <alexander.deucher@amd.com> 6220Date: Fri Jan 24 10:17:08 2014 -0500 6221 6222 Require glamor 0.6.0 6223 6224 This is required for Xv support and a number of important 6225 performance improvements. 6226 6227 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6228 6229commit f2a0a5cf6c5a21e2a02280e110a4eb8e6609dace 6230Author: Michel Dänzer <michel.daenzer@amd.com> 6231Date: Wed Jan 22 11:04:42 2014 +0900 6232 6233 Don't require the glamoregl module to be pre-loaded with xserver >= 1.15 6234 6235 The issues with loading it on demand have been fixed in xserver 1.15. 6236 6237 Inspired by Jérôme Glisse on IRC. 6238 6239commit 3213df16d61302148be0088c8f93c6a5a88558f1 6240Author: Michel Dänzer <michel.daenzer@amd.com> 6241Date: Wed Jan 8 11:30:59 2014 +0900 6242 6243 dri2: Make last_vblank_seq local unsigned to match dpms_last_seq 6244 6245 Without this, I was occasionally running into gnome-shell hangs due to 6246 wildly off vblank sequence values. Doesn't seem to happen anymore with 6247 this change. 6248 6249 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6250 6251commit bcc454ea2fb239e13942270faec7801270615b9c 6252Author: Alex Deucher <alexander.deucher@amd.com> 6253Date: Mon Jan 6 09:52:50 2014 -0500 6254 6255 radeon/exa: Always use a scratch surface for UTS to vram 6256 6257 If we don't, we may hit a buffer that crosses the 6258 visible vram boundary resulting in a sigbus when the 6259 CPU accesses the buffer beyond the PCI aperture. 6260 This will introduce an extra copy in certain cases. 6261 6262 This is based on Michel's patch from bug 44099 updated 6263 for all asic families. 6264 6265 Bug: 6266 https://bugs.freedesktop.org/show_bug.cgi?id=44099 6267 6268 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6269 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 6270 6271commit 04ef035c9315b4a6fbf1b14720be87cee4099a9f 6272Author: Alex Deucher <alexander.deucher@amd.com> 6273Date: Tue Dec 24 15:14:35 2013 -0500 6274 6275 drm/radeon: fix SUMO2 pci id 6276 6277 0x9649 is sumo2, not sumo. 6278 6279 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6280 6281commit bfbff3b246db509c820df17b8fcf5899882ffcfa 6282Author: Robert Millan <rmh@freebsd.org> 6283Date: Fri Dec 20 11:03:14 2013 +0000 6284 6285 radeon: Restore kernel module load on FreeBSD. 6286 6287 Since the introduction of a call to drmCheckModesettingSupported() 6288 in radeon_kernel_mode_enabled(), with abort condition if such call 6289 fails, the drmOpen() call in radeon_open_drm_master() no longer 6290 takes the responsibility of loading the radeon kernel module. 6291 6292 However at least on FreeBSD (and GNU/kFreeBSD), X is still relied 6293 on to load the modules it needs. This commit restores the old 6294 behaviour of loading kernel modules on these systems. 6295 6296 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=72852 6297 Signed-off-by: Robert Millan <rmh@freebsd.org> 6298 6299commit 796c9a0cb587f528326bede11fa3f3eb7d3edaf1 6300Author: Robert Millan <rmh@freebsd.org> 6301Date: Fri Dec 20 10:55:09 2013 +0000 6302 6303 radeon: Set first parameter of drmOpen() to NULL 6304 6305 Since the introduction of a call to drmCheckModesettingSupported() 6306 in radeon_kernel_mode_enabled(), with abort condition if such call 6307 fails, the first argument to drmOpen() call in radeon_open_drm_master() 6308 has become a no-op red herring. 6309 6310 Such argument (a kernel module name) is supposed to result in load 6311 of specified kernel module. However, this will never happen. The 6312 problem is that if the code containing drmOpen() call is reached, it 6313 means that drmCheckModesettingSupported() check has previously 6314 succeeded, which implies the module is already loaded. 6315 6316 So, drmOpen() will never load a kernel module. But it gives the 6317 impression that it will. 6318 6319 In order to avoid this confusion, this commit replaces it with NULL, 6320 like xf86-video-intel driver does. 6321 6322 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=72849 6323 Signed-off-by: Robert Millan <rmh@freebsd.org> 6324 6325commit d571d6af70ef27efd1ed6420eb892bdde963ed7a 6326Author: Alex Deucher <alexander.deucher@amd.com> 6327Date: Tue Sep 24 11:39:10 2013 -0400 6328 6329 radeon/kms: add Hawaii pci ids 6330 6331 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6332 6333commit e38a92e00d015a6b80a1f3a16d58c61f084b066f 6334Author: Alex Deucher <alexander.deucher@amd.com> 6335Date: Tue Sep 24 11:35:00 2013 -0400 6336 6337 radeon: add support for Hawaii 6338 6339 Disabled by default until the acceleration code stablizes. 6340 6341 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6342 6343commit e4cd0f4392ea11c93088ad429f36eaaf9bcbf505 6344Author: Alex Deucher <alexander.deucher@amd.com> 6345Date: Wed Oct 30 15:21:26 2013 -0400 6346 6347 radeon: enable tiling on SI by default (v2) 6348 6349 Now that mesa 9.2 is out with support for tiling 6350 on SI asics, we can enable it here. Tiling improves 6351 memory bandwidth utilization. 6352 6353 V2: update man page 6354 6355 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6356 6357commit 3b38701a72fa1cad1e4610a2f4330b3da4cc6391 6358Author: Vadim Girlin <vadimgirlin@gmail.com> 6359Date: Fri Nov 1 10:36:39 2013 -0400 6360 6361 radeon: disable 2D tiling on buffers < 128 pixels 6362 6363 Seems to run into alignment problems with certain 6364 card configurations. 6365 6366 bug: 6367 https://bugs.freedesktop.org/show_bug.cgi?id=70675 6368 6369 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6370 6371commit 0c921edf0162fed616cea9d02e168b719243bcd2 6372Author: Jerome Glisse <jglisse@redhat.com> 6373Date: Wed Oct 30 13:01:02 2013 -0400 6374 6375 radeon/glamor: with new pixmap for dri2 drawable no need to create new texture 6376 6377 When creating a new pixmap/bo for dri2 drawable there is no need to create a 6378 new texture, instead the texture associated with the new pixmap should be use 6379 otherwise there is a missmatch between the bo backing the texture for the 6380 drawable and the bo used by dri2 client. 6381 6382 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 6383 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6384 6385commit f1dc677e79cd7a88d7379a934ebc7d87a3b18805 6386Author: Christopher James Halse Rogers <raof@ubuntu.com> 6387Date: Mon Sep 23 12:25:29 2013 -0700 6388 6389 EXA/evergreen: Paranoia around linear tiling. (v2) 6390 6391 The last two bytes of tiling_mode contain the actual tiling mode; the rest are 6392 extra tiling configuration bits. These configuration bits are not necessarily 6393 zero for a linear buffer, so mask them out before checking for linearity 6394 6395 v2: Also fix up evergreen_textured_videofuncs.c 6396 6397 agd5f: remove trailing whitespace 6398 6399 Signed-off-by: Christopher James Halse Rogers <raof@ubuntu.com> 6400 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6401 6402commit 67fb82a3f0759b171fea21b475a70fa825693570 6403Author: Alex Deucher <alexander.deucher@amd.com> 6404Date: Tue Oct 1 09:35:30 2013 -0400 6405 6406 radeon: fix the non-glamor build harder... 6407 6408 I need to stop pushing patches first thing in the morning. 6409 6410 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6411 6412commit afc0374fdec3e24ece25805724459463e9a19f5e 6413Author: Alex Deucher <alexander.deucher@amd.com> 6414Date: Tue Oct 1 09:32:02 2013 -0400 6415 6416 drm/radeon: fix non-glamor build 6417 6418 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6419 6420commit 2d791370dfc5570eb74d7a1fb3baf4d4c8ecf243 6421Author: Dave Airlie <airlied@redhat.com> 6422Date: Mon Sep 23 07:57:15 2013 +0100 6423 6424 radeon: use glamor Xv support if present. 6425 6426 This creates adaptors using glamor if possible. 6427 6428 Signed-off-by: Dave Airlie <airlied@redhat.com> 6429 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6430 6431commit cbb99f659ee7b18ded0008a606e41ded38c1a194 6432Author: Alex Deucher <alexander.deucher@amd.com> 6433Date: Mon Sep 30 08:55:16 2013 -0400 6434 6435 Revert "radeon: add glamor Xv support (v2)" 6436 6437 This causes problems if glamor doesn't have Xv support 6438 enabled. I just noticed that Dave has a better version, 6439 so use that instead. 6440 6441 This reverts commit 4fc1fa920584ace2c84d75af82d06962d0c84ec8. 6442 6443commit 4fc1fa920584ace2c84d75af82d06962d0c84ec8 6444Author: Alex Deucher <alexander.deucher@amd.com> 6445Date: Thu Sep 26 11:02:55 2013 -0400 6446 6447 radeon: add glamor Xv support (v2) 6448 6449 v2: guard new glamor Xv bits with USE_GLAMOR 6450 6451 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6452 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 6453 6454commit 41dfe327ac8740ac2cd84def96b5947224e422e7 6455Author: Alex Deucher <alexander.deucher@amd.com> 6456Date: Thu Sep 26 11:00:49 2013 -0400 6457 6458 radeon: fix limit handling for cards with >4G of ram 6459 6460 We can overflow the 32-bit limit. 6461 6462 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6463 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 6464 6465commit c45e728107269c6f51599dad4f6a02ccfef703f1 6466Author: Michel Dänzer <michel.daenzer@amd.com> 6467Date: Wed Sep 18 10:57:52 2013 +0200 6468 6469 DRI2: Install client callback only once 6470 6471 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=60182 6472 Acked-by: Alex Deucher <alexander.deucher@amd.com> 6473 6474commit b955ddd0d41801e4ca0c30a70a5d0b27c3f366c8 6475Author: Alex Deucher <alexander.deucher@amd.com> 6476Date: Mon Sep 16 10:58:22 2013 -0400 6477 6478 glamor: require 0.5.1 or newer 6479 6480 0.5.1 fixes a number of issues. 6481 6482 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6483 6484commit 282587cd0709850e7bacb1d8307065d95dc2c97d 6485Author: Alex Deucher <alexander.deucher@amd.com> 6486Date: Mon Sep 16 09:41:13 2013 -0400 6487 6488 radeon: enable glamor by default (v4) 6489 6490 It's required for newer asics, so enable it by 6491 default. 6492 6493 v2: update the autoconf help. Require --disable-glamor 6494 to build without glamor support. 6495 v3: default to yes rather than auto 6496 v4: fix help text 6497 6498 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6499 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 6500 6501commit fdb7563a5cbc736b09c2864b67a93b475c98b2bd 6502Author: Alex Deucher <alexander.deucher@amd.com> 6503Date: Thu Jan 24 21:17:11 2013 -0500 6504 6505 radeon/kms: add berlin pci ids 6506 6507 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6508 6509commit 8927d33f76ee12bc618fecfc59fc7ff1fcedcd5e 6510Author: Mark Kettenis <kettenis@openbsd.org> 6511Date: Thu Aug 22 10:32:46 2013 -0400 6512 6513 Fix shadowfb on big-endian machines 6514 6515 For shadowfb, the framebuffer needs to have the RADEON_TILING_SURFACE 6516 flag set, otherwise the appropriate byte swapping won't happen. 6517 6518 See https://bugs.freedesktop.org/show_bug.cgi?id=66663 6519 6520 Signed-off-by: Mark Kettenis <kettenis@openbsd.org> 6521 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6522 6523commit fa83d3d1636c315bc43dd622d407eb058e3ef976 6524Author: Alex Deucher <alexander.deucher@amd.com> 6525Date: Tue Aug 20 09:29:23 2013 -0400 6526 6527 radeon: disallow glamor on pre-R600 asics 6528 6529 I'm not sure they can handle the shaders properly, especially 6530 only older parts like r300. 6531 6532 This will avoid display corruption problems reported by people 6533 using glamor on older asics by falling back to EXA if they try 6534 and enable glamor. 6535 6536 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6537 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 6538 6539commit d0323622ee9b97a5f246baffbb2c65930a78ed14 6540Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 6541Date: Wed Aug 7 11:27:07 2013 +0200 6542 6543 bump version post release 6544 6545 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 6546 6547commit 9c97cca5c24409ca8447c99f051a12fd2d494e79 6548Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 6549Date: Wed Aug 7 10:48:17 2013 +0200 6550 6551 radeon: bump version for release 6552 6553 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 6554 6555commit 16270cfb202ab67dd152644ef019b2f1ee4d0341 6556Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 6557Date: Wed Aug 7 10:29:33 2013 +0200 6558 6559 add bicubic_table.py to EXTRA_DIST 6560 6561 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 6562 6563commit 2cb9197ca7a337c911f38b5de562a2364b922b86 6564Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 6565Date: Wed Aug 7 10:28:52 2013 +0200 6566 6567 kill unused radeon_driver.c 6568 6569 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 6570 6571commit c5cbfcf575b0b4aea6f797558ae974c1453c8e07 6572Author: Alex Deucher <alexander.deucher@amd.com> 6573Date: Tue Jul 30 10:08:25 2013 -0400 6574 6575 drmmode: add support for multi-screen reverse optimus 6576 6577 Initial reverse optimus didn't consider multiple screens, so 6578 this overhauls the code to use the new X server interface, 6579 and allows for multiple outputs on the dGPU to be used with 6580 the iGPU doing the rendering. Ported from Dave's nouveau 6581 patch. 6582 6583 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6584 6585commit 429d5b797769895eb4f5fef816ce4e2f3a342031 6586Author: Dave Airlie <airlied@redhat.com> 6587Date: Tue Jan 8 15:56:37 2013 +1000 6588 6589 radeon: add support for reverse prime (v2) 6590 6591 This adds support for reverse prime configurations 6592 6593 v2: fix compilation with older xservers 6594 6595 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 6596 6597commit 4de9356a2900ae0fb380a2350791ef045629cd05 6598Author: Alex Deucher <alexander.deucher@amd.com> 6599Date: Mon Aug 5 17:57:16 2013 -0400 6600 6601 radeon: fix naming clashes with multiple GPUs (v3) 6602 6603 The compat naming code for UMS causes problems 6604 with multiple GPU as you may end up with the same 6605 output name on multiple GPUs. Adjust the naming on 6606 secondary GPUs to avoid conflicts. 6607 6608 v2: integrate Dave's fixes for nouveau 6609 v3: keep compat with existing naming on primary GPU 6610 6611 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6612 6613commit 2ae6bb18fefddb309920fa69c9b56c3a7f3db7b4 6614Author: Grigori Goronzy <greg@chown.ath.cx> 6615Date: Wed Jul 31 12:01:20 2013 +0200 6616 6617 EXA/evergreen/ni: replace magic number 6618 6619 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 6620 6621commit 6a278369c05a298a4367306d986467a9ceacae8c 6622Author: Raul Fernandes <rgfernandes@gmail.com> 6623Date: Tue Jul 30 09:26:05 2013 -0400 6624 6625 EXA/6xx/7xx: optimize non-overlapping Copy 6626 6627 In case dst and src rectangles of a Copy operation in the same surface 6628 don't overlap, it is safe to skip the scratch surface. This is a 6629 common case. 6630 6631 Based on evergreen/ni patch from Grigori Goronzy. 6632 6633 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6634 6635commit 4375a6e75e5d41139be7031a0dee58c057ecbd07 6636Author: Grigori Goronzy <greg@chown.ath.cx> 6637Date: Mon Jul 22 02:30:28 2013 +0200 6638 6639 EXA/evergreen/ni: accelerate PictOpOver with component alpha 6640 6641 Subpixel text rendering is typically done with a solid src and a 6642 pixmap mask. Traditionally, this cannot be accelerated in a single 6643 pass and requires two passes [1]. However, we can cheat a little 6644 with a constant blend color. 6645 6646 We can use: 6647 6648 const.A = src.A / src.A 6649 const.R = src.R / src.A 6650 const.G = src.G / src.A 6651 const.B = src.B / src.A 6652 6653 dst.A = const.A * (src.A * mask.A) + (1 - (src.A * mask.A)) * dst.A 6654 dst.R = const.R * (src.A * mask.R) + (1 - (src.A * mask.R)) * dst.R 6655 dst.G = const.G * (src.A * mask.G) + (1 - (src.A * mask.G)) * dst.G 6656 dst.B = const.B * (src.A * mask.B) + (1 - (src.A * mask.B)) * dst.B 6657 6658 This only needs a single source value. src.A is cancelled down in 6659 the right places. 6660 6661 [1] http://anholt.livejournal.com/32058.html 6662 6663commit 94d0d14914a025525a0766669b556eaa6681def7 6664Author: Grigori Goronzy <greg@chown.ath.cx> 6665Date: Thu Jul 18 16:06:23 2013 +0200 6666 6667 EXA/evergreen/ni: fast solid pixmap support 6668 6669 Solid pixmaps are currently implemented with scratch pixmaps, which 6670 is slow. This replaces the hack with a proper implementation. The 6671 Composite shader can now either sample a src/mask or use a constant 6672 value. 6673 6674commit 5bb04351c43a91a1d60348b7293544da05d75e72 6675Author: Grigori Goronzy <greg@chown.ath.cx> 6676Date: Fri Jul 27 17:31:53 2012 +0200 6677 6678 EXA/evergreen/ni: optimize non-overlapping Copy 6679 6680 In case dst and src rectangles of a Copy operation in the same surface 6681 don't overlap, it is safe to skip the scratch surface. This is a 6682 common case. 6683 6684commit c08e09b7bec441c4bf93b4cae4de1260754bf940 6685Author: Grigori Goronzy <greg@chown.ath.cx> 6686Date: Sat May 18 13:46:03 2013 +0200 6687 6688 Fix RADEON_FALLBACK logging 6689 6690commit c16c59f8f9b6aa7a4a6a6465582ad98f02a3606a 6691Author: Mark Kettenis <kettenis@openbsd.org> 6692Date: Sun Jul 7 13:44:13 2013 +0200 6693 6694 Always go through DFS/UTS when byte swapping is needed. 6695 6696 Before commit ef9bfb262db7004bef3704e5d914687e50d3fca4 and 6697 e5bd99faa3b6629a55168386d5dfa936ee4e97ae, byte swapping for the front buffer 6698 used to be done in hardware. Now that this no longer happens we need to let 6699 DFS/UTS ihandle the byte swapping. 6700 6701 See https://bugs.freedesktop.org/show_bug.cgi?id=66663 . 6702 6703 Signed-off-by: Mark Kettenis <kettenis@openbsd.org> 6704 6705commit 1239dbbd8c8d9b55756c1de52cad353171a06522 6706Author: Alex Deucher <alexander.deucher@amd.com> 6707Date: Sun Jul 7 20:22:10 2013 -0400 6708 6709 radeon: bump libdrm requirement to 2.4.46 for CIK support 6710 6711 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6712 6713commit 869e0b187b2d07e637c4482fe50d9f1b0c09f4e4 6714Author: Alex Deucher <alexander.deucher@amd.com> 6715Date: Thu Jan 24 21:12:28 2013 -0500 6716 6717 radeon/kms: add kabini pci ids 6718 6719 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6720 6721commit 0edcefc4c97e4572431c38fe170032470e22e0b7 6722Author: Alex Deucher <alexander.deucher@amd.com> 6723Date: Fri Jun 7 14:56:26 2013 -0400 6724 6725 radeon/kms: add bonaire pci ids 6726 6727 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6728 6729commit ebc32b27af25b23604e725eb50d844a8d26116bb 6730Author: Alex Deucher <alexander.deucher@amd.com> 6731Date: Fri Aug 31 18:17:24 2012 -0400 6732 6733 radeon: update cursor handling for CIK 6734 6735 CIK asics have 128x128 hw cursors 6736 6737 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6738 6739commit 4adaea996454b1ad5185f9c0f37667dbfc266495 6740Author: Alex Deucher <alexander.deucher@amd.com> 6741Date: Fri Jun 7 14:47:36 2013 -0400 6742 6743 radeon: add family ids for CIK 6744 6745 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 6746 6747commit 3626ab147b6749a355e7d21710447f13b9796c77 6748Author: Leo Liu <leo.liu@amd.com> 6749Date: Fri Jun 14 10:27:20 2013 -0400 6750 6751 radeon: fix a memory leak in get_modes() callback. 6752 6753 leak happens when looping xrandr prop. 6754 6755 Signed-off-by: Leo Liu <leo.liu@amd.com> 6756 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 6757 6758commit c0f7d03cb29166c6c00dd6fb698ba3f70c59dbdf 6759Author: Alex Deucher <alexdeucher@gmail.com> 6760Date: Mon Jun 17 11:38:27 2013 -0400 6761 6762 man: fix up previous commit 6763 6764 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 6765 6766commit 03b5edc9dd3859eed2504155970613aaf88df852 6767Author: Eric S. Raymond <esr@thyrsus.com> 6768Date: Fri Jun 7 13:23:14 2013 -0400 6769 6770 Fix some incomplete list entries. 6771 6772 This necessarily involved adding some content, which I was able to 6773 Google for. Please review, correct, and make more specific. 6774 6775 Bodiless .TP entries screw up doclifter's page parsing. Please 6776 don't do that anymore. 6777 6778commit 7ce43639a8e53eefb94e5255a39c1232b6bdbf66 6779Author: Michel Dänzer <michel.daenzer@amd.com> 6780Date: Thu Jun 6 09:58:27 2013 +0200 6781 6782 autogen.sh: Restore passing --enable-maintainer-mode to configure 6783 6784 Looks like this was accidentally dropped when the GNOME Build API was 6785 implemented. 6786 6787commit 365e571d9de0b5979425c34210789afca4ea5f68 6788Author: Samuel Li <samuel.li@amd.com> 6789Date: Wed Mar 20 11:59:58 2013 -0400 6790 6791 radeon: use direct mapping for fast fb access. 6792 6793 Signed-off-by: Samuel Li <samuel.li@amd.com> 6794 6795commit bd2557ea5ef84b975060e929d5ece53ec464336f 6796Author: Ilija Hadzic <ilijahadzic@gmail.com> 6797Date: Wed May 8 22:39:48 2013 -0400 6798 6799 DRI2: add interpolated blanks to frame number in event handlers 6800 6801 The 'frame' argument passed to event handlers is the plain 6802 CRTC vblank counter that stops progressing when the 6803 associated display is in DPMS-off mode. If we have a 6804 DPMS-off period the frame counter and MSC will 6805 diverge, which can cause some higher-level functions 6806 to return incorrect values. 6807 6808 This patch fixes the problem by adding interpolated 6809 vblanks to the frame counter before using it in handler 6810 functions. 6811 6812 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 6813 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6814 6815commit 385a92b4fbe8f53b359ef6c463704414d00476fa 6816Author: Ilija Hadzic <ilijahadzic@gmail.com> 6817Date: Wed May 8 22:39:47 2013 -0400 6818 6819 DRI2: support scheduling emulated events with zero delay 6820 6821 Now that we fully emulating a running CRTC through DPMS-off state 6822 it is possible to come up with a zero delay when scheduling 6823 a swap or MSC-wait (e.g., if a call into respective wait function 6824 was entered very late). This patch wraps the TimerSet function 6825 into our own radeon_dri2_schedule_event such that the latter 6826 calls the event right away if zero delay is specified. 6827 6828 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 6829 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6830 6831commit e87b52e6ad41ad7a87e43b818d80e7d522d9c68d 6832Author: Ilija Hadzic <ilijahadzic@gmail.com> 6833Date: Wed May 8 22:39:46 2013 -0400 6834 6835 DRI2: make wait_msc and get_msc work on disabled CRTCs 6836 6837 Now that the running CRTC is emulated through DPMS-off 6838 states, wait_msc and get_msc must also work on disabled 6839 CRTCs. When CRTC is disabled, we must extrapolate the 6840 MSC count from present time and last MSC when CRTC was 6841 running and also use timers facility in wait_msc. 6842 6843 v2: CRTC-private now stores frame rate instead of nominal 6844 vblank period. 6845 6846 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 6847 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6848 6849commit f1584152bbeda98cc1907359f1c7954a84c8837d 6850Author: Ilija Hadzic <ilijahadzic@gmail.com> 6851Date: Wed May 8 22:39:45 2013 -0400 6852 6853 DRI2: move rename and rework radeon_dri2_deferred_swap 6854 6855 radeon_dri2_deferred_swap will be used to generate 6856 real events (not just fallbacks) so now it needs to 6857 generate real timestamp and frame counter. Also 6858 this function will be used both by schedule_swap 6859 and wait_msc, so give it a more generic name: 6860 radeon_dri2_deferred_event 6861 6862 v2: - Extrapolate the frame number from the time 6863 of actual execution of the function instead 6864 of using the MSC calculated (extrapolated) 6865 at event scheduling time. 6866 - CRTC-private now stores frame rate instead of 6867 nominal vblank period. 6868 6869 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 6870 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6871 6872commit 84bce7943b735aee19e26be4d47fdf140564f3f5 6873Author: Ilija Hadzic <ilijahadzic@gmail.com> 6874Date: Wed May 8 22:39:44 2013 -0400 6875 6876 DRI2: track CRTC in event record 6877 6878 Tracking the CRTC associated with an event will save us 6879 some lookups later in event handlers. 6880 6881 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 6882 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6883 6884commit e40d5390b3efdea3e02267413350410d8e6a2970 6885Author: Ilija Hadzic <ilijahadzic@gmail.com> 6886Date: Wed May 8 22:39:43 2013 -0400 6887 6888 DRI2: hook up vblank extrapolation to schedule_swap 6889 6890 This patch hooks up swap-scheduling function with 6891 vblank-extrapolation function. Rather than waiting for 6892 fixed time, we calculate exactly how much we should wait 6893 and what we should update target_msc to using 6894 radeon_dri2_extrapolate_msc_delay helper function 6895 and schedule the swap completion using DIX's timer facility. 6896 6897 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 6898 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6899 6900commit ffaa5abf207415159cdb28e90da49b95f497ef61 6901Author: Ilija Hadzic <ilijahadzic@gmail.com> 6902Date: Wed May 8 22:39:42 2013 -0400 6903 6904 DRI2: add vblank extrapolation function 6905 6906 Implement a helper function that will be called when emulating 6907 the running CRTC. The function should be called only when CRTC 6908 is in DPMS-off state. It will look at the vblank count and the 6909 time that was recorded last time the CRTC was running and 6910 calculate how long one must wait (from present time) until 6911 the target_msc is reached if the CRTC were running. 6912 6913 v2: - CRTC-private now stores frame rate instead of nominal 6914 vblank period. 6915 - DIX's timer facility can sometimes wake up the scheduled 6916 functions more than a millisecond earlier. To avoid 6917 generating an old MSC, we have to add more margin when 6918 converting the delay in microseconds to milliseconds. 6919 6920 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 6921 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6922 6923commit 80ae2291d082b57c70d27a80182f00d760fb3d3a 6924Author: Ilija Hadzic <ilijahadzic@gmail.com> 6925Date: Wed May 8 22:39:41 2013 -0400 6926 6927 DRI2: adjust MSC by the number of interpolated vblanks 6928 6929 Normally, MSC value equals the vblank count of the CRTC 6930 on which the drawable resides. However, we are now interpolating 6931 vblanks through DPMS-off state, so if the CRTC is turned 6932 off at some point, the vlbank count does not change, but 6933 MSC does change by the amount that equals the cumulative 6934 number of interpolated vblanks. 6935 6936 So each time we interact with the CRTC's vblank facility, 6937 we have to adjust the vblank counter: in the request, we 6938 have to subtract the number of interpolated vblanks from 6939 MSC and in the reply we have to add the number of interpolated 6940 vblanks to the MSC. 6941 6942 This patch only makes the MSC correct when CRTC is on. 6943 The subsequent patch will construct the MSC when CRTC is not 6944 running. 6945 6946 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 6947 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6948 6949commit 5fd2eb5d12cea32927a9e6c6ce4afd18aa7d046a 6950Author: Ilija Hadzic <ilijahadzic@gmail.com> 6951Date: Wed May 8 22:39:40 2013 -0400 6952 6953 drmmode: calculate interpolated vblanks while in dpms-off state 6954 6955 This adds provisions for interpolating vblanks while the CRTC 6956 is in DPMS-off state. When entering DPMS-off state, we 6957 record the last vblank time, sequence number and frame rate 6958 in CRTC-private structure. 6959 6960 When going back to DPMS-on state we read the current 6961 time and calculate how long we have been off. Then we derive 6962 how many vblanks that would have been had the CRTC remained 6963 running. These are the interpolated vblanks. 6964 6965 Finally, we accumulate the number of interpolated vblanks 6966 in CRTC-private structure to get the number of interpolated 6967 vblanks over the system lifetime. 6968 6969 v2: Track frame rate instead of vblank period. The former 6970 eliminates some roundoff errors. 6971 6972 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 6973 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6974 6975commit e41ad30d09be4962cfb1942b0b9f63875dbb2d2a 6976Author: Ilija Hadzic <ilijahadzic@gmail.com> 6977Date: Wed May 8 22:39:39 2013 -0400 6978 6979 drmmode: remove dead code from drmmode_crtc_dpms 6980 6981 We are about to modify this function, so let's clean it 6982 up first. 6983 6984 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 6985 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6986 6987commit 373671d2eed30e30b3bdee7e40426cf58c127234 6988Author: Ilija Hadzic <ilijahadzic@gmail.com> 6989Date: Wed May 8 22:39:38 2013 -0400 6990 6991 drmmode: add drmmode_get_current_ust function 6992 6993 The new helper function retrieves current time in the format 6994 that is compatible with vblank timestamps. 6995 6996 v2: - fix an incorrect statement in a comment 6997 - add a #define so that don't depend on libdrm patches that 6998 have not yet been accepted upstream 6999 7000 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 7001 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 7002 7003commit 006fbbd1d38a089b50ab3197d32815689ed249fa 7004Author: Ilija Hadzic <ilijahadzic@gmail.com> 7005Date: Wed May 8 22:39:37 2013 -0400 7006 7007 DRI2: make populate_vbl_request_type external 7008 7009 We'll need to use this function in other .c files so 7010 remove static annotation from it. Also, add radeon_ 7011 prefix because this function is now part of global namespace, 7012 so it would be nice to know where it is coming from. 7013 7014 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 7015 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 7016 7017commit 34660d8757ff16acd1686ff2872f4600d92a68bf 7018Author: Ilija Hadzic <ilijahadzic@gmail.com> 7019Date: Wed May 8 22:39:36 2013 -0400 7020 7021 DRI2: change signature of populate_vbl_request_type 7022 7023 This function no longer uses info argument, so nuke it. 7024 7025 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 7026 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 7027 7028commit ab6e85d1f5dd4657591b314b85fd2453f99c3712 7029Author: Michel Dänzer <michel.daenzer@amd.com> 7030Date: Fri Mar 22 10:51:05 2013 +0100 7031 7032 glamor: Keep old private in radeon_pixmap_set_bo 7033 7034commit 0a84f2760eccf07781ae6c3f55d2d246c155c9ce 7035Author: Alex Deucher <alexander.deucher@amd.com> 7036Date: Mon May 13 17:06:00 2013 -0400 7037 7038 radeon: add Hainan pci ids 7039 7040 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 7041 7042commit 6cee1eaebfcef2b2d2c51ec30c31260fd98b564e 7043Author: Alex Deucher <alexander.deucher@amd.com> 7044Date: Mon May 13 17:01:30 2013 -0400 7045 7046 radeon: add hainan chip family 7047 7048 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 7049 7050commit 8a9e9b21cbb44c23d37bf18c983e2c3f0fe33e87 7051Author: Michel Dänzer <michel.daenzer@amd.com> 7052Date: Thu Apr 25 17:10:12 2013 +0200 7053 7054 glamor: Fix build against xserver < 1.13 7055 7056 Those didn't define CREATE_PIXMAP_USAGE_SHARED. 7057 7058commit 6216b5aa3793abe71b864170bc9d4d745f027328 7059Author: Alex Deucher <alexander.deucher@amd.com> 7060Date: Thu Apr 25 14:19:37 2013 -0400 7061 7062 radeon: add new richland pci ids 7063 7064 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 7065 7066commit fd5903f15b36555bf1ccc1c3865c9fffd499128a 7067Author: Alex Deucher <alexander.deucher@amd.com> 7068Date: Thu Apr 25 14:17:51 2013 -0400 7069 7070 radeon: add new SI pci ids 7071 7072 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 7073 7074commit b2826580e4d246d45871b413e4deecb070162577 7075Author: Jerome Glisse <jglisse@redhat.com> 7076Date: Mon Apr 8 13:45:18 2013 -0400 7077 7078 radeonsi: add 2d tiling support 7079 7080 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 7081 7082commit 6e74aacc5e5da3b51744153dad1645caa6ea4ce3 7083Author: Michel Dänzer <michel.daenzer@amd.com> 7084Date: Wed Mar 27 11:57:29 2013 +0100 7085 7086 glamor: Only enable for depth >= 24 7087 7088 It can't handle lower depths yet. 7089 7090 Reported-by: Igor Vagulin <igor.vagulin@gmail.com> 7091 7092commit 33d8408eec806355c2e55726679ec50ef3b769f1 7093Author: Michel Dänzer <michel.daenzer@amd.com> 7094Date: Tue Mar 12 19:03:06 2013 +0100 7095 7096 glamor: Enable by default on SI 7097 7098 We should now handle its major initialization failure cases gracefully. 7099 7100 Reviewed-by: Jerome Glisse <jglisse@redhat.com> 7101 7102commit ad1d8874fab30257ba8ec8bee8ffa42a8fa9b5c0 7103Author: Michel Dänzer <michel.daenzer@amd.com> 7104Date: Fri Mar 15 18:28:27 2013 +0100 7105 7106 glamor: Initial PRIME pixmap sharing hooks. 7107 7108 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=57200 7109 7110 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 7111 Tested-by: Christoph Haag <haagch.christoph@googlemail.com> 7112 Tested-by: linedot <linedot@xcpp.org> 7113 7114commit 35b384a00a0c76ea556f55787fccc95ecd51bc0c 7115Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7116Date: Thu Mar 21 15:34:21 2013 +0100 7117 7118 drmmode: handle failure of drmModeGetConnector 7119 7120 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7121 7122commit 751c7259cf23a29fdd2653881e2db63ce3cf1ca5 7123Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7124Date: Thu Mar 21 10:40:42 2013 +0100 7125 7126 Move fd cleanup to correct place. 7127 7128 Oops, just after previous commit I realized that it 7129 needs to be done from RADEONFreeRec, because when preinit fails 7130 info pointer can be null in FreeScreen. 7131 7132 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7133 7134commit 1643b8cd2dc53ed36916e11003590c7037b4ddd6 7135Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7136Date: Thu Mar 21 10:33:19 2013 +0100 7137 7138 radeon: add refcounts to fix up zaphod open/close. 7139 7140 Oops, turns out my previous commits were buggy. 7141 Adding proper refcounts will handle this correctly. 7142 7143 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7144 7145commit a4d1bc43d4d10e2ca1e512c1010e962c587d6c16 7146Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7147Date: Wed Mar 20 17:13:44 2013 +0100 7148 7149 Cleanup drmmode in CloseScreen. 7150 7151 Prevents a use-after-free. 7152 7153 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7154 7155commit 3cb3ab8704813d7c14b9e547a30e37598d70c85a 7156Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7157Date: Wed Mar 20 17:12:29 2013 +0100 7158 7159 Close drm_fd in CloseScreen. 7160 7161 This prevents a small fd leak. 7162 7163 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7164 7165commit 462d8d3b86038438ab854072e546cc7a9efacdd7 7166Author: Michel Dänzer <michel.daenzer@amd.com> 7167Date: Tue Mar 12 18:59:19 2013 +0100 7168 7169 glamor: Bail if the glamoregl module wasn't loaded early 7170 7171 Trying to initialize glamor in that case crashes. 7172 7173 Reviewed-by: Jerome Glisse <jglisse@redhat.com> 7174 7175commit 3e5350bedf50a1a140e0eba08c39ed2451fb9115 7176Author: Alex Deucher <alexander.deucher@amd.com> 7177Date: Fri Mar 8 14:07:42 2013 -0500 7178 7179 radeon: add Richland pci ids 7180 7181 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 7182 7183commit bdbf766d8170f6ea6c16b9780ef5a4007730588b 7184Author: Alex Deucher <alexander.deucher@amd.com> 7185Date: Fri Mar 8 14:00:09 2013 -0500 7186 7187 radeon: mark trinity APUs with IGP flag 7188 7189 Was missing on trinity. 7190 7191 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 7192 7193commit 5ac335eaae135c0e66cc4643908f41174e2f1123 7194Author: Michel Dänzer <michel.daenzer@amd.com> 7195Date: Fri Mar 8 17:37:41 2013 +0100 7196 7197 Only parse Option "AccelMethod" if glamor was enabled at build time 7198 7199 Will hopefully make it more obvious when it wasn't. 7200 7201commit f1059e3f0934ccf1ae183099dbd4de700a7c81ac 7202Author: Colin Walters <walters@verbum.org> 7203Date: Wed Jan 4 17:37:06 2012 -0500 7204 7205 autogen.sh: Implement GNOME Build API 7206 7207 http://people.gnome.org/~walters/docs/build-api.txt 7208 7209 Signed-off-by: Adam Jackson <ajax@redhat.com> 7210 7211commit 343b01c9bf35b125cd0c3df8db7c01a5fb227bda 7212Author: Alex Deucher <alexander.deucher@amd.com> 7213Date: Thu Jan 24 19:57:23 2013 -0500 7214 7215 radeon/kms: add oland pci ids 7216 7217 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 7218 7219commit 9980f1e0e0a5df5fb0c791687e5c98c13089d7d1 7220Author: Alex Deucher <alexander.deucher@amd.com> 7221Date: Thu Jan 24 19:52:04 2013 -0500 7222 7223 radeon: add oland chip family 7224 7225 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 7226 7227commit d89e545826969910226cb96e73b64f18e92d2064 7228Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7229Date: Wed Jan 30 17:21:05 2013 +0100 7230 7231 bump version post release 7232 7233 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7234 7235commit 21a4f3eaca98c1c54bffcdf3926c631cec9c409d 7236Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7237Date: Wed Jan 30 16:55:27 2013 +0100 7238 7239 radeon: bump version for release 7240 7241 And fixup distcheck.. 7242 7243 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 7244 7245commit 102ee4a24e4c9ba0c0b65d3ca8599dc7e0216c1e 7246Author: Alex Deucher <alexdeucher@gmail.com> 7247Date: Fri Jan 11 09:52:32 2013 -0500 7248 7249 radeon/r200: program RE_WIDTH_HEIGHT properly 7250 7251 Values are inclusive. Mesa already sets these 7252 correctly. Also bump EXA/Xv limits from 2047 to 7253 2048. 7254 7255 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 7256 7257commit 47689ef8b878a56ea094e21a6090b228e67d605e 7258Author: Michel Dänzer <michel.daenzer@amd.com> 7259Date: Fri Jan 11 15:41:48 2013 +0100 7260 7261 Make radeon_glamor_create_textured_pixmap() return TRUE without USE_GLAMOR. 7262 7263 Otherwise rotation won't work when glamor support isn't built in. 7264 7265 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7266 7267commit 1a8c05fd5559be842d8a038df73de34d4b261c15 7268Author: Michel Dänzer <michel.daenzer@amd.com> 7269Date: Fri Jan 11 13:22:24 2013 +0100 7270 7271 Make CRTC shadow pixmaps usable by glamor. 7272 7273 Fixes rotation with glamor. 7274 7275 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7276 7277commit e5e22d788b103c04d3460a221a3b9dc3be69e0a4 7278Author: Michel Dänzer <michel.daenzer@amd.com> 7279Date: Thu Jan 10 17:00:16 2013 +0100 7280 7281 DRI2/r[345]xx: 16 bit depth buffers use square micro tiling. 7282 7283 Fixes piglit sanity.tests (with 3D drivers which allocate depth buffers from 7284 the X server). 7285 7286 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 7287 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7288 7289commit f5d2682584c74587890a9795ece7a8917214a83f 7290Author: Michel Dänzer <michel.daenzer@amd.com> 7291Date: Thu Jan 10 17:00:15 2013 +0100 7292 7293 DRI2: Properly use buffer depth passed in by client. 7294 7295 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 7296 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7297 7298commit 4e35b2f530e2ca8c7b7220cacd05c661de43d20d 7299Author: Michel Dänzer <michel.daenzer@amd.com> 7300Date: Thu Jan 10 12:10:52 2013 +0100 7301 7302 Drop support for X servers older than 1.7. 7303 7304 The main purpose is to drop DRI2 compatibility code paths which are getting 7305 awkward to deal with and at this point are probably only tested lightly if at 7306 all. xserver 1.7 was released more than 3 years ago, and 1.6.2 was already 7307 required before. 7308 7309 Mostly mechanical, there's probably potential for more cleanups. 7310 7311 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 7312 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7313 7314commit 857d729eb3672443f89c0516f8226eacb37dad2b 7315Author: Dave Airlie <airlied@redhat.com> 7316Date: Mon Jan 7 14:13:33 2013 +1000 7317 7318 radeon: fix damage reporting for slave pixmaps 7319 7320 This is definitely more correct, not sure if it'll fix any bugs. 7321 7322 Signed-off-by: Dave Airlie <airlied@redhat.com> 7323 7324commit 6981a5c087165b126c15ba0025cffdba218ab652 7325Author: Ilija Hadzic <ihadzic@research.bell-labs.com> 7326Date: Wed Dec 19 10:35:43 2012 -0500 7327 7328 DRI2: limit the swap rate when CRTC is in DPMS-off state 7329 7330 If drawable is displayed on a CRTC and relevant CRTC is in 7331 DPMS off state, defer the swap by a fixed (hard-coded) time. 7332 7333 This patch fixes a bug that caused an application to render 7334 at uncontrolled rate when CRTC goes into DPMS "off" state, 7335 thus thrashing the GPU and CPU and likely offsetting the 7336 power savings achieved by shutting off the display. 7337 7338 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 7339 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 7340 7341commit 3657672207322be651cdb94a811337b7c5668c84 7342Author: Ilija Hadzic <ihadzic@research.bell-labs.com> 7343Date: Wed Dec 19 10:35:42 2012 -0500 7344 7345 DRI2: change signature of radeon_dri2_drawable_crtc 7346 7347 Return pointer to the CRTC instead of CRTC ID and expose 7348 consider_disabled option in arguments of this function. 7349 7350 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 7351 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 7352 7353commit 61d0aec40e2521488c2fe43e7a6823e5c87d94d7 7354Author: Ilija Hadzic <ihadzic@research.bell-labs.com> 7355Date: Wed Dec 19 10:35:41 2012 -0500 7356 7357 video: add option to include disabled CRTCs in best CRTC search 7358 7359 This patch adds an option called consider_disabled to 7360 radeon_pick_best_crtc function. If this option is set 7361 and searching for best-fit CRTC yields nothing, the search 7362 will be widened to include CRTCs in DPMS "off" state. 7363 The new option is not used yet; it will be in the patches 7364 to follow. 7365 7366 v2: Do not look at the CRTC mode, we only care about its box. 7367 7368 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 7369 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 7370 7371commit a60d2152e928a7011fc7c44a885a34c3cdd4f0fe 7372Author: Michel Dänzer <michel.daenzer@amd.com> 7373Date: Thu Dec 27 12:13:57 2012 +0100 7374 7375 glamor: Enable full 2D acceleration for SI as well. 7376 7377 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7378 7379commit 04ebe14fe775d18b534e89b50b90ee293ae1910e 7380Author: Michel Dänzer <michel.daenzer@amd.com> 7381Date: Thu Dec 27 11:54:27 2012 +0100 7382 7383 glamor: Release the drawable after passing to glamor_push_pixels 7384 7385 Ported from xf86-video-intel commit ca5c028c2b4d9bf02002acd484054fe427ea8d09 . 7386 7387 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7388 7389commit cee7d61b144162974238b282d76a70aa2f14cc0d 7390Author: Alex Deucher <alexdeucher@gmail.com> 7391Date: Tue Dec 18 15:49:47 2012 -0500 7392 7393 exa/Xv: fix coordinate limits on AVIVO IGPs 7394 7395 RS6xx asics are r4xx derived, but seem to have r3xx 7396 limitations as far as clipping is concerned. Spotted 7397 by Michel on IRC. 7398 7399 Fixes: 7400 https://bugs.freedesktop.org/show_bug.cgi?id=58469 7401 7402 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 7403 7404commit 3c7d024c2671dac541b8f2daed55040f4fd5d62d 7405Author: Michel Dänzer <michel.daenzer@amd.com> 7406Date: Wed Dec 12 10:43:55 2012 +0100 7407 7408 Remove dead option definitions. 7409 7410 These were only used with UMS. 7411 7412 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7413 7414commit bdbd84a6c99943f4e012c92ad9cb920cbd676621 7415Author: Michel Dänzer <michel.daenzer@amd.com> 7416Date: Mon Dec 10 23:35:41 2012 +0100 7417 7418 Remove local variables total_size_bytes. 7419 7420 Not actually used for anything. Refactoring gone wrong? 7421 7422 Also remove a comment that doesn't seem to make sense anymore. 7423 7424 Stumbled over this due to a warning by clang: 7425 7426 ../../src/radeon_kms.c:1562:6: warning: variable 'total_size_bytes' is uninitialized when used here [-Wuninitialized] 7427 total_size_bytes += (64 * 4 * 64); 7428 ^~~~~~~~~~~~~~~~ 7429 ../../src/radeon_kms.c:1557:25: note: initialize the variable 'total_size_bytes' to silence this warning 7430 int total_size_bytes; 7431 ^ 7432 = 0 7433 7434 Reviewed-by: Alex Deucher <aleander.deucher@amd.com> 7435 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7436 7437commit cf0a81547bcde32bdd2b080c0a546a49dd2eb628 7438Author: Michel Dänzer <michel.daenzer@amd.com> 7439Date: Tue Dec 11 11:48:33 2012 +0100 7440 7441 Remove unused local variable 'path'. 7442 7443 ../../src/radeon_probe.c: In function 'radeon_platform_probe': 7444 ../../src/radeon_probe.c:270:11: error: unused variable 'path' [-Werror=unused-variable] 7445 7446 Reviewed-by: Alex Deucher <aleander.deucher@amd.com> 7447 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7448 7449commit d268038a5bdd1655f7533c8dd2edf74fbc82bbee 7450Author: Michel Dänzer <michel.daenzer@amd.com> 7451Date: Mon Jul 2 09:07:14 2012 +0200 7452 7453 Remove dead code flagged by gcc -Wunused-but-set-variable. 7454 7455 Reviewed-by: Alex Deucher <aleander.deucher@amd.com> 7456 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7457 7458commit 793e1b0ea579cce7489aee57a95f20384d6ecce8 7459Author: Alex Deucher <alexdeucher@gmail.com> 7460Date: Thu Dec 6 09:21:19 2012 -0500 7461 7462 radeon: fix ring count in R300PrepareTexturedVideo 7463 7464 Fixes: 7465 https://bugs.freedesktop.org/show_bug.cgi?id=31364 7466 7467 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 7468 7469commit f402805b22e4f4f735924d59c69474b24e623a25 7470Author: Jerome Glisse <jglisse@redhat.com> 7471Date: Fri Nov 30 10:07:51 2012 -0500 7472 7473 radeon: avoid copying over itself ddx fb 7474 7475 In some rare case it seems that the old fb could already be the 7476 ddx fb. As copying from the same bo will trigger a ttm deadlck 7477 detection which will result in infinite kernel/userspace loop. 7478 Avoid doing any copy as anyway it's useless. 7479 7480 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 7481 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 7482 7483commit 60cd6ceaf44b506433aebf6b3a639a17604dfddd 7484Author: Alex Deucher <alexdeucher@gmail.com> 7485Date: Wed Nov 21 18:42:56 2012 -0500 7486 7487 radeon: add new SI pci id 7488 7489 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 7490 7491commit 53af6aa99dfb67b5223c28d5129a79694165048d 7492Author: Jerome Glisse <jglisse@redhat.com> 7493Date: Wed Nov 14 12:15:47 2012 -0500 7494 7495 radeon: avoid segfault when pixmap exceed GPU capabilities 7496 7497 We might get a request for a pixmap/drawable that is too big 7498 for GPU capabilities in the dri2 get buffer path. In face of 7499 such things just return NULL to dri2 get buffer request. The 7500 GL driver should then use something like transparent black to 7501 avoid something too ugly on the screen. 7502 7503 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 7504 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 7505 7506commit 71eeb0833489ece745f1e5dc41379ec63161ee81 7507Author: Alex Deucher <alexdeucher@gmail.com> 7508Date: Tue Nov 6 09:43:55 2012 -0500 7509 7510 bump verion post release 7511 7512 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 7513 7514commit cf377c991f4c1cd4eee1a8530ba9fdf6f9f63301 7515Author: Alex Deucher <alexdeucher@gmail.com> 7516Date: Tue Nov 6 09:12:39 2012 -0500 7517 7518 radeon: bump version for release 7519 7520 7.0.0, RIP UMS 7521 7522 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 7523 7524commit 20bfc652ce40008ea561db2984bccf137409c7fd 7525Author: Ilija Hadzic <ihadzic@research.bell-labs.com> 7526Date: Mon Sep 24 17:19:09 2012 -0400 7527 7528 radeon/radeon_platform_probe: fix Zaphod mode breakage 7529 7530 Using radeon_platform_probe function breaks the Zaphod mode because 7531 it attempts to call xf86AddEntityToScreen multiple times, but nobody 7532 calls xf86SetEntityShared prior to that. Consequently, calls for all 7533 but first device instance fail. 7534 7535 Prior to introduction of platform bus, the logic was that the Probe 7536 function would make the entity sharable, which would cause Xserver 7537 to later make it shared prior to adding it to screen. With the 7538 platform bus loading, add to screen happens in the probe 7539 function so we have to make it shared there. 7540 7541 v2: do not make the entity shared if it was not previously marked 7542 sharable. Should fix: 7543 7544 https://bugs.freedesktop.org/show_bug.cgi?id=56680 7545 https://bugs.freedesktop.org/show_bug.cgi?id=56663 7546 7547 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 7548 7549commit b2807b53083af5db0b85645666b0127be16623c7 7550Author: Alex Deucher <alexdeucher@gmail.com> 7551Date: Fri Nov 2 17:57:18 2012 -0400 7552 7553 man: note that 2D tiling requires Mesa 9.0 or newer 7554 7555 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 7556 7557commit 1038e29162d869e2ce37f9b8f5f7c8ea37c3e53d 7558Author: Alex Deucher <alexdeucher@gmail.com> 7559Date: Fri Nov 2 10:04:46 2012 -0400 7560 7561 Revert "radeon/radeon_platform_probe: fix Zaphod mode breakage" 7562 7563 This reverts commit 39eac3104c2f08b4d78aab3f88fd104301eb4711. 7564 7565 This breaks non-zaphod configurations with xserver 1.13. 7566 See: 7567 https://bugs.freedesktop.org/show_bug.cgi?id=56680 7568 https://bugs.freedesktop.org/show_bug.cgi?id=56663 7569 7570commit d748894c84e891662cde6d7a5748f74a5aa5175e 7571Author: Alex Deucher <alexdeucher@gmail.com> 7572Date: Fri Nov 2 08:50:16 2012 -0400 7573 7574 man: update man pages for trinity (ARUBA) 7575 7576 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 7577 7578commit 6740e8e90b934611c3bb0f9e8732d66e25b8c727 7579Author: Andreas Boll <andreas.boll.dev@gmail.com> 7580Date: Fri Nov 2 13:30:44 2012 +0100 7581 7582 radeon/man: ColorTiling2D is by default on for r6xx-cayman 7583 7584 Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com> 7585 7586commit 39eac3104c2f08b4d78aab3f88fd104301eb4711 7587Author: Ilija Hadzic <ihadzic@research.bell-labs.com> 7588Date: Mon Sep 24 17:41:32 2012 -0400 7589 7590 radeon/radeon_platform_probe: fix Zaphod mode breakage 7591 7592 Using radeon_platform_probe function breaks the Zaphod mode because 7593 it attempts to call xf86AddEntityToScreen multiple times, but nobody 7594 calls xf86SetEntityShared prior to that. Consequently, calls for all 7595 but first device instance fail. 7596 7597 Prior to introduction of platform bus, the logic was that the Probe 7598 function would make the entity sharable, which would cause Xserver 7599 to later make it shared prior to adding it to screen. With the 7600 platform bus loading, add to screen happens in the probe 7601 function so we have to make it shared there. 7602 7603 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 7604 7605commit 2e96aa641bc8a5c263a5b643961383cdf530d044 7606Author: Marek Olšák <maraeo@gmail.com> 7607Date: Sun Oct 28 12:50:49 2012 -0400 7608 7609 radeon: turn on 2D tiling by default on r6xx-cayman 7610 7611 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 7612 7613commit e830c039479539735d0836082abef24f6841c803 7614Author: Marek Olšák <maraeo@gmail.com> 7615Date: Sun Oct 28 12:49:15 2012 -0400 7616 7617 radeon: log if 2D tiling is enabled 7618 7619 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 7620 7621commit ce1b745dcb60dc516ad999756240b78e72a7aa54 7622Author: Alex Deucher <alexander.deucher@amd.com> 7623Date: Tue Oct 16 13:05:46 2012 -0400 7624 7625 radeon: add new pci ids for SI 7626 7627 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 7628 7629commit e3519be285bdab50ef2193f98826a351a4414fae 7630Author: Dave Airlie <airlied@redhat.com> 7631Date: Tue Oct 16 16:09:23 2012 +1000 7632 7633 radeon/dri2: fix the window->pixmap math properly. 7634 7635 Okay kwin did things different, and I really hope this time this is the 7636 correct transformation operation. It works here with gears under kwin 7637 and under mutter and under xcompmgr. 7638 7639 Signed-off-by: Dave Airlie <airlied@redhat.com> 7640 7641commit bd9e2c064d93afbe4ab84febb2a43ae6a95ab17b 7642Author: Dave Airlie <airlied@redhat.com> 7643Date: Thu Oct 11 14:36:11 2012 +1000 7644 7645 radeon/dri2: fix damage reporting on pageflip 7646 7647 when you use MPX or slave usb devices, the damage posting was incorrect, 7648 and resulted in misrendered boxes on the screen. 7649 7650 Fix it by reporting damage correctly. 7651 7652 Signed-off-by: Dave Airlie <airlied@redhat.com> 7653 7654commit e8cb0b721e6ea251f85c799ca0563bfa59a2d37c 7655Author: Adam Jackson <ajax@redhat.com> 7656Date: Wed Sep 26 09:27:54 2012 -0400 7657 7658 Remove mibstore.h 7659 7660 Signed-off-by: Adam Jackson <ajax@redhat.com> 7661 7662commit 8637f772347c958fa3beea6c0dc9c22255db70e5 7663Author: Michel Dänzer <michel.daenzer@amd.com> 7664Date: Thu Sep 13 19:50:35 2012 +0200 7665 7666 glamor: Force acceleration for DRI2 copies. 7667 7668 Improves performance of non-pageflipped 3D apps on SI. 7669 7670 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7671 7672commit 40b87f0e9db2642ca8a70e994fb4d3742e199640 7673Author: Michel Dänzer <michel.daenzer@amd.com> 7674Date: Wed Sep 12 18:44:10 2012 +0200 7675 7676 glamor: Fix DRI2 crash introduced by PRIME changes. 7677 7678 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7679 7680commit 646ffd49d2cf0d3130301e8ad2f49f620bd9ef1b 7681Author: Dave Airlie <airlied@redhat.com> 7682Date: Mon Sep 10 15:31:06 2012 +1000 7683 7684 radeon: fix build against old servers 7685 7686 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=54711 7687 7688 Reported-by: wolput@onsneteindhoven.nl 7689 Signed-off-by: Dave Airlie <airlied@redhat.com> 7690 7691commit 7c7f2775604da61537c24c249884aaa220e06e13 7692Author: Dave Airlie <airlied@redhat.com> 7693Date: Mon Sep 10 11:27:45 2012 +1000 7694 7695 radeon: don't fail to load if we have no modes and gpu driver. 7696 7697 This ensures radeon loads on output less GPUs as a GPU driver. 7698 7699 There are server bugs that also need to be fixed. 7700 7701 Signed-off-by: Dave Airlie <airlied@redhat.com> 7702 7703commit 7fe16dd037ac688726869816888bd77fe4356f8e 7704Author: Dave Airlie <airlied@gmail.com> 7705Date: Wed Sep 5 10:41:51 2012 +1000 7706 7707 radeon: fix typo in glamor paths 7708 7709 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=54519 7710 7711 Reported-By: scaty@dcinformatique.com 7712 Signed-off-by: Dave Airlie <airlied@redhat.com> 7713 7714commit cdc985e9190ce54dc6274ebc0e87ab954b15bb0d 7715Author: Dave Airlie <airlied@gmail.com> 7716Date: Tue Sep 4 18:18:50 2012 +1000 7717 7718 radeon: fix build against older servers 7719 7720 Another silly one. 7721 7722 Signed-off-by: Dave Airlie <airlied@redhat.com> 7723 7724commit bd0e9c51407cb99f1fe4f0acfedc37ee0c354040 7725Author: Dave Airlie <airlied@gmail.com> 7726Date: Tue Sep 4 18:17:06 2012 +1000 7727 7728 radeon/dri2: fix build against older servers. 7729 7730 Older servers won't have DRI2UpdatePrime, so fix build. 7731 7732 bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=54433 7733 Signed-off-by: Dave Airlie <airlied@redhat.com> 7734 7735commit 75c1f62f44d7b3709c1fa103d59733866967b841 7736Author: Dave Airlie <airlied@redhat.com> 7737Date: Mon Sep 3 13:09:02 2012 +1000 7738 7739 radeon: add capability and provider support 7740 7741 Signed-off-by: Dave Airlie <airlied@redhat.com> 7742 7743commit ca8b61e916998458203dfa4b9d0c997fa7d90ba9 7744Author: Dave Airlie <airlied@redhat.com> 7745Date: Mon Sep 3 13:08:37 2012 +1000 7746 7747 radeon: allocated shared pixmaps as untiled gtt. 7748 7749 Signed-off-by: Dave Airlie <airlied@redhat.com> 7750 7751commit 8c1bf9d8fe3948b72795984e625ef46b2f0bf654 7752Author: Dave Airlie <airlied@redhat.com> 7753Date: Mon Sep 3 13:08:05 2012 +1000 7754 7755 radeon: add pixmap sharing hooks. 7756 7757 This hooks into EXA and the dirty tracking to add sharing and output 7758 offload support. 7759 7760 Signed-off-by: Dave Airlie <airlied@redhat.com> 7761 7762commit 3add8df8122697acfe126d4857f3946ce44b8305 7763Author: Dave Airlie <airlied@redhat.com> 7764Date: Mon Sep 3 13:07:07 2012 +1000 7765 7766 radeon: add dri2 offload support. 7767 7768 This adds the new dri2 interfaces required for radeon to be offloaded to. 7769 7770 Signed-off-by: Dave Airlie <airlied@redhat.com> 7771 7772commit f7502a11c8ef9c453ceb40d26109977116df88c2 7773Author: Dave Airlie <airlied@redhat.com> 7774Date: Mon Sep 3 12:56:56 2012 +1000 7775 7776 radeon: add shared support to pixmaps. 7777 7778 this just adds the interface and shared support to the pixmap. 7779 7780 Signed-off-by: Dave Airlie <airlied@redhat.com> 7781 7782commit f71139a2afe8fffb628331402bf829a6d67c9fff 7783Author: Dave Airlie <airlied@redhat.com> 7784Date: Mon Sep 3 12:37:22 2012 +1000 7785 7786 radeon: add platform bus loading support. 7787 7788 This allows the radeon driver to be loaded via the platform bus mechanism. 7789 7790 Signed-off-by: Dave Airlie <airlied@redhat.com> 7791 7792commit 9911e72ea79e000ee0492d19a434aa98b83b5066 7793Author: Dave Airlie <airlied@redhat.com> 7794Date: Fri Aug 17 17:02:51 2012 +1000 7795 7796 radeon: damage full screen after pageflip 7797 7798 in order for offload to work we need to cause a full 7799 screen damage to be registered after pageflip. 7800 7801 Signed-off-by: Dave Airlie <airlied@redhat.com> 7802 7803commit 0873b8cef21abf1e7286731758f0db1e7f2e7c97 7804Author: Dave Airlie <airlied@redhat.com> 7805Date: Fri Aug 3 16:45:50 2012 +1000 7806 7807 ati: fix stub driver loader to load all drivers properly 7808 7809 Signed-off-by: Dave Airlie <airlied@redhat.com> 7810 7811commit cdf522f5992adf7b76f9153c777e6dddcf7758a6 7812Author: Adam Jackson <ajax@redhat.com> 7813Date: Wed Aug 15 13:38:28 2012 -0400 7814 7815 Implement ->driverFunc 7816 7817 Copied from fbdev, makes it so we can run without iopl. 7818 7819 Signed-off-by: Adam Jackson <ajax@redhat.com> 7820 7821commit 763f387ccac2399f1d58ec112c8bed45196bea5e 7822Author: Adam Jackson <ajax@redhat.com> 7823Date: Wed Aug 15 12:49:10 2012 -0400 7824 7825 Don't bother to include vbe.h 7826 7827 The only VBE code here went away when we dropped UMS. 7828 7829 Signed-off-by: Adam Jackson <ajax@redhat.com> 7830 7831commit aabcf5ea30195274c79ae94c0fe6edc965f5c123 7832Author: Tomas Chvatal <tchvatal@suse.cz> 7833Date: Mon Aug 13 22:58:18 2012 +0200 7834 7835 Add switch for udev to disable automagic detection. 7836 7837 Signed-off-by: Tomas Chvatal <tchvatal@suse.cz> 7838 7839commit aef90993621331f01d17b95a751d4924d77eed0c 7840Author: Alex Deucher <alexdeucher@gmail.com> 7841Date: Mon Aug 6 11:00:45 2012 -0400 7842 7843 radeon: add some new SI pci ids 7844 7845 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 7846 7847commit 6ef1ad6a46348d3aecd8d1f5e94431ca2298853c 7848Author: Michel Dänzer <michel.daenzer@amd.com> 7849Date: Fri Jul 13 11:15:25 2012 +0200 7850 7851 Deal more gracefully with DRI2 being unavailable at build or run time. 7852 7853 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7854 7855commit ef8a404391036d8aa814dbda2407c789b8a64b92 7856Author: Michel Dänzer <michel.daenzer@amd.com> 7857Date: Thu Jul 5 20:14:48 2012 +0200 7858 7859 Initial SI support. 7860 7861 Defaults to shadowfb. 3D acceleration is available with glamor. 2D 7862 acceleration is disabled until the radeonsi driver can handle glamor's 7863 shaders. 7864 7865 v2: add chip flags (Alex Deucher) 7866 7867 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7868 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 7869 7870commit e9edd2f5002c642b59f028b3ec076d604ae8ce9d 7871Author: Michel Dänzer <michel.daenzer@amd.com> 7872Date: Wed Jun 20 08:40:07 2012 +0200 7873 7874 Initial glamor support. 7875 7876 Enable at build time with --enable-glamor and runtime with 7877 7878 Option "AccelMethod" "glamor" 7879 7880 The most notable lack of functionality is XVideo. Use something like VDPAU for 7881 now. 7882 7883 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7884 7885commit 9eac8021f3d33a63156f9f5d43a220e88bb3f8db 7886Author: Michel Dänzer <michel.daenzer@amd.com> 7887Date: Wed Jun 27 14:48:47 2012 +0200 7888 7889 EXA: Factor out pixmap BO allocation into a helper function. 7890 7891 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7892 7893commit 060c7836e7f7777bacca4e23f57c5985beab33bc 7894Author: Michel Dänzer <michel.daenzer@amd.com> 7895Date: Wed Jul 4 11:49:18 2012 +0200 7896 7897 Some cosmetic fixups for the radeon manpage. 7898 7899 ATI -> ATI/AMD 7900 PCIE -> PCIe 7901 7902 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7903 7904commit c3a9f64fe40c78331b9340c3832d344f38b11564 7905Author: Michel Dänzer <michel.daenzer@amd.com> 7906Date: Wed Jul 4 11:46:30 2012 +0200 7907 7908 Drop UMS specific parts from the radeon manpage. 7909 7910 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 7911 7912commit 5ebebfdb4fadada157d1369f35a8610c0fa72e02 7913Author: Dave Airlie <airlied@redhat.com> 7914Date: Fri Jun 29 15:03:43 2012 +0100 7915 7916 radeon: fix conflict with fix in master. 7917 7918 Signed-off-by: Dave Airlie <airlied@redhat.com> 7919 7920commit ff12ddf613ad42e4566d6e10fc6cf33a4e445736 7921Author: Dave Airlie <airlied@redhat.com> 7922Date: Fri Jun 29 15:00:44 2012 +0100 7923 7924 configure: bump to 6.99.99 7925 7926 this is in prep for 7.0 release. 7927 7928 Signed-off-by: Dave Airlie <airlied@redhat.com> 7929 7930commit 50689ec8dbd4a68527b2ac16cecac298b8d441d0 7931Merge: 6a60fcdf0 ae682a6a1 7932Author: Dave Airlie <airlied@redhat.com> 7933Date: Fri Jun 29 14:59:47 2012 +0100 7934 7935 Merge remote-tracking branch 'origin/kms-only' 7936 7937 This merges the removal of the UMS code and subsequent simplifications. 7938 7939 Hell yes. 7940 7941 * origin/kms-only: (24 commits) 7942 radeon: fix radeonchipsets properly 7943 radeon: fix RADEONChipsets 7944 radeon: migrate remainder of radeon_driver.c to rest of driver 7945 radeon: avoid including large static struct twice. 7946 radeon: drop some more unused macros/inlines 7947 radeon: drop another unused struct member. 7948 radeon: more unused stuff 7949 radeon: drop cardType shouldn't matter to userspace. 7950 radeon: drop radeonGetPixmapOffset 7951 radeon: move more functions into texture video 7952 radeon: drop legacy memory, merge into radeon video 7953 radeon: drop unused txoffset vars. 7954 radeon: drop vb_mc_addr, not needed anymore 7955 radeon: drop offset member of r600 accel object 7956 radeon: move macros into radeon.h 7957 radeon: drop more wrapper macros. 7958 radeon: further macro cleanups. 7959 radeon: make exa copy funcs static. 7960 radeon: drop FUNC_NAME macro and ONCE_ONLY stuff 7961 radeon: drop radeon_commonfuncs. 7962 ... 7963 7964 Conflicts: 7965 configure.ac 7966 7967commit 6a60fcdf060a0f553d8d4f7939c2a05aa04bedfa 7968Author: Dave Airlie <airlied@redhat.com> 7969Date: Fri Jun 29 14:57:58 2012 +0100 7970 7971 configure: bump version after release. 7972 7973 Signed-off-by: Dave Airlie <airlied@redhat.com> 7974 7975commit 008eb209d506ed7195fd544dfa91be9783946fd5 7976Author: Dave Airlie <airlied@redhat.com> 7977Date: Fri Jun 29 14:46:55 2012 +0100 7978 7979 configure: bump version number to 6.14.6 pre-release 7980 7981 Signed-off-by: Dave Airlie <airlied@redhat.com> 7982 7983commit 72fea2635a56c4f572c07fc50fc58f703a21eb4f 7984Author: Dave Airlie <airlied@redhat.com> 7985Date: Fri Jun 29 14:46:17 2012 +0100 7986 7987 radeon: require libdrm 2.4.36 for KMS support. 7988 7989 This is due to some commits to the surface manager that fix bugs. 7990 7991 Signed-off-by: Dave Airlie <airlied@redhat.com> 7992 7993commit 179b035835bfa99283d44ba2d5ee17e686196535 7994Author: Michel Dänzer <michel.daenzer@amd.com> 7995Date: Wed Jun 27 19:26:51 2012 +0200 7996 7997 Fix up displayWidth vs. virtualX confusion in radeon_setup_kernel_mem(). 7998 7999 It was using the pitch (displayWidth) for the virtual width (virtualX). This 8000 prevented using page flipping in some cases, as displayWidth was already 8001 overaligned for virtualX, so the DRI2 front and back buffers ended up having 8002 a different pitch. 8003 8004 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8005 Reviewed-by: Jerome Glisse <jglisse@redhat.com> 8006 8007commit 5dc5012320536f7b1e86a81ed3b823e3626828e1 8008Author: Dave Airlie <airlied@redhat.com> 8009Date: Mon Jun 25 09:23:04 2012 +0100 8010 8011 radeon: dixPrivateKeyRegistered is only in server 1.9 8012 8013 just fallback to the old behaviour on older servers. 8014 8015 should fix: https://bugs.freedesktop.org/show_bug.cgi?id=51388 8016 Review-over-irc: Michel Dänzer <michel.daenzer@amd.com> 8017 Signed-off-by: Dave Airlie <airlied@redhat.com> 8018 8019commit ae682a6a1781ab023ab36e3154d6b3e88da63d64 8020Author: Dave Airlie <airlied@redhat.com> 8021Date: Mon Jun 18 13:11:31 2012 +0100 8022 8023 radeon: fix radeonchipsets properly 8024 8025commit 6514bcdc0fc199b4f3a8b4273bd29e0fe255d232 8026Author: Dave Airlie <airlied@redhat.com> 8027Date: Mon Jun 18 13:02:03 2012 +0100 8028 8029 radeon: fix RADEONChipsets 8030 8031 reported on irc by okias. 8032 8033 Signed-off-by: Dave Airlie <airlied@redhat.com> 8034 8035commit dd4ab5aba9047d522dbcfe8a341c368312e1a471 8036Author: Dave Airlie <airlied@redhat.com> 8037Date: Sat Jun 16 08:47:32 2012 +0100 8038 8039 radeon: migrate remainder of radeon_driver.c to rest of driver 8040 8041 This helps make a few more things static and the driver generally 8042 smaller. 8043 8044 Signed-off-by: Dave Airlie <airlied@redhat.com> 8045 8046commit a19237fad2492d70777167c631f4e6e1fae0e908 8047Author: Dave Airlie <airlied@redhat.com> 8048Date: Sat Jun 16 08:19:35 2012 +0100 8049 8050 radeon: avoid including large static struct twice. 8051 8052 Just extern this in the one file that needs access. 8053 8054 Signed-off-by: Dave Airlie <airlied@redhat.com> 8055 8056commit c8a5c5a0dd9adc46b03b898c9c1719dbbde05ba1 8057Author: Dave Airlie <airlied@redhat.com> 8058Date: Fri Jun 15 19:52:14 2012 +0100 8059 8060 radeon: drop some more unused macros/inlines 8061 8062 These are no longer used anywhere. 8063 8064 Signed-off-by: Dave Airlie <airlied@redhat.com> 8065 8066commit 3097f273443b5b8fce097e2b1e09a17b02d06a40 8067Author: Dave Airlie <airlied@redhat.com> 8068Date: Fri Jun 15 19:45:15 2012 +0100 8069 8070 radeon: drop another unused struct member. 8071 8072 Signed-off-by: Dave Airlie <airlied@redhat.com> 8073 8074commit cb97e75acab84b67b7b7358860788638efc9b344 8075Author: Dave Airlie <airlied@redhat.com> 8076Date: Fri Jun 15 19:38:45 2012 +0100 8077 8078 radeon: more unused stuff 8079 8080 Remove all CurrentLayout stuff. 8081 8082 Signed-off-by: Dave Airlie <airlied@redhat.com> 8083 8084commit 81593e7deb688fa3108a0589c1418459ec0df4de 8085Author: Dave Airlie <airlied@redhat.com> 8086Date: Fri Jun 15 18:52:52 2012 +0100 8087 8088 radeon: drop cardType shouldn't matter to userspace. 8089 8090 Signed-off-by: Dave Airlie <airlied@redhat.com> 8091 8092commit 997b0dddfeb180043ece5eff0ac0d6a1a751f9b5 8093Author: Dave Airlie <airlied@redhat.com> 8094Date: Fri Jun 15 18:43:01 2012 +0100 8095 8096 radeon: drop radeonGetPixmapOffset 8097 8098 no longer needed. 8099 8100 Signed-off-by: Dave Airlie <airlied@redhat.com> 8101 8102commit 639061edb3ac35a24e4d6633c9b6384434dcac64 8103Author: Dave Airlie <airlied@redhat.com> 8104Date: Fri Jun 15 18:38:48 2012 +0100 8105 8106 radeon: move more functions into texture video 8107 8108 make more thing static. 8109 8110 Signed-off-by: Dave Airlie <airlied@redhat.com> 8111 8112commit 85711a0a6a553232b603dbbb2b2d234c35391578 8113Author: Dave Airlie <airlied@redhat.com> 8114Date: Fri Jun 15 17:26:50 2012 +0100 8115 8116 radeon: drop legacy memory, merge into radeon video 8117 8118 This merges these two functions into textured video code. 8119 8120 Signed-off-by: Dave Airlie <airlied@redhat.com> 8121 8122commit 445ca1d66ce7b37762731fe67407cc1b978bb129 8123Author: Dave Airlie <airlied@redhat.com> 8124Date: Fri Jun 15 17:13:48 2012 +0100 8125 8126 radeon: drop unused txoffset vars. 8127 8128 These vars aren't being used anymore. 8129 8130 Signed-off-by: Dave Airlie <airlied@redhat.com> 8131 8132commit 832244de9e6d7aa1b8fbeb29c9a6a86923744d60 8133Author: Dave Airlie <airlied@redhat.com> 8134Date: Fri Jun 15 17:09:00 2012 +0100 8135 8136 radeon: drop vb_mc_addr, not needed anymore 8137 8138 This field is totally unused now, so drop it. 8139 8140 Signed-off-by: Dave Airlie <airlied@redhat.com> 8141 8142commit 6559b6e28bf111b64c3aeb5ec1260acfeb3e0b7f 8143Author: Dave Airlie <airlied@redhat.com> 8144Date: Fri Jun 15 17:05:54 2012 +0100 8145 8146 radeon: drop offset member of r600 accel object 8147 8148 This is always 0 now. 8149 8150 Signed-off-by: Dave Airlie <airlied@redhat.com> 8151 8152commit 857179e538c9752fab239a21312d0fcc80022151 8153Author: Dave Airlie <airlied@redhat.com> 8154Date: Fri Jun 15 16:55:23 2012 +0100 8155 8156 radeon: move macros into radeon.h 8157 8158 Drop separate header file, there isn't many separate macros now. 8159 8160 Signed-off-by: Dave Airlie <airlied@redhat.com> 8161 8162commit 148e889149f5ecc3dc2473f2a9e01572eeb037d3 8163Author: Dave Airlie <airlied@redhat.com> 8164Date: Fri Jun 15 16:52:24 2012 +0100 8165 8166 radeon: drop more wrapper macros. 8167 8168 Signed-off-by: Dave Airlie <airlied@redhat.com> 8169 8170commit ad8629e3323efe484dd1421d94dd934e1bfafe56 8171Author: Dave Airlie <airlied@redhat.com> 8172Date: Fri Jun 15 16:49:13 2012 +0100 8173 8174 radeon: further macro cleanups. 8175 8176 drop some more of the wrapper macros. 8177 8178 Signed-off-by: Dave Airlie <airlied@redhat.com> 8179 8180commit 96644f96f1a2df1d7d1ca8cbdc9a105affdb4f32 8181Author: Dave Airlie <airlied@redhat.com> 8182Date: Fri Jun 15 16:34:53 2012 +0100 8183 8184 radeon: make exa copy funcs static. 8185 8186 These are no longer used outside this area. 8187 8188 Signed-off-by: Dave Airlie <airlied@redhat.com> 8189 8190commit 83904b820b81d226940ec6d663246f243f92e7ba 8191Author: Dave Airlie <airlied@redhat.com> 8192Date: Fri Jun 15 16:30:41 2012 +0100 8193 8194 radeon: drop FUNC_NAME macro and ONCE_ONLY stuff 8195 8196 Drop all the CP vs MMIO macros that name functions. 8197 8198 Signed-off-by: Dave Airlie <airlied@redhat.com> 8199 8200commit b422d8085aab4e01a1ac4ffb247c73fe51aafa0d 8201Author: Dave Airlie <airlied@redhat.com> 8202Date: Fri Jun 15 16:14:54 2012 +0100 8203 8204 radeon: drop radeon_commonfuncs. 8205 8206 modify all the macros and drop it into radeon_accel.c 8207 8208 Signed-off-by: Dave Airlie <airlied@redhat.com> 8209 8210commit 29dd3f84eecd2892a438f004403521de8986a91d 8211Author: Dave Airlie <airlied@redhat.com> 8212Date: Fri Jun 15 16:11:03 2012 +0100 8213 8214 radeon: misc cleanups. 8215 8216 Signed-off-by: Dave Airlie <airlied@redhat.com> 8217 8218commit 678c1f3425c0fa7e98795c8e351da77d04652e02 8219Author: Dave Airlie <airlied@redhat.com> 8220Date: Fri Jun 15 16:06:27 2012 +0100 8221 8222 make pci chipsets static 8223 8224 Signed-off-by: Dave Airlie <airlied@redhat.com> 8225 8226commit 95cbbe999418c72dc407069d31e759d70f49ab16 8227Author: Dave Airlie <airlied@redhat.com> 8228Date: Fri Jun 15 16:01:53 2012 +0100 8229 8230 drop RING_LOCALS/ACCEL_PREAMBLE 8231 8232 these were unused now. 8233 8234 Signed-off-by: Dave Airlie <airlied@redhat.com> 8235 8236commit 18d5ae3bd9075ac1a2ee21b071ac133e2e634b62 8237Author: Dave Airlie <airlied@redhat.com> 8238Date: Fri Jun 15 10:05:03 2012 +0100 8239 8240 radeon: drop all UMS/DRI1/XAA/overlay support. 8241 8242 This overhauls the radeon driver and removes all the old UMS-only code, 8243 it drops all the UMS, DRI1, XAA, overlay Xv, video capture, tv tuners 8244 8245 There are probably a lot more cleanups that will fall out of this afterwards. 8246 8247 So far this is compile/build tested. 8248 8249 Signed-off-by: Dave Airlie <airlied@redhat.com> 8250 8251commit 248e912c487636d7352cfad43c03fc9f19fc2215 8252Author: Michel Dänzer <michel.daenzer@amd.com> 8253Date: Fri Jun 8 12:18:48 2012 +0200 8254 8255 UMS: Fix CRTC DPMS state check. 8256 8257 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8258 8259commit 58883711136fa6671d4f3250d4ee973e2953ebb4 8260Author: Alex Deucher <alexdeucher@gmail.com> 8261Date: Thu Jun 7 11:47:21 2012 -0400 8262 8263 configure: bump version post release 8264 8265 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8266 8267commit 3a9103f855c67bbcd7b97ee2e96836353e0d789d 8268Author: Alex Deucher <alexdeucher@gmail.com> 8269Date: Thu Jun 7 11:27:08 2012 -0400 8270 8271 configure: bump for release 8272 8273 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8274 8275commit f1693ee09e9b251eb75e3e5fe0ee7699fec6fb98 8276Author: Alex Deucher <alexdeucher@gmail.com> 8277Date: Thu Jun 7 11:23:31 2012 -0400 8278 8279 configure: bump libdrm_radeon requirement 8280 8281 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8282 8283commit 77a056c1725c9c44a328ed324687bdf35144e9e2 8284Author: Michel Dänzer <michel.daenzer@amd.com> 8285Date: Wed Jun 6 12:29:37 2012 +0200 8286 8287 Don't page-flip or wait on a CRTC while we're VT-switched away. 8288 8289 We don't know what the CRTC state is. 8290 8291 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8292 8293commit 1d9f1eb7e65c57c52759ccbfa51e8bdd5a10acac 8294Author: Michel Dänzer <michel.daenzer@amd.com> 8295Date: Wed Jun 6 12:16:08 2012 +0200 8296 8297 Don't wait on a CRTC which has been disabled via DPMS (bug #49761). 8298 8299 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=49761 . 8300 8301 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8302 8303commit 4603285aa8efaf2614f15a38254465cec2075f11 8304Author: Alex Deucher <alexdeucher@gmail.com> 8305Date: Tue Jun 5 09:48:59 2012 -0400 8306 8307 radeon: add new PCI ids 8308 8309 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8310 8311commit 9307609420b4b209767d2057b4803bcb16cc1455 8312Author: Dave Airlie <airlied@redhat.com> 8313Date: Sat Jun 2 17:16:34 2012 +0100 8314 8315 radeon: improve smooth startup fallbacks. 8316 8317 If we can't handover the framebuffer, memset it to black. 8318 8319 mostly ported from nouveau. 8320 8321 Signed-off-by: Dave Airlie <airlied@redhat.com> 8322 8323commit 89a4c79a43a2b6963e41d7812e8fe5beedb556af 8324Author: Dave Airlie <airlied@redhat.com> 8325Date: Sat Jun 2 17:14:14 2012 +0100 8326 8327 radeon: fix smooth startup with tiling enabled. 8328 8329 We need to use the surface we worked out when we allocated the front bo, 8330 not work out a new surface from scratch. 8331 8332 This fixes smooth handover from plymouth to gdm on F17. 8333 8334 Signed-off-by: Dave Airlie <airlied@redhat.com> 8335 8336commit c1b9b2c9d71a87a0b98ccdb542da8a30c8c4d99a 8337Author: Dave Airlie <airlied@redhat.com> 8338Date: Fri Jun 1 12:22:06 2012 +0100 8339 8340 radeon: fix compat api for newest planned API. 8341 8342 Signed-off-by: Dave Airlie <airlied@redhat.com> 8343 8344commit 68c51d148b22ee8cbe98b5d5441e3494804ee808 8345Author: Dave Airlie <airlied@redhat.com> 8346Date: Fri Jun 1 12:21:34 2012 +0100 8347 8348 drmmode: drop flags arg to adjust frame. 8349 8350 Signed-off-by: Dave Airlie <airlied@redhat.com> 8351 8352commit b1e9c3081d2b975682a62389aa29063b79f9655c 8353Author: Dave Airlie <airlied@redhat.com> 8354Date: Thu May 24 17:54:41 2012 +0100 8355 8356 radeon: move include down further 8357 8358 fixes tinderbox build. 8359 8360 Signed-off-by: Dave Airlie <airlied@redhat.com> 8361 8362commit 5ed959585f35a4eff2cd803ae222b46d164f210d 8363Author: Dave Airlie <airlied@redhat.com> 8364Date: Thu May 24 08:17:11 2012 +0100 8365 8366 radeon: update compat-api.h for block handler + enable/disable fb. 8367 8368 This updates the compat stuff for the latest block handler code, 8369 and the enable/disable interface. 8370 8371 Signed-off-by: Dave Airlie <airlied@redhat.com> 8372 8373commit 2de17efdcb502f91fb145130c12be3cb164f03c5 8374Author: Dave Airlie <airlied@redhat.com> 8375Date: Thu May 24 08:16:39 2012 +0100 8376 8377 radeon/xvmc: fix build with new API. 8378 8379 This was missing the compat include. 8380 8381 Signed-off-by: Dave Airlie <airlied@redhat.com> 8382 8383commit a1f47768309aa8b14b13f05ba01bb490024c4cbc 8384Author: Dave Airlie <airlied@redhat.com> 8385Date: Thu May 24 07:55:14 2012 +0100 8386 8387 radeon: fix a few more RHDAtomBiosFunc uses 8388 8389 I missed these in my initial search/replace for some reason. 8390 8391 Signed-off-by: Dave Airlie <airlied@redhat.com> 8392 8393commit c26e40cf60eb6a2bffb5ec7cba1f1bbffcbf1d53 8394Author: Dave Airlie <airlied@redhat.com> 8395Date: Wed May 23 12:12:54 2012 +0100 8396 8397 radeon/atombios: rewrite atombios parser code to not use xf86Screens. 8398 8399 Just pass the ScrnInfoPtr around instead. 8400 8401 Signed-off-by: Dave Airlie <airlied@redhat.com> 8402 8403commit 14122214edabd2a574b66577289a7c8732f4e953 8404Author: Dave Airlie <airlied@redhat.com> 8405Date: Wed May 23 11:57:35 2012 +0100 8406 8407 radeon/generic_bus: stop passing scrnIndex 8408 8409 Just pass a pointer to the screen, removes usage of xf86Screens lookup 8410 8411 Signed-off-by: Dave Airlie <airlied@redhat.com> 8412 8413commit 2101e485480e3ffd6a17ed12f5287b8b1e4e0223 8414Author: Dave Airlie <airlied@redhat.com> 8415Date: Tue May 22 16:17:55 2012 +0100 8416 8417 compat for new server API 8418 8419commit 9d2549057ffced996bfaeb8df4901ef57da636c7 8420Author: Dave Airlie <airlied@redhat.com> 8421Date: Wed May 16 15:45:53 2012 +0100 8422 8423 ati: convert to new screen conversion APIs 8424 8425 The compat header takes care of the old server vs new server. 8426 8427 this commit was autogenerated from util/modular/x-driver-screen-scrn-conv.sh 8428 8429 Signed-off-by: Dave Airlie <airlied@redhat.com> 8430 8431commit 80a45bfc26e20d1d481f13cb15c0720d165a8306 8432Author: Dave Airlie <airlied@redhat.com> 8433Date: Wed May 16 15:42:59 2012 +0100 8434 8435 radeon: add compat-api.h 8436 8437commit 84c3900b5179a8d2cc5f925e94ed5b2bbdee1ad9 8438Author: Dave Airlie <airlied@redhat.com> 8439Date: Wed May 23 08:49:45 2012 +0100 8440 8441 radeon/xaa: drop scrnIndex parameter to some functions 8442 8443 This isn't needed, and makes api changes later easier. 8444 8445 Signed-off-by: Dave Airlie <airlied@redhat.com> 8446 8447commit 3fb694b308ebadd1b849836059b6b56bb19385f7 8448Author: Thierry Vignaud <thierry.vignaud@gmail.com> 8449Date: Wed May 16 14:43:53 2012 +0200 8450 8451 UMS/EXA: Add reminder for potential solid picture performance issue. 8452 8453commit 4b9bad959438725e4434e2aa4f142542d8dcc260 8454Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 8455Date: Fri May 11 05:25:32 2012 +0200 8456 8457 radeon: use GB_GR and BG_RG formats for packed yuv video for r600+ 8458 8459 Those formats were invented for exactly that purpose so use them. 8460 This saves some code and also some hw resources (only need one 8461 sampler instead of two for packed yuv). 8462 Only tested on EG. 8463 8464commit 37786e9027b8c8d1f9ec9928915784dd28853766 8465Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 8466Date: Fri May 11 05:25:33 2012 +0200 8467 8468 radeon: avoid rounding errors in texture coords for textured xv on EG+ 8469 8470 make sure the division is done with floats, otherwise the coordinate 8471 can be wrong up to 1 texel. 8472 Particularly visible with clipping and small source scaled up (since one 8473 texel can be a shift of several pixels) but could be seen even unscaled. 8474 Should provide more accurate coords without clipping too depending on the 8475 scale factor probably. 8476 This is a straight port of 688c8a54a00b01e73a11970ad2abe858f8c7c5c4 8477 when I apparently forgot the eg code... 8478 8479commit faea3aafa8c9a7c1bc866ffcd847972c5569c8bc 8480Author: Anisse Astier <anisse@astier.eu> 8481Date: Thu May 10 17:22:59 2012 +0200 8482 8483 Fail more gracefully when drm surface manager can't be initialized 8484 8485 Should make bugs like https://bugs.freedesktop.org/show_bug.cgi?id=48138 8486 easier to diagnose. 8487 8488 [ Michel Dänzer: Appended newline to error message. ] 8489 8490 Signed-off-by: Anisse Astier <anisse@astier.eu> 8491 Singed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8492 8493commit b0b7d8d26fd107df342b5c87b0a38e5bb08101a9 8494Author: Michel Dänzer <michel.daenzer@amd.com> 8495Date: Wed May 9 11:08:49 2012 +0200 8496 8497 EXA/UMS: Synchronize to the GPU before writing solid colour to scratch pixmap. 8498 8499 UMS doesn't do this automagically. It's a big hammer that will probably suck 8500 for performance, but I don't have any better ideas right now. 8501 8502 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8503 8504commit 6bda7ceda645e838723883d133d614def1511d16 8505Author: Michel Dänzer <michel.daenzer@amd.com> 8506Date: Thu May 3 15:07:30 2012 +0200 8507 8508 EXA: Attempt to fix solid picture acceleration with UMS. 8509 8510 Only compile tested, but should fix 8511 https://bugs.freedesktop.org/show_bug.cgi?id=49182 . 8512 8513 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8514 8515commit b33cf9cf9dce5add8aaccab10ba5aaa64247c8ce 8516Author: Alex Deucher <alexdeucher@gmail.com> 8517Date: Wed May 2 09:55:16 2012 -0400 8518 8519 radeon: add MacModel entry for SAM440ep embedded board 8520 8521 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8522 8523commit 0bda305f7ab2a4720b3fea3f318ab2a73be151e5 8524Author: Alex Deucher <alexander.deucher@amd.com> 8525Date: Fri Apr 13 13:28:10 2012 -0400 8526 8527 EXA: Support acceleration of solid pictures on R2xx. 8528 8529 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 8530 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 8531 8532commit a7754b076e355fef3aea082b6e2d3aefbb8e7a9b 8533Author: Alex Deucher <alexander.deucher@amd.com> 8534Date: Fri Apr 13 13:24:46 2012 -0400 8535 8536 EXA: Support acceleration of solid pictures on R1xx. 8537 8538 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 8539 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 8540 8541commit cac2d7ed54918579418fc762558497d3ec547fad 8542Author: Alex Deucher <alexander.deucher@amd.com> 8543Date: Sat Apr 14 08:53:39 2012 -0400 8544 8545 EXA: Support acceleration of solid pictures on Evergreen/NI. 8546 8547 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 8548 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 8549 8550commit d88b9700137ee407c483f263bb55c77cd6f92fef 8551Author: Michel Dänzer <michel.daenzer@amd.com> 8552Date: Fri Apr 13 09:35:38 2012 +0200 8553 8554 EXA: Support acceleration of solid pictures on R3xx-R7xx. 8555 8556 Allocate 1x1 scratch pixmaps to hold the solid picture colours. 8557 8558 This works around https://bugs.freedesktop.org/show_bug.cgi?id=47266 and might 8559 improve performance in other cases as well. 8560 8561 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8562 8563commit 66b586b9b9cdaf70f0fcd547b5a04f044d848d44 8564Author: Michel Dänzer <michel.daenzer@amd.com> 8565Date: Mon Apr 16 15:33:36 2012 +0200 8566 8567 RADEONCopySwap: Fix RADEON_HOST_DATA_SWAP_16BIT case. 8568 8569 It was the same code as for RADEON_HOST_DATA_SWAP_32BIT. This caused bus errors 8570 on FreeBSD/PPC, but I'm not sure how it could not cause problems anywhere... 8571 8572 Reported-by: Andreas Tobler <andreast@fgznet.ch> 8573 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8574 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 8575 8576commit 1e656cd5973bbcf85b63406638ede0bccd65e28f 8577Author: Michel Dänzer <michel.daenzer@amd.com> 8578Date: Mon Apr 16 15:44:01 2012 +0200 8579 8580 Make radeon_setup_kernel_mem failures more graceful and verbose. 8581 8582 So that bugs like https://bugs.freedesktop.org/show_bug.cgi?id=48138 can be 8583 diagnosed more easily. 8584 8585 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8586 8587commit d282719a9c2fb0ee32830aa75b8dfbb9392954ed 8588Author: Jerome Glisse <jglisse@redhat.com> 8589Date: Wed Apr 4 17:08:30 2012 -0400 8590 8591 r6xx-r9xx: force 1D tiling for buffer with height < 64 8592 8593 Due to some old kernel issue, height is 8 aligned insided the ddx 8594 For buffer with height btw 57 & 63 this lead ddx to believe it can 8595 allocate a 2D tiled surface while mesa will not align height and 8596 will assume 1D tiled leading to disagreement and rendering issue. 8597 This patch force buffer with height < 64 to be 1D tiled. 8598 8599 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 8600 8601commit de2419e5509398328903ca61f4ea80852cba6bc4 8602Author: Alex Deucher <alexdeucher@gmail.com> 8603Date: Thu Mar 29 00:28:39 2012 -0400 8604 8605 configure: bump version post release 8606 8607 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8608 8609commit 9425c50e93903fb64d9e569cfdc1e2c35d16ce25 8610Author: Alex Deucher <alexdeucher@gmail.com> 8611Date: Thu Mar 29 00:19:12 2012 -0400 8612 8613 configure: bump version for release 8614 8615 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8616 8617commit dc18d771713ecd893c7d5833da6e0661093161dc 8618Author: Alex Deucher <alexdeucher@gmail.com> 8619Date: Wed Mar 28 17:32:53 2012 -0400 8620 8621 configure: bump libdrm requirement for TN support 8622 8623 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8624 8625commit 03535904a3e1542b3924d0a062c4b022ca196888 8626Author: Alex Deucher <alexdeucher@gmail.com> 8627Date: Tue Mar 27 09:48:28 2012 -0400 8628 8629 radeon: man page updates 8630 8631 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8632 8633commit 6ed191c33caa33c12c2c6dafcba3a5ab1bf4a02f 8634Author: Alex Deucher <alexdeucher@gmail.com> 8635Date: Tue Mar 20 19:57:53 2012 -0400 8636 8637 radeon/kms: add TN pci ids 8638 8639 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8640 8641commit 55d65fcf33eb383e3fbc7a1d469ab68a70a7ab37 8642Author: Alex Deucher <alexdeucher@gmail.com> 8643Date: Tue Mar 20 19:54:57 2012 -0400 8644 8645 radeon/kms: add support for TN (trinity) APUs 8646 8647 - KMS only 8648 - Includes full EXA/Xv support 8649 8650 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8651 8652commit b5cf9bd693cf4090956add4c33c4fae9c3069a03 8653Author: Marek Olšák <maraeo@gmail.com> 8654Date: Wed Mar 7 11:01:40 2012 -0500 8655 8656 r6xx: initialize SX_MISC 8657 8658 If Mesa set it to 1, the DDX would not render anything = the monitor would 8659 basically freeze. 8660 8661 agd5f: update emit count as well. 8662 8663 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8664 8665commit 355dc4295912c153f5333421594fa90aa119a056 8666Author: Michel Dänzer <michel.daenzer@amd.com> 8667Date: Tue Mar 6 15:52:40 2012 +0100 8668 8669 DRI2: Unreference buffers immediately when event wait info is invalidated. 8670 8671 Deferring this could result in trying to unreference buffers from a previous 8672 server generation, i.e. accessing freed memory. 8673 8674 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8675 Tested-by: Christian König <Christian.koenig@amd.com> 8676 8677commit fe51469b2e02e4d565050bab077985270fb58a9b 8678Author: Michel Dänzer <michel.daenzer@amd.com> 8679Date: Tue Mar 6 15:52:40 2012 +0100 8680 8681 Re-register DRM FD wakeup handler for each server generation. 8682 8683 Fixes hang when trying to use DRI2 swap scheduling after a server reset. 8684 8685 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8686 Tested-by: Christian König <Christian.koenig@amd.com> 8687 8688commit 878454ae8d8e96dd27a19d0b30940d014c4cd7e2 8689Author: Hans Verkuil <hverkuil@xs4all.nl> 8690Date: Fri Feb 24 09:35:39 2012 -0500 8691 8692 Fix ConnectorTable crash in radeon_output.c 8693 8694 The sam440ep PPC board requires a ConnectorTable xorg.conf option, but putting 8695 in that option causes the radeon driver to crash. I finally traced it to a 8696 copy-and-paste bug in radeon_output.c as a result of a major rework in commit 8697 82f12e5a40c1fbcb91910a0f8b725c34fff02aae. 8698 8699 The actual crash occurred in RADEONPrintPortMap(). 8700 8701 Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl> 8702 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8703 8704commit 688c8a54a00b01e73a11970ad2abe858f8c7c5c4 8705Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 8706Date: Sat Feb 18 21:12:34 2012 +0100 8707 8708 radeon: avoid rounding errors in texture coords for textured xv 8709 8710 make sure the division is done with floats, otherwise the coordinate 8711 can be wrong up to 1 texel. 8712 Particularly visible with clipping and small source scaled up (since one 8713 texel can be a shift of several pixels) but could be seen even unscaled. 8714 Should provide more accurate coords without clipping too depending on the 8715 scale factor probably. 8716 Changed for r100-r600, though only tested on r300. 8717 8718commit 2778b56252124ef6f636a493d2e1457b43911c37 8719Author: Jerome Glisse <jglisse@redhat.com> 8720Date: Mon Feb 13 20:42:57 2012 -0500 8721 8722 radeon: r6xx-eg use linear general when using scratch bo 8723 8724 In path where we need to use scratch bo as temporary area, 8725 consider it as linear buffer. Not linear aligned. Fix some 8726 case such as in bugs: 8727 8728 https://bugs.freedesktop.org/show_bug.cgi?id=45827 8729 8730 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 8731 8732commit c66ae235700f5efe64eb168327551b8f1d153c9c 8733Author: Michel Dänzer <michel.daenzer@amd.com> 8734Date: Mon Feb 13 10:43:58 2012 +0100 8735 8736 Handle new xorg_list API. 8737 8738 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=45937 8739 8740 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8741 8742commit 7ff277e22c629308915307bbee96eb25ff77f8b9 8743Author: Alex Deucher <alexdeucher@gmail.com> 8744Date: Fri Feb 10 13:04:59 2012 -0500 8745 8746 radeon: fix crash in drmmode_create_bo_pixmap() 8747 8748 Only init surface on r6xx+. Return NULL rather than 8749 FALSE. 8750 8751 Fixes: 8752 https://bugs.freedesktop.org/show_bug.cgi?id=45829 8753 8754 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8755 8756commit 60b949f34df5db05e0e102cc3daa33469aa50cfc 8757Author: Alex Deucher <alexdeucher@gmail.com> 8758Date: Fri Feb 10 13:11:26 2012 -0500 8759 8760 radeon/kms: reusing fd message is not an error 8761 8762 It's standard behavior. 8763 8764 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8765 8766commit e20284409937d784847339b5d466a95012d85940 8767Author: Michel Dänzer <michel.daenzer@amd.com> 8768Date: Fri Feb 3 12:21:59 2012 +0100 8769 8770 EXA/r6xx+: Only set write domain or read domains, not both. (Bug #43893) 8771 8772 Avoids an accounting bug in libdrm_radeon 2.4.31 or older. 8773 8774 See https://bugs.freedesktop.org/show_bug.cgi?id=43893 8775 8776 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8777 8778commit 5a7f64c5170ca424c9dca739662018e30df13413 8779Author: Michel Dänzer <michel.daenzer@amd.com> 8780Date: Wed Feb 8 11:35:35 2012 +0100 8781 8782 Fix UMS build failure. 8783 8784 And some UMS specific warnings. 8785 8786 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8787 8788commit 87f776b966f4200c97a989536d4b71822ae4c0b3 8789Author: Michel Dänzer <michel.daenzer@amd.com> 8790Date: Wed Feb 8 11:10:27 2012 +0100 8791 8792 Remove unused local variable 'height'. 8793 8794 Pointed out by gcc -Wunused-variable. 8795 8796 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8797 8798commit f63262e2f112a348c45f0dcecd891c8b6d9c5ee8 8799Author: Michel Dänzer <michel.daenzer@amd.com> 8800Date: Wed Feb 8 10:28:45 2012 +0100 8801 8802 evergreen: Initialize source surface member for textured video. 8803 8804 Fixes crash reported by Ole Salscheider on IRC. 8805 8806 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8807 8808commit c8f104d38870f14049402bbc14f662c151caeeef 8809Author: Jerome Glisse <jglisse@redhat.com> 8810Date: Tue Feb 7 15:04:37 2012 -0500 8811 8812 radeon: fix tiling for weird resolution 8813 8814 Should also fix xv for some case. 8815 8816 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 8817 8818commit 615033f2b5e3817e335e9d022fc9fdcf8ac8b11a 8819Author: Jerome Glisse <jglisse@redhat.com> 8820Date: Tue Dec 13 11:08:19 2011 -0500 8821 8822 r600-evergreen: use common surface allocator for tiling v11 8823 8824 Use libdrm common surface code so mesa,ddx have same idea 8825 about tiling surface and what their pitch should be and 8826 the alignment constraint. 8827 8828 v2 fix remaining issue add new option to conditionaly enable 8829 v3 fix fbcon copy and r600 exa copy path 8830 v4 fix non tiled path 2D tiling on GPU >= R600, set it to false 8831 as default 8832 v5 adapt to pixel/element size split of libdrm/radeon 8833 v6 update to properly handle falling back to 1d tiled 8834 v6 final fix to tile split value on evergreen and newer 8835 v7 fix default array mode on r6xx, fix height alignment issue 8836 on evergreen 8837 v8 fix tile split value 8838 v9 add stencil tile split support, simplify dri2 for stencil 8839 with evergreen 8840 v10 Try to fix xv path regarding tiling. Adapt to libdrm API 8841 change. Try to fix case where there is no surface which 8842 means non tiled bo. 8843 v11 check for proper libdrm 8844 8845 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 8846 8847commit 36c190671081967bac6fff48aaf66d67b639a48c 8848Author: Michel Dänzer <michel.daenzer@amd.com> 8849Date: Wed Feb 1 13:21:02 2012 +0100 8850 8851 Fix vline range calculations. 8852 8853 The range passed in is in pixmap coordinates, so the CRTC offset needs to be 8854 added to the clamping limits and subtracted from the clamped range for 8855 pre-AVIVO display engines. 8856 8857 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8858 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 8859 8860commit bb0e1531ac6949d38025d7dcb19234fee33b2acf 8861Author: Michel Dänzer <michel.daenzer@amd.com> 8862Date: Wed Feb 1 13:07:11 2012 +0100 8863 8864 Check for empty vline ranges after clamping. 8865 8866 The clamping could turn a previously non-empty range into an empty one. 8867 8868 Also, start == stop means the range is empty. 8869 8870 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8871 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 8872 8873commit 13b3aed4ef9afbcbaea1dcf0ed1acb162b240a3f 8874Author: Alex Deucher <alexdeucher@gmail.com> 8875Date: Tue Jan 10 09:35:09 2012 -0500 8876 8877 EXA/r6xx+: fix rop setting for overlapping copies 8878 8879 Need to use GXCopy for the src to temp copy, then 8880 the original rop for the temp to dest copy. 8881 8882 Noticed by: Frank Huang 8883 8884 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8885 8886commit eb6d769a087b2ed5952f477fc3f0b0625810a287 8887Author: Egbert Eich <eich@suse.de> 8888Date: Tue Nov 15 18:50:56 2011 +0100 8889 8890 DPMS: Split non-modeset CRTC DPMS function. 8891 8892 RADEONRestore() calls crtc->funcs->dpms() after most of the mode setting 8893 subsystems have been restored. This function enables the CRTCs but does 8894 more: it calls DRM pre- and post-modeset ioctls and sets up the palettes 8895 (LUTs). 8896 None of these two things are needed. Accessing the palette registers after 8897 restoring the PLLs can even lead to lockups. 8898 Thus the CRTC DPMS function is split into two parts: one that just enables 8899 /disables the CRTC and one which wraps this function and does the rest. 8900 Now the inner function can be called directly from RADEONRestore() as 8901 there is no need to go thru the RandR hooks in this function while the 8902 RandR hook uses the wrappering function so the full functionality is 8903 preserved from an RandR point of view. 8904 8905 Signed-off-by: Egbert Eich <eich@freedesktop.org> 8906 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 8907 8908commit ac51e331895b216d288bc7bd108a38b362214668 8909Author: Egbert Eich <eich@suse.de> 8910Date: Mon Nov 14 19:10:01 2011 +0100 8911 8912 UMS: Fix lockups in palette save/restore on pre-AVIVO chips. 8913 8914 The reintroduction of palette save/restore in 5efdf514 causes some 8915 pre-AVIVO chips to lock up. An investigation revealed that accessing 8916 palette registers when the associated PLL is not running is causing 8917 this. With UMS the PLL setup that is saved has been done by the BIOS 8918 typically. 8919 A similar issue was observed when VGA palette save/restore had 8920 been reinitroduced with 80eee856: 8921 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=480312 8922 and has been worked around for Linux without further investigation 8923 by 87e66ce7. 8924 To fix the issue we now 8925 a. introduce 'on-demand' palette saving (ie the palette is 8926 saved before it is first altered). This guarantees that 8927 the palette register are only associated when the associated 8928 CRTC is active and thus the PLLs are powered up and running. 8929 b. move palette restore before PLL restore. 8930 c. eliminate generic VGA palette save/restore which seems to be 8931 unneeded when the palette is restored natively. 8932 It is believed that this caused the behavior described in 8933 https://bugs.freedesktop.org/show_bug.cgi?id=18407#c27 8934 8935 Signed-off-by: Egbert Eich <eich@freedesktop.org> 8936 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 8937 8938commit 0a8d04eeac95f4db9d03ee31070bd825a7feb0b2 8939Author: Matthieu Herrb <matthieu.herrb@laas.fr> 8940Date: Sun Jan 1 18:27:54 2012 +0100 8941 8942 Update for new vgaHW API. 8943 8944 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 8945 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 8946 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8947 8948commit ae45d7e6d8e6844cd4586c9ee97c21b257fa788f 8949Author: Michel Dänzer <michel.daenzer@amd.com> 8950Date: Wed Dec 28 11:48:36 2011 +0100 8951 8952 DRI2: Can't use page flipping for pixmaps. (bug #42913) 8953 8954 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=42913 . 8955 8956 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 8957 8958commit aacbd629b02cbee3f9e6a0ee452b4e3f21376bd3 8959Author: Alex Deucher <alexdeucher@gmail.com> 8960Date: Mon Dec 12 09:32:30 2011 -0500 8961 8962 radeon: add some new pci ids 8963 8964 fixes: 8965 https://bugs.freedesktop.org/show_bug.cgi?id=43739 8966 8967 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 8968 8969commit bc54e415e2fd344939c5c788ea0686133a7e2c69 8970Author: Dave Airlie <airlied@redhat.com> 8971Date: Tue Dec 6 15:47:45 2011 +0000 8972 8973 radeon: add original radeon to always tiled. 8974 8975 and actually enable it for M7, previous commit only did one function. 8976 8977 Signed-off-by: Dave Airlie <airlied@redhat.com> 8978 8979commit ba46c7b0cf72d157748981eb3224d5eefb6200aa 8980Author: Dave Airlie <airlied@redhat.com> 8981Date: Tue Dec 6 13:42:49 2011 +0000 8982 8983 radeon: refine always tiled depth check 8984 8985 So it appears the M7 family always tiles its depth buffer also. 8986 8987 Signed-off-by: Dave Airlie <airlied@redhat.com> 8988 8989commit 98b2d5fe1722a43c4bbe7711ed7180a3fb65305f 8990Author: Dave Airlie <airlied@redhat.com> 8991Date: Mon Dec 5 18:44:28 2011 +0000 8992 8993 radeon: r200 depth buffers are always tiled 8994 8995 When we do the allocations we need to make sure the always tiled 8996 nature is taken into account. 8997 8998 Signed-off-by: Dave Airlie <airlied@redhat.com> 8999 9000commit 7dcefc69d9fbceae27cd03083c815e01a19b527e 9001Author: Alex Deucher <alexdeucher@gmail.com> 9002Date: Mon Dec 5 09:21:48 2011 -0500 9003 9004 Xv: Evergreen+ asics support 16k surfaces 9005 9006 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9007 9008commit 72e386d42516e7cd3c2cbf2fffc9174cd3ec8451 9009Author: Alex Deucher <alexander.deucher@amd.com> 9010Date: Wed Nov 30 19:38:35 2011 -0500 9011 9012 radeon: add some new pci ids 9013 9014 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 9015 9016commit 3853c3020d05175ae180b9a188dec7c425bdd0b8 9017Author: Dave Airlie <airlied@redhat.com> 9018Date: Mon Nov 28 18:38:30 2011 +0000 9019 9020 fixup xinerama since 9151f3b1c2ebcc34e63195888ba696f2183ba5e2 9021 9022 since the driver would call RRFirstOutput without checking if randr has 9023 been enabled, and it would crash in privates code. 9024 9025 reported by vereteran on #radeon 9026 9027 Signed-off-by: Dave Airlie <airlied@redhat.com> 9028 Acked-on-irc-by: Michel Dänzer <michel.daenzer@amd.com> 9029 9030commit d669c34f140c000f88c4b4e464e44e6c8694f581 9031Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 9032Date: Mon Nov 21 11:35:40 2011 +1100 9033 9034 ddx/evergreen: Fix endian of ALU constants 9035 9036 The constants are written directly into a buffer object shared with the 9037 card and we "forget" to swap them. This patch fixes it by doing the swap 9038 in evergreen_set_alu_consts() in-place (ie, it modifies the buffer), 9039 which should be fine with the way we use it in the ddx. 9040 9041 This makes everything work fine on my caicos card on a G5 including some 9042 quik tests with Xv, gnome3 shell, etc... 9043 9044 Thanks a lot to Jerome Glisse for holding my hand through debugging that 9045 (and finding the actual bug). 9046 9047 Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> 9048 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9049 9050commit 422bdd4fe6cb728e1dd08a56f6ee2d0f009cbfcb 9051Author: Alex Deucher <alexdeucher@gmail.com> 9052Date: Mon Nov 14 09:39:16 2011 -0500 9053 9054 radeon: add missing FireMV pci id 9055 9056 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9057 9058commit 534fb6e413a909a9d1afd57d1c711844b7c5ddf4 9059Author: Dave Airlie <airlied@redhat.com> 9060Date: Fri Nov 11 10:26:51 2011 +0000 9061 9062 ati: enable bg none when fbcon succeeds and we are built against ABI after 10. 9063 9064 One less patch to keep carrying in Fedora. 9065 9066 Signed-off-by: Dave Airlie <airlied@redhat.com> 9067 9068commit 89452c08048c98fb5cc3dc551b3824be40d52cf2 9069Author: Michel Dänzer <michel.daenzer@amd.com> 9070Date: Tue Nov 8 11:23:11 2011 +0100 9071 9072 UMS: Guard references to PCITAG / pciTag with XSERVER_LIBPCIACCESS (bug #42690) 9073 9074 Should fix https://bugs.freedesktop.org/show_bug.cgi?id=42690 . 9075 9076 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9077 9078commit 5ec34ed95948f7164184551615c1fc4c3eef3b98 9079Author: Ilija Hadzic <ihadzic@research.bell-labs.com> 9080Date: Thu Nov 3 20:16:47 2011 -0400 9081 9082 DRI/DRI2: remove hard-coded limitation to 6 crtcs 9083 9084 DRM's hard limit to the number of CRTCs is 32. ATI DDX unnecessarily 9085 clips this limit to 6 by hard coding initial assumption for 9086 output->possible_crtcs mask to 0x7f (before it gets trimmed down to 9087 what's really possible for a given output) and by allocating only 6 9088 entries for for cursor_bo[] array in RADEONInfoRec. 9089 9090 Fix this and thus allow the ATI DDX to deal with as many CRTCs 9091 as the DRM allows (32), so it is ready if anything with >6 CRTCs 9092 comes out. 9093 9094 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 9095 9096commit 4853ab2cdc3b97948c7cd69eaf4fff54f59774fc 9097Author: Michel Dänzer <michel.daenzer@amd.com> 9098Date: Fri Nov 4 12:15:53 2011 +0100 9099 9100 Turn compile time check into runtime check. 9101 9102 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9103 9104commit bcdb54fe16ebf2e239b84eebf20e8adfe5094bff 9105Author: Alex Deucher <alexdeucher@gmail.com> 9106Date: Thu Oct 20 11:11:35 2011 -0400 9107 9108 check for xserver 1.9.4.901 to enable tiling by default 9109 9110 Previous xservers had a bug in the EXA code which caused 9111 display corruption in some cases. 9112 9113 See: 9114 https://bugs.freedesktop.org/show_bug.cgi?id=33929 9115 9116 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9117 9118commit edde94cba5321e6e51e7fd4d79dde5abc4944495 9119Author: Michel Dänzer <michel.daenzer@amd.com> 9120Date: Wed Nov 2 13:40:05 2011 +0100 9121 9122 Bump version post release. 9123 9124commit 93459f842c2d8dc178a1954b8e05150fcb96ac9a 9125Author: Michel Dänzer <michel.daenzer@amd.com> 9126Date: Wed Nov 2 12:51:15 2011 +0100 9127 9128 Bump version for 6.14.3 release. 9129 9130commit fe3ce559b7f07d39ded39abb38576846d6eb185b 9131Author: Jeremy Huddleston <jeremyhu@apple.com> 9132Date: Tue Nov 1 20:27:03 2011 -0700 9133 9134 Build fix for -Werror=int-to-pointer-cast 9135 9136 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 9137 9138commit d3d6a5da07aaec5961e51c9a8f90c1490ee101b6 9139Author: Tormod Volden <debian.tormod@gmail.com> 9140Date: Sat Oct 29 23:08:46 2011 +0200 9141 9142 radeon: do not include xf86PciInfo.h 9143 9144 We already use atipciids.h instead most places. 9145 9146 Signed-off-by: Tormod Volden <debian.tormod@gmail.com> 9147 9148commit 70da7001e81363ed6ef2c4727c512daf53ae29fe 9149Author: Jeremy Huddleston <jeremyhu@apple.com> 9150Date: Sat Oct 29 20:15:09 2011 -0700 9151 9152 Use malloc/calloc/realloc/free directly 9153 9154 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 9155 9156commit a3bb07efb1757c33d70e2e1928219d12a4dd6498 9157Author: Michel Dänzer <michel.daenzer@amd.com> 9158Date: Tue Oct 25 17:43:58 2011 +0200 9159 9160 EXA >= R6xx / KMS: Avoid running out of CS space at inconvenient times. 9161 9162 Otherwise we may end up with things not properly set up at the beginning of the 9163 next CS. 9164 9165 Fixes http://bugs.debian.org/645007 . 9166 9167 In contrast to the Composite code for < R6xx, this isn't necessary with UMS, 9168 as the draw packet only uses constant space in the indirect buffer, and nothing 9169 else can mess with the 3D state between indirect buffers. 9170 9171 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9172 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 9173 9174commit 23788c4a5e3b6affb9b183e1393edd0e5ca4550e 9175Author: Michel Dänzer <michel.daenzer@amd.com> 9176Date: Thu Oct 27 19:12:22 2011 +0200 9177 9178 EXA < R6xx: Make sure 2D state is re-emitted after running out of CS space. 9179 9180 Otherwise it's basically luck what the 2D state ends up being at the beginning 9181 of the next CS. 9182 9183 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9184 9185commit 6e0e1a821accc6ca95f4134e49b66a6b168c1934 9186Author: Michel Dänzer <michel.daenzer@amd.com> 9187Date: Tue Oct 25 17:39:57 2011 +0200 9188 9189 Make radeon_dri2_create_buffer(s) more robust. (Bug #30047) 9190 9191 In particular, handle and propagate failure to allocate GPU accessible memory, 9192 instead of crashing. Fixes https://bugs.freedesktop.org/show_bug.cgi?id=30047 . 9193 9194 Also take care not to leak resources in error paths. 9195 9196 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9197 9198commit 856583dbca9319c77fed40daa9956e81a0068f9e 9199Author: Jerome Glisse <jglisse@redhat.com> 9200Date: Thu Oct 20 14:17:14 2011 -0400 9201 9202 radeon/kms: fallback to vesa if GPU is not supported by UMS 9203 9204 For GPU not supported by UMS, test in probe so that we properly 9205 fallback to vesa. 9206 9207 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 9208 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 9209 9210commit f6f1b4f7789cfef763c063e671b76b60be8bdd8e 9211Author: Maarten Lankhorst <m.b.lankhorst@gmail.com> 9212Date: Tue Oct 18 14:30:39 2011 +0200 9213 9214 dri2: Add vdpau driver name entry 9215 9216 libvdpau has a driver loading mechanism that looks for a dri2 driver 9217 first before falling back to nvidia, so lets use that. 9218 9219 Allows use of libvdpau_rx00 without having to set things up separately, 9220 similar to the patch to xf86-video-nouveau. 9221 9222 Signed-off-by: Maarten Lankhorst <m.b.lankhorst@gmail.com> 9223 Reviewed-by: Christian König <deathsimple@vodafone.de> 9224 Reviewed-by: Michel Dänzer <michel@daenzer.net> 9225 Tested-by: Michel Dänzer <michel@daenzer.net> 9226 9227commit a6b2bd2d184f10d4c56c4ee17186aedb238a36ec 9228Author: Alex Deucher <alexdeucher@gmail.com> 9229Date: Wed Sep 21 17:00:16 2011 -0400 9230 9231 UMS: fix DDIA enable on some rs690 systems 9232 9233 DVOOutputControl checks the value of of bios scratch reg 3 9234 on some tables and assumes the encoder is already enabled 9235 if the DFP2_ACTIVE bit is set. Clear that bit so the table 9236 sets the DDIA enable bit properly. 9237 9238 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9239 9240commit d78860ba53d9bfcf6c28e1cfd2d970709b5e20fa 9241Author: Michel Dänzer <michel.daenzer@amd.com> 9242Date: Fri Aug 12 11:21:33 2011 +0200 9243 9244 Only call radeon_dri2_close_screen() if DRI2 was enabled. 9245 9246 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9247 9248commit 11330ca5dc61a70fe4507e63230f9133ca22d891 9249Author: Michel Dänzer <michel.daenzer@amd.com> 9250Date: Fri Aug 12 11:21:32 2011 +0200 9251 9252 Remove dead variable remain_size_bytes. 9253 9254 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9255 9256commit fcf0cca9c0ab0f692b222f619aee8f1cdad3b519 9257Author: Michel Dänzer <michel.daenzer@amd.com> 9258Date: Tue Sep 20 12:34:05 2011 +0200 9259 9260 KMS Color Tiling requires xserver which supports EXA_MIXED_PIXMAPS. 9261 9262 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9263 9264commit c96e6fb8a5f5be2319fdb4c431c1ba5279997fe8 9265Author: Alex Deucher <alexdeucher@gmail.com> 9266Date: Sat Sep 17 08:26:12 2011 -0400 9267 9268 man: note that the list of marketing names is non-exhaustive 9269 9270 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9271 9272commit 1a51fce6f6ab169c882a86b936909c0820f27a68 9273Author: Alex Deucher <alexdeucher@gmail.com> 9274Date: Thu Sep 15 17:52:25 2011 -0400 9275 9276 update man page with new marking names 9277 9278 Note that the driver support all currently 9279 shipping asics and the names in the man page 9280 are just a sampling. 9281 9282 Fixes: 9283 https://bugs.freedesktop.org/show_bug.cgi?id=40808 9284 9285 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9286 9287commit 64f237a4cf3ef5bcb3163e88e1447ff275a1eefa 9288Author: Michel Dänzer <michel.daenzer@amd.com> 9289Date: Thu Aug 18 19:11:08 2011 +0200 9290 9291 Convert register ranges for >= r6xx from enums to defines. 9292 9293 Avoids lots of "comparison between 'enum <anonymous>' and 'enum <anonymous>'" 9294 warnings with newer versions of gcc. See 9295 https://bugs.freedesktop.org/show_bug.cgi?id=38238 . 9296 9297 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9298 9299commit 95991fcce45f0dd904f76b9f98f9c7ed5708e4fa 9300Author: Michel Dänzer <michel.daenzer@amd.com> 9301Date: Wed Aug 17 11:10:34 2011 +0200 9302 9303 Bail if we're trying to start up in UMS mode on KMS. 9304 9305 Ideally, the display manager will start the X server again, and everything 9306 will be fine and dandy. But in the worst case, at least we won't hit the 9307 hardware behind the KMS driver's back. 9308 9309 (This change intentionally makes (ab)use of the fact that Bool is defined as 9310 int). 9311 9312 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9313 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 9314 9315commit f95a41b7851565c282d22f8d679db1377428f165 9316Author: Michel Dänzer <michel.daenzer@amd.com> 9317Date: Thu Aug 11 11:22:57 2011 +0200 9318 9319 video: Don't round up bottom/right edge for clipping source width/height. 9320 9321 It's not necessary: If the top/left edge was rounded down, this will be 9322 compensated by the subtraction. 9323 9324 Worse, if the original source width/height is odd, rounding up may result in 9325 reading past the end of the source data. 9326 9327 Fixes http://bugs.debian.org/637258 . 9328 9329 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9330 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 9331 9332commit 93fc0843a1e31dc9237433bc2bf17df79e956d26 9333Author: Michel Dänzer <michel.daenzer@amd.com> 9334Date: Wed Aug 10 17:44:37 2011 +0200 9335 9336 Change my e-mail address to something that still works, and always will, I hope. 9337 9338commit 9151f3b1c2ebcc34e63195888ba696f2183ba5e2 9339Author: Michel Dänzer <michel.daenzer@amd.com> 9340Date: Tue Aug 9 19:13:26 2011 +0200 9341 9342 Prefer the CRTC of the primary output for synchronization. 9343 9344 See https://bugs.freedesktop.org/show_bug.cgi?id=39696 . 9345 9346 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9347 9348commit 3b9fdc807dd7e52af0576299cefba596040f6f2f 9349Author: Alex Deucher <alexdeucher@gmail.com> 9350Date: Wed Aug 3 16:20:13 2011 -0400 9351 9352 r5xx+: Fix vline setup with crtc offsets 9353 9354 On r5xx+, vline is relative to to the viewport, not 9355 the scanlines. Based on initial patch and investigation 9356 from Herbert Pötzl (Bertl) on IRC. 9357 9358 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9359 9360commit 104b2d7c071f29266b1bc4184a74e9714d14febc 9361Author: Alex Deucher <alexander.deucher@amd.com> 9362Date: Mon Aug 1 10:05:30 2011 -0400 9363 9364 kms: fix possible leak in pageflip code 9365 9366 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 9367 9368commit d29bab632e9ecccba518d4107d52620bf75eb1cf 9369Author: Ville Syrjala <syrjala@sci.fi> 9370Date: Wed May 4 23:51:27 2011 +0300 9371 9372 kms: Move flip_count and co. to a per swap structure 9373 9374 If multiple drawables are doing page flipping, the global drmmode 9375 structure can't be used to keep per swap information. For example 9376 flip_count can increase prematurely due to another swap request, 9377 and then the previous swap request never gets completed, leading to a 9378 stuck client. Move the relevant pieces of data to a strucuture that 9379 gets allocated once per swap request and shared by all involved CRTCs. 9380 9381 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 9382 9383commit 9493563c1ef4b51af0ee8a44cb4e7c5bb280347e 9384Author: Ville Syrjala <syrjala@sci.fi> 9385Date: Wed May 4 23:51:26 2011 +0300 9386 9387 dri2: Update front buffer pixmap and name before exchanging buffers 9388 9389 Buffer exchange assumes that the front buffer pixmap and name 9390 information is accurate. That may not be the case eg. if the window 9391 has been (un)redirected since the buffer was created. 9392 9393 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 9394 9395commit 8c9266ed2da22a510243f9a952c14d4423f48a2b 9396Author: Alex Deucher <alexander.deucher@amd.com> 9397Date: Fri Jul 15 10:44:57 2011 -0400 9398 9399 radeon: add some new NI pci ids 9400 9401 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 9402 9403commit f59c3b294b0f715fc96e2bbe25893f2b31aa488b 9404Author: Christian König <deathsimple@vodafone.de> 9405Date: Thu Jul 14 11:49:06 2011 +0200 9406 9407 Register XvMC video decoding acceleration 9408 9409commit e8d0d437957b15252dfad775796a3949ed50dbcf 9410Author: Dave Airlie <airlied@redhat.com> 9411Date: Tue Jul 12 11:43:25 2011 -0400 9412 9413 evergreen: Emit SQ_LDS_RESOURCE_MGMT 9414 9415 Avoids rendering problems when compute changes this reg. 9416 9417 Fixes: 9418 https://bugs.freedesktop.org/show_bug.cgi?id=39119 9419 9420 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9421 9422commit 9bb31158466e6168116d841d12c8b4303f11c4a6 9423Author: Alex Deucher <alexdeucher@gmail.com> 9424Date: Wed Jun 22 12:24:28 2011 -0400 9425 9426 evergreen: fix num_banks for 2D tiling config 9427 9428 The field is encoded. 9429 9430 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9431 9432commit 122b471f734aa07427b01d4bec35ff1ac28290b5 9433Author: Alex Deucher <alexdeucher@gmail.com> 9434Date: Fri Jun 17 04:02:16 2011 -0400 9435 9436 dri2: fix copy pasto in a6154c00c64932332e8f6e334661ffd579cfd894 9437 9438 Reported-by: Nils Wallménius <nils.wallmenius@gmail.com> 9439 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9440 9441commit 122bedcbcf45cb583cf51b2fd04ed2805e0ca60b 9442Author: Alex Deucher <alexdeucher@gmail.com> 9443Date: Thu Jun 16 12:57:11 2011 -0400 9444 9445 dri2: missing bit from a6154c00c64932332e8f6e334661ffd579cfd894 9446 9447 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9448 9449commit a6154c00c64932332e8f6e334661ffd579cfd894 9450Author: Alex Deucher <alexdeucher@gmail.com> 9451Date: Thu Jun 16 12:48:43 2011 -0400 9452 9453 dri2/eg+: fix size and alignment of depth/stencil buffers 9454 9455 Base alignment may be 256B or 512B depending on the group 9456 size. Also need to check against front size for virtualX. 9457 9458 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9459 9460commit 41eb1fbb3d9da64feb4a96df7c575e44136b1538 9461Author: Alex Deucher <alexdeucher@gmail.com> 9462Date: Mon Jun 13 12:50:39 2011 -0400 9463 9464 kms/man: update ColorTiling info 9465 9466 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9467 9468commit 481baa5c74271cd7ce38bae3965d2bc4b8809058 9469Author: Alex Deucher <alexdeucher@gmail.com> 9470Date: Mon Jun 13 12:44:07 2011 -0400 9471 9472 kms: enable ColorTiling by default on r6xx-cayman asics 9473 9474 Currently only 1D tiling as 2D tiling still has some corner 9475 cases to fix up. 9476 9477 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9478 9479commit cbcc57b0fa6f581be777bef648f2bf3efe7443ee 9480Author: Alex Deucher <alexdeucher@gmail.com> 9481Date: Mon Apr 4 12:52:00 2011 -0400 9482 9483 radeon: add llano pci ids 9484 9485 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9486 9487commit 36afd1e1055eeadb2396dadcc68b214655bd90a9 9488Author: Alex Deucher <alexdeucher@gmail.com> 9489Date: Tue May 31 16:03:36 2011 -0400 9490 9491 radeon: add support for llano APUs 9492 9493 - KMS only 9494 - Includes full EXA/Xv support 9495 9496 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9497 9498commit 557f46dc2f18734ecf1f18dee7e951e0bf062e63 9499Author: Dave Airlie <airlied@redhat.com> 9500Date: Fri May 27 07:22:08 2011 +1000 9501 9502 bump version after release 9503 9504commit 2fca40ea65d9f2a6f8451c324bb4b82786f34f76 9505Author: Dave Airlie <airlied@redhat.com> 9506Date: Thu May 26 12:52:21 2011 +1000 9507 9508 radeon: bump version for release of 6.14.2 9509 9510commit fab868c5f7d8cafdb0176d2751f216819a5ba66a 9511Author: Alex Deucher <alexdeucher@gmail.com> 9512Date: Wed May 25 01:09:12 2011 -0400 9513 9514 EXA: make evergreen_fix_scissor_coordinates static 9515 9516 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9517 9518commit fe5c42f5155361006b687da824181418f688809f 9519Author: Alex Deucher <alexdeucher@gmail.com> 9520Date: Tue May 24 22:32:01 2011 -0400 9521 9522 cayman: endian fixes for shaders 9523 9524 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9525 9526commit 470ecd02347c32e79316046d01a7d5dad0e2fe99 9527Author: Alex Deucher <alexdeucher@gmail.com> 9528Date: Tue May 24 18:31:47 2011 -0400 9529 9530 EXA/Xv: add workarounds for eg/cayman scissors bugs 9531 9532 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9533 9534commit b913e7ba2b60d47a6660699210cc3cf6f5dc52c2 9535Author: Alex Deucher <alexdeucher@gmail.com> 9536Date: Tue May 24 18:02:52 2011 -0400 9537 9538 Revert "cayman: add scissors workaround." 9539 9540 Needs a proper workaround for a hw bug. 9541 9542 This reverts commit b77d374b0d11f48c33cfffdb4157c4ec4b05ea15. 9543 9544commit 82cb33c3f0e1ba802d7a94f3159b3c5c86cd4043 9545Author: Dave Airlie <airlied@redhat.com> 9546Date: Wed May 18 14:49:17 2011 +1000 9547 9548 cayman: enable all accel 9549 9550commit b77d374b0d11f48c33cfffdb4157c4ec4b05ea15 9551Author: Dave Airlie <airlied@redhat.com> 9552Date: Tue May 24 15:53:58 2011 +1000 9553 9554 cayman: add scissors workaround. 9555 9556 wasted a lot of time getting to this. 9557 9558commit ffeab7a7058298e15294a3b2c740c731e36dda1d 9559Author: Alex Deucher <alexdeucher@gmail.com> 9560Date: Mon Apr 18 18:16:51 2011 -0400 9561 9562 cayman: fix dword counts default state 9563 9564 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9565 9566commit 3cbfae361bf5e779d3364f0f31cfd25bd0f59e65 9567Author: Alex Deucher <alexdeucher@gmail.com> 9568Date: Wed Mar 2 20:48:19 2011 -0500 9569 9570 cayman: add spi state to default state 9571 9572 changed in e3145801b80fd4be4cf770128876e86e89bda66f 9573 9574 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9575 9576commit b8ade97c9d0fa5aacb0e3166868bb72e9bc679a6 9577Author: Alex Deucher <alexdeucher@gmail.com> 9578Date: Wed Mar 2 20:44:19 2011 -0500 9579 9580 cayman: first pass at exa/Xv shaders 9581 9582 Main differences with evergreen: 9583 - 4-way rather than 5-way 9584 - END_OF_PROGRAM bit removed from CF istructions, use 9585 CF_INST_END instead. 9586 - MEGA_FETCH* fields removed from VTX commands 9587 - no more VC, all fetches go through the TC 9588 9589 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9590 9591commit 01b646ed800732985c1638b147716641a99082f9 9592Author: Alex Deucher <alexdeucher@gmail.com> 9593Date: Wed Mar 2 20:39:38 2011 -0500 9594 9595 cayman: add a default state function 9596 9597 The rest of the state functions can be shared 9598 with evergreen. I've noted where there are 9599 differences. 9600 9601 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9602 9603commit 42eecc6f4fb1570769490bdaeac06817c6c36a7e 9604Author: Alex Deucher <alexdeucher@gmail.com> 9605Date: Wed Mar 2 20:20:56 2011 -0500 9606 9607 cayman: add 3D register headers 9608 9609 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9610 9611commit e1d28e011f4a5139cbc778973c63158ed2746716 9612Author: Alex Deucher <alexdeucher@gmail.com> 9613Date: Wed Mar 2 20:13:50 2011 -0500 9614 9615 kms/cayman: stub out exa support 9616 9617 Just fallbacks for now. 9618 9619 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9620 9621commit 21e44a20b8b1b64079ee77f45aaa5010206ed7b6 9622Author: Michel Dänzer <daenzer@vmware.com> 9623Date: Thu May 12 09:23:38 2011 +0200 9624 9625 UMS: Fix comparison of unsigned variable against < 0. 9626 9627 Pointed out by clang: 9628 9629 ../../src/radeon_crtc.c:242:18: error: comparison of unsigned expression < 0 is always false [-Werror,-Wtautological-compare] 9630 error = error < 0 ? 0xffffffff : error; 9631 ~~~~~ ^ ~ 9632 9633 If a UMS regression is bisected to this commit, the assignment should probably 9634 just be removed, as it's a no-op in the current form. 9635 9636commit 3b893d81982c9381393c92625e308541e0071b05 9637Author: Michel Dänzer <daenzer@vmware.com> 9638Date: Fri May 13 09:10:02 2011 +0200 9639 9640 KMS: Fix output properties logic error. 9641 9642 Pointed out by clang: 9643 9644 ../../src/drmmode_display.c:1023:30: error: use of logical && with constant operand; switch to bitwise & or remove constant [-Werror,-Wconstant-logical-operand] 9645 if (props && (props->flags && DRM_MODE_PROP_ENUM)) { 9646 ^ ~~~~~~~~~~~~~~~~~~ 9647 9648 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 9649 9650commit f83d58cf5b33686139067f8f898b8e566ba5c253 9651Author: Nicolas Kaiser <nikai@nikai.net> 9652Date: Fri May 13 00:56:31 2011 +0200 9653 9654 man: fix typos 9655 9656 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 9657 9658commit 90abffbd30f44b9cf76a6e28103ddcb5419b4522 9659Author: Ilija Hadzic <ihadzic@research.bell-labs.com> 9660Date: Fri May 6 09:45:23 2011 -0400 9661 9662 DRI2: fix high-crtc/vblank oversight/bug 9663 9664 improvements to high-crtc handling done in 9665 f0b7d7b449cc77bb2b281d81108507f8bc2e6018 introduced a bug that caused 9666 the populate_vbl_request_type to never use the high-crtc field even 9667 when it should. The reason is that the offending patch put the code 9668 under #ifdef DRM_VBLANK_HIGH_CRTC_MASK which is not visible outside the 9669 enum type, so #else was always taken in compilation type. This patch 9670 fixes it by basing #ifdef on (pre-processor visible) 9671 DRM_VBLANK_HIGH_CRTC_SHIFT constant 9672 9673 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 9674 9675commit 62a4cd180fe884dca24586d453395472516e6496 9676Author: Alex Deucher <alexdeucher@gmail.com> 9677Date: Wed May 4 01:13:55 2011 -0400 9678 9679 fusion: fix tiling enable logic 9680 9681 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9682 9683commit 76638ca687b02d3b1494b9868f817fd4fd892c64 9684Author: Alex Deucher <alexdeucher@gmail.com> 9685Date: Wed May 4 01:06:22 2011 -0400 9686 9687 fusion: enable tiling if DFS works 9688 9689 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9690 9691commit a6d2dba6573a3512d550d7e442bf42ea03012bbc 9692Author: Dave Airlie <airlied@redhat.com> 9693Date: Wed May 4 10:44:43 2011 +1000 9694 9695 radeon: add add hw DFS support for fusion 9696 9697 Fusion had a bug setting up the VM on earlier kernels so we need to work 9698 around that and only enable accel on a new enough kernel. 9699 9700 Signed-off-by: Dave Airlie <airlied@redhat.com> 9701 9702commit 859e052af49e68a826b77a9135c7f067dc331a06 9703Author: Alex Deucher <alexdeucher@gmail.com> 9704Date: Tue May 3 15:15:04 2011 -0400 9705 9706 radeon: add some new pci ids 9707 9708 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9709 9710commit 8f8bbf628c6eed037f57bc8c155f0ecdacbebad1 9711Author: Alex Deucher <alexdeucher@gmail.com> 9712Date: Sun May 1 13:19:15 2011 -0400 9713 9714 man: add cayman to man page 9715 9716 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9717 9718commit 903e90c31cf0319be9297529aa7b8daa1756cf63 9719Author: Alex Deucher <alexdeucher@gmail.com> 9720Date: Wed Apr 20 03:10:08 2011 -0400 9721 9722 EXA/Xv: used cached bo tiling flags for accel setup on 6xx+ 9723 9724 This avoids calling into the kernel for each bo in the accel 9725 code. This is a follow on to: 9726 cc7d1fa39da40a532fcdbe6c7924ca47a879e66a 9727 9728 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9729 9730commit 982c22f16c8eeee9be81779fbfe17d8d3f9b6897 9731Author: Adam Jackson <ajax@redhat.com> 9732Date: Thu Apr 14 16:04:50 2011 -0400 9733 9734 R520: Fix textures larger than 2k 9735 9736 Ported from the equivalent fix in Mesa. 9737 9738commit cc7d1fa39da40a532fcdbe6c7924ca47a879e66a 9739Author: Michel Dänzer <daenzer@vmware.com> 9740Date: Mon Apr 4 17:37:12 2011 +0200 9741 9742 EXA: Cache BO tiling flags. 9743 9744 Calling into the kernel every time is quite expensive, and nobody else should 9745 ever change the tiling flags. 9746 9747 There's still more to do along the same lines for >= R6xx. 9748 9749commit f0b7d7b449cc77bb2b281d81108507f8bc2e6018 9750Author: Michel Dänzer <daenzer@vmware.com> 9751Date: Tue Apr 5 13:36:01 2011 +0200 9752 9753 DRI2: Some cleanups for the scheduling mess. 9754 9755 * Fix build against libdrm that doesn't define *_VBLANK_HIGH_CRTC*. 9756 * If we have more than two CRTCs but can't use DRM_VBLANK_HIGH_CRTC_MASK, don't 9757 enable scheduling in the first place rather than relying on 9758 DRM_VBLANK_SECONDARY magically doing something sensible for higher CRTCs. 9759 * Only set up client state tracking when scheduling is enabled. 9760 * Only declare pRADEONEnt when it's needed, and break long lines. 9761 9762commit 7acf9bc833de539fa2259a051c66a99445a54bc4 9763Author: Alex Deucher <alexdeucher@gmail.com> 9764Date: Mon Apr 4 11:08:37 2011 -0400 9765 9766 radeon: add some additional ontario pci ids 9767 9768 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9769 9770commit 0af6386f85a0f1ba14864a3334164733a10a6cb8 9771Author: Ilija Hadzic <ihadzic@research.bell-labs.com> 9772Date: Thu Mar 24 13:33:27 2011 -0400 9773 9774 xf86-video-ati: (revised #2) add support for vblank on crtc > 1 9775 9776 Hi Alex, 9777 9778 Enclosed is a revised version of two patches sent on Mar 18 and Mar 22, 9779 respectively. Details summarized in these two threads: 9780 http://lists.freedesktop.org/archives/dri-devel/2011-March/009463.html 9781 http://lists.freedesktop.org/archives/dri-devel/2011-March/009582.html 9782 9783 This patch reconciles the DDX with the change in libdrm sent to this list 9784 earlier today. Specifically, it refers to a symbol that has been renamed 9785 from DRM_CAP_HIGH_CRTC to DRM_CAP_VBLANK_HIGH_CRTC. It *supersedes* the 9786 previous patch (i.e. apply it to the master branch as it exists at the 9787 time of this writing, not as an incremental patch to the one sent previously). 9788 9789 Regards, 9790 9791 Ilija 9792 9793 Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com> 9794 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9795 9796commit fe2e0ad3ffa58f40311319c950b842e2928a5740 9797Author: matthew green <mrg@eterna.com.au> 9798Date: Mon Mar 21 12:17:58 2011 -0400 9799 9800 bug fix for r6xx/r7xx UMS 9801 9802 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9803 9804commit 83978ad5fe37581e2b2f3fbd9c073d91b2ae1d50 9805Author: Alex Deucher <alexdeucher@gmail.com> 9806Date: Thu Mar 17 19:07:43 2011 -0400 9807 9808 bump version post release 9809 9810 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9811 9812commit 38d9368e59b2990bf32a028ece2132451b402350 9813Author: Alex Deucher <alexdeucher@gmail.com> 9814Date: Thu Mar 17 18:59:55 2011 -0400 9815 9816 bump for release 9817 9818 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9819 9820commit cdfc007eccc9b130fc08012deef304488eb6e469 9821Author: Dave Airlie <airlied@redhat.com> 9822Date: Wed Mar 16 10:40:42 2011 +1000 9823 9824 radeon/exa: correct function name 9825 9826 this corrects the function name so it matches the contents. 9827 9828 Signed-off-by: Dave Airlie <airlied@redhat.com> 9829 9830commit b93d7658f669f6dc1cfacebcfe955a1e113a537c 9831Author: Dave Airlie <airlied@redhat.com> 9832Date: Wed Mar 16 10:37:22 2011 +1000 9833 9834 radeon/exa: fix scaling check 9835 9836 This caused a regression in firefox, as these numbers are xFixed 9837 values hence 1 is actually 65536. 9838 9839 Should fix: https://bugzilla.redhat.com/show_bug.cgi?id=685084 9840 9841 Signed-off-by: Dave Airlie <airlied@redhat.com> 9842 9843commit 8567b8cd9e136ef69906e02286b3752db2404741 9844Author: Alex Deucher <alexdeucher@gmail.com> 9845Date: Tue Mar 15 13:13:06 2011 -0400 9846 9847 APU: no tiling yet 9848 9849 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9850 9851commit 4d350497012fa31a417ada662006e2d64db2a4b5 9852Author: Dave Airlie <airlied@redhat.com> 9853Date: Tue Mar 15 10:32:19 2011 +1000 9854 9855 radeon: exa shaders don't handle scaling either. 9856 9857 rendercheck tsrccoords test fails. 9858 9859 Signed-off-by: Dave Airlie <airlied@redhat.com> 9860 9861commit 6319a33cb7bd8dba1eaf47bee1e2608cee5d7600 9862Author: Alex Deucher <alexdeucher@gmail.com> 9863Date: Wed Mar 2 19:38:04 2011 -0500 9864 9865 kms: add cayman pci ids 9866 9867 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9868 9869commit fd65def45a57e3242596274d270791df8abc8e29 9870Author: Alex Deucher <alexdeucher@gmail.com> 9871Date: Thu Jan 13 17:02:31 2011 -0500 9872 9873 kms: no accel yet for cayman 9874 9875 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9876 9877commit 585e2a3b803b6d3e070f2427e9cb49bab0949239 9878Author: Alex Deucher <alexdeucher@gmail.com> 9879Date: Thu Jan 13 16:58:01 2011 -0500 9880 9881 kms: add cayman chip family 9882 9883 KMS only. No UMS support for cayman. 9884 9885 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9886 9887commit 6835e22256ba7a74c02504383f63baaba7e87156 9888Author: Alex Deucher <alexdeucher@gmail.com> 9889Date: Wed Mar 2 12:20:45 2011 -0500 9890 9891 r6xx+: truncate point sampled coordinates 9892 9893 By default the hardware rounds texcoords. However, 9894 for point sampled textures, the expected behavior is 9895 to truncate. When we have point sampled textures, 9896 set the truncate bit in the sampler. 9897 9898 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9899 9900commit 8a8862e4165dc5be779bce5e327a31da4826db3b 9901Author: Sedat Dilek <sedat.dilek@googlemail.com> 9902Date: Sat Feb 26 21:04:22 2011 +0100 9903 9904 kms: Fix warning XNFprintf is deprecated (v2) 9905 9906 The first take of my patch missed the breakage against pre-1.10 9907 xservers and now I know XNFasprintf() is not a 1:1 replacement 9908 for XNFprintf(). 9909 The idea for the fix seemed to be so obvious, but as usual "really" 9910 test the patch you are sending to any ML. 9911 I am awfully sorry for the overhasty bad hack. 9912 Thanks to Alan and Alex for the help to fix it correctly, now. 9913 9914 Tested-By: Sedat Dilek <sedat.dilek@gmail.com> 9915 Signed-off-by: Sedat Dilek <sedat.dilek@gmail.com> 9916 9917commit 88304fe9c0ce69012d44a77a5368c35236511dac 9918Author: Alex Deucher <alexdeucher@gmail.com> 9919Date: Mon Feb 28 11:29:19 2011 -0500 9920 9921 kms: don't rotate if acceleration is not enabled 9922 9923 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9924 9925commit 99141761e947e2153ed1a3bc485dc5716a22acb1 9926Author: Alex Deucher <alexdeucher@gmail.com> 9927Date: Mon Feb 28 11:20:44 2011 -0500 9928 9929 kms: don't enable tiling if accel is off 9930 9931 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9932 9933commit 7928c120fb629a782846bcd5734d9fa14cbb099d 9934Author: Dave Airlie <airlied@redhat.com> 9935Date: Mon Feb 28 14:11:57 2011 +1000 9936 9937 xv: fix height alignments for U/V planes 9938 9939 The kernel CS checks showed we were incorrectly aligning the 9940 U/V allocations for when we used them as texture sources. 9941 9942 This should fix 9943 https://bugs.freedesktop.org/show_bug.cgi?id=34567 9944 9945 Signed-off-by: Dave Airlie <airlied@redhat.com> 9946 9947commit e55520c0e582f2a500b3b01c2cb78750e15256ca 9948Author: Michel Dänzer <daenzer@vmware.com> 9949Date: Sat Feb 26 15:27:59 2011 +0100 9950 9951 Revert "kms: Fix warning XNFprintf is deprecated" 9952 9953 This reverts commit cb4dc7ab66016d0c980800f8dbf71bbe993889d3. 9954 9955 Looks like this caused https://bugs.freedesktop.org/attachment.cgi?id=43852 , 9956 but even if it didn't, the following needs to be addressed first: 9957 9958 cc1: warnings being treated as errors 9959 ../../src/radeon_kms.c: In function ‘radeon_open_drm_master’: 9960 ../../src/radeon_kms.c:413: error: passing argument 1 of ‘XNFasprintf’ from incompatible pointer type 9961 /usr/local/include/xorg/Xprintf.h:57: note: expected ‘char **’ but argument is of type ‘char *’ 9962 ../../src/radeon_kms.c:413: error: passing argument 2 of ‘XNFasprintf’ makes pointer from integer without a cast 9963 /usr/local/include/xorg/Xprintf.h:57: note: expected ‘const char * restrict’ but argument is of type ‘uint16_t’ 9964 ../../src/radeon_kms.c:413: error: assignment makes pointer from integer without a cast 9965 make[2]: *** [radeon_kms.lo] Error 1 9966 9967commit cb4dc7ab66016d0c980800f8dbf71bbe993889d3 9968Author: Sedat Dilek <sedat.dilek@gmail.com> 9969Date: Fri Feb 25 17:21:52 2011 -0500 9970 9971 kms: Fix warning XNFprintf is deprecated 9972 9973 Noticed when building ati/radeon ddx against xserver 1.10-rc3. 9974 9975 [ build.log ] 9976 ../../src/radeon_kms.c:412:5: warning: 'XNFprintf' is deprecated (declared at /usr/include/xorg/os.h:273) 9977 9978 agd5f: don't break the build against pre-1.10 xservers. 9979 9980 Signed-off-by: Sedat Dilek <sedat.dilek@gmail.com> 9981 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 9982 9983commit ecfdb209afe2aafc378baab8c511f5df7b000270 9984Author: Sedat Dilek <sedat.dilek@googlemail.com> 9985Date: Fri Feb 25 21:48:14 2011 +0100 9986 9987 UMS: Fix build against xserver 1.10-rc3 9988 9989 This issue was introduced due to last minute backout of RandR-1.4 9990 in xserver 1.10-rc3. 9991 9992 Switch to "#ifdef RANDR_14_INTERFACE" as suggested by Keith Packard. 9993 See also <http://lists.x.org/archives/xorg-devel/2011-February/019643.html>. 9994 9995 Note: 9996 The ddx needs a rebuild as the X video driver ABI changed to version 10.0. 9997 9998 Reported-by: Alex Deucher <alexdeucher@gmail.com> 9999 CC: Keith Packard <keithp@keithp.com> 10000 Signed-off-by: Sedat Dilek <sedat.dilek@gmail.com> 10001 10002commit 04293a3a84acd96ccdf8e5e865cf973325ca43a1 10003Author: Alex Deucher <alexdeucher@gmail.com> 10004Date: Thu Feb 24 02:33:41 2011 -0500 10005 10006 dri2: disable pageflipping for transformed displays 10007 10008 Based on Ben's nouveau patch. 10009 10010 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10011 10012commit 91070cfd75d5607c4a72ace780f830f0ddb40e84 10013Author: Alex Deucher <alexdeucher@gmail.com> 10014Date: Wed Feb 23 00:55:28 2011 -0500 10015 10016 Xv: use aligned height to calculate Xv buffer size 10017 10018 Make sure it's large enough for the tiling mode so the 10019 CS checker doesn't complain. Should fix: 10020 https://bugs.freedesktop.org/show_bug.cgi?id=34567 10021 10022 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10023 10024commit acd54a48a6680956ed903fc890204c0c910b6712 10025Author: Mark Kettenis <kettenis@openbsd.org> 10026Date: Sun Feb 20 22:51:04 2011 +0100 10027 10028 Properly restore CLOCK_CNTL_INDEX register 10029 10030 RADEONRestore() restores the CLOCK_CNTL_INDEX register before calling 10031 RADEONRestorePLLRegisters(), which modifies the CLOCK_CNTL_INDEX register 10032 to access the PLL registers. As a result we may end up with the wrong clock 10033 being selected when exiting X. This happens on platforms where the driver 10034 doesn't save and restore the VGA state. 10035 10036 Signed-off-by: Mark Kettenis <kettenis@openbsd.org> 10037 10038commit d7c2ba4286ea32401c0026cb92ec167f7b5753d8 10039Author: Alex Deucher <alexdeucher@gmail.com> 10040Date: Wed Feb 16 18:30:34 2011 -0500 10041 10042 man: add btc chips to the list 10043 10044 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10045 10046commit d0f8bffbbc26818eccdd305ef142387dc4975bba 10047Author: Alex Deucher <alexdeucher@gmail.com> 10048Date: Wed Feb 16 15:09:20 2011 -0500 10049 10050 kms: add tiling support for evergreen/NI 10051 10052 - requires 2.6.37 or newer 10053 - 1D only at the moment 10054 - disabled by default 10055 10056 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10057 10058commit 0471d8412acd82e281a35fc4c6bb2d53b1ff5802 10059Author: Alex Deucher <alexdeucher@gmail.com> 10060Date: Wed Feb 16 15:07:35 2011 -0500 10061 10062 kms: EXA/Xv tiling fixes 10063 10064 - properly set tiling flags for temp surfaces 10065 - fix CB non_disp_tiling bits on evergreen 10066 10067 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10068 10069commit 4d7e1498f7d9eb50e2eddabca193fc27bde24f0e 10070Author: Alex Deucher <alexdeucher@gmail.com> 10071Date: Tue Feb 15 11:24:23 2011 -0500 10072 10073 fix Xv on pre-r6xx asics 10074 10075 broken by: 10076 a9c1358099:"Xv: fix textured video alignment" 10077 10078 Reported by: 10079 Artur Skawina <art.08.09@gmail.com> 10080 10081 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10082 10083commit a9a59717d11af37a2dda5555f6a83c5b65449527 10084Author: Alex Deucher <alexdeucher@gmail.com> 10085Date: Sat Feb 12 05:34:50 2011 -0500 10086 10087 kms: use worst case base/pitch align if we don't have drm tiling info 10088 10089 To avoid CS rejection. 10090 10091 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10092 10093commit 5f9b7fdb4ce27f8c6a64134bb4d5ebeb1c958cca 10094Author: Alex Deucher <alexdeucher@gmail.com> 10095Date: Sat Feb 12 05:19:12 2011 -0500 10096 10097 kms: always get the tiling info even if we aren't tiling 10098 10099 We need the proper group size to calculate alignment so the 10100 ddx and drm agree. 10101 10102 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10103 10104commit 355f0f7a67ce239b7fd2f2e81bcaaa88b2a09f03 10105Author: Alex Deucher <alexdeucher@gmail.com> 10106Date: Sat Feb 12 04:52:52 2011 -0500 10107 10108 kms: fix rotate pitch align 10109 10110 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10111 10112commit f1dc419c989addc4737aed06ec8b8acdb4d40063 10113Author: Alex Deucher <alexdeucher@gmail.com> 10114Date: Fri Feb 11 17:21:10 2011 -0500 10115 10116 kms: evergreen/ni big endian accel support 10117 10118 Based on 6xx/7xx patches from Cédric Cano. 10119 10120 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10121 10122commit e8dc728a549323f1babe337b9d42ad504af1ca39 10123Author: Cédric Cano <ccano@interfaceconcept.com> 10124Date: Fri Feb 11 17:06:47 2011 -0500 10125 10126 ums: atombios endian fixes 10127 10128 agd5f: fix a few more cases. 10129 10130 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10131 10132commit 48ffad957f1dbca909515ffa00629f4caa68706b 10133Author: Cédric Cano <ccano@interfaceconcept.com> 10134Date: Fri Feb 11 17:00:31 2011 -0500 10135 10136 kms: 6xx/7xx big endian accel support 10137 10138 agd5f: minor cleanups 10139 10140 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10141 10142commit 151b22bd7c3b1002a7261538611fb2b468815c86 10143Author: Cédric Cano <ccano@interfaceconcept.com> 10144Date: Fri Feb 11 16:33:16 2011 -0500 10145 10146 kms/6xx+: endian swap cursor uploads 10147 10148 agd5f: remove kms surface for cursor since 10149 we now do the swap in the cursor load to avoid 10150 breaking pre-6xx cards with BE. 10151 10152 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10153 10154commit 309fac041f998e20338d320aabaeda5cc1bb6348 10155Author: Alex Deucher <alexdeucher@gmail.com> 10156Date: Fri Feb 11 14:32:45 2011 -0500 10157 10158 kms/exa/xv: fix alignment checking in accel state setup 10159 10160 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10161 10162commit 25fe3ec4c6dfcdf6ba72a032bc9cb2c84d84029a 10163Author: Alex Deucher <alexdeucher@gmail.com> 10164Date: Fri Feb 11 14:03:56 2011 -0500 10165 10166 kms/exa: UTS/DFS base alignment fixes 10167 10168 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10169 10170commit 3d10278ce511f5dabb68ed86ee43eaaf43585983 10171Author: Alex Deucher <alexdeucher@gmail.com> 10172Date: Fri Feb 11 13:51:11 2011 -0500 10173 10174 Xv: fix textured video alignment harder 10175 10176 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10177 10178commit a9c135809945fb372ce390f06918b9fd8e455b3f 10179Author: Alex Deucher <alexdeucher@gmail.com> 10180Date: Fri Feb 11 13:21:43 2011 -0500 10181 10182 Xv: fix textured video alignment 10183 10184 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10185 10186commit e5bd99faa3b6629a55168386d5dfa936ee4e97ae 10187Author: Alex Deucher <alexdeucher@gmail.com> 10188Date: Fri Feb 11 12:14:55 2011 -0500 10189 10190 kms: remove RADEON_TILING_SURFACE flag on front bo 10191 10192 Should have been removed in: 10193 ef9bfb262db7004bef3704e5d914687e50d3fca4 10194 10195 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10196 10197commit 92ffce1892d24a98d789aad57a4230cadb0c812a 10198Author: Alex Deucher <alexdeucher@gmail.com> 10199Date: Fri Feb 11 03:12:43 2011 -0500 10200 10201 EXA/r6xx+: properly account for height alignment in copy temp buffer 10202 10203 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10204 10205commit 39104c6e8461cf49c1bb03a18858ad75a9d98b46 10206Author: Alex Deucher <alexdeucher@gmail.com> 10207Date: Thu Feb 10 14:31:43 2011 -0500 10208 10209 remove EVERGREENSetAccelState() 10210 10211 It was a duplicate of the R600 variant. 10212 10213 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10214 10215commit 7ac3a2e0bcdadff7c7172a9f833f526b526da16b 10216Author: Alex Deucher <alexdeucher@gmail.com> 10217Date: Thu Feb 10 14:24:50 2011 -0500 10218 10219 6xx+: switch to linear aligned rather than linear general 10220 10221 linear aligned is supposedly more performant, but more 10222 importantly, linear general only works on the CB without 10223 slices. The texture blocks technically don't support 10224 linear general although, I think linear general gets 10225 upgraded to linear aligned in the hw which is why it 10226 currently works. 10227 10228 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10229 10230commit e3145801b80fd4be4cf770128876e86e89bda66f 10231Author: Alex Deucher <alexdeucher@gmail.com> 10232Date: Thu Feb 10 14:14:55 2011 -0500 10233 10234 evergreen/NI: consolidate spi setup 10235 10236 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10237 10238commit be67ded05621aff9c85525372fd119071d3278ec 10239Author: Alex Deucher <alexdeucher@gmail.com> 10240Date: Thu Feb 10 14:06:38 2011 -0500 10241 10242 6xx/7xx: consolidate spi setup 10243 10244 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10245 10246commit 2c5ae1724307e0dba5d0306fe27c1e15a7390a2f 10247Author: Alex Deucher <alexdeucher@gmail.com> 10248Date: Thu Feb 10 13:47:53 2011 -0500 10249 10250 evergreen/ni: consolidate CB state handling 10251 10252 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10253 10254commit c52d817b51e13447802fe338be2d247ffba2a669 10255Author: Alex Deucher <alexdeucher@gmail.com> 10256Date: Thu Feb 10 13:35:02 2011 -0500 10257 10258 6xx/7xx: consolidate remaining CB state 10259 10260 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10261 10262commit 5f9bc127d302404432b631e3e774192950f57b42 10263Author: Alex Deucher <alexdeucher@gmail.com> 10264Date: Thu Feb 10 03:28:01 2011 -0500 10265 10266 kms/r6xx+: clean up pitch/height alignment in EXA UTS/DFS 10267 10268commit a4899db96029acde6cd400fc0541693a487898e7 10269Author: Alex Deucher <alexdeucher@gmail.com> 10270Date: Sat Feb 5 14:11:42 2011 -0500 10271 10272 UMS: fix spelling in error message 10273 10274 Noticed by: Warren Block <wblock@wonkity.com> 10275 10276 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10277 10278commit 84450ea9221f32296c1f786a6aa126a08bb9756f 10279Author: Alex Deucher <alexdeucher@gmail.com> 10280Date: Sat Feb 5 14:05:48 2011 -0500 10281 10282 UMS/DCE3.2: fix segfault 10283 10284 atombios_pick_dig_encoder() needs to be called 10285 with a valid crtc assigned. 10286 10287 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10288 10289commit a27b5dbd9fba689f1814c0888d1b7459b6c6b1a1 10290Author: Alex Deucher <alexdeucher@gmail.com> 10291Date: Thu Feb 3 19:06:22 2011 -0500 10292 10293 Bump version post release 10294 10295 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10296 10297commit 0a1a0513a61f392580bde39cca4880f2c19abc8d 10298Author: Alex Deucher <alexdeucher@gmail.com> 10299Date: Thu Feb 3 18:09:17 2011 -0500 10300 10301 bump version for release 10302 10303 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10304 10305commit ef9bfb262db7004bef3704e5d914687e50d3fca4 10306Author: Alex Deucher <alexdeucher@gmail.com> 10307Date: Tue Feb 1 19:24:44 2011 -0500 10308 10309 kms/pre-6xx: fix pageflipping with tiling 10310 10311 Use UTS/DFS to tile/untile as appropriate for sw access on 10312 pre-6xx. 10313 10314 Fixes: 10315 https://bugs.freedesktop.org/show_bug.cgi?id=33738 10316 10317 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10318 10319commit bb16dd77321e5a64a3cb2d2ca9982117799ac1a8 10320Author: Alex Deucher <alexdeucher@gmail.com> 10321Date: Tue Feb 1 19:20:00 2011 -0500 10322 10323 kms/r6xx+: fix tiling and pageflipping harder 10324 10325 Thanks for Michel for final fix. Fixes: 10326 https://bugs.freedesktop.org/show_bug.cgi?id=33738 10327 for r6xx+ asics. A similar approach for pre-r6xx 10328 asics is pending. 10329 10330 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10331 10332commit 9c124f781049ef11a8b19894a29b7f62975b3011 10333Author: Alex Deucher <alexdeucher@gmail.com> 10334Date: Tue Feb 1 18:59:40 2011 -0500 10335 10336 radeon: remove device 0x4243 10337 10338 0x4243 is a PCI bridge, not a gpu. 10339 10340 See: 10341 https://bugs.freedesktop.org/show_bug.cgi?id=33815 10342 10343 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10344 10345commit 66eb81b62e5ae8e1d7bd44ed8a179e5ec1ca69af 10346Author: Michel Dänzer <daenzer@vmware.com> 10347Date: Mon Jan 31 12:26:20 2011 +0100 10348 10349 UMS: Slightly improve xserver version check. 10350 10351 This will ensure the xserver definitions necessary to compile the new code are 10352 there. xf86CrtcSetMode will be unresolved at runtime with a few xserver 10353 commits, but I'm afraid this is the best we can do. 10354 10355 See https://bugs.freedesktop.org/show_bug.cgi?id=33719 . 10356 10357commit 3dc28c86eb57cac819e4ed650acfe1f7df6ef565 10358Author: Alex Deucher <alexdeucher@gmail.com> 10359Date: Thu Jan 27 13:09:26 2011 -0500 10360 10361 ums: fix build against xserver >= 1.10 10362 10363 xf86CrtcSetMode was removed in: 10364 http://cgit.freedesktop.org/xorg/xserver/commit/?id=8b35118c03590a7ad3786d3284bafb3f40fcb8cc 10365 10366commit f58e1354b78bf6b70120bddfe1566da3b0723f72 10367Author: Jerome Glisse <jglisse@redhat.com> 10368Date: Wed Jan 26 16:13:30 2011 -0500 10369 10370 dp: fix displayport support by syncing with KMS code 10371 10372 Warning the dp clock value are divided by 10 in ddx (10 times 10373 bigger than kernel value) this is somethings very picky. 10374 10375 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 10376 10377commit 0f90133ac61c5e1d80b6fa9256e079349940029f 10378Author: Alex Deucher <alexdeucher@gmail.com> 10379Date: Mon Jan 24 12:40:22 2011 -0500 10380 10381 vbo: warn if we add a bo with a pointer to the wait list 10382 10383commit 42529603ecf86fcfd0a8f3495d9db9ce1ee9b4c3 10384Author: Alex Deucher <alexdeucher@gmail.com> 10385Date: Mon Jan 24 12:30:56 2011 -0500 10386 10387 vbo: always flush the cbuf bo in flush_indirect() 10388 10389 Always flush the cbuf in case we end up with a cbuf mapped in 10390 Prepare*(), but never end up issuing a draw call since the cbuf 10391 may be in use by multiple ops. 10392 10393 The CS check for the cbuf bo is no longer necessay because 10394 the cbuf bo is NULL at this point due to the radeon_vbo_put() 10395 just prior. 10396 10397 Fixes: 10398 https://bugs.freedesktop.org/show_bug.cgi?id=32188 10399 10400commit 9b0e03e98ab739efb4031f81cc4a1a50b3d87a42 10401Author: Alex Deucher <alexdeucher@gmail.com> 10402Date: Mon Jan 24 00:13:25 2011 -0500 10403 10404 vbo: avoid extra cs revalidation in vbo alloc 10405 10406 Make sure we have a valid bo before revalidating. 10407 10408commit fadee0409a8e13b78bbccb83dd70f590fee23d57 10409Author: Alex Deucher <alexdeucher@gmail.com> 10410Date: Fri Jan 21 17:30:28 2011 -0500 10411 10412 vbo: balance radeon_vbo_get() and radeon_vbo_put() 10413 10414 bo_ref() and bo_map() in radeon_vbo_get() 10415 bo_unmap() and bo_unref() in radeon_vbo_put() 10416 10417 rather than doing the bo_map() separately in 10418 radeon_vbo_space(). 10419 10420commit af7d81625a8cf873e6efc881489b3eda9861bd03 10421Author: Alex Deucher <alexdeucher@gmail.com> 10422Date: Fri Jan 21 17:01:25 2011 -0500 10423 10424 evergreen: use vb_offset rather than vb_start_op for cbuf offset 10425 10426commit 4817fac5f728b777939e2e2bee16b842c9e1367b 10427Author: Alex Deucher <alexdeucher@gmail.com> 10428Date: Fri Jan 21 15:59:32 2011 -0500 10429 10430 vbo: remove useless radeon_vb_discard() 10431 10432 We already reset vb_start_op to -1 in the 10433 UMS/KMS ib discard functions. 10434 10435commit 0a03f03a65aad925ba2d9c76b1d3356184607bf9 10436Author: Alex Deucher <alexdeucher@gmail.com> 10437Date: Thu Jan 20 17:07:32 2011 -0500 10438 10439 evergreen/ni: fix Xv artifacts 10440 10441 Port of the 6xx/7xx fix to evergreen. Bad texture size 10442 for texture cache flush. 10443 10444commit f41cf83750ba9a2e0797fdb21ea9104b85ce53aa 10445Author: Marton Balint <cus@fazekas.hu> 10446Date: Thu Jan 20 17:04:56 2011 -0500 10447 10448 r6xx/7xx: fix Xv artifacts 10449 10450 bad textures size for cache flushes. 10451 10452 Fixes: 10453 https://bugs.freedesktop.org/show_bug.cgi?id=22007 10454 10455commit 57fbddfc21d8c6794f378489b764cc2a0ad4a48c 10456Author: Michel Dänzer <daenzer@vmware.com> 10457Date: Tue Jan 18 16:07:52 2011 +0100 10458 10459 Fix crash freeing KMS video memory. 10460 10461 Where's that brown paper bag? :} 10462 10463commit 6548bb9836253c586023ffe5ad1497ddabaa50fc 10464Author: Michel Dänzer <daenzer@vmware.com> 10465Date: Tue Jan 18 10:23:41 2011 +0100 10466 10467 Fix KMS textured video leaks (bug #33193). 10468 10469 v2: Fix radeon_legacy_free_memory() argument type error pointed out by 10470 Marton Balint, refactor video memory freeing logic into helper function. 10471 10472 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=33193 . 10473 10474commit edc3496b55577ee8509ddd9188e6f2bcdf7169a1 10475Author: Dave Airlie <airlied@redhat.com> 10476Date: Mon Jan 17 07:55:29 2011 +1000 10477 10478 radeon: add back flush in block handler. 10479 10480 this is also needed for certain things like stipple rendering. 10481 10482 Signed-off-by: Dave Airlie <airlied@redhat.com> 10483 10484commit 63d7dece3f4be8b14012b3a2bedd850831c0437b 10485Author: Alex Deucher <alexdeucher@gmail.com> 10486Date: Thu Jan 13 11:24:26 2011 -0500 10487 10488 rs880 fix typo in HD 4250 string 10489 10490 Noticed by Nigel Taylor 10491 10492 Fixes: 10493 https://bugs.freedesktop.org/show_bug.cgi?id=33057 10494 10495commit c5b3db18d888552328e9718ea022794fc5bde352 10496Author: Alex Deucher <alexdeucher@gmail.com> 10497Date: Tue Jan 11 16:21:18 2011 -0500 10498 10499 kms: fix pitch aligment for scanout 10500 10501 Display has slightly stricter pitch alignment requirements 10502 than other blocks. Factor that in when aligning pitch. 10503 10504 Fixes: 10505 https://bugs.freedesktop.org/show_bug.cgi?id=32997 10506 10507commit bbd7adce889359b5eb3239b73e904b3ede283e12 10508Author: Alex Deucher <alexdeucher@gmail.com> 10509Date: Tue Jan 11 15:41:03 2011 -0500 10510 10511 radeon: fix yet another pitch align 10512 10513commit af2e6d7d2f1b3d8f8f6b0acfb2b7b0cfaff7bcdb 10514Author: Alex Deucher <alexdeucher@gmail.com> 10515Date: Tue Jan 11 14:42:17 2011 -0500 10516 10517 radeon: fix pitch align in pageflip code 10518 10519commit e5d0a400d08da2358fac9c2ad12042f125525736 10520Merge: 0e432dff9 e27e9b4e5 10521Author: Alex Deucher <alexdeucher@gmail.com> 10522Date: Mon Jan 10 18:16:49 2011 -0500 10523 10524 Merge branch 'kms-pflip' of git+ssh://git.freedesktop.org/git/xorg/driver/xf86-video-ati 10525 10526commit 0e432dff9e06a183acaeb20db29cbd03ff0f4b82 10527Author: Alex Deucher <alexdeucher@gmail.com> 10528Date: Thu Jan 6 20:56:45 2011 -0500 10529 10530 NI: add pci ids 10531 10532commit 34d61785b9bc13287bd7ab8bdd8a1b99a3df4440 10533Author: Alex Deucher <alexdeucher@gmail.com> 10534Date: Tue Nov 23 20:54:57 2010 -0500 10535 10536 Add EXA/Xv acceleration support for NI chips 10537 10538commit 97322c00faf08461b941edf97efe86d8b082a0ce 10539Author: Alex Deucher <alexdeucher@gmail.com> 10540Date: Wed Nov 24 12:18:04 2010 -0500 10541 10542 Bail on NI cards with UMS 10543 10544 NI cards require KMS. 10545 10546commit c678b79f99238473df6ff1bedbae840950cdf88c 10547Author: Alex Deucher <alexdeucher@gmail.com> 10548Date: Tue Nov 23 20:45:25 2010 -0500 10549 10550 add NI family ids 10551 10552commit e27e9b4e50ad42885ad2e25be897cdf29aa59712 10553Author: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 10554Date: Sun Dec 19 03:56:17 2010 +0100 10555 10556 ddx/ati: Bugfix for pageflip consistency check. 10557 10558 target_msc wasn't passed from vblank event handler 10559 to pageflip routine, due to a missing initalization. 10560 Now fixed. 10561 10562 Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 10563 10564commit 30c5f217f91ed1ee15b5fff0002521d98682efb7 10565Author: Alex Deucher <alexdeucher@gmail.com> 10566Date: Fri Dec 17 13:40:40 2010 -0500 10567 10568 exa/dfs/ontario: always use the CPU for DFS for now 10569 10570 There are issues with rendering to gart memory that haven't 10571 been sorted out completely yet. 10572 10573commit e97079c1b621433b4aa3ecfdd2aa0884aac5df76 10574Author: Alex Deucher <alexdeucher@gmail.com> 10575Date: Thu Dec 16 18:49:55 2010 -0500 10576 10577 DCE4.1: fix DIGx routing 10578 10579 Works more like DCE4.0. With only two DIGx blocks 10580 the routing is hardcoded based on link. 10581 This fixes blank screen issues when changing crtc 10582 routing due to incorrect crtc to dig mapping. 10583 10584 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10585 10586commit 3b1fb12dec7df4e9434b2c51568ac8d997638922 10587Author: Alex Deucher <alexdeucher@gmail.com> 10588Date: Sun Dec 12 19:16:47 2010 -0500 10589 10590 man: minor formatting fixes 10591 10592commit b1af52e522158788a9613d6f1a2309a3987ad59f 10593Author: Alex Deucher <alexdeucher@gmail.com> 10594Date: Tue Dec 7 19:17:09 2010 -0500 10595 10596 man: add rs880 info 10597 10598commit f9bbb26dd97254b66de11bb2abd821aa293ecba5 10599Author: Matt Turner <mattst88@gmail.com> 10600Date: Thu Dec 2 17:00:11 2010 -0500 10601 10602 Add missing call to vbeFree (bug 4417) 10603 10604 Also kill unneeded vbe.h include from radeon_bios.c. 10605 10606 Signed-off-by: Matt Turner <mattst88@gmail.com> 10607 10608commit 9fc16ff3f8806971cb495e8838af1e5f9149d34f 10609Author: Alex Deucher <alexdeucher@gmail.com> 10610Date: Fri Dec 3 02:04:28 2010 -0500 10611 10612 radeon: remove duplicated define 10613 10614 xfixedtofloat macro was defined several times 10615 10616commit f074762fff4253a457cb48519bb33a2e2c90f8b9 10617Author: Alex Deucher <alexdeucher@gmail.com> 10618Date: Fri Dec 3 01:25:41 2010 -0500 10619 10620 radeon/exa: fallback for non-affine transforms 10621 10622 We need to implement a texture lookup with perspective 10623 divide for non-affine transforms. For now just fallback. 10624 Should fix: 10625 https://bugs.freedesktop.org/show_bug.cgi?id=31799 10626 although it appears either EXA or the xserver gets this 10627 wrong too. 10628 10629commit 774b114baa932078f8bf91ea302850a8c54c2c07 10630Author: Nicolas Joly <njoly@pasteur.fr> 10631Date: Thu Dec 2 16:02:44 2010 -0500 10632 10633 ati: remove leftover word in ati.man 10634 10635 fixes: 10636 https://bugs.freedesktop.org/show_bug.cgi?id=28297 10637 10638commit 69639ef377a9d6701cdef902f8a1c5e0b58cf833 10639Author: Alex Deucher <alexdeucher@gmail.com> 10640Date: Thu Dec 2 02:12:24 2010 -0500 10641 10642 radeon/kms: pageflipping man page updates 10643 10644commit f48af8a6cfa1ac665f07b8f9712e94b77bc4f5e9 10645Author: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 10646Date: Tue Nov 23 04:41:53 2010 +0100 10647 10648 ddx/ati: Fix reporting of pageflip completion events on multi-head. 10649 10650 When a drawable is page-flipped on multiple crtc's (fullscreen 10651 drawable on mirror-mode or multi-head x-screen), only one pageflip 10652 event is finally delivered, after the last participating crtc signals 10653 flip completion, this to avoid visual corruption. 10654 10655 Old code returned vblank count and timestamps of flip completion 10656 of this last crtc, instead of the values of the "master crtc", the 10657 one that was used for initially scheduling/triggering the pagflip 10658 via vblank events. (master = radeon_dri2_drawable_crtc(drawable)) 10659 10660 This patch makes sure that the pageflip completion values of the 10661 "master" crtc are returned, otherwise client applications will 10662 get confused by the random (msc, ust) values returned by whichever 10663 crtc was the last to complete its flip. Without this, the returned 10664 values change randomly and jump forward and backward in time and 10665 count. 10666 10667 The patch also implements a consistency check on returned vblank 10668 count values of pageflip completion. Impossible values are detected, 10669 a x-warning is logged and returned (msc,ust) values are marked invalid, 10670 so clients could perform error handling. Such a warning would indicate 10671 bugs in the pageflip completion routine of future kms drivers or the 10672 ddx to aid driver debugging. 10673 10674 Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 10675 10676commit 122536ee0aeb1eef1a9d80d5e464dcb423dc2837 10677Author: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 10678Date: Mon Nov 22 04:11:07 2010 +0100 10679 10680 ddx/ati: Add option "SwapbuffersWait" to control vsync of DRI2 swaps. 10681 10682 A new optional kms driver option "SwapbuffersWait" is defined 10683 for xorg.conf, which defaults to "on". If "on", DRI2 bufferswaps 10684 will be synchronized to vsync, otherwise not. 10685 10686 This currently only affects copy-swaps, not pageflipped swaps. 10687 It also requires a swap_interval setting of zero by the OpenGL 10688 client. 10689 10690 Ideally, we'd provide a way for dri2 to pass the current swap 10691 interval to the ddx so we could change this dynamically. 10692 10693 Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 10694 10695commit 0de680730294bd623f6b3e189faa7b88a09d3a2a 10696Author: Alex Deucher <alexdeucher@gmail.com> 10697Date: Wed Nov 17 17:39:39 2010 -0500 10698 10699 radeon/kms: add pageflip support 10700 10701 requires radeon drm 2.8.0 or higher 10702 10703 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10704 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 10705 10706commit fccdca8db34010f566bd068c74cdef0f4a8cb7f5 10707Author: Alex Deucher <alexdeucher@gmail.com> 10708Date: Wed Nov 17 17:37:25 2010 -0500 10709 10710 radeon/kms: allow tiled front buffer on 6xx/7xx 10711 10712 Use UTS/DFS to tile/untile as appropriate for sw access. 10713 Also enables pageflipping with tiling enabled. 10714 10715commit 035f7f3ab529ca19b853066792af8a23d08a0f53 10716Author: Dave Airlie <airlied@redhat.com> 10717Date: Wed Dec 1 10:23:14 2010 +1000 10718 10719 radeon: add some missing math.h and assert.h includes 10720 10721commit 0f924e83d7be6fca7e024acdf973089efdf3b627 10722Author: Dave Airlie <airlied@redhat.com> 10723Date: Wed Dec 1 10:23:33 2010 +1000 10724 10725 radeon: make sure EXA is off when EXA is disabled with --disable-exa 10726 10727commit e142e55c56d8440d7185b594c696ec9eeb699cda 10728Author: Alex Deucher <alexdeucher@gmail.com> 10729Date: Tue Nov 30 15:51:21 2010 -0500 10730 10731 evergreen: set default group_bytes to 256 10732 10733 512 seems to cause issues on certain cards. 10734 10735 Fixes: 10736 https://bugs.freedesktop.org/show_bug.cgi?id=31724 10737 10738commit 90f831361844f1b80b3f6bb718ff5ac584d73d48 10739Author: Alex Deucher <alexdeucher@gmail.com> 10740Date: Mon Nov 29 18:09:05 2010 -0500 10741 10742 evergreen: use dot4 for transforms 10743 10744commit 3cae361d0448b6e231c80f53d64bdbbdd74dc4cf 10745Author: Alex Deucher <alexdeucher@gmail.com> 10746Date: Mon Nov 29 17:44:47 2010 -0500 10747 10748 6xx/7xx: clean up gpr/const handling in shaders 10749 10750commit d9bcac516f2a810acb300b29169e56a2df0b47ac 10751Author: Alex Deucher <alexdeucher@gmail.com> 10752Date: Mon Nov 29 17:23:30 2010 -0500 10753 10754 r6xx/r7xx use dot4 for transforms 10755 10756commit 5d3f33729be0639cef17372345b2dab6127e39d9 10757Author: Alex Deucher <alexdeucher@gmail.com> 10758Date: Wed Nov 24 10:59:09 2010 -0500 10759 10760 radeon/kms: fix output enumeration 10761 10762 If a system has a mix of DVI-I, and DVI-D connectors, or a mix 10763 of HDMI-A and HDMI-B connectors, the enumeration was per 10764 connector type, but the naming was the same for multiple 10765 connectors (DVI-[0-n] for DVI-I, -D, -A, and HDMI-[0-n] 10766 for HDMI-A, and -B), you'd end up with multiple connectors 10767 with the same name. Fix this by tracking the number of 10768 connectors across similar connector types. 10769 10770commit dc470b5382ffa27bd4a673b21fb6b978d3716ca0 10771Author: Alex Deucher <alexdeucher@gmail.com> 10772Date: Tue Nov 23 01:46:15 2010 -0500 10773 10774 radeon/man: divide driver options into UMS and KMS 10775 10776commit 4cbb7100df4b983d77c20a77431cf2f5101bde36 10777Author: Alex Deucher <alexdeucher@gmail.com> 10778Date: Tue Nov 23 01:22:04 2010 -0500 10779 10780 update man page to include ontario 10781 10782commit 718dfd443897d70d97fc19afddfbcfe1e39a9c59 10783Author: Alex Deucher <alexdeucher@gmail.com> 10784Date: Mon Nov 22 13:11:50 2010 -0500 10785 10786 Ontario: add workaround for small pixmap issues with DFS 10787 10788 On past asics these kind a problems were usually due to a bad 10789 backend map config. I suspect in this case, and the previous 10790 cases it may be related to using linear buffers rather than 10791 tiled. 10792 10793 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10794 10795commit 61c97b0812c820564c7bc9d64a998c109bcd46b3 10796Author: Alex Deucher <alexdeucher@gmail.com> 10797Date: Mon Nov 22 13:10:16 2010 -0500 10798 10799 Add EXA/Xv acceleration support for Ontario Fusion APUs 10800 10801 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10802 10803commit 0659f35dfbeda7dd3ff58da1deb6484fabadc332 10804Author: Alex Deucher <alexdeucher@gmail.com> 10805Date: Mon Nov 22 13:07:00 2010 -0500 10806 10807 Add Ontario fusion APU pci ids 10808 10809 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10810 10811commit fa0cec331e6c94cebf82270dcd6c179b15851c5e 10812Author: Alex Deucher <alexdeucher@gmail.com> 10813Date: Mon Nov 22 12:56:45 2010 -0500 10814 10815 ontario: add UMS modesetting support 10816 10817 Accel not enabled yet. 10818 10819 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 10820 10821commit 7ad1f01573e930aa6238eaeb0e11708ae548dfce 10822Author: Jerome Glisse <jglisse@redhat.com> 10823Date: Fri Nov 19 10:27:40 2010 -0500 10824 10825 evergreen: fix typo for fetch resource offset 10826 10827 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 10828 10829commit 3455a3b58532ea3ad901a317126968ab6cbb21b7 10830Author: Alex Deucher <alexdeucher@gmail.com> 10831Date: Wed Nov 17 17:32:41 2010 -0500 10832 10833 radeon/kms: fix buffer base alignment for tiling 10834 10835 On r6xx+, 2D tiling can require larger than 4k base alignment. 10836 10837commit beb7fecd0191e38fb238134ba612985062cf9770 10838Author: Dave Airlie <airlied@redhat.com> 10839Date: Mon Nov 15 11:59:58 2010 +1000 10840 10841 dce4: improve the VT switching code. 10842 10843 add in lots more blocks of regs to save/restore 10844 10845commit 51e51f864a691e51b5193d102398ced667e0beb1 10846Author: Dave Airlie <airlied@redhat.com> 10847Date: Thu Nov 11 18:48:40 2010 +1000 10848 10849 evergreen: remove warnings since last commit 10850 10851commit abbf73ee990512ac16ca77e8bb23288495e1f9f4 10852Author: Dave Airlie <airlied@redhat.com> 10853Date: Thu Nov 11 09:38:15 2010 +1000 10854 10855 evergreen: add UMS VT switch support. 10856 10857 This isn't perfect, but it brings back text VTs here on the 10858 DAC and DVI outputs. 10859 10860commit e1dfaf93d06bc5eafdbc2e1823d19204ce8f242c 10861Author: Dave Airlie <airlied@redhat.com> 10862Date: Thu Nov 11 09:15:27 2010 +1000 10863 10864 avivo: use arrays to store the crtc/pll info for save/restore 10865 10866 this is just prep work for evergreen VT save/restore 10867 10868commit 0c2834e67df6e143a7d15f373faaddca1dda6b18 10869Author: Michel Dänzer <daenzer@vmware.com> 10870Date: Fri Nov 5 09:30:40 2010 +0100 10871 10872 Fix incorrect CS size in one textured video path. (Bug #31364) 10873 10874 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=31364 . 10875 10876commit 2a7b080727f61559cf8444c93a5088f4155a8fdf 10877Author: Alex Deucher <alexdeucher@gmail.com> 10878Date: Fri Oct 29 16:56:45 2010 -0400 10879 10880 radeon/kms: more alignment fixes 10881 10882commit bf60af579382a0d48f7a65a4cec88759cc8b683c 10883Author: Dave Airlie <airlied@redhat.com> 10884Date: Thu Oct 28 13:06:54 2010 +1000 10885 10886 dri2: reference count the client privates key/callback 10887 10888 This lets multi-screen work better, but still having issues after server 10889 recycle, but it doesn't crash at least. 10890 10891commit fb22d0c06a7dc42216230e198ff443d8035e9d21 10892Author: Dave Airlie <airlied@redhat.com> 10893Date: Thu Oct 28 11:10:03 2010 +1000 10894 10895 radeon: only register block handler once per fd. 10896 10897 should fix https://bugs.freedesktop.org/show_bug.cgi?id=29726 10898 10899 the problem is of course the second head instance tries to access the 10900 fd and fails, however I think this might break syncing on the second 10901 head but not sure, but its better than just hanging up the X server 10902 10903commit 040140b5d7cbeaf5d83687bc87e4a236ff9f0cbd 10904Author: Alex Deucher <alexdeucher@gmail.com> 10905Date: Wed Oct 27 20:16:17 2010 -0400 10906 10907 drm/radeon: unify fb bo alignment harder 10908 10909 More duplicated paths discoved... 10910 10911commit f07f9b7b61c05f2de1d61bb0e2f71bd017c8d36a 10912Author: Alex Deucher <alexdeucher@gmail.com> 10913Date: Wed Oct 27 12:37:42 2010 -0400 10914 10915 kms/radeon: unify fb bo alignment handling 10916 10917 Previously there were 3 different paths with what should 10918 have had duplicated code: 10919 - EXACreatePixmap2 10920 - Initial front buffer creation 10921 - Randr resize 10922 10923 This patch attempts to unify the alignment across all 3. 10924 10925 This may fix tiling issues in some cases and should make 10926 buffer pitches match for pageflipping. 10927 10928commit d31046ba6c8eee9b7decc3875697d37c38bc38f3 10929Author: Michel Dänzer <daenzer@vmware.com> 10930Date: Wed Oct 27 12:12:04 2010 +0200 10931 10932 Re-use result of GetScreenPixmap call. 10933 10934 Fixes compile warning due to local variable ppix being unused when building 10935 against current xserver Git. 10936 10937commit 4dd7f835c17b00707f1d8d4e36a24380ba52761e 10938Author: Michel Dänzer <daenzer@vmware.com> 10939Date: Wed Oct 27 12:09:03 2010 +0200 10940 10941 Fix another stray xfree() call. 10942 10943commit f815e9c3c61b1ed73e7dd9383587efcdd8dfc07f 10944Author: Michel Dänzer <daenzer@vmware.com> 10945Date: Wed Oct 27 11:29:12 2010 +0200 10946 10947 Don't try to delete DRI2 event list entries that were never added. (Bug #31086) 10948 10949 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=31086 . 10950 10951commit df92245e1fd0ead2581b675cd96ac8bcc8cb65d7 10952Author: Alex Deucher <alexdeucher@gmail.com> 10953Date: Sun Oct 24 14:27:27 2010 -0400 10954 10955 r100: add some missing bits in SE_COORD_FMT 10956 10957 Texture coordinates work fine with or without these, 10958 but this should be more correct I think although 10959 I don't think it matters since we aren't sending w 10960 anyway. 10961 10962commit fc7aa6e3bbb86eaa2f34b827629f7e82cde65ac7 10963Author: Alex Deucher <alexdeucher@gmail.com> 10964Date: Sun Oct 24 11:56:48 2010 -0400 10965 10966 UMS/radeon: fix typo in restore palette 10967 10968commit 5efdf514ab0e5439114e8dd7a49105812155eb69 10969Author: Alex Deucher <alexdeucher@gmail.com> 10970Date: Sun Oct 24 11:45:19 2010 -0400 10971 10972 UMS/radeon: fix screen dimming on VT switch 10973 10974 Save and restore the palettes on VT switch. The restore 10975 has to be done after the vga restore to work properly as 10976 determined by Jonathan Kollasch. 10977 10978 Fixes: 10979 https://bugs.freedesktop.org/show_bug.cgi?id=18407 10980 10981commit 1e3ede62c39307d0210cb2044badd619a4f44fec 10982Author: Alex Deucher <alexdeucher@gmail.com> 10983Date: Sun Oct 24 11:24:07 2010 -0400 10984 10985 UMS/atom: default to DVI mode if we don't have monitor hdmi info 10986 10987 Fixes: 10988 https://bugs.freedesktop.org/show_bug.cgi?id=30330 10989 10990commit 51f26a7f01b68e373867c03881868ca8830664d6 10991Author: Mark Schreiber <mark7@alumni.cmu.edu> 10992Date: Sat Apr 10 10:55:08 2010 -0700 10993 10994 Correct spelling and grammar 10995 10996commit 0f575984594e6977b4a7cb794dad8c6e56c850f1 10997Author: Alex Deucher <alexdeucher@gmail.com> 10998Date: Tue Oct 19 19:21:39 2010 -0400 10999 11000 Xv: limit overlay to 2047x2047 11001 11002 Fixes: 11003 https://bugs.freedesktop.org/show_bug.cgi?id=15391 11004 11005commit 0f184630cd4760199430320215e4280438fc4ce5 11006Author: Alex Deucher <alexdeucher@gmail.com> 11007Date: Tue Oct 19 17:13:06 2010 -0400 11008 11009 radeon: fix compilation on xserver >= 1.10 11010 11011 Fixes: 11012 https://bugs.freedesktop.org/show_bug.cgi?id=30451 11013 11014 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11015 11016commit 179b0e15924ae2757efaece6488ae835ba6436e3 11017Author: Dave Airlie <airlied@redhat.com> 11018Date: Tue Oct 19 13:37:22 2010 +1000 11019 11020 r600: for 2d tiles we need to align to channels no banks. 11021 11022commit 809cacbc23ab37aa7952665f9b2e37c03a34f173 11023Author: Adam Jackson <ajax@redhat.com> 11024Date: Mon Oct 18 12:32:11 2010 -0400 11025 11026 kms: Let the server know if we got more than 1 EDID block 11027 11028 Otherwise things like xf86MonitorIsHDMI() won't work right. 11029 11030 Signed-off-by: Adam Jackson <ajax@redhat.com> 11031 11032commit 042037e485396981cf4a420c247104ba5d016ca0 11033Author: Alex Deucher <alexdeucher@gmail.com> 11034Date: Thu Oct 14 17:00:51 2010 -0400 11035 11036 evergreen: requires same cursor fix as avivo chips 11037 11038commit 0f37458c253aec46d98db6661c4def711df8fc17 11039Author: Adam Jackson <ajax@redhat.com> 11040Date: Thu Oct 14 14:37:34 2010 -0400 11041 11042 Don't allow the config file to override BIOS location 11043 11044 Signed-off-by: Adam Jackson <ajax@redhat.com> 11045 11046commit 2918df3a312fc600e83f84c298b9653d2eb3a8d5 11047Author: Alex Deucher <alexdeucher@gmail.com> 11048Date: Wed Oct 13 17:16:04 2010 -0400 11049 11050 evergreen: increase XV_CRTC Xv attribute range 11051 11052 Cover all 6 possible crtcs. 11053 11054commit 0d1f9fd03d0196dda5c1ce34b2e68b007c1874f2 11055Author: Alex Deucher <alexdeucher@gmail.com> 11056Date: Sun Oct 10 12:06:22 2010 -0400 11057 11058 DCE3+: UMS modesetting fixes 11059 11060 - add mising LVTMA case statement for DCE3.0 dig encoder 11061 - some DCE4 systems have EN/DISABLE_OUTPUT actions 11062 11063commit acf4024aa0ef192355b2bd8281507b1c2e6b7de4 11064Author: Alex Deucher <alexdeucher@gmail.com> 11065Date: Fri Oct 8 16:30:36 2010 -0400 11066 11067 ATOM DDC fixes for UMS 11068 11069 - Header size was already subtraced from table size 11070 - Only hw capable ddc pads are shared with aux 11071 11072commit 74fd2b91477106a26a2d9fb4b11c885910996041 11073Author: Alex Deucher <alexdeucher@gmail.com> 11074Date: Thu Oct 7 17:33:00 2010 -0400 11075 11076 DCE3.2 UMS: fix duallink 11077 11078 Should fix: 11079 https://bugs.freedesktop.org/show_bug.cgi?id=30686 11080 11081commit 8c544272f218fd8bd93a898b96afedc2498b27bb 11082Author: Alex Deucher <alexdeucher@gmail.com> 11083Date: Thu Oct 7 17:25:47 2010 -0400 11084 11085 fix the non-kms build 11086 11087 Should fix: 11088 https://bugs.freedesktop.org/show_bug.cgi?id=30685 11089 11090commit 81360adffb2a66b9a95a38671f9227a9718c9841 11091Author: Oldřich Jedlička <oldium.pro@seznam.cz> 11092Date: Mon Sep 6 21:14:52 2010 +0200 11093 11094 radeon: proper DRI2 pending events handling when client gone. (v6) 11095 11096 Properly handle asynchronous DRI2 events for disconnected clients. 11097 Track client's pending requests and mark them as invalid when the 11098 client disconnects. 11099 11100 This is based on the version from Alban Browaeys in bug #29065. 11101 11102 v1 (Alban Browaeys): Based upon a detailed explanation from Oldřich 11103 Jedlička and comments from Christopher James Halse Rogers. 11104 on http://lists.x.org/archives/xorg-driver-ati/2010-August/016780.html . 11105 11106 v2: Updated version to apply on master. Removed unnecessary 11107 client_index field from _DRI2FrameEvent. Added freeing/removing from 11108 list to failed paths of radeon_dri2_schedule_wait_msc and 11109 radeon_dri2_schedule_swap. 11110 11111 v3: Adopt to older xorg-server that doesn't have dixRegisterPrivateKey. 11112 11113 v4: Conditional include of list.h, unreachable return removed. 11114 11115 v5: Distribute list.h as xorg_list.h, remove xorg-server version check. 11116 Use the version from xorg-server when available (checked in 11117 configure.ac). 11118 11119 v6: Removed xorg_list.h, made DRI2 scheduling features dependent on 11120 list.h presence. 11121 11122commit adee138f007e26307f1aab7f8fe066150c12e55d 11123Author: Jesse Adkins <jesserayadkins@gmail.com> 11124Date: Tue Sep 28 13:29:50 2010 -0700 11125 11126 Purge cvs tags. 11127 11128 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 11129 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 11130 11131commit f1bf9cbb471d60ee4b0422bc4e896e3b86f4d75f 11132Author: Alex Deucher <alexdeucher@gmail.com> 11133Date: Wed Oct 6 11:44:42 2010 -0400 11134 11135 radeon: man page updates 11136 11137 - tv-out is not supported on carding using 11138 the rage theatre chip for tv-out 11139 - add evergreen cards to the list 11140 11141commit 0921ecc1c751df0dd56e0b1d0b78ab53d7164904 11142Author: Alex Deucher <alexdeucher@gmail.com> 11143Date: Mon Oct 4 12:34:47 2010 -0400 11144 11145 evergreen: use EXACreatePixmap2 if available 11146 11147commit 5bdb6434975584eef90eb9e5955b9c2a14b7f327 11148Author: Alex Deucher <alexdeucher@gmail.com> 11149Date: Mon Oct 4 12:31:51 2010 -0400 11150 11151 evergreen: port Karl's UTS/DFS changes to evergreen 11152 11153commit be8f45cbd313b68ad663f303c64edb4525b8f981 11154Merge: e843faf35 ff5f466e6 11155Author: Alex Deucher <alexdeucher@gmail.com> 11156Date: Mon Oct 4 12:36:56 2010 -0400 11157 11158 Merge branch 'evergreen_accel' of git+ssh://git.freedesktop.org/git/xorg/driver/xf86-video-ati 11159 11160commit ff5f466e6ca0ee541a255facff6c9198976dfd52 11161Author: Alex Deucher <alexdeucher@gmail.com> 11162Date: Fri Oct 1 15:46:43 2010 -0400 11163 11164 evergreen/dri2: fix depth allocation for depth+stencil 11165 11166 evergreen uses separate allocations for depth and stencil, 11167 so to handle that, create a depth buffer large enough to 11168 handle both. This is required for using the stencil 11169 buffer in mesa. 11170 11171commit e843faf355c864beab81e74f0e39f8ad53d4c2bf 11172Author: Alex Deucher <alexdeucher@gmail.com> 11173Date: Thu Sep 30 19:30:35 2010 -0400 11174 11175 radeon: fix fbLocation for >32 bit MC addresses 11176 11177 If the fbLocation was at an address >32 bits, we'd fail. 11178 Change fbLocation to uint64_t and properly cast when needed. 11179 11180commit 886febc882053e09294225e85b102f965041b62b 11181Author: Alex Deucher <alexdeucher@gmail.com> 11182Date: Thu Sep 30 19:20:17 2010 -0400 11183 11184 r6xx: fix bad mask when setting up HDP_NONSURFACE_BASE 11185 11186 This fails for MC addresses >32 bits 11187 11188commit 16589b862feb796e15dc6e471d25604a4f5e7c1c 11189Author: Alex Deucher <alexdeucher@gmail.com> 11190Date: Mon Sep 27 18:28:43 2010 -0400 11191 11192 bump version post release 11193 11194commit cc5005af61f45a3552f7358dc5aa711e42f5af54 11195Author: Alex Deucher <alexdeucher@gmail.com> 11196Date: Mon Sep 27 18:20:53 2010 -0400 11197 11198 bump version for release 11199 11200commit 7f8820fcec8c90bf2f823170bd08a23e8b4ff7af 11201Author: Michael Cree <mcree@orcon.net.nz> 11202Date: Mon Sep 27 13:39:12 2010 -0400 11203 11204 Fix some unaligned 32bit accesses in the AtomBios code. 11205 11206 On the Alpha architecture unaligned 32bit accesses incur a software 11207 trap to the kernel and pollute the kernel logs. Fixed by use of the 11208 ldl_u() interface. 11209 11210 Signed-off-by: Michael Cree <mcree@orcon.net.nz> 11211 11212commit c4f834cdfbe96aa47ac5fb039f9dd7aa9730c8a3 11213Author: Nicolas Reinecke <nr@das-labor.org> 11214Date: Mon Sep 27 13:33:55 2010 -0400 11215 11216 radeon: Convert remaining x(c)alloc/xfree to m/calloc/free. 11217 11218 Fixes deprecation warnings missed out by 11219 f7a91ece264af9f3fd2fc18e99aefcda93ce9f5c 11220 11221commit f8fb9312d791af1f77020e8c2d35bb30841ed9aa 11222Author: Karl Tomlinson <karlt+@karlt.net> 11223Date: Sun Aug 22 22:46:33 2010 +1200 11224 11225 RADEONPrepareAccess_CS: fallback to DFS when pixmap is in VRAM 11226 11227 This avoids costly CPU VRAM reads and lets EXA manage a system memory cache 11228 of the portions of pixmaps needed for unaccelerated operations. 11229 11230 https://bugs.freedesktop.org/show_bug.cgi?id=27139 11231 11232commit 35c4ff936601ee083f51510a5192fb97d622a483 11233Author: Karl Tomlinson <karlt+@karlt.net> 11234Date: Sun Aug 22 22:28:06 2010 +1200 11235 11236 radeon: complete UTS and DFS even when a scratch BO is not necessary 11237 11238 Turns on the big-endian paths even for little-endian systems, and adds 11239 similar paths to the r6xx/r7xx functions. 11240 11241 This makes UTS and DFS reliable, which will let PrepareAccess (with 11242 mixed pixmaps) choose to fail based on whether the pixmap is in VRAM 11243 (to avoid CPU reads). 11244 11245commit d46381a3a6bf10903803f5acaa7aa0ce06373b96 11246Author: Karl Tomlinson <karlt+@karlt.net> 11247Date: Sun Aug 22 21:02:45 2010 +1200 11248 11249 radeon: complete big endian UTS and DFS even when scratch allocation fails. 11250 11251 On big endian systems, PrepareAccess will fail when byte-swapping is 11252 required so UploadToScreen and DownloadFromScreen cannot rely on 11253 fallback to PrepareAccess. 11254 11255 When scratch BO space allocation fails, this patch merely adds simple 11256 fallback to direct CPU access without any GPU blit. This sometimes 11257 requires a CS flush even in UploadToScreen. 11258 (No allocation retry after a flush is added here.) 11259 11260commit 4ced4e1eff67946e306c0c67c9ed59dd5f3c4ba9 11261Author: Karl Tomlinson <karlt+@karlt.net> 11262Date: Sun Aug 22 20:04:42 2010 +1200 11263 11264 RADEONDownloadFromScreenCS: flush CS writes before mapping BO for read 11265 11266 If unflushed CS operations write to the pixmap BO, then these need to be 11267 flushed before mapping the BO for read. This currently only affects big 11268 endian systems and only when the operation writes to the GTT domain. 11269 11270commit a4eef8faffbb1ea2f742273ee855f4e6f992e5c8 11271Author: Karl Tomlinson <karlt+@karlt.net> 11272Date: Sat Aug 21 22:29:34 2010 +1200 11273 11274 FinishAccess_CS: set bo_mapped to FALSE on unmap 11275 11276 This is actually only necessary when PrepareAccess may behave differently on 11277 different calls with the same pixmap, which currently doesn't happen. 11278 11279 However resetting bo_mapped is necessary to let PrepareAccess (with mixed 11280 pixmaps) choose to fail based on whether the pixmap is in VRAM (to avoid CPU 11281 reads). 11282 11283commit bfebe039af0c0282d04eb6234b6e6d1e02097146 11284Author: Karl Tomlinson <karlt+@karlt.net> 11285Date: Sat Aug 21 21:44:39 2010 +1200 11286 11287 DownloadFromScreenCS: download via a scratch BO if pixmap domain is unknown 11288 11289 radeon_bo_is_busy() may return without setting the domain out-parameter. 11290 If this happens, then download via a scratch GTT BO to avoid CPU VRAM read. 11291 11292commit e26a59e9db8067882327f872e3d2d760ce4c66f3 11293Author: Mathieu Bérard <mathieu@mberard.eu> 11294Date: Thu Sep 9 19:32:38 2010 -0400 11295 11296 Fix NULL possible deref in evergreen_cp_wait_vline_sync 11297 11298commit b90cb61ccb0f4f80e0627141f223354a9371d47d 11299Author: Alex Deucher <alexdeucher@gmail.com> 11300Date: Tue Sep 7 11:51:16 2010 -0400 11301 11302 radeon: set interlaced and doublescan enabled for randr outputs 11303 11304 interlaced used to work without setting these parameters. Changes 11305 in the xserver seem to require them now. 11306 11307 Should fix: 11308 https://bugs.freedesktop.org/show_bug.cgi?id=29591 11309 11310commit 7cc0d3311f23ad569af004e0e4a0e2efbac107e5 11311Author: Alex Deucher <alexdeucher@gmail.com> 11312Date: Fri Sep 3 00:32:27 2010 -0400 11313 11314 evergreen: fix Xv 11315 11316 VS const buffer offset was wrong. 11317 11318 fixes: 11319 https://bugs.freedesktop.org/show_bug.cgi?id=29788 11320 11321commit 7c6ce4e62693f446d7d3c8a86502ccc03c0e55b0 11322Author: Alex Deucher <alexdeucher@gmail.com> 11323Date: Thu Sep 2 20:34:56 2010 -0400 11324 11325 evergreen: add additional default state 11326 11327 Note, you also need a drm patch to fix the GPU hangs: 11328 drm/radeon/kms/evergreen: fix gpu hangs in userspace accel code 11329 11330 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11331 11332commit 2b98ec1f7e931019a4ab699a56d5dfaa395946fb 11333Author: Alex Deucher <alexdeucher@gmail.com> 11334Date: Wed Sep 1 13:24:19 2010 -0400 11335 11336 r6xx/r7xx: fix typos 11337 11338 some stray - signs 11339 11340commit 966ac1be81da76c8aa4ea46b63f3ca5358a2c021 11341Author: Alex Deucher <alexdeucher@gmail.com> 11342Date: Fri Aug 27 18:22:21 2010 -0400 11343 11344 evergreen: work around bad data in some i2c tables 11345 11346 The 7th entry in a lot of evergreen i2c gpio tables is partially 11347 zeroed. Fix the entry. 11348 11349 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11350 11351commit 91f707d308d4bbf16c3d62d046cf280fef5a8f4c 11352Author: Heikki Lindholm <holin@iki.fi> 11353Date: Fri Aug 27 02:26:24 2010 -0400 11354 11355 xv: fix non-kms/non-dri Xv column ordering on big endian systems 11356 11357 Column order is wrong on big endian systems, primarly because of a 11358 bits / bytes mix up with the bpp variable. Fix tested with r100 and 11359 r300, screen depth 16 and 32 with YV12 and YUY2 (overlay, textured video), 11360 RGBA and RGBT (overlay). 11361 11362 Should fix: https://bugs.freedesktop.org/show_bug.cgi?id=29041 11363 11364 Signed-off-by: Heikki Lindholm <holin@iki.fi> 11365 11366commit 6a2c8587a4e05a8be2a2e975a6660942cfe115d6 11367Author: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 11368Date: Fri Aug 27 13:14:33 2010 +1000 11369 11370 dri2: Reference count DRI2 buffers 11371 11372 When a client calls ScheduleSwap we set up a kernel callback when the 11373 relevent vblank event occurs. However, it's possible for the client 11374 to go away between calling ScheduleSwap and the vblank event, 11375 resulting in the buffers being destroyed before they're passed to 11376 radeon_dri2_frame_event_handler. 11377 11378 Add reference-counting to the buffers and take a reference in 11379 radeon_dri2_schedule_swap to ensure the buffers won't be destroyed 11380 before the vblank event is dealt with. 11381 11382 This parallels the approach taken by the Intel DDX in commit 11383 0d2392d44aae95d6b571d98f7ec323cf672a687f. 11384 11385 Fixes: http://bugs.freedesktop.org/show_bug.cgi?id=29065 11386 11387 v2: Don't write completion events to the client if it has quit. 11388 v3: Don't try to unref the NULL buffers from a DRI2_WAITMSC event. 11389 Take a ref in schedule_swap earlier, so the offscreen fallback 11390 doesn't incorrectly destroy the buffers. 11391 11392 Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 11393 Signed-off-by: Dave Airlie <airlied@redhat.com> 11394 11395commit e9928fe036e9382fd7bc353f3f05531445f08977 11396Author: Dave Airlie <airlied@redhat.com> 11397Date: Wed Aug 25 10:42:39 2010 +1000 11398 11399 remove explicit buffer submit from copy region 11400 11401 port of 0be3e95c844247746742805830860ace9f546d99 from intel driver. 11402 11403 Remove explicit batchbuffer submit in DRI2 copyregion 11404 11405 Now that we submit from the flush callback chain, we know we'll always 11406 submit before the client receives the reply or event that blocks it from 11407 rendering the next frame. 11408 11409 Signed-off-by: Dave Airlie <airlied@redhat.com> 11410 11411commit 9f13049ddf06f6f2138851a548cfb82f12a52f42 11412Author: Dave Airlie <airlied@redhat.com> 11413Date: Wed Aug 25 08:56:37 2010 +1000 11414 11415 radeon: add correct flushing for direct rendered 11416 11417 this is a port of 69d65f9184006eac790efcff78a0e425160e95aa from the Intel 11418 driver. 11419 11420 Submit batch buffers from flush callback chain 11421 11422 There are a few cases where the server will flush client output buffers 11423 but our block handler only catches the most common (before going into select 11424 If the server flushes client buffers before we submit our batch buffer, 11425 the client may receive a damage event for rendering that hasn't happened yet 11426 11427 Instead, we can hook into the flush callback chain, which the server will 11428 invoke just before flushing output. This lets us submit batch buffers 11429 before sending out events, preserving ordering. 11430 11431 Fixes 28438: [bisected] incorrect character in gnome-terminal under compiz 11432 https://bugs.freedesktop.org/show_bug.cgi?id=28438 11433 11434 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 11435 11436 Signed-off-by: Dave Airlie <airlied@redhat.com> 11437 11438commit 141cbc283fddeb67c2a6c47a0f0f5c2aa2bfb430 11439Author: Alex Deucher <alexdeucher@gmail.com> 11440Date: Tue Aug 24 18:01:05 2010 -0400 11441 11442 evergreen Xv: remove stray line that caused a segfault 11443 11444 fixes: 11445 https://bugs.freedesktop.org/show_bug.cgi?id=29788 11446 11447commit f170dddd3538a587f2363ef5fa10a4b484e762da 11448Author: Alex Deucher <alexdeucher@gmail.com> 11449Date: Mon Aug 23 18:58:12 2010 -0400 11450 11451 evergreen: use vbo pool for constant buffers 11452 11453commit eede93b057bbdddcde5da3220a3f8c6f73784a7e 11454Author: Alex Deucher <alexdeucher@gmail.com> 11455Date: Mon Aug 23 13:16:27 2010 -0400 11456 11457 radeon: move vbo data to a separate struct 11458 11459 this way we can share the vbo code with const buffers 11460 11461commit d8abf27dbd14f5eb746c5e8b8b1436ad292d8ec6 11462Author: Alex Deucher <alexdeucher@gmail.com> 11463Date: Mon Aug 23 01:31:27 2010 -0400 11464 11465 evergreen: fix dword counts in default state 11466 11467commit 5a7c9d94733a0db1d3565447acc9f0e751db5950 11468Author: Alex Deucher <alexdeucher@gmail.com> 11469Date: Mon Aug 23 00:43:48 2010 -0400 11470 11471 radeon: fix legacy lvds dpms sequence 11472 11473 Take from my kms commit. 11474 11475 Should fix: 11476 https://bugs.freedesktop.org/show_bug.cgi?id=19459 11477 11478 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11479 11480commit 6930d2c981221757b1e11ef194809f085753a611 11481Author: Alex Deucher <alexdeucher@gmail.com> 11482Date: Fri Aug 20 16:55:21 2010 -0400 11483 11484 Add initial EXA and Xv support for evergreen 11485 11486 Based on the r6xx/r7xx code updated for evergreen. 11487 Still causes GPU hangs in some cases. We haven't 11488 tracked down why yet. Might be related to constant 11489 buffer persistence. 11490 11491 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11492 11493commit bdd41fecdb19c83c6c7b793016b61d38065dcd13 11494Author: Alex Deucher <alexdeucher@gmail.com> 11495Date: Fri Aug 20 01:17:13 2010 -0400 11496 11497 evergreen: set encoder type to DVI for HDMI 11498 11499 Fixes: 11500 http://bugs.freedesktop.org/show_bug.cgi?id=27452 11501 11502 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11503 11504commit ad8ea1f6e5fcb4f163622cf5eab953ea812b5829 11505Author: Alex Deucher <alexdeucher@gmail.com> 11506Date: Thu Aug 19 17:04:35 2010 -0400 11507 11508 DCE3+: switch pads to ddc mode when doing i2c 11509 11510 The pins for ddc and aux are shared so you need to switch the 11511 mode when doing ddc. The ProcessAuxChannel table already sets 11512 the pin mode to DP. This should fix unreliable ddc issues 11513 on DP ports using non-DP monitors. 11514 11515 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11516 11517commit fd686668289258ffaf6b81057545e50612aac6a8 11518Author: Dave Airlie <airlied@redhat.com> 11519Date: Thu Aug 12 12:59:18 2010 +1000 11520 11521 radeon: fixup non-kms build 11522 11523commit 5a9865d90c23c4ce0f46d380ea9119eac87a99eb 11524Author: Dave Airlie <airlied@redhat.com> 11525Date: Thu Aug 12 11:18:06 2010 +1000 11526 11527 evergreen: fix multi-head cursor support 11528 11529commit 8351bb9f085fde5dc47b115375efcc61adc23859 11530Author: Dave Airlie <airlied@redhat.com> 11531Date: Tue Aug 10 15:05:45 2010 +1000 11532 11533 radeon: take 8/10 encoding into a/c on displayport link 11534 11535commit b91e0efcb24eea32e6978c146c5409eeeeac0a62 11536Author: Dave Airlie <airlied@redhat.com> 11537Date: Tue Aug 10 15:05:25 2010 +1000 11538 11539 dce32: remove rmx workaround 11540 11541commit 94bc1b7156cd0866566dc44a823c7e051bb45175 11542Author: Dave Airlie <airlied@redhat.com> 11543Date: Tue Aug 10 13:24:52 2010 +1000 11544 11545 atombios: fixup set crtc source like KMS 11546 11547 This removes a bunch of strict aliasing warnings and fixes the 11548 codepaths up like the latest KMS code, including a workaround for a bug 11549 on evergreen. 11550 11551commit 9bc716eb62d4e0eed2902b92437a42634eef6ba1 11552Author: Dave Airlie <airlied@redhat.com> 11553Date: Tue Aug 10 13:23:21 2010 +1000 11554 11555 atombios: move adjust pixel clock around to follow KMS code flow 11556 11557 This reworks the pixel clock adjusting code to follow the KMS style, 11558 also fixes warnings in this code. 11559 11560commit 31de43bf9d9eb93cc2b2150474ea7404beabe49d 11561Author: Dave Airlie <airlied@redhat.com> 11562Date: Tue Aug 10 13:21:39 2010 +1000 11563 11564 displayport: retry on timeout 11565 11566 this is ported from KMS 11567 11568commit 9ef67335583d36080d227e8bce1966afe08e0486 11569Author: Dave Airlie <airlied@redhat.com> 11570Date: Tue Aug 10 13:21:01 2010 +1000 11571 11572 evergreen: don't call YUV table on evergreen 11573 11574commit 1cce55e8ba43e7958cb67147aeaeb068826ab99f 11575Author: Dave Airlie <airlied@redhat.com> 11576Date: Tue Aug 10 13:20:13 2010 +1000 11577 11578 evergreen: add support to parse firmware info for ext dp clk 11579 11580commit bbffd67d3296344e8735b007cdee83146d38369c 11581Author: Dave Airlie <airlied@redhat.com> 11582Date: Tue Aug 10 13:14:54 2010 +1000 11583 11584 atombios: realign digital transmitter/encoder setup with kms 11585 11586commit 6244153467665f5007e2fc7786b4bcc4b0b96030 11587Author: Dave Airlie <airlied@redhat.com> 11588Date: Tue Aug 10 12:57:22 2010 +1000 11589 11590 update atombios.h to latest from kernel 11591 11592commit bb7c77ca75e857f90791b0dd1c04c8e2f19d0e3c 11593Author: Alex Deucher <alexdeucher@gmail.com> 11594Date: Mon Aug 9 22:18:53 2010 -0400 11595 11596 atom: upstream parser update 11597 11598 fixes digital output problems on evergreen asics 11599 11600commit cc0a167ff2db9b097891883ded147af1d67e4407 11601Author: Alex Deucher <alexdeucher@gmail.com> 11602Date: Tue May 25 18:17:15 2010 -0400 11603 11604 r6xx/r7xx: default to micro (1D) tiling for now 11605 11606 SW de-tiling used in the r600c 3D driver has issues with 11607 certain configurations. 11608 11609 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11610 11611commit b5bfdbd70d9671250957ccd41dfc8818850d257e 11612Author: Alex Deucher <alexdeucher@gmail.com> 11613Date: Thu Aug 5 17:26:28 2010 -0400 11614 11615 r6xx/r7xx: add support for tiling with kms (v3) 11616 11617 Requires radeon drm 2.6.0 and updated mesa. 11618 11619 v2: - fix lockup due to not emiting DB_DEPTH_INFO 11620 https://bugs.freedesktop.org/show_bug.cgi?id=28342 11621 - fix drm minor version to deal with evergreen accel 11622 v3: rebase on current ddx git 11623 11624 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11625 11626commit a3c59c6f6be7067421e348142da0ca13428dcd57 11627Author: Alex Deucher <alexdeucher@gmail.com> 11628Date: Wed Aug 4 12:07:36 2010 -0400 11629 11630 radeon: add new pci ids 11631 11632commit a456587b77ae357750179a50f8db2a17c0f2738e 11633Author: Alex Deucher <alexdeucher@gmail.com> 11634Date: Mon Aug 2 14:24:41 2010 -0400 11635 11636 r6xx/r7xx: move syrface sync emit to the functions that emit surface info 11637 11638 reduces code duplication. 11639 11640commit 8eba977cab1878ba247da8160771d41194d8014f 11641Author: Alex Deucher <alexdeucher@gmail.com> 11642Date: Thu Jun 10 13:34:59 2010 -0400 11643 11644 EXA: move more common helpers to exa_shared 11645 11646commit 71c1a2704af23b61439cee5ce784f7fe267a8a26 11647Author: Alex Deucher <alexdeucher@gmail.com> 11648Date: Thu Jun 10 11:52:43 2010 -0400 11649 11650 remove unused reg headers from radeon_exa_shared.c 11651 11652commit d73aef78919005369af1b60df138439b4b6105c3 11653Author: Alex Deucher <alexdeucher@gmail.com> 11654Date: Mon Aug 2 13:27:12 2010 -0400 11655 11656 r6xx/r7xx: set VGT_MAX_VTX_INDX to a larger value 11657 11658commit dacaf5d827b58c39f9e5a7ac0530f9ea6e257347 11659Author: Dave Airlie <airlied@redhat.com> 11660Date: Mon Aug 2 08:33:51 2010 +1000 11661 11662 fix make distcheck 11663 11664commit c79ce215a01b45fc63b483da167ae37ec7aefad6 11665Author: Dave Airlie <airlied@redhat.com> 11666Date: Sun Aug 1 16:51:48 2010 +1000 11667 11668 radeon/r600: restructure exa + vbo to provide more sharing 11669 11670 This is a precursor for r300/500 vbo support. 11671 11672 Signed-off-by: Dave Airlie <airlied@redhat.com> 11673 11674commit 82254b59268140c4102ae3cd713743ae2be15c00 11675Author: Alex Deucher <alexdeucher@gmail.com> 11676Date: Fri Jul 30 17:15:05 2010 -0400 11677 11678 r6xx/r7xx: unify composite mask and non-mask pixel shader 11679 11680commit 1c17f3a192f644e8e38b5cfb1470f49434bfba27 11681Author: Alex Deucher <alexdeucher@gmail.com> 11682Date: Fri Jul 30 16:34:54 2010 -0400 11683 11684 r6xx/r7xx: clean up composite vertex shader 11685 11686 keep CF, ALU, Fetch instructions in separate groups 11687 11688commit f9d6c0de231357f96e2e0de71e6c9221bcb36bd4 11689Author: Gaetan Nadon <memsize@videotron.ca> 11690Date: Fri Jul 23 13:28:42 2010 -0400 11691 11692 The local copy of the modes code is no longer required. 11693 11694 The server 1.2 as shipped in the tarball on the web does not contain the 11695 modes code. It was added just after and found in git branch server-1.2-branch. 11696 11697 The modes code was initially included in version ati 6.8.0 and fails to compile 11698 with server 1.2 as it requires randr 1.2. The modes code is included in server 11699 versions 1.3 and later, so there is no need to provide an unknown version of 11700 the modes code in the ati driver tarball. It will never be used. 11701 11702 This patch makes the ati driver requiring server 1.3 or later. 11703 Version 6.8.0 configures and builds ok on server 1.3 11704 Master branch post 6.13.1 configures and builds ok on server 1.3 11705 11706 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11707 11708commit 0028419acb0762eeb950de5fe702c93e70301612 11709Author: Gaetan Nadon <memsize@videotron.ca> 11710Date: Sat Jul 24 10:51:18 2010 -0400 11711 11712 config: add AM_PROG_CC_C_O for per-target compilation flags 11713 11714 Per-target compilation flags (theatre200_drv_la_CFLAGS) are required 11715 when multiple targets which require different compiler flags, 11716 are build in the same makefile. 11717 11718 Automake issues a command with -c and -o flags which not all compilers 11719 support. The object fles are prefixed with theatre200_drv_la. 11720 The macro AM_PROG_CC_C_O must then be used to provide this feature 11721 on compilers that do not have it. If not, a warning is issued at make time. 11722 11723 This macros checks for compiler support and if missing, uses a "compile" 11724 script it generates in the package root directory. 11725 11726 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11727 11728commit aca0a8669b538d58f018f95c9b22e6b3ec1ffe50 11729Author: Gaetan Nadon <memsize@videotron.ca> 11730Date: Wed Jul 21 16:49:04 2010 -0400 11731 11732 config: add comments for main statements 11733 11734commit 66e614f7115efeec237b3b916d9637e8b3e8985c 11735Author: Gaetan Nadon <memsize@videotron.ca> 11736Date: Wed Jul 21 16:07:00 2010 -0400 11737 11738 config: replace deprecated use of AC_OUTPUT with AC_CONFIG_FILES 11739 11740 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11741 11742commit 16e5510c90ef1ba2bbaab78d18943f080b86d809 11743Author: Gaetan Nadon <memsize@videotron.ca> 11744Date: Wed Jul 21 14:37:41 2010 -0400 11745 11746 config: replace deprecated AC_HELP_STRING with AS_HELP_STRING 11747 11748 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11749 11750commit b36d8c09e91382f4cfa71635374ec88f5b676d1c 11751Author: Gaetan Nadon <memsize@videotron.ca> 11752Date: Wed Jul 21 14:05:22 2010 -0400 11753 11754 config: replace deprecated AM_CONFIG_HEADER with AC_CONFIG_HEADERS 11755 11756 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11757 11758commit cd9351b04c2d6982b28c647a63d550eb3e1937eb 11759Author: Alex Deucher <alexdeucher@gmail.com> 11760Date: Wed Jul 21 13:48:24 2010 -0400 11761 11762 r6xx/r7xx: group op variable state 11763 11764 Group the op variable state into one emit block, re-order 11765 to reduce dwords emitted. 11766 11767commit 5f838c664e8010f4e51afecd4100d73a96fe1209 11768Author: Gaetan Nadon <memsize@videotron.ca> 11769Date: Wed Jul 21 09:27:42 2010 -0400 11770 11771 config: complete AC_INIT m4 quoting 11772 11773 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11774 11775commit 48ec2e65c268b426ab9a3e214d174447cf5b5936 11776Author: Gaetan Nadon <memsize@videotron.ca> 11777Date: Wed Jul 21 08:33:38 2010 -0400 11778 11779 config: remove unrequired AC_SUBST for LIBDRM and LIBUDEV 11780 11781 These macros are called by the PKG_CHECK_MODULES macro. 11782 11783 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11784 11785commit 1aabb7eb7d8f06c7481151145db3b9a722ce4ef0 11786Author: Gaetan Nadon <memsize@videotron.ca> 11787Date: Tue Jul 20 21:54:11 2010 -0400 11788 11789 config: remove unrequired AC_SUBST([DRI_CFLAGS]) 11790 11791 This macro is called by PKG_CHECK_MODULES 11792 11793 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11794 11795commit 8f92b349821a3ee5ed8df55273d905b9605385aa 11796Author: Gaetan Nadon <memsize@videotron.ca> 11797Date: Tue Jul 20 21:44:57 2010 -0400 11798 11799 config: remove unrequired AC_SUBST([XORG_CFLAGS]) 11800 11801 This macro is called by PKG_CHECK_MODULES 11802 11803 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11804 11805commit 5b483b832f9c4a5b92ffb7f72a470669201f1fba 11806Author: Gaetan Nadon <memsize@videotron.ca> 11807Date: Tue Jul 20 20:24:42 2010 -0400 11808 11809 config: remove unrequired AC_HEADER_STDC 11810 11811 Autoconf says: 11812 "This macro is obsolescent, as current systems have conforming 11813 header files. New programs need not use this macro". 11814 11815 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11816 11817commit 6574e3a16eb3631ee7e00ee60a8c9ba95c8b84ef 11818Author: Gaetan Nadon <memsize@videotron.ca> 11819Date: Tue Jul 20 19:41:30 2010 -0400 11820 11821 config: remove AC_PROG_CC as it overrides AC_PROG_C_C99 11822 11823 XORG_STRICT_OPTION from XORG_DEFAULT_OPTIONS calls 11824 AC_PROG_C_C99. This sets gcc with -std=gnu99. 11825 If AC_PROG_CC macro is called afterwards, it resets CC to gcc. 11826 11827 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11828 11829commit 593eff2924c2ad161d8b490fbf6d7e433fbe2a80 11830Author: Gaetan Nadon <memsize@videotron.ca> 11831Date: Tue Jul 20 18:45:18 2010 -0400 11832 11833 config: update AC_PREREQ statement to 2.60 11834 11835 Unrelated to the previous patches, the new value simply reflects 11836 the reality that the minimum level for autoconf to configure 11837 all x.org modules is 2.60 dated June 2006. 11838 11839 ftp://ftp.gnu.org/gnu/autoconf/autoconf-2.60.tar.gz 11840 11841 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 11842 11843commit c2ab6ffc25aa6759cbbb4c1fbbd4a136b38983bf 11844Author: Gaetan Nadon <memsize@videotron.ca> 11845Date: Tue Jul 20 16:15:29 2010 -0400 11846 11847 config: upgrade to util-macros 1.8 for additional man page support 11848 11849 Use MAN_SUBST now supplied in XORG_MANPAGE_SECTIONS 11850 The value of MAN_SUBST is the same for all X.Org packages. 11851 11852commit cdeb1949c820242f05a8897d3ddd0718f204dacf 11853Author: Jerome Glisse <jglisse@redhat.com> 11854Date: Thu Jul 15 16:21:41 2010 -0400 11855 11856 kms: don't call cursor helper if using software cursor 11857 11858 Fix : 11859 https://bugzilla.redhat.com/show_bug.cgi?id=601713 11860 https://bugzilla.redhat.com/show_bug.cgi?id=598358 11861 11862 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 11863 11864commit 06691376b1ee963c711420edaf5a03eab6f5658f 11865Author: Dave Airlie <airlied@redhat.com> 11866Date: Wed Jul 7 13:15:03 2010 +1000 11867 11868 fix build on non-kms 11869 11870commit 052cf0169ae70d5448af6dc4db840b2fc195569b 11871Author: Dave Airlie <airlied@redhat.com> 11872Date: Wed Jul 7 11:10:46 2010 +1000 11873 11874 configure.ac: bump version post release 11875 11876commit ad999e633ff41d27eed9d2c6535e163a7181b0bd 11877Author: Dave Airlie <airlied@redhat.com> 11878Date: Wed Jul 7 10:49:22 2010 +1000 11879 11880 set version for release 11881 11882commit 37b348059b1c15d7b381cd3df3db52bd9ee6613e 11883Author: Alex Deucher <alexdeucher@gmail.com> 11884Date: Wed Jun 30 12:56:48 2010 -0400 11885 11886 remove rv100 quirk 11887 11888 Some RV100 cards with 2 VGA ports show up with DVI+VGA, however 11889 some boards with DVI+VGA have the same subsystem ids. Better 11890 to have a VGA port show up as DVI than having a non-useable 11891 DVI port. 11892 11893 reported by DHR in irc. 11894 11895 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11896 11897commit c8ea942bd0e9d1c055d50b94440aa4dae425f20b 11898Author: Cooper Yuan <cooperyuan@gmail.com> 11899Date: Tue Jun 29 20:34:57 2010 +0800 11900 11901 Remove HDP_SOFT_RESET function, there is no need to reset HDP block. 11902 11903 This commit can fix an issue reported on DELL server, system gets hang 11904 during soft resetting while another application tries to access PCI 11905 configuration space. 11906 11907commit 139b38bf67ec10d876cc56df833541d497ae4fa4 11908Author: Alex Deucher <alexdeucher@gmail.com> 11909Date: Thu Jun 24 14:55:09 2010 -0400 11910 11911 r6xx shader: use ADDR() for CF_DWORD0 11912 11913 no change in functionality 11914 11915commit 801e83227a59a29eea425ea612083bbf2b536c30 11916Author: Wolfram <bugzilla1@malloc.de> 11917Date: Mon Jun 21 18:59:19 2010 -0400 11918 11919 r6xx/r7xx: fix ums cmd buffer leak 11920 11921 Fixes: 11922 https://bugs.freedesktop.org/show_bug.cgi?id=27957 11923 11924commit b13d719080b75fc6db4d15d2d323b8fce8b7ad06 11925Author: Alex Deucher <alexdeucher@gmail.com> 11926Date: Mon Jun 21 18:49:44 2010 -0400 11927 11928 r6xx/r7xx: fix miscount in state emit 11929 11930commit c3c5c8e2cc91b51a24effdffb85281216eed731d 11931Author: Alex Deucher <alexdeucher@gmail.com> 11932Date: Mon Jun 21 14:30:17 2010 -0400 11933 11934 r6xx/r7xx accel: add back some additional default state 11935 11936 This adds back everything removed in c29157bbf5b0dd26857675282ab094082fbaed0d 11937 except CB_FOG_*, CB_CLEAR_* and the VPORT transforms. Those shouldn't 11938 be needed as we aren't using fog or viewport transforms. We probably don't 11939 need all the state that was added back either but I can't reproduce any 11940 problems here, so it's hard to say which parts are problematic. 11941 11942 Should fix: 11943 https://bugs.freedesktop.org/show_bug.cgi?id=28629 11944 and several corruption reports on #radeon. 11945 11946 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11947 11948commit 800cb2088fec698d0626063a9ab198ff534938c0 11949Author: Michel Dänzer <daenzer@vmware.com> 11950Date: Mon Jun 21 08:15:14 2010 +0200 11951 11952 DRI2: Fix up confusion between windows and pixmaps. 11953 11954 Fixes crashes with xserver master, where looking up a pixmap private on a 11955 window no longer happens to work. 11956 11957commit f7a91ece264af9f3fd2fc18e99aefcda93ce9f5c 11958Author: Michel Dänzer <daenzer@vmware.com> 11959Date: Mon Jun 14 08:52:16 2010 +0200 11960 11961 Convert x(c)alloc/xfree to m/calloc/free. 11962 11963 Fixes deprecation warnings with xserver master and should also work with older 11964 xservers. 11965 11966commit ea37d24b1b6d4cbcf73e680846de25b72af216e3 11967Author: Dave Airlie <airlied@redhat.com> 11968Date: Mon Jun 21 13:55:15 2010 +1000 11969 11970 radeon: fix support for 1.9 server master. 11971 11972 This moves pixmap and mode set into CSR where its allowed. Should work fine on 11973 old servers also. 11974 11975 Signed-off-by: Dave Airlie <airlied@redhat.com> 11976 11977commit fdd8ecafd054f65842351aee6ee6fba7af6613b2 11978Author: Alex Deucher <alexdeucher@gmail.com> 11979Date: Wed Jun 16 19:02:10 2010 -0400 11980 11981 r6xx/r7xx: macro safety fixes 11982 11983commit 4651d77211b508cb6b76931807780e317f232220 11984Author: Alex Deucher <alexdeucher@gmail.com> 11985Date: Wed Jun 16 12:28:36 2010 -0400 11986 11987 radeon: fix depth 16 with ums 11988 11989 Fixes: 11990 https://bugs.freedesktop.org/show_bug.cgi?id=28494 11991 https://bugzilla.redhat.com/show_bug.cgi?id=554967 11992 11993 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 11994 11995commit 1e1d6a515428b6884fea586d180346fc74ef75a1 11996Author: Alex Deucher <alexdeucher@gmail.com> 11997Date: Wed Jun 16 12:20:03 2010 -0400 11998 11999 r3xx-r5xx Xv: disable bicubic filtering by default 12000 12001 - makes Xv more consistent with r1xx/r2xx/r6xx/r7xx 12002 - Xv attributes like brightness, contrast, hue, etc. only work 12003 when bicubic is disabled. 12004 - avoids performance issues on some systems when sampling from textures 12005 in GART with kms. 12006 - It can be re-enabled with using xvattr 12007 12008 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 12009 12010commit c29157bbf5b0dd26857675282ab094082fbaed0d 12011Author: Alex Deucher <alexdeucher@gmail.com> 12012Date: Tue Jun 15 19:39:42 2010 -0400 12013 12014 r6xx/r7xx: remove unnecessary state emit 12015 12016 No need to emit state that doesn't impact that we use for EXA/Xv. 12017 12018 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 12019 12020commit 35280545fcfb911f189d7657bd1040a28450fe7b 12021Author: Alex Deucher <alexdeucher@gmail.com> 12022Date: Tue Jun 15 19:21:32 2010 -0400 12023 12024 r6xx/r7xx: reorg default and clipping state emit 12025 12026 Group ordered registers to save command buffer space. 12027 Reduces the default and clipping state from 256 to 160 dwords. 12028 12029 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 12030 12031commit a2528a734c1d4e8639c49e5d222e3630a93ffbfd 12032Author: Alex Deucher <alexdeucher@gmail.com> 12033Date: Tue Jun 15 17:00:34 2010 -0400 12034 12035 r6xx/r7xx accel: remove some duplicate emits and minor clean up 12036 12037 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 12038 12039commit 5f093357f18eb9bea641394ab86a92a1766d8f2e 12040Author: Gaetan Nadon <memsize@videotron.ca> 12041Date: Sat Jun 12 15:11:46 2010 -0400 12042 12043 COPYING: replace stub file with actual copyright notices 12044 12045 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 12046 12047commit 0c2118586d3edeecc2473b5d685472df4b5e70fa 12048Author: Dave Airlie <airlied@redhat.com> 12049Date: Fri Jun 11 15:53:18 2010 +1000 12050 12051 radeon: fixup last fix, use CURSOR_WIDTH not RADEON_CURSOR_WIDTH. 12052 12053commit 9a117f768cb8261327fd3c324da9c98875785cc1 12054Author: Dave Airlie <airlied@redhat.com> 12055Date: Fri Jun 11 14:28:24 2010 +1000 12056 12057 radeon: set proper stride for cursor in tiling flags. 12058 12059 definitely incorrect, but hope it doesn't break anything. 12060 12061 Signed-off-by: Dave Airlie <airlied@redhat.com> 12062 12063commit b6346ede94f0d0b11ee04770cf52508cb0a5e6c6 12064Author: Dave Airlie <airlied@redhat.com> 12065Date: Thu Jun 10 20:13:58 2010 -0400 12066 12067 drmmode: fix big endian issue with properties. 12068 12069 On a power machine with an rn50, this was causing load detection to be turned off after the first X run, subsequent X runs would find nothing connected. 12070 12071 Signed-off-by: Dave Airlie <airlied@redhat.com> 12072 12073commit 1971dc6d758dea13b9fb6d1c8f516c165628b5e5 12074Author: Alex Deucher <alexdeucher@gmail.com> 12075Date: Tue Jun 8 11:18:28 2010 -0400 12076 12077 radeon: fix rn50 cloning with kms 12078 12079 Since they only have one crtc sometimes the xserver doesn't assign 12080 a crtc to one of the outputs even though both outputs have common modes 12081 which results in only one monitor being enabled. Assign a crtc in 12082 preinit so that both outputs light up. 12083 12084 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 12085 12086commit 426114b4a99d37b394efe3336968bb0ab9b6e9ae 12087Author: Dave Airlie <airlied@redhat.com> 12088Date: Tue Jun 8 11:34:35 2010 +1000 12089 12090 xv: fix Xv on M6/RV100 under KMS. 12091 12092 pRADEONEnt->HasCRTC2 wasn't setup under KMS. 12093 12094 Signed-off-by: Dave Airlie <airlied@redhat.com> 12095 12096commit f64bf0de8e2de7c1bf9cc0c614603dd23c9060ad 12097Author: Alex Deucher <alexdeucher@gmail.com> 12098Date: Thu Jun 3 14:48:21 2010 -0400 12099 12100 evergreen: explicitly disable accel on evergreen 12101 12102 Previously we relied on the drm not having accel enabled 12103 to make sure evergreen used shadowfb, now we when we enable 12104 accel in the drm, we need to make sure the ddx doesn't try 12105 and use it until it's implemented. 12106 12107 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 12108 12109commit 428125c095b18c2a2864c1aef24ac0f384b6be54 12110Author: Marek Olšák <maraeo@gmail.com> 12111Date: Sun May 16 20:11:17 2010 +0000 12112 12113 r3xx-r5xx: enable color tiling by default on KMS 12114 12115 Signed-off-by: Marek Olšák <maraeo@gmail.com> 12116 12117commit 6712ce4f4715c8ce7c2fcddb52ca9b103bdd634c 12118Author: Marek Olšák <maraeo@gmail.com> 12119Date: Sun May 16 20:09:31 2010 +0000 12120 12121 rs740: follow macro-switch too 12122 12123 Signed-off-by: Marek Olšák <maraeo@gmail.com> 12124 12125commit 842fa162e9d3bbad2bc44d3732bbc8e5a54402ea 12126Author: Alex Deucher <alexdeucher@gmail.com> 12127Date: Tue May 18 11:21:01 2010 -0400 12128 12129 dri2: use radeon_pick_best_crtc() 12130 12131 It's already exposed and used by exa and Xv and has 12132 the same functionality. radeon_covering_crtc() can 12133 be dropped eventually. 12134 12135commit 64e6858aaf5d0e39ecc9f9804689012f3a38660a 12136Author: Alex Deucher <alexdeucher@gmail.com> 12137Date: Mon May 17 12:50:44 2010 -0400 12138 12139 r2xx exa: fix typo in 27c3326863deae36bc35e2c3b73fffd400208ff1 12140 12141 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 12142 12143commit 27c3326863deae36bc35e2c3b73fffd400208ff1 12144Author: Alex Deucher <alexdeucher@gmail.com> 12145Date: Mon May 17 09:53:32 2010 -0400 12146 12147 r1xx/r2xx EXA: handle RepeatNone properly with transforms 12148 12149 Fallback with unsupported ops like r3xx, r6xx. 12150 12151 Should fix fdo bug 28142. 12152 12153 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 12154 12155commit 30591320ec46e491ba20904cc64f3405b51c6505 12156Author: Jerome Glisse <jglisse@redhat.com> 12157Date: Fri May 7 17:35:09 2010 +0200 12158 12159 kms: add support for the MSC swap & sync API 12160 12161 This patch is mostly a port over of Intel ddx code for 12162 MSC support. It needs a radeon KMS module with version 12163 2.4 which has the query for hw crtc id. 12164 12165 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 12166 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 12167 12168commit 766024dcc61c83490540910ce752f9bfe6dddba4 12169Author: Marek Olšák <maraeo@gmail.com> 12170Date: Mon May 3 22:56:27 2010 +0200 12171 12172 r3xx-r5xx: fix texturing with small macrotiled pixmaps 12173 12174 Pixmaps smaller than a macrotile cannot be used as textures because hardware 12175 automatically switches to macro-linear and therefore sampled pixels are 12176 messed up. This behavior is known as MACRO_SWITCH. 12177 12178 The only sane workaround seems to be not using macrotiling for small pixmaps. 12179 12180 The function RADEONMacroSwitch has been ported from r300g and implements 12181 MACRO_SWITCH the same way it's implemented in hardware. It's been well tested 12182 in r300g. 12183 12184 This commit also fixes blit-based framebuffer reads, which are used for tiled 12185 surfaces in r300g, when ColorTiling is enabled. 12186 12187 Signed-off-by: Marek Olšák <maraeo@gmail.com> 12188 12189commit 6aedd57f6c2f08d0151a8bd1c5893d40d3db709a 12190Author: Alex Deucher <alexdeucher@gmail.com> 12191Date: Tue May 4 14:30:52 2010 -0400 12192 12193 radeon: fix pll_out_min default for pre-avivo 12194 12195 Fixes fdo bug 27960 12196 12197commit b2ecd1bb507ed4e0d9ba6ebd498c51da2e8136fa 12198Author: Michel Dänzer <daenzer@vmware.com> 12199Date: Fri Apr 23 18:28:22 2010 +0200 12200 12201 FB/MMIO_cnt should be ints, not pointers. 12202 12203 Pointed out by compiler warnings. 12204 12205 https://bugs.freedesktop.org/show_bug.cgi?id=27817 12206 12207commit 1ad41054989f7c3edec373ccf09dceda6d7c94b4 12208Author: Alex Deucher <alexdeucher@gmail.com> 12209Date: Thu Apr 22 16:48:21 2010 -0400 12210 12211 r3xx: fix gb pipe setup for SE cards 12212 12213 Noticed by Tormod Volden 12214 12215commit 59e287d7c484b4addd4a06d013670577639c7ae2 12216Author: Alex Deucher <alexdeucher@gmail.com> 12217Date: Thu Apr 22 02:05:40 2010 -0400 12218 12219 r3xx-r5xx: fix vertex units 12220 12221 Noticed by Tormod Volden. 12222 12223 RV3xx is 2, RV560,RV570 is 8 12224 12225commit 2a6b409496f26da0436972b5feae6ea035dde08d 12226Author: Owain Ainsworth <zerooa@googlemail.com> 12227Date: Mon Apr 19 14:37:33 2010 +0100 12228 12229 Reference count shared driver mappings. 12230 12231 With MMIO it wasn't *such* a bit deal if we leaked the smallish mapping. 12232 with FB it could be a larger deal. So instead of worrying about this, 12233 reference count the mappings in the entity structure and unmap them when 12234 no one cares anymore. 12235 12236 Prompted by a discussion with airlied 12237 12238 Signed-off-by: Dave Airlie <airlied@redhat.com> 12239 12240commit 761f0de5556e46f166280476185977f720efe586 12241Author: Owain G. Ainsworth <zerooa@googlemail.com> 12242Date: Mon Apr 19 04:05:36 2010 +0100 12243 12244 Make consistent use of fbOffset and share fb mappings. 12245 12246 What we were doing previously was mapping the framebuffer for zaphod for 12247 only this driver instances chunk, however, fbOffset was (rightly) set to 12248 the offset into the whole framebuffer we were using. 12249 12250 Since in some cases we did operations on the FB virtual address + 12251 fbOffset (for example zeroing the framebuffer on entervt) we were 12252 actually pissing all over ourselves in those cases. 12253 12254 Fix this by implementing shared fb mappings like we do for MMIO already, 12255 and whenever we wish to refer to our area of FB space we always use 12256 fbOffset. Fixes zaphod for some users on r600 chipsets, my 4870 is still 12257 behaving strangely on screen 0, but I suspect that is another bug. 12258 12259 Once calculation (in PreInitAccel) is now wrong because of this, however 12260 dri on zaphod does now happen so this is irrelavent, add a comment to 12261 that effect. 12262 12263 Signed-off-by: Dave Airlie <airlied@redhat.com> 12264 12265commit 2059d628c5fb03222a62502cc1b5724bf296a7b3 12266Author: Owain G. Ainsworth <zerooa@googlemail.com> 12267Date: Mon Apr 19 04:05:35 2010 +0100 12268 12269 Kill per-context SAREA support. 12270 12271 It has never been used, and since the world is changing it almost 12272 certainly never will be. Good riddance. 12273 12274 Signed-off-by: Dave Airlie <airlied@redhat.com> 12275 12276commit c4b46f56279cc18c8114235ca41cfb41ffede67f 12277Author: Owain G. Ainsworth <zerooa@googlemail.com> 12278Date: Mon Apr 19 04:05:34 2010 +0100 12279 12280 Remove remnants of the ShowCache option. 12281 12282 The rest of it died a long time ago. 12283 12284 Signed-off-by: Dave Airlie <airlied@redhat.com> 12285 12286commit 03b34c5fb12c4df180d4df113d2b654f6afd7230 12287Author: Dave Airlie <airlied@redhat.com> 12288Date: Mon Apr 19 17:58:07 2010 +1000 12289 12290 radeon: add FireMV 2400 pci id. 12291 12292 From RH#581927 - thanks to Carsten Clasohm 12293 12294 Signed-off-by: Dave Airlie <airlied@redhat.com> 12295 12296commit 6c07816025f70e522986614c29c300ea13e5d932 12297Author: Alex Deucher <alexdeucher@gmail.com> 12298Date: Fri Apr 16 18:50:17 2010 -0400 12299 12300 r5xx texvid: deal with large numbers of verts 12301 12302 should fix fdo bug 25884 12303 12304commit 5ee320a6b1b4c65fe592c8ac4d1981799242d59e 12305Author: Alex Deucher <alexdeucher@gmail.com> 12306Date: Fri Apr 16 18:42:56 2010 -0400 12307 12308 r3xx texvid: deal with large numbers of verts 12309 12310 should fix fdo bug 25884 12311 12312commit 47af3f4f266232517486238917d82fc5ca9c82e6 12313Author: Alex Deucher <alexdeucher@gmail.com> 12314Date: Fri Apr 16 20:50:18 2010 -0400 12315 12316 r2xx texvid: deal with large numbers of verts 12317 12318 should fix fdo bug 25884 12319 12320commit 57577d5cd0641b7cad02242478699bcfece59227 12321Author: Alex Deucher <alexdeucher@gmail.com> 12322Date: Fri Apr 16 20:40:56 2010 -0400 12323 12324 r1xx texvid: deal with large numbers of verts 12325 12326 should fix fdo bug 25884 12327 12328commit 31a888e78fa403f2bddacee098a29f36eaa969cb 12329Author: Alex Deucher <alexdeucher@gmail.com> 12330Date: Fri Apr 16 15:52:59 2010 -0400 12331 12332 r5xx textured video: split into prepare and draw functions 12333 12334commit f7ebe4a127965bb9e46d20c60331bcc4dab76fc5 12335Author: Alex Deucher <alexdeucher@gmail.com> 12336Date: Fri Apr 16 15:46:24 2010 -0400 12337 12338 r3xx textured video: split into prepare and draw functions 12339 12340commit 10799b4ce0c18c5a7d92d688549e8e777344f111 12341Author: Alex Deucher <alexdeucher@gmail.com> 12342Date: Fri Apr 16 15:38:49 2010 -0400 12343 12344 r2xx textured video: split into prepare and draw functions 12345 12346commit bd1cf42201a1f918080bf34786a6de327cb31b5f 12347Author: Alex Deucher <alexdeucher@gmail.com> 12348Date: Fri Apr 16 15:31:40 2010 -0400 12349 12350 r1xx textured video: split into prepare and draw functions 12351 12352commit c1136f94b80346065893f8a43c0fbf60c8c9b057 12353Author: Alex Deucher <alexdeucher@gmail.com> 12354Date: Fri Apr 16 15:23:14 2010 -0400 12355 12356 Xv: track vtx_count, is_planar in port priv 12357 12358commit 2f680b631357661511456f07b7b4ba1aa2cbf30c 12359Author: Alex Deucher <alexdeucher@gmail.com> 12360Date: Fri Apr 16 13:43:41 2010 -0400 12361 12362 radeon/kms: fix possible crtc mask for evergreen 12363 12364commit 4656f5dff1ed72fa2c7a1484305f2aef7b65ff2b 12365Author: Alex Deucher <alexdeucher@gmail.com> 12366Date: Fri Apr 16 13:36:25 2010 -0400 12367 12368 radeon/kms: fix crash when using more than two heads 12369 12370 fixes fdo bug 27692 12371 12372commit a69e749d0562887af6bd236c38802472e54640c4 12373Author: Dave Airlie <airlied@redhat.com> 12374Date: Fri Apr 16 13:58:54 2010 +1000 12375 12376 kms: add uevent support. 12377 12378 When we get a hotplug event from the kernel we should notify the client side to reconfigure displays. 12379 12380 based on work by ajax in intel driver. 12381 12382 Signed-off-by: Dave Airlie <airlied@redhat.com> 12383 12384commit 22a46dddd375b2b9399e12fdf168fa5292ff17a4 12385Author: Alex Deucher <alexdeucher@gmail.com> 12386Date: Tue Apr 13 11:08:59 2010 -0400 12387 12388 radeon: add support for pll algo selection 12389 12390 tv-out on atom systems is very particular about it's 12391 dividers. force it to use the old algo. 12392 12393 Should fix fdo bug 27593. 12394 12395commit c1b817c45477c21234abaaebe78feb7ea4fd92b1 12396Author: Alex Deucher <alexdeucher@gmail.com> 12397Date: Mon Apr 12 16:32:04 2010 -0400 12398 12399 atom: disable TV encoder when VGA is in use 12400 12401 Switching between TV and VGA caused problems on some 12402 systems since the TV encoder was left enabled. 12403 12404 Should fix fdo bug 25520. 12405 12406commit 878814aeab7031ff6169fcaf2be869b7595c8699 12407Author: Alex Deucher <alexdeucher@gmail.com> 12408Date: Mon Apr 12 11:16:14 2010 -0400 12409 12410 Xv overlay: set scalerwidth to 1920 on r3xx/r4xx chips 12411 12412 reported on phoronix forums. 12413 12414commit 276fe739f1caa95778213054375a65288b80f320 12415Author: Alex Deucher <alexdeucher@gmail.com> 12416Date: Fri Apr 9 20:03:09 2010 -0400 12417 12418 radeon: add some new pci ids 12419 12420commit eb5665688ef9b52f03f61546351d0848cab54740 12421Author: Alex Deucher <alexdeucher@gmail.com> 12422Date: Thu Apr 8 00:31:52 2010 -0400 12423 12424 radeon: don't setup Xv on rn50 12425 12426 It has no overlay or 3D engine. 12427 12428 fixes fdo bug 27528 12429 12430commit 3c6c6afcbfc6f2c582e13a7c37fe2014b0875b22 12431Author: Alex Deucher <alexdeucher@gmail.com> 12432Date: Mon Apr 5 12:12:56 2010 -0400 12433 12434 change version post release 12435 12436commit fb7911912e60b2cdbc2152b96847775b767ca3ef 12437Author: Alex Deucher <alexdeucher@gmail.com> 12438Date: Mon Apr 5 11:59:24 2010 -0400 12439 12440 bump version for release 12441 12442commit 476a1c6e8b42807b897b8c6f8550ec42b2d5c10e 12443Author: Alex Deucher <alexdeucher@gmail.com> 12444Date: Thu Apr 1 12:42:56 2010 -0400 12445 12446 radeon: fix possible gpio i2c table overrun 12447 12448 GPIO_I2C_INFO does not always have ATOM_MAX_SUPPORTED_DEVICE 12449 entries. Should fix Novell bug 589022. 12450 12451commit 6baa96c44ca93b88acf5233335cee233e59d5af4 12452Author: Dave Airlie <airlied@redhat.com> 12453Date: Thu Apr 1 11:24:32 2010 +1000 12454 12455 r600/exa: further cleanup, use the object struct in the accel state. 12456 12457 This cleans up the accel state objects as well. 12458 12459 Signed-off-by: Dave Airlie <airlied@redhat.com> 12460 12461commit 87f49a24019967552c5478d6af89ba438b4261c8 12462Author: Dave Airlie <airlied@redhat.com> 12463Date: Thu Apr 1 10:52:58 2010 +1000 12464 12465 r600: cleanup wasteful variables. 12466 12467 Just assign directly to the structs. 12468 12469 Signed-off-by: Dave Airlie <airlied@redhat.com> 12470 12471commit 478319e55d52fce241d040912fea952071e3b742 12472Author: Dave Airlie <airlied@redhat.com> 12473Date: Thu Apr 1 10:45:52 2010 +1000 12474 12475 r600: reduce function call overhead. 12476 12477 Create a small accel object that can be used to reduce the amount 12478 of parameters passed to SetAccelState. This can be cleaned up a lot more. 12479 12480 Signed-off-by: Dave Airlie <airlied@redhat.com> 12481 12482commit bc93395b3eb5e3511c1b62af90693269f4fa6e13 12483Author: Alex Deucher <alexdeucher@gmail.com> 12484Date: Wed Mar 31 22:30:24 2010 -0400 12485 12486 r6xx+ EXA: fix swapped domains in kms UTS 12487 12488 Noticed by Dave. 12489 12490 Should fix fdo bug 27284 12491 12492commit 3a33b1a92c328b690bdc1aafa0011e4326109815 12493Author: Owain G. Ainsworth <zerooa@googlemail.com> 12494Date: Tue Mar 30 17:14:02 2010 +0100 12495 12496 Add support for MacModel autodetection on OpenBSD. 12497 12498 This allows radeon-equipped macs to work without an xorg.conf on openbsd. 12499 Patch originally written by Mark Kettenis (kettenis@openbsd.org) 12500 12501commit c7eeda8c3f3514ba95ebf2893fbe124bf526b3df 12502Author: Alex Deucher <alexdeucher@gmail.com> 12503Date: Fri Mar 26 01:38:36 2010 -0400 12504 12505 radeon: add support eDP connectors with and LCD device tag 12506 12507 should fix fdo bug 27322 12508 12509commit ff323e11485fa5030ac278c836bf80a535c22249 12510Author: Alex Deucher <alexdeucher@gmail.com> 12511Date: Fri Mar 26 01:24:16 2010 -0400 12512 12513 radeon: fix name string for eDP 12514 12515commit 53ac06161eb2b8cffb1b88e24a9a21cfd12e5883 12516Author: Alex Deucher <alexdeucher@gmail.com> 12517Date: Tue Mar 23 13:34:38 2010 -0400 12518 12519 r6xx+ EXA/Xv: add a R600SetAccelState function 12520 12521 This moves CS bo checking and alignment checks into 12522 a central location. It also cleans up the code. 12523 12524commit 7a044472dfea7cf05ba4c2b87be30e94e2ae0b62 12525Author: Alex Deucher <alexdeucher@gmail.com> 12526Date: Tue Mar 23 01:27:22 2010 -0400 12527 12528 r6xx+ EXA: always use a temp surface for overlapping copies 12529 12530 The scanline based fallback code was complex and added a lot of 12531 overhead. It also didn't work with kms. 12532 12533commit d33bddaedae81126ab7f0023af1c6443bb0b4c4f 12534Author: Alex Deucher <alexdeucher@gmail.com> 12535Date: Mon Mar 22 18:15:23 2010 -0400 12536 12537 r6xx+ EXA: always use the accel_state state in DoPrepareCopy 12538 12539commit dda3f5a99e7a2dc5d57860f4d07df3498e1e21df 12540Author: Alex Deucher <alexdeucher@gmail.com> 12541Date: Mon Mar 22 18:05:36 2010 -0400 12542 12543 r6xx EXA/Xv: track src/dst domains 12544 12545 Much of the code is shared, so track the src/dst 12546 domains so we make sure the uses consistent domains 12547 for each bo. 12548 12549commit 5c256808cb5fea955eea96ffe9196473715156aa 12550Author: Alex Deucher <alexdeucher@gmail.com> 12551Date: Wed Mar 17 23:47:50 2010 -0400 12552 12553 XAA: disable render accel 12554 12555 It's been reported broken for a while. Should fix 12556 fdo bug 27151, others. 12557 12558commit a0683be5cc082bdbdd3bc4e9b52f39f423650946 12559Author: Dave Airlie <airlied@redhat.com> 12560Date: Thu Mar 18 12:36:25 2010 +1000 12561 12562 radeon: avoid using DRI1 init path on DRI2 driver. 12563 12564 I was playing with multi-seat and found this code, fixed 12565 it up to be sane and more DRI2 like. 12566 12567 Signed-off-by: Dave Airlie <airlied@redhat.com> 12568 12569commit 080a5414593e9b59ed923f26aa6057747b0c868f 12570Author: Alex Deucher <alexdeucher@gmail.com> 12571Date: Tue Mar 16 12:33:39 2010 -0400 12572 12573 kms: fix ums naming compat for DisplayPort 12574 12575commit 819b4015349b5d8c5ffa5f979097599774fce5bb 12576Author: Alex Deucher <alexdeucher@gmail.com> 12577Date: Mon Mar 15 13:47:29 2010 -0400 12578 12579 dump version for rc release 12580 12581commit d3482a947e3731be4ed0b00b4e3079470700dc4c 12582Author: Michael Cree <mcree@orcon.net.nz> 12583Date: Fri Mar 12 22:23:31 2010 +1300 12584 12585 Fix some word accesses in AtomBios to work on all architectures. 12586 12587 The UINT16LE_TO_CPU(), etc., macros are used in the AtomBios code to 12588 fix up endian issues but they do not address bad alignment or assist 12589 architectures that cannot perform hardware byte or word accesses. 12590 This patch inserts use of the ldw_u(), etc., interface of the Xserver 12591 into certain AtomBios accesses to address alignment issues. 12592 12593 This resolves Debian bug 572311, namely that the driver when compiled 12594 for generic Alpha architecture (i.e. doesn't use the byte-word extension) 12595 resulted in no display output on certain Radeon cards. 12596 12597 Signed-off-by: Michael Cree <mcree@orcon.net.nz> 12598 12599commit 488c9fd8300505cc6c0c2f8f0f00849f27cc5d63 12600Author: Alex Deucher <alexdeucher@gmail.com> 12601Date: Mon Mar 15 12:25:57 2010 -0400 12602 12603 r6xx/r7xx: fix domain handling in accel code 12604 12605 Noticed by Pauli and Michel on IRC. 12606 12607 Improves GetImage performace by a factor of ~10. 12608 12609commit 2ace2591d92fb6d3ce7a6453edb04b36a6c49a32 12610Author: Alex Deucher <alexdeucher@gmail.com> 12611Date: Mon Mar 15 12:03:58 2010 -0400 12612 12613 radeon: remove some leftover debugging output 12614 12615commit 67e81c8f17ddde6eba633d2a5aef528e1d598d89 12616Author: Andrzej Hajda <andrzej.hajda@wp.pl> 12617Date: Wed Mar 10 18:19:35 2010 -0500 12618 12619 radeon: add support for pal on legacy IGP chips 12620 12621 Based on my initial non-working patch. 12622 12623 Fixes some element of fdo bug 12007 12624 12625commit 3a44f1cb0d2bb748692b1024003de8ee88ca77a5 12626Author: Alex Deucher <alexdeucher@gmail.com> 12627Date: Tue Mar 9 09:44:01 2010 -0500 12628 12629 atom: i2c gpio fixes 12630 12631 Basically a port of my kms patch. This allows us 12632 to remove some quirks. 12633 12634commit d85cb40f516c67305e818302bec7ee817df4144c 12635Author: Matt Turner <mattst88@gmail.com> 12636Date: Sun Mar 7 14:24:35 2010 -0500 12637 12638 Don't check for Xinerama. 12639 12640 It doesn't seem to be used anywhere, so don't require it. 12641 12642 CC: Jerome Glisse <jglisse@redhat.com> 12643 CC: Alex Deucher <alexdeucher@gmail.com> 12644 CC: Dave Airlie <airlied@redhat.com> 12645 Signed-off-by: Matt Turner <mattst88@gmail.com> 12646 12647commit e7b41f8cb082ed462d29bf3fc440072424cbd852 12648Author: Alex Deucher <alexdeucher@gmail.com> 12649Date: Fri Mar 5 19:16:11 2010 -0500 12650 12651 radeon: disable frac fb div with new pll code 12652 12653 fixes fdo bug 26897 12654 12655commit 14aff767490c253cbcdd411f812e50b91673119e 12656Author: Alex Deucher <alexdeucher@gmail.com> 12657Date: Wed Mar 3 13:31:19 2010 -0500 12658 12659 radeon: add new RS880 pci id 12660 12661commit e6dc886634b38e4a36af7b5f0b23299d5acd7244 12662Author: Dave Airlie <airlied@redhat.com> 12663Date: Tue Mar 2 10:25:15 2010 +1000 12664 12665 radeon: bump configure.ac 12666 12667commit 4975658f05c387b39b3e96a292a7683f17645c2c 12668Author: Dave Airlie <airlied@redhat.com> 12669Date: Sat Feb 27 16:47:19 2010 +1000 12670 12671 pciids: hopefully fix HP 12672 12673commit e76b90b399c3cc0f0998c0209300c46f97505498 12674Author: Alex Deucher <alexdeucher@gmail.com> 12675Date: Fri Feb 26 15:01:28 2010 -0500 12676 12677 rv740: disable dfs workaround for drm 1.32+ 12678 12679 rv740 pipe setup was fixed in 2.6.33 12680 12681commit 2de0af5f5b806f9dbfdb1e9b6a5cf96d9433961d 12682Author: Matt Turner <mattst88@gmail.com> 12683Date: Wed Feb 24 22:46:28 2010 -0500 12684 12685 Use RADEON_ALIGN instead of open coding it. 12686 12687 Also fix some RADEON_ALIGN(x, 63), which would return incorrect results 12688 for odd x. Though this shouldn't happen, it's still not right. You 12689 wouldn't ever write (x + 62) & ~62 which is clearly wrong (and what it 12690 expands to). 12691 12692 CC: Jerome Glisse <jglisse@redhat.com> 12693 CC: Alex Deucher <alexdeucher@gmail.com> 12694 CC: Dave Airlie <airlied@redhat.com> 12695 Signed-off-by: Matt Turner <mattst88@gmail.com> 12696 12697commit c7e81d2f3a372e0d5f751dd0c5091aec2b56d936 12698Author: Matt Turner <mattst88@gmail.com> 12699Date: Wed Feb 24 22:46:27 2010 -0500 12700 12701 Use/define RADEON_GPU_PAGE_SIZE instead of sprinkling 4096 everywhere. 12702 12703 Also, define RADEON_BUFFER_ALIGN in terms of it, and replace some 12704 RADEON_ALIGN(x, RADEON_BUFFER_ALIGN) with RADEON_ALIGN(x, 12705 RADEON_GPU_PAGE_SIZE) since this is really what was intended. 12706 12707 CC: Jerome Glisse <jglisse@redhat.com> 12708 CC: Alex Deucher <alexdeucher@gmail.com> 12709 CC: Dave Airlie <airlied@redhat.com> 12710 Signed-off-by: Matt Turner <mattst88@gmail.com> 12711 12712commit c0a5c9403dff254e1669df606a4193794270aeff 12713Author: Dave Airlie <airlied@redhat.com> 12714Date: Fri Feb 26 19:17:28 2010 +1000 12715 12716 pci: add HP vendor id 12717 12718commit 3970cc5563900f08ba5236b23f765c926616c345 12719Author: Cooper Yuan <cooperyuan@gmail.com> 12720Date: Fri Feb 26 16:14:12 2010 +0800 12721 12722 add more triple-head server ID 12723 12724commit b499eee7506f7478649e0000e9f8fcd00ac9d88d 12725Author: James Le Cuirot <chewi@aura-online.co.uk> 12726Date: Thu Feb 25 21:13:18 2010 -0500 12727 12728 clean up kms zaphod handling 12729 12730 Taken from James patch on bug 24523 12731 12732 agd5f: adapt to the current code 12733 12734commit 433c8617341f5768255826435a2b09afba684f02 12735Author: Alex Deucher <alexdeucher@gmail.com> 12736Date: Thu Feb 25 21:03:13 2010 -0500 12737 12738 update ZaphodHeads option 12739 12740 Now that the screen section mapping is fixed, 12741 Make the option per-instance and allow multiple 12742 outputs to be specified; e.g., DVI-0 and HDMI-0 12743 associated with instance 0 and LVDS and VGA-0 12744 associated with instance 1. 12745 12746commit f3e20b01c77f02bdf3483ceee27a1f00d240c33d 12747Author: James Le Cuirot <chewi@aura-online.co.uk> 12748Date: Thu Feb 25 19:10:19 2010 -0500 12749 12750 zaphod: Fix entity instance assignment. 12751 12752 Each screen needs to have a pointer back to its 12753 device's entity instance, Currently Screen1 is 12754 pointing to Screen0's instance so when it comes 12755 to match up the screen sections with the detected 12756 screens, Screen0 gets chosen in both cases. 12757 12758 See bug 24523 12759 12760commit 1b7e9a2e50f77819b3aff4e37ba39eaec69ff932 12761Author: Alex Deucher <alexdeucher@gmail.com> 12762Date: Tue Feb 23 11:39:55 2010 -0500 12763 12764 radeon: fixes for zaphodheads option 12765 12766 Needed for systems with more than two outputs. 12767 Both KMS and non-KMS zaphod work on systems with 12768 more than two outputs now. 12769 12770commit bd83e5f1c75b8c00b43846c7225568917fc9a8ce 12771Author: Dave Airlie <airlied@redhat.com> 12772Date: Tue Feb 23 19:27:24 2010 +1000 12773 12774 zaphod: hopefully fix kms + zaphod 12775 12776commit 0c3468d812e3790ce03d9e76779ae81e7b7b82d5 12777Author: Alex Deucher <alexdeucher@gmail.com> 12778Date: Mon Feb 22 17:34:51 2010 -0500 12779 12780 AVIVO: set frac fb divider pll flag 12781 12782 matches previous behavior 12783 12784commit 212e152536c4325e6799018891d9aee132681f48 12785Author: Alex Deucher <alexdeucher@gmail.com> 12786Date: Mon Feb 22 17:27:24 2010 -0500 12787 12788 radeon: update new pll algo 12789 12790 - add support for pre-avivo chips 12791 - add support for fixed post/ref dividers 12792 - add support for non-fractional fb dividers 12793 12794 By default avivo chips use the new algo and 12795 pre-avivo chips use the old algo. Use the 12796 "NewPLL" option to toggle between them (set to 12797 TRUE for the new algo, FALSE for the old). 12798 12799commit e68d3a3890fc81c51f2006b5548da1e8756ad2fd 12800Author: Alex Deucher <alexdeucher@gmail.com> 12801Date: Thu Feb 18 14:02:00 2010 -0500 12802 12803 RS600: add connector quirk 12804 12805 System lists DVI port as HDMI. 12806 12807 fixes fdo bug 26605 12808 12809commit a3b730eceb522c7ac1ef3dd6f6c7d773118d03f7 12810Author: Jerome Glisse <jglisse@redhat.com> 12811Date: Thu Feb 18 14:13:48 2010 +0100 12812 12813 r6xx/kms: when reseting BO force default state emission 12814 12815 In KMS world each cs need to fully initialize the 3D engine 12816 when we were erasing cs in IBDiscard for r6xx we weren't 12817 forcing reemission of default state. This patch force this 12818 emission. Fix : 12819 12820 http://bugs.freedesktop.org/show_bug.cgi?id=26603 12821 http://bugzilla.kernel.org/show_bug.cgi?id=15284 12822 12823commit 579cdcf9b4e38c791a497b747a055fc0a07d8dd6 12824Author: Alex Deucher <alexdeucher@gmail.com> 12825Date: Wed Feb 17 12:22:48 2010 -0500 12826 12827 radeon: add ZaphodHeads option 12828 12829 Allows users that want to use zaphod mode to select 12830 which xrandr outputs are assigned to which head. E.g., 12831 12832 Option "ZaphodHeads" "LVDS,VGA-0" 12833 will assign LVDS to the first zaphod driver instance 12834 and VGA-0 to the second instance. 12835 12836commit 47136fa347d1756523239746b4c74cd5278a1118 12837Author: Michel Dänzer <daenzer@vmware.com> 12838Date: Mon Feb 15 16:44:39 2010 +0100 12839 12840 EXA: Fix order of coordinates passed to radeon_pick_best_crtc(). 12841 12842 The wrong order prevented tearing avoidance from working with EXA/DRI2. 12843 12844commit 6e95905f759b844373be62ec513c78153ed00222 12845Author: Michel Dänzer <daenzer@vmware.com> 12846Date: Mon Feb 15 17:02:16 2010 +0100 12847 12848 Allocate BOs for 'video offscreen surfaces' and 'TV-in' in VRAM. 12849 12850 I think these can only really work in VRAM, though right now there's probably 12851 nothing to prevent them from getting evicted... 12852 12853commit 78e7047c5235b09858b66dd3688d39aaa27d7589 12854Author: Pauli Nieminen <suokkos@gmail.com> 12855Date: Mon Feb 15 13:40:37 2010 +0200 12856 12857 Allocate Xv buffers to GTT. 12858 12859 KMS doesn't have acceleration for upload to vram. memcpy/memmove to VRAM 12860 directly is very slow (40M/s in benchmark) which causes visible problems 12861 to video. 12862 12863 Allocating video buffer in GTT will give good performance (350-450M/s) 12864 for memmove operation. This is nice performance boost for Xv under KMS. 12865 12866 Signed-off-by: Pauli Nieminen <suokkos@gmail.com> 12867 12868commit 3ec25e59854b6b03ad763bc374d3475a50f562d8 12869Author: Pauli Nieminen <suokkos@gmail.com> 12870Date: Mon Feb 15 13:45:28 2010 +0200 12871 12872 Use memcpy instead of memove for RADEONSwapCopy. 12873 12874 memcpy has about 25% better performance than memmove when 12875 destination is GTT (wc caching). Changing memmove to memcpy 12876 in SwapCopy will reduce CPU time spent moving data between 12877 host and GPU. 12878 12879 memcpy will specially help Xv when playing high resolution 12880 videos. 12881 12882 The swap path doesn't support overlapping memory copies either 12883 so extra safety is not worth the performance hit in commonly 12884 used code path. 12885 12886 Signed-off-by: Pauli Nieminen <suokkos@gmail.com> 12887 12888 [ Michel Dänzer: Fixed up whitespace ] 12889 12890commit 221ef11b31756deb7134801730e76c040e841f5c 12891Author: Alex Deucher <alexdeucher@gmail.com> 12892Date: Fri Feb 12 14:15:10 2010 -0500 12893 12894 r6xx/r7xx: implement EXA vline support 12895 12896commit 8ad4025affe1fb2f417e3a3031d74f83be5df253 12897Author: Alex Deucher <alexdeucher@gmail.com> 12898Date: Fri Feb 12 11:34:47 2010 -0500 12899 12900 radeon: consolidate crtc selector for vline wait 12901 12902 Use the Xv version as it takes into account the area 12903 covered by the op rather than just picking the largest 12904 crtc area. 12905 12906commit 4f9d1714a7382594b834d446bbe502663f6a2778 12907Author: Alex Deucher <alexdeucher@gmail.com> 12908Date: Fri Feb 5 12:27:51 2010 -0500 12909 12910 r500: fix relocs for Xv planar video 12911 12912 fixes fdo bug 26445 12913 12914commit b7ca1ab11ac0e4e4be5a1e7789ae8633b460a2d4 12915Author: Dave Airlie <airlied@redhat.com> 12916Date: Fri Feb 5 13:37:20 2010 +1000 12917 12918 rs400/rs480: mc idle bit is bit 2 like original radeon 12919 12920 backport from a KMS fix, the rs400/480 mc idle is bit 2 not bit 4. 12921 12922 Signed-off-by: Dave Airlie <airlied@redhat.com> 12923 12924commit 97c387e269672b146a95b45fbef5c8c18e473e36 12925Author: Gaetan Nadon <memsize@videotron.ca> 12926Date: Fri Jan 29 16:08:17 2010 -0500 12927 12928 config: remove dead LINUXDOC macro usage 12929 12930 This module does not generate LINUXDOC documentation. 12931 12932 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 12933 12934commit 05fa2dee0c63d56fd523ee1ebab479bcb9f5e89d 12935Author: Alex Deucher <alexdeucher@gmail.com> 12936Date: Thu Feb 4 13:13:58 2010 -0500 12937 12938 evergreen: minor clean up in transmitter setup 12939 12940 both link and encoder must be set for linkb 12941 12942commit 76eea5e02776f7beeb8f4491a374bd36792eda92 12943Author: Alex Deucher <alexdeucher@gmail.com> 12944Date: Thu Feb 4 10:52:14 2010 -0500 12945 12946 r600: reduce number of cache flushes 12947 12948 We don't need to flush so often. Next step 12949 would be to move the flushing to the drm and only 12950 flush after each command buffer rather than each 12951 draw. 12952 12953commit 8d63d70f7ebaf9d250f0449d3720ef47516c05df 12954Author: Alex Deucher <alexdeucher@gmail.com> 12955Date: Wed Feb 3 15:22:22 2010 -0500 12956 12957 evergreen: use external clock source for DP PHY 12958 12959 DP CRTC clock always comes from DCPLL. This frees 12960 up PPll1/2 for non-DP-mode PHYs and CRTCs 12961 12962commit ecbc26431914216a8b207e81451282ea07c8b92f 12963Author: Alex Deucher <alexdeucher@gmail.com> 12964Date: Wed Feb 3 03:52:07 2010 -0500 12965 12966 evergreen: blank/unblank DP in dpms calls 12967 12968commit 710a2fd07235349f084ec40626cc28fbae523d27 12969Author: Alex Deucher <alexdeucher@gmail.com> 12970Date: Wed Feb 3 03:00:24 2010 -0500 12971 12972 evergreen: pll fixes 12973 12974 SetPixelClock needs to be called for DP outputs with a 12975 valid PLL. 12976 12977commit 9ab2377ee23c5dc360ba135a77aa6b181a1862e6 12978Author: Alex Deucher <alexdeucher@gmail.com> 12979Date: Tue Feb 2 15:15:36 2010 -0500 12980 12981 evergreen: DP requires coherent mode 12982 12983commit 634da80056525a48ec17ffb81e2fb59b168bea9e 12984Author: Alex Deucher <alexdeucher@gmail.com> 12985Date: Tue Feb 2 15:09:57 2010 -0500 12986 12987 evergreen: fix units on frac_fb_div 12988 12989commit a887818f491f6c7315c56c4e0d0b702c4c6aa4ac 12990Author: Alex Deucher <alexdeucher@gmail.com> 12991Date: Mon Feb 1 11:01:47 2010 -0500 12992 12993 evergreen: add pci ids 12994 12995commit cb2772b69480268c059766c4f6b209ce590ede0e 12996Author: Alex Deucher <alexdeucher@gmail.com> 12997Date: Fri Jan 29 12:59:46 2010 -0500 12998 12999 evergreen: add atombios crtc/pll functions 13000 13001commit bd8e04cb7b39f38b6958273582a9b324a9f0759a 13002Author: Alex Deucher <alexdeucher@gmail.com> 13003Date: Mon Feb 1 10:07:43 2010 -0500 13004 13005 evergreen: add atom support for digital outputs 13006 13007 analog is already supported by the existing code. 13008 13009commit 8ad40d3c32ad5b91725bd37fcade6bed504df421 13010Author: Alex Deucher <alexdeucher@gmail.com> 13011Date: Tue Jan 26 15:39:44 2010 -0500 13012 13013 evergreen: add crtc set base/format support 13014 13015commit 2c9cf0a07ff9e5a4989861bc2fdfd71d841013a8 13016Author: Alex Deucher <alexdeucher@gmail.com> 13017Date: Thu Dec 10 15:29:14 2009 -0500 13018 13019 evergreen: add lut support 13020 13021commit 65246545c3dd9bfef73e8a0f200bd5909b89a167 13022Author: Alex Deucher <alexdeucher@gmail.com> 13023Date: Tue Jan 26 12:10:02 2010 -0500 13024 13025 evergreen: add hw cursor support 13026 13027commit c05cad56b69d239fa2e69905d15f4f08b9db4c55 13028Author: Alex Deucher <alexdeucher@gmail.com> 13029Date: Thu Jan 28 01:28:52 2010 -0500 13030 13031 evergreen: add base asic support 13032 13033commit ed63e1b1abe8810b5da6b4140892337eef08a9ea 13034Author: Alex Deucher <alexdeucher@gmail.com> 13035Date: Fri Jan 29 12:42:33 2010 -0500 13036 13037 evergreen: add register and utility defines 13038 13039commit 019260ec4a9f9fbf2ac63a2ca3314aa308471f7e 13040Author: Alex Deucher <alexdeucher@gmail.com> 13041Date: Thu Dec 10 14:57:39 2009 -0500 13042 13043 evergreen: add chip enums 13044 13045commit e86ae564d8b94d2cc8d3c7cceb452a197ab6387a 13046Author: Alex Deucher <alexdeucher@gmail.com> 13047Date: Mon Feb 1 12:03:23 2010 -0500 13048 13049 atom: update to the latest upstream atombios.h changes 13050 13051commit 77b13a02c70842a58e0590d0243f0ae016c5a640 13052Author: Alex Deucher <alexdeucher@gmail.com> 13053Date: Mon Feb 1 12:16:20 2010 -0500 13054 13055 r500: fix CS section size missmatch in Xv code 13056 13057commit 5c11264517d1b9808415d47e9dedc345610f0f86 13058Author: Dave Airlie <airlied@linux.ie> 13059Date: Fri Jan 29 15:10:33 2010 +1000 13060 13061 radeon: further digital encoder cleanups. 13062 13063 this just cleans up a few more bits of the digital encoder setup. 13064 13065 Signed-off-by: Dave Airlie <airlied@redhat.com> 13066 13067commit 872a11fa9efb0b76cef83089ea1bb80792b94930 13068Author: Alex Deucher <alexdeucher@gmail.com> 13069Date: Thu Jan 28 01:19:23 2010 -0500 13070 13071 radeon: clean up dig encoder handling 13072 13073 - switch the var name to dig_encoder 13074 - quiet coherent messages 13075 - clean up dig encoder selection 13076 13077 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13078 13079commit c1a73ff51b371ecc6f57a81a33f644344b4e3f70 13080Author: Dave Airlie <airlied@linux.ie> 13081Date: Thu Jan 28 12:58:29 2010 +1000 13082 13083 radeon: add support for picking the digitial encoder 13084 13085 Testing on a W500 laptop found the LVDS + DisplayPort would conflict 13086 over the choice of digital encoder, and only one would work. 13087 13088 This patch fixes the DCE3 case where LVTMA requires the second digital 13089 encoder and lets DP/DVI use the first in most cases. 13090 13091 Signed-off-by: Dave Airlie <airlied@redhat.com> 13092 13093commit 95b56275d80dcee48a7927df124602c34cd72235 13094Author: Alex Deucher <alexdeucher@gmail.com> 13095Date: Wed Jan 27 15:48:25 2010 -0500 13096 13097 avivo: add some hotplug detect regs 13098 13099commit 5a4327f7784361933484895c9af751ccfa242d48 13100Author: Alex Deucher <alexdeucher@gmail.com> 13101Date: Tue Jan 26 16:06:00 2010 -0500 13102 13103 radeon: minor pll updates 13104 13105 add new fixed post divider option 13106 13107commit cbd642af7c76469d701471daea0c8d167567ccf8 13108Author: Alex Deucher <alexdeucher@gmail.com> 13109Date: Tue Jan 26 12:03:45 2010 -0500 13110 13111 radeon: use mmio bar size rather than hardcoded number for register ops 13112 13113 newer asics have larger mmio bars 13114 13115commit e5933fd74923a07aa401bdb4ecd8d1d641148440 13116Author: Alex Deucher <alexdeucher@gmail.com> 13117Date: Mon Jan 25 13:18:47 2010 -0500 13118 13119 KMS: fix no accel option 13120 13121 NO_ACCEL wasn't handled correctly with KMS. 13122 13123 fixes fdo bug 26198 13124 13125commit ba99a82ced1a6101365876723d22bcfec4607537 13126Author: Dave Airlie <airlied@redhat.com> 13127Date: Mon Jan 25 13:50:13 2010 +1000 13128 13129 radeon/kms: only enable EXA pixmaps under kms if we support render accel 13130 13131 I've seen RN50s with 64MB of RAM that are slow as molasses, this 13132 should fix them. 13133 13134 RH bug 556400 13135 13136 Signed-off-by: Dave Airlie <airlied@redhat.com> 13137 13138commit de40d36264364434541377c7f65f3836dc514fc0 13139Author: Kusanagi Kouichi <slash@ac.auone-net.jp> 13140Date: Fri Jan 22 17:18:51 2010 +0900 13141 13142 Remove unnecessary checks. 13143 13144 Signed-off-by: Kusanagi Kouichi <slash@ac.auone-net.jp> 13145 13146commit 30a19b75cc82b8e04c45e6684b84f9a4ccc0505b 13147Author: Alan Coopersmith <alan.coopersmith@sun.com> 13148Date: Tue Jan 19 14:54:04 2010 -0800 13149 13150 Fix configure with --disable-dri 13151 13152 AM_CONDITIONAL calls can't be made inside if statements, since 13153 automake won't know what to do when the if isn't true. 13154 13155 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 13156 13157commit bd011784c0609f1fd03dcb784744c8f5cdef33da 13158Author: Alex Deucher <alexdeucher@gmail.com> 13159Date: Tue Jan 19 11:59:38 2010 -0500 13160 13161 ATOM: Upstream parser updates 13162 13163commit 95d63e408cc88b6934bec84a0b1ef94dfe8bee7b 13164Author: Jerome Glisse <jglisse@redhat.com> 13165Date: Mon Jan 18 10:46:50 2010 +0100 13166 13167 r6xx/r7xx: emit relocation for FRAG & TILE buffer 13168 13169 FRAG & TILE buffer are unused but still they need 13170 to be associated with a valid relocation so that 13171 userspace can't try to abuse them to overwritte 13172 GART and then try to write anywhere in system 13173 memory. 13174 13175commit bbaf71fb46e3e0acdadcb91e9b1c73396bd9aa8e 13176Author: Alan Coopersmith <alan.coopersmith@sun.com> 13177Date: Fri Jan 15 15:09:27 2010 -0800 13178 13179 Update Sun license notices to current X.Org standard form 13180 13181 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 13182 13183commit 3d158716a60d5a113a541ea2f680b81a1be41ad2 13184Author: Dave Airlie <airlied@redhat.com> 13185Date: Thu Jan 14 11:21:33 2010 +1000 13186 13187 displayport: fix DDC on DVI->DP convertors. 13188 13189 Seems we have to not do auxch DDC if we aren't talking to a DP 13190 sink. 13191 13192commit 6a363f68415d37c302151581f2a86855dba39b67 13193Author: Alex Deucher <alexdeucher@gmail.com> 13194Date: Mon Jan 11 15:13:45 2010 -0500 13195 13196 rv100: fix typo in fix for bug 25992 13197 13198 Noticed by Maarten Maathuis. 13199 13200commit 7b01e1ee29f681bf1735ecded6445d12beeb52d8 13201Author: Alex Deucher <alexdeucher@gmail.com> 13202Date: Mon Jan 11 12:40:06 2010 -0500 13203 13204 rv100: reject modes >135 Mhz with DVI 13205 13206 Due to heat issues. fixes bug 25992 13207 13208 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13209 13210commit 74da9a66a420a40bb5d562f07e97ce6ac2eb010f 13211Author: Alex Deucher <alexdeucher@gmail.com> 13212Date: Fri Jan 8 16:51:43 2010 -0500 13213 13214 radeon/kms: add new strings for eDP and TV 13215 13216commit 61977f61f516b54d1fd9df34d0f0892fbf5227d7 13217Author: Alex Deucher <alexdeucher@gmail.com> 13218Date: Thu Jan 7 02:15:31 2010 -0500 13219 13220 radeon: fallback to i2c for edid if aux fails 13221 13222 In case of a DP->DVI adapter for example. 13223 13224commit 958d073869404f60e56dc0cc70b3e7de85904694 13225Author: Alex Deucher <alexdeucher@gmail.com> 13226Date: Thu Jan 7 02:12:00 2010 -0500 13227 13228 radeon: add initial support for eDP (embedded DisplayPort) 13229 13230 Should fix fdo bug 25931 13231 13232commit 944ae73f7cd5f9a1622b38e06a58565812cb102f 13233Author: Alex Deucher <alexdeucher@gmail.com> 13234Date: Thu Jan 7 01:52:35 2010 -0500 13235 13236 radeon: updated ObjectID.h 13237 13238commit 48aa5064aff4b9adf768e480df2312d4375e9c40 13239Author: Alex Deucher <alexdeucher@gmail.com> 13240Date: Mon Jan 4 11:05:44 2010 -0500 13241 13242 radeon: fix LVDS power sequence on Mac cards 13243 13244 Noticed by John R. Dunning. Fix taken from radeonfb. 13245 13246 I'm not sure if this sequence would be useful on any PC 13247 laptops or not so make it mac specific for now. 13248 13249commit 36bd69affc996c92c40b7360a7fbaa1a3a46abfd 13250Author: Gaetan Nadon <memsize@videotron.ca> 13251Date: Wed Dec 16 15:52:18 2009 -0500 13252 13253 configure.ac: remove unused sdkdir=$(pkg-config...) statement 13254 13255 The sdkdir variable isn't use, so remove the statement. 13256 13257 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 13258 13259 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 13260 13261commit 9d0f3af7278dc939fd4e6f3ea69d9f488a9fbed7 13262Author: Alex Deucher <alexdeucher@gmail.com> 13263Date: Tue Dec 22 08:45:27 2009 -0500 13264 13265 radeon: add cvt timing if we only have panel w/h 13266 13267 fixes mac laptops without an edid 13268 13269 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13270 13271commit 6e1f5553c6d7e3b5d089af2e3d587efe95936855 13272Author: Alex Deucher <alexdeucher@gmail.com> 13273Date: Mon Dec 21 17:32:40 2009 -0500 13274 13275 ATOM: add new power table defs 13276 13277commit 4b05c47ac657f9a93d76221269761ed64c81f716 13278Author: Dave Airlie <airlied@redhat.com> 13279Date: Thu Dec 17 14:30:23 2009 +1000 13280 13281 radeon: use new libdrm api 13282 13283commit 1d620800bd79290967fa487decf798e318c45f25 13284Author: Ingmar Vanhassel <ingmar@exherbo.org> 13285Date: Sun Dec 20 21:04:56 2009 +0100 13286 13287 Replace shave with automake silent-rules 13288 13289 Which are enabled by default in Xorg's util-macros 1.3 if automake-1.11 13290 is used. 13291 13292 Signed-off-by: Ingmar Vanhassel <ingmar@exherbo.org> 13293 13294commit 52f6a2ff682ad1f8e26ef1cd7d380493dcd57752 13295Author: Gaetan Nadon <memsize@videotron.ca> 13296Date: Wed Dec 16 15:29:53 2009 -0500 13297 13298 src/Makefile.am: missing 3 header files in tarball 13299 13300 radeon_dummy_bufmgr.h 13301 radeon_vbo.h 13302 simple_list.h 13303 Replace leading spaces with tab for radeon_atomwrapper.h 13304 13305 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 13306 13307commit 3a30210d50b27f8772fc5045133940246764fce9 13308Author: Matthias Hopf <mhopf@suse.de> 13309Date: Tue Dec 15 10:53:48 2009 -0500 13310 13311 fix 200M freezes on VT switch if CRTC is disabled 13312 13313 It appears that RS4xx chips need to have the crtc 13314 enabled when the timing is programmed. 13315 13316 agd5f: minor fixes/cleanup of the original patch 13317 13318commit 299d395bd3f294239dee58ab7d607d7d2c657f61 13319Author: Corbin Simpson <MostAwesomeDude@gmail.com> 13320Date: Sun Dec 13 14:51:17 2009 -0800 13321 13322 dri: be more useful if disabling DRI due to KMS/UMS mismatch. 13323 13324 It's only fair. 13325 13326 Compile-tested only. 13327 13328 Signed-off-by: Dave Airlie <airlied@redhat.com> 13329 13330commit 0e5c9d87b5d7e0751df71cc8958ca5ccaed25104 13331Author: Alex Deucher <alexdeucher@gmail.com> 13332Date: Thu Dec 10 14:27:43 2009 -0500 13333 13334 ATOM: split set base/format into separate functions 13335 13336 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13337 13338commit 46630da5fd6f45bb8ea150b870162997480d69c7 13339Author: Alex Deucher <alexdeucher@gmail.com> 13340Date: Wed Dec 9 12:55:25 2009 -0500 13341 13342 AVIVO: add new PLL code 13343 13344 This should hopefully help the problems with flickering 13345 and blinking monitors reported on some systems. If there 13346 are problems, the old PLL algorithm can be selected with: 13347 Option "NewPLL" "FALSE" 13348 in the device section of your X config. 13349 13350 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13351 13352commit f082b1693d6f7f763ccf5a8436a89890ca2c6129 13353Author: Matthijs Kooijman <matthijs@stdin.nl> 13354Date: Thu Dec 3 12:10:14 2009 -0500 13355 13356 radeon: fix crtc2 dpms 13357 13358 noticed by Matthijs Kooijman on fdo bug 22140 13359 13360commit 3a96fbf3b84522cf1ba1b176e82b662222331c14 13361Author: Alex Deucher <alexdeucher@gmail.com> 13362Date: Wed Dec 2 18:02:33 2009 -0500 13363 13364 radeon: only read RADEON_PPLL_REF_DIV for ref div on pre-avivo 13365 13366 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13367 13368commit f03450796d2e9247a1228c4e2abb1dfad7aecddf 13369Author: Kusanagi Kouichi <slash@ac.auone-net.jp> 13370Date: Wed Dec 2 01:36:37 2009 -0500 13371 13372 radeon: Lift hardcoded limit from RADEONQueryImageAttributes 13373 13374 The dimension of an XvImage is limited to 2048 x 2048 even if an adaptor 13375 supports larger image. 13376 XvCreateImage and XvShmCreateImage lower the width or height of an image. 13377 XvPutImage and XvShmPutImage return BadValue. 13378 The cause is that 2048 is hardcoded in RADEONQueryImageAttributes. 13379 13380commit efbc2c80ab02879edf3b7b3d65b16c45ddce5017 13381Merge: 88a50a30d a612813c5 13382Author: Dave Airlie <airlied@redhat.com> 13383Date: Wed Dec 2 14:30:36 2009 +1000 13384 13385 Merge remote branch 'origin/displayport' 13386 13387commit 88a50a30df11a06263209340a42251851f8e2334 13388Author: Dave Airlie <airlied@redhat.com> 13389Date: Mon Nov 30 15:32:12 2009 +1000 13390 13391 r600: fix multi-operation in single batch support. 13392 13393 This ports the mesa DMA buffer handling with the 3 lists, 13394 13395 Signed-off-by: Dave Airlie <airlied@redhat.com> 13396 13397commit b2597deea3a3953ff50d54ff37e3c043eac409f4 13398Author: Kusanagi Kouichi <slash@ac.auone-net.jp> 13399Date: Mon Nov 30 11:10:50 2009 -0500 13400 13401 R600: fix some warnings 13402 13403commit 4d90dc3cb248e9d61c0c490bba80f6727fafd318 13404Author: Alex Deucher <alexdeucher@gmail.com> 13405Date: Mon Nov 30 11:05:38 2009 -0500 13406 13407 atom: pull misc mode info for lvds panel mode from bios tables 13408 13409 sync polarity, etc. This will likely fix LVDS problems 13410 on some laptops. 13411 13412 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13413 13414commit a612813c5f329f00a9271e7bb69abd2630a49403 13415Author: Alex Deucher <alexdeucher@gmail.com> 13416Date: Fri Nov 27 17:20:30 2009 -0500 13417 13418 radeon: clean up DP code 13419 13420 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13421 13422commit a8dbf7c23481501987971a9c0b6cb0760f86127f 13423Author: Dave Airlie <airlied@redhat.com> 13424Date: Fri Nov 27 10:16:00 2009 +1000 13425 13426 r600/accel: cleanup and merge a lot of the accel functions 13427 13428 All of the drawing ops were the exact same modulo the vtx size, 13429 this along with the vertex buffer wrapping code could all be consolidated 13430 into a smaller set of functions. 13431 13432 This also adds 2 VBO which we switch between, and merges a #define to 13433 enable the multiple operations in one CS under KMS mode. 13434 13435 Multi-operation still isn't working though. 13436 13437 Signed-off-by: Dave Airlie <airlied@redhat.com> 13438 13439commit 971e463b2cfdc1ba1c78a24fb439d33dd9155dfc 13440Author: Alex Deucher <alexdeucher@gmail.com> 13441Date: Thu Nov 26 12:45:15 2009 -0500 13442 13443 dce3: CV1/TV1OutputControl tables didn't go away until dce3.2 13444 13445 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13446 13447commit 95385f0906f371a1ed6e60eb3e597e699c7a3222 13448Author: Jerome Glisse <jglisse@redhat.com> 13449Date: Thu Nov 26 12:21:21 2009 +0100 13450 13451 kms: Fix resizing when acceleration is disabled 13452 13453 When acceleration is disabled we need to reallocate 13454 a new shadow framebuffer and we should also avoid 13455 calling any EXA function as EXA is disabled in such 13456 case. 13457 13458commit 8195385d3cedf70d89ca576371af6bca5045bbd2 13459Author: Dave Airlie <airlied@redhat.com> 13460Date: Thu Nov 26 15:43:54 2009 +1000 13461 13462 dce3: no CV1 or TV1 Output Control tables anymore. 13463 13464 According to the DCE3 docs we should only use DAC1/2 not 13465 CV1/TV OutputControl, also my rv730 bios doesn't have any 13466 CV1/TV tables. 13467 13468 Signed-off-by: Dave Airlie <airlied@redhat.com> 13469 13470commit 381b14d5422131ea8ff942ab06a328f6beca0d3f 13471Author: Dave Airlie <airlied@redhat.com> 13472Date: Thu Nov 26 15:28:15 2009 +1000 13473 13474 atombios: CV mode retrival was broken. 13475 13476 a) crev and frev are reversed 13477 b) my rv730 bios only has one mode in it, so 13478 bounds check the table size. 13479 13480commit a67a13c315a9235be052662e2a3270686cccc0f0 13481Author: Dave Airlie <airlied@redhat.com> 13482Date: Thu Nov 26 13:41:36 2009 +1000 13483 13484 radeon: fix typo in unused as of yet CV code 13485 13486commit cd9bfb5892eebef6dff14ac414dff074db5ddee4 13487Author: Dave Airlie <airlied@redhat.com> 13488Date: Thu Nov 26 06:43:13 2009 +1000 13489 13490 fixes displayport for me here 13491 13492commit 256de862a3c017dcb4ed0d8689294a552eda0d18 13493Author: Dave Airlie <airlied@redhat.com> 13494Date: Thu Nov 26 06:35:39 2009 +1000 13495 13496 displayport: cleanup call dp from dpms on instead. 13497 13498commit 05551295c5e0946745163f17e5c1d3d41b94bcbf 13499Author: Kristian Høgsberg <krh@bitplanet.net> 13500Date: Tue Nov 24 21:09:03 2009 -0500 13501 13502 dri2: Use drmGetDeviceNameFromFd() instead of open coded loop. 13503 13504commit 0061c4db1d3aecdca13efb5133ab8784dc37df95 13505Author: Dave Airlie <airlied@redhat.com> 13506Date: Wed Nov 25 15:39:10 2009 +1000 13507 13508 Revert "r600: enable multiple operations in one CS" 13509 13510 turn this off for now, on my rv635 desktop, I started getting blanks 13511 in places in firefox and the odd bit of font corruption, need to 13512 track that down. 13513 13514 This reverts commit 985a065518b1d33599de33f7fe082d3302db58a6. 13515 13516commit 8b28534bcf877557a5681fa3b4f107c74615d3de 13517Author: Dave Airlie <airlied@redhat.com> 13518Date: Wed Nov 25 13:24:44 2009 +1000 13519 13520 radeon/exa: change option to determine exa pixmap usage. 13521 13522 This moves to a boolean instead of using VRAM sizing. 13523 13524 as per Michel's suggestions on list. 13525 13526 Signed-off-by: Dave Airlie <airlied@redhat.com> 13527 13528commit f0acb16fa19844453adc6db3399977fba7c8a0db 13529Author: Dave Airlie <airlied@redhat.com> 13530Date: Wed Nov 25 12:04:47 2009 +1000 13531 13532 r600/xv: drop inited 3d false in xv code 13533 13534commit 985a065518b1d33599de33f7fe082d3302db58a6 13535Author: Dave Airlie <airlied@redhat.com> 13536Date: Wed Nov 25 11:54:08 2009 +1000 13537 13538 r600: enable multiple operations in one CS 13539 13540 This switches on multiple ops in a single CS under KMS/DRI2. 13541 13542 It gets for on a Pentium D 3 + rv740 from 330,000 to 500,000 13543 with x11perf -aa10text. 13544 13545 It also knocks a couple of seconds of gtkperf -a 13546 13547 Signed-off-by: Dave Airlie <airlied@redhat.com> 13548 13549commit af816ac752820255f245793b53a7cca5a4a49cd4 13550Author: Dave Airlie <airlied@redhat.com> 13551Date: Wed Nov 25 11:53:07 2009 +1000 13552 13553 r600: fixup problems with EXA operation reset for multiple ops 13554 13555 To put multiple ops into one CS, you can't just discard the whole 13556 IB. This add supports for reset the CS cdw to the correct place 13557 after an op discards. 13558 13559 Still doesn't enable the final accel bits. 13560 13561commit 3d8dcbc29323a3c644100bec13aa93f024653bd3 13562Author: Dave Airlie <airlied@redhat.com> 13563Date: Wed Nov 25 11:43:57 2009 +1000 13564 13565 kms: flush cs on close screen. 13566 13567 this is needed for server recycle. 13568 13569 Signed-off-by: Dave Airlie <airlied@redhat.com> 13570 13571commit 19f1a357944f9d8f4567a2691a68067ec033ccb7 13572Author: Dave Airlie <airlied@redhat.com> 13573Date: Wed Nov 25 11:41:52 2009 +1000 13574 13575 radeon: fix check for no work in operation 13576 13577commit 3a460a14b9603159f10d89da27b559c36a184e27 13578Author: Dave Airlie <airlied@redhat.com> 13579Date: Wed Nov 25 10:33:17 2009 +1000 13580 13581 r600: refactor code to help future acceleration speedups. 13582 13583 This changes the vertex buffer index to be an offset, and 13584 records the start of the vb for each operation and uses 13585 that to set the operations up. 13586 13587 This still flushes after each operation to make sure we have 13588 no regressions in non-kms/kms cases. 13589 13590 Signed-off-by: Dave Airlie <airlied@redhat.com> 13591 13592commit 797a3f0c71c94477eec565ea2c95553c6f66d9fd 13593Author: Dave Airlie <airlied@redhat.com> 13594Date: Wed Nov 25 10:07:59 2009 +1000 13595 13596 kms: allow prepare copy to fail without dying 13597 13598commit ba76acc76e0d0a51a1c488e4d5494a71e155a05b 13599Author: Dave Airlie <airlied@redhat.com> 13600Date: Tue Nov 24 15:38:47 2009 +1000 13601 13602 radeon/kms: drop special r600 indirect flush 13603 13604 This just merges it with the main kms cs flush 13605 13606commit 3c527d752d3cced219c6113acb932511b16e35fb 13607Author: Dave Airlie <airlied@redhat.com> 13608Date: Tue Nov 24 15:11:12 2009 +1000 13609 13610 r600: enable mixed pixmaps 13611 13612commit b9eb2380df5145c85b2ef694a04628c055c47429 13613Author: Alex Deucher <alexdeucher@gmail.com> 13614Date: Mon Nov 23 12:12:02 2009 -0500 13615 13616 DCE3.2: fix uniphy2 dvi issues 13617 13618 In some cases the atom transmitter table sets the 13619 golden value of this reg differently which some monitors 13620 don't like. I haven't had time to dig further, so this 13621 works around it for now. 13622 13623 Fixes fdo bug 24313 13624 13625 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13626 13627commit d0dd5122d8d75b7c4a72df07dc7f562b2e75dd5e 13628Author: Gaetan Nadon <memsize@videotron.ca> 13629Date: Mon Nov 23 09:25:05 2009 -0500 13630 13631 Makefile.am: add ChangeLog and INSTALL on MAINTAINERCLEANFILES 13632 13633 Now that the INSTALL file is generated. 13634 Allows running make maintainer-clean. 13635 13636commit 3d7888aa0493f46d826f4d80dea36f1f642f3c6b 13637Author: Dave Airlie <airlied@redhat.com> 13638Date: Mon Nov 23 15:26:15 2009 +1000 13639 13640 radeon: fix warning in printf 13641 13642commit 3a1a8b7b73424a769baf917bba2223ea6515b596 13643Author: Dave Airlie <airlied@redhat.com> 13644Date: Mon Nov 23 15:07:10 2009 +1000 13645 13646 radeon: surround mixed pixmaps usage with ifdef check for old servers 13647 13648commit e20af9c94982ec6487dae42c141d33cf6a7a2eb4 13649Author: Dave Airlie <airlied@redhat.com> 13650Date: Fri Nov 20 12:15:02 2009 +1000 13651 13652 radeon: avoid using hw pixmaps when we have little VRAM. 13653 13654 This patch returns NULL for pixmap creation when we are using 13655 mixed pixmaps and the pixmap has a size. 13656 13657 The size check is necessary for the front buffer. 13658 13659 We add a flag to force pixmap creation for certain pixmaps 13660 that need to be hw, like the DRI2 and Xv ones. 13661 13662 Idea from Michel and workarounds from Ben Skeggs. 13663 13664 v2: add Option "EXALowVRAM" to allow configuring this, value in MBs. 13665 13666 Signed-off-by: Dave Airlie <airlied@redhat.com> 13667 13668commit 543338842ba7fe7da4902947d386ea1cabd1cdf0 13669Author: Alex Deucher <alexdeucher@gmail.com> 13670Date: Fri Nov 20 16:38:43 2009 -0500 13671 13672 dp fix timing 13673 13674 aux channel atom command table delay is 10 usec units 13675 this gets the link training working, but alas, no image. 13676 13677commit dd3eab848cf352bb96c3d01fe6028d8a4a8e451e 13678Author: Alex Deucher <alexdeucher@gmail.com> 13679Date: Fri Nov 20 01:06:49 2009 -0500 13680 13681 r300: render target limit is 2560 13682 13683 limit was increased to 4021 in r4xx. 13684 13685 fixes fdo bug 25191. 13686 13687 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13688 13689commit df35d40d4ec260cdab57a9b36e4e96b2a6c2a827 13690Author: Alex Deucher <alexdeucher@gmail.com> 13691Date: Thu Nov 19 20:01:42 2009 -0500 13692 13693 DP: hack that makes DP work. 13694 13695 Something's wrong with the status check. 13696 13697commit f42412ec4dbd21f562a7fe4239c0a1f17ef1b8f2 13698Author: Alex Deucher <alexdeucher@gmail.com> 13699Date: Thu Nov 19 19:40:54 2009 -0500 13700 13701 wip 13702 13703commit eb9bc133fc426e67b397e661bfd22bf62009d9d3 13704Author: Dave Airlie <airlied@redhat.com> 13705Date: Fri Nov 20 09:23:31 2009 +1000 13706 13707 kms: recalculate the flush limits after screen resize. 13708 13709 When we resize the front buffer we need to reduce the flush limits 13710 appropriately. 13711 13712 Signed-off-by: Dave Airlie <airlied@redhat.com> 13713 13714commit f7f58ef4c042e492618665a6c5555e8e67387ab3 13715Author: Dave Airlie <airlied@redhat.com> 13716Date: Fri Nov 20 09:22:39 2009 +1000 13717 13718 radeon: r100/r200 have a 2047 scissor limit 13719 13720 We were overflowing this in my case with a 2704 width desktop, 13721 so videos were stopping around 700 bytes across the screen. 13722 13723 can I haz shatter already? 13724 13725 Signed-off-by: Dave Airlie <airlied@redhat.com> 13726 13727commit 54c09778d79de1516ee511d6d7d5b54dd4bf9130 13728Author: Alex Deucher <alexdeucher@gmail.com> 13729Date: Thu Nov 19 13:32:30 2009 -0500 13730 13731 DP wip 13732 13733 - use coherent 13734 - fix up lane/clock numbers 13735 13736commit 02d017a0552db2c8a48506c5b26c48a1a2ab75b6 13737Author: Gaetan Nadon <memsize@videotron.ca> 13738Date: Wed Oct 28 14:41:41 2009 -0400 13739 13740 INSTALL, NEWS, README or AUTHORS files are missing/incorrect #24206 13741 13742 Automake 'foreign' option is specified in configure.ac. 13743 Remove from Makefile.am 13744 13745commit 396ee9b7e7e3b1dd64532270c4e32cdb33857596 13746Author: Gaetan Nadon <memsize@videotron.ca> 13747Date: Wed Oct 28 14:09:09 2009 -0400 13748 13749 INSTALL, NEWS, README or AUTHORS files are missing/incorrect #24206 13750 13751 Add missing INSTALL file. Use standard GNU file on building tarball 13752 README may have been updated 13753 Remove AUTHORS file as it is empty and no content available yet. 13754 Remove NEWS file as it is empty and no content available yet. 13755 13756commit 1dcda0c4d5d506e18a7989ff05f2ad8569a73f23 13757Author: Gaetan Nadon <memsize@videotron.ca> 13758Date: Mon Oct 26 12:54:21 2009 -0400 13759 13760 Several driver modules do not have a ChangeLog target in Makefile.am #23814 13761 13762 The git generated ChangeLog replaces the hand written one. 13763 Update configure.ac to xorg-macros level 1.3. 13764 Use XORG_DEFAULT_OPTIONS which replaces four XORG_* macros 13765 Update Makefile.am to add ChangeLog target if missing 13766 Remove ChangeLog from EXTRA_DIST or *CLEAN variables 13767 This is a pre-req for the INSTALL_CMD 13768 13769commit 5b517bcc353e7111d2365a213505f27cfae30924 13770Author: Gaetan Nadon <memsize@videotron.ca> 13771Date: Thu Oct 22 12:34:16 2009 -0400 13772 13773 .gitignore: use common defaults with custom section # 24239 13774 13775 Using common defaults will reduce errors and maintenance. 13776 Only the very small or inexistent custom section need periodic maintenance 13777 when the structure of the component changes. Do not edit defaults. 13778 13779commit 605c869076985b69350c893cf2bbcc90de8dc6bf 13780Author: Alex Deucher <alexdeucher@gmail.com> 13781Date: Wed Nov 18 01:51:12 2009 -0500 13782 13783 more dp cleanup 13784 13785 - clean up handling of linkb, etc. 13786 - add support for ATOM_TRANSMITTER_ACTION_SETUP_VSEMPH 13787 to existing transmitter control function 13788 - fix lots of warnings 13789 13790commit bce1911f66f446b793125ac8cd5463bf6292629e 13791Author: Alex Deucher <alexdeucher@gmail.com> 13792Date: Wed Nov 18 00:04:33 2009 -0500 13793 13794 radeon: pull previous displayport work into master 13795 13796commit 437113124bbd6fb166825169eabec4dfde900dd9 13797Author: Alex Deucher <alexdeucher@gmail.com> 13798Date: Tue Nov 17 14:16:22 2009 -0500 13799 13800 radeon: deal with connectors sourced to the same encoder 13801 13802 Some systems have multiple connectors connected to the 13803 same encoder; e.g., DVI and HDMI connected to the same 13804 encoder with the same ddc line. Since we expose 13805 connectors as xrandr outputs, randr treats them separately 13806 which results in it trying to source the same encoder to 13807 different crtcs. If we have an HDMI and DVI-D port on the 13808 same encoder, pick the one to be considered connected 13809 based on the edid (HDMI if edid indicates HDMI, DVI 13810 otherwise). Also, don't turn off (dpms) encoders that are 13811 shared. 13812 13813 Fixes fdo bug 21767. 13814 13815 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13816 13817commit 9b8605d4e2bf7703c7848f3e2022a8c7328a3acd 13818Author: Alex Deucher <alexdeucher@gmail.com> 13819Date: Tue Nov 17 11:45:09 2009 -0500 13820 13821 radeon: fix potential memory leak in ddc code 13822 13823 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13824 13825commit ce8299962003de572122561a6f7f61eaccf633b2 13826Author: Dave Airlie <airlied@redhat.com> 13827Date: Tue Nov 17 15:23:30 2009 +1000 13828 13829 kms: check for 0 mask for cloned outputs. 13830 13831 Signed-off-by: Dave Airlie <airlied@redhat.com> 13832 13833commit 06d68f7fa3870f9601e837e1834d33e8395008b1 13834Author: Dave Airlie <airlied@redhat.com> 13835Date: Tue Nov 17 15:10:04 2009 +1000 13836 13837 kms: add cloning support. 13838 13839 We have to translate from the kernel encoder cloning to the randr 13840 connector cloning, we do this by constructing an encoder mask per output 13841 and an encoder cloning mask per output then comparing them. 13842 13843 Signed-off-by: Dave Airlie <airlied@redhat.com> 13844 13845commit 7587ce16ab0380337fe801f457c1d1d9b4141cc5 13846Author: Alex Deucher <alexdeucher@gmail.com> 13847Date: Mon Nov 16 19:16:03 2009 -0500 13848 13849 r600: remove un-needed format conversions 13850 13851 we do the normalization and xforms in the vertex 13852 shader. 13853 13854 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13855 13856commit ba605b5c9ff48765f2b66960aa4cc81b6235d012 13857Author: Alex Deucher <alexdeucher@gmail.com> 13858Date: Mon Nov 16 19:00:52 2009 -0500 13859 13860 r600: fix num format in vtx fetch 13861 13862 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13863 13864commit 0c4710c67a2fee2061fc3da43c9f908585693cfa 13865Author: Alex Deucher <alexdeucher@gmail.com> 13866Date: Thu Nov 12 11:11:17 2009 -0500 13867 13868 radeon: man page updates 13869 13870commit 2af2744c80f3307e6d84edba9479be5bd36e1d03 13871Author: Alex Deucher <alexdeucher@gmail.com> 13872Date: Wed Nov 11 13:12:24 2009 -0500 13873 13874 radeon: fix support for external tmds on some legacy chips 13875 13876 Should fix bug 11801. 13877 13878 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13879 13880commit f17657fdc83d8f4c0386d2c7dade98de5b94acbe 13881Author: Alex Deucher <alexdeucher@gmail.com> 13882Date: Mon Nov 9 16:59:11 2009 -0500 13883 13884 r600/r700: typo, fix mask of DB_ALPHA_TO_MASK 13885 13886 noticed by glisse. 13887 13888commit 0ee7763fb33bf7fbc4ae1e8372cbc74578bfb720 13889Author: Alex Deucher <alexdeucher@gmail.com> 13890Date: Mon Nov 9 15:09:34 2009 -0500 13891 13892 IGP: some IGP chips report as AGP 13893 13894 Set bus type appropriately. fixes bug 25002 13895 13896 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13897 13898commit da1fcddaade58e371c7bee555ce32944c689f810 13899Author: Alex Deucher <alexdeucher@gmail.com> 13900Date: Thu Nov 5 14:13:49 2009 -0500 13901 13902 radeon: properly fix unused label warnings 13903 13904commit 6e496f46fdc90d0a6cef41e9d45a6458aaed1eca 13905Author: Alex Deucher <alexdeucher@gmail.com> 13906Date: Thu Nov 5 14:11:55 2009 -0500 13907 13908 Revert "Fix warning about unused defined labels in radeon_exa_funcs.c." 13909 13910 This reverts commit b6ac42cce2156880ac18b32e15acbff40b2f53e9. 13911 13912 This patch is very broken. 13913 13914commit b6ac42cce2156880ac18b32e15acbff40b2f53e9 13915Author: Tomáš Chvátal <scarabeus@gentoo.org> 13916Date: Sat Oct 31 23:07:07 2009 +0100 13917 13918 Fix warning about unused defined labels in radeon_exa_funcs.c. 13919 13920 Signed-off-by: Tomáš Chvátal <scarabeus@gentoo.org> 13921 13922commit 87d7235790866f9c19ef08972d5237f09d940cd9 13923Author: Alex Deucher <alexdeucher@gmail.com> 13924Date: Thu Nov 5 10:23:03 2009 -0500 13925 13926 EXA: fallback if no pMaskPicture->pDrawable 13927 13928 A solid or gradient mask could be used for blending 13929 the source picture onto the destination picture. 13930 13931 Fixes fdo bug 24838 13932 13933 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13934 13935commit 873897c54429e230a23011a9dd2f1069f4eb7752 13936Author: Alex Deucher <alexdeucher@gmail.com> 13937Date: Thu Nov 5 00:51:11 2009 -0500 13938 13939 DCE3+: call transmitter init on mode set 13940 13941 Generally this is done at post, but might not always 13942 be done with softboot or for connectors on docking 13943 stations. 13944 13945 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13946 13947commit 3028374488cc0f34942ac372c8d05cf15898a613 13948Author: Alex Deucher <alexdeucher@gmail.com> 13949Date: Wed Nov 4 18:48:03 2009 -0500 13950 13951 ATOM/DCE3+: fix up usPixelClock calculation for Transmitter tables 13952 13953 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13954 13955commit 52973e8f2c4104de60d00c0e5c29a31f77bf3de7 13956Author: Alex Deucher <alexdeucher@gmail.com> 13957Date: Sun Nov 1 13:58:03 2009 -0500 13958 13959 radeon/atom: fix tv-out on r5xx 13960 13961 fixes bug 24796 13962 13963 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13964 13965commit 19b4f5f9b647cd470616b8ec1d34f1e4afa357ad 13966Author: Alex Deucher <alexdeucher@gmail.com> 13967Date: Tue Oct 27 11:55:20 2009 -0400 13968 13969 ATOM: fix up DVO for DCE 3.x 13970 13971 AdjustDisplayPll takes care of this for us. 13972 13973commit 5a0019126a57138ee506d9a66738c9e8b75cbb96 13974Author: Alex Deucher <alexdeucher@gmail.com> 13975Date: Tue Oct 27 11:39:06 2009 -0400 13976 13977 radeon: fix DVO on AVIVO chips 13978 13979 DVO generally requires 2x ppll on AVIVO cards due to the 13980 way the tmds chip is wired up. Because of this, disable 13981 cloning of DVO outputs. 13982 13983 fixes fdo bug 21857. 13984 13985 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 13986 13987commit 2a6eeec72d6bb04a32225883f431c1d2e8cff123 13988Author: Alex Deucher <alexdeucher@gmail.com> 13989Date: Tue Oct 27 11:18:44 2009 -0400 13990 13991 atom: loosen pll min output limits 13992 13993 Limiting the pll output range is a good thing generally as 13994 it limits the number of possible pll combinations for a given 13995 frequency presumably to the ones that work best on each card. 13996 That's why the limits are in the bios tables. However, certain 13997 duallink DVI monitors seem to like pll combinations that would 13998 be limited by this at least on pre-DCE 3.0 r6xx hardware. This 13999 might need to be adjusted per family or per clock range in the 14000 future. 14001 14002 Fixes fdo bug 24727. 14003 14004 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 14005 14006commit f0d9d80fee4176eaba9435a9539f29d0eefe2a87 14007Author: Luca Tettamanti <kronos.it@gmail.com> 14008Date: Sat Oct 24 16:45:23 2009 -0400 14009 14010 Disable color tiling on r600/r700 14011 14012 Not supported yet. 14013 14014commit ac499c3c0ea92f83ebd8127bb8cb5de625de6c94 14015Author: Alex Deucher <alexdeucher@gmail.com> 14016Date: Fri Oct 23 10:48:55 2009 -0400 14017 14018 radeon: clamp the internal FB map to the aperture size v2 14019 14020 We don't use the invisible memory yet and on cards with 14021 large amounts of vram this can cause the top of GART 14022 calculation to overflow. 14023 14024 Fixes bug fdo bug 24301: 14025 http://bugs.freedesktop.org/show_bug.cgi?id=24301 14026 14027 v2: only clamp cards with more than 512 MB. This seems 14028 to cause problems on some older cards due to the way the 14029 drm and ddx set up the internal memory map. 14030 14031commit e57b54daf318ff59315a1c9ed21934f288c5c782 14032Author: Alex Deucher <alexdeucher@gmail.com> 14033Date: Thu Oct 22 16:07:58 2009 -0400 14034 14035 R7xx: program additional CUR/GRPH regs for 40 bit addresses 14036 14037 The *_HIGH regs are reversed. The secondary ones are in the 14038 primary block and vice versa. 14039 14040commit 4cf06dfba617529291ce4b4c306c4fc1bba110ee 14041Author: Alex Deucher <alexdeucher@gmail.com> 14042Date: Thu Oct 22 02:08:45 2009 -0400 14043 14044 ATOM: use indirect addressing for regs >= 0x10000 14045 14046commit 66b194a78c470cb3978f310828dd96c3f3e96944 14047Author: Alex Deucher <alexdeucher@gmail.com> 14048Date: Wed Oct 21 15:58:18 2009 -0400 14049 14050 ATOM: fix up get clock info 14051 14052 Newer revisions of the atom firmware table have 14053 changed. This helps select better pll dividers 14054 in some cases. Noticed by Mathias Froehlich. 14055 14056commit bd89b7501f294ac645390ef144df569953c81dc4 14057Author: Alex Deucher <alexdeucher@gmail.com> 14058Date: Sun Oct 18 16:21:36 2009 -0400 14059 14060 radeon: fixup CustomEDID option 14061 14062 - add support for analog option to force edid to analog 14063 mode 14064 - fix compilation on older xservers 14065 - updated man page 14066 14067commit 255ac8d54e3dc12e579e05edb41b40946a00eab8 14068Author: Alex Deucher <alexdeucher@gmail.com> 14069Date: Tue Oct 13 12:27:52 2009 -0400 14070 14071 radeon: add XV_CRTC attribute for textured video 14072 14073 This attribute allows the user to override which 14074 crtc is synced with when XV_VSYNC is enabled. This 14075 is useful for clone modes where the user can might want 14076 to override the default. 14077 14078commit 801d900add4731f3c424ebb780ad2dbd7a1c6a25 14079Author: Alex Deucher <alexdeucher@gmail.com> 14080Date: Mon Oct 12 10:00:57 2009 -0400 14081 14082 ATOM: minor cleanups to pll setup 14083 14084commit f4407962cd7b272e0860319f11f6a6583ef226c2 14085Author: Csillag Kristof <csillag.kristof@united-consult.hu> 14086Date: Fri Oct 9 18:00:09 2009 -0400 14087 14088 Fix some issues with CustomEDID support 14089 14090 - Fixes a memory allocation problem introduced with the previous 14091 version of the CustomEDID patch (now in GIT), and 14092 - Enhances the syntax of the CustomEDID option so that I can force 14093 digital output, even if the EDID was acquired over an analog link. 14094 14095commit de55995e82c3875f70b6394fff440d695d062113 14096Author: Alex Deucher <alexdeucher@gmail.com> 14097Date: Fri Oct 9 11:07:30 2009 -0400 14098 14099 r600 EXA: fix up mask reg mixup 14100 14101commit a168caf5349b757873ad89d977772feaa7fb6e58 14102Author: Alex Deucher <alexdeucher@gmail.com> 14103Date: Fri Oct 9 10:32:26 2009 -0400 14104 14105 r600 EXA: fix some compile warnings 14106 14107commit fb8dda2928bd0def2b34c8c8e414129e650d6eb5 14108Author: Alex Deucher <alexdeucher@gmail.com> 14109Date: Fri Oct 9 10:30:21 2009 -0400 14110 14111 r600 EXA: properly check planemask 14112 14113 We only support per-component masks. 14114 14115commit 1b25bad2d8ebe481ffb4b0fd3a85ab636bd4fec2 14116Author: Dave Airlie <airlied@linux.ie> 14117Date: Fri Oct 9 20:09:52 2009 +1000 14118 14119 Revert "radeon: clamp the internal FB map to the aperture size" 14120 14121 This reverts commit 5f846360c46f5a989f5d0fde6d251cdbd61d4968. 14122 14123 Numerous reports of system hangs since this, I'm guessing 14124 some sort of conflict with the drm memory setup. This code 14125 has always been fragile between kernel/userspace drivers. 14126 14127commit a956d478043b44c49aed39c7aa3d576ef32cadc5 14128Author: Dave Airlie <airlied@linux.ie> 14129Date: Fri Oct 9 20:08:31 2009 +1000 14130 14131 Revert "kms: setup colormap and gamma correctly." 14132 14133 This reverts commit 60d9685abddccec17c1a9a5ec48cbe9c92543e0f. 14134 14135 which breaks colormap on 1.6 X server 14136 14137 Conflicts: 14138 14139 src/drmmode_display.c 14140 14141commit a493feb94c0d84aaf5bfd84c6940d7f03974dd86 14142Author: Dave Airlie <airlied@redhat.com> 14143Date: Fri Oct 9 15:05:25 2009 +1000 14144 14145 kms: don't use scratch pixmaps at all 14146 14147 use our own wrapper to allocate a pixmap for wrapping a bo in. 14148 14149commit 0caffbd6fd6b356b9b7d715c015285eaa91e3e36 14150Author: Dave Airlie <airlied@redhat.com> 14151Date: Fri Oct 9 15:04:56 2009 +1000 14152 14153 drmmode: reload cursors on modeset 14154 14155commit 02e12ae6be7bc1976f63848fa2854d320d5ab36e 14156Author: Alex Deucher <alexdeucher@gmail.com> 14157Date: Wed Oct 7 19:56:58 2009 -0400 14158 14159 ATOM: rework crtc modeset 14160 14161 - clean up tv timing handling 14162 - unify SetCRTCTiming and SetCRTCDTDTiming interfaces 14163 14164commit d499eeaf22f77a1294e99aa38a50aa6810bb684a 14165Author: Alex Deucher <alexdeucher@gmail.com> 14166Date: Wed Oct 7 17:43:11 2009 -0400 14167 14168 ATOM: reorder crtc dpms based on bios recommendations 14169 14170commit 4ccd2a21f9c50515b4246f35454f76ef49006c76 14171Author: Alex Deucher <alexdeucher@gmail.com> 14172Date: Wed Oct 7 16:30:03 2009 -0400 14173 14174 ATOM: add support for AdjustDisplayPll table 14175 14176 Depending on the output and clock, this table will 14177 adjust the pixelclock accordingly. 14178 14179commit 5f846360c46f5a989f5d0fde6d251cdbd61d4968 14180Author: Alex Deucher <alexdeucher@gmail.com> 14181Date: Wed Oct 7 17:01:55 2009 -0400 14182 14183 radeon: clamp the internal FB map to the aperture size 14184 14185 We don't use the invisible memory yet and on cards with 14186 large amounts of vram this can cause the top of GART 14187 calculation to overflow. 14188 14189 Fixes bug fdo bug 24301: 14190 http://bugs.freedesktop.org/show_bug.cgi?id=24301 14191 14192commit 186bf719c394116341abff471fb4b4803650da9a 14193Author: Dave Airlie <airlied@redhat.com> 14194Date: Wed Oct 7 17:23:36 2009 +1000 14195 14196 radeon: fix rotation since mixed pixmaps. 14197 14198 Passing null in here stops the mixed pixmap code doing it wrong, 14199 but I'm not 100% sure how it is meant to work, we have the 14200 same issue with the real front bo by the looks of it. 14201 14202commit e08411af1aa8c7d4233ba593b84360397cdbb307 14203Author: Roel Kluin <roel.kluin@gmail.com> 14204Date: Tue Oct 6 18:31:17 2009 -0400 14205 14206 radeon: Fix duplicated bit settings 14207 14208 [agd5f: adapted from kms patch] 14209 14210 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 14211 14212commit e59ae08270711512e64b70f79b6476cc2c52d230 14213Author: Michel Dänzer <daenzer@vmware.com> 14214Date: Tue Oct 6 00:13:09 2009 +0200 14215 14216 Don't call radeon_cs_flush_indirect() without KMS. 14217 14218 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=24316 . 14219 14220commit 457646d734558672402b6bcd15cadb66741edbed 14221Author: Jerome Glisse <jglisse@redhat.com> 14222Date: Mon Oct 5 15:52:42 2009 +0200 14223 14224 kms: assume accel is working is info ioctl return -EINVAL 14225 14226 -EINVAL should only happen if the info ioctl doesn't support 14227 the get accel query. This patch assume that on such case accel 14228 is working. 14229 14230commit 9d596562496863d65850306d2126d8df98464de4 14231Author: Dave Airlie <airlied@linux.ie> 14232Date: Mon Oct 5 18:35:20 2009 +1000 14233 14234 kms: only touch gamma stuff on 1.7 14235 14236 for correct colormap operation you probably want 1.7 14237 14238commit 876f43d4682df4d47d3e158f75e0ca6040f41c70 14239Author: Alex Deucher <alexdeucher@gmail.com> 14240Date: Mon Oct 5 02:23:03 2009 -0400 14241 14242 R1/2xx EXA: Minimise number of draw primitives used for Composite operations. 14243 14244commit 86cafb8affea448bdf58300044e755201b822d2a 14245Author: Dave Airlie <airlied@redhat.com> 14246Date: Mon Oct 5 14:17:25 2009 +1000 14247 14248 kms: don't use scratch pixmaps when copying fbcon. 14249 14250 scratch pixmaps seem to interact badly with mixed pixmaps, it appears 14251 some state may be getting left around in the privates somewhere, since 14252 scratch pixmap headers don't get destroyed. 14253 14254commit f8471512ea9f1d38140dfe98a0f832e9f935f51b 14255Author: Csillag Kristof <csillag.kristof@united-consult.hu> 14256Date: Sun Oct 4 18:18:56 2009 -0400 14257 14258 radeon: add support for Custom EDID 14259 14260 Allows you to specify an edid per output from a file 14261 to override what is detected by DDC. Useful for 14262 problematic monitors or KVM switches that block 14263 DDC. Specifying an EDID that is not compatible with 14264 your monitor could damage your monitor so use with 14265 caution. 14266 14267 agd5f: cache the custom edid at startup so we don't 14268 have to read it from file every time the output is 14269 queried. 14270 14271commit 60d9685abddccec17c1a9a5ec48cbe9c92543e0f 14272Author: Dave Airlie <airlied@redhat.com> 14273Date: Mon Oct 5 12:41:30 2009 +1000 14274 14275 kms: setup colormap and gamma correctly. 14276 14277 We need to program the gamma at modeset. no need to use 14278 old cmap interface at all. 14279 14280commit b0b10e548fd69b3ff67a0f9b71456b89c61cb4d6 14281Author: Alex Deucher <alexdeucher@gmail.com> 14282Date: Sun Oct 4 17:07:32 2009 -0400 14283 14284 R6/7xx EXA: Add support for BGRA picture formats. 14285 14286 Also add mysteriously missing case statements for ABGR destination formats. 14287 14288commit 8404c37b2ff95aba31db97676f106ec2f30fda7f 14289Author: Dave Airlie <airlied@linux.ie> 14290Date: Sun Oct 4 18:18:51 2009 +1000 14291 14292 radeon: fix build without kms 14293 14294commit 9460ea864b12ec1fbd11c5d9a20bb5a4279d9d3d 14295Author: Michel Dänzer <daenzer@vmware.com> 14296Date: Sat Oct 3 16:33:33 2009 +0200 14297 14298 Fix KMS on big endian machines. 14299 14300 Requires at least xserver 1.7 to work properly. 14301 14302 Also make sure the front buffer is and stays tiled if colour tiling is enabled. 14303 14304commit aee6b6f2c120baf477b4845ddc1a94637c31db2f 14305Author: Michel Dänzer <daenzer@vmware.com> 14306Date: Sat Oct 3 16:33:33 2009 +0200 14307 14308 EXA: Bail earlier from PrepareCopy hook on trivially unsupportable conditions. 14309 14310commit 30031b169def5caadb394225919a6079a0957142 14311Author: Michel Dänzer <daenzer@vmware.com> 14312Date: Sat Oct 3 16:33:33 2009 +0200 14313 14314 R3/5xx EXA: Add support for BGRA picture formats. 14315 14316 Also add mysteriously missing case statements for ABGR destination formats. 14317 14318commit 4b4ce36081ca151c24e028c54b59986f41731a73 14319Author: Michel Dänzer <daenzer@vmware.com> 14320Date: Sat Oct 3 16:33:32 2009 +0200 14321 14322 R3/5xx EXA: Minimise number of draw primitives used for Composite operations. 14323 14324 This should reduce the kernel CS checker overhead, if nothing else. 14325 14326 I'll leave porting this to other chipset families to others who can test it. 14327 14328commit eade1e5be159c9f2965d611925596d33cab11d6d 14329Author: Michel Dänzer <daenzer@vmware.com> 14330Date: Sat Oct 3 16:33:32 2009 +0200 14331 14332 EXA: Don't always flush when switching between 2D/3D engines with KMS. 14333 14334 This seems to work fine here now and help 2D performance quite a bit, let's see 14335 if it breaks anybody else's setup... 14336 14337commit 926e414fc835ace141c066830d11d8ce32dbb06c 14338Author: Michel Dänzer <daenzer@vmware.com> 14339Date: Sat Oct 3 16:33:32 2009 +0200 14340 14341 KMS: Double-buffer textured video source image upload. 14342 14343 In order to avoid stalling on previous frame. 14344 14345 OTOH without KMS we can't do this but have to wait for the previous frame to 14346 finish rendering. 14347 14348commit 6a3b75fae4147244212298ff55c7ab36c3d80d30 14349Author: Michel Dänzer <daenzer@vmware.com> 14350Date: Sat Oct 3 16:33:31 2009 +0200 14351 14352 KMS: Don't hide HW cursor when it's updated. 14353 14354 Inspired by the intel driver; might avoid/reduce HW cursor flicker in some 14355 cases. 14356 14357commit c4ba15b4871c70cced7f1761e24be692cdb28d2b 14358Author: Michel Dänzer <daenzer@vmware.com> 14359Date: Sat Oct 3 16:33:31 2009 +0200 14360 14361 Fix some compiler warnings. 14362 14363commit cc45856a18dd3e6f7e44d9eb507b31419da70977 14364Author: Kenneth Graunke <kenny@whitecape.org> 14365Date: Fri Oct 2 14:31:36 2009 -0400 14366 14367 radeon: Remove gatos message from driver startup 14368 14369 Remove a message on driver startup which directs people to the old GATOS 14370 website; said site contains highly out of date software which is no longer 14371 necessary. 14372 14373commit d302481493ff61eb7f8648b7afe7407b6b653cf7 14374Author: Jerome Glisse <jglisse@redhat.com> 14375Date: Wed Sep 16 15:18:40 2009 +0200 14376 14377 radeon/kms: fallback to shadowfb if kernel report acceleration is off 14378 14379 This will fallback to shadowfb is GPU accel failed somewhere during 14380 KMS kernel init. 14381 14382commit 7968e1fb89f6b59d1654df48249bf4b81990c008 14383Author: Alex Deucher <alexdeucher@gmail.com> 14384Date: Fri Sep 25 19:05:38 2009 -0400 14385 14386 radeon: fix segfault in MMIO path in RADEONInit3DEngine 14387 14388 fixes fdo bug 24158. 14389 14390 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 14391 14392commit 0380a9cf98d34e88dd81370ade5525680ec89c02 14393Author: Alex Deucher <alexdeucher@gmail.com> 14394Date: Fri Sep 25 16:46:43 2009 -0400 14395 14396 kms/r600: add support for vline relocs 14397 14398 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 14399 14400commit 3efecebb10de7f7bacf9f8c57ae20fd508097294 14401Author: Alex Deucher <alexdeucher@gmail.com> 14402Date: Fri Sep 25 14:44:41 2009 -0400 14403 14404 radeon: fix vline handling for kms 14405 14406 drm crtc ids do not correspond to actual hw crtcs, 14407 as such the vline stuff was never enabled for Xv. 14408 14409 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 14410 14411commit 9733dcde0a21b7503aa20254724f2910b541b990 14412Author: Dave Airlie <airlied@linux.ie> 14413Date: Sat Sep 26 06:15:58 2009 +1000 14414 14415 r600: fix build 14416 14417commit ff18595eace42ddfc931f4f948cb5adf59ebcf52 14418Author: Dave Airlie <airlied@linux.ie> 14419Date: Sat Sep 26 06:04:51 2009 +1000 14420 14421 r600/xv: fix dst bo write domain 14422 14423commit 718a611f3cc7f5eea95cd73a9dcc7913c922fcbc 14424Author: Dave Airlie <airlied@linux.ie> 14425Date: Sat Sep 26 06:03:51 2009 +1000 14426 14427 radeon: fix zaphod 14428 14429commit 86c367be517e992aabc1bcfea2a2dde1a1ae520c 14430Author: Alex Deucher <alexdeucher@gmail.com> 14431Date: Fri Sep 25 13:03:44 2009 -0400 14432 14433 r600: fix bo accounting for Xv 14434 14435 fixes Xv after changes in: 14436 8f80e37eed3ec028718b4e71bbb9b598847fd94e 14437 14438 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 14439 14440commit 8f80e37eed3ec028718b4e71bbb9b598847fd94e 14441Author: Dave Airlie <airlied@linux.ie> 14442Date: Fri Sep 25 11:37:38 2009 +1000 14443 14444 r600: space check wasn't correct. 14445 14446 the r6xx space check code was incorrect for the DFS case, since 14447 the dst bo was in GTT but the hardcoded cp_start function 14448 assumed the bos were where it thought. 14449 14450 Ripped out assumptions and replaced with code more like other radeons. 14451 14452commit 90669f6cb5ada9067d6fccc7f54c69b367862f42 14453Author: Alex Deucher <alexdeucher@gmail.com> 14454Date: Thu Sep 24 10:40:40 2009 -0400 14455 14456 radeon: remove old pre-randr rotation cruft 14457 14458 leftover from the randr 1.2 conversion. 14459 14460 Signed-off-by: Alex Deucher <alexdeucher@gmail.com> 14461 14462commit 97a4e747bfac14f34646c55ddf639e8fe22f2f55 14463Author: Michael Olbrich <m.olbrich@pengutronix.de> 14464Date: Sun Sep 20 14:19:41 2009 +0200 14465 14466 use AC_CHECK_HEADER instead of AC_CHECK_FILE 14467 14468 AC_CHECK_FILE is not possible when cross-compiling. Use AC_CHECK_HEADER / 14469 AC_PREPROC_IFELSE instead. 14470 14471 [ Michel Dänzer: Shuffled things around slightly to make it work on my setup ] 14472 14473 Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de> 14474 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 14475 14476commit 579070ed9f0218a975e840331b6fe9fa8fd05789 14477Author: Alex Deucher <alexdeucher@gmail.com> 14478Date: Sat Sep 19 19:15:14 2009 -0400 14479 14480 kms: disable dga 14481 14482 Can be dangerous with kms: 14483 http://lists.freedesktop.org/archives/xorg-devel/2009-September/002113.html 14484 14485commit da7487f6ac7b32ed7866af53da2925aa52ed7aae 14486Author: Alex Deucher <alexdeucher@gmail.com> 14487Date: Sat Sep 19 13:08:02 2009 -0400 14488 14489 kms: init dga using xf86DiDGAInit 14490 14491 fixes cursor problems in games like UT2004 and Penumbra Overture 14492 14493 see fdo bug 24034 14494 14495commit 52279251fae9df99c569c16e2522bbd346d8ec38 14496Author: Dave Airlie <airlied@redhat.com> 14497Date: Fri Sep 18 14:29:32 2009 +1000 14498 14499 avivo: disable VGA rendering core when starting X. 14500 14501 this thing can be rendering to VRAM when we don't expect it. 14502 turn it off. 14503 14504 Signed-off-by: Dave Airlie <airlied@redhat.com> 14505 14506commit 2c46bafcb77c4125a27c18ad6ca2f6de5f143a2d 14507Author: Pauli Nieminen <suokkos@gmail.com> 14508Date: Thu Sep 17 15:13:51 2009 +0300 14509 14510 Fix configuration to disable KMS if not finding libdrm_radeon. 14511 14512 Most of KMS building was disabled if LIBDRM_RADEON was not detected. But at 14513 least configure was reporting KMS enabled even tough it was realy disabled. 14514 14515commit ee9f6d802988a4d7f7985687036db854f4a14fe5 14516Author: Pauli Nieminen <suokkos@gmail.com> 14517Date: Mon Sep 14 16:36:48 2009 +0300 14518 14519 radeon: Add missing libdrm_radeon cflags to compiler paramaters. 14520 14521 This fixes compilation if libdrm_radeon is installed to non-standard location like /opt. 14522 14523commit ac853ca0a950d6147cd6a0b4a1b089f5d2f080ea 14524Author: Alex Deucher <alexdeucher@gmail.com> 14525Date: Fri Sep 11 14:25:31 2009 -0400 14526 14527 atom: fix typo in asus quirks 14528 14529 Should be DVI-I, not DVI-D 14530 14531commit a88f60e7b789cd23a79e6e18a7e3fa008d15aa77 14532Author: Alex Deucher <alexdeucher@gmail.com> 14533Date: Thu Sep 10 13:42:31 2009 -0400 14534 14535 r600: add CS UTS and DFS hooks 14536 14537commit 0bb0ff0e55aaf8e64017d204e376a14f0a7384a3 14538Author: Dave Airlie <airlied@linux.ie> 14539Date: Thu Sep 10 11:48:00 2009 +1000 14540 14541 r6xx: fix EXA crash with gtkperf -a 14542 14543 unmapping was getting unbalanced. 14544 14545commit 10a58d54857484b6e89763ecf463f54dc290c702 14546Author: Dave Airlie <airlied@redhat.com> 14547Date: Wed Sep 9 14:24:49 2009 +1000 14548 14549 kms: add property support. 14550 14551 this adds support to the userspace DDX to get properties from the kernel 14552 and expose them over randr. Its most liberated from the Intel driver. 14553 14554commit fad9ee6cdae5ae45e6406e3071a84276f1b59a85 14555Author: Alex Deucher <alexdeucher@gmail.com> 14556Date: Tue Sep 8 17:50:50 2009 -0400 14557 14558 radeon: replace dga code with DiDGA 14559 14560commit 651fe5a47435c6a5cc1cdfb33c5c6601d692adcb 14561Author: Adam Jackson <ajax@redhat.com> 14562Date: Tue Sep 8 13:21:07 2009 +1000 14563 14564 radeon: fix dri2 struct zeroing 14565 14566commit 6990f2ac6478bf92929a4400ef84fb2142699204 14567Merge: 917f2d7cd 853f4c3d1 14568Author: Dave Airlie <airlied@redhat.com> 14569Date: Tue Sep 8 11:26:32 2009 +1000 14570 14571 Merge branch 'r6xx-cs' 14572 14573commit 853f4c3d1ea8f975ab2855f18d3ae336a4095091 14574Author: Dave Airlie <airlied@redhat.com> 14575Date: Tue Sep 8 11:25:39 2009 +1000 14576 14577 r600: more alignment fixups + vb map/unmap 14578 14579 I'm not so sure the vb map/unmap is a good idea, I think 14580 it pretty much locksteps the cpu/gpu, so we should really 14581 work out if we really need to flush this often, since 14582 mesa doesn't have to and we are just doing 3D ops. 14583 14584commit 917f2d7cd1815279710939edc27133ca6609460f 14585Author: Alex Deucher <alexdeucher@gmail.com> 14586Date: Sun Sep 6 10:53:29 2009 -0400 14587 14588 RV280: add agp quirk 14589 14590 deb bug 545040 14591 14592commit 78fcbf577ad6eba6399cc39f74b7ce5f9c8e265e 14593Author: Michel Dänzer <daenzer@vmware.com> 14594Date: Thu Sep 3 14:55:05 2009 +0200 14595 14596 EXA: Check for solid/gradient pictures the same way for all generations. 14597 14598 In particular, also catch them for >= R300. 14599 14600commit 794ae743c305331741e7cf45dd71c755fe325ed5 14601Author: Michel Dänzer <daenzer@vmware.com> 14602Date: Wed Sep 2 09:19:17 2009 +0200 14603 14604 EXA: Fix Composite restart dst/src switcheroo. 14605 14606 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=22636 . 14607 14608 Now, where's that brown paper bag? :} 14609 14610commit c4ab50c5eafca3d04129a72453496eb8afb26b34 14611Author: Alex Deucher <alexdeucher@gmail.com> 14612Date: Tue Sep 1 13:43:00 2009 -0400 14613 14614 radeon: add some notes about DVO setup 14615 14616commit 22fbd6476f4b3513378e848affc4dd269e9d4498 14617Author: Dave Airlie <airlied@itt42.(none)> 14618Date: Tue Sep 1 15:22:57 2009 +1000 14619 14620 r100/r200: seriously WRAP is illegal for rectangular textures 14621 14622 unless we scale all the bos up to POT 14623 14624commit 8f4196e88855f10762254fca9e0a0988e7b5562f 14625Author: Alex Deucher <alexdeucher@gmail.com> 14626Date: Mon Aug 31 19:41:59 2009 -0400 14627 14628 r6xx/r7xx: various CS fixes from Dave 14629 14630commit a32384f9b9ed73c14484a380dfa14652ed478ad6 14631Author: Alex Deucher <alexdeucher@gmail.com> 14632Date: Mon Aug 31 13:26:05 2009 -0400 14633 14634 r1xx: fix textured video 14635 14636 Copy/paste bug from when kms support was added. 14637 14638commit 3d6bf0f8cf128ee67d448491be44d345547b75c5 14639Author: Pauli Nieminen <suokkos@gmail.com> 14640Date: Fri Aug 28 17:05:44 2009 +0300 14641 14642 radeon: Fix DRI2BufferPtr to be DRI2Buffer2Ptr for xserver 1.6. 14643 14644 Bump requirement for xserver in KMS mode to 1.6.2 for DRI2Buffer2Ptr. 14645 14646commit 6521dbf3884185ca5692d35d879d251890cc4f08 14647Author: Roland Scheidegger <sroland@tungstengraphics.com> 14648Date: Sat Aug 29 12:11:03 2009 +0200 14649 14650 r100/r200: fix section size mismatch for textured video 14651 14652 this fixes the driver complaining though reportedly it still does 14653 not work (on r100) 14654 14655commit 39dfac41eebbcdacb1ae0c17cea325ea2966c4a5 14656Author: Michel Dänzer <daenzer@vmware.com> 14657Date: Fri Aug 28 12:20:04 2009 +0200 14658 14659 EXA: RENDER repeat fix and more cleanup. 14660 14661 Apparently changing pPict->repeatType causes badness, see 14662 http://bugs.freedesktop.org/show_bug.cgi?id=23560 . 14663 14664 Now we just use a normalized repeatType value derived from pPict->repeat and 14665 pPict->repeatType everywhere. 14666 14667commit 7d36f13afa5da742aa6c989a2652e92c96e5f80e 14668Author: Michel Dänzer <daenzer@vmware.com> 14669Date: Thu Aug 27 09:52:11 2009 +0200 14670 14671 KMS: Remove explicit radeon_bo_wait calls before radeon_bo_map. 14672 14673 libdrm_radeon always takes care of this implicitly now. 14674 14675commit 066c90e8f5b93c80195dc43486ad3efa083a1f3a 14676Author: Michel Dänzer <daenzer@vmware.com> 14677Date: Thu Aug 27 09:52:11 2009 +0200 14678 14679 KMS: Don't use a blit for UploadToScreen if the pixmap BO will be idle anyway. 14680 14681 Direct CPU writes should be at least as fast in that case. 14682 14683commit 174b61bb786a841cebd354e23b4b0caa85b541aa 14684Author: Michel Dänzer <daenzer@vmware.com> 14685Date: Thu Aug 27 09:52:11 2009 +0200 14686 14687 KMS: DownloadFromScreen improvements. 14688 14689 * Drop superfluous RADEONDownloadFromScreenGTT function, EXA does the same 14690 thing when we return FALSE. 14691 * Take unflushed operations into account for determining which GEM domain the 14692 pixmap BO will end up in. 14693 * Only use a blit if it ends up in VRAM. 14694 14695commit 7623e169e1f7d5afbd4108de03f28098bca519db 14696Author: Michel Dänzer <daenzer@vmware.com> 14697Date: Thu Aug 27 09:52:11 2009 +0200 14698 14699 EXA: Allocate pixmap BOs in VRAM. 14700 14701 This is the intention with EXA in xserver Git, and IME it improves performance 14702 significantly even with older versions. 14703 14704commit b90b823790657707b5bf7249095833bef959b0e8 14705Author: Michel Dänzer <daenzer@vmware.com> 14706Date: Thu Aug 27 09:52:11 2009 +0200 14707 14708 DRI2: Set tiling for depth/stencil buffers with all DRI2 interface versions. 14709 14710commit 27d9664f8862d1948dd32021f8e9bcb94f23e4ef 14711Author: Michel Dänzer <daenzer@vmware.com> 14712Date: Thu Aug 27 09:52:10 2009 +0200 14713 14714 EXA: RENDER repeat cleanups. 14715 14716 We can't rely on the server setting pPict->repeatType to something sensible 14717 when pPict->repeat is FALSE, but we can do it ourselves. 14718 14719commit 9b5517fcf7f2397a1a814d396fc9dc24769a6d9f 14720Author: Michel Dänzer <daenzer@vmware.com> 14721Date: Thu Aug 27 09:52:10 2009 +0200 14722 14723 Throttle DRI2 buffer swaps / frontbuffer flushes. 14724 14725commit e87f0f50f31a59ca1f60d4582d4a57ed00854fb7 14726Author: Alex Deucher <alexdeucher@gmail.com> 14727Date: Wed Aug 26 02:13:38 2009 -0400 14728 14729 r6xx/r7xx: set EXA_HANDLES_PIXMAPS 14730 14731commit 3212c26b90c0f6f1a7248b4da3ed985a9c2e9381 14732Author: Alex Deucher <alexdeucher@gmail.com> 14733Date: Wed Aug 26 01:42:10 2009 -0400 14734 14735 r6xx/r7xx: more WIP 14736 14737commit 5a08e68cc254fb255e631b456e331c32456ef0e7 14738Author: Alex Deucher <alexdeucher@gmail.com> 14739Date: Tue Aug 25 19:24:41 2009 -0400 14740 14741 r6xx/r7xx: fix some define problems in Xv code 14742 14743commit 9aa214e125b7927d62b9fe124a851d0373c24d7e 14744Author: Alex Deucher <alexdeucher@gmail.com> 14745Date: Tue Aug 25 18:45:49 2009 -0400 14746 14747 r6xx/r7xx: fix reloc for vtx buffer 14748 14749commit bba51187055932ecd466f5f817428d6c773747b9 14750Author: Alex Deucher <alexdeucher@gmail.com> 14751Date: Tue Aug 25 18:37:15 2009 -0400 14752 14753 R6xx/r7xx: unmap vb bo when done 14754 14755commit 599adfc1f5e6d708be7ad30f4871de3046775727 14756Author: Alex Deucher <alexdeucher@gmail.com> 14757Date: Tue Aug 25 18:13:14 2009 -0400 14758 14759 r6xx/r7xx: fix flipped domains 14760 14761commit 2e83cca8d7efaf1a6836cfb9ea5893fd9d70175f 14762Author: Alex Deucher <alexdeucher@gmail.com> 14763Date: Tue Aug 25 18:05:43 2009 -0400 14764 14765 r6xx/r7xx: more cs exa wip 14766 14767commit 65852de027989c105246fa4e4eed432f29525a22 14768Author: Alex Deucher <alexdeucher@gmail.com> 14769Date: Tue Aug 25 17:29:42 2009 -0400 14770 14771 r6xx/r7xx EXA: WIP 14772 14773commit 69ec7a35e2a0a3d802ec093a6aab2d7ed2cc88be 14774Author: Alex Deucher <alexdeucher@gmail.com> 14775Date: Tue Aug 25 16:14:02 2009 -0400 14776 14777 r6xx/r7xx: first pass at kms accel support 14778 14779 Adapted from various patches from Dave and Jerome. 14780 14781commit 04692f80678ffea15ac33820381ce45865e697b8 14782Author: Dave Airlie <airlied@linux.ie> 14783Date: Wed Aug 26 05:31:56 2009 +1000 14784 14785 radeon: afaics these aren't radeon regs but r128 regs. 14786 14787 so remove them from radeon use. 14788 14789commit 9cf965bbc977f0523437c0ecf1d7363b17de2468 14790Author: Alex Deucher <alexdeucher@gmail.com> 14791Date: Tue Aug 25 12:47:34 2009 -0400 14792 14793 R6xx/r7xx: add begin/end batch macros 14794 14795commit b6368cc572c79bce9a9366242c727c13cab3f006 14796Author: Alex Deucher <alexdeucher@gmail.com> 14797Date: Tue Aug 25 12:14:33 2009 -0400 14798 14799 r6xx/r7xx: move more common state to default state setup 14800 14801commit fc74e1194c980d978667e02c60a29a761a694bde 14802Author: Alex Deucher <alexdeucher@gmail.com> 14803Date: Tue Aug 25 11:39:56 2009 -0400 14804 14805 RV740: disable small DFS transfers 14806 14807 Seems problematic on all rv740 chips, so until we 14808 can find a solution, disable them. 14809 14810commit 1c909ecd88e75c6e16af7aa068e32a7a60b0af37 14811Author: Jerome Glisse <jglisse@redhat.com> 14812Date: Tue Aug 25 11:21:41 2009 +0200 14813 14814 radeon: Fix X recycling with KMS 14815 14816 On X recycling CloseScreen drop master so grab it again 14817 in InitScreen and avoid to reallocate buffer & various 14818 others cs/bo utilities on X recycling. 14819 14820commit 39aaf9addc923852598ba6fb914459671b78e492 14821Author: Dave Airlie <airlied@linux.ie> 14822Date: Mon Aug 24 18:59:22 2009 +1000 14823 14824 r600: use exa fallback code for solid/copy codepaths. 14825 14826 Solid seem to try and handle 1bpp which really no. 14827 14828commit e4b2d57ddca8739ec0b452565713398d036be0d5 14829Author: Dave Airlie <airlied@redhat.com> 14830Date: Mon Aug 24 14:11:24 2009 +1000 14831 14832 r600: fix bad unit in texture setup 14833 14834commit bac224912c750dc1c85ff2d9b8526dad6c23b572 14835Author: Dave Airlie <airlied@redhat.com> 14836Date: Sat Aug 22 21:17:59 2009 +1000 14837 14838 radeon: don't spec any initial placement for pixmaps. 14839 14840 allow the first use to decide placement. 14841 14842commit 77f98717d825162da106c6898cdbcbdf5c984ae6 14843Author: Dave Airlie <airlied@redhat.com> 14844Date: Sat Aug 22 21:16:25 2009 +1000 14845 14846 exa/cs: add DFS from GTT optimisation 14847 14848 This uses the new libdrm busy interface, once I had this in place 14849 I added a error if this happened and it does on my desktop here, 14850 so may as well add the optimisation that used to be in my old KMS tree. 14851 14852 Signed-off-by: Dave Airlie <airlied@redhat.com> 14853 14854commit b1b77a4d6fb7404af9568644e1a8e050fdfa956e 14855Author: Alex Deucher <alexdeucher@gmail.com> 14856Date: Wed Aug 19 13:29:04 2009 -0400 14857 14858 rs600: add support for DisplayPriority HIGH 14859 14860 Still haven't implemented the actual watermark 14861 calculation, but this forces the display requests 14862 to urgent if the user specifies DisplayPriority HIGH. 14863 14864commit 72e0d1b2cb11a67b8e4be4c74913ee44dc051c5b 14865Author: Dave Airlie <airlied@redhat.com> 14866Date: Wed Aug 19 19:01:30 2009 +1000 14867 14868 radeon/kms: add initial colortiling support (disabled by default). 14869 14870 This requires an X server from git with createpixmap2 support fixed up in it. 14871 14872 On 1.6 and previous it won't do any tiling, if you are running git server 14873 please upgrade to latest git. 14874 14875 Option "AllowColorTiling" "true" to enable and do some testing 14876 14877commit 479a6daefe46f985c415b0d000b1b1b820f3924e 14878Author: Alex Deucher <alexdeucher@gmail.com> 14879Date: Tue Aug 18 02:53:29 2009 -0400 14880 14881 AVIVO: add support for DisplayPriority HIGH 14882 14883 If the DisplayPriority option is set to HIGH, 14884 force the diplay mem requests to urgent. 14885 14886commit 8fd6f69c84103b277de70181988e843da23c85b0 14887Author: Dave Airlie <airlied@redhat.com> 14888Date: Mon Aug 17 15:50:21 2009 +1000 14889 14890 radeon: move detected monitor type debugging. 14891 14892 We should print this after DAC detection to be useful for TV 14893 14894commit d25185f33d06a8162469e18688e95ee19a5d94d8 14895Author: Dave Airlie <airlied@redhat.com> 14896Date: Mon Aug 17 15:31:16 2009 +1000 14897 14898 atom: recall crtc source script after tv dpms on 14899 14900 For some reason we lost the DACB source when we dpms tv back on. 14901 14902commit a971f1cfbbda1e3d16fb41f4a07f3b43fd59bc0d 14903Author: Dave Airlie <airlied@redhat.com> 14904Date: Mon Aug 17 15:25:17 2009 +1000 14905 14906 atombios: store dac detect result into save register. 14907 14908 We end up writing back an empty save register over the dac 14909 detection results so the atombios table can no longer 14910 figure out the S-video vs composite so sets the wrong one up. 14911 14912 this gets tv-out on my r580 with a composite connector working. 14913 14914commit efef7c01ad38e078de2fa3f9e528e4ef7d05d00a 14915Author: Dave Airlie <airlied@redhat.com> 14916Date: Mon Aug 17 13:44:20 2009 +1000 14917 14918 atom/tvout: set the correct scaler up for tv out 14919 14920 If we need to use the second scaler actually set the second scaler 14921 up. 14922 14923commit 032a8a06315d12e66240e7ef74b5b165f6d0d1a9 14924Author: Dave Airlie <airlied@redhat.com> 14925Date: Mon Aug 17 12:18:23 2009 +1000 14926 14927 kms: make tv out match the connector name 14928 14929commit 36e51c4d63836863dd7a17cbf6e5a9b7c1bbd31a 14930Author: Dave Airlie <airlied@redhat.com> 14931Date: Mon Aug 17 12:17:31 2009 +1000 14932 14933 r100/r200: dont emit wrong clamp modes. 14934 14935 for rect textures you don't want to use wrap clamping which 14936 is the default. 14937 14938commit 595f7c3a60ca982ca2836a0174e1f36f9d1a4fec 14939Author: Dave Airlie <airlied@itt42.(none)> 14940Date: Mon Aug 17 09:38:09 2009 +1000 14941 14942 radeon/tv: fix typo in crtc register modifications 14943 14944commit 92baeafde8f0c1eed2b6769e96f8093e3d6d7478 14945Author: Alex Deucher <alexdeucher@gmail.com> 14946Date: Tue Aug 11 15:29:17 2009 -0400 14947 14948 RV280: add agp quirk 14949 14950 from Hartmut Niemann 14951 14952commit 04288fdb8b8c62fc2eb32fd467fdf05151ad121a 14953Author: Jerome Glisse <jglisse@redhat.com> 14954Date: Tue Aug 11 19:32:52 2009 +0200 14955 14956 radeon/r600: abstract vertex buffer for r600 & r700 hardware 14957 14958 Abstract vertex buffer, preliminary work before others change 14959 to bring cs support along IB support. 14960 14961commit 72daee2a18bfb809260d1bcbe80cc14afea3b0fa 14962Author: Adam Jackson <ajax@redhat.com> 14963Date: Mon Aug 10 13:59:49 2009 -0400 14964 14965 Fix misleading filenames in DRI setup failure messages 14966 14967commit cd99d9f0d715f1f74de4fe22f2fd30046f2c7568 14968Author: Michel Dänzer <daenzer@vmware.com> 14969Date: Fri Aug 7 11:56:50 2009 +0200 14970 14971 Properly let radeon_cs_flush_indirect() re-emit 2D state if necessary. 14972 14973 Also remove a superfluous assignment. 14974 14975commit 577ff3ce922e457cc32f80d4365cb1da81552e72 14976Author: Michel Dänzer <daenzer@vmware.com> 14977Date: Fri Aug 7 11:56:49 2009 +0200 14978 14979 Add support for EXA_MIXED_PIXMAPS in xserver master. 14980 14981commit e755fa56d03aa338d0c6345ed41e32aa3115ad4b 14982Author: Michel Dänzer <daenzer@vmware.com> 14983Date: Fri Aug 7 11:56:49 2009 +0200 14984 14985 Minor fixes for KMS EXA DownloadFromScreen hook. 14986 14987 Check we can handle the bpp, and remove superfluous flush. 14988 14989commit 4cebafae81c1738a1c330d52a2c5248869f06411 14990Author: Alex Deucher <alexdeucher@gmail.com> 14991Date: Fri Aug 7 11:56:40 2009 +0200 14992 14993 Add KMS EXA UploadToScreen hook. 14994 14995 Fixups by Michel Dänzer: Doesn't seem to be a win in terms of raw numbers, but 14996 should allow more pipelining and avoiding BO memory waste for small glyph 14997 pixmaps. 14998 14999commit 9243791322e36b9231e6a3f04024ad66325385e3 15000Author: Dave Airlie <airlied@redhat.com> 15001Date: Fri Aug 7 14:07:32 2009 +1000 15002 15003 r200: fixup scissors for DDX. 15004 15005 a) turn of R200_RE_CNTL - SCISSOR_ENABLE - this save us emitting R200_RE_TOP_LEFT, note scissor is still enabled. 15006 b) disable aux scissors. 15007 15008commit bd03977e320591ca55b1a2fbb32414c53cb3f72e 15009Author: Alex Deucher <alexdeucher@gmail.com> 15010Date: Tue Aug 4 15:21:44 2009 -0400 15011 15012 radeon: fix compile with !kms 15013 15014commit 4fc1e67e606daf9227e74518c54315b60cd38301 15015Author: Alex Deucher <alexdeucher@gmail.com> 15016Date: Tue Aug 4 15:17:39 2009 -0400 15017 15018 radeon: re-add asus connector quirks 15019 15020 fixes bug 19943 15021 15022commit 22074cf0e58fddba743924532625e6fca49b6bdc 15023Author: Jerome Glisse <jglisse@redhat.com> 15024Date: Tue Aug 4 21:09:25 2009 +0200 15025 15026 radeon/kms: add simple DownloadFromScreen implementation 15027 15028 What we want to do is add userspace object support to radeon 15029 kernel modesetting. Also this DFS is dumb and might endup doing 15030 blit from GTT to GTT. 15031 15032commit fce31b61a88522733863a9b4e9f1c935c439cb4e 15033Author: Alex Deucher <alexdeucher@gmail.com> 15034Date: Tue Aug 4 11:16:05 2009 -0400 15035 15036 RS880: enable accel 15037 15038commit 6b1b4b9214d58175727f7992fc42bc7244c42bf2 15039Author: Alex Deucher <alexdeucher@gmail.com> 15040Date: Mon Aug 3 15:55:04 2009 -0400 15041 15042 radeon: add some new r7xx pci ids 15043 15044commit f564460e94c9d0f1cf3ff4b8535481b2b8b4e9c1 15045Author: Alex Deucher <alexdeucher@gmail.com> 15046Date: Sat Aug 1 16:53:47 2009 -0400 15047 15048 radeon: use XAA in some cases 15049 15050 Use XAA in low memory situations or when the DRI is disabled. 15051 Using shadowfb might also be a viable option, maybe even a better option... 15052 fixes bug 21683 15053 15054commit c71b2f050e8996787eae463eddbfdb5864ffa65a 15055Author: Bryce Harrington <bryce@canonical.com> 15056Date: Sat Aug 1 16:50:43 2009 -0400 15057 15058 radeon: AGPMode quirk needed for SiS 15059 15060 fixes bug 23065 15061 15062commit e3659ed06fc5bb8817f1dbd7c2d6bc94c67b30f7 15063Author: Bryce Harrington <bryce@canonical.com> 15064Date: Sat Aug 1 16:48:48 2009 -0400 15065 15066 radeon: AGPMode quirk needed for IBM Thinkpad T40 with Mobility M7 LW 15067 15068 fixes bug 23064 15069 15070commit 2391531ed6b7c11ddd5ab91b2369821cc5f8b8a7 15071Author: Bryce Harrington <bryce@canonical.com> 15072Date: Sat Aug 1 16:41:41 2009 -0400 15073 15074 radeon: AGPMode quirk needed for HP Omnibook 6200 15075 15076 fixed bug 23063 15077 15078commit 474eda02257152ced52364f38cbad24c20aebbc0 15079Author: Alex Deucher <alexdeucher@gmail.com> 15080Date: Sat Aug 1 16:06:52 2009 -0400 15081 15082 radeon: reload bicubic Xv texture on VT switch 15083 15084 fixed bicubic filtering after VT switch or suspend/resume 15085 15086commit 447a2ce1b88aa2d6d5713e93174c4002617720f7 15087Author: Dave Airlie <airlied@redhat.com> 15088Date: Thu Jul 30 09:54:01 2009 +1000 15089 15090 radeon: fix r600 shadow fb mode with cursor allocation in wrong place 15091 15092 From RHEL QE testing, we could end up with the cursor at 0 since 15093 we think EXA is in use when really it isn't. the info->useEXA = FALSe 15094 might be unnecessary but better to be explicit 15095 15096 Signed-off-by: Dave Airlie <airlied@redhat.com> 15097 15098commit 2b9a5be3bb22fad1d52ec9eea7733c5e956250b7 15099Author: Alex Deucher <alexdeucher@gmail.com> 15100Date: Wed Jul 29 03:34:11 2009 -0400 15101 15102 radeon: fix typo in object header to connector type conversion 15103 15104 Should fix bug 19943 15105 15106commit 63c873cbd4d1d21d9f688028c0900c79fadc42c1 15107Author: Dave Airlie <airlied@redhat.com> 15108Date: Tue Jul 28 15:22:40 2009 +1000 15109 15110 ati: change to using ABI version check 15111 15112commit 21a621c297ac71c65c239ea960c38706e718b91c 15113Author: Dave Airlie <airlied@redhat.com> 15114Date: Tue Jul 28 13:32:28 2009 +1000 15115 15116 ati: update for resources/RAC API removal 15117 15118commit 5ad2519c2a7c0df389b2cd7cf7151c7e4b7252a8 15119Author: Michel Dänzer <daenzer@vmware.com> 15120Date: Mon Jul 27 14:22:02 2009 +0200 15121 15122 KMS: Explicitly specify VRAM and GTT domains when allocating BOs for pixmaps. 15123 15124 Without this, apparently they are initially allocated in system RAM (or 15125 possibly GTT), so any GPU rendering to them first triggers a copy of the whole 15126 uninitialized data to VRAM... 15127 15128 We may want to explore more sophisticated schemes in the future, but for now 15129 this seems to improve KMS 2D performance quite a bit. 15130 15131commit 57f2c83a22f27567506c555af431f89e6031204c 15132Author: Alex Deucher <alexdeucher@gmail.com> 15133Date: Sat Jul 25 14:19:38 2009 -0400 15134 15135 radeon: Set PCI/PCIE bus type properly per asic 15136 15137 - r1xx-rv350 chips have the old pci gart 15138 - rv380+ chips have newer pcie gart 15139 15140 Select the right kind regardless of whether the user selects 15141 PCI or PCIE. 15142 15143commit 2afc46fa74ce7730f766a3370d323c6b59694186 15144Author: Jerome Glisse <jglisse@redhat.com> 15145Date: Thu Jul 23 20:13:15 2009 +0200 15146 15147 radeon: fix KMS shadowfb for r6xx and newer hw. 15148 15149 We need to map front buffer and also to avoid EXA initialization 15150 when falling back to shadowfb on newer r6xx/r7xx when KMS is enabled. 15151 15152commit 328e4c816348b6f867df30b39856dbe78c8e0dcc 15153Author: Alex Deucher <alexdeucher@gmail.com> 15154Date: Thu Jul 23 13:28:36 2009 -0400 15155 15156 RS780: fill in MC access functions 15157 15158 Also, r6xx/r7xx don't have indirect MC space, so 15159 don't try to access it. 15160 15161commit f96d5b255425fbd02be2cad26edb590d474a5640 15162Author: Alex Deucher <alexdeucher@gmail.com> 15163Date: Thu Jul 23 13:25:04 2009 -0400 15164 15165 R6xx/R7xx: clip rendering to destination surface dimensions 15166 15167commit e372f845b0defaf2d2c9ef3cbbf7498e09d9372e 15168Author: Michel Dänzer <daenzer@vmware.com> 15169Date: Sun Jul 19 16:05:29 2009 +0200 15170 15171 Guard reference to CRTC active field. 15172 15173 Fixes build against older xserver. 15174 15175commit e38305aebdc95f80f5b4b3e5ba541ea67dc05f01 15176Author: Michel Dänzer <daenzer@vmware.com> 15177Date: Sat Jul 18 22:23:45 2009 +0200 15178 15179 Also drop DRM master in KMS CloseScreen. 15180 15181 The LeaveVT hook isn't always called when the server dies, e.g. when quitting 15182 from the GDM greeter. This may cause existing servers to fall over if the dying 15183 server process still exists when they try to re-acquire master and set a mode. 15184 15185 Also use drmSet/DropMaster() rather than ioctl() directly. 15186 15187commit a43c660a00147bfae5ca601f4720b2680b75211f 15188Author: Michel Dänzer <daenzer@vmware.com> 15189Date: Sat Jul 18 22:21:59 2009 +0200 15190 15191 Set CRTC active flag to TRUE after a successful KMS mode set. 15192 15193 Otherwise some things like changing the colour map won't work properly. 15194 15195commit f32069f42b3e4643f15148d0b96164def00dcc74 15196Author: Michel Dänzer <daenzer@vmware.com> 15197Date: Sat Jul 18 22:20:08 2009 +0200 15198 15199 Set the STALL bit in the CRTC_GUI_TRIG_VLINE registers. 15200 15201 As recommended by the register reference when using the WAIT_CRTC_VLINE bit in 15202 the WAIT_UNTIL register, as we are. 15203 15204commit 27bd9fc9b2c0eeff488c1f26f1355a7fa3655520 15205Author: Michel Dänzer <daenzer@vmware.com> 15206Date: Sat Jul 18 22:14:27 2009 +0200 15207 15208 Enable vsync for DRI2 region copies. 15209 15210commit 4b1b28865ffb17f58a70d955212fd51cb1fb55ab 15211Author: Alex Deucher <alexdeucher@gmail.com> 15212Date: Thu Jul 16 17:09:33 2009 -0400 15213 15214 R6xx/r7xx: reduce the amount of default state setup 15215 15216 Most of the default state setup was extraneous or duplicated 15217 in the accel ops. There's no need to reset all the consts 15218 everytime you reset the default state for example. This leads 15219 to a 10x reduction in default state. 15220 15221commit 7e6557b81026d8a4d6c837839e68c28b151f8a97 15222Author: Peter Hutterer <peter.hutterer@who-t.net> 15223Date: Thu Jul 16 11:33:57 2009 +1000 15224 15225 Update to xextproto 7.1 support. 15226 15227 DPMS header was split into dpms.h (client) and dpmsconst.h (server). Drivers 15228 need to include dpmsconst.h if xextproto 7.1 is available. 15229 15230 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15231 15232commit 95d431e8b260dd43cf6d93b90eecab1dc0ec7c26 15233Author: Dave Airlie <airlied@redhat.com> 15234Date: Wed Jul 15 12:07:01 2009 +1000 15235 15236 radeon: emit colorpitch relocs. 15237 15238 This causes relocs to be emitted for the colorpitch. 15239 15240 Shouldn't have no effect on current kernels, but will keep DDX 15241 going on kms kernels when tiling lands 15242 15243 Also contains a missing reloc space for r200 15244 15245commit 8312763977b22e108b7d003cbf3256e0165e4439 15246Author: Alex Deucher <alexdeucher@gmail.com> 15247Date: Tue Jul 14 18:13:27 2009 -0400 15248 15249 Add an r420 AGP quirk 15250 15251 fixes bug 22726 15252 15253commit 5473eeeb45468fd200fec49a8506281628a5e4b5 15254Author: Alex Deucher <alexdeucher@gmail.com> 15255Date: Tue Jul 14 14:05:13 2009 -0400 15256 15257 R3/4/5xx: only upload the bicubic texture once 15258 15259 Upload the bicubic texture once during textured video init 15260 rather than once per frame. Suggested by Michel Daenzer on 15261 IRC. 15262 15263commit 0485f27bc3d75cb6ab320e8164dbe6ea2713c78e 15264Author: Alex Deucher <alexdeucher@gmail.com> 15265Date: Tue Jul 14 13:11:12 2009 -0400 15266 15267 R3/4/5xx: fix bicubic Xv filtering with KMS changes 15268 15269 fixes bug 22730 15270 15271commit ca4a9efef987108266ae59afbb5142d4d2000c8e 15272Author: Alex Deucher <alexdeucher@gmail.com> 15273Date: Mon Jul 13 11:28:41 2009 -0400 15274 15275 Add configure option to disable kms support 15276 15277 Useful to building with mixed kms/non-kms setups 15278 15279commit 035e8d1d5593c12828bb079de4e663cf1b1f1674 15280Author: Edward O'Callaghan <eocallaghan@auroraux.org> 15281Date: Sun Jul 12 13:45:29 2009 +0200 15282 15283 Fix a warning 15284 15285 Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com> 15286 15287commit 76af48c43f829e7aebacc9f2a623823fa26ee22b 15288Author: Gaetan Nadon <memsize@videotron.ca> 15289Date: Wed Jul 8 09:58:56 2009 -0400 15290 15291 xf86-video-ati: use XORG_CHANGELOG macro to create ChangeLog. #22611 15292 15293 Build break: Makefile.am: command not found: git-log 15294 Adding the macro in configure.ac and use it in Makefile.am 15295 Refer to: https://bugs.freedesktop.org/show_bug.cgi?id=22611 15296 Tested: running autogen.sh, make and 'make dist' 15297 15298 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 15299 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15300 15301commit 43db263d301082e84e9bc304816bcbb206fe280e 15302Author: Alex Deucher <alexdeucher@gmail.com> 15303Date: Thu Jul 9 12:48:06 2009 -0400 15304 15305 RV280: another AGP quirk 15306 15307 fdo bug 12544 15308 15309commit 69b5e5496f10a9f566d2e563862c96cb41952eb6 15310Author: Alex Deucher <alexdeucher@gmail.com> 15311Date: Thu Jul 9 12:42:45 2009 -0400 15312 15313 RV280: Add an AGP quirk 15314 15315 lp bug 370205 via fdo bug 12544 15316 15317commit a6d6c900e60a3685ee9e93c34eb7f6d237c45fd8 15318Author: Alex Deucher <alexdeucher@gmail.com> 15319Date: Thu Jul 9 11:30:57 2009 -0400 15320 15321 RS740: fixup display latency setting as per rs690 15322 15323 Both chips program the same in this regard 15324 15325commit ea407570d39ace3162d372eda56bf791bfd80c24 15326Author: Dave Airlie <airlied@redhat.com> 15327Date: Thu Jul 9 10:46:23 2009 +1000 15328 15329 radeon: add shave support. 15330 15331 clean up output a lot 15332 15333commit 8c03c1fdb5ea35570064946557050c87ca30582a 15334Author: Alex Deucher <alexdeucher@gmail.com> 15335Date: Wed Jul 8 18:34:57 2009 -0400 15336 15337 R6xx/R7xx: fix hangs on x2 cards with PM options 15338 15339 Changing the PCIE lanes on x2 cards results in a hang, 15340 so for now, disable it. 15341 15342 Fixes fdo bug 22669 15343 15344commit 9645838c57f6b40837fdce23ce7f9faefb3d9966 15345Author: Michel Dänzer <daenzer@vmware.com> 15346Date: Wed Jul 8 19:18:02 2009 +0200 15347 15348 Further non-KMS fixes / cleanups. 15349 15350 Only compile tested. 15351 15352commit 2d86ec4ddb1fb83be95ed9cdcbabf4d568a6d7b6 15353Author: Alex Deucher <alexdeucher@gmail.com> 15354Date: Wed Jul 8 12:33:34 2009 -0400 15355 15356 Fix the build for the !XF86DRM_MODE case 15357 15358 compile tested only. 15359 15360commit a68afc2a80b8581cc3d67e9b5a95a10b0d4ead0e 15361Author: Michel Dänzer <daenzer@vmware.com> 15362Date: Wed Jul 8 09:15:28 2009 +0200 15363 15364 Fix RADEON_TRACE_FALL build. 15365 15366commit 9ca0aeaad01b7c682396b30af48668f287a59d6e 15367Author: Michel Dänzer <daenzer@vmware.com> 15368Date: Wed Jul 8 09:13:36 2009 +0200 15369 15370 Pass on CS relocation offset values. 15371 15372 Not actually used ATM, but just in case... 15373 15374commit f718dd794477fbdd5eeb4183f43ff25517c63034 15375Author: Michel Dänzer <daenzer@vmware.com> 15376Date: Wed Jul 8 09:11:18 2009 +0200 15377 15378 Don't set EXA UploadTo/DownloadFromScreen hooks when they'd always return FALSE. 15379 15380commit 42e76d0b24b76fbcc6b08a4ea5003dbec368645c 15381Author: Michel Dänzer <daenzer@vmware.com> 15382Date: Wed Jul 8 08:51:10 2009 +0200 15383 15384 Don't always flush CS when switching from 2D to 3D. 15385 15386 This works fine here and doubles x11perf -aa10text scores, but may cause 15387 problems on other setups according to Dave. Let's see... 15388 15389commit 94fe049c35187ddc3ec0472e9309c3a22879f415 15390Author: Michel Dänzer <daenzer@vmware.com> 15391Date: Wed Jul 8 08:36:22 2009 +0200 15392 15393 KMS CS fixes, take two. 15394 15395 Avoid flushing CS in the middle of (setting up state for) an operation, 15396 properly finish/restart the operation. 15397 15398commit 0519f15af2fe5e7e5a6c94e7203a96fe363c643e 15399Author: Michel Dänzer <daenzer@vmware.com> 15400Date: Tue Jul 7 10:04:09 2009 +0200 15401 15402 Fix DRI2 with current xserver Git. 15403 15404commit 7ad11fec2579807371360cb680e593f723ebaa1a 15405Author: Dave Airlie <airlied@redhat.com> 15406Date: Mon Jul 6 18:43:08 2009 +1000 15407 15408 radeon: add make dist headers 15409 15410commit e68fbf743ec88e0fa3a9f1ffe83daf251e250b21 15411Author: Dave Airlie <airlied@redhat.com> 15412Date: Mon Jul 6 18:32:32 2009 +1000 15413 15414 git: fix git-log to git log 15415 15416commit 5eeb1fd19c6797cc76ce308570dffe34cb82b24d 15417Author: Dave Airlie <airlied@linux.ie> 15418Date: Mon Jul 6 17:45:50 2009 +1000 15419 15420 radeon: oops fix compile without libdrm 15421 15422commit 98824f37fc74dc1fec2677515d0dc3cfc263f00a 15423Author: Dave Airlie <airlied@redhat.com> 15424Date: Mon Jul 6 15:29:53 2009 +1000 15425 15426 radeon: move state changes into flush function 15427 15428commit 503b51343267248b4c6994b88238dd5384528f03 15429Author: Dave Airlie <airlied@redhat.com> 15430Date: Mon Jul 6 15:20:13 2009 +1000 15431 15432 radeon: taken from kms-support + CS fixes commit. 15433 15434 re-emit 2D state and reset 3D state emit on flush 15435 15436commit d78d50591851d14543f0935d051a59ef29751bbc 15437Author: Dave Airlie <airlied@redhat.com> 15438Date: Mon Jul 6 14:45:07 2009 +1000 15439 15440 radeon: port to new space checking in libdrm interface 15441 15442 This uses a new libdrm interface which shares code with mesa. 15443 15444 It also fixes the bo to flush when full instead of never flushing. 15445 15446 It survives gtkperf -a here which the driver didn't before now 15447 15448commit e1200cb89218930d01330ba0114e013438655cce 15449Author: Dave Airlie <airlied@redhat.com> 15450Date: Mon Jul 6 15:10:29 2009 +1000 15451 15452 Revert "KMS CS fixes." 15453 15454 This reverts commit 11cf8a28eb46f48d101888552266921bee8b63fc. 15455 15456 I'll bring this back in a few minutes, I've got a big change introducing 15457 a new cs space accounting scheme which this messes up and I think 15458 probably conflicts with what I've done. 15459 15460 I promise i'll go over this and pick the bits that are still useful :) 15461 15462commit 11cf8a28eb46f48d101888552266921bee8b63fc 15463Author: Michel Dänzer <daenzer@vmware.com> 15464Date: Sun Jul 5 15:14:59 2009 +0200 15465 15466 KMS CS fixes. 15467 15468 * Flush CS before running out of space. 15469 * Don't flush CS every time when switching between 2D and 3D engine, but make 15470 sure to re-initialize 2D/3D state after flushing. 15471 15472 Doesn't seem quite perfect yet (some rendering seems to get dropped on the floor 15473 sometimes), but this should fix a lot of CS failures people have been seeing. 15474 15475commit 8d2f712eaf1e569fd92bbe2db5aceb43b7b367d1 15476Author: Dave Airlie <airlied@linux.ie> 15477Date: Fri Jul 3 10:40:59 2009 +1000 15478 15479 radeon: don't set up overlay under KMS until we figure out how. 15480 15481commit 2a622168606d98c404708b879e6417c8f7d7305c 15482Author: Michel Dänzer <daenzer@vmware.com> 15483Date: Thu Jul 2 15:03:53 2009 +0200 15484 15485 Test BO pointer before passing it to radeon_bo_unref(). 15486 15487commit 36a4ea5a9c7062f4c8aa7480bb7488ce2062390f 15488Author: Michel Dänzer <daenzer@vmware.com> 15489Date: Thu Jul 2 14:52:44 2009 +0200 15490 15491 Warning fixes. 15492 15493commit 65465d56c1bf943113518a5b2f57b3170213bab9 15494Author: Dave Airlie <airlied@linux.ie> 15495Date: Thu Jul 2 18:38:38 2009 +1000 15496 15497 radeon: fix up configure check better 15498 15499commit ad17108b72c9adf773a122b4c054c7b92e6a8b9d 15500Author: Dave Airlie <airlied@linux.ie> 15501Date: Thu Jul 2 17:25:57 2009 +1000 15502 15503 radeon: fix it properly. 15504 15505 I blame the flu 15506 15507commit e1b9cb9ea665eaa84e64a386299efa0819d55df4 15508Author: Dave Airlie <airlied@redhat.com> 15509Date: Thu Jul 2 17:18:47 2009 +1000 15510 15511 radeon: fix compile without kms 15512 15513commit de427fac6912a625f7e4d8b982a1f7a1ba80dfbe 15514Author: Dave Airlie <airlied@redhat.com> 15515Date: Thu Jul 2 15:41:43 2009 +1000 15516 15517 radeon/dri2: fix dri2 - needs newest libdrm_radeon 15518 15519commit 2b671adba54ebff1756dd990123f4bade6e35d61 15520Author: Dave Airlie <airlied@redhat.com> 15521Date: Thu Jul 2 15:25:19 2009 +1000 15522 15523 radeon/dri2: call proper gem bo function 15524 15525commit 851b2b8cab883e4422745cca1827aa8ac9f193a3 15526Author: Dave Airlie <airlied@redhat.com> 15527Date: Thu Jul 2 15:17:13 2009 +1000 15528 15529 radeon/kms: add textured video BO size checks 15530 15531commit 9d092060236e8587697fdd57a7a01f3c0dddd680 15532Author: Dave Airlie <airlied@redhat.com> 15533Date: Thu Jul 2 15:17:03 2009 +1000 15534 15535 kms: drop DGA 15536 15537commit 207174ea964c561076bdd5297a2fb6f136b4b645 15538Author: Dave Airlie <airlied@redhat.com> 15539Date: Thu Jul 2 15:16:47 2009 +1000 15540 15541 radeon: fix CS imbalance 15542 15543commit 4d24a9d2c0be0c025266cc8cdf7e8d7bb065cdb1 15544Author: Dave Airlie <airlied@redhat.com> 15545Date: Thu Jul 2 14:55:45 2009 +1000 15546 15547 radeon: enable KMS build 15548 15549commit 9bcd3ae7966eb7a17b922bcd33e0d6f016245614 15550Author: Dave Airlie <airlied@redhat.com> 15551Date: Thu Jul 2 14:55:18 2009 +1000 15552 15553 radeon/kms: fixup 3d state emission for KMS 15554 15555 This fixes font rendering under KMS (should have looked at dmesg) 15556 15557commit ac1607bb23f0c7584648407a3ac3fb27142e186e 15558Author: Dave Airlie <airlied@redhat.com> 15559Date: Thu Jul 2 14:55:05 2009 +1000 15560 15561 kms: no need to reference fifo slots 15562 15563commit 1782ce28953184776c90eb1255208a3e0ad245f0 15564Author: Dave Airlie <airlied@redhat.com> 15565Date: Tue Jun 30 16:55:26 2009 +1000 15566 15567 radeon: add KMS support (still disabled) 15568 15569 This adds DRI2 + KMS + driver pixmaps support to the driver. 15570 15571 I've decided to just do a completely separate KMS driver file 15572 instead of hacking the crap out of radeon_driver.c. So now 15573 I do the KMS check in radeon_probe.c time and set the DDX 15574 pointed up to a completely different set at this stage. 15575 15576 This avoids a lot of if (kms) type crap in the code at 15577 the expense of making sure we make changes to both files 15578 if necessary. 15579 15580 This code is still disabled in configure.ac as I broke EXA composite 15581 rendering somehow in KMS mode 15582 15583commit ac4bd24a66c1bdda0293f770a3f891e2b88cc8ee 15584Author: Alex Deucher <alexdeucher@gmail.com> 15585Date: Tue Jun 30 16:00:06 2009 -0400 15586 15587 EXA: fix the build on servers without EXA_HANDLES_PIXMAPS 15588 15589commit 49a0ca19ee4109cdd39a43fbf4bedf3f0557d6ec 15590Author: Michel Dänzer <daenzer@vmware.com> 15591Date: Tue Jun 30 17:22:50 2009 +0200 15592 15593 Warning fixes. 15594 15595commit e932836691aeaec37794fdaed2dabb22710fd171 15596Author: Dave Airlie <airlied@redhat.com> 15597Date: Tue Jun 30 16:24:37 2009 +1000 15598 15599 radeon: initial preparation for kms patch. 15600 15601 This patch contains most of the changes to the EXA and texture video 15602 accel code. 15603 15604 It adds a few bits of pixmap support but doesn't actually do anything 15605 useful KMS yet. 15606 15607 Testing this should not have any regressions over what we have already, 15608 biggest worries are r6xx, I've fixed a textured video one, but no idea 15609 what other might lurk 15610 15611 It won't build against libdrm radeon yet either 15612 15613commit bb04b450ed00ca4b1aa44c33085567d47b33b547 15614Author: Fredrik Höglund <fredrik@kde.org> 15615Date: Mon Jun 29 22:38:26 2009 +0200 15616 15617 Check if the composite op is supported in R200CheckComposite. 15618 15619commit f39cafc5c4a93a16ac93756473ebb1f33c7881db 15620Author: Alex Deucher <alexdeucher@gmail.com> 15621Date: Sun Jun 28 20:42:49 2009 -0400 15622 15623 radeon: call hdmi-b connectors dvi in randr 15624 15625 Technically HDMI-B is a glorfied DL DVI, but this is confusing 15626 to users, so call it DVI when reporting randr outputs. This 15627 allows us to remove two quirks where the bios was actually 15628 correct. 15629 15630commit 407e676b04117c7ca3dcc0cc072f3552b03c64e5 15631Author: Roland Scheidegger <sroland@tungstengraphics.com> 15632Date: Sun Jun 28 16:54:32 2009 -0400 15633 15634 R6xx/R7xx: fix pixel centers 15635 15636 Make sure we are using GL mode (centers at 0.5) rather 15637 than D3D mode (centers at 0.0). This also fixes fdo 15638 bug 21963. 15639 15640commit 68001981f22173ff949720055dba89291f284474 15641Author: Alex Deucher <alexdeucher@gmail.com> 15642Date: Sun Jun 28 16:20:10 2009 -0400 15643 15644 Add some missing M96 pci ids 15645 15646 fixes fdo bug 22404 15647 15648commit 06d5d8d7fd92b58b2467c782e57a80a3520a2c90 15649Author: Adam Jackson <ajax@redhat.com> 15650Date: Thu Jun 25 13:51:36 2009 -0400 15651 15652 Add a (harmlessly) missing break statement. 15653 15654commit 7d1d968fdc8de1e0ab0be916a1f3a4a398af47aa 15655Author: Adam Jackson <ajax@redhat.com> 15656Date: Thu Jun 25 13:49:18 2009 -0400 15657 15658 Cosmetic cleanup to digital output detection 15659 15660 No functional change, but this logic would need extending anyway for 15661 DisplayID, so unify it now. 15662 15663commit 96581746eea2f4e29a04f9266ec9649f25480424 15664Author: Adam Jackson <ajax@redhat.com> 15665Date: Wed Jun 24 17:05:47 2009 -0400 15666 15667 atom: Fix phantom VGA connector when HDMI-B present. 15668 15669 HDMI-B is really a glorified DL-DVI connector, the analog pins work fine. 15670 15671commit 496adc4dc7cf53fc6acfecb1158f1491d5f36168 15672Author: Michel Dänzer <daenzer@vmware.com> 15673Date: Fri Jun 19 14:37:01 2009 +0200 15674 15675 Revert "Build fixes for -Werror that aren't ready for prime time yet." 15676 15677 This reverts commit ab20ea1530ec8a3a9fd4089775c4e157e50fe1bf. 15678 15679 Whoops, pushed wrong branch... 15680 15681commit ab20ea1530ec8a3a9fd4089775c4e157e50fe1bf 15682Author: Michel Dänzer <daenzer@vmware.com> 15683Date: Sun Jun 14 16:50:07 2009 +0200 15684 15685 Build fixes for -Werror that aren't ready for prime time yet. 15686 15687commit 77e3537d312175a25f0e21cc07c3a96f78c3b35a 15688Author: Markus Gapp <markus.gapp@gmx.net> 15689Date: Sat Jun 13 11:51:29 2009 -0400 15690 15691 Add quirk for asus hd3450 board 15692 15693 Fixes bug 22266 15694 15695commit e8e0bb1ba29761c94d473e42ca6ada161cfff557 15696Author: Alex Deucher <alexdeucher@gmail.com> 15697Date: Fri Jun 12 11:07:43 2009 -0400 15698 15699 Fix count in 3d state setup after 43374c7420e378918bec062f4cbd581f16adb6f0 15700 15701commit 808c90a24c48da7fa97e15e2f12be5bb8fd8cc96 15702Author: Alex Deucher <alexdeucher@gmail.com> 15703Date: Thu Jun 11 16:09:40 2009 -0400 15704 15705 Only enable frac fb divs on rs600/rs690/rs740 for now 15706 15707 fixes bug 21413 15708 15709commit d095bb3157faaee10b0130ea9b1032e0b533aaee 15710Author: Alex Deucher <alexdeucher@gmail.com> 15711Date: Thu Jun 11 12:34:15 2009 -0400 15712 15713 Re-enable frac feedback dividers on AVIVO hardware 15714 15715 Preferring a freq <= target frequency seems to fix the 15716 previous issues with frac fb divs. Appears to work fine 15717 on all hardware I've tested on. See fdo bug 22229 for 15718 details. Should fix bugs: 22229, 21553, 21413, and 15719 possibly other ones related to unstable images at certain 15720 resolutions. 15721 15722commit 55fbdbae83d1563b472f49d0436c9298e390be66 15723Author: Alex Deucher <alexdeucher@gmail.com> 15724Date: Thu Jun 11 12:31:53 2009 -0400 15725 15726 Add PLL flag to prefer frequencies <= the target freq 15727 15728 This appears to be needed when using fractional feedback 15729 dividers. Based on a patch from Tom Hirst. See fdo 15730 bug 22229 for more details. 15731 15732commit 43374c7420e378918bec062f4cbd581f16adb6f0 15733Author: Alex Deucher <alexdeucher@gmail.com> 15734Date: Tue Jun 9 14:39:54 2009 -0400 15735 15736 r1xx/r2xx: clip 3D rendering to destination buffer 15737 15738commit 74b0a38c75c1d6a6344015d32d84b43d9e80df1d 15739Author: Alex Deucher <alexdeucher@gmail.com> 15740Date: Tue Jun 9 14:26:21 2009 -0400 15741 15742 Consolidate indentical RADEONDoneSolid/RADEONDoneCopy functions 15743 15744commit f6dff8836729cc9d29264080c9cfa0e1df6182cf 15745Author: Alex Deucher <alexdeucher@gmail.com> 15746Date: Tue Jun 9 13:31:15 2009 -0400 15747 15748 r3xx/r4xx/r5xx: clip 3D rendering to destination buffer 15749 15750commit 48dbf09e1309c815ac8b733145a0310083a29933 15751Author: Alex Deucher <alexdeucher@gmail.com> 15752Date: Tue Jun 9 13:12:40 2009 -0400 15753 15754 r3xx/r4xx/r5xx: refactor exa prepare composite 15755 15756 - consolidate code duplicated between r3xx/r4xx and r5xx 15757 - make pixel shader setup easier to read and less 15758 redundant 15759 15760commit ec1a3ccdf61c2226617b571cc4a209a542e9d7e3 15761Author: Alex Deucher <alexdeucher@gmail.com> 15762Date: Mon Jun 8 15:31:57 2009 -0400 15763 15764 R6xx/r7xx: don't clamp tex coords in composite VS 15765 15766 Fixes broken repeat modes. see fdo bugs 21818, 22139 15767 15768commit 32c218c5633fb7ffeb19e717bf645c1b25dd4ef5 15769Author: Alan Coopersmith <alan.coopersmith@sun.com> 15770Date: Thu Jun 4 07:23:03 2009 -0700 15771 15772 Add link to patch submission instructions to README 15773 15774 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 15775 15776commit 39f736d6cb65fc7c3fd769478023fc87c1af5979 15777Author: Alex Deucher <alexdeucher@gmail.com> 15778Date: Thu Jun 4 09:57:22 2009 -0400 15779 15780 atom: connector quirk 15781 15782 Some r6xx cards have a VGA and an HDMI port with 15783 a shared ddc line listed as an HDMI port with both 15784 analog and digital encoders. 15785 15786 Fixes bug 19943 15787 15788commit 7599dc40855c33a5fbd8e9bbc2b4cd62752fb7df 15789Author: Alex Deucher <alexdeucher@gmail.com> 15790Date: Fri May 29 16:29:50 2009 -0400 15791 15792 Add RV740 (HD4770) support 15793 15794 Verified by ernstp on IRC 15795 15796commit ee0424078a3ea26f58c2998c9e58e065fb63f1a0 15797Author: Alex Deucher <alexdeucher@gmail.com> 15798Date: Wed May 27 18:37:58 2009 -0400 15799 15800 radeon: remove dead code 15801 15802commit b34df233115c0d82d7bcf82e041afbc55981ce82 15803Author: Alex Deucher <alexdeucher@gmail.com> 15804Date: Tue May 26 17:47:01 2009 -0400 15805 15806 R3xx/R4xx: increase EXA render target limits 15807 15808 now that we are using 1/12 subpixel mode 15809 15810commit 1457bc089d1395d18cf582a4d52444005f013753 15811Author: Matthieu Herrb <matthieu.herrb@laas.fr> 15812Date: Tue May 26 17:21:07 2009 -0400 15813 15814 radeon: fix segfault if Option "NoAccel" is set 15815 15816commit 4079f97958cf1d5fd76a83aa61b93d6289bb9b1e 15817Author: Michel Dänzer <daenzer@vmware.com> 15818Date: Sat May 23 13:51:00 2009 +0200 15819 15820 Bump version to mark Git snapshot. 15821 15822 The fourth version component isn't visible in the log file. 15823 15824commit 88a9e98341d96e5e7f48b69aed597d1bada6313a 15825Author: Alex Deucher <alexdeucher@gmail.com> 15826Date: Tue May 19 17:11:34 2009 -0400 15827 15828 Pre-atom: Improve engine clock setup code 15829 15830 Also remove SetMemoryClk() for pre-atom cards for now 15831 as it requires quite a bit more asic specific work. 15832 To set the mclk we'll need to use the mem reset/dll tables 15833 in the bios. 15834 15835commit 1429133ffdbfe046249fcaaa4cbee432a1012e55 15836Author: Michel Dänzer <daenzer@vmware.com> 15837Date: Mon May 18 17:59:00 2009 +0200 15838 15839 EXA: Set new EXA_SUPPORTS_OFFSCREEN_OVERLAPS flag when we have a 2D engine. 15840 15841commit e9cca1e7dc521d3756bb2d97f1d703b9ad65cb47 15842Author: Alex Deucher <alexdeucher@gmail.com> 15843Date: Mon May 18 10:48:25 2009 -0400 15844 15845 r4xx: Add some missing pci ids 15846 15847 Noticed by Romain in bug 21798 15848 15849commit 76fffd51027820b3a6ea10de2be67d5fd4f3e843 15850Author: Alex Deucher <alexdeucher@gmail.com> 15851Date: Fri May 15 16:35:13 2009 -0400 15852 15853 R100: switch to normalized texture coordinates 15854 15855 for consistency with the other asics 15856 15857commit 1a1eef31e36f98b7a30e27e567c1b66e0e4cf7dc 15858Author: Alex Deucher <alexdeucher@gmail.com> 15859Date: Thu May 14 15:48:54 2009 -0400 15860 15861 radeon: try harder when checking if a card is posted 15862 15863 fixes debian bug 524280, possibly others. 15864 15865commit 22e39392297fa11003df90c175db3c449d8f9853 15866Author: Michel Dänzer <daenzer@vmware.com> 15867Date: Thu May 14 10:24:07 2009 +0200 15868 15869 Pre-R600 EXA: Fix coordinate limits off-by-one error. 15870 15871 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=21598 . 15872 15873commit fa09b058c7a17689989e600ffd465856a058579d 15874Author: Alex Deucher <alexdeucher@gmail.com> 15875Date: Wed May 13 16:10:40 2009 -0400 15876 15877 R6xx/R7xx Xv: normalize texture coordinates in the vertex shader 15878 15879commit 026b6f820d6caea17d2a082193e850713d5770a8 15880Author: Alex Deucher <alexdeucher@gmail.com> 15881Date: Wed May 13 15:48:32 2009 -0400 15882 15883 R6xx/R7xx: do EXA transforms in the vertex shader 15884 15885commit cd89241396d1931b04cfbdd8d553be16dbf9c360 15886Author: Alex Deucher <alexdeucher@gmail.com> 15887Date: Tue May 12 17:30:02 2009 -0400 15888 15889 R3xx-R5xx: do EXA transforms in the vertex shader 15890 15891commit 15ee78d37f8b64c3e6d234e7ab37a31e82327b6b 15892Author: Alex Deucher <alexdeucher@gmail.com> 15893Date: Mon May 11 02:22:13 2009 -0400 15894 15895 RV770: add missing pci id 15896 15897 Fixes bug 21209 15898 15899commit 5ff51ea83f9fdd836e830c77e3561eaef4b83183 15900Author: Alex Deucher <alexdeucher@gmail.com> 15901Date: Fri May 8 13:19:26 2009 -0400 15902 15903 R7xx: add support for reading disabled bios images 15904 15905commit 76fdae9018330aa5af0d57339baff3ca4834e1b5 15906Author: Alex Deucher <alexdeucher@gmail.com> 15907Date: Thu May 7 02:59:32 2009 -0400 15908 15909 R3xx/R4xx tex vid: no need to check if IS_R300_3D 15910 15911 function is r3xx/r4xx specific already 15912 15913commit 23a8ada06a5b1ccc975e89f129c00382ecbf29df 15914Author: Alex Deucher <alexdeucher@gmail.com> 15915Date: Thu May 7 02:34:22 2009 -0400 15916 15917 R3xx/R4xx tex vid: increase the guardband limit for rendering with a tri 15918 15919 limit goes from 2880 to 4021 now that we've switched to 1/12 subpixel mode. 15920 15921 This allows us to render with a clipped tri at higher resolutions preventing 15922 diagonal tearing. 15923 15924commit 0c3541f7013c53fb8096ada63cc7b3577524aa4c 15925Author: Alex Deucher <alexdeucher@gmail.com> 15926Date: Wed May 6 12:24:35 2009 -0400 15927 15928 AVIVO: set default cursor mode in cursor init 15929 15930 Should fix bug 21557 15931 15932commit da58e351b6398fa18b1d9c1a57a6e49b285f610f 15933Author: Alex Deucher <alexdeucher@gmail.com> 15934Date: Mon May 4 16:03:53 2009 -0400 15935 15936 AVIVO: fix cursoe corruption bands for real 15937 15938 Don't have to leave both cursors enabled, just have to use 15939 the same mode for both cursors whether or not they are enabled. 15940 15941commit 2c8e130f73c680d4a7381b2ef37982b82c6ee478 15942Author: Alex Deucher <alexdeucher@gmail.com> 15943Date: Mon May 4 13:40:17 2009 -0400 15944 15945 radeon: switch to EXA by default 15946 15947commit 8155a62df6817358a12cc2d613c1470ca55996ef 15948Author: Alex Deucher <alexdeucher@gmail.com> 15949Date: Mon May 4 12:57:36 2009 -0400 15950 15951 AVIVO: move cusor offscreen when disabling 15952 15953 Adapted from Yang's patch. Setting size to zero is 15954 actually a 1x1 cursor. 15955 15956commit a1c64ea5224009779ccad66b0f84d861eae966ac 15957Author: Alex Deucher <alexdeucher@gmail.com> 15958Date: Sun May 3 19:42:44 2009 -0400 15959 15960 r3xx-r5xx: switch to 1/12 subpixel precision 15961 15962 - based on similar patch in Jerome's cs ddx tree 15963 - also fix clipping offsets 15964 - should eventually allow for 4k render targets 15965 - mesa driver uses 1/12 mode, this avoids changing the 15966 subpixel mode when switching between ddx and mesa 15967 15968commit 3880bc3c95764acccd62cfa438932458df95daf6 15969Author: Alex Deucher <alexdeucher@gmail.com> 15970Date: Sun May 3 19:01:28 2009 -0400 15971 15972 Remove some old debugging info 15973 15974commit 3362381aa3a97c1da564f8a479ed4e7b0d6368b8 15975Author: Alex Deucher <alexdeucher@gmail.com> 15976Date: Sun May 3 18:52:00 2009 -0400 15977 15978 r2xx-r4xx: fix typo in last i2c fix and clarify hw i2c pin sel 15979 15980commit e01b8dc827ce1913927e1a4fad05a99a2dd9e805 15981Author: Alex Deucher <alexdeucher@gmail.com> 15982Date: Sun May 3 13:09:39 2009 -0400 15983 15984 RS690: gpio/connector fixes 15985 15986 Fixes bug 21521 15987 15988commit 1f70c9f05df9017d87b37f887e1eccd6d0568a02 15989Author: Alex Deucher <alexdeucher@gmail.com> 15990Date: Thu Apr 30 19:18:13 2009 -0400 15991 15992 R3xx-R5xx: don't set TX_OFFSET_RS in RS_INST_COUNT 15993 15994 Isn't necessary and seems to cause problems for RS690 users 15995 15996commit de17d36c1f710305870e3c636cae3f742f1cb6d6 15997Author: Alex Deucher <alexdeucher@gmail.com> 15998Date: Thu Apr 30 19:13:09 2009 -0400 15999 16000 AVIVO: first pass at display watermark setup 16001 16002 This helps with flickering and blanking when 16003 there is contention for MC bandwidth. 16004 16005commit 7d9f643ae3d07e51e644a5979ca90bc2c102bc89 16006Author: Alex Deucher <alexdeucher@gmail.com> 16007Date: Wed Apr 29 12:38:15 2009 -0400 16008 16009 radeon: fix several dpms issues 16010 16011 - atom dpms was unblanking for standby/suspend 16012 - return if r600+ in radeon_crtc_modeset_ioctl() 16013 - remove seprate standby/suspend handling in legacy 16014 crtc dpms. we turn the outputs off, so turn the 16015 crtcs off too. 16016 - disable the crtcs in legacy crtc dpms 16017 - move radeon_crtc_modeset_ioctl() calls to radeon_crtc_dpms() 16018 so they will get called for both atom and legacy paths 16019 16020 Should fix bug 21321 16021 16022commit 35183ef528411eb0122cf48550d93f921d291d12 16023Author: Alex Deucher <alexdeucher@gmail.com> 16024Date: Wed Apr 29 01:55:10 2009 -0400 16025 16026 AVIVO: better fix for cursor flickering/corruption issues 16027 16028 Should prevent ghost cursor from being left on screen. 16029 16030commit 53e75cbd0ba8f39799d7b776cdc59d60aa2a6916 16031Author: Alex Deucher <alexdeucher@gmail.com> 16032Date: Wed Apr 29 01:05:31 2009 -0400 16033 16034 R6xx/r7xx: fix CURSOR_SWAPPING_* macros 16035 16036 r6xx/r7xx have different swapper regs 16037 16038commit 4e948e2aaa59267a9aa797396add2c6b7b4c1e7c 16039Author: Alex Deucher <alexdeucher@gmail.com> 16040Date: Tue Apr 28 19:28:25 2009 -0400 16041 16042 AVIVO: fix cursor corruption/flickering issues 16043 16044 When both crtcs are enabled, both cursors need to be enabled 16045 or you get corruption bands. 16046 16047 Thanks to Yang Zhao for figuring this out. 16048 16049commit 4d0beaf02b28d5a5dac4ff6053593a2f979f92ec 16050Author: Alex Deucher <alexdeucher@gmail.com> 16051Date: Tue Apr 28 12:33:24 2009 -0400 16052 16053 RV410: SE variants only have 1 quad pipe 16054 16055 Should fix EXA corruption with the 3D engine. 16056 Mesa and drm patches forthcoming. 16057 16058 Reported by Kano on IRC. 16059 16060commit abd565ae71d9844d5afcd658628ca7778ce26ab1 16061Author: Alex Deucher <alexdeucher@gmail.com> 16062Date: Mon Apr 27 14:21:38 2009 -0400 16063 16064 AVIVO: disable fractional feedback divs 16065 16066 Doesn't seem to work well in all cases. Needs more investigation. 16067 fixes bug 21413 16068 16069commit 93626b1277f57bd559cf71b263221209b02e0046 16070Author: Alex Deucher <alexdeucher@gmail.com> 16071Date: Mon Apr 27 03:35:24 2009 -0400 16072 16073 AVIVO: make sure cursor width isn't negative 16074 16075commit 32af0f097bccc4499d67e969d46ba35e7487c5e2 16076Author: Dave Airlie <airlied@linux.ie> 16077Date: Sat Apr 25 19:39:34 2009 +1000 16078 16079 Revert "r300/r500: set R clamp mode." 16080 16081 This reverts commit dab3d61619d91ae6a3da3e9b29b8258313653813. 16082 16083commit b2c8f57bccbd46164907049a26ad2dd209d7b108 16084Author: Dave Airlie <airlied@linux.ie> 16085Date: Sat Apr 25 19:39:19 2009 +1000 16086 16087 Revert "r3xx/r5xx: fix typo in last commit" 16088 16089 This reverts commit 1a7b982457ead5f21c6e62a409ca0dbf86005607. 16090 16091commit 253731dce6ab25122f2eda559233c098536b7143 16092Author: Dave Airlie <airlied@linux.ie> 16093Date: Sat Apr 25 19:39:05 2009 +1000 16094 16095 Revert "radeon: attempt to fix R clamping for repeat again" 16096 16097 This reverts commit 5998f262a52cb85b334fcc1fe24c7b2ae474ce93. 16098 16099commit 5998f262a52cb85b334fcc1fe24c7b2ae474ce93 16100Author: Dave Airlie <airlied@linux.ie> 16101Date: Fri Apr 24 17:30:57 2009 +1000 16102 16103 radeon: attempt to fix R clamping for repeat again 16104 16105commit 7ed27b926f28a1f088eace5034a7d9985f13752d 16106Author: Alex Deucher <alexdeucher@gmail.com> 16107Date: Thu Apr 23 14:06:06 2009 -0400 16108 16109 r2xx/r3xx/r4xx: further i2c fixups 16110 16111 - hw i2c engine has pin selection on r2xx/r2xx/r3xx chips 16112 - also switch hw i2c pin sel for external tmds 16113 16114commit 0da80f34b5aa5999d2030ffbc5187328fa4e4ae5 16115Author: Alex Deucher <alexdeucher@gmail.com> 16116Date: Wed Apr 22 18:42:36 2009 -0400 16117 16118 radeon: minor cleanup and warning fix 16119 16120commit 1a7b982457ead5f21c6e62a409ca0dbf86005607 16121Author: Alex Deucher <alexdeucher@gmail.com> 16122Date: Thu Apr 23 10:09:22 2009 -0400 16123 16124 r3xx/r5xx: fix typo in last commit 16125 16126 Noticed by osiris on IRC 16127 16128commit dab3d61619d91ae6a3da3e9b29b8258313653813 16129Author: Dave Airlie <airlied@redhat.com> 16130Date: Thu Apr 23 15:46:44 2009 +1000 16131 16132 r300/r500: set R clamp mode. 16133 16134 On rs690 without this set, we appear to end up fetching from well 16135 outside the texture sometimes into memory that isn't mapped. 16136 16137 This "shouldn't" have any effect. 16138 16139commit bbfd4625478e37d10fd58b23f46153f349370a52 16140Author: Alex Deucher <alexdeucher@gmail.com> 16141Date: Wed Apr 22 14:47:16 2009 -0400 16142 16143 radeon PM: update quirks 16144 16145commit e3c564114eea4d032f920fcebd4040f1934c21b0 16146Author: Alex Deucher <alexdeucher@gmail.com> 16147Date: Wed Apr 22 14:31:44 2009 -0400 16148 16149 AVIVO: fix for cursor corruption issue 16150 16151commit 24e4b73b4fbbb2c790e6120ede3caaa4e7e58359 16152Author: Alex Deucher <alexdeucher@gmail.com> 16153Date: Wed Apr 22 11:42:18 2009 -0400 16154 16155 radeon pll: add support for fractional feedback divs 16156 16157 Allows us to hit dot clocks much closer, especially on 16158 chips with non-27 Mhz reference clocks like most IGP chips. 16159 16160 This should fix most flickering and blanking problems with 16161 non-exact dot clocks. 16162 16163commit efa0825a86a8dc0f03ebb42c576ed26189e9d4bb 16164Author: Alex Deucher <alexdeucher@gmail.com> 16165Date: Tue Apr 21 13:51:11 2009 -0400 16166 16167 Tex vid: fix top clipping 16168 16169commit aa59b5b55e6ac64940fb3f0fa61f7b8fe36ba745 16170Author: Alex Deucher <alexdeucher@gmail.com> 16171Date: Tue Apr 21 12:06:24 2009 -0400 16172 16173 RS690: add quirk for acer board 16174 16175 thanks to Gino Badouri for testing 16176 16177commit 0e5164d968316bbb59b4484a7df087854738068f 16178Author: Alex Deucher <alexdeucher@gmail.com> 16179Date: Tue Apr 21 04:47:43 2009 -0400 16180 16181 Tex vid: rework part 2 16182 16183 - use pPriv->w/h directly for tex coords 16184 - take src x/y offset into account when calculating tex coords 16185 - when copying data, only optimize for y clipping. In order 16186 to deal with the x clipping optimization, the copy routines 16187 or tex coords would need to be fixed. This should fix clipping 16188 problems with the current code. 16189 16190commit 06a1b4169b506f69dd807380d9ce7210b3a3bc23 16191Author: Alex Deucher <alexdeucher@gmail.com> 16192Date: Tue Apr 21 01:21:51 2009 -0400 16193 16194 R4xx: avoid a PM related hang 16195 16196commit 228b9ac87668241b09980fca17e9d866891285e6 16197Author: Alex Deucher <alexdeucher@gmail.com> 16198Date: Tue Apr 21 01:02:34 2009 -0400 16199 16200 radeon PM: add some quirks 16201 16202commit b6febf7368d3df3f1d3fee6a87aa71c54e77cdcc 16203Author: Alex Deucher <alexdeucher@gmail.com> 16204Date: Mon Apr 20 18:04:18 2009 -0400 16205 16206 Tex vid: cleanup and unify common code 16207 16208commit d096d1189c8dc5e19f68cc605ccffe1a8fef048b 16209Author: Alex Deucher <alexdeucher@gmail.com> 16210Date: Mon Apr 20 13:05:18 2009 -0400 16211 16212 Endian fixes for object table 16213 16214commit 101ce28ec4af696549b7d6defebb5b4ad412176c 16215Author: Alex Deucher <alexdeucher@gmail.com> 16216Date: Sun Apr 19 12:13:18 2009 -0400 16217 16218 Remove old messages 16219 16220 R5xx/r6xx support isn't really experimental anymore 16221 16222commit db177c70ce88af19c8e05eb916a15f3e124876eb 16223Author: Alex Deucher <alexdeucher@gmail.com> 16224Date: Fri Apr 17 01:05:15 2009 -0400 16225 16226 Update Xv info in man page 16227 16228commit 00266177bed2dc8693df497ca3ec19f2dc4adc05 16229Author: Alex Deucher <alexdeucher@gmail.com> 16230Date: Thu Apr 16 20:33:28 2009 -0400 16231 16232 R3xx/R5xx: only apply Xv attributes if bicubic is disabled 16233 16234 Provides consistent output 16235 16236commit 85323a7f84381fef7fad20c7f7ec601637af9aa7 16237Author: Alex Deucher <alexdeucher@gmail.com> 16238Date: Thu Apr 16 11:10:51 2009 -0400 16239 16240 R3xx/R4xx: set tex caching for Y texture when doing planar rendering 16241 16242 Doesn't affect performance, but docs indicate its the right 16243 thing to do. 16244 16245commit 5ea5df22c038fc8f00984acc760e9d8c962bf902 16246Author: Alex Deucher <alexdeucher@gmail.com> 16247Date: Wed Apr 15 20:26:34 2009 -0400 16248 16249 Tex vid: remove remnants of XV_HWPLANAR 16250 16251 no longer needed as bicubic is the only thing that uses 16252 the old csc code. 16253 16254commit 9091b3f5f13dbea83ffd89679dac600e9f280bb2 16255Author: Alex Deucher <alexdeucher@gmail.com> 16256Date: Wed Apr 15 20:07:42 2009 -0400 16257 16258 R3xx/R4xx: fix up planar shader 16259 16260 We were overwriting the coord fetch address with the first 16261 tex fetch. Seemed to work however, luck I guess. Reorder 16262 the fetches to write to temp0 last. 16263 16264commit 14c13faeb9f9b7717a25fcc1ca97d46cc6ee0031 16265Author: Alex Deucher <alexdeucher@gmail.com> 16266Date: Wed Apr 15 19:53:12 2009 -0400 16267 16268 R5xx: add shader-based csc 16269 16270 - native planar support 16271 - Xv attributes 16272 16273commit 832efc7b90f5eb2da99512fcb902ab4838d2dcd1 16274Author: Alex Deucher <alexdeucher@gmail.com> 16275Date: Tue Apr 14 17:56:49 2009 -0400 16276 16277 R3xx/R4xx: Implement shader-based csc for packed formats 16278 16279commit 32625118c27041265d25811c00d25ab7e82fb340 16280Author: Alex Deucher <alexdeucher@gmail.com> 16281Date: Tue Apr 14 16:38:40 2009 -0400 16282 16283 tex vid: fix attribute setup typo for XV_COLORSPACE 16284 16285commit adf0912006b4f1597784dbfcc563d5c6d1c5667d 16286Author: Alex Deucher <alexdeucher@gmail.com> 16287Date: Tue Apr 14 16:16:33 2009 -0400 16288 16289 R6xx/R7xx: implement Xv attributes 16290 16291 - brightness, contrast, hue, etc. 16292 - TODO: implement gamma 16293 16294commit 8810fe92b5aed08888584c6914482586b59f71ab 16295Author: Alex Deucher <alexdeucher@gmail.com> 16296Date: Tue Apr 14 11:26:35 2009 -0400 16297 16298 R200: clean code, always use shader based csc 16299 16300 - consolidate common r2xx csc shader code 16301 - always use shader based csc for both packed and planar 16302 formats 16303 - always use native planar csc on r1xx 16304 16305commit 17685fefba68d188c7c0fe7a079180ec0722c046 16306Author: Alex Deucher <alexdeucher@gmail.com> 16307Date: Mon Apr 13 20:06:11 2009 -0400 16308 16309 Tex vid: general cleanup 16310 16311 - convert macros to more meaningful VTX_OUT_4 and VTX_OUT_6 16312 names to reflect that they actually do 16313 - fix indenting 16314 16315commit 093ab4c9a33b0b396b78c061c3321dc044bdccdc 16316Author: Alex Deucher <alexdeucher@gmail.com> 16317Date: Mon Apr 13 19:48:35 2009 -0400 16318 16319 R1xx: add support for native planar textured Xv 16320 16321commit ec0cb51df81c6c9a1de640d227fa9c9c33161083 16322Author: Alex Deucher <alexdeucher@gmail.com> 16323Date: Mon Apr 13 17:21:20 2009 -0400 16324 16325 R2xx tex vid: append verts for clip boxes 16326 16327 rather than sending a new draw packet for each rect 16328 16329commit fde075a30a8ee2c333aa1bbe8fbd177258b085ba 16330Author: Alex Deucher <alexdeucher@gmail.com> 16331Date: Mon Apr 13 17:13:51 2009 -0400 16332 16333 R1xx tex vid: append verts for clip boxes 16334 16335 rather than sending a new draw packet for each rect 16336 16337commit 12839fc17a2cca4ac14b9757bdaa63ba4679f96f 16338Author: Alex Deucher <alexdeucher@gmail.com> 16339Date: Mon Apr 13 17:04:31 2009 -0400 16340 16341 Tex vid: split by family 16342 16343commit a30737b337edb31528174b483c9094941a5d41bb 16344Author: Roland Scheidegger <sroland@tungstengraphics.com> 16345Date: Mon Apr 13 15:36:07 2009 -0400 16346 16347 r200/r300: implement brightness/contrast/hue/saturation/gamma controls for textured video 16348 16349 This implements 16350 contrast/brightness/hue/saturation controls for r200/r300 plus gamma (same 16351 gamma value for all channels used though separate values would be trivial) 16352 control for r300. 16353 Some issues left: 16354 - only r200/r300 16355 - still can't be combined with bicubic 16356 - controls will silently cease to work if the format used is packed and not 16357 planar (except for rv250) 16358 - gamma range is from 100 to 10000 corresponding to 0.1 and 10.0 like used in 16359 overlay. However, usable range is far smaller. Over 2.0 picture gets dark 16360 pretty quickly, and below 0.6 or so black seems to turn into purple (I've 16361 verified that even with gamma 1.0 black actually often seems to be RGB 1/0/1 so 16362 this explains this since that gets amplified by low gamma values - not sure if 16363 this is a rounding problem somewhere, bogus reference values or is somehow 16364 expected). 16365 - gamma adds a bit too many instructions for my taste (7) though the 16366 alternative (3 texture lookups + some swizzling instructions) doesn't seem any 16367 better. 16368 16369commit 9dd33cc8346073300312cd5f68de72d9bea1dd6b 16370Author: Dave Airlie <airlied@redhat.com> 16371Date: Fri Apr 17 10:30:21 2009 +1000 16372 16373 radeon: tweak ddc fix for all r300s. 16374 16375 Alex seems to think this might affect some other chips 16376 16377commit a846ec14dd65bb12f314f83ffb45c4c4be564a5e 16378Author: Dave Airlie <airlied@redhat.com> 16379Date: Fri Apr 17 10:24:15 2009 +1000 16380 16381 radeon: fix DDC on rv410 VGA ports for non-atom use. 16382 16383 The rv410 hw i2c block appears to hold down whatever i2c lines 16384 it is connected to when in reset, so when doing DDC change hw i2c to 16385 point at different i2c lines. 16386 16387commit efa1476fc0765b53298624198137dc5cc71d37a8 16388Author: Alex Deucher <alexdeucher@gmail.com> 16389Date: Thu Apr 16 16:43:02 2009 -0400 16390 16391 PM: fix reversed logic in pm block handler 16392 16393commit 2b95de17781959457a809c8fecc6bbb08336c83f 16394Author: Alex Deucher <alexdeucher@gmail.com> 16395Date: Thu Apr 16 12:52:41 2009 -0400 16396 16397 RS400/RS480: attempt to fix ClockGating on RS4xx chips 16398 16399commit c0215d2f2a27755fb04d20aaa4320a5e7da7ee04 16400Author: Alex Deucher <alexdeucher@gmail.com> 16401Date: Thu Apr 16 11:58:19 2009 -0400 16402 16403 RS690/RS740: fix connector enumeration in some cases 16404 16405 Should fix Masta-G's issue reported on IRC. 16406 16407commit 573339114660f06e5f257648ba413bab0d6083ce 16408Author: Alex Deucher <alexdeucher@gmail.com> 16409Date: Thu Apr 16 11:19:01 2009 -0400 16410 16411 R3xx/R4xx EXA: set explicit tex cache regions 16412 16413 When using a mask, set explicit cache partitions for 16414 each texture. Gives 1% performance improvement in 16415 x11perf with anti-aliased text. 16416 16417commit eea9800890b56bac9c07b7bd9c9e33fae2938af3 16418Author: Alex Deucher <alexdeucher@gmail.com> 16419Date: Thu Apr 16 10:41:28 2009 -0400 16420 16421 Revert "DCE3/3.2: cleanup dpms after previous changes" 16422 16423 This reverts commit f8c7d6a6162196a743f6885ecaf63ba50de1722a. 16424 16425 This is apparently still needed for some setups, however, I can't 16426 reproduce this locally anymore. 16427 16428commit 099cb27d760af61b967239d960bc77f71e828655 16429Author: Alex Deucher <alexdeucher@gmail.com> 16430Date: Wed Apr 15 16:23:01 2009 -0400 16431 16432 Change ForceLowPowerMode to 4 pcie lanes 16433 16434 This fixes corruption for some users 16435 16436commit 5abcea88d1f3307a7d08ae7757c4b535df1503bf 16437Author: Alex Deucher <alexdeucher@gmail.com> 16438Date: Wed Apr 15 15:52:04 2009 -0400 16439 16440 PM: don't touch clock gating regs unless ClockGating option is set 16441 16442 Fixes an M26 hang reported by tormod on IRC 16443 16444commit bae88f81e2a01c0d547132b5b91e717ff9d0497d 16445Author: Alex Deucher <alexdeucher@gmail.com> 16446Date: Wed Apr 15 14:38:35 2009 -0400 16447 16448 man page: fix missing break 16449 16450commit 4b3a3785f7b06205eeec6ca6b8913ecc2b94bd1f 16451Author: Alex Deucher <alexdeucher@gmail.com> 16452Date: Sat Apr 4 11:39:35 2009 -0400 16453 16454 Unify ForceLowPowerMode and DynamicPM Options 16455 16456 Force lower power mode and switch to a lower mode 16457 when idle. 16458 16459commit 63cb57c5eb0d6bc083f54eb100cb972b0e7bca69 16460Author: Alex Deucher <alexdeucher@gmail.com> 16461Date: Wed Apr 1 16:22:02 2009 -0400 16462 16463 Add DynamicPM Option 16464 16465 Dyanmically switch between power states. Switch to a low 16466 power state when the system is idle (DPMS off). 16467 16468commit a34a8b37afbea6ed4bf8ca42364195e174250c48 16469Author: Alex Deucher <alexdeucher@gmail.com> 16470Date: Wed Apr 1 15:23:17 2009 -0400 16471 16472 Set default low power PCIE lanes to 2 16473 16474 1 lane seems to cause occasional corruption when 16475 blitting to/from gart memory. 16476 16477commit 1f0dc778dc25f4f85fedd73c55c847cab2c79fc5 16478Author: Alex Deucher <alexdeucher@gmail.com> 16479Date: Wed Apr 1 15:12:27 2009 -0400 16480 16481 Add support for setting the number PCIE lanes 16482 16483commit 7e10b6222e8f44a3ecc6aaea55a7a7680d133bb3 16484Author: Alex Deucher <alexdeucher@gmail.com> 16485Date: Wed Apr 1 15:05:38 2009 -0400 16486 16487 Add ForceLowPowerMode Option 16488 16489 Force the chip to a low power mode at the expense 16490 of performance. 16491 16492commit 533d01a5933cd491bbc09cd463ea62475abf4bf2 16493Author: Alex Deucher <alexdeucher@gmail.com> 16494Date: Tue Mar 31 17:24:29 2009 -0400 16495 16496 radeon: Add functions to set sclk/mclk using atombios 16497 16498commit adb099409768e695b9928fa6aa5760f93dadd9af 16499Author: Alex Deucher <alexdeucher@gmail.com> 16500Date: Tue Mar 31 17:13:11 2009 -0400 16501 16502 radeon: Add functions to set sclk/mclk on r1xx-r4xx 16503 16504commit 68e2a959ccc3d1a5d0731f1b55fdf1b2412635b2 16505Author: Alex Deucher <alexdeucher@gmail.com> 16506Date: Tue Mar 31 17:02:45 2009 -0400 16507 16508 Add unified ClockGating Option 16509 16510 - replaces DynamicClocks Option as the name was misleading 16511 - unified interface for atom and com based bioses 16512 - fix up clock gating code for newer r3xx asics 16513 16514commit 336cee0e6d19068582b189b2c747d92cb2139d2f 16515Author: Dave Airlie <airlied@redhat.com> 16516Date: Mon Apr 13 18:11:48 2009 +1000 16517 16518 radeon: legacy bandwidth calcs for later r3xx/r4xx cards. 16519 16520 This is an attempt to rationalise the code using the register info 16521 files available to me here. 16522 16523 Further info is required: 16524 r350 check for stop_req > 15 then subtract 10: what other chips need it? 16525 get rs480 working instead of using magic. 16526 16527commit 367ba7af76e7f63f3186a6e8d9bbe21aa31fc6d1 16528Author: Alex Deucher <alexdeucher@gmail.com> 16529Date: Fri Apr 10 09:37:15 2009 -0400 16530 16531 radeon: fix tmds_pll randr property 16532 16533commit 15f1a2f31a7bc33f276a65283dbad0c106bee73e 16534Author: Alex Deucher <alexdeucher@gmail.com> 16535Date: Thu Apr 9 20:17:03 2009 -0400 16536 16537 radeon: remove old crtc size heuristics 16538 16539 IIRC, the old randr code used to use this to use for front buffer sizing, 16540 but it has since been changed. 16541 16542commit ce9077b2f5a5e749c0078dfb79d26ef8711a823a 16543Author: Alex Deucher <alexdeucher@gmail.com> 16544Date: Thu Apr 9 19:59:19 2009 -0400 16545 16546 R6xx/R7xx: cleanup macro 16547 16548commit 6532aeddacdeda3aa534264741c8648e1c449e76 16549Author: Michel Dänzer <daenzer@vmware.com> 16550Date: Thu Apr 9 09:41:56 2009 +0200 16551 16552 Clear the framebuffer before initializing modes. 16553 16554 Fixes garbage being visible shortly on server startup or when VT switching back 16555 to X. 16556 16557commit 0c19053975925fed637aaa41d440e0cab8f6b596 16558Author: Alex Deucher <alexdeucher@gmail.com> 16559Date: Wed Apr 8 13:34:23 2009 -0400 16560 16561 radeon: chip family typo 16562 16563commit 2888dd9ae9689b1cd72115dc0ceea1f5957299b0 16564Author: Alex Deucher <alexdeucher@gmail.com> 16565Date: Wed Apr 8 13:10:01 2009 -0400 16566 16567 R6xx/R7xx: set proper 3D client driver name for r600 16568 16569commit 61a083116a2c89e145aa19da951243536e07542e 16570Author: Alex Deucher <alexdeucher@gmail.com> 16571Date: Wed Apr 8 10:44:30 2009 -0400 16572 16573 Bump version post release 16574 16575commit 1b02b93895c31a0c9d641e47b46dce43b40edd97 16576Author: Alex Deucher <alexdeucher@gmail.com> 16577Date: Wed Apr 8 10:34:44 2009 -0400 16578 16579 Bump for release 16580 16581commit da74b94a5da8f0bcf4ae99b20e594d2f0f4fc6f6 16582Author: Alex Deucher <alexdeucher@gmail.com> 16583Date: Tue Apr 7 18:59:06 2009 -0400 16584 16585 ATOM: code cleanup 16586 16587 - remove cases that aren't possible (e.g., no TMDSA on DCE3 cards) 16588 - condense duplicate cases 16589 16590commit a24aa7250d09ae68ceaaba987314388b5e52096c 16591Author: Alex Deucher <alexdeucher@gmail.com> 16592Date: Tue Apr 7 18:52:08 2009 -0400 16593 16594 DCE3/3.2: further fixes 16595 16596 disable the dig transmitter and encoder before setup and enable 16597 16598commit 284e1943f6726ed08e796f99e49fbd90ac7fa8a4 16599Author: Alex Deucher <alexdeucher@gmail.com> 16600Date: Wed Apr 1 15:07:01 2009 -0400 16601 16602 Cast info->sclk to int 16603 16604commit b580e388e0e31843308c783c1ca859672dc6ca0e 16605Author: Alex Deucher <alexdeucher@gmail.com> 16606Date: Wed Apr 1 14:59:43 2009 -0400 16607 16608 Fix missing ')' in PCIE macro 16609 16610commit 8ec27be3c8043a6ddc816b0b1480dd19f8d73832 16611Author: Alex Deucher <alexdeucher@gmail.com> 16612Date: Tue Apr 7 11:33:44 2009 -0400 16613 16614 radeon: fix Xv vsync for multi-head 16615 16616commit f8c7d6a6162196a743f6885ecaf63ba50de1722a 16617Author: Alex Deucher <alexdeucher@gmail.com> 16618Date: Sat Apr 4 16:05:20 2009 -0400 16619 16620 DCE3/3.2: cleanup dpms after previous changes 16621 16622commit a707d355c3c6ff92252c5a060a1fc32d97547552 16623Author: Alex Deucher <alexdeucher@gmail.com> 16624Date: Sat Apr 4 15:20:01 2009 -0400 16625 16626 DCE3/3.2: fix up transmitter/encoder setup 16627 16628 Should fix bug 21050 16629 16630commit 215e12f9c0e8ac62c23af1add776ef88f9a0dc54 16631Author: Eduard Fuchs <edfuchs@uni-kassel.de> 16632Date: Thu Apr 2 12:36:59 2009 -0400 16633 16634 Enable byte swapping for r6xx/r7xx Hardware 16635 16636 Patch from Eduard Fuchs with some cleanup from me. 16637 16638 Tested at 32 bpp on MPC8641HPCN board (PowerPC) with 16639 HD2400 PCIe card 16640 16641commit 0ded9fab6bc3fecc1976ae8369b788659e4d7793 16642Author: Alex Deucher <alexdeucher@gmail.com> 16643Date: Thu Apr 2 12:09:00 2009 -0400 16644 16645 Add RV790 (HD 4890) support 16646 16647commit 0c6f20a9404c6a4210c5180bdca61c13fc186754 16648Author: Alex Deucher <alexdeucher@gmail.com> 16649Date: Tue Mar 31 15:51:11 2009 -0400 16650 16651 Add PCIE register accessors 16652 16653commit 229d4e20e5adcdf9207944c19ca721477936ff02 16654Author: Alex Deucher <alexdeucher@gmail.com> 16655Date: Tue Mar 31 15:39:42 2009 -0400 16656 16657 Unify mc_idle() 16658 16659commit f44b486848094b75479abef61c1aa075aede835c 16660Author: Alex Deucher <alexdeucher@gmail.com> 16661Date: Tue Mar 31 15:33:11 2009 -0400 16662 16663 COMBIS: Clarify case 0 for LCD DDC table 16664 16665commit e6f3c48899a92174a55f6fe7327412bc88039d87 16666Author: Alex Deucher <alexdeucher@gmail.com> 16667Date: Tue Mar 31 15:30:40 2009 -0400 16668 16669 COMBIOS: fix default sclk/mclk from bios 16670 16671commit 75bf5272936093bc024559e5744d97d8e79943b1 16672Author: Alex Deucher <alexdeucher@gmail.com> 16673Date: Tue Mar 31 15:17:37 2009 -0400 16674 16675 mac mini: 3rd time's the charm 16676 16677 fixes 10fbbac4bed59e12ad794ac873dd415d9ee4e340 16678 fixes bugs 20796 and 20979 16679 16680commit 10fbbac4bed59e12ad794ac873dd415d9ee4e340 16681Author: Alex Deucher <alexdeucher@gmail.com> 16682Date: Tue Mar 31 15:14:27 2009 -0400 16683 16684 Mac mini: fix connector setup for real 16685 16686 fixes f061308e7abcf93d1612681b935387fee23e0a24 16687 fixes bugs 20796 and 20979 16688 16689commit 8f2a03bfc86422669a46c2853050979f099a17a8 16690Author: Alex Deucher <alexdeucher@gmail.com> 16691Date: Mon Mar 30 11:30:34 2009 -0400 16692 16693 r1xx-r3xx TV: fix white noise 16694 16695 should fix bug 13872 16696 16697commit f061308e7abcf93d1612681b935387fee23e0a24 16698Author: Alex Deucher <alexdeucher@gmail.com> 16699Date: Fri Mar 27 01:14:12 2009 -0400 16700 16701 mac mini: fix connector setup regression 16702 16703 fixes bug 20796 16704 16705commit 47134ca77988f8c3b42d55ac4fcbb5dc182878d4 16706Author: Alex Deucher <alexdeucher@gmail.com> 16707Date: Thu Mar 26 17:18:47 2009 -0400 16708 16709 radeon: re-enable r/g/b intensity attributes for overlay 16710 16711 Noticed by sroland on IRC. 16712 16713commit e2b5e48e2652daf03fcae852d8368f29ea4abcd9 16714Author: Alex Deucher <alexdeucher@gmail.com> 16715Date: Thu Mar 26 13:42:25 2009 -0400 16716 16717 R6xx/R7xx EXA: rework composite pixel shader 16718 16719 - move to vram storage 16720 - move swizzle logic to tex setup 16721 16722commit 24dbd3f93c640f819a9f5029f5a49e41ab875bd4 16723Author: Alex Deucher <alexdeucher@gmail.com> 16724Date: Thu Mar 26 12:01:07 2009 -0400 16725 16726 R6xx/R7xx: clean up logic in EXA Comp PS setup 16727 16728commit 076e4e3ac1c81086aecb4e70dd30fd5d26ec7bc8 16729Author: Alex Deucher <alexdeucher@gmail.com> 16730Date: Thu Mar 26 11:36:26 2009 -0400 16731 16732 R6xx/R7xx: clean up bool const code 16733 16734 3 regs: 1 bit per bool, 32 bools per ps/vs/gs 16735 16736commit d2c3964fe04be42fe538f36439ed5ffca96e436a 16737Author: Roland Scheidegger <sroland@tungstengraphics.com> 16738Date: Wed Mar 18 01:55:12 2009 +0100 16739 16740 fix textured video allocation bug 16741 16742 size needs to be calculated after dstPitch adjustments, got already fixed 16743 for earlier than R600 chips by planar textured yuv patches, clean this up. 16744 16745commit 18e56eb179fde28477487c63e6f9ebf7579e2cd5 16746Author: Roland Scheidegger <sroland@tungstengraphics.com> 16747Date: Thu Mar 5 02:07:46 2009 +0100 16748 16749 don't convert planar yuv to packed for r200 16750 16751 uses 3 textures for planar yuv and does yuv->rgb conversion in the shader. 16752 Similar to r300 code, but might have precision issues - hardware alu should 16753 have enough precision but hardware consts are only 8bit and we'd want 16754 at least 11. 16755 This also enables textured video on rv250 (and also supports packed yuv 16756 on that chip by using basically the same shader with packed data). 16757 16758commit 58530bf4912800f9e09ebaea42a13cff8a80c19e 16759Author: Roland Scheidegger <sroland@tungstengraphics.com> 16760Date: Sat Feb 21 04:46:31 2009 +0100 16761 16762 don't convert planar yuv to packed for r300 16763 16764 uses 3 textures for planar yuv and does yuv->rgb conversion in the shader. 16765 small performance advantage, but manual texture cache setting is necessary 16766 otherwise it may be measurably slower (but probably not relevant) in some 16767 cases. 16768 Unlike some other drivers, using MADs instead of DP3s, since this requires 16769 less instructions due to no MOVs are required, the end result is the same 16770 though the constants need to be different. 16771 Use of this is user settable for now (XV_HWPLANAR attrib). 16772 16773commit 97e19d96ba65a3df2fa3bbf73cfcc01b6dc3e796 16774Author: Roland Scheidegger <sroland@tungstengraphics.com> 16775Date: Tue Dec 30 22:23:39 2008 +0100 16776 16777 clip fixes 16778 16779 This fixes some oddities observed when the video is only partly visible. 16780 Instead of recalculating the geometry of the video, always use the same. 16781 Also fixes a assignment present twice, and another issue (bring in line with 16782 what the overlay code does). 16783 16784commit 7aaf5f08e4147b75c4975e6d680d3f5a93424164 16785Author: Alex Deucher <alexdeucher@gmail.com> 16786Date: Mon Mar 23 12:55:02 2009 -0400 16787 16788 R4xx ATOM: fix crtc routing for CRT1 16789 16790 should fix bug 20814 16791 16792commit 2fe6b393d9dbb193d4f0a3351ad76618cda07e2c 16793Author: Alex Deucher <alexdeucher@gmail.com> 16794Date: Sun Mar 22 01:31:30 2009 -0400 16795 16796 radeon: add new chip ids 16797 16798commit 1ddee7cd6fd267b2fc86f21af27c5425eb0835a4 16799Author: Alex Deucher <alexdeucher@gmail.com> 16800Date: Fri Mar 20 13:44:00 2009 -0400 16801 16802 DCE3.2: use RMX for for non-native modes on DVI 16803 16804 DVI seems to have issues with low dotclocks, so 16805 use the scaler instead. 16806 16807 Fixes bug 20754 16808 16809commit 44081e8fd2bf974272d5447b45dd7b206cf939c1 16810Author: Alex Deucher <alexdeucher@gmail.com> 16811Date: Thu Mar 19 20:48:16 2009 -0400 16812 16813 R6xx/R7xx EXA: fix maxPitchBytes 16814 16815 should now allow accel up the hw max of 8192x8192 16816 16817commit e8e2367b66d9a4f5977d95f52b37cdbad75e409a 16818Author: Alex Deucher <alexdeucher@gmail.com> 16819Date: Thu Mar 19 20:09:32 2009 -0400 16820 16821 Add new pci ids 16822 16823commit 68e83f10adb23fc359c1da881f5d666cd08b4067 16824Author: Alex Deucher <alexdeucher@gmail.com> 16825Date: Wed Mar 18 20:58:19 2009 -0400 16826 16827 bump version post release 16828 16829commit a196aae19603c8651dc502f5cb0df4d55befe163 16830Author: Alex Deucher <alexdeucher@gmail.com> 16831Date: Wed Mar 18 20:52:12 2009 -0400 16832 16833 bump for release 16834 16835commit 537da7018b0fd2ad263fc5cdc55b3b85cf8e9205 16836Author: Alex Deucher <alexdeucher@gmail.com> 16837Date: Wed Mar 18 19:22:55 2009 -0400 16838 16839 ATOM: fix up tv-out support 16840 16841 - fix purple hue when using ATOMTvOut option 16842 - fix TV load detection on newer r6xx/r7xx chips 16843 - disable CV for now 16844 16845commit c0e2513ab128ddd5be0ed626d9e31777a98983ef 16846Author: Alex Deucher <alexdeucher@gmail.com> 16847Date: Wed Mar 18 14:08:30 2009 -0400 16848 16849 R6xx/R7xx AGP: disable gart data transfers 16850 16851 UTS/DFS/Xv 16852 16853 Doesn't seem to be reliable on AGP. 16854 16855 fixes bugs: 20436, 20236, several reports on ML and IRC 16856 16857commit e3e1f0c7b892b094b292081021a0e97932b8e079 16858Author: Fabio <fabio.ped@libero.it> 16859Date: Wed Mar 18 12:13:53 2009 -0400 16860 16861 man page update 16862 16863commit 6ae54d3ea023968887bf0cb0ec1be2115621d554 16864Author: Alex Deucher <alexdeucher@gmail.com> 16865Date: Wed Mar 18 11:50:22 2009 -0400 16866 16867 AVIVO: Fix dac load detection 16868 16869commit 04e36b794c216e4fa77c780fd989df1ef1c7f0d9 16870Author: Fabio <fabio.ped@libero.it> 16871Date: Tue Mar 17 11:29:26 2009 -0400 16872 16873 man page updates 16874 16875 from bug 20648 16876 16877commit a6855c370194b6df307ea33724fe17a85d67607e 16878Author: Alex Deucher <alexdeucher@gmail.com> 16879Date: Mon Mar 16 13:52:42 2009 -0400 16880 16881 radeon: add another AGP quirk 16882 16883 bug 20647 16884 16885commit a55ced5ee20c07e743c7c0978803fd10589c1531 16886Author: Alex Deucher <alexdeucher@gmail.com> 16887Date: Sun Mar 15 12:08:04 2009 -0400 16888 16889 Fix fallout from r6xx/r7xx EXA merge in IB handling 16890 16891 Noticed Christiaan van Dijk. 16892 Should fix bug 20510 16893 16894commit a065eeb9b301ab452cc8fbd3cac1fd9a3841c4e1 16895Author: Alex Deucher <alexdeucher@gmail.com> 16896Date: Sat Mar 14 18:40:28 2009 -0400 16897 16898 atom: fix up bios scartch register usage 16899 16900 the bios scratch registers keep 3 types of state for each output: 16901 1. dpms 16902 2. active 16903 3. connected 16904 16905 some of the command tables use the active flags. We used to set the active 16906 flags based on dpms state, but instead set them based on whether they are 16907 connected or not. 16908 16909commit 5c4ef924d082fd3b279f62fda7a97f563b009d38 16910Author: Alex Deucher <alexdeucher@gmail.com> 16911Date: Sat Mar 14 18:16:49 2009 -0400 16912 16913 radeon: fix typo in bios scratch reg setup 16914 16915commit 96fce5a56269509a935e8ee621a5e17cc9005ddc 16916Author: Alex Deucher <alexdeucher@gmail.com> 16917Date: Fri Mar 13 19:29:58 2009 -0400 16918 16919 Bump post-release 16920 16921commit 8da1b82a70de7b3ef38905bd77c58a20fd44ddd6 16922Author: Alex Deucher <alexdeucher@gmail.com> 16923Date: Fri Mar 13 19:23:19 2009 -0400 16924 16925 bump for release 16926 16927commit 28d9d699dcdff394aa561505f7f54f08db800828 16928Author: Alex Deucher <alexdeucher@gmail.com> 16929Date: Fri Mar 13 18:18:31 2009 -0400 16930 16931 radeon: man page updates 16932 16933commit 06963796d00b5cb6704412a7d4417fcefae4aa4d 16934Author: Adam Jackson <ajax@redhat.com> 16935Date: Fri Mar 13 18:14:43 2009 -0400 16936 16937 uniphy: start adding DisplayPort setup 16938 16939commit 1db81afe4f8db55e02a65cd22ad84c6bbb27e8ef 16940Author: Alex Deucher <alexdeucher@gmail.com> 16941Date: Fri Mar 13 17:51:56 2009 -0400 16942 16943 radeon: add support for 30 bit LUTs 16944 16945 All radeons have them. Thanks to Yang Zhao for figuring 16946 this out. 16947 16948commit 5bb326c64ad3bf5c93943280dcbcb25c8910d1aa 16949Author: Alex Deucher <alexdeucher@gmail.com> 16950Date: Fri Mar 13 14:10:58 2009 -0400 16951 16952 R6xx/R7xx: EXA VSync Option not supported yet 16953 16954commit 8c2d40894448329e721360811c124998a17ba2ba 16955Author: Alex Deucher <alexdeucher@gmail.com> 16956Date: Fri Mar 13 14:03:43 2009 -0400 16957 16958 R6xx/R7xx: fix up vline stuff along the lines of previous chips 16959 16960 Avoids an additional function call. 16961 16962commit 176c1b43fc3ee48e87c134446754360b3683c24b 16963Author: Alex Deucher <alexdeucher@gmail.com> 16964Date: Fri Mar 13 13:28:42 2009 -0400 16965 16966 R6xx/R7xx: return in RADEONWaitForIdleMMIO() if accel is off 16967 16968 Don't wait for engine idle if we haven't initialized the engine. 16969 Behavior should be consistent with versions of the driver prior 16970 to accel support being added. 16971 16972 Should fix bug 20645 16973 16974commit d538a4508248adf2debaf0cc870a2d1c48abd986 16975Author: Alex Deucher <alexdeucher@gmail.com> 16976Date: Fri Mar 13 13:22:03 2009 -0400 16977 16978 R6xx/r7xx: clarify accel messages 16979 16980 We only support EXA and and only with DRI. 16981 16982commit 1b7090a4c0d6a38701fbfb928fac8e15d40fb81f 16983Author: Adam Jackson <ajax@redhat.com> 16984Date: Fri Mar 13 17:41:59 2009 -0400 16985 16986 Add some more cases to the DVI single-link checks. 16987 16988commit 104e13da74de37477e3a31a5de96c5ebb8fe888b 16989Author: Cooper Yuan <cooperyuan@gmail.com> 16990Date: Fri Mar 13 11:47:43 2009 -0400 16991 16992 radeon: save bios scratch registers in Preinit() 16993 16994 The display detect routines can change these which causes 16995 havok with some bioses. 16996 16997commit f70d7a0fd9ac3659d93511ab362eb80a7ff49777 16998Author: Dave Airlie <airlied@redhat.com> 16999Date: Fri Mar 13 12:56:29 2009 +1000 17000 17001 r600: reload shaders into VRAM on resume 17002 17003 As VRAM gets zeroed out over s/r, we need to reload the 17004 shaders. 17005 17006commit 1fe0dcfe98dfe30499ba53c1f208c4ef10bef001 17007Author: Yang Zhao <yang@yangman.ca> 17008Date: Thu Mar 12 18:50:32 2009 -0400 17009 17010 R6xx/R7xx shader: Fix OFFSET_[XYZ] macro for TEX_DWORD2 to accept floats 17011 17012 Values for OFFSET_[XYZ] are 5-bits two's-complement fixed-point with 17013 one-bit after decimal point. Values in [-8.0, 7.5] are valid. Inputs 17014 that do not exactly land on 0.5 increments are rounded towards 0 to the 17015 nearest increment. 17016 17017commit f356ca7795531e0c080775efb598141eb7db44ba 17018Author: Alex Deucher <alexdeucher@gmail.com> 17019Date: Thu Mar 12 03:42:40 2009 -0400 17020 17021 RS600: fix page table size for rs600 as well 17022 17023commit ef69336584c45e35aabb797a06c4030e75d66507 17024Author: Dave Airlie <airlied@redhat.com> 17025Date: Thu Mar 12 14:54:18 2009 +1000 17026 17027 r600: fix sizing of PCI GART table for r600 17028 17029commit 945ccbbd4fa2b65ccdfb23716c178c95b036734d 17030Author: Alex Deucher <alexdeucher@gmail.com> 17031Date: Tue Mar 10 09:53:14 2009 -0400 17032 17033 radeon: clean more thoroughly in RADEONFreeRec() 17034 17035commit 03ec2099fdc859b23e5fa56ac8696ec045f1b110 17036Author: 오유연(Yu-yeon Oh) <yuyeon_oh@tmax.co.kr> 17037Date: Tue Mar 10 09:51:08 2009 -0400 17038 17039 radeon_driver.c small memory bug 17040 17041commit 60dac878cc8626b0821c4c7a6a6dfa12a5597cfa 17042Author: Alex Deucher <alexdeucher@gmail.com> 17043Date: Mon Mar 9 20:48:39 2009 -0400 17044 17045 R4xx: add R4xxATOM option 17046 17047 If set, enables the use of atombios for modesetting 17048 on r4xx cards. 17049 17050commit 758fcc33c232c32e7b66261fca51de161d1a4800 17051Author: Dave Airlie <airlied@redhat.com> 17052Date: Tue Mar 10 10:34:13 2009 +1000 17053 17054 r600: enable DRI by default 17055 17056 fallback to shadowfb if DRI fails pre init 17057 17058commit 7735e411954fd2cb6105c30ebd5a1b209974ce22 17059Author: Alex Deucher <alexdeucher@gmail.com> 17060Date: Mon Mar 9 17:49:25 2009 -0400 17061 17062 R6xx/r7xx: clarify EXA message 17063 17064commit 1d355798f3d421cabf2a3a0ba32a580483476ab0 17065Author: Alex Deucher <alexdeucher@gmail.com> 17066Date: Mon Mar 9 15:13:01 2009 -0400 17067 17068 AVIVO: fix panning 17069 17070 low 2 bits of X need to be 0. 17071 17072commit d1add18f3d238c755f9875b5e5c211e4af742482 17073Author: root <root@benzylpiperazine.boston.devel.redhat.com> 17074Date: Fri Mar 6 17:53:59 2009 -0500 17075 17076 output: Filter out dual-link modes from DP->DVI connections 17077 17078 There's not enough pins on a DP->DVI connector for this to possibly work. 17079 17080commit 04481bf450650d1ea5aadabc9213b899b0fe8119 17081Author: root <root@benzylpiperazine.boston.devel.redhat.com> 17082Date: Fri Mar 6 17:32:15 2009 -0500 17083 17084 atom: Enable DisplayPort source to DVI/HDMI sink 17085 17086 DP source to DP sink still doesn't work yet, of course. 17087 17088commit 0ba73e040b94590867f8b1071a26da2526a3c375 17089Author: Adam Jackson <ajax@redhat.com> 17090Date: Fri Mar 6 15:38:00 2009 -0500 17091 17092 output: Allow for multiple DisplayPort outputs. 17093 17094commit 8f7c17a6b77ca368f067c2acb803cc8f9ef19351 17095Author: Tormod Volden <debian.tormod@gmail.com> 17096Date: Fri Mar 6 09:42:39 2009 -0500 17097 17098 RV350: Add AGPMode quirk for Thinkpad 17099 17100 fixes bug 15096 (and Ubuntu bug #195051) 17101 17102commit 5dc4b69fbdd0e8ca48cc00321f193260a8f10bd0 17103Author: Alex Deucher <alexdeucher@gmail.com> 17104Date: Thu Mar 5 12:26:04 2009 -0500 17105 17106 radeon: add a few more default common modes for lvds 17107 17108commit 3b0fc22ad9e5c0f120a74c4f3d8e48c270f0ff29 17109Author: Alex Deucher <alexdeucher@gmail.com> 17110Date: Wed Mar 4 19:04:34 2009 -0500 17111 17112 R6xx/R7xx: wire up DMAForXv option like older asics 17113 17114 Select between SW and HW-assisted uploads 17115 17116commit 52f06ace04ad8141effc45fb6a0107a05bb46a73 17117Author: Alex Deucher <alexdeucher@gmail.com> 17118Date: Wed Mar 4 18:51:44 2009 -0500 17119 17120 R6xx/R7xx: disable XV_BICUBIC attribute 17121 17122 It's not implemented yet. 17123 17124commit b1fd883b59b85fed8782e035890098908902f4ce 17125Author: Alex Deucher <alexdeucher@gmail.com> 17126Date: Wed Mar 4 11:17:16 2009 -0500 17127 17128 RN50: fix up cloning on servers 17129 17130 Since the rn50 only has one crtc, but often has multiple outputs, 17131 occasionally one of the outputs doesn't get a crtc assigned eventhough 17132 the outputs may have common modes. If we see an output with modes, but 17133 no crtc, assign it so all the attached outputs light up. 17134 17135 fixes bugs 19457 and similar ones 17136 17137commit 592bcac52f113a95923a8f1cb8427e7552d5670b 17138Author: Alex Deucher <alexdeucher@gmail.com> 17139Date: Wed Mar 4 10:53:20 2009 -0500 17140 17141 ATOM: don't use fixed ref div for LVDS 17142 17143 The bios doesn't provide one. If there are problems 17144 with certain panels, we may need to adjust the PLL flags 17145 in certain cases. 17146 17147 Should fix bug 20458 17148 17149commit e88b0eac8cab8b54b769fd7b4ad3b3b65a90de09 17150Author: Alex Deucher <alexdeucher@gmail.com> 17151Date: Wed Mar 4 02:52:36 2009 -0500 17152 17153 radeon: just add some common modes for LVDS 17154 17155 avoids needing to hack around with the edid 17156 17157commit 2bb319fd80d20a3d4e5cc9416891bc7e4c3e431e 17158Author: Alex Deucher <alexdeucher@gmail.com> 17159Date: Wed Mar 4 02:50:45 2009 -0500 17160 17161 Revert "radeon: adjust LVDS so that default modes get added" 17162 17163 This reverts commit 1a2b16561d19ec9c027c562902f5fc086c856994. 17164 17165commit 1a2b16561d19ec9c027c562902f5fc086c856994 17166Author: Alex Deucher <alexdeucher@gmail.com> 17167Date: Tue Mar 3 20:21:17 2009 -0500 17168 17169 radeon: adjust LVDS so that default modes get added 17170 17171 we can scale, so add the default modes. for panels 17172 with an EDID, set the continous freq bit. for panels without 17173 and EDID, add a FAKE edid with the continous freq bit set. 17174 17175 Based on similar code in the Intel driver. 17176 17177commit 71117970df36cbe689ef15e9a6cca24439b4cd62 17178Author: Alex Deucher <alexdeucher@gmail.com> 17179Date: Tue Mar 3 19:40:30 2009 -0500 17180 17181 AVIVO: add aspect scaling mode 17182 17183 No luck yet for aspect on pre-avivo chips 17184 17185commit d586a2c6f821c821a4a7708a3382acb63187534f 17186Author: Alex Deucher <alexdeucher@gmail.com> 17187Date: Tue Mar 3 12:17:30 2009 -0500 17188 17189 Rotation: don't rotate if acceleration is not active 17190 17191commit fdc64948119e2218921b4dc0a492522442668875 17192Author: Alex Deucher <alexdeucher@gmail.com> 17193Date: Mon Mar 2 16:33:46 2009 -0500 17194 17195 R6xx/r7xx: remove some unneeded code I missed in the last commit 17196 17197commit 4ad1c4decfee653dbbc1ea2ca4270487be622382 17198Author: Alex Deucher <alexdeucher@gmail.com> 17199Date: Mon Mar 2 15:38:35 2009 -0500 17200 17201 R6xx/R7xx EXA: cleanup composite texture setup 17202 17203commit b7164ac4ad55e5d0fc474df8ae762b469b91ba30 17204Author: Alex Deucher <alexdeucher@gmail.com> 17205Date: Mon Mar 2 04:08:09 2009 -0500 17206 17207 R6xx/R7xx EXA: combine composite mask/non-mask VS 17208 17209 Also fix set_bool_const() 17210 the CF bool consts are not contiguous by shader type 17211 There are 96 boolean constants (32 each for PS, VS, GS) and 17212 they are ordered as follows: 17213 ps, vs, gs ... ps, vs, gs 17214 17215commit fa98f424de739be2c6005b740a74bbf1ee968a8b 17216Author: Alex Deucher <alexdeucher@gmail.com> 17217Date: Mon Mar 2 02:28:57 2009 -0500 17218 17219 R6xx/R7xx Xv: combine packed and planar shaders 17220 17221 use a bool const to select the tex fetch routine 17222 17223commit ccde35c3eda3fff0de29eb8c6fdc392629724a34 17224Author: Christian Koenig <deathsimple@vodafone.de> 17225Date: Sun Mar 1 23:38:37 2009 -0500 17226 17227 R6xx/R7xx: move shaders to r600_shader.c and fixup Xv PS 17228 17229 patches from Christian Koenig with some adjustments from me 17230 17231commit a8e631c1b1c9b46602aeca66f8e7e68154d0bfc8 17232Author: Alex Deucher <alexdeucher@gmail.com> 17233Date: Sun Mar 1 22:30:31 2009 -0500 17234 17235 R6xx/R7xx: code cleanups 17236 17237commit 528061d51bb4248d6fabec9579dead32a730467a 17238Author: Alex Deucher <alexdeucher@gmail.com> 17239Date: Sun Mar 1 18:28:13 2009 -0500 17240 17241 R6xx/R7xx: write vertexes directly to the IB 17242 17243 Reduces the vertex buffer setup overhead 17244 17245commit 96a0bafa16dbb67a3d10fa6fe45be13930495ddc 17246Author: Alex Deucher <alexdeucher@gmail.com> 17247Date: Sun Mar 1 17:58:36 2009 -0500 17248 17249 R6xx/R7xx: switch emit functions to macros 17250 17251 This improves performance due to decreased function call 17252 overhead. 17253 17254commit cbd579c41a16bddcbb8793eac2a41f2e958bef4d 17255Author: Alex Deucher <alexdeucher@gmail.com> 17256Date: Sun Mar 1 13:16:20 2009 -0500 17257 17258 Print a message when we have a shared DDC line 17259 17260commit a7f465f73363fce409870f62173d518b1bc02ae6 17261Author: Alex Deucher <alexdeucher@gmail.com> 17262Date: Sun Mar 1 12:39:31 2009 -0500 17263 17264 R300: Add AGP quirk 17265 17266 bug 7770 17267 17268commit 017bc7f34de2e40919f0e711205b78ee4e3643f6 17269Author: Tormod Volden <debian.tormod@gmail.com> 17270Date: Sat Feb 28 10:33:35 2009 -0500 17271 17272 M9+: Add AGP quirk for Sony Vaio 17273 17274commit 49b57767d0d2c041517b0764c2ed2d2ba5a7092c 17275Author: Bryce Harrington <bryce@bryceharrington.org> 17276Date: Fri Feb 27 20:31:03 2009 -0500 17277 17278 Quirk for RV280 on 82865G/PE/P DRAM Controller/Host-Hub 17279 17280 Solves issue where display suffers from line corruption. 17281 (See https://edge.launchpad.net/ubuntu/+bug/300304) 17282 17283commit 150abd4123664eab65ea43585dbdb9b3151c46af 17284Author: Tormod Volden <lists.tormod@gmail.com> 17285Date: Fri Feb 27 11:23:32 2009 -0500 17286 17287 Janitor: cosmetic clean-up of AGPMode quirk table 17288 17289commit c7d344f0ad255761d5ae30da59cf0a8b532dfc68 17290Author: Michel Dänzer <daenzer@vmware.com> 17291Date: Fri Feb 27 17:05:01 2009 +0100 17292 17293 EXA: Make sure Prepare/FinishAccess hooks can handle EXA_PREPARE_AUX* indices. 17294 17295 And signal this to the EXA core using the new EXA_SUPPORTS_PREPARE_AUX flag. 17296 17297commit f4d57105fc869e9318f447a1514f1b284a0a4ec0 17298Author: Michel Dänzer <daenzer@vmware.com> 17299Date: Fri Feb 27 09:23:24 2009 +0100 17300 17301 Revert "EXA: Adapt to EXA changes in xserver Git." 17302 17303 This reverts commit de358736dc696559ba99c71cf5b2a97508201630. 17304 17305commit 3285a4af17b19f2d1c24eebfb6aa51cddcf3ca50 17306Author: Alex Deucher <alexdeucher@gmail.com> 17307Date: Thu Feb 26 17:31:13 2009 -0500 17308 17309 DCE3.2: fix up Save()/Restore() 17310 17311commit 4fc978f3d83ef6370c9715e7ca5350640ccbfe10 17312Author: Alex Deucher <alexdeucher@gmail.com> 17313Date: Thu Feb 26 12:10:08 2009 -0500 17314 17315 R6xx/R7xx: use shadowfb if DRI is disabled 17316 17317commit 000756e052a291230e5c95e48b69a5aa9c4fab0e 17318Merge: 22d7746bf 8373f4399 17319Author: Alex Deucher <alexdeucher@gmail.com> 17320Date: Thu Feb 26 11:44:13 2009 -0500 17321 17322 Merge branch 'r6xx-r7xx-support' of git+ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati 17323 17324commit 22d7746bfc1bcbb3269afac88971254088ea9ece 17325Author: Alex Deucher <alexdeucher@gmail.com> 17326Date: Thu Feb 26 11:00:46 2009 -0500 17327 17328 RS600: enable the DRI by default 17329 17330commit 8373f4399b03961f2c928a9275d47e9f41bd92bb 17331Author: Alex Deucher <alexdeucher@gmail.com> 17332Date: Wed Feb 25 09:16:38 2009 -0500 17333 17334 R6xx/R7xx EXA: same surface and same coords equals nop 17335 17336 should fix bug 20305 17337 17338commit c74727015453ff3c3d6d06b812ebca9eb19a9767 17339Author: Alex Deucher <alexdeucher@gmail.com> 17340Date: Tue Feb 24 11:08:46 2009 -0500 17341 17342 R6xx/R7xx EXA: init copy_area to NULL 17343 17344commit 95ce13572dc2d9f5dd6cf55c23411e275c0aadf1 17345Author: Alex Deucher <alexdeucher@gmail.com> 17346Date: Tue Feb 24 10:51:13 2009 -0500 17347 17348 R6xx/R7xx EXA: Optimize temp surface for overlapping copies 17349 17350 - allocate temp surface in PrepareCopy() 17351 - fall back to old OverlapCopy() path if we are not able 17352 to allocate a temp surface 17353 17354commit 1a7db3fc2a0277d724d60d028064d8ef75019c28 17355Author: Mark van Doesburg <mark.vandoesburg@hetnet.nl> 17356Date: Tue Feb 24 10:44:19 2009 -0500 17357 17358 R6xx/R7xx EXA: use a temp surface for overlapping copy 17359 17360commit 957c2ba3459af152ad16a3f6dd4631044f9df6f9 17361Author: Alex Deucher <alexdeucher@gmail.com> 17362Date: Tue Feb 24 09:43:58 2009 -0500 17363 17364 RS600: fix MC addr mask 17365 17366 Noticed by osiris on IRC 17367 17368commit de358736dc696559ba99c71cf5b2a97508201630 17369Author: Michel Dänzer <daenzer@vmware.com> 17370Date: Tue Feb 24 08:30:23 2009 +0100 17371 17372 EXA: Adapt to EXA changes in xserver Git. 17373 17374 The PrepareAccess hook can now get six different indices. 17375 17376 Also remove superfluous instances of #include "exa.h". 17377 17378 Note that you may need to run autoheader manually to get config.h.in updated. 17379 17380commit 752b0218af36a45b333a12f66dfdce8c1e662aa8 17381Author: Michel Dänzer <daenzer@vmware.com> 17382Date: Tue Feb 24 08:13:06 2009 +0100 17383 17384 EXA: Pass pScrn and info into RadeonCompositeTile. 17385 17386 Saves some cycles looking them up. 17387 17388commit e2e508bc08d993fc108297e9cd2fa0ce183282bc 17389Author: Michel Dänzer <daenzer@vmware.com> 17390Date: Tue Feb 24 08:12:42 2009 +0100 17391 17392 Only call RADEONWaitForVLine if it might actually do anything useful. 17393 17394 Otherwise it and RADEONBiggerCrtcArea show up pretty high in x11perf -aa*text 17395 profiles. 17396 17397commit 5ab8699c97c050424cf813b9a60550aba32446a2 17398Author: Michel Dänzer <daenzer@vmware.com> 17399Date: Tue Feb 24 08:12:23 2009 +0100 17400 17401 Fix compile warning when building without EXA. 17402 17403commit 132e61741185f4a1114e66a34bdbf21e08e4d5d8 17404Author: Alex Deucher <alexdeucher@gmail.com> 17405Date: Mon Feb 23 12:13:41 2009 -0500 17406 17407 radeon: re-enable load detection output attribute for TV/CV 17408 17409commit 918d511f6f26d2fdeb63d6e32cd348707b4a2140 17410Author: Alex Deucher <alexdeucher@gmail.com> 17411Date: Mon Feb 23 10:22:19 2009 -0500 17412 17413 RS600: fix up MC setup 17414 17415 Please test if you have an RS600 17416 17417commit 27f8ca2cce65be2bcb3375231886d5444d251808 17418Author: Alex Deucher <alexdeucher@gmail.com> 17419Date: Thu Feb 19 21:18:07 2009 -0500 17420 17421 R6xx/R7xx: add wait for idle MMIO path 17422 17423commit 1f783117f79f95f1744681ccc7d80dc0e7b53057 17424Author: Alex Deucher <alexdeucher@gmail.com> 17425Date: Thu Feb 19 21:07:06 2009 -0500 17426 17427 RBBM_GUICNTL is pre-r6xx only 17428 17429commit f3933ed200dbf61b05a114f7697f2bb68d92abe7 17430Author: Alex Deucher <alexdeucher@gmail.com> 17431Date: Thu Feb 19 20:12:33 2009 -0500 17432 17433 Don't write new HDP location until we've written the new FB location 17434 17435commit 7f6801abb4bbd79830999d99ccdbd672a9984df6 17436Author: Alex Deucher <alexdeucher@gmail.com> 17437Date: Thu Feb 19 19:02:52 2009 -0500 17438 17439 radeon: one more 32 -> 64 just to be safe 17440 17441 missed this in cadae2a8d9979b2d61eae7d1f7a9ab8f5eb28e83 17442 17443commit ca22390c8a5256bbc8eeff775d5eebf0e892d890 17444Author: Alex Deucher <alexdeucher@gmail.com> 17445Date: Thu Feb 19 18:58:40 2009 -0500 17446 17447 R6xx/R7xx: fix up a few more paths 17448 17449commit cadae2a8d9979b2d61eae7d1f7a9ab8f5eb28e83 17450Author: Alex Deucher <alexdeucher@gmail.com> 17451Date: Thu Feb 19 18:50:49 2009 -0500 17452 17453 Fix MC setup on systems with more than 512 MB of VRAM 17454 17455 many thanks to app4des on IRC for helping me sort this out. 17456 17457commit 42492c70c481c88911337eefb97ba9d030adf7b6 17458Author: Alex Deucher <alexdeucher@gmail.com> 17459Date: Thu Feb 19 10:20:28 2009 -0500 17460 17461 R6xx/R7xx: wait for MC idle when changing the MC 17462 17463 should fix bug 19984 17464 17465commit e6475282486f4895bc68f6b093ecbb1aa6d25f72 17466Author: Alex Deucher <alexdeucher@gmail.com> 17467Date: Wed Feb 18 20:21:13 2009 -0500 17468 17469 R6xx/R7xx Xv: fix some missing bits from last commit 17470 17471commit adff8906c9899dde7711382577a63f4a726437ca 17472Author: Alex Deucher <alexdeucher@gmail.com> 17473Date: Wed Feb 18 20:01:57 2009 -0500 17474 17475 R6xx/R7xx EXA/Xv: properly deal with running out of vertex buffer space 17476 17477 As noted by mhopf, if VGT_MAX/MIN_INDX, etc. regs change, you need to re-emit 17478 CB blocks to avoid a hang. So, just set the VGT_MAX_INDX to a reasonably large value 17479 in the default state and don't touch them when drawing. When we run out of VB space, 17480 just draw the current buffer, grab a new one, and continue. 17481 17482commit cf85d4a1d43a3209c7ca9307aede2c2c243f7130 17483Author: Alex Deucher <alexdeucher@gmail.com> 17484Date: Wed Feb 18 19:02:56 2009 -0500 17485 17486 R6xx/R7xx: reset 3D state after VT switch 17487 17488commit 6545f0a2967414fa518a78440060b8b69c0146ee 17489Author: Alex Deucher <alexdeucher@gmail.com> 17490Date: Wed Feb 18 18:15:18 2009 -0500 17491 17492 R6xx/R7xx: fixup accel paths 17493 17494commit 61311b1de019c5a2344928a47347a966c229b08a 17495Author: Alex Deucher <alexdeucher@gmail.com> 17496Date: Wed Feb 18 17:27:05 2009 -0500 17497 17498 Bump version post release 17499 17500commit 0d3d1f1f7b450dabd3ad7d2df26ad1c72709b29b 17501Author: Alex Deucher <alexdeucher@gmail.com> 17502Date: Wed Feb 18 17:24:57 2009 -0500 17503 17504 bump for release 17505 17506commit 76a32f8391317513538dafee49cbb61c0d756356 17507Author: Alex Deucher <alexdeucher@gmail.com> 17508Date: Wed Feb 18 16:31:44 2009 -0500 17509 17510 R6xx: Connector quirk for asus board 17511 17512 bug 19943 17513 17514commit 97b8482dba4e99088b59d3ab1396be98a70e823e 17515Author: Alex Deucher <alexdeucher@gmail.com> 17516Date: Wed Feb 18 13:27:15 2009 -0500 17517 17518 Set default RMX type to FULL on LVDS 17519 17520commit 7d22b1799b34010d34c3600d6cb02ffe839a0780 17521Author: Alex Deucher <alexdeucher@gmail.com> 17522Date: Wed Feb 18 12:55:40 2009 -0500 17523 17524 DCE3.2+: allow output cloning 17525 17526commit 1a237a40958c006c56b80850bd77b2ac6c17e030 17527Author: Alex Deucher <alexdeucher@gmail.com> 17528Date: Wed Feb 18 11:46:20 2009 -0500 17529 17530 ATOM: reset crtc initialized flag on CloseScreen() 17531 17532 additional fix needed for bug 16781 17533 17534commit 9a108f0a0b7203458673ce6221e747a166d39617 17535Author: Alex Deucher <alexdeucher@gmail.com> 17536Date: Tue Feb 17 19:52:27 2009 -0500 17537 17538 ATOM: don't unblank uninitialized crtcs 17539 17540 If the crtc timing isn't setup, you might get stuck in a loop 17541 in the BlankCRTC command table 17542 17543 fixes bug 16781 17544 17545commit 3dff20e276615e8b77177689a4a5f8d91b3e8eac 17546Author: Alex Deucher <alexdeucher@gmail.com> 17547Date: Tue Feb 17 13:14:38 2009 -0500 17548 17549 R6xx/R7xx EXA: switch to surface sync packet 17550 17551commit 31bd6d28dd35021e0e706d4f094989deb856c26a 17552Author: Alex Deucher <alexdeucher@gmail.com> 17553Date: Tue Feb 17 11:24:02 2009 -0500 17554 17555 Fix crtc routing on pre-DCE3.2 systems 17556 17557 This should fix peterz and bug 20074 17558 17559commit 3c9e00c7f2ed494976713e8f77b729e0d76409db 17560Author: Alex Deucher <alexdeucher@gmail.com> 17561Date: Mon Feb 16 16:18:42 2009 -0500 17562 17563 RV350: add another AGP quirk 17564 17565commit 036a17bec5dabbb91640b907811da060dac154d6 17566Author: Alex Deucher <alexdeucher@gmail.com> 17567Date: Mon Feb 16 13:10:45 2009 -0500 17568 17569 RV280: add another agp quirk 17570 17571commit 00ac9d2f8d58725d0d3e6b1bf7f728d9ae7e6109 17572Author: David Miller <davem@davemloft.net> 17573Date: Sun Feb 15 18:26:14 2009 +1000 17574 17575 GART: Save/restore GART table consistently. 17576 17577 Always save the GART table with the RADEON_SURFACE_CNTL register 17578 zero'd out to make sure we always use the same endinanness. 17579 17580 fixed up as per Michel's suggestion for endianness. 17581 17582 Signed-off-by: David S. Miller <davem@davemloft.net> 17583 Signed-off-by: Dave Airlie <airlied@linux.ie> 17584 17585commit 2de90c8349409eb892edc57fd24b61b679eb7719 17586Author: David Miller <davem@davemloft.net> 17587Date: Sun Feb 15 18:24:01 2009 +1000 17588 17589 DRI: Fix page size used in RADEONDRIGetPciAperTableSize(). 17590 17591 The ATI GART has a fixed size of 4096 bytes. So using 17592 the system page size here is wrong. 17593 17594 Signed-off-by: David S. Miller <davem@davemloft.net> 17595 Signed-off-by: Dave Airlie <airlied@linux.ie> 17596 17597commit 7cde00f49649e25fd5816927c7a5e28b608fabcd 17598Author: Alex Deucher <alexdeucher@gmail.com> 17599Date: Thu Feb 12 17:39:44 2009 -0500 17600 17601 R6xx/R7xx: switch to drm for wait for idle 17602 17603 THIS REQUIRES AN UPDATED DRM 17604 17605commit e22cd4011b9be437ba89bff568e7fb82b4907d99 17606Author: Yang Zhao <yang@yangman.ca> 17607Date: Thu Feb 12 14:46:53 2009 -0500 17608 17609 R6xx/R7xx EXA: Further optimizations to overlapping copy 17610 17611 Diagonal overlapping copies can be reduced to either horizontal- or 17612 vertical-only offset, and the one with fewer copies is picked. 17613 17614commit da08b760bcf3d04d775c4440fafec10657bb1863 17615Author: Alex Deucher <alexdeucher@gmail.com> 17616Date: Thu Feb 12 13:53:11 2009 -0500 17617 17618 R6xx/R7xx EXA: handle running out of vertex space in the copy path 17619 17620commit e3be312b0b73982c24f1f5d9cf76d7caafae0853 17621Author: Alex Deucher <alexdeucher@gmail.com> 17622Date: Thu Feb 12 13:48:36 2009 -0500 17623 17624 R6xx/R7xx EXA: properly handle non repeat cases in the texture setup 17625 17626commit 3ff1eb7e50fef22adb6280dd7de77c3ecafdb451 17627Author: Michel Dänzer <daenzer@vmware.com> 17628Date: Thu Feb 12 19:24:41 2009 +0100 17629 17630 EXA: If making a pixmap offscreen fails, return ~0ULL as texture offset. 17631 17632 This allows AIGLX to fall back to the non-zero-copy code rather than messing up 17633 the 3D driver. 17634 17635commit c19aa4fc8da7ac4745624098b146bcc42c0436dd 17636Author: Michel Dänzer <daenzer@vmware.com> 17637Date: Thu Feb 12 19:22:12 2009 +0100 17638 17639 EXA: The source tiling code can't handle RepeatReflect yet. 17640 17641commit adc35636aa1e936ec74cba8931ccffe2b5c11656 17642Author: Alex Deucher <alexdeucher@gmail.com> 17643Date: Thu Feb 12 13:10:53 2009 -0500 17644 17645 R1xx/R2xx EXA: fix non repeat texture setup 17646 17647commit c6e1bf75c399807716b99486ab3301aeba60fa79 17648Author: Alex Deucher <alexdeucher@gmail.com> 17649Date: Thu Feb 12 12:49:24 2009 -0500 17650 17651 R3xx-R5xx EXA: fix texture setup for non-repeat case 17652 17653 fixes bug 19923 17654 17655commit 0804dde9012cf33f614a1c2ee6d5d0def040a9af 17656Author: Alex Deucher <alexdeucher@gmail.com> 17657Date: Thu Feb 12 12:02:57 2009 -0500 17658 17659 AVIVO: fix dualhead/rotation for real 17660 17661commit 3e5ac32f747bef70903e9cdf41652a827eece707 17662Author: Alex Deucher <alexdeucher@gmail.com> 17663Date: Wed Feb 11 17:40:20 2009 -0500 17664 17665 Revert "R6xx/R7xx: handle running out of vertex buffer space" 17666 17667 This reverts commit 4fd7228de7ad88edf825dbc3039df877795a9479. 17668 17669 Causes lockups. 17670 17671commit bd141aa73a77f68301715fb3b5664e2082327a80 17672Author: Alex Deucher <alexdeucher@gmail.com> 17673Date: Wed Feb 11 15:31:47 2009 -0500 17674 17675 R6xx/R7xx Xv: switch packed over to Yang's new shader code 17676 17677commit cb4a1ceaa32d6847b146a31291772e1183972ee7 17678Author: Yang Zhao <yang@yangman.ca> 17679Date: Wed Feb 11 15:18:00 2009 -0500 17680 17681 R6xx/R7xx Xv: Planar - Properly scale Y'CbCr values before converting to RGB 17682 17683 According to MPEG-2 spec, Y' and Cb/Cr values are scaled to [16, 235] 17684 and [16, 240], respectively, when packed into bytes. Properly take care 17685 of the reverse scaling before translating to RGB. 17686 17687 Conversion matrix has been simplified to remove 3rd column, as the fitting 17688 to [-0.5, 0.5] can be done with scaling. 17689 17690 Redundant MOV instructions were also removed, and now only 3 GPRs are required. 17691 17692commit 4fd7228de7ad88edf825dbc3039df877795a9479 17693Author: Alex Deucher <alexdeucher@gmail.com> 17694Date: Wed Feb 11 14:46:02 2009 -0500 17695 17696 R6xx/R7xx: handle running out of vertex buffer space 17697 17698 draw current VB and start a new one. 17699 17700commit ec60ef094762901ede4df5cde55c3f162e8c667b 17701Author: Alex Deucher <alexdeucher@gmail.com> 17702Date: Wed Feb 11 13:05:27 2009 -0500 17703 17704 R6xx/R7xx: be more verbose about what function ran out of VB space 17705 17706commit e7dedbc355970407eddb38370b58a8c96d204c1d 17707Author: Alex Deucher <alexdeucher@gmail.com> 17708Date: Wed Feb 11 12:45:14 2009 -0500 17709 17710 R6xx/R7xx: Move engine idle to sync functions 17711 17712commit 4a759a907ecd571460c7ff64cd6288380359b04b 17713Author: Alex Deucher <alexdeucher@gmail.com> 17714Date: Wed Feb 11 12:29:50 2009 -0500 17715 17716 adjust alignment 17717 17718commit d98359b20e1eb8eae893581ab199c299c8352dc4 17719Author: Alex Deucher <alexdeucher@gmail.com> 17720Date: Wed Feb 11 12:03:56 2009 -0500 17721 17722 Revert "Radeon EXA: wait for the engine to be idle before sw access" 17723 17724 This reverts commit 0c4694c58f2ec152b19023626bc7ced891d9da65. 17725 17726 This patch just adds latency and thus isn't a real solution 17727 17728commit d6c50b221cbab6e726948e2310c1def9fc38da64 17729Author: Alex Deucher <alexdeucher@gmail.com> 17730Date: Wed Feb 11 11:05:02 2009 -0500 17731 17732 R6xx/R7xx Xv: add support for packed uploads 17733 17734commit 8e437e996cc3f2c424c342701f4aa6bcf72ad08e 17735Author: Alex Deucher <alexdeucher@gmail.com> 17736Date: Wed Feb 11 10:53:50 2009 -0500 17737 17738 R6xx/R7xx Xv: Add native support for packed formats 17739 17740commit 0c4694c58f2ec152b19023626bc7ced891d9da65 17741Author: Alex Deucher <alexdeucher@gmail.com> 17742Date: Tue Feb 10 17:08:37 2009 -0500 17743 17744 Radeon EXA: wait for the engine to be idle before sw access 17745 17746 Really we need proper fencing with timestamps but this should 17747 help in the short term. 17748 17749 This should help with minor corruption issues and possibly 17750 lockups. 17751 17752commit 13ab61d90bb58112bef4ce7267ef127935315f80 17753Author: Fabio <fabio.ped@libero.it> 17754Date: Tue Feb 10 10:50:30 2009 -0500 17755 17756 man page updates 17757 17758 bug 20030 17759 17760commit 615f2f202cee5f0dd2c6b29329f843bb1c9c70d5 17761Author: Alex Deucher <alexdeucher@gmail.com> 17762Date: Tue Feb 10 10:17:47 2009 -0500 17763 17764 Bump version post RC release 17765 17766commit 1b4afc1c9f8458f9ab3434418f4ccf959f532ac1 17767Author: Alex Deucher <alexdeucher@gmail.com> 17768Date: Mon Feb 9 19:50:37 2009 -0500 17769 17770 R6xx/R7xx: Add checks to make sure we don't overrun VB space 17771 17772commit ff6efe513e0c3148812d6a9578d2265522417053 17773Author: Alex Deucher <alexdeucher@gmail.com> 17774Date: Mon Feb 9 17:43:24 2009 -0500 17775 17776 ATOM: warning fixes 17777 17778commit 87ad9006f3dafb45e131c9a3d6df713321f39998 17779Author: Alex Deucher <alexdeucher@gmail.com> 17780Date: Mon Feb 9 17:40:41 2009 -0500 17781 17782 RV350: add AGP quirk 17783 17784 bug 19981 17785 17786commit 610b8b65fd7aa97f9c03b2ff6ae607aa959eab0b 17787Author: Alex Deucher <alexdeucher@gmail.com> 17788Date: Mon Feb 9 16:33:20 2009 -0500 17789 17790 Bump for rc release 17791 17792commit e85b7a1c5948b5123eddf9dfbb9e002d74fbe9fb 17793Author: Alex Deucher <alexdeucher@gmail.com> 17794Date: Mon Feb 9 14:13:16 2009 -0500 17795 17796 R6xx/R7xx Xv: fix cache flush buffer size for planar 17797 17798commit 2a893bac1faffd28dce6d9a9693108196c71e6d3 17799Author: Alex Deucher <alexdeucher@gmail.com> 17800Date: Mon Feb 9 13:18:17 2009 -0500 17801 17802 R6xx/R7xx Xv: add accelerated uploads for planar formats 17803 17804commit 231aee18a73805be2f6c962e94a8345dd89fd0df 17805Author: Alex Deucher <alexdeucher@gmail.com> 17806Date: Mon Feb 9 13:02:27 2009 -0500 17807 17808 R6xx/R7xx Xv: implement native shader for planar formats 17809 17810commit 6c76bfe8105e3cf4e7e6ea1bfe1235be2079110f 17811Author: Alex Deucher <alexdeucher@gmail.com> 17812Date: Mon Feb 9 09:47:22 2009 -0500 17813 17814 R6xx/R7xx UTS: move actual upload to separate function 17815 17816 So it can be shared with Xv 17817 17818commit 04859362ce64ed4ae48305a83c92059899c1db4b 17819Author: Alex Deucher <alexdeucher@gmail.com> 17820Date: Sun Feb 8 20:13:06 2009 -0500 17821 17822 Add some missing r6xx/r7xx pci ids 17823 17824commit 132e4c575dc4675f4995e45f08c53c26bffd999a 17825Author: Yang Zhao <yang@yangman.ca> 17826Date: Sat Feb 7 13:54:51 2009 -0500 17827 17828 R6xx/R7xx EXA: Optimize overlapping copy 17829 17830 Overlapping copy is now done in chunks proportional to the 17831 non-overlapping area. 17832 17833 Diagonal overlaps are also handled properly. 17834 17835commit 8e9ef8ff581892cbe1b7ea56d48b9a1abd70179d 17836Author: Pierre Ossman <pierre@ossman.eu> 17837Date: Sat Feb 7 18:57:47 2009 +0100 17838 17839 Xv vsync support on r6xx/r7xx cards. 17840 17841commit 2222f0fd700f100b2e91fac2babe7d1b53f56c3e 17842Author: Pierre Ossman <pierre@ossman.eu> 17843Date: Sat Feb 7 18:56:42 2009 +0100 17844 17845 Fix bad range adjustment in VLINE code. 17846 17847commit 8ce6c024e31f6a3f5ae6c882738b1e64ae2944a1 17848Author: Alex Deucher <alexdeucher@gmail.com> 17849Date: Sat Feb 7 10:32:01 2009 -0500 17850 17851 R6xx/R7xx Xv: fix typos in cache flushing commands 17852 17853commit bba5d806cc46328ad9e4c521bfeeb2a59f96862b 17854Author: Alex Deucher <alexdeucher@gmail.com> 17855Date: Fri Feb 6 19:44:17 2009 -0500 17856 17857 AVIVO: better fix for rotation 17858 17859 This should handle initial rotation as well. 17860 17861commit 28cf3492f431f325cb08d7c6aade43d8886df3f6 17862Author: Alex Deucher <alexdeucher@gmail.com> 17863Date: Fri Feb 6 19:20:18 2009 -0500 17864 17865 AVIVO: fix rotation 17866 17867 When rotation is active, virtualX and virtualY change. 17868 17869commit 1b45936ae614244aa49b1a5d3c7fc39773c4f9b6 17870Author: Alex Deucher <alexdeucher@gmail.com> 17871Date: Fri Feb 6 18:29:55 2009 -0500 17872 17873 R6xx/R7xx EXA: Fix typo in DFS 17874 17875 noticed by pzad in IRC 17876 17877commit c06d89e16d5b2553142e8641e66080e1770c1563 17878Author: Alex Deucher <alexdeucher@gmail.com> 17879Date: Fri Feb 6 18:07:22 2009 -0500 17880 17881 R6xx/R7xx EXA: fallback on overlapping blits for now 17882 17883 Leave this disabled until we get a proper solution. 17884 17885commit 1d5fc3febf3470b94c423a1eda5e0683856909df 17886Author: Alex Deucher <alexdeucher@gmail.com> 17887Date: Fri Feb 6 14:43:30 2009 -0500 17888 17889 Revert "r6xx/r7xx EXA: Optimize overlapping copy" 17890 17891 This reverts commit 0dfadc1843e0d14b9cc1ee19a72f4fd60a2c495b. 17892 17893 This doesn't always work properly. 17894 17895commit def317e22d072405cd95ddb19d17eacd784ffd9e 17896Author: Alex Deucher <alexdeucher@gmail.com> 17897Date: Fri Feb 6 11:04:59 2009 -0500 17898 17899 R6xx/R7xx EXA: add accelerated UTS/DFS hooks 17900 17901 I'm not sure how much of a win these are. I need to 17902 do some benchmarking. 17903 17904commit 0dfadc1843e0d14b9cc1ee19a72f4fd60a2c495b 17905Author: Yang Zhao <yang@yangman.ca> 17906Date: Fri Feb 6 10:29:39 2009 -0500 17907 17908 r6xx/r7xx EXA: Optimize overlapping copy 17909 17910 When source and destination blocks are only offset horizontally, it 17911 appears to be unnecessary to perform careful, segment-by-segment copy. 17912 The code path that does this is taken out completely. 17913 17914 For the case where offset is only vertical, copying is now done by 17915 height of the non-overlapping area each time, instead of always 17916 line-by-line. 17917 17918commit 3d17bd199423e92d201c20f047d5e699942af976 17919Author: Alex Deucher <alexdeucher@gmail.com> 17920Date: Thu Feb 5 19:25:03 2009 -0500 17921 17922 Revert "R6xx/R7xx EXA: improve overlapping copy performance" 17923 17924 This reverts commit b24827c9d211e8a35da53b665385a7733d19910d. 17925 17926 This seems to cause corruption in some cases. 17927 17928commit bf74055f543e7f0664741620fb1fe827ebc12711 17929Author: Alex Deucher <alexdeucher@gmail.com> 17930Date: Thu Feb 5 17:31:43 2009 -0500 17931 17932 r6xx/r7xx EXA: fix corruption when doing sw access 17933 17934 need to wait until the engine is idle. Ideally we wait 17935 on a timestamp shadowed in memory, but polling the 17936 GRBM_STATUS reg will do for now. 17937 17938commit b24827c9d211e8a35da53b665385a7733d19910d 17939Author: Alex Deucher <alexdeucher@gmail.com> 17940Date: Thu Feb 5 16:21:37 2009 -0500 17941 17942 R6xx/R7xx EXA: improve overlapping copy performance 17943 17944 send vertices for each line of the copy, but only draw once 17945 17946commit 729fe756f809a41494dc161252d82313956e4a4b 17947Author: Alex Deucher <alexdeucher@gmail.com> 17948Date: Thu Feb 5 16:08:35 2009 -0500 17949 17950 r6xx/r7xx EXA: cleanup overlapping copy 17951 17952commit 16e01a5796a8e8b86ad25d3aa45b9e7044dc72d1 17953Author: Alex Deucher <alexdeucher@gmail.com> 17954Date: Tue Feb 3 19:46:10 2009 -0500 17955 17956 Fix encoder accounting 17957 17958 Should fix bug 19924 17959 17960commit d7bf7b9c17add31dc9a115b56a563c7f9bf2cdb6 17961Author: Alex Deucher <alexdeucher@gmail.com> 17962Date: Tue Feb 3 18:28:38 2009 -0500 17963 17964 EXA: fix and re-enable Solid() on R7xx 17965 17966 missing last bit in alu clause in solid PS 17967 17968commit 9abb09d5f43320bd27b336612d39d264fe93549f 17969Author: Alex Deucher <alexdeucher@gmail.com> 17970Date: Tue Feb 3 11:17:13 2009 -0500 17971 17972 Fix bad rv710 pci id 17973 17974commit e5b916770946a9eebcb4bd1e6f698220db8c718a 17975Author: Alex Deucher <alexdeucher@gmail.com> 17976Date: Tue Feb 3 10:48:41 2009 -0500 17977 17978 Allow rotation on r6xx/r7xx 17979 17980commit d1f071c7f1dad6babfbcfcc2cb2b722a4987f372 17981Author: Alex Deucher <alexdeucher@gmail.com> 17982Date: Tue Feb 3 10:44:10 2009 -0500 17983 17984 Initial R6xx/R7xx EXA and textured video support 17985 17986commit c88c3ef6f3db266c1aacba5297b8dfc8b66bf00e 17987Author: Wolke Liu <wolke.liu@amd.com> 17988Date: Mon Feb 2 17:01:34 2009 -0500 17989 17990 AVIVO: Save/restore vga pll registers 17991 17992 This fixes some VT switch issues on some chips 17993 17994commit 6fac3cefd1f46161c1e276ba40e72da2823aa9f6 17995Author: Alex Deucher <alexdeucher@gmail.com> 17996Date: Sat Jan 31 14:21:02 2009 -0500 17997 17998 Return NULL for encoder if no active device is assigned 17999 18000 fixes bug 19855 18001 18002commit db178c852d00e9b85513d750b5f8f7495f9e3360 18003Author: Alan Coopersmith <alan.coopersmith@sun.com> 18004Date: Fri Jan 30 20:37:41 2009 -0800 18005 18006 Add README with pointers to mailing list, bugzilla & git repos 18007 18008commit fa8e5a4fc236f8f15f462cb0d6164b194a65a118 18009Author: Michel Dänzer <daenzer@vmware.com> 18010Date: Fri Jan 30 16:16:26 2009 +0100 18011 18012 EXA: Accelerate Composite of RepeatPad/Reflect pictures when possible. 18013 18014commit 3a6e958c59b4b44fe620c8abe760ed1abf0007b1 18015Author: airlied <airlied@roo.(none)> 18016Date: Fri Jan 30 07:56:14 2009 +1000 18017 18018 rs780: include RS780 in the InitMemory to leave alone 18019 18020 This stops the MC_FB_LOCATION getting reassigned to 0, which 18021 is really bad thing to happen. 18022 18023 I've had reported memory corruption on these cards so hopefully this 18024 fixes it. 18025 18026commit bd8021d46e9066e4cd116c03a7b7adcfe2557aff 18027Merge: 4f88dd83f 20d5dd387 18028Author: Alex Deucher <alexdeucher@gmail.com> 18029Date: Wed Jan 28 14:33:14 2009 -0500 18030 18031 Merge branch 'atom-rework' and fix conflicts 18032 18033commit 20d5dd387da555e895e2b73fb53e2b026dd91003 18034Author: Alex Deucher <alexdeucher@gmail.com> 18035Date: Wed Jan 28 14:12:26 2009 -0500 18036 18037 Move encoder specific data to encoder dev_priv 18038 18039commit 4f88dd83f290fb38e41256a7f1804008c0f28139 18040Merge: 1c5678fc4 2ca970c7a 18041Author: Alex Deucher <alexdeucher@gmail.com> 18042Date: Wed Jan 28 13:13:13 2009 -0500 18043 18044 Merge branch 'atom-tvout' of ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati 18045 18046 Conflicts: 18047 18048 src/radeon_atombios.c 18049 18050commit 1a62360e8260eaf5f98fa6be3f89f64c8cae7d32 18051Author: Alex Deucher <alexdeucher@gmail.com> 18052Date: Mon Jan 26 10:11:28 2009 -0500 18053 18054 Encoders not assigned yet, use supported devices 18055 18056 need to move this to encoder setup 18057 18058commit 44d9ceacb1fa6e1859d5d8bd1c55405aea24ad85 18059Author: Alex Deucher <alexdeucher@gmail.com> 18060Date: Mon Jan 26 03:47:32 2009 -0500 18061 18062 Fix legacy output setup 18063 18064 don't have assigned encoders at this point. Need to 18065 eventually move this stuff to encoder dev_priv. 18066 18067commit 0c24cbcea56c8f36374aca0f8ba849adfbf0ecfe 18068Author: Alex Deucher <alexdeucher@gmail.com> 18069Date: Mon Jan 26 03:24:19 2009 -0500 18070 18071 ATOM: print useful output info for DPMS events 18072 18073commit 2f92b885457c5cee2098ebbf8c6fe5a5eb98ec48 18074Author: Alex Deucher <alexdeucher@gmail.com> 18075Date: Mon Jan 26 03:12:44 2009 -0500 18076 18077 Warning fixes 18078 18079commit 4cbf6189f3a7945a1141b0451e4f7881ceaaadcd 18080Author: Alex Deucher <alexdeucher@gmail.com> 18081Date: Mon Jan 26 02:59:51 2009 -0500 18082 18083 Move legacy output setup functions to legacy_output.c 18084 18085commit 93ed767255b60fbdf5b416b9bd06c366036a5141 18086Author: Alex Deucher <alexdeucher@gmail.com> 18087Date: Mon Jan 26 02:40:41 2009 -0500 18088 18089 Fix off by one when printing encoder name 18090 18091 - also cleanup some debugging output 18092 18093commit b515eee4b1cafae78ed39cd91c18b4f5e671b20b 18094Author: Alex Deucher <alexdeucher@gmail.com> 18095Date: Mon Jan 26 01:50:18 2009 -0500 18096 18097 Additional output cleanup 18098 18099commit 1c5678fc452dbc93135e9ece50a4cff6583f6078 18100Author: Michel Dänzer <daenzer@vmware.com> 18101Date: Sat Jan 24 11:10:00 2009 +0100 18102 18103 Drop memcpy fallbacks from EXA UploadToScreen and DownloadFromScreen hooks. 18104 18105 While in theory it's possible for the PrepareAccess hook to fail on big endian 18106 platforms, in practice it isn't at this point because there's no other users of 18107 the surface registers. 18108 18109commit 5cbd6082b9f8d7009ee9ef5e036673f99e1bfb05 18110Author: Michel Dänzer <daenzer@vmware.com> 18111Date: Sat Jan 24 11:07:12 2009 +0100 18112 18113 Don't transform EXA Composite mask coordinates when there's no mask. 18114 18115commit 7e54c4b8acafead8de5d0dd6fdc1b619f832dd6f 18116Author: Thomas Jaeger <ThJaeger@gmail.com> 18117Date: Fri Jan 23 20:03:27 2009 -0500 18118 18119 Fall back to software for unsupported repeat modes 18120 18121commit 289b4a377cfff0c2a8ec938951cfbae1e270570e 18122Author: Alex Deucher <alexdeucher@gmail.com> 18123Date: Fri Jan 23 20:00:43 2009 -0500 18124 18125 Remove OutputType and other cruft 18126 18127commit 5f98508effbdc1fd62d0df15798722c147d88da1 18128Author: Maciej Cencora <m.cencora@gmail.com> 18129Date: Fri Jan 23 19:33:25 2009 -0500 18130 18131 Make sure gb_num_pipes is initialized when DRI is disabled 18132 18133 Found with valgrind 18134 18135commit 0faab02cc732414af3cfb2014d3ddc1139aff063 18136Author: Alex Deucher <alexdeucher@gmail.com> 18137Date: Fri Jan 23 17:35:01 2009 -0500 18138 18139 Remove some unused cruft 18140 18141commit 025082160c53c01a44f4d681cc63bf26ac4561d5 18142Author: Alex Deucher <alexdeucher@gmail.com> 18143Date: Fri Jan 23 17:31:18 2009 -0500 18144 18145 track encoder state 18146 18147commit 82f12e5a40c1fbcb91910a0f8b725c34fff02aae 18148Author: Alex Deucher <alexdeucher@gmail.com> 18149Date: Fri Jan 23 16:21:33 2009 -0500 18150 18151 Remove TMDSType, DACType, LVDSType from output rec 18152 18153commit 7b42b57d748f77ff6ef6f8ade7cbc983a98d7204 18154Author: Alex Deucher <alexdeucher@gmail.com> 18155Date: Wed Jan 21 16:28:00 2009 -0500 18156 18157 few more logic pasto's bits I missed 18158 18159commit fef1c12fbe4ceb4676258850d1fa5cac994f7c21 18160Author: Alex Deucher <alexdeucher@gmail.com> 18161Date: Wed Jan 21 15:34:01 2009 -0500 18162 18163 Fix compilation with RADEON_TRACE_FALL set 18164 18165 noticed by curaga on phoronix forum 18166 18167commit 1305376b009d17d65b4935896dc35cb7530f7c52 18168Author: Alex Deucher <alexdeucher@gmail.com> 18169Date: Wed Jan 21 15:20:47 2009 -0500 18170 18171 Move active_device setup to detect() 18172 18173 Also add encoder debugging printouts 18174 18175commit 9e0cc0ca5d7bc68756ae92ee3b5b2392366be71a 18176Author: Alex Deucher <alexdeucher@gmail.com> 18177Date: Wed Jan 21 14:47:08 2009 -0500 18178 18179 Fix logic cut and paste error 18180 18181commit 12010ae9296635ffa5f42ce0cd14482aaa290cf7 18182Author: Alex Deucher <alexdeucher@gmail.com> 18183Date: Wed Jan 21 13:52:11 2009 -0500 18184 18185 More legacy rework 18186 18187commit 3f8bbede8bf639050f773780888b89cbdcb425b8 18188Author: Alex Deucher <alexdeucher@gmail.com> 18189Date: Wed Jan 21 13:45:11 2009 -0500 18190 18191 fix legacy crtc routing and add some debugging info 18192 18193commit 80e0162ea2c1f793964731268251aba2d3243127 18194Author: Alex Deucher <alexdeucher@gmail.com> 18195Date: Wed Jan 21 13:28:14 2009 -0500 18196 18197 ATOM: fix encoder init 18198 18199commit aea27d9caabb0e54e868a21e6e547733a4ca2709 18200Author: Alex Deucher <alexdeucher@gmail.com> 18201Date: Wed Jan 21 12:33:49 2009 -0500 18202 18203 Switch legacy output code to use new encoder objects 18204 18205commit 9b44a4b57b00b150d041a5d82cb2df42159c062b 18206Author: Alex Deucher <alexdeucher@gmail.com> 18207Date: Wed Jan 21 11:32:24 2009 -0500 18208 18209 ATOM: more output cleanup 18210 18211commit 9e283fa9de34407fa6f0d5a8006a611b104f4bf2 18212Author: Alex Deucher <alexdeucher@gmail.com> 18213Date: Wed Jan 21 03:16:20 2009 -0500 18214 18215 Fixup encoder setup on pre-ATOM chips 18216 18217commit 5850abbdcd22b70bdbccd3642f8ba1a27e38e7cd 18218Author: Alex Deucher <alexdeucher@gmail.com> 18219Date: Wed Jan 21 02:28:47 2009 -0500 18220 18221 clean up encoder setup 18222 18223 - make sure not to duplicate encoders 18224 - use single encoder array 18225 18226commit 152d1d4c179c262be4d0a5618400aa13b4820342 18227Author: Alex Deucher <alexdeucher@gmail.com> 18228Date: Tue Jan 20 00:58:18 2009 -0500 18229 18230 First pass at converting legacy code to encoder objects 18231 18232commit 7cc7ba8bf9ce2f0eebfd9b48aa648c416c333ff4 18233Author: Alex Deucher <alexdeucher@gmail.com> 18234Date: Mon Jan 19 18:42:12 2009 -0500 18235 18236 ATOM: round 1 of output rework 18237 18238commit 25f3878858e6efb486888e5271b60d4ba08c73c6 18239Author: Alex Deucher <alexdeucher@gmail.com> 18240Date: Mon Jan 19 15:54:42 2009 -0500 18241 18242 start to re-org outputs 18243 18244commit cbeb310d35e5ea25ba7a45a2c555468391cdf29e 18245Author: Alex Deucher <alexdeucher@gmail.com> 18246Date: Fri Jan 16 19:53:49 2009 -0500 18247 18248 ATOM: switch to define for external tmds 18249 18250commit d0d157abced43f822885fa26f14a4924766594cd 18251Author: Alex Deucher <alexdeucher@gmail.com> 18252Date: Fri Jan 16 19:51:07 2009 -0500 18253 18254 ATOM: combine DAC setup functions 18255 18256 plus a few other clean-ups 18257 18258commit f6f969338f8187529c7ae9b2adc980b71a3f5837 18259Author: Alex Deucher <alexdeucher@gmail.com> 18260Date: Fri Jan 16 18:10:34 2009 -0500 18261 18262 DCE30: LVTMA requires DIG2 encoder 18263 18264 Fixes bug 19544 18265 18266commit 549ae63aecc2624a69b41ce4db75883d03d59c23 18267Author: Alex Deucher <alexdeucher@gmail.com> 18268Date: Thu Jan 15 19:26:40 2009 -0500 18269 18270 RV280 Add another AGP quirk 18271 18272commit 0e69068a8ed82e3fb7349d8224933451955e5f16 18273Author: Tormod Volden <debian.tormod@gmail.com> 18274Date: Tue Jan 13 20:31:17 2009 -0500 18275 18276 [PATCH] Add yet another AGP quirk for RV280 18277 18278commit b437da34cb4f7c637674fb432ad81b9bf32244a3 18279Author: Christiaan van Dijk <dvbmail@xs4all.nl> 18280Date: Mon Jan 12 11:54:15 2009 -0500 18281 18282 R3xx/R4xx: Maximize the use of clipped triangles for Xv rendering 18283 18284 While trying to get tear free video on my RS690 I found some problems in the code for rendering the 18285 video as a clipped triangle. The size limitations are quite strict and scaling the video output window 18286 creates all kinds of weird effects. This patch maximizes the use of this function and resolves the 18287 scaling issues. There could be a problem with zero target height/width, not sure if this is handled 18288 elsewhere. 18289 18290commit 3d106b990fa8daba96e590c57b164640a2faae7f 18291Author: Alan Coopersmith <alan.coopersmith@sun.com> 18292Date: Fri Jan 9 16:26:04 2009 -0800 18293 18294 Remove xorgconfig & xorgcfg from See Also list in man page 18295 18296commit 14d29a223028c183ba11e717a3737285d0ac8de1 18297Author: Alex Deucher <alexdeucher@gmail.com> 18298Date: Wed Jan 7 17:04:42 2009 -0500 18299 18300 RV280: add another AGP quirk 18301 18302commit 1795706a6bb13272a58e8fa1bd5ad2603d553dde 18303Author: Nicos Gollan <gtdev@spearhead.de> 18304Date: Wed Jan 7 14:25:24 2009 -0500 18305 18306 [PATCH] Fixed enumerations in radeon-output.c 18307 18308commit f69ad5aa17bb907105c1657803628d52b5ff3f11 18309Author: Alex Deucher <alexdeucher@gmail.com> 18310Date: Wed Jan 7 11:33:53 2009 -0500 18311 18312 Bump version post release 18313 18314commit a0dd5d7ee3f038a9bfe051db8dbfac4934a81581 18315Author: Alex Deucher <alexdeucher@gmail.com> 18316Date: Tue Jan 6 19:08:28 2009 -0500 18317 18318 Bump version for release 18319 18320commit c3fb8bb279959512a4ced644a64cc660c5cd97bd 18321Author: Alex Deucher <alexdeucher@gmail.com> 18322Date: Tue Jan 6 17:41:22 2009 -0500 18323 18324 ATOM: rework encoder/transmitter setup 18325 18326commit 9bb6b8ceaf772e1ce63610812a12260ee3860c27 18327Author: Alex Deucher <alexdeucher@gmail.com> 18328Date: Tue Jan 6 16:25:26 2009 -0500 18329 18330 ATOM: refactor output dpms 18331 18332commit 467995d960bdefb1911eabfc2a047cf816013101 18333Author: Alex Deucher <alexdeucher@gmail.com> 18334Date: Tue Jan 6 15:18:41 2009 -0500 18335 18336 ATOM: Adjust PLL setup for recent atom changes 18337 18338commit ee6310174c44809c761dbdb423e56e61392a0e10 18339Author: Alex Deucher <alexdeucher@gmail.com> 18340Date: Tue Jan 6 14:59:22 2009 -0500 18341 18342 ATOM: handle cases where TMDS uses linkb 18343 18344 fixes bug 18564 18345 18346commit 690999b682c0f954e27d9f1b3a3d8db4ef865c21 18347Author: Alex Deucher <alexdeucher@gmail.com> 18348Date: Tue Jan 6 14:55:49 2009 -0500 18349 18350 ATOM: rework object table parsing 18351 18352commit e23baa2fa8b25b68cd473aec7a3d8cf3ee7b0ea8 18353Author: Alex Deucher <alexdeucher@gmail.com> 18354Date: Tue Jan 6 14:38:06 2009 -0500 18355 18356 R100: Fix typo in e55598cc1b35d398c9eb2c3946df71456775f621 18357 18358 Noticed by Andreas Stenglein 18359 18360commit d5cb24075928690377fd2dfaa32abfebff9990df 18361Author: Alex Deucher <alexdeucher@gmail.com> 18362Date: Sun Jan 4 21:04:52 2009 -0500 18363 18364 ATOM: add format rev 2 for digitial command tables 18365 18366 should fix bug 18911 18367 18368commit 1813456af3def38c6f53103054976f3092491746 18369Author: Alex Deucher <alexdeucher@gmail.com> 18370Date: Sun Jan 4 21:01:05 2009 -0500 18371 18372 ATOM: fail on more tables with unhandled revisions 18373 18374commit 90ce00a3dbf37833a71db234372695460c8f0342 18375Author: Alex Deucher <alexdeucher@gmail.com> 18376Date: Sun Jan 4 19:13:53 2009 -0500 18377 18378 AGP: Print both device and subsytem pci ids for agp 18379 18380 helpful for adding quirks with libpciaccess 18381 18382commit 006d736d890a6b2f1d7eb8ee20ccf467a939516f 18383Author: Alex Deucher <alexdeucher@gmail.com> 18384Date: Sun Jan 4 19:08:37 2009 -0500 18385 18386 Another AGP quirk 18387 18388 debian bug 510208 18389 18390commit c39ca66a7e216b7492fc5e3680ac535831cfb9a1 18391Author: Alex Deucher <alexdeucher@gmail.com> 18392Date: Tue Dec 30 19:05:58 2008 -0500 18393 18394 R2xx: add another AGP quirk 18395 18396 from http://www.phoronix.com/forums/showthread.php?t=14626 18397 18398commit 877be1b268f275b8ad83d5e245a0f0d3b1a05091 18399Author: Fabio <fabio.ped@libero.it> 18400Date: Mon Dec 29 18:11:10 2008 -0500 18401 18402 man page updates 18403 18404 bug 19329 18405 18406commit 837f7e8a48ea35d21de24b620830a2c6db854861 18407Author: Dave Airlie <airlied@linux.ie> 18408Date: Tue Dec 30 06:50:54 2008 +1000 18409 18410 atombios: endian fix in r600 object table handling 18411 18412 This fixes the r600 object table on r600. 18413 18414commit 45f8b911659dba7f43b9b70bde4b9c3be51b0b5b 18415Author: Alex Deucher <alexdeucher@gmail.com> 18416Date: Mon Dec 29 10:25:33 2008 -0500 18417 18418 RS600: remove gpio quirk for ddc 18419 18420 Doesn't seem necessary on RS600. I'll try and find out more. 18421 18422commit e913990030c536427bfecdb7b3d1e18f1beb90d2 18423Author: Alex Deucher <alexdeucher@gmail.com> 18424Date: Mon Dec 29 10:22:20 2008 -0500 18425 18426 Revert "radeon: maybe special handling only needed for DVI port not HDMI one." 18427 18428 This reverts commit c877450a75d3fd4a2f200af197c691669966f6f4. 18429 18430 This breaks RS690 setups. I think for now we just apply this quirk to 18431 RS690/RS740. I'll follow up with the bios folks about RS600. 18432 18433commit c877450a75d3fd4a2f200af197c691669966f6f4 18434Author: Dave Airlie <airlied@linux.ie> 18435Date: Mon Dec 29 09:05:45 2008 +1000 18436 18437 radeon: maybe special handling only needed for DVI port not HDMI one. 18438 18439 fixes ciaranm's rs600 board on radeon DDC 18440 18441commit 2a9bc6be3246ec230b181391221e3a46beb94a23 18442Author: Dave Airlie <airlied@linux.ie> 18443Date: Mon Dec 29 09:05:16 2008 +1000 18444 18445 radeon: a-bit motherboard has no DVI output 18446 18447 thanks to ciaranm for finding this. 18448 18449commit eccf3fed2be54f5c02fefef32e39c4084acbbe22 18450Author: Alex Deucher <alexdeucher@gmail.com> 18451Date: Fri Dec 26 19:51:58 2008 -0500 18452 18453 Make sure to include radeon_drm.h in the tarball 18454 18455commit e55598cc1b35d398c9eb2c3946df71456775f621 18456Author: Alex Deucher <alexdeucher@gmail.com> 18457Date: Fri Dec 26 19:31:55 2008 -0500 18458 18459 R2xx EXA: limit composite to 2047 like r1xx 18460 18461 See bug 19269 18462 18463commit 631123d144d088d4f77a0599c34deaa31d551d71 18464Author: Dave Airlie <airlied@redhat.com> 18465Date: Tue Dec 23 10:08:46 2008 +1000 18466 18467 radeon: update to 6.9.0.91 18468 18469 do a prerelease for end of year. 18470 18471commit 047f7603174f6047090f4fafbe6488d697eb78bc 18472Author: Dave Airlie <airlied@linux.ie> 18473Date: Tue Dec 23 08:16:25 2008 +1000 18474 18475 radeon: pass distcheck 18476 18477commit c0c33dab44e6966b1702d4e8cfba3537fc6e2d5c 18478Author: Patrick Haller <patrick.haller@haller-consult.com> 18479Date: Mon Dec 22 03:06:23 2008 -0500 18480 18481 Fix off by one in EXA composite limit checking 18482 18483 Patch from Patrick, with some updates from me: 18484 - fix r200 limits 18485 - note about r100 limits 18486 18487commit d01a609a2b07da4ca0f182e79459432584ec61f6 18488Author: Alex Deucher <alexdeucher@gmail.com> 18489Date: Mon Dec 22 02:18:43 2008 -0500 18490 18491 R3xx-R5xx: better fix for xv primitive tearing issues 18492 18493 R5xx: always use single clipped triangle 18494 R3xx/R4xx: use single clipped triangle up to guardband limit, 18495 then use quad. 18496 18497commit 4d009d14c9c72f6e1bc8100ec8d9b8037ee4e514 18498Merge: d52882d81 4e96278b5 18499Author: Dave Airlie <airlied@redhat.com> 18500Date: Mon Dec 22 16:00:01 2008 +1000 18501 18502 Merge branch 'atom-tvout' 18503 18504commit 4e96278b581e296c1203d97a6d7aa3bff3977222 18505Author: Dave Airlie <airlied@redhat.com> 18506Date: Mon Dec 22 15:58:35 2008 +1000 18507 18508 atombios/tv: add an option to enable atom tv-out for users. 18509 18510 This code is still experimental but we will allow users to enable it for experimental reasons 18511 18512commit d52882d8188830dd52fa112dadcf5ea7f3e5fd5e 18513Author: Dave Airlie <airlied@redhat.com> 18514Date: Mon Dec 22 09:27:47 2008 +1000 18515 18516 radeon: setup 3D engine even when no DRI. 18517 18518 This should fix some missing font issues in EXA without DRI cases. 18519 18520commit 2346fd7cbd90dcdce2b361d374a53da064fa6a12 18521Author: Dave Airlie <airlied@redhat.com> 18522Date: Thu Dec 18 16:35:13 2008 +1000 18523 18524 radeon: add all new pci ids for rv730/rv710 families 18525 18526commit 2ca970c7aa1f3f212f03bf3287d6b60eeafd68ae 18527Author: Alex Deucher <alexdeucher@gmail.com> 18528Date: Wed Dec 17 13:07:19 2008 -0500 18529 18530 Allow arbitrary tv-out modes 18531 18532commit 5da147eec70abb7c9125a8940089fd3ff8bdea95 18533Author: Alex Deucher <alexdeucher@gmail.com> 18534Date: Wed Dec 17 05:37:58 2008 -0500 18535 18536 missed one in last commit 18537 18538commit 6bfbb930c549e67f68969f7f0483380417f755b5 18539Author: Alex Deucher <alexdeucher@gmail.com> 18540Date: Wed Dec 17 05:36:09 2008 -0500 18541 18542 Make sure we hit the right bios reg 18543 18544commit 6ed0e68060fd2775e2d4027ff1533729e87848ba 18545Author: Alex Deucher <alexdeucher@gmail.com> 18546Date: Wed Dec 17 04:54:50 2008 -0500 18547 18548 properly handle EnableYUV 18549 18550 A bit of a hack. we really need to restructure the 18551 bios scratch reg handling. 18552 18553commit f21e673aced1bb5138bbb80e6d9724530eb95ebf 18554Author: Alex Deucher <alexdeucher@gmail.com> 18555Date: Wed Dec 17 03:17:45 2008 -0500 18556 18557 Fix colors on tv-out 18558 18559 Call EnableYUV to set the transform and enable/disable it. 18560 should also fix green tint with tv connected at boot issues. 18561 18562commit b2b43905a5385a8bb0b59b8e50952863d8dacb59 18563Author: Alex Deucher <alexdeucher@gmail.com> 18564Date: Wed Dec 17 00:59:52 2008 -0500 18565 18566 Pre-avivo: fix FP setup 18567 18568 - make sure to clear various shadow timing bits 18569 - crtc1 select bit was set wrong. 18570 - should fix bugs like 19100 18571 18572commit 3754cc23abac99880eda87d06aa42e16081b6c8a 18573Author: Alex Deucher <alexdeucher@gmail.com> 18574Date: Wed Dec 17 00:56:19 2008 -0500 18575 18576 R2xx: switch EXA composite to rects rather quads 18577 18578 For conistency with Xv path and to reduce potential diagonal tearing 18579 18580commit be12cf8fcc7a5a3c3ee2c630474e2f9d9368f4d4 18581Author: Dave Airlie <airlied@redhat.com> 18582Date: Wed Dec 17 15:28:25 2008 +1000 18583 18584 radeon: r500 tv-out force scaler values to nice set that looks correct 18585 18586commit 01f7a50d76f706a456de7a4eb8323939d905c82b 18587Author: Dave Airlie <airlied@redhat.com> 18588Date: Wed Dec 17 13:43:21 2008 +1000 18589 18590 r500: re-enable TV out 18591 18592commit 45b9ff5e250d69de0df41f68b33524f7e2ce13b7 18593Author: Dave Airlie <airlied@redhat.com> 18594Date: Wed Dec 17 13:42:10 2008 +1000 18595 18596 radeon: r500 PAL timings are slightly incorrect 18597 18598 this stablises the PAL image at least 18599 18600commit 435d996949185201a97493e131f9920a32863acd 18601Author: Dave Airlie <airlied@redhat.com> 18602Date: Wed Dec 17 11:50:48 2008 +1000 18603 18604 radeon: tv timings have crev/frev wrong way around for some reason. 18605 18606commit b9549e89cbdf8fff4aa7bd9f3adfb7d5732508cc 18607Author: Dave Airlie <airlied@redhat.com> 18608Date: Wed Dec 17 11:40:53 2008 +1000 18609 18610 atombios: re-enable TV dpms 18611 18612commit db85ff2e0c52a030546c0cc8ed0b851418f3e045 18613Author: Dave Airlie <airlied@redhat.com> 18614Date: Wed Dec 17 11:37:17 2008 +1000 18615 18616 atom/r600: re-enable TV outputs 18617 18618commit ff7b52c9a48f2e417acf6bec3b72546ad9b33471 18619Author: Dave Airlie <airlied@redhat.com> 18620Date: Wed Dec 17 11:36:38 2008 +1000 18621 18622 radeon: use latest atombios constants for TV encoders in set_crtc_source 18623 18624commit 1e0c7d3f0bf6be67b976fdea8fa2df0b4b16435f 18625Author: Dave Airlie <airlied@redhat.com> 18626Date: Wed Dec 17 11:35:06 2008 +1000 18627 18628 atom: spc3 only requires ucEncoderMode for DIG outputs 18629 18630commit 105cfd0b8696347ef71ad9daa941abb308bd6c2c 18631Author: Dave Airlie <airlied@redhat.com> 18632Date: Wed Dec 17 11:31:15 2008 +1000 18633 18634 atom/tv: enable scaler can set the TV output scaler up. 18635 18636 Add the tv standard to the scaler enable for TV use. 18637 18638commit 2a768abbd2f18abe203571213260bccc225f70fd 18639Author: Dave Airlie <airlied@redhat.com> 18640Date: Wed Dec 17 11:29:07 2008 +1000 18641 18642 atom/tv: fixup the tv out default if bios doesn't specify 18643 18644commit b7fdd5da08bfec705691363b8890699c794dce53 18645Author: Dave Airlie <airlied@redhat.com> 18646Date: Wed Dec 17 11:24:56 2008 +1000 18647 18648 atombios: fixup parsing of TV tables on r500 and r600 cards. 18649 18650 The TV tables changed between for later cards. 18651 18652commit e3b08cb59716b525b3063b184eab70e32b9d66f0 18653Author: Dave Airlie <airlied@redhat.com> 18654Date: Tue Dec 16 15:43:16 2008 +1000 18655 18656 atombios: correct fix for previous issue. 18657 18658 I committed an old patch. 18659 18660commit 1431bc6900b0e0d7f486b5b3ba281873472b3fdc 18661Author: Dave Airlie <airlied@redhat.com> 18662Date: Tue Dec 16 15:31:34 2008 +1000 18663 18664 atombios: fix issue with Clearing of MM_INDEX 0. 18665 18666 While playing with tv-out I discovered this issue with clearing 18667 the MM_INDEX register. 18668 18669commit f72367aff2d09161d30066a00673ba291b292a36 18670Author: Dave Airlie <airlied@redhat.com> 18671Date: Tue Dec 16 12:32:24 2008 +1000 18672 18673 radeon: add initial rv730 pciids. 18674 18675 Lots more to come, this just adds a few to start. 18676 18677commit 4d2429f3ec4d7d3e1512b268c7ff5cbff8d186b4 18678Author: Dave Airlie <airlied@redhat.com> 18679Date: Tue Dec 16 12:19:00 2008 +1000 18680 18681 DCE32: add support for DCE3.2 digital outputs. 18682 18683 This adds support for the digital outputs on the RV730. 18684 18685commit 4d3792fd1b9f733107582436540f483d8993199f 18686Author: Dave Airlie <airlied@redhat.com> 18687Date: Fri Dec 12 10:34:57 2008 +1000 18688 18689 radeon: fixes from Alex for some output engines 18690 (cherry picked from commit 1e8ac6ea7b35ccbc3649c96ead60f69158f66ebc) 18691 18692commit f892154dbd0d8b6dd8abdf2196ab0c7c18327412 18693Author: Alex Deucher <alexdeucher@gmail.com> 18694Date: Mon Dec 15 19:41:32 2008 -0500 18695 18696 R3xx-R5xx: switch back to quad rendering for Xv 18697 18698 The large clipped triangle trick hits the guardband limits on 18699 large displays. This gets textured video working again 18700 until we sort out how to deal with the situation better. 18701 18702 See bug 19046 18703 18704commit 62344e460f5d0cac81f94c535cda77d3d76528a9 18705Author: Alex Deucher <alexdeucher@gmail.com> 18706Date: Mon Dec 15 11:30:42 2008 -0500 18707 18708 Remove remnants of unused DDCMode option 18709 18710commit f1efee6a11dc9436276fb12cfb90aa79ada1192b 18711Author: Alex Deucher <alexdeucher@gmail.com> 18712Date: Mon Dec 15 11:27:17 2008 -0500 18713 18714 Remove mention of MergedFB from server log 18715 18716 noticed by Fabio, bug 19098 18717 18718commit 94bffac6968057cdedc792f8540fe44cb6ecfd7c 18719Author: Fabio <fabio.ped@libero.it> 18720Date: Mon Dec 15 11:22:56 2008 -0500 18721 18722 man page updates 18723 18724 from bug 19098 with some minor fixes from me 18725 18726commit c7899905ec824071f9cabbd1e271c4d76010b733 18727Author: Fabio <fabio.ped@libero.it> 18728Date: Mon Dec 15 10:58:23 2008 -0500 18729 18730 Remove duplicate ATI in some device names 18731 18732 from bug 19098 18733 18734commit 89a8fd5337c77abf1f3166e69ed061d85b5c20f0 18735Author: Dave Airlie <airlied@redhat.com> 18736Date: Mon Dec 15 14:41:10 2008 +1000 18737 18738 r600: don't idle engine 18739 (cherry picked from commit b77729448e487d37707ce028d8edc59352d1cc2d) 18740 18741commit 4ce255a9db8027f2471ba50e57724d3e4d5b16ba 18742Author: Fabio <fabio.ped@libero.it> 18743Date: Fri Dec 12 09:53:33 2008 -0500 18744 18745 Fix formatting in man page 18746 18747 fixes bug 19039 18748 18749commit 0e99017cb9a6d46b891705df73678c2705439532 18750Author: Joseph Adams <joeyadams3.14159@gmail.com> 18751Date: Thu Dec 11 00:57:03 2008 -0500 18752 18753 Add more eMac modes, fix crash in last patch 18754 18755commit fb46c30d316ab3ec54c54f3aec91d5164070a423 18756Author: Alex Deucher <alexdeucher@gmail.com> 18757Date: Wed Dec 10 17:16:56 2008 -0500 18758 18759 Add support for Apple eMacs 18760 18761 Thanks to Joseph Adams for helping me sort this out. 18762 Verified on eMac G4/1.0 with radeon 7500 18763 Later eMacs with radeon 9200 or 9600 chips may have 18764 different ddc setups. Need to verify. 18765 18766commit c0bcea9150ef215fa614733cef9a5b71a55a33bd 18767Author: Alex Deucher <alexdeucher@gmail.com> 18768Date: Wed Dec 10 11:40:47 2008 -0500 18769 18770 Add another AGP quirk 18771 18772 fixes bug 12544 18773 18774commit 607b02d6abf78147902a5944bd006897ad95e982 18775Author: Alex Deucher <alexdeucher@gmail.com> 18776Date: Wed Dec 10 03:54:35 2008 -0500 18777 18778 Add randr 1.3 panning support 18779 18780commit 24007acbf36bcfd09cb5e74ce4a8d918d393cf34 18781Author: Benjamin Close <Benjamin.Close@clearchain.com> 18782Date: Wed Dec 10 02:17:56 2008 -0500 18783 18784 [PATCH] xf86misc extension has been removed, update driver to reflect it 18785 18786commit 69e445f7f2ebd299070e44457b434bbc28b7e05c 18787Author: Dave Airlie <airlied@linux.ie> 18788Date: Tue Dec 9 18:59:11 2008 +1000 18789 18790 radeon: memset the pll setting function 18791 18792 Thanks to Peter Zijlstra for making me debug this and providing 18793 feedback. 18794 18795commit c72b732a2f5597cd72bd4ebe207799ae07f2681e 18796Author: Adrian Friedli <adi@koalatux.ch> 18797Date: Mon Dec 8 10:42:46 2008 -0500 18798 18799 Add another AGP quirk 18800 18801commit 979ad04d901676cb41aea5cf47a8a2e9c19064dd 18802Author: Alex Deucher <alexdeucher@gmail.com> 18803Date: Fri Dec 5 17:53:57 2008 -0500 18804 18805 Add missing .TP in man page 18806 18807commit b89d434011787fb71593395f462d142375783e35 18808Author: Pierre Ossman <pierre@ossman.eu> 18809Date: Fri Dec 5 17:52:02 2008 -0500 18810 18811 Fix scissor setup for Xv 18812 18813 We need to adjust the scissoring for each clip box, so move the setup 18814 to inside the while loop. 18815 18816commit 83c1115333cbe763f86422fa5aabf4ab21951e68 18817Author: Alex Deucher <alexdeucher@gmail.com> 18818Date: Fri Dec 5 11:52:18 2008 -0500 18819 18820 R200: Remove scissor setup from init3d() 18821 18822 We aren't using them, so no need to set them up. 18823 18824commit 7593d5399cdb2f42ecbcb5193580385b286a9db9 18825Author: Alex Deucher <alexdeucher@gmail.com> 18826Date: Fri Dec 5 11:18:15 2008 -0500 18827 18828 set coherent mode consistently (enabled for now) 18829 18830 It's set to enabled in the output setup, but then was set to 18831 disabled when teh output properties were setup. Set it up 18832 consistently. 18833 18834commit ad2579f8898251105a6b36b745afd1ce1dab103e 18835Author: Dave Airlie <airlied@linux.ie> 18836Date: Fri Dec 5 20:03:53 2008 +1000 18837 18838 radeon: damn you bios, HDMI cannot have DACs. 18839 18840commit bae30856f2cb3884387ef347327af6ff00b75854 18841Author: Alex Deucher <alexdeucher@gmail.com> 18842Date: Fri Dec 5 08:15:59 2008 +0100 18843 18844 Make VSync for EXA and Xv configurable 18845 18846commit 33638d9e388b330e2f4eb4debd05ba09924cf176 18847Author: Pierre Ossman <pierre@ossman.eu> 18848Date: Thu Dec 4 23:29:31 2008 +0100 18849 18850 Optimise RADEONWaitForVLine 18851 18852 Only avoid the vlines we are rendering to, instead of the entire 18853 screen. This way we don't stall the card for longer than we 18854 absolutely have to. 18855 18856 EXA calls fixed by Alex Deucher. 18857 18858commit d1690f5cc096e2f735c8b407c370a1c1cd7a8afc 18859Author: Pierre Ossman <pierre@ossman.eu> 18860Date: Wed Dec 3 21:49:37 2008 +0100 18861 18862 Improve tearing avoidance for Xvideo in two steps 18863 18864 - Fix up VLINE handling to trigger whenever scanout is outside the 18865 visible area. 18866 18867 - Render the video as a scissored triangle as R300+ cannot render a 18868 quad in a single pass. 18869 18870commit 4d98acbca2e630056bf56cdcd0e23007fded2ced 18871Author: Alex Deucher <alexdeucher@gmail.com> 18872Date: Wed Jul 16 03:14:06 2008 -0400 18873 18874 First pass at tear-free accel 18875 18876 if the dest pixmap is the front buffer, stall the pipe 18877 until the vline is outside the active area. 18878 For EXA, pick crtc based on the larger mode area; 18879 ideally we'd have one pixmap per crtc. 18880 For Xv, use dst window area to determine crtc. 18881 18882commit 82d67b70d0fb747cf7b1eb0b64244ae56cb84572 18883Author: Pierre Ossman <pierre@ossman.eu> 18884Date: Thu Dec 4 15:38:13 2008 -0500 18885 18886 Switch r200 Xv to use rect lists rather than quads to avoid tearing 18887 18888 Patch from Pierre with fixes from me. 18889 18890commit 9eccd0ad0827a1b8ec9abb2cebf5768d204a5ce9 18891Author: Alex Deucher <alexdeucher@gmail.com> 18892Date: Thu Dec 4 12:28:12 2008 -0500 18893 18894 Fix up US setup after r3xx bicubic merge 18895 18896commit e13fba853ba19e6b0f081b9b3d9fa76c38a0f82b 18897Merge: d29633708 79bbdd984 18898Author: Alex Deucher <alexdeucher@gmail.com> 18899Date: Thu Dec 4 12:25:29 2008 -0500 18900 18901 Merge branch 'bicubic' of git://git.infradead.org/users/drzeus/xf86-video-ati 18902 18903commit d29633708b5451f5541e88371d831ae03019a7f5 18904Author: Dave Airlie <airlied@redhat.com> 18905Date: Thu Dec 4 09:08:48 2008 +1000 18906 18907 radeon: apply atom quirks to object table 18908 18909commit 941ac0d3ab839dc5aaaef2f3b7f5c7085b2894c5 18910Author: Alex Deucher <alexdeucher@gmail.com> 18911Date: Wed Dec 3 17:31:25 2008 -0500 18912 18913 Make sure 3D state is emited when switching to 3D IB 18914 18915 - fixes bug 18864 18916 18917commit 79bbdd984c925e37f5b3db2605339f1a21377fcf 18918Author: Pierre Ossman <pierre@ossman.eu> 18919Date: Wed Dec 3 20:33:36 2008 +0100 18920 18921 Change the XV_BICUBIC attribute to a tristate, where the third state 18922 automatically turns the filter on and off as appropriate. 18923 18924commit bbdf821c179242266fd86edf9e688a6be9cf985f 18925Author: Pierre Ossman <pierre@ossman.eu> 18926Date: Wed Dec 3 19:42:50 2008 +0100 18927 18928 Make the R300 bicubic shader program a bit easier to understand by 18929 having comments that describe the operation, not how it has to be 18930 implemented in the hardware. 18931 18932commit 8f4b22932c91f1d51fcbf1d1a99d852fcdaafdcc 18933Author: Pierre Ossman <pierre@ossman.eu> 18934Date: Wed Dec 3 19:31:31 2008 +0100 18935 18936 Now that we have a bicubic code path that messes up the pixel shader, 18937 we need to restore it in the non-bicubic code. 18938 18939commit 0851daa6e7db77e440188d774ef8b3236146bcb9 18940Author: Pierre Ossman <pierre@ossman.eu> 18941Date: Wed Dec 3 19:17:36 2008 +0100 18942 18943 Fix macros so that they are safe to use with expressions as parameters. 18944 18945commit e04bcb0066781f61e97787856261e3380036d304 18946Author: Pierre Ossman <pierre@ossman.eu> 18947Date: Wed Dec 3 19:16:03 2008 +0100 18948 18949 Fix node setup on R300 bicubic pixel shader. It was backwards and used 18950 the wrong base for sizes. 18951 18952commit 927aee181f3637bda1b761caf93fbc0652e1207b 18953Author: Alex Deucher <alexdeucher@gmail.com> 18954Date: Wed Dec 3 01:39:38 2008 -0500 18955 18956 Only flush IB when direct rendering is enabled 18957 18958commit ca74ebf8813b7fab8d08465ef9600f001c948e94 18959Author: Alex Deucher <alexdeucher@gmail.com> 18960Date: Tue Dec 2 22:26:21 2008 -0500 18961 18962 move 3d state init after RADEON_SWITCH_TO_3D() 18963 18964commit fc079c5267baf431bbecee7744e484783d393152 18965Author: Dave Airlie <airlied@redhat.com> 18966Date: Tue Dec 2 22:11:57 2008 -0500 18967 18968 Don't mix 2D and 3D in the same IB 18969 18970 fix from radeon-gem-cs 18971 18972commit 678693aabd0b4c1812ab44bf5abb4c8394b254bb 18973Author: Dave Airlie <airlied@redhat.com> 18974Date: Tue Dec 2 22:02:21 2008 -0500 18975 18976 Stability fixes from radeon-gem-cs 18977 18978commit f231e4d9b2c683dc16f42cf09484fae926a892c1 18979Author: Dave Airlie <airlied@redhat.com> 18980Date: Wed Dec 3 11:39:04 2008 +1000 18981 18982 radeon: fix HDMI having DACs in some BIOSes 18983 18984commit fe73d9a7dfe8ec5c8f1a8dc08e14b4e138aa9276 18985Author: Bryce Harrington <bryce@canonical.com> 18986Date: Tue Dec 2 19:47:45 2008 -0500 18987 18988 Add another AGP quirk 18989 18990commit 6e0c48958c6a50a639368e369ce72d813256aee4 18991Author: Alex Deucher <alexdeucher@gmail.com> 18992Date: Mon Dec 1 18:50:40 2008 -0500 18993 18994 Add quirk for r200 card with the primary dac wired to both ports 18995 18996 refactor load detection setup as a byproduct 18997 should fix bug 18719 18998 18999commit fa496d7b0397d9be57db90d0860928e9ced73cca 19000Author: Pierre Ossman <pierre@ossman.eu> 19001Date: Mon Dec 1 08:29:29 2008 +0100 19002 19003 Fix comments for R500 fragment shader to reflect the code. 19004 19005commit 0d95fca114194664d9ab612c90d3222912cf41ff 19006Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19007Date: Sat Aug 30 12:38:43 2008 -0700 19008 19009 Fix indentation on IS_R300_3D Xv code. 19010 19011commit 36a62b8c47def3f23a0b735875f59d23ed3c095a 19012Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19013Date: Sat Aug 30 11:15:20 2008 -0700 19014 19015 r3xx: Various Xv fixes. 19016 19017commit 01ed5462fad56f5b7bf867041d00443bdf81c653 19018Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19019Date: Thu Aug 7 18:08:54 2008 -0700 19020 19021 Switch to Mesa-style 24-bit float packing. 19022 Seems like Mesa's got it right, so we should follow suit. 19023 19024commit 85bbe889045aa434f71a3eb0ba8153b7d2cc5023 19025Author: Maciej Cencora <m.cencora@gmail.com> 19026Date: Thu Aug 7 07:06:46 2008 -0700 19027 19028 Enable bicubic Xv on r3xx+. Still some buggies in the actual display, but looks pretty good. 19029 19030commit bf6263cff50bb85c19ff6dfebb739333fa8aca18 19031Author: Maciej Cencora <m.cencora@gmail.com> 19032Date: Thu Aug 7 07:00:36 2008 -0700 19033 19034 Add r3xx-specific bicubic filtering code. FP, RS, mostly, everything else is shared with r5xx. 19035 19036commit da021c36bbdf3bca31ee50ebe01cdb9495c09b36 19037Author: Dave Airlie <airlied@linux.ie> 19038Date: Sun Nov 30 18:44:24 2008 +1000 19039 19040 radeon_drm.h: remove kernel defines 19041 19042 remove these we should update from an installed kernel copy not from the 19043 original kernel copy. 19044 19045commit 783cdb7374941bb1d2b63eea375fbf2f1b808cc3 19046Author: Dave Airlie <airlied@redhat.com> 19047Date: Thu Nov 27 15:29:06 2008 +1000 19048 19049 r100/r200: EXA misrenders dst == a8 + dst alpha use. 19050 19051 Fedora BZ#469556 19052 19053 I've played with various dst and texture formats and rewritten large 19054 parts of this to no avail, so I'm punting for now. 19055 19056commit 2d7cfc421713286edd5121b56999246f9ae70286 19057Author: Eygene Ryabinkin <rea-fbsd@codelabs.ru> 19058Date: Tue Nov 25 13:25:33 2008 -0500 19059 19060 Catch unsubstituted macro for PCI region base access 19061 19062 also fix up some warnings. 19063 19064commit af7690b9c300d37c3a0c51e27e0ad2ca009224a2 19065Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 19066Date: Tue Nov 25 02:57:30 2008 -0500 19067 19068 [PATCH] radeon: Fix PCI usage of 32-bit driver on 64-bit platform 19069 19070 The radeon driver is storing PCI addresses in unsigned long's which 19071 won't work well on 32-bit platforms with 64-bit physical address space 19072 such as PowerPC 4xx. This fixes it by using unsigned long long instead. 19073 19074commit 065938617c0feab17f4274a5350de02a692ba065 19075Author: Ramon van der Stelt <rvanderstelt@xs4all.nl> 19076Date: Tue Nov 25 02:23:46 2008 -0500 19077 19078 Interlaced mode fixes 19079 19080 see bug 12626 19081 19082commit 36a7dc6ea1e1929e986ab1159497c71521cb2f10 19083Author: Bryce Harrington <bryce@canonical.com> 19084Date: Tue Nov 25 01:04:37 2008 -0500 19085 19086 Additional AGP quirks 19087 19088 bug 18693 19089 19090commit 2acb8e4ed220fccb21b22a53c5142ab3a35d32ef 19091Author: Alex Deucher <alexdeucher@gmail.com> 19092Date: Tue Nov 25 15:07:09 2008 +1000 19093 19094 radeon: fixup shared DDC lines for some rv610 cards. 19095 19096 Some cards share DDC between a HDMI and VGA output, and expect the 19097 driver to use load detect or EDID to figure it out. 19098 19099 airlied- shipped in RHEL5 19100 19101 Signed-off-by: Dave Airlie <airlied@redhat.com> 19102 19103commit b01e35168de40d192fd7a9ce6884b9c7419afbd4 19104Author: Alex Deucher <alexdeucher@gmail.com> 19105Date: Mon Nov 24 21:06:42 2008 -0500 19106 19107 Fix up posted logic 19108 19109 noticed by benh on IRC 19110 19111commit d7a03e93372b17d642b42f5490464616ae7c99c1 19112Author: Alex Deucher <alexdeucher@gmail.com> 19113Date: Mon Nov 24 15:52:05 2008 -0500 19114 19115 Get hardcoded edid from the bios for servers that support it 19116 19117 should fix bug 11300 19118 19119commit 3858e31fa9d77d52a5cabb02f53f06385eab40f7 19120Author: Alex Deucher <alexdeucher@gmail.com> 19121Date: Mon Nov 24 14:39:41 2008 -0500 19122 19123 Don't support rotation is accel is not available 19124 19125 Fixes bug 18068 19126 19127commit 3d27876d91ef69d71f17b6cf8c07c26bf12857b4 19128Author: Dave Airlie <airlied@redhat.com> 19129Date: Mon Nov 10 15:47:13 2008 +1000 19130 19131 radeon: always align dstPitch to 64 bytes for textured video 19132 19133 bz# 18454 19134 19135commit 6c8b352264540ccc4622a7c0b704ce452dcd707d 19136Author: Dave Airlie <airlied@redhat.com> 19137Date: Wed Oct 29 15:53:57 2008 +1000 19138 19139 pciid: add HD3300 - 790GX chipset. 19140 19141 RH BZ (#466706) 19142 19143commit 902eaf768142c6c7dcc487e10775027b84cd1f9a 19144Author: Alex Deucher <alexdeucher@gmail.com> 19145Date: Thu Nov 6 15:46:43 2008 -0500 19146 19147 Check for LVDS on all IGP chips 19148 19149 - fixes bug 18395 19150 19151commit 44abaa0ea8e5e3bbe833fa73e6d011618bb6d7e7 19152Author: Adam Jackson <ajax@redhat.com> 19153Date: Mon Nov 3 15:09:28 2008 -0500 19154 19155 Fix HDMI output setup on DIG1/DIG2. 19156 19157 The only reliable way to detect an HDMI sink is to check for the 19158 appropriate EDID extension block, so this will only work right on 19159 servers with xf86DoEEDID(). RANDR 1.2 will call this internally for 19160 us if it exists, but pre-1.6 servers are out of luck. 19161 19162commit 937b7ac2a259cf504a19dcf62a58b1db1afb8eb9 19163Author: Bryce Harrington <bryce@canonical.com> 19164Date: Wed Oct 29 01:31:50 2008 -0400 19165 19166 Add AGP quirk table 19167 19168 - set known good default agp modes for problematic chip/bridge combos 19169 - should fix bug 17360, others 19170 19171commit 67bc2ae7f227a2afedc8e699cf0458fa7960f301 19172Author: Julien Cristau <jcristau@debian.org> 19173Date: Sun Oct 26 16:05:04 2008 +0100 19174 19175 Bug#7148: Fix manpage formatting 19176 19177commit 98fdd78a1229584b922f816f331a45bfc178c961 19178Author: Dave Airlie <airlied@linux.ie> 19179Date: Tue Oct 21 18:36:22 2008 +1000 19180 19181 atombios: oops typo 19182 19183commit 763dff6c117a10b6a57ba1854c5b798359789e44 19184Author: Dave Airlie <airlied@linux.ie> 19185Date: Tue Oct 21 18:35:44 2008 +1000 19186 19187 atombios: set missing pixel clock 19188 19189commit 435cf7da68186f2601c4b888296117d4f652c625 19190Author: Alex Deucher <alexdeucher@gmail.com> 19191Date: Wed Oct 15 10:52:14 2008 -0400 19192 19193 Add LCD dithering quirk for macbook pro 19194 19195 fixes bug 17897. 19196 19197commit c0e6cb6d0eeef8f2ea60d840e1cd668fa92cd7f9 19198Author: Alex Deucher <alexdeucher@gmail.com> 19199Date: Fri Oct 10 11:09:27 2008 -0400 19200 19201 Fix grey level dithering setup 19202 19203commit 9eb2b6cd28224f0cbbabf3df451696ae1a845b9c 19204Author: Dave Airlie <airlied@redhat.com> 19205Date: Fri Oct 10 16:18:05 2008 +1000 19206 19207 radeon: hopefully fix plls for 30" monitors. 19208 19209 Only do the HIGH trick on the DCE3 cards. Tested on rv610/630/770. 19210 19211commit 0975e007ec7933acf9cd8ad2f281313ef8d32ac9 19212Author: Dave Airlie <airlied@redhat.com> 19213Date: Thu Oct 9 17:16:28 2008 +1100 19214 19215 radeon: rename radeon_memory to radeon_legacy_memory. 19216 19217 This is a setup patch for the kms/memory management code. 19218 19219commit d744dc013cf5ca3141c0fdf55e36d713fe9f53ba 19220Author: Adam Jackson <ajax@redhat.com> 19221Date: Wed Oct 8 15:50:25 2008 -0400 19222 19223 Fix connector table mapping for lvtma/kaleidoscope/uniphy. 19224 19225 Fixes DPMS on LVDS. 19226 19227commit dc795ba66a8c0eb2c3fdb86bd42dfc17e0aa3b6e 19228Author: Alex Deucher <alexdeucher@gmail.com> 19229Date: Tue Oct 7 12:32:47 2008 -0400 19230 19231 Fixups from last DFP/LCD refactor 19232 19233 - fix DFP1 -> LCD1 19234 - use _INDEX rather than _SUPPORT defs 19235 19236commit 4a9e8d9fe24278b56aef51677fd35289d01584a7 19237Author: Alex Deucher <alexdeucher@gmail.com> 19238Date: Mon Oct 6 18:00:08 2008 -0400 19239 19240 set grey levels correctly for temporal dithering 19241 19242commit 5708624d1151877dd66a462d4a75fed6774604a7 19243Author: Alex Deucher <alexdeucher@gmail.com> 19244Date: Mon Oct 6 15:11:56 2008 -0400 19245 19246 Turn coherent mode off by default on DVI 19247 19248commit a5c5ce96279d01eb519bfb92b94c06a58acb7f07 19249Author: Alex Deucher <alexdeucher@gmail.com> 19250Date: Mon Oct 6 15:07:31 2008 -0400 19251 19252 Refactor atom LCD/DFP output setup 19253 19254 Simplify the code, properly handle dithering, coherent mode, encoding, etc. 19255 Should fix bug 17897. 19256 19257commit f9826a5694b7adb6920eb5bdf45d840d8fb14d53 19258Author: Alex Deucher <alexdeucher@gmail.com> 19259Date: Fri Oct 3 15:24:12 2008 -0400 19260 19261 Add support for DDC via atom commands for RV410 19262 19263 The atom calls use the hw i2c engine for DDC. For some 19264 reason, sw i2c doesn't seem to work on the VGA GPIO on 19265 RV410 chips, so we use atom in that case. 19266 19267 This fixes the longstanding VGA DDC problems on RV410/M26 19268 chips. 19269 19270commit 23c17c0121d43c2fd11bb5cfbaa4b02abac2b16d 19271Author: Alex Deucher <alexdeucher@gmail.com> 19272Date: Fri Oct 3 14:30:33 2008 -0400 19273 19274 man page updates 19275 19276 Update render accel and page flipping information. 19277 fixes bug 17887. 19278 19279commit 4ac6e499089aec3c79af5dbede5fe631b105ce3d 19280Author: Alex Deucher <alexdeucher@gmail.com> 19281Date: Wed Oct 1 18:53:15 2008 -0400 19282 19283 proper fix for initial rotation 19284 19285 The initial modeset was setting the MC regs, so just set them 19286 rather than the full modeset. 19287 19288commit 346228b20c69a965c9e7d67157f42c1d36a5b6c3 19289Author: Alex Deucher <alexdeucher@gmail.com> 19290Date: Wed Oct 1 14:38:27 2008 -0400 19291 19292 Workaround to make initial rotation work 19293 19294 Acceleration needs to be initialized before setting the mode 19295 for initial rotation to work. Changing the order in RADEONScreenInit() 19296 doesn't work because RADEONDRIKernelInit() hangs in the ioctl to 19297 initialize the CP if the mode is not set prior (even if the heads 19298 are turned off after setting the mode). The workaround is to set the 19299 modes twice, once before accel and once after. 19300 19301 This needs to be looked into further. 19302 19303commit c359c2a31caf9f75b9fc6b6bcbc3e9dc1fe404ba 19304Author: Michel Dänzer <michel@tungstengraphics.com> 19305Date: Tue Sep 30 11:04:54 2008 +0200 19306 19307 Pass base offset into RADEONDisplayVideo() explicitly. 19308 19309 This prevents the base offset from flipping back and forth if double buffering 19310 is enabled and the two buffers start on different sides of a 4 MB boundary. 19311 19312 Should fix http://bugs.freedesktop.org/show_bug.cgi?id=17254 ; please don't 19313 reopen that bug report without attaching a log file captured after reproducing 19314 the problem with this fix. 19315 19316commit ba21d5ae7fbbf5828c44c41655aff8b182453e2e 19317Author: Michel Dänzer <michel@tungstengraphics.com> 19318Date: Tue Sep 30 11:03:04 2008 +0200 19319 19320 Update GEN_INT_CNTL register value after calling DRM modeset ioctl. 19321 19322 Fixes sync-to-vblank hangs after mode switch with DRM vblank-rework. 19323 19324commit 3110646daf05fa09faaa70066cc8c3db1bb07cec 19325Author: Michel Dänzer <michel@tungstengraphics.com> 19326Date: Sun Sep 14 11:46:36 2008 +0200 19327 19328 Fix -Werror build. 19329 19330commit d82f2938f69402c641a1c8362fdc513419b27659 19331Author: Alex Deucher <alexdeucher@gmail.com> 19332Date: Fri Sep 26 13:51:24 2008 -0400 19333 19334 atombios updates from upstream 19335 19336commit d0d58b39e49c5ce00bc8bd12f721ad8c7f554b91 19337Author: Alex Deucher <alexdeucher@gmail.com> 19338Date: Wed Sep 24 11:54:34 2008 -0400 19339 19340 radeon: Further pll adjustments 19341 19342commit d100a6af8f828eb94f8ba6e8a96c24389b5cf46f 19343Author: Alex Deucher <alexdeucher@gmail.com> 19344Date: Fri Sep 19 14:04:59 2008 -0400 19345 19346 cleanup macbook quirk 19347 19348commit 1218e38c75351a20f7414ba6fcb3545689339442 19349Author: Alex Deucher <alexdeucher@gmail.com> 19350Date: Fri Sep 19 13:52:07 2008 -0400 19351 19352 more macbook fixes 19353 19354commit 83e1f132965deadedda7a622184eb400cd4c120c 19355Author: Alex Deucher <alexdeucher@gmail.com> 19356Date: Fri Sep 19 09:36:59 2008 -0400 19357 19358 Fix copy/paste error in the last commit 19359 19360 DVI-D -> DVI-I 19361 19362commit 4af825f7b13cba8c3b204e72a07cab5009331240 19363Author: Alex Deucher <alexdeucher@gmail.com> 19364Date: Fri Sep 19 00:29:20 2008 -0400 19365 19366 radeon: Add quirk for macbook 19367 19368 - should fix bug 17625 19369 19370commit 38aa450fce6a91b35c191fd07112490a62cc29c6 19371Author: Alex Deucher <alexdeucher@gmail.com> 19372Date: Tue Aug 26 21:50:23 2008 -0400 19373 19374 fix some fallout from the common allocator 19375 19376 should fix bug 17317 19377 19378commit a3cc1d7a421456186024c5c069e403d374a0f0b9 19379Author: Dave Airlie <airlied@linux.ie> 19380Date: Tue Aug 26 08:03:20 2008 +1000 19381 19382 radeon: fix powerpc build 19383 19384commit 6cebfe257f7ddad855ee743e4eb899bd6fac7f46 19385Author: Alex Deucher <alexdeucher@gmail.com> 19386Date: Fri Jul 11 19:32:06 2008 -0400 19387 19388 Switch EXA path back to static cursor allocation 19389 19390 pre-AVIVO cards have address limits for the cursor offset 19391 19392commit 4dff54a3c8d7c9f2d6ec50354ff0b92f1b7fcbdf 19393Author: Alex Deucher <alexdeucher@gmail.com> 19394Date: Fri Jul 11 17:28:03 2008 -0400 19395 19396 Switch cursors over to generic allocator 19397 19398commit 5b1978a4796bcc31ac2f01d303dc8f8f44323025 19399Author: Alex Deucher <alexdeucher@gmail.com> 19400Date: Mon Aug 25 10:05:28 2008 -0400 19401 19402 Bicubic fixes from the last cherry-pick 19403 19404commit 7daba77ded1c718e93ae8c372a39a6e85228d513 19405Author: Alex Deucher <alexdeucher@gmail.com> 19406Date: Fri Jul 11 02:02:38 2008 -0400 19407 19408 Convert randr, Xv to a common allocator 19409 19410commit 610fe1a937da78f4ac813ac919c158dde8f42442 19411Author: Alex Deucher <alexdeucher@gmail.com> 19412Date: Thu Jul 10 22:47:35 2008 -0400 19413 19414 Switch to using radeon_drm.h from the drm 19415 19416 modelled after Matthias' similar rhd change 19417 19418commit a6561f2ec673b38907f7181235386f32e60c32ba 19419Author: Alex Deucher <alexdeucher@gmail.com> 19420Date: Thu Jul 10 22:31:13 2008 -0400 19421 19422 Move DRI to a separate struct 19423 19424 based on Matthias' similar work in rhd 19425 19426commit 75ef8dc214715d3c5c50996b293933842903ba65 19427Author: Egbert Eich <eich@freedesktop.org> 19428Date: Thu Jul 10 21:49:55 2008 -0400 19429 19430 Cleanups from rhd port 19431 19432 - remove unused vars 19433 - remove static exa render vars 19434 19435commit 71ad140fa11f3a504c38d6bddf40e3a3c0a20e60 19436Author: Alex Deucher <alexdeucher@gmail.com> 19437Date: Thu Jul 10 21:24:16 2008 -0400 19438 19439 Move accel state to a separate struct 19440 19441commit 5b2e095c31b88d8495a4f86e6cb46b49fa4acd65 19442Author: Alex Deucher <alexdeucher@gmail.com> 19443Date: Thu Jul 10 20:07:44 2008 -0400 19444 19445 Move CP into a separate struct 19446 19447commit 2145309230e36aee7758bd244deb1e72ada0c065 19448Author: Alex Deucher <alexdeucher@gmail.com> 19449Date: Mon Aug 25 08:47:59 2008 -0400 19450 19451 match textured video macro names with other accel code 19452 19453 OUT_VIDEO_REG() -> OUT_ACCEL_REG() 19454 etc. 19455 19456commit 01daef0f095fbbaee701d5fe97f3dd7838b5f915 19457Author: Alex Deucher <alexdeucher@gmail.com> 19458Date: Mon Aug 25 08:26:16 2008 -0400 19459 19460 Additional cleanups and re-arragement following bicubic merge 19461 19462commit 0a51d08c24af040fe48690662b1a912acad51700 19463Author: Alex Deucher <alexdeucher@gmail.com> 19464Date: Mon Aug 25 06:43:14 2008 -0400 19465 19466 Whitespace cleanup from bucubic merge 19467 19468commit c0170fef510b53a6ca2c6ea7a99119235229c929 19469Author: Maciej Cencora <m.cencora@gmail.com> 19470Date: Thu Aug 7 06:53:39 2008 -0700 19471 19472 Add needed FP registers, etc. for r3xx bicubic Xv. 19473 19474commit d9c38326cf70f57ab777ffdf9520b8cdea9d9cb6 19475Author: Dennis Kasprzyk <onestone@opencompositing.org> 19476Date: Tue Aug 5 18:45:38 2008 +0200 19477 19478 Fix typos. 19479 19480commit d38ceba62aa5cee76baa342ce7719a983a79f09e 19481Author: Dennis Kasprzyk <onestone@opencompositing.org> 19482Date: Tue Aug 5 16:21:57 2008 +0200 19483 19484 Fix bicubic fp calculation. 19485 19486commit 6f9c7fde75edc0399559f975db9a6c4dc22714ae 19487Author: Dennis Kasprzyk <onestone@opencompositing.org> 19488Date: Tue Aug 5 15:38:42 2008 +0200 19489 19490 Fixed bicubic fragment program comments. 19491 19492commit 48b09ca40ccb28b5584069316fd38786a78c1dd3 19493Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19494Date: Sat Aug 2 02:36:21 2008 -0700 19495 19496 Switch from 32-bit floats to 16-bit half-floats. 19497 Massive bandwidth savings, or so I'm told. Yay? 19498 19499commit ebbb7fb634fcadf28ff99c1df2c3db89fd56932d 19500Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19501Date: Sat Aug 2 01:45:43 2008 -0700 19502 19503 Change floats to uint32_t hex. 19504 Useful for moving to 16-bit half-floats. 19505 19506commit 3b46162d8a90be3524cb513d42b9ad3d0bea45f5 19507Author: Dennis Kasprzyk <onestone@opencompositing.org> 19508Date: Fri Aug 1 15:24:01 2008 +0200 19509 19510 Move some ALU instructions after the TEX instructions, so we can do something usefull while we are waiting for the texture values. 19511 19512commit 2ecdec4bafc97212dde4d6908ee4ccf618adc0e1 19513Author: Dennis Kasprzyk <onestone@opencompositing.org> 19514Date: Fri Aug 1 04:03:12 2008 +0200 19515 19516 Another uneeded SEM_AQUIRE. 19517 19518commit e93b5d1b80d6203f63543b7b678e2f1d9221b5b3 19519Author: Dennis Kasprzyk <onestone@opencompositing.org> 19520Date: Fri Aug 1 03:00:26 2008 +0200 19521 19522 Smarter usage of the texture semaphore. 19523 19524commit 286f7aa18f19bd21b263701adab38b736dbeda0f 19525Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19526Date: Thu Jul 31 17:01:44 2008 -0700 19527 19528 Enable bicubic filtering for all r5xx HW. 19529 19530commit 242aa4f630b4c60aefa3c12dc459a4d4d0b334a0 19531Author: Dennis Kasprzyk <onestone@opencompositing.org> 19532Date: Fri Aug 1 01:11:28 2008 +0200 19533 19534 Remove one constant. 19535 19536commit 413eacb0538977b0b3c92df074d40510f4539abc 19537Author: Dennis Kasprzyk <onestone@opencompositing.org> 19538Date: Fri Aug 1 00:52:09 2008 +0200 19539 19540 Heavy optimizations. 19541 19542commit a0c4a949cb49e5ac1e857aef08a8742b9f7b49da 19543Author: Dennis Kasprzyk <onestone@opencompositing.org> 19544Date: Thu Jul 31 21:19:18 2008 +0200 19545 19546 Remove uneeded negations. 19547 19548commit abb2b2e757d92591ab9277824b9d9746ba98f875 19549Author: Dennis Kasprzyk <onestone@opencompositing.org> 19550Date: Thu Jul 31 20:31:03 2008 +0200 19551 19552 Implement LPR in one instruction. 19553 19554commit c370b74bec13194573348583c38adf710b880e79 19555Author: Dennis Kasprzyk <onestone@opencompositing.org> 19556Date: Thu Jul 31 19:50:49 2008 +0200 19557 19558 Set helper texture filter correctly. 19559 19560commit 5e85d5a5d0c50b29086ec0c219c8b52d25dbc2e9 19561Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19562Date: Wed Jul 30 19:45:15 2008 -0700 19563 19564 Update bicubic tables. 19565 Now including the 2048x1 texture as well. 19566 19567commit 17e5e9573e59c3d82d51c261b9c5005f6aec7d43 19568Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19569Date: Wed Jul 30 11:57:25 2008 -0700 19570 19571 Force R580-only for bicubic. 19572 Initial reading of docs suggest RV560 and RV570 can't handle it, 19573 but they're welcome to test. 19574 19575commit 83b52473d0e7102265918b07be053fcda17a14b4 19576Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19577Date: Wed Jul 30 01:03:57 2008 -0700 19578 19579 Bump bicubic cutoff to R580. 19580 RV530 chipsets can't handle fullscreen bicubic... 19581 19582commit 69a4998d1286bcdd7bfe874cd5628bc1cc232bae 19583Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19584Date: Wed Jul 30 01:03:46 2008 -0700 19585 19586 Oops, made a mistake with vertices. 19587 19588commit a4ec30a677906ec2ff9824c7ddca586655f6d1a8 19589Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19590Date: Wed Jul 30 00:54:32 2008 -0700 19591 19592 Merge upstream changes to vertices, and also add Xv attributes for textured video, including bicubic filtering. 19593 19594commit f3b81c7582aed307fa44e134ee161cd8a3158657 19595Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19596Date: Mon Jul 28 19:50:10 2008 -0700 19597 19598 Fix constants. 19599 19600commit 7dbb7023ba023ec1a38be63af9c9f49e40222b7b 19601Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19602Date: Mon Jul 28 17:30:41 2008 -0700 19603 19604 Finally got the fragment program fully working for bicubic filtering on r5xx. 19605 19606commit 0e4dd73b9ebc6f608eeff945b4d463a00c02e07c 19607Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19608Date: Mon Jul 28 12:16:50 2008 -0700 19609 19610 More bicubic FP buggies. 19611 19612commit a87647e0c27e0950f4d0d8203a1242a994ad3419 19613Author: Dennis Kasprzyk <onestone@compiz-fusion.org> 19614Date: Sun Jul 27 10:43:01 2008 -0700 19615 19616 Fix texture size, texture filter, vertex offsets, etc. 19617 19618commit 232aa3e943fef4c4037b255c3b64a0aaff90ab5c 19619Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19620Date: Sat Jul 26 10:56:56 2008 -0700 19621 19622 Make vertices emit properly. 19623 *bangs head against wall* 19624 19625commit 8c84f67b93d926095633830aa8d95930a48b1c7b 19626Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19627Date: Sun Jul 20 14:25:44 2008 -0700 19628 19629 Try to get tex coords from the VAP to RS properly when bicubic filtering is enabled. 19630 I'm soo bad at this... :c 19631 19632commit a4a7d5f5967c51c394229de5eccaec44cfec8f50 19633Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19634Date: Sat Jul 19 13:02:02 2008 -0700 19635 19636 Upload pixel shader to card for r5xx. This was ridiculous. Also it doesn't work yet... 19637 19638commit 20c1db2d7c110ab5c1117a57b169baa1ab070518 19639Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19640Date: Fri Jul 18 13:48:08 2008 -0700 19641 19642 Upload bicubic filter to card. This was a LOT easier than I had feared, to be honest. 19643 19644commit a760e628134c6d7d42ec3c98118b6e5f6fcd3e7f 19645Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19646Date: Fri Jul 18 12:21:20 2008 -0700 19647 19648 Allocate memory for the bicubic filter texture. 19649 19650commit b6c9e2bb5365de82315c6814f915e57b0c4fa444 19651Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19652Date: Fri Jul 18 11:24:14 2008 -0700 19653 19654 Fixed typos in the bicubic texture tables. Whoops, looks like I'm still asleep. 19655 19656commit ca51f4f37e1dbf53bf7ffc0e8f612e9609e11209 19657Author: Corbin Simpson <MostAwesomeDude@gmail.com> 19658Date: Fri Jul 18 11:06:34 2008 -0700 19659 19660 Add bicubic texture table, as well as the script used to (re)generate it. To regenerate, just run "python bicubic_table.py > bicubic_table.h". 19661 19662commit 1cf7a5494fa94e8d9f30f9b2905dfbe6d4faa445 19663Author: Bryce Harrington <bryce@bryceharrington.org> 19664Date: Wed Aug 20 09:46:59 2008 -0400 19665 19666 radeon: Fix pasto in connector table setup for vga powerbooks 19667 19668 fixes bug 17214 19669 19670commit a55e85f742d1334bf88e4681e553f025d2de38df 19671Author: Michel Dänzer <michel@tungstengraphics.com> 19672Date: Thu Aug 14 15:21:51 2008 -0400 19673 19674 Make sure video offerlay offsets don't exceed the hardware limit of 128 MB. 19675 19676 Always set the overlay base address such that the buffer offsets are as small 19677 as possible. This could still break in theory if the buffers were more than 19678 128 MB apart, but in reality this can't happen ATM because we always allocate 19679 a single memory area for all buffers. 19680 19681 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=16845 . 19682 19683commit 92ee21df344a989778e37369c7beb3904a00ead6 19684Author: Alex Deucher <alexdeucher@gmail.com> 19685Date: Thu Aug 14 14:49:45 2008 -0400 19686 19687 PLL adjustments 19688 19689 Seems higher dotclocks prefer a higher FB div. 19690 Someone with a lot of should try and find out where 19691 the div sweet spots are for various dotclock ranges. 19692 fixes bug 17125 19693 19694commit 0d5e0347af4322713075193154b8a348de4a0b52 19695Author: Alex Deucher <alexdeucher@gmail.com> 19696Date: Wed Aug 13 14:17:34 2008 -0400 19697 19698 Remove reset of 3D scissor registers when using the CP in the ddx 19699 19700 They should only affect 3D and init3d() should take care of that case 19701 noticed by libv on IRC. 19702 19703commit 7e456e9c427d48151b166bcac08c3e8a8b180f01 19704Author: Dave Airlie <airlied@linux.ie> 19705Date: Tue Aug 12 08:18:30 2008 +1000 19706 19707 radeon: add 0x9441 for hd 4870 x2 19708 19709 thanks to Michael Larabel for testing 19710 19711commit 4dbdeea7c9316575fba26b41fd347452e42cdcf2 19712Author: Alex Deucher <alexdeucher@gmail.com> 19713Date: Fri Aug 8 17:39:48 2008 -0400 19714 19715 Further cleanup and unification of i2c code 19716 19717 - unify the ddc and i2c code 19718 - add gpio mask support for legacy chips 19719 - remove the magic gpio dance for ancient monitors 19720 (if you have an ancient monitor that ddc stops 19721 working on let me know). This should speed up DDC 19722 on legacy chips. 19723 19724 -- radeon sw gpio i2c -- 19725 4 sets of gpio registers for clk and data and corresponding bit masks 19726 mask - locks the gpio for use by sw 19727 get - reads the value off the gpio pad 19728 put - sets the gpio direction to output 19729 a - "other stuff" On legacy chips you clear them if you want 19730 to use a gpio for i2c. In some cases they are used for the 19731 output value when the gpio in the output state. 19732 19733commit 268c848130ec1770bb645a74197b6aca7fc95abc 19734Author: Alex Deucher <alexdeucher@gmail.com> 19735Date: Fri Aug 8 15:50:07 2008 -0400 19736 19737 Fix VT switching on M6 chips 19738 19739 Some M6 chips have a faulty MEM_SIZE register that in 19740 some cases reports 0 on 8 MB cards. On EnterVT we check 19741 the MEM_SIZE reg as a check to see if the card is posted or 19742 not. Since this reg returns 0, the driver attempts to post 19743 the card which can lead to a hang. Switch this to check if 19744 either crtc is active as is done in the bios init code. 19745 fixes bug 13994 19746 19747commit 33f88f7fc90d9d93fdcbba9ad59dd70a6596bc3f 19748Author: Alex Deucher <alexdeucher@gmail.com> 19749Date: Thu Aug 7 17:57:07 2008 -0400 19750 19751 RS4xx DDC fixes take 3 19752 19753commit df0d1ef53100f0a19c5b5fdc349f5186c8d9bd87 19754Author: Alex Deucher <alexdeucher@gmail.com> 19755Date: Wed Aug 6 19:13:56 2008 -0400 19756 19757 RS4xx: Fix up ddc gpio 19758 19759 - I mixed up the regs previously 19760 19761commit 18429390440a829fb24ed3afd99ccf8278138496 19762Author: Alex Deucher <alexdeucher@gmail.com> 19763Date: Wed Aug 6 00:35:15 2008 -0400 19764 19765 Remove un-needed dac check for single crtc cards 19766 19767 Connector tables should be setup properly already and this 19768 check breaks systems with 2 DACs and 1 crtc like the es1000. 19769 19770commit 001c535687d8588873037ee5363d0a709f44b418 19771Author: Alex Deucher <alexdeucher@gmail.com> 19772Date: Tue Aug 5 17:14:48 2008 -0400 19773 19774 IGP: fix typo in IGP quirk handling 19775 19776commit eb65ddf70d182b6457e1ef5ebb820456039e8f6d 19777Author: Alex Deucher <alexdeucher@gmail.com> 19778Date: Tue Aug 5 15:01:33 2008 -0400 19779 19780 IGP: attempt to get external TMDS working 19781 19782 RS4xx chips have a tmds init table in the mobile info table 19783 19784commit 8b8990917809b9a35c6e9c1b9e3b12ff81c6dbb3 19785Author: Alex Deucher <alexdeucher@gmail.com> 19786Date: Tue Aug 5 12:16:06 2008 -0400 19787 19788 IGP: Updated quirk handling 19789 19790 - Hopefully finally fix DDC on RS4xx chips 19791 - RS3xx/RS4xx do not have internal TMDS 19792 - general quirk handling cleanup 19793 - we don't currently support mm gpio for external TMDS 19794 so return false. 19795 19796commit cb0deba5412a575d36f2f99377120b123506c946 19797Author: Calvin Fong <hoiwai930@gmail.com> 19798Date: Mon Aug 4 02:04:15 2008 -0400 19799 19800 IGP: add support for NTSC tv-out on legacy IGP chips 19801 19802 Patch from Calvin with some cleanups from me. 19803 No support for PAL yet. 19804 19805commit 942b18aca91819fa65d853cd15ffd1cd720cbd68 19806Author: Dave Airlie <airlied@redhat.com> 19807Date: Mon Aug 4 14:38:21 2008 +1000 19808 19809 radeon: make r600 use i2c table lookup for ddc. 19810 19811 This may fix DDC on rv770 cards. 19812 19813commit 1f3eee3682f3598a303c9c3accfbe01b245cacf9 19814Author: Alex Deucher <alexdeucher@gmail.com> 19815Date: Tue Jul 29 20:29:32 2008 -0400 19816 19817 Change prim types for exa and textured video to help avoid tearing 19818 19819 - r1xx - switch from tri fan to rect list 19820 - r2xx/r3xx/r4xx/r5xx - switch from tri fan to quad list 19821 19822commit 5b5441f8cc119db0d1e03dd35bd06015a26270dd 19823Author: Brice Goglin <bgoglin@debian.org> 19824Date: Tue Jul 29 23:31:14 2008 +0200 19825 19826 Increase default GART size to 32MB on >=R300 19827 19828commit 0a505297f09aefb1b4432176a263bfdf6f256f77 19829Author: Alex Deucher <alexdeucher@gmail.com> 19830Date: Mon Jul 28 17:12:41 2008 -0400 19831 19832 Fix error in driver connector table for powerbook w/ vga 19833 19834commit d5799ac53c3e1b2ca1da97780b171a44060c3aad 19835Author: Alex Deucher <alexdeucher@gmail.com> 19836Date: Mon Jul 28 11:09:10 2008 -0400 19837 19838 Add quirk for oem x300 card 19839 19840 - debian bug 492457 19841 19842commit 64dbadddcf6d069c0472f37afeab89d3e31e937d 19843Author: Owain Gordon Ainsworth <oga@openbsd.org> 19844Date: Mon Jul 28 10:40:58 2008 +1000 19845 19846 radeon: don't call ioctl unless DRI is initialised 19847 19848commit 1c5858484da4fb1c9bc3ac3b4d7a97863ab99730 19849Author: Alex Deucher <alexdeucher@gmail.com> 19850Date: Mon Jul 21 23:47:45 2008 -0400 19851 19852 First pass at InitDispBandwidth() for AVIVO chips 19853 19854 - support for LB allocation 19855 - MC priority bumps for display1/2 on RV515 variants and RS690 19856 If you are having display underflow problems (flickering on sides of 19857 screen in high res modes, etc.) on RV515 or RS690 boards, try setting: 19858 Option "DisplayPriority" "HIGH" in your config. 19859 - still no support for full display watermark programming yet 19860 19861 Something similar might be useful in rhd as well. 19862 19863commit b0378bb145c8a915c943bef7d17f2cdecfccc891 19864Author: Alex Deucher <alexdeucher@gmail.com> 19865Date: Mon Jul 21 13:47:09 2008 -0400 19866 19867 Interlaced mode fixups for AVIVO chips 19868 19869commit c18fad622a3c4f9572051120d83af68b625b5686 19870Author: Alex Deucher <alexdeucher@gmail.com> 19871Date: Mon Jul 21 10:36:48 2008 -0400 19872 19873 Clear display priority bits before resetting them 19874 19875commit dc231ff8e063313d2bcf5acccad67a9f8a7e3314 19876Author: Alex Deucher <alexdeucher@gmail.com> 19877Date: Mon Jul 21 10:30:41 2008 -0400 19878 19879 Clean up legacy display watermark setup 19880 19881 - makes crtc1 and crtc2 watermark setup independant. 19882 - fixes the case where only crtc2 is active 19883 19884commit f9034214f070fe3054cd9cefd7034252de234f38 19885Author: Michel Dänzer <michel@tungstengraphics.com> 19886Date: Mon Jul 21 09:09:02 2008 +0200 19887 19888 Call DRM modeset ioctl after the IRQ has been initialized. 19889 19890 This lets the DRM know it can safely disable the vblank interrupts. 19891 19892commit ecb6347a3d7071890600c98c2addef3a8ca260ee 19893Author: Alex Deucher <alexdeucher@gmail.com> 19894Date: Sat Jul 19 11:34:16 2008 -0400 19895 19896 Add oem quirk for external tmds setup on Dell Inspiron 8600 19897 19898 Noticed by fnord42 on IRC. 19899 19900commit df53d12a06fad41f00cff849458cb358ab5e2098 19901Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 19902Date: Thu Jul 17 14:37:05 2008 +1000 19903 19904 Fix console switch on R500 19905 19906 This patch fixes the console switch for me on R5xx. 19907 19908 There are two aspects to it: 19909 19910 - Fix the ordering of avivo_restore() to better match what's 19911 happening in the driver & ATOM, properly locking/unlocking and 19912 only enabling the CRTCs after everything has been properly 19913 programmed. 19914 19915 - Don't ASIC_INIT if the card has any CRTC enabled. This is the 19916 best I came up with for avoiding spurrious ASIC_INIT on cards that 19917 -are- POSTed but don't have the BIOS coming from c0000 on x86. The 19918 problem with spurrious ASIC_INIT is that we do it before we do 19919 RADEONSave(), so that screws up the console switch. 19920 19921 Note that I think we also should save/restore the palette, I don't think 19922 we do. right now, it's a minor issue for me because I fixed offb to be 19923 able to set it on AVIVO's but it might still have to be done in the long 19924 run. 19925 19926 Tested with a VGA analog setup on DACA and a DVI setup on TMDS 0. I 19927 haven't tested any other combo but that should get us going. 19928 19929 Cheers, 19930 Ben. 19931 19932 Signed-off-by: Dave Airlie <airlied@redhat.com> 19933 19934commit e8c07270e1fc2ee455320370692aaf90abac2773 19935Author: Alex Deucher <alexdeucher@gmail.com> 19936Date: Tue Jul 15 17:50:32 2008 -0400 19937 19938 ATOM: fix for r4xx cards with no VRAM_Usage tables 19939 19940 fixes bug 15954 19941 19942commit da41e71efd41907d9347a902720bce60b32550e5 19943Author: Alex Deucher <alexdeucher@gmail.com> 19944Date: Mon Jul 14 13:21:38 2008 -0400 19945 19946 ATOM: Add support for UpdateCRTC_DoubleBufferRegisters cmd table 19947 19948 locks/unlocks the crtc/grph/mode regs before updating the crtc 19949 19950commit a5e0cf13dc7ace6cf0e44e18b73b9a9266e740ab 19951Author: Julien Cristau <jcristau@debian.org> 19952Date: Sun Jul 6 12:17:28 2008 +0200 19953 19954 Link with -lpciaccess and -ldrm if needed 19955 19956 This makes sure the driver ends up with a DT_NEEDED reference to 19957 the libraries it's using. 19958 19959commit f8da849fd89322a54cc4d722767da755358eab70 19960Author: Michel Dänzer <michel@tungstengraphics.com> 19961Date: Sun Jul 13 11:31:03 2008 +0200 19962 19963 Only declare local variable when it's used. 19964 19965commit a6db4dc65aec560f3b1f37ff73ec6b0bd8207295 19966Author: Dave Airlie <airlied@linux.ie> 19967Date: Sat Jul 12 18:47:07 2008 +1000 19968 19969 atombios: fix typo in mode conversion for panel 19970 19971commit 19a0d5879851eff23a3e7d5cdea55bd784051115 19972Author: Dave Airlie <airlied@linux.ie> 19973Date: Sat Jul 12 17:04:25 2008 +1000 19974 19975 atombios: use macro to get offset into table 19976 19977commit a9746114369d516072d841ec356ec3ba3d0be71a 19978Author: Dave Airlie <airlied@linux.ie> 19979Date: Sat Jul 12 10:46:36 2008 +1000 19980 19981 pciid: add radeon HD3850. 19982 19983 pci id from legume on #radeon 19984 19985commit ab14f725676e4d4e45278c64b03fe2d328a3e439 19986Author: Alex Deucher <alexdeucher@gmail.com> 19987Date: Fri Jul 11 19:05:00 2008 -0400 19988 19989 R300: NUM_FPU adjustments for VAP_CNTL 19990 19991commit 810c28cc2660b73e03e4d27be97988fc5321a06f 19992Author: Alex Deucher <alexdeucher@gmail.com> 19993Date: Fri Jul 11 15:31:57 2008 -0400 19994 19995 Make sure cursor offsets are properly aligned when using EXA 19996 19997commit 7e67d0163579a44f104e8f354a929ac9b33e4c21 19998Author: Alex Deucher <alexdeucher@gmail.com> 19999Date: Fri Jul 11 14:30:17 2008 -0400 20000 20001 Fix cursor with multi-head and rotation 20002 20003commit 9086d008fb8c3cde0633b90ce19ffbf4eded388d 20004Author: Alex Deucher <alexdeucher@gmail.com> 20005Date: Tue Jul 8 09:57:04 2008 -0400 20006 20007 Add quirk for Falcon NW laptop 20008 20009commit 61f82ace0210251beb0bcc492218a75a193e1deb 20010Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 20011Date: Tue Jul 8 21:58:43 2008 +1000 20012 20013 atombios: add support for other endians. 20014 20015 This is a cleaned up (in as much as atombios can be..) of benh's patch. 20016 20017 airlied - removed benh's debugging for now, it might need to be put back later.. 20018 20019commit b4d1a47b5c6eafda5e274398eebe2701b030f22e 20020Author: Dave Airlie <airlied@linux.ie> 20021Date: Tue Jul 8 11:09:01 2008 +1000 20022 20023 rv770 initial support 20024 20025 adds pci ids and one register from AMD code 20026 20027commit 7ae4cec8cc8c90aee5dc4fa7abcce22321d4f4eb 20028Author: Roland Scheidegger <sroland@tungstengraphics.com> 20029Date: Mon Jul 7 14:39:47 2008 -0400 20030 20031 clamp tex coords (r100/r200) for textured video 20032 20033 fixes bug 14904 20034 20035commit 879f9a37856642b337e3c0233e38fd5443338eef 20036Author: Alex Deucher <alexdeucher@gmail.com> 20037Date: Fri Jul 4 13:54:49 2008 -0400 20038 20039 Revert "rotate: fix rotation in conf file." 20040 20041 This reverts commit 22c6ca0d5d226c725ba448e5db63fdba9e32332c. 20042 20043 There seems to be an ordering issue in that causes a nasty hang 20044 if the DRI is enabled. This needs more investigation. 20045 20046commit 22c6ca0d5d226c725ba448e5db63fdba9e32332c 20047Author: Dave Airlie <airlied@redhat.com> 20048Date: Thu Jul 3 20:14:59 2008 +1000 20049 20050 rotate: fix rotation in conf file. 20051 20052 need to set modes after init accel architecture 20053 20054 reported-by: Christoph Bartoschek <bartoschek@or.uni-bonn.de> 20055 20056commit ce1a3edd52b647744cfa4433301befb437d211bb 20057Author: Dave Airlie <airlied@redhat.com> 20058Date: Thu Jul 3 20:05:54 2008 +1000 20059 20060 radeon: drop all use of CPMode. 20061 20062 We never test the other codepath and I don't think I've ever recommended it 20063 for anyone. 20064 20065commit c037b4ce8769ad840a257e22b1e4ad58b8ed96fa 20066Author: Wolke Liu <wolke.lui@amd.com> 20067Date: Tue Jul 1 10:45:39 2008 -0400 20068 20069 Add pci id for FireMV 2400 20070 20071commit 72992668ca96717bf69bcea95a5deeaf7c57e826 20072Author: Alex Deucher <alex@botchco.com> 20073Date: Mon Jun 30 14:09:09 2008 -0400 20074 20075 RS4xx: enable the DRI by default on all variants 20076 20077 Some RS4xx chips had the DRI disabled by default. 20078 With the recent drm and ddx changes these chips have been 20079 reported to work now with the DRI. 20080 20081commit 0378c8ed88c829f09b5c6d51314325eaf2685fdb 20082Author: Michel Dänzer <michel@tungstengraphics.com> 20083Date: Sat Jun 28 14:06:45 2008 +0200 20084 20085 Restore versioning of interface for Mesa DRI driver. 20086 20087 This doesn't have anything to do with the package version. 20088 20089commit cfaa23d925e3c062cf87ea844566ac11ea02d69e 20090Author: Alex Deucher <alex@botchco.com> 20091Date: Fri Jun 27 20:36:54 2008 -0400 20092 20093 RADEON: IGP VGA fixes take 2 20094 20095 It seems only RS300/350/400/480 are missing the primary dac 20096 See bug 15708 20097 20098commit 31c27ffcb3c4c5334cf726ecd4e293a678b2a1ea 20099Author: Alex Deucher <alex@botchco.com> 20100Date: Fri Jun 27 20:29:04 2008 -0400 20101 20102 Revert "IGP: attempt to fix VGA on IGP chips" 20103 20104 This reverts commit e78e8a21b4040cd7f1983c241c860d9209398396. 20105 20106commit af79a77e6d8206b18bba97cf4e0770921dc439a7 20107Author: Brice Goglin <Brice.Goglin@ens-lyon.org> 20108Date: Fri Jun 27 17:09:13 2008 +0200 20109 20110 use PACKAGE_VERSION_* for radeon version 20111 20112 now radeon shows verion 6.9.0 instead of 4.3.0. 20113 20114 RADEON_VERSION_MAJOR_TILED is dropped from ddxDriverMajorVersion 20115 since RADEON_VERSION_MAJOR is bigger now. 20116 20117commit c83fbdfa076c107012b7dfbbfbbb2feede00542b 20118Author: Alex Deucher <alex@botchco.com> 20119Date: Thu Jun 26 19:48:45 2008 -0400 20120 20121 Bump for 6.9.0 release 20122 20123commit 32e1d1daf4f68ea37624afaa2bc1ea7742e1163b 20124Author: Brad Smith <brad@comstyle.com> 20125Date: Thu Jun 26 10:22:03 2008 -0400 20126 20127 RADEON: sys/endian.h needs sys/types.h on BSDs 20128 20129 See bug 16512 20130 20131commit dca522355a9039eca6efaba3b36397b246800f94 20132Author: Brad Smith <brad@comstyle.com> 20133Date: Thu Jun 26 10:13:59 2008 -0400 20134 20135 RADEON: fix copy/paste error in accel code 20136 20137commit f4292e110105910d0c19bf2db28e2682b27af2c7 20138Author: Dave Airlie <airlied@redhat.com> 20139Date: Fri Jun 20 15:31:43 2008 +1000 20140 20141 Revert "Revert "ATOM: disable TV-out for now"" 20142 20143 Oops I really didn't mean to do this, I was testing something and it slipped 20144 past. 20145 20146 This reverts commit dd18caa4b2efc430eaae0c4362b65641f9bef440. 20147 20148commit bb1cfcbfbc1ace1f257fc3bf6401501f1a6da827 20149Author: Alex Deucher <alex@botch2.com> 20150Date: Wed Jun 25 15:50:16 2008 -0400 20151 20152 RADEON: adjustments to Jerome's last commit 20153 20154 - Flush caches and wait for idle after drawing 20155 - Make sure 3D is idle too (after composite or textured video) 20156 20157commit 07be302f6dba61dd15918963fae66502230c74b7 20158Merge: 8c9b8de03 52459745e 20159Author: Jerome Glisse <glisse@freedesktop.org> 20160Date: Wed Jun 25 10:30:45 2008 +0200 20161 20162 Merge branch 'master' of git://anongit.freedesktop.org/xorg/driver/xf86-video-ati 20163 20164commit 8c9b8de0373797cba6f0a27e6b6f461e1070fef9 20165Author: Jerome Glisse <glisse@freedesktop.org> 20166Date: Wed Jun 25 10:28:26 2008 +0200 20167 20168 radeon: flush & wait for 2d & dma idle after 2d blit 20169 20170 This should help to avoid 2d & 3d engine to step on each 20171 other dma transaction. 20172 20173commit 52459745ec05de88adbc087e9566fe6d97ef424b 20174Author: Dave Airlie <airlied@linux.ie> 20175Date: Wed Jun 25 17:09:11 2008 +1000 20176 20177 ati: hopefully fix byteswap mess for those other OSes. 20178 20179commit 67a6ac0001bc9d062aa426384a11a41fa7a1c09a 20180Author: Alex Deucher <alex@botch2.com> 20181Date: Tue Jun 24 21:06:37 2008 -0400 20182 20183 bump for rc release 20184 20185commit bd68507d2d66e03d8bcde5f6e7ea9b2dbfe8b8a0 20186Author: Alex Deucher <alex@botch2.com> 20187Date: Tue Jun 24 20:59:58 2008 -0400 20188 20189 RADEON: warning fix 20190 20191commit 8e534d69c9b19fc085f7c5ca9e18f5ea04f6fc12 20192Author: Alex Deucher <alex@botch2.com> 20193Date: Tue Jun 24 20:08:35 2008 -0400 20194 20195 RADEON: cleanups 20196 20197 - fix some warnings 20198 - RS400 and RS480 are separate families now (update default tmds and dac2 20199 tables) 20200 20201commit e78e8a21b4040cd7f1983c241c860d9209398396 20202Author: Alex Deucher <alex@botch2.com> 20203Date: Tue Jun 24 19:53:28 2008 -0400 20204 20205 IGP: attempt to fix VGA on IGP chips 20206 20207 VGA has never worked on some IGP chips. While the chip only has 20208 one DAC, it appears to use a mix of Primary DAC and TVDAC controls. 20209 20210 See bug 15708 20211 20212commit faa4b4b8dbe9b8a452cfa60d53874350bb04e0cb 20213Author: Jiří Paleček <jpalecek@web.de> 20214Date: Tue Jun 24 19:20:50 2008 -0400 20215 20216 RADEON: remove extraneous line from new pll code 20217 20218commit 72feaa37ea07620f5f2ead438dbc72a1c8883cd3 20219Author: Jiří Paleček <jpalecek@web.de> 20220Date: Mon Jun 23 15:53:58 2008 -0400 20221 20222 RADEON: PLL tweaks 20223 20224 Patch from Jiří Paleček (see debian bug 465864) with some tweaks 20225 by me. 20226 20227 - abort rather than programming bad dividers if no pll dividers can be found 20228 - improve the pll selection algorithm 20229 - in general, prefer lower ref dividers 20230 20231 I've tested this patch on a wide variety of chips (r1xx-r6xx) and clocks. 20232 20233commit 9c2f909ea437a63a408d2398ecabe0b378dbb982 20234Author: Alex Deucher <alex@botch2.com> 20235Date: Mon Jun 23 10:38:15 2008 -0400 20236 20237 RADEON: adjust randr crtc/output prepare/commit ordering 20238 20239 This fixes some occasional mode change problems with multiple heads active. 20240 It seems radeons generally like to turn on the whole output/crtc setup 20241 in one shot. 20242 20243commit aea9bf75cf0774afd3e65fcf3fd3851f5fb21ca3 20244Author: Alex Deucher <alex@botch2.com> 20245Date: Sat Jun 21 10:57:05 2008 -0400 20246 20247 0x1002 0x5657 is actually RV410 20248 20249 See bug 14289 20250 20251 com_bios_scratch.diff 20252 20253commit 38ce8a984f96056b7345bcc06505ba27e0e6d5b4 20254Author: Dave Airlie <airlied@redhat.com> 20255Date: Fri Jun 20 14:16:22 2008 +1000 20256 20257 legacy: use entity MMIO for dpms as this can crossover between zaphod infos 20258 20259commit 1a7d9bc53512b0a5240176c949e6abf1ae2fb1fd 20260Author: Dave Airlie <airlied@redhat.com> 20261Date: Fri Jun 20 14:14:21 2008 +1000 20262 20263 atombios: use MMIO from the entity not the info 20264 20265commit ef624b88903b1a87ef5b6388e18291f75776b93d 20266Author: Alex Deucher <alex@botch2.com> 20267Date: Thu Jun 19 18:20:52 2008 -0400 20268 20269 RADEON: fix read past the end of an array 20270 20271commit cfe814a481d8cf2005d738a0ca9782f1ed4177f5 20272Author: Dave Airlie <airlied@linux.ie> 20273Date: Fri Jun 20 07:51:27 2008 +1000 20274 20275 r600: don't add fb offset here to make shadowfb work. 20276 20277 discovered on irc with wpwrak. 20278 20279commit 5b323a2cbbc412b6dd4e782438b64ee996558d05 20280Merge: dd18caa4b eed241553 20281Author: Dave Airlie <airlied@linux.ie> 20282Date: Fri Jun 20 07:49:16 2008 +1000 20283 20284 Merge branch 'master' of ssh://git.freedesktop.org/git/xorg/driver/xf86-video-ati 20285 20286commit eed241553748125e902c44dcc8cf8e0605be071b 20287Author: Alex Deucher <alex@botch2.com> 20288Date: Tue Jun 17 17:42:31 2008 -0400 20289 20290 radeon: X_ERROR -> X_WARNING for num gb pipes 20291 20292 This just means your drm is old, not fatal or anything 20293 20294commit dd18caa4b2efc430eaae0c4362b65641f9bef440 20295Author: Dave Airlie <airlied@linux.ie> 20296Date: Sat Jun 14 18:37:56 2008 +1000 20297 20298 Revert "ATOM: disable TV-out for now" 20299 20300 This reverts commit effa245914823371e052cd9aa1143a02350891e7. 20301 20302commit b416e97a1f16ef700ba1aaca317dee82b7a3cf64 20303Author: Alex Deucher <alex@botch2.com> 20304Date: Fri Jun 13 02:21:31 2008 -0400 20305 20306 RADEON: man page updates 20307 20308commit 88ce87d79069662de28ecb3c10c140919927d584 20309Author: Alex Deucher <alex@botchco.com> 20310Date: Thu Jun 12 15:10:47 2008 -0400 20311 20312 Bump for 6.9.0rc1 release 20313 20314commit ada41723fdc8cbeeda12bb4ae2d16e8d1ba215de 20315Author: Alex Deucher <alex@botch2.com> 20316Date: Wed Jun 11 14:29:36 2008 -0400 20317 20318 R3xx: use DSTCACHE_CTLSTAT rather than RB2D_DSTCACHE_CTLSTAT 20319 20320 According to the hw guys, you should use DSTCACHE_CTLSTAT to 20321 flush the 2D dst cache rather than RB2D_DSTCACHE_CTLSTAT. 20322 20323commit 3d70c398afa0831387af81559453b77c218a54f9 20324Author: Dave Airlie <airlied@redhat.com> 20325Date: Thu Jun 12 06:18:26 2008 +1000 20326 20327 atom: fix another endian define 20328 20329commit f1794e4c34d801371461afed37e7bb7d1fc79305 20330Author: Dave Airlie <airlied@redhat.com> 20331Date: Thu Jun 12 06:10:30 2008 +1000 20332 20333 atom: fixup includes in cd operations to get correct endian bits in atombios.h 20334 20335commit f29976d12cc2020cd80c90fa3fd4b25c031559ba 20336Author: Dave Airlie <airlied@ppcg5.localdomain> 20337Date: Wed Jun 11 19:33:02 2008 +1000 20338 20339 radeon: add lots of endian bits into the atombios parser 20340 20341commit 09eb220971b5d2bfd7d1ff6f552060967a133152 20342Author: Dave Airlie <airlied@ppcg5.localdomain> 20343Date: Wed Jun 11 19:21:22 2008 +1000 20344 20345 atombios: fixup atombios bitfields for endianness 20346 20347 reverse the bits for big endian 20348 20349commit b80a3e85bf9dbfc8d96a745876d75d38bea97c1a 20350Author: Alex Deucher <alex@botchco.com> 20351Date: Mon Jun 9 20:29:47 2008 -0400 20352 20353 R5xx: fix RADEONSetAgpBase() for R5xx chips 20354 20355 Is there any reason to still do this in the ddx? 20356 Maybe real old drms? 20357 20358commit ab7936708dfbee4d3676262f8010e001b73a4a38 20359Author: Alex Deucher <alex@botchco.com> 20360Date: Mon Jun 9 17:05:04 2008 -0400 20361 20362 RS4xx: RC410 chips are actually RS400 based 20363 20364commit a54ca3c699c103e2e3df87101dca929a3eee4377 20365Author: Dave Airlie <airlied@linux.ie> 20366Date: Thu Jun 5 18:31:53 2008 +1000 20367 20368 radeon: fix typo in memory calcs 20369 20370 noticed by ghepeu on irc. 20371 20372commit 6e4e6d2a8f29f92efc219dca24ea31d1f37d5a0f 20373Author: Alex Deucher <alex@botchco.com> 20374Date: Tue Jun 3 20:32:20 2008 -0400 20375 20376 RADEON: minor cleanups 20377 20378commit 9d3afbf5fa4110928a9f965df2733c79db92ea99 20379Author: Michel Dänzer <michel@tungstengraphics.com> 20380Date: Tue Jun 3 11:40:49 2008 +0200 20381 20382 Call DRM_IOCTL_MODESET_CTL ioctl to avoid problems with DRM post vblank-rework. 20383 20384commit effa245914823371e052cd9aa1143a02350891e7 20385Author: Alex Deucher <alex@botchco.com> 20386Date: Sun Jun 1 20:47:50 2008 -0400 20387 20388 ATOM: disable TV-out for now 20389 20390 It only sort of works in NTSC mode and isn't handled correctly 20391 in all cases yet. Until we sort out the remaining details leave it 20392 disabled. 20393 See bugs: 16186, 16178, 16185 20394 20395commit 8504c6b0e40477ee544ad7f5366d569bdc53d6f0 20396Author: Alex Deucher <alex@botch2.com> 20397Date: Fri May 30 11:19:03 2008 -0400 20398 20399 RADEON: update RADEONGetVRAMType() for newer chips 20400 20401commit 7cb695a329a9b543ad61af08c2d7d6eaf56de35e 20402Author: Alex Deucher <alex@botch2.com> 20403Date: Fri May 30 10:10:56 2008 -0400 20404 20405 RADEON: add Int10 option 20406 20407 On some radeons you can't read the bios without initializing int10. 20408 On ATOM-based secondary cards, intitializing int10 tends to hang the card. 20409 20410commit 714b2c63ec1c4a1410a6d521c03e9d1f90937c01 20411Author: Alex Deucher <alex@botch2.com> 20412Date: Thu May 29 02:31:00 2008 -0400 20413 20414 RADEON: don't enable legacy bios table based init yet 20415 20416 Needs more work and we need to figure out how best to decide 20417 when to use it. 20418 20419commit e20b08525a64888287ec4a369d8f7dbde95c655d 20420Author: Alex Deucher <alex@botch2.com> 20421Date: Wed May 28 14:43:47 2008 -0400 20422 20423 ATOM: fail of we are not able to get a bios image 20424 20425commit d4fa17b426b0cafac3184b2ea9b675e2ff154ee9 20426Author: Alex Deucher <alex@botch2.com> 20427Date: Wed May 28 09:50:02 2008 -0400 20428 20429 R1xx: Gamma fixes for overlay 20430 20431commit c7e6a50f60a04bbbf1c1ddbf4738a5f75c94f950 20432Author: Alex Deucher <alex@botchco.com> 20433Date: Wed May 28 09:05:19 2008 -0400 20434 20435 RADEON: only attempt to read vbios from legacy ISA space if it's primary 20436 20437commit faea008806802ec0e045754ec1eca492ebae320e 20438Author: Alex Deucher <alex@botch2.com> 20439Date: Tue May 27 18:36:01 2008 -0400 20440 20441 R3/4/5xx: use get_param to get the num_gb_pipes from the drm 20442 20443commit 965a5dbcd9dc4bf1cdd7f2bbdec15e9733b2e090 20444Author: Alex Deucher <alex@botch2.com> 20445Date: Tue May 27 16:48:41 2008 -0400 20446 20447 RADEON: improve support for secondary cards 20448 20449 this should fix bugs 16115, 16035 20450 20451commit 5f951a5573f0c7572230c9aa4d3f75d67f91ed71 20452Author: Alex Deucher <alex@botch2.com> 20453Date: Wed May 21 00:35:14 2008 -0400 20454 20455 ATOM: remove duplicate code 20456 20457commit edce33e87fb79a540d8c285f205d4c3f2c2bc9f4 20458Author: Dave Airlie <airlied@redhat.com> 20459Date: Wed May 28 06:43:40 2008 +1000 20460 20461 radeon: fix surface access on avivo chips. 20462 20463 This should fix VT switch on vesafb 20464 20465commit 307bd65a25ee425d8359dd30572b002ce2338f91 20466Author: Michael Babej <mbabej@redhat.com> 20467Date: Tue May 27 19:13:49 2008 +1000 20468 20469 r600: fix vt switch issue 20470 20471 RH BZ 441492 20472 20473commit bf48c9eb8ec592515be9d1732d60283af715674d 20474Author: Dave Airlie <airlied@linux.ie> 20475Date: Sat May 24 11:51:20 2008 +1000 20476 20477 atombios: for LVDS set use ref div as per legacy 20478 20479commit 04500c8419b1aeaeac7968492b82e1d7cad1d05b 20480Author: Alan Coopersmith <alan.coopersmith@sun.com> 20481Date: Tue May 20 20:34:31 2008 -0700 20482 20483 Strip ^M's from src/AtomBios/includes/ObjectID.h 20484 20485commit 521a0488d1c414209c3534dc242000faa332a441 20486Author: Alan Coopersmith <alan.coopersmith@sun.com> 20487Date: Tue May 20 20:33:37 2008 -0700 20488 20489 On non-gcc compilers, use C99's __func__ instead of gcc's __FUNCTION__ 20490 20491commit b7c80d0c86646105d2bce5d4d59ba6c45aa7cafc 20492Author: Tobias Diedrich <ranma+freedesktop@tdiedrich.de> 20493Date: Tue May 20 19:16:54 2008 -0400 20494 20495 R200/R300: fix gamma setup for overlay 20496 20497 More pending. See bug 16001 20498 20499commit 69423482e2e94637142a9ba675589a1449a346a8 20500Author: Alex Deucher <alex@botch2.com> 20501Date: Tue May 20 18:57:13 2008 -0400 20502 20503 RADEON: cleanup connection detection and remove duplicate code 20504 20505commit 55e4469f59c82bb5762673de5f3f27d18b0bd9a3 20506Author: Alex Deucher <alex@botch2.com> 20507Date: Tue May 20 17:46:58 2008 -0400 20508 20509 RADEON: enable cloning on multi-crtc cards 20510 20511commit 12f185634071980041aaac6265d89708b181b215 20512Author: Alex Deucher <alex@botch2.com> 20513Date: Tue May 20 16:05:41 2008 -0400 20514 20515 ATOM: Ignore invalid connector entries 20516 20517commit 19e97f74e39fc2b35727708ac429de33d0b70162 20518Author: Egbert Eich <eich@freedesktop.org> 20519Date: Tue May 20 11:16:15 2008 +0200 20520 20521 Change RMX code to follow the programming algorithm suggested by ATI. 20522 20523 ATI provides the following algorithm to calculate the RMX scaling ratios 20524 in its programming specs: 20525 when RMX_AUTO_RATIO_HORZ_INC set to 1,Horizontal auto ratio result trucated, 20526 and then incremented by 1. 20527 Horz_Ratio = ( ((Active display width in characters (including overscan) + 1) 20528 / (Panel width in characters)) x 4096 + 1 ) 20529 else 20530 Horz_Ratio = ( ((Active display width in characters (including overscan)) 20531 / (Panel width in characters)) x 4096 + 1 ) 20532 20533 when RMX_AUTO_RATIO_VERT_INC set to 1, Vertical auto ratio result trucated, 20534 and then incremented by 1. 20535 Vert_Ration = ( ((Active display width in characters (including overscan) + 1) 20536 / (Panel width in characters)) x 4096 + 1) 20537 else 20538 Vert_Ration = ( ((Active display width in characters (including overscan)) 20539 / (Panel width in characters)) x 4096 + 1) 20540 20541 This patch implements this behavor. Additionally it avoids the use of floats. 20542 20543commit a4f3d0088ba763ed8eab1d331959b5ecde8262e8 20544Author: Alex Deucher <alex@botch2.com> 20545Date: Tue May 20 11:42:53 2008 -0400 20546 20547 R3/4/5xx: fixup logic from last commit 20548 20549 Spotted by otaylor and MrCooper 20550 20551commit 130e55738047f2a073bcc47e3e1400f7b694a81b 20552Author: Alex Deucher <alex@botch2.com> 20553Date: Tue May 20 11:32:42 2008 -0400 20554 20555 R3/4/5xx: fix EXA rotation 20556 20557 xrandr uses PictOpSrc for rotation which we were falling back on since 20558 render semanties require alpha=0 for REPEAT_NONE when there is no alpha 20559 channel and there is a transform. If the dst has no alpha channel we 20560 should be ok I think. 20561 20562 Otayler and I discussed this on IRC. the more general fix would be 20563 to clip the dst to the src and mask for bounded ops like in/add/over 20564 in the pixel exact transform case. 20565 20566commit cc9f510770700228e5d597c872e926a4e99bd950 20567Author: Corbin Simpson <MostAwesomeDude@gmail.com> 20568Date: Wed May 14 16:49:01 2008 -0400 20569 20570 RADEON: missing stdint.h includes 20571 20572commit 71fa57f871dba03260dba2180ce1dab44048ac1a 20573Author: Alex Deucher <alex@cube.(none)> 20574Date: Wed May 14 13:36:38 2008 -0400 20575 20576 Add RS600 support 20577 20578commit 582c1a1b2c7b1032e9f9f54ca36100c57f580c5c 20579Author: Alex Deucher <alex@cube.(none)> 20580Date: Tue May 13 20:50:25 2008 -0400 20581 20582 RS4xx: Split out RS400 and RS480 as separate families 20583 20584 RS400 (intel based IGP) and RS480 (AMD based IGP) have different 20585 MC setups and need to be handled differently 20586 20587commit 708e7c98f636738fbcc47a597bc94b309a4dc1c4 20588Author: Michel Dänzer <michel@tungstengraphics.com> 20589Date: Tue May 13 11:00:38 2008 +0200 20590 20591 Reinstate CARD* types that are part of external interfaces. 20592 20593 The compiler pointed this out... 20594 20595commit 908b7b940e3ef296836bc94680ebb8ab67650e31 20596Author: Matt Turner <mattst88@gmail.com> 20597Date: Mon May 12 12:06:33 2008 -0400 20598 20599 Replace CARD{8,16,32} with uint{8,16,32}_t 20600 20601 As has been done with xf86-video-intel, replace all CARD* datatypes with 20602 uint*_t datatypes available from stdint.h. 20603 20604commit 87e66ce76430890ab4939ffcd42f72b9288eb598 20605Author: Avi Rozen <avi.rozen@gmail.com> 20606Date: Mon May 12 08:33:38 2008 -0400 20607 20608 RADEON: fix lockup on start 20609 20610 see debian bug 480312 20611 20612commit 94bf8f01bd43cb103fffecfe04d04a214f892baf 20613Author: Dave Airlie <airlied@redhat.com> 20614Date: Mon May 12 20:02:51 2008 +1000 20615 20616 radeon: fix set_cursor_colours remove fb offset 20617 20618commit 1d0f1d31e2ed1d91ee87cb3e02ce48c8c07aa418 20619Author: Dave Airlie <airlied@redhat.com> 20620Date: Mon May 12 15:44:35 2008 +1000 20621 20622 radeon: rs485 vs rs485m.. mobile? non-mobile? bi? 20623 20624 So it appears we have in the wild two chipsets with the same PCI ID (0x1002:0x5974) 20625 that are mobile and non-mobile. 20626 20627 the RH bug references is a desktop system. 20628 The Dell Vostro 1100 also has this chipset with LVDS bits. 20629 20630commit 25e0c3945a51ae8c14b3a847ec75a256e1397f24 20631Author: Alex Deucher <alexdeucher@gmail.com> 20632Date: Mon May 12 10:28:33 2008 +1000 20633 20634 radeon: add initial support for cloning outputs. (single-CRTC cards only) 20635 20636 airlied - This code was originally written by Alex, and I've modified it to 20637 only run on single-CRTC cards for now until we can test it some more. 20638 20639commit 94405eb1c9e4e0ababc6aef03b753d6ed9eb7838 20640Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 20641Date: Fri May 9 05:28:44 2008 -0400 20642 20643 RADEON: fixup ifdef from last commit 20644 20645commit 18e20bc22a55ecfee9798c01079d7b24d19f0051 20646Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 20647Date: Thu May 8 08:45:51 2008 -0400 20648 20649 RADEON: disable MMIO paths for EXA composite/texvid on IGP/R5xx 20650 20651 The MMIO paths eventually lead to a hang on r5xx/IGP. I haven't 20652 been able to find out why yet. 20653 20654commit c3532268875fd24e6519bea2fb1b814d612bbdb4 20655Author: Dave Airlie <airlied@linux.ie> 20656Date: Wed May 7 02:37:18 2008 +1000 20657 20658 radeon: fix zaphod EXA with texture video 20659 20660commit ffc437f3606ab8ceba1ff152e4bb08988a58b54c 20661Author: Dave Airlie <airlied@linux.ie> 20662Date: Wed May 7 02:30:28 2008 +1000 20663 20664 avivo: fix zaphod cursor in theory 20665 20666commit e36ef14e3a1087e1fe41baa26ade2937f396001f 20667Author: Dave Airlie <airlied@linux.ie> 20668Date: Wed May 7 01:39:28 2008 +1000 20669 20670 radeon: fix textured-xv on zaphod 20671 20672commit 8fc19bee27c0f151d2ab3354f6ac0992b358436d 20673Author: Dave Airlie <airlied@linux.ie> 20674Date: Wed May 7 01:38:24 2008 +1000 20675 20676 radeon: zaphod: fix render accel for EXA 20677 20678commit fc41b9042a5220a8419cc7b69ca3850cae6b903c 20679Author: Dave Airlie <airlied@linux.ie> 20680Date: Wed May 7 01:32:23 2008 +1000 20681 20682 radeon: fix EXA pixmap offset on zaphod 20683 20684commit 4568cb820d567c6909a4be956d7e79b91232535e 20685Author: Dave Airlie <airlied@linux.ie> 20686Date: Wed May 7 01:19:39 2008 +1000 20687 20688 radeon: zaphod fixes for pciaccess not allowing multiple MMIO maps 20689 20690commit ca81fa086b21633a7fd926833fb6d1d4fa080646 20691Author: Dave Airlie <airlied@linux.ie> 20692Date: Wed May 7 01:12:01 2008 +1000 20693 20694 radeon: zaphod fix for cursor on second head 20695 20696 We don't need to add fbOffset here as the mmap we have of the framebuffer 20697 starts half way. 20698 20699commit 24b60c8965f6a0b3f0c2bb1e7236b4d6642c5918 20700Author: Julien Cristau <jcristau@debian.org> 20701Date: Fri May 2 15:30:45 2008 -0400 20702 20703 Add a test for __GLIBC__ to CD_Common_Types.h. 20704 20705 Atombios redefines the standard types but the definitions conflict 20706 with the ones from glibc (causes build failures on GNU/Hurd 20707 and GNU/kFreeBSD). 20708 20709commit f051359ac09c6b9416e39b9ca7d9dc0880aa1557 20710Author: thegraze <thegraze@googlemail.com> 20711Date: Fri May 2 14:02:16 2008 -0400 20712 20713 ATOM: add support for DragonFlyBSD 20714 20715commit 3d469cbc3225d890a895dac7cbc1ab7e08054b48 20716Author: Alex Deucher <alex@cube.(none)> 20717Date: Wed Apr 30 18:33:04 2008 -0400 20718 20719 RADEON: lock the cursors when updating 20720 20721 this should fix occasional corruption seen when updating 20722 the cursor. 20723 20724commit 445b71021843665ba32f37b2ce5c9d2857c07cc7 20725Author: Alex Deucher <alex@cube.(none)> 20726Date: Tue Apr 29 21:01:41 2008 -0400 20727 20728 RADEON: assorted fixes 20729 20730 - free rotate pixmaps on VT switch 20731 - save crtc/output status so we only turn on 20732 crtcs/outputs if they are off 20733 - show/hide cursors when changing modes 20734 20735commit 070cce5255a5c311f9d8b85ec54bd56655014933 20736Author: Stephan Wolf <stephan@letzte-bankreihe.de> 20737Date: Mon Apr 28 11:26:37 2008 -0400 20738 20739 R3xx+: further fix for IGP chips 20740 20741 see bug 15538 20742 20743commit 211e0041c7fc2df494b77428553943a2b526ee4e 20744Author: Alex Deucher <alex@cube.(none)> 20745Date: Sun Apr 27 21:08:00 2008 -0400 20746 20747 IGP: fix EXA composite corruption 20748 20749commit 656b06bdde129ca4fc370f5a2cf7311c9179b0ff 20750Author: Alex Deucher <alex@botch2.com> 20751Date: Sun Apr 27 20:20:49 2008 -0400 20752 20753 RADEON: remove duplicate register define 20754 20755 Also add more bit defs to wait_until register 20756 20757commit 8a9820a3aa49bc667f90ac291a27e4d7b4ae38b3 20758Author: Alex Deucher <alex@botch2.com> 20759Date: Sun Apr 27 19:02:22 2008 -0400 20760 20761 RADEON: decrease crtc/output verbosity 20762 20763commit c5d62fa0e8f52c3264ff9db3ff10cdf5a806bfc0 20764Author: Owen Taylor <otaylor@huygens.home.fishsoup.net> 20765Date: Thu Apr 17 13:14:53 2008 +0200 20766 20767 Emulate repeats by drawing in tiles 20768 20769 When we can't turn on hardware repeats, because the texture 20770 is non-power-of-two, or has padding at the ends of lines, 20771 try to draw the image in multiple tiles rather than falling 20772 back to software. (We can only do this when there is no 20773 transform.) 20774 20775commit eeb7b74bb6c813b0e3afa4b704f6ffb0d0aab92b 20776Author: Owen Taylor <otaylor@huygens.home.fishsoup.net> 20777Date: Thu Apr 17 13:14:25 2008 +0200 20778 20779 Turn on wrapping when repeating on R100 + R200 20780 20781 Actually enable repeats for R100 and R200. This corresponds 20782 to a R300 change made in the patch in: 20783 http://bugs.freedesktop.org/show_bug.cgi?id=15333 20784 20785commit e511f39dfef503006cf249b9f6934091eaade9b5 20786Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 20787Date: Thu Apr 17 05:04:34 2008 -0400 20788 20789 R300+: move more common code into init3d() 20790 20791 - pre-load r3xx tex instructions 20792 - setup RS instructions in init3d() 20793 20794commit 99435b7c18d931ea620044d0fdb4cc93dfcc6331 20795Author: Owen Taylor <otaylor@redhat.com> 20796Date: Thu Apr 17 02:46:11 2008 -0400 20797 20798 Radeon: Omit mask coordinates 20799 20800 Adapted from Owen's patch on bug 15546 20801 This fixes the slowness with aatext on r300 20802 and may speed up other chips marginally. 20803 20804commit 37614e1db9a595fbe8a21d7a045895e11d272db9 20805Author: Alex Deucher <alex@botch2.com> 20806Date: Tue Apr 15 09:48:16 2008 -0400 20807 20808 fix up some things from the last commit 20809 20810commit 1286fe5ce1c77453d57817b9b26b1bdb32ca7bc8 20811Author: Alex Deucher <alex@botch2.(none)> 20812Date: Mon Apr 14 20:02:14 2008 -0400 20813 20814 R300+: properly setup vap_cntl 20815 20816 this fixes tcl/pvs on RV515 among other things 20817 20818commit f72a4b805db26f10f69330b88459cbeae661189b 20819Author: Alex Deucher <alex@botch2.com> 20820Date: Mon Apr 14 14:10:40 2008 -0400 20821 20822 EXA: Don't wait for 3D idle after each Composite() 20823 20824 wait in CompositeDone() instead 20825 20826commit 4cd4acf1092aeb696b086a382a033aee471d2de9 20827Author: Alex Deucher <alex@botch2.com> 20828Date: Mon Apr 14 11:50:59 2008 -0400 20829 20830 R300: move more common code to init3d() 20831 20832commit 3c523c9a07402e17dff588fad842224c57e98223 20833Author: Alex Deucher <alex@botch2.com> 20834Date: Mon Apr 14 11:21:42 2008 -0400 20835 20836 R3xx+: 3D engine documentation and minor cleanups 20837 20838 - document the R300 exa/textured video code 20839 - minor cleanups of textured video code to clarify meaning 20840 20841commit ce025bbb2496d4de94b8d4ac450c64441b64ee04 20842Author: Alex Deucher <alex@botch2.com> 20843Date: Sat Apr 12 21:22:03 2008 -0400 20844 20845 R300+: consolidate some tcl/non-tcl paths 20846 20847 - Move more code to init3d() 20848 - MMIO textured video seems more reliable now on newer chips 20849 20850commit 11b54a319c7c9dd52e3fb13372697059dafe1cd3 20851Author: Alex Deucher <alex@botch2.com> 20852Date: Sat Apr 12 16:50:22 2008 -0400 20853 20854 R3xx+: fix XAA + textured video on non-TCL path 20855 20856commit dd15a2f5906725116b8cd9954243099055e88e37 20857Author: Alex Deucher <alex@botch2.com> 20858Date: Sat Apr 12 16:49:03 2008 -0400 20859 20860 R3xx+: more fixes to 2D/3D engine init 20861 20862commit f3e68d4b7afd2e23675bf6361c496814c9cb4b94 20863Author: Alex Deucher <alex@botch2.com> 20864Date: Fri Apr 11 10:59:07 2008 -0400 20865 20866 Fix exa glyph corruption on newer chips 20867 20868commit b59686d6427cbf8b35e36b020cbbc6a0c5149b22 20869Author: Alex Deucher <alex@botch2.com> 20870Date: Fri Apr 11 10:15:25 2008 -0400 20871 20872 R300+: pre-load vertex programs in init3D() 20873 20874commit acc5833a35ad6c29a57f659607afb27eebdc2ea5 20875Author: Alex Deucher <alex@botch2.com> 20876Date: Thu Apr 10 17:52:52 2008 -0400 20877 20878 R3xx+: consolidate more tcl code 20879 20880commit 6f8f75bd19ef1919c0291141675be2d0e29b3251 20881Author: Alex Deucher <alex@botch2.com> 20882Date: Thu Apr 10 17:08:50 2008 -0400 20883 20884 R3xx+: consolidate some common 3D code 20885 20886commit 4b9234e1c4f7c7f419cb4245d64f3f9756c98bb6 20887Author: Alex Deucher <alex@botch2.com> 20888Date: Thu Apr 10 16:58:22 2008 -0400 20889 20890 R3xx+: tcl wip 20891 20892commit 865c463e3afb4759758f569132be8bf1386da5cc 20893Author: Alex Deucher <alex@botch2.com> 20894Date: Thu Apr 10 16:51:04 2008 -0400 20895 20896 R300+: textured video tcl cleanup 20897 20898commit 79c8d4ca36a1c3e5fe759d4ccc379c36af8f1676 20899Author: Alex Deucher <alex@botch2.com> 20900Date: Thu Apr 10 16:28:18 2008 -0400 20901 20902 RADEON: cleanup 20903 20904commit c4821a287d29a65f3bcb7d60dc72ec13c0384008 20905Author: Alex Deucher <alex@botch2.com> 20906Date: Thu Apr 10 16:20:17 2008 -0400 20907 20908 Revert "R3xx/R5xx: move more VAP, etc. state setup into common init3d() function" 20909 20910 This reverts commit 305a3310963a5dd07b3495015b06aa8c7c4e6b02. 20911 20912 Conflicts: 20913 20914 src/radeon_commonfuncs.c 20915 src/radeon_exa_render.c 20916 src/radeon_textured_videofuncs.c 20917 20918commit 0032c80bf30bab189204e3e6929e18a19d753138 20919Author: Alex Deucher <alex@botch2.com> 20920Date: Thu Apr 10 14:35:00 2008 -0400 20921 20922 RADEON: store tcl status in driver rec 20923 20924commit 9e2ffe66d106abe34a670d2edc9905ed62c485e8 20925Author: Alex Deucher <alex@botch2.com> 20926Date: Thu Apr 10 14:24:04 2008 -0400 20927 20928 R3xx+: use the right register for engine flush 20929 20930commit e1a9f26c2d2cbca9ad159e723ec95b95be1ef349 20931Author: Alex Deucher <alex@botch2.com> 20932Date: Thu Apr 10 14:12:15 2008 -0400 20933 20934 R3xx+: minor textured video fixes 20935 20936 - set shader output swizzling correctly 20937 - flush the right cache register on r3xx+ 20938 20939commit d79040906cd25bd494feb5901f465bbd050aa923 20940Author: Alex Deucher <alex@botch2.com> 20941Date: Thu Apr 10 13:59:58 2008 -0400 20942 20943 R3xx+: EXA/textured video fixes 20944 20945 - get pipe config based on GB_PIPE_SELECT where applicable 20946 (adapted from a similar patch from Dave) 20947 - only flush the dst cache after submitting vertices, freeing 20948 the cache lines stalls the pipe 20949 - no need to wait for 3D idle after submitting vertices 20950 - fix PURGE_CACHE() and PURGE_ZCACHE() for r3xx+ 20951 - fix depth 16 with EXA composite 20952 20953commit 0a96173cc38e506728d4c3f2dd383ba56e856578 20954Author: Michel Dänzer <michel@tungstengraphics.com> 20955Date: Mon Apr 7 18:15:34 2008 +0200 20956 20957 Increase default CP timeout. 20958 20959 Helps avoid spurious timeouts causing problems, see 20960 http://bugs.freedesktop.org/show_bug.cgi?id=15203 . 20961 20962commit 255fbf465f5e7db2609a5a151bfa810249db52a0 20963Author: Owen W. Taylor <otaylor@fishsoup.net> 20964Date: Thu Apr 3 02:25:41 2008 -0400 20965 20966 Fix rendering of transformed sources for REPEAT_NONE with EXA on >= R300. 20967 20968 Use the border color when possible, otherwise fall back to software. 20969 20970commit bc0407e53237d7968808110bc0243076377acf6e 20971Author: Alex Deucher <alex@cube.(none)> 20972Date: Fri Apr 4 18:40:16 2008 -0400 20973 20974 ATOMBIOS: Add support for DynamicClocks option 20975 20976 This patch adds support for dynamic clock gating and static 20977 power management using the atom command tables. In some cases 20978 the bios may already set this up during post, so YMMV. 20979 20980 I was only able to test on desktop cards, so I haven't tested 20981 to see how much (if any) power this saves or how it affects the 20982 thermal footprint. 20983 20984commit 5f5e21bb50555c56bd371576074c28c929307ff1 20985Author: Alex Deucher <alex@cube.(none)> 20986Date: Fri Apr 4 14:29:45 2008 -0400 20987 20988 RADEON: warning fixes 20989 20990commit c8e9a973aaded24aad567a0e36d0c78a05d6b2fd 20991Author: Alex Deucher <alex@cube.(none)> 20992Date: Fri Apr 4 14:26:19 2008 -0400 20993 20994 RADEON: add some quirks 20995 20996commit 091963a635b79884afe77c026eabb48972fbe175 20997Author: Alex Deucher <alex@botch2.com> 20998Date: Thu Apr 3 22:35:16 2008 -0400 20999 21000 Minor cleanup 21001 21002commit 950e9860643c20acde0eca4e4ff26baacc1f2b69 21003Author: Alex Deucher <alex@botch2.com> 21004Date: Thu Apr 3 22:11:48 2008 -0400 21005 21006 Revert "RADEON: memmap rework 1" 21007 21008 This reverts commit dd8ee1b444f4b973a1e0fadca5f943f2162b5e94. 21009 21010 Conflicts: 21011 21012 src/radeon.h 21013 src/radeon_driver.c 21014 21015 This rework seems to have caused more trouble than it was worth. 21016 21017commit 88a1fe4a94c5d11aff22734b21c89890e4428cd5 21018Author: Alex Deucher <alex@botch2.com> 21019Date: Thu Apr 3 22:04:43 2008 -0400 21020 21021 Revert "RADEON: remove driver rec copies of mc info, use save rec directly" 21022 21023 This reverts commit be0858a84fbdf74c0b844f462933a221d48c707d. 21024 21025 Conflicts: 21026 21027 src/radeon_driver.c 21028 21029commit c40a7aa3989576a8144213e2f31b892d21df8686 21030Author: Owen W. Taylor <otaylor@fishsoup.net> 21031Date: Thu Apr 3 14:43:55 2008 -0400 21032 21033 R3xx/R5xx: Fix pitch and clamp mode for repeating textures 21034 21035 - We can always use TXPITCH on a R300 even when repeating, 21036 (previous check for pitch matching width was also wrong) 21037 - Fix clamp mode for repeating textures to be WRAP 21038 21039commit a8593482c1f2e0f2dbac06c2e5325ba8c83ed9ff 21040Author: Dave Airlie <airlied@redhat.com> 21041Date: Wed Apr 2 09:58:05 2008 +1000 21042 21043 atombios: fix the dual-head hopefully. 21044 21045 tested on r600 with DVI and VGA 21046 21047commit 61d883d116fab3e9b513432d65e705afc5bb39f1 21048Author: Dave Airlie <airlied@redhat.com> 21049Date: Wed Apr 2 09:57:38 2008 +1000 21050 21051 Revert "Revert "atombios: fixup the width/height to use the mode values not the scrn ones"" 21052 21053 This reverts commit fc9af578997b6f22ee8b17e83f37d98689291b0e. 21054 21055 I see your revert and raise you one... 21056 21057commit fc9af578997b6f22ee8b17e83f37d98689291b0e 21058Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 21059Date: Tue Apr 1 09:25:45 2008 -0400 21060 21061 Revert "atombios: fixup the width/height to use the mode values not the scrn ones" 21062 21063 This reverts commit c2b1c8b706a6c7c1fd0af80091958473133d54e7. 21064 21065 These registers hold surface size. Using the mode values 21066 breaks dualhead. 21067 21068commit 959509dd54de053f526b534e379a46934127231f 21069Author: Dave Airlie <airlied@redhat.com> 21070Date: Mon Mar 31 14:29:44 2008 +1000 21071 21072 radeon: use correct DDC interfaces so quirks get applied 21073 21074 Radeon seemed to mess up applying certain quirks, hopefully this will fix it. 21075 21076commit 18f5f1cd2f52afed89fc11ade0920f3dfea87306 21077Author: Dave Airlie <airlied@redhat.com> 21078Date: Mon Mar 31 14:11:49 2008 +1000 21079 21080 radeon: split quirks into separate function and new quirk for IBM RN50 21081 21082 Add a connector table quirk for the IBM RN50. 21083 21084commit c2b1c8b706a6c7c1fd0af80091958473133d54e7 21085Author: Dave Airlie <airlied@redhat.com> 21086Date: Sun Mar 30 11:44:14 2008 +1000 21087 21088 atombios: fixup the width/height to use the mode values not the scrn ones 21089 21090 this fixes it properly, legacy appears to be okay. 21091 21092commit c5edea3d8c9254d3a21e390b8309e39e4c9635db 21093Author: Dave Airlie <airlied@redhat.com> 21094Date: Sun Mar 30 11:11:22 2008 +1000 21095 21096 r500/r600: fix rotation to fill screen 21097 21098 I'm not 100% sure this is the correct fix (maybe we shouldn't be using scrn 21099 virtualX/Y)... this will fix it for now until I get more time. 21100 21101commit 9c62c820ba45ebc14d5f36f5d7885863800b6adb 21102Author: Michel Dänzer <michel@tungstengraphics.com> 21103Date: Fri Mar 28 12:37:29 2008 +0100 21104 21105 Include config.h, so FGL_LINUX can actually be defined when it's tested... 21106 21107commit a00d9260a85b94a522c442aee24bc5ea4dc31c5c 21108Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 21109Date: Thu Mar 27 20:03:13 2008 -0400 21110 21111 RADEON: fix lid issues on AVIVO chips for real this time :) 21112 21113commit f0e89c09074b2c7e641f73692bb39b0bf68eb49c 21114Author: Alex Deucher <alex@botch2.com> 21115Date: Thu Mar 27 19:15:18 2008 -0400 21116 21117 Revert "RADEON: attempt to fix lid issues" 21118 21119 This reverts commit 9b4473c1d830b88866dd22e8174a07195bd6fcf4. 21120 This doesn't help. 21121 21122commit 1442d396b938049b83f009a78ddabe2bf85641b6 21123Author: Dave Airlie <airlied@redhat.com> 21124Date: Thu Mar 27 14:02:51 2008 +1000 21125 21126 radeon: size bios to max of bar vs 64k. 21127 21128 reported by dwmw2: rhbz 438299 21129 21130commit de2f609ff0004ef8b74727bfebc2c74fb91205ea 21131Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 21132Date: Wed Mar 26 18:35:21 2008 -0400 21133 21134 AVIVO: no need to call PreinitXv() on AVIVO chips as they have no overlay 21135 21136commit 75884c257bc2bcfa5b498a77d4c403f09face036 21137Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 21138Date: Wed Mar 26 18:16:47 2008 -0400 21139 21140 XAA: update message about render so as to not confuse users 21141 21142commit 9b4473c1d830b88866dd22e8174a07195bd6fcf4 21143Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 21144Date: Wed Mar 26 18:01:29 2008 -0400 21145 21146 RADEON: attempt to fix lid issues 21147 21148 On some laptops the bios attempts to re-program the chip 21149 when a lid event comes in. This should hopefully prevent 21150 the bios from doing that. 21151 21152commit 8b144830fe9b4a0cee4745023de5e7d387070f60 21153Author: Alex Deucher <alex@samba.(none)> 21154Date: Tue Mar 25 01:15:05 2008 -0400 21155 21156 RV250: disable textured video due to HW bug 21157 21158 The YUV->RGB conversion in the texture engine is broken 21159 on RV250 so the colors come out wrong. 21160 21161commit 1789f11ab91633d3928f8b71988d51ff44bda9d1 21162Author: Alex Deucher <alex@samba.(none)> 21163Date: Mon Mar 24 19:03:30 2008 -0400 21164 21165 R3xx/R5xx: flush PVS state before enabling pvs-bypass 21166 21167commit 305a3310963a5dd07b3495015b06aa8c7c4e6b02 21168Author: Alex Deucher <alex@samba.(none)> 21169Date: Mon Mar 24 14:25:03 2008 -0400 21170 21171 R3xx/R5xx: move more VAP, etc. state setup into common init3d() function 21172 21173 Also some minor code cleanups 21174 21175commit 399b1d405e602c62d6deebea6d7e1f38886cd8e2 21176Author: Alex Deucher <alex@samba.(none)> 21177Date: Mon Mar 24 13:04:57 2008 -0400 21178 21179 R3xx/R5xx: use non VAP/TCP for textured video 21180 21181 Just extra state to emit. 21182 21183commit cd77ec18f32a7b36acb655c927bbfd7044019f97 21184Author: Dave Airlie <airlied@redhat.com> 21185Date: Mon Mar 24 18:42:21 2008 +1000 21186 21187 r300: don't bother with VAP/TCL for render. 21188 21189 We just send more data to the card to process per transaction, without getting 21190 any actual gains, as we already pre-compute the vertices without needing 21191 any clipping or transforms from the card. 21192 21193 Perhaps some stuff could be done on-card, but so far the code is a lot 21194 faster if we avoid sending this extra info. 21195 21196 pre: 150000 glyphs/sec 21197 post: 185000 glyphs/sec 21198 21199commit 301c6739b88676a0c78fc72194e993f894b8dc28 21200Author: Alex Deucher <alex@botch2.com> 21201Date: Sun Mar 23 11:14:02 2008 -0400 21202 21203 RS4xx: Revert back to previous fifo settings for now 21204 21205 Setup of these registers needs more investigation. 21206 21207commit 9bea60b3eb378de5e1d44cc02a2763f4feae7882 21208Author: Alex Deucher <alex@botch2.com> 21209Date: Sat Mar 22 11:46:15 2008 -0400 21210 21211 RS4xx: more work on disp/disp2 fifo setup 21212 21213commit 90f11c3986c28daa7b600b9662da145af325d264 21214Author: Alex Deucher <alex@botch2.com> 21215Date: Sat Mar 22 11:29:51 2008 -0400 21216 21217 RS4xx: missed this on the last commit. 21218 21219commit 6d5066a451017a2683addc9e2496987626795dda 21220Author: Alex Deucher <alex@samba.(none)> 21221Date: Fri Mar 21 16:21:54 2008 -0400 21222 21223 RS4xx: attempt to set up disp/disp2 fifos correctly 21224 21225 If you have an XPRESS chip, please test!!! 21226 21227commit fb1cffac05ae20c8365b25a2042b0ae961880faf 21228Author: Alex Deucher <alex@samba.(none)> 21229Date: Fri Mar 21 15:24:36 2008 -0400 21230 21231 RS4xx: attempt to fix TMDS/DVO support 21232 21233 XPRESS chips added a second set of FP control registers. 21234 I don't have the hw to test however. 21235 21236commit 5e3b21284482df9974c9a58f248f0100def2bb0c 21237Author: Alex Deucher <alex@samba.(none)> 21238Date: Wed Mar 19 19:15:05 2008 -0400 21239 21240 Disable the setting of HARDWARE_CURSOR_BIT_ORDER_MSBFIRST 21241 21242 See bug 11796 21243 21244commit 17cd42ed31814ba329a6a68edd0d75390a7da40e 21245Author: Matt Turner <mattst88@gmail.com> 21246Date: Wed Mar 19 18:17:10 2008 -0400 21247 21248 Enable BSR in Log2 functions 21249 21250 This patch edits RADEONLog2 and ATILog2 to use the x86 BSR instruction instead 21251 of looping through bits. It should provide a somewhat of a speed increase in 21252 this function on x86 and AMD64 architectures. 21253 21254 Note: the BSR instruction was added with the 80386 CPU and is therefore not 21255 compatible with earlier CPUs, though I highly doubt it's even possible to use a 21256 286 in conjunction with a Radeon. 21257 21258 The inline assembly also works with Intel's compiler (icc). 21259 21260commit c83827b4d2b6f03c54429e757a756eb99ff8be6b 21261Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 21262Date: Wed Mar 19 17:58:34 2008 -0400 21263 21264 [PATCH] Compile warning fixes. 21265 21266 Minor changes to avoid declarations mixed with code. 21267 Ansified functions with empty prototype to specify they don't 21268 receive arguments. 21269 Added some prototypes to radeon.h, and major reorder on radeon.h 21270 adding prototypes in alphabetical order and specifying to file that 21271 defines it. 21272 21273commit bed9754ad21d6c0a7f61067b04ba31c430a7cecb 21274Merge: 55e446b5b f71ac0e40 21275Author: Alex Deucher <alex@samba.(none)> 21276Date: Wed Mar 19 16:06:41 2008 -0400 21277 21278 Merge branch 'master' of ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati into r3xx-render 21279 21280commit 55e446b5bc091e6c7b3c2e9ae20b45130555c246 21281Author: Alex Deucher <alex@samba.(none)> 21282Date: Wed Mar 19 13:15:32 2008 -0400 21283 21284 R3xx/R5xx: Make sure to clamp the output of the FS 21285 21286commit b6aa4279cbe68cc8e4523795e9714fb798b62d98 21287Author: Alex Deucher <alex@samba.(none)> 21288Date: Wed Mar 19 12:45:01 2008 -0400 21289 21290 R5xx: bump textured video limits to 4096 21291 21292commit 4a445a3e8c4c5ecd9d4ef8daa26906c3ceaa94a1 21293Author: Alex Deucher <alex@samba.(none)> 21294Date: Wed Mar 19 12:31:51 2008 -0400 21295 21296 RADEON: add new macros to distinguish between R3xx and R5xx 3D 21297 21298commit 85d0c9e8d22ccc72bec87b3fd44da5d7609293e0 21299Author: Alex Deucher <alex@samba.(none)> 21300Date: Wed Mar 19 12:07:33 2008 -0400 21301 21302 RADEON: fixed textured video with XAA and tiling 21303 21304commit f5951db7b3522e0fe6af7f46a170c9c9a60a9bff 21305Author: Alex Deucher <alex@samba.(none)> 21306Date: Wed Mar 19 12:01:50 2008 -0400 21307 21308 RV515: fix textured video and EXA Composite 21309 21310 There seems to be an issue with the PVS setup on RV515, but 21311 bypassing it seems to work fine. 21312 21313commit 13573879fe56368ad06234712b677c23fabc56c6 21314Author: Dave Airlie <airlied@redhat.com> 21315Date: Wed Mar 19 15:06:47 2008 +1000 21316 21317 r500: make it work from startup. 21318 21319 I'm not sure why this worked or what is going wrong here, really the 21320 VAP internal architecture escapes me :) 21321 21322commit d331dd64d644a18ec99a2136cd0943b5edca1f03 21323Author: Alex Deucher <alex@samba.(none)> 21324Date: Tue Mar 18 19:44:26 2008 -0400 21325 21326 R3xx/R5xx: remove extra return after last commit 21327 21328commit bc34df7a9c35cdd38c49d5c22471f3f487a33d6e 21329Author: Alex Deucher <alex@samba.(none)> 21330Date: Tue Mar 18 19:39:47 2008 -0400 21331 21332 R3xx/R5xx: switch an ErrorF() to RADEONFALLBACK() 21333 21334commit 6f03f8fe0ecf4181dcf125049cf63bece0451fb2 21335Author: Alex Deucher <alex@samba.(none)> 21336Date: Tue Mar 18 19:36:05 2008 -0400 21337 21338 R3xx: we only use 2 temps, not 3 21339 21340commit 8bb71ab4a3eb4fb6ef7f709e87c8df387cb70ee3 21341Author: Tilman Sauerbeck <tilman@code-monkey.de> 21342Date: Tue Mar 18 14:36:08 2008 -0400 21343 21344 R3xx/R5xx: fix up a8-src-something_with_colors 21345 21346commit c362591d9b496df30668543158e4de44de742dc3 21347Author: Alex Deucher <alex@samba.(none)> 21348Date: Tue Mar 18 11:15:17 2008 -0400 21349 21350 R3xx/R5xx: remove some cruft 21351 21352commit 89fe6d2c7d7471e6088558130f6e49f46c31dd47 21353Author: Dave Airlie <airlied@linux.ie> 21354Date: Tue Mar 18 09:43:43 2008 -0400 21355 21356 R5xx: fix typ in r5xx render accel 21357 21358 This gets render working on r5xx 21359 21360commit 79b40ebcd8dedfc83e484c1024beeeaccc6124f3 21361Author: Alex Deucher <alex@samba.(none)> 21362Date: Tue Mar 18 02:46:49 2008 -0400 21363 21364 R5xx: first pass at render support (untested) 21365 21366commit 71292c8f193230255d1d980c2e996bb01d04fab6 21367Author: Alex Deucher <alex@samba.(none)> 21368Date: Tue Mar 18 00:45:37 2008 -0400 21369 21370 R5xx: bump tex/dst limits to 4096 21371 21372commit 30b52f8aa6a471455284f59b5b27252743892b13 21373Author: Alex Deucher <alex@samba.(none)> 21374Date: Mon Mar 17 23:20:10 2008 -0400 21375 21376 R3xx/R5xx: whitespace cleanup and cruft removal 21377 21378commit 9c9f1b538ed710c3066775fba0a8e936b63087b1 21379Author: Alex Deucher <alex@samba.(none)> 21380Date: Mon Mar 17 23:01:37 2008 -0400 21381 21382 R3xx: get masks working and cleanup 21383 21384 RS offset was wrong for mask texture 21385 21386commit ef94febd74f8ee63081b61e42f093a5a2b8fbf1e 21387Author: Alex Deucher <alex@samba.(none)> 21388Date: Mon Mar 17 22:27:19 2008 -0400 21389 21390 R3xx: minor adjustments 21391 21392commit f71ac0e40b9d950bcb3bba42a75d41f45b6ed1bf 21393Author: Alban Browaeys <prahal@yahoo.com> 21394Date: Mon Mar 17 20:48:48 2008 -0400 21395 21396 RADEON: Revert to old behavior when resetting the memmap on VT switch 21397 21398 Not sure why this needs to be done twice. Should fix bug 14980 21399 Probably needs more investigation. 21400 21401commit bedbbf196dc97ee5142e7dfae16fb6f317fca5a7 21402Author: Alex Deucher <alex@samba.(none)> 21403Date: Mon Mar 17 20:16:25 2008 -0400 21404 21405 R3xx: some progress 21406 21407commit af0e626c132de2dd9958fec657fcc85d4c0fe5e1 21408Author: Alex Deucher <alex@samba.(none)> 21409Date: Mon Mar 17 18:07:12 2008 -0400 21410 21411 R3xx: fix errant w 21412 21413commit 29ea5bfc0eb3194e2454fc3ee863df54f0300880 21414Author: Alex Deucher <alex@cube.(none)> 21415Date: Mon Mar 17 16:41:57 2008 -0400 21416 21417 RADEON: fix typo in RADEONAdjustMemMapRegisters() 21418 21419commit ab317e85c5ab1a249a510c34aeb3a908be1a66fc 21420Author: Alex Deucher <alex@cube.(none)> 21421Date: Mon Mar 17 15:28:09 2008 -0400 21422 21423 RADEON: make sure var is initialized properly in RADEONAdjustMemMapRegisters() 21424 21425commit 208d307227e15f37a6af5194398ed23266ff743a 21426Author: Dave Airlie <airlied@linux.ie> 21427Date: Sun Mar 16 19:39:23 2008 +1000 21428 21429 radeon: the 0x5974 appears to be a mobility chip... 21430 21431 After debugging with partymola on #radeon, adding this allowed his 21432 Dell Vostro 1000 to work properly 21433 21434commit 9bc7c2ec4048e1677547c1d60c51ccb954f7589a 21435Author: Alex Deucher <alex@samba.(none)> 21436Date: Fri Mar 14 20:12:22 2008 -0400 21437 21438 R3xx: odds and ends... 21439 21440 still not working. 21441 - swizzle US output for BGR formats 21442 - no need to write to temps in ALU ops, 21443 write to output only 21444 - flush the PVS before updating 21445 21446commit 96bea7906c4706fcd57a9cd8f1ce3feab6ac676d 21447Author: Alex Deucher <alex@samba.(none)> 21448Date: Fri Mar 14 15:59:36 2008 -0400 21449 21450 R3xx: theoretical support for component alpha 21451 21452 masks are still broken so... 21453 21454commit cffe3dcc8991cd7c457a9c1a9f41055aa9ea3436 21455Author: Alex Deucher <alex@samba.(none)> 21456Date: Fri Mar 14 14:37:43 2008 -0400 21457 21458 R3xx: VS WIP 21459 21460commit b73f52a50dfd6ff8d92f04d6b510c39582c6ac67 21461Author: Alex Deucher <alex@samba.(none)> 21462Date: Fri Mar 14 14:20:49 2008 -0400 21463 21464 R3xx/R5xx: enable VS for mask texture 21465 21466commit 569a14ca9be1e18fe9921edc816ac3dc32d6cca7 21467Author: Alex Deucher <alex@samba.(none)> 21468Date: Fri Mar 14 13:32:12 2008 -0400 21469 21470 R3xx/R5xx: Fix magic numbers in vertex shaders 21471 21472commit 4878997529601d62e257aa1c9112bd460561de73 21473Author: Alex Deucher <alex@samba.(none)> 21474Date: Thu Mar 13 21:23:40 2008 -0400 21475 21476 R3xx: make sure to set the FS code size correctly 21477 21478commit 22f46b88ef05afb6a6b6d70007ac4980a446430e 21479Author: Alex Deucher <alex@samba.(none)> 21480Date: Thu Mar 13 20:25:33 2008 -0400 21481 21482 R3xx: attempt to setup the rasterizer properly for mask texture 21483 21484 Not working yet 21485 21486commit 081fc9e892fa3d2e07b7db65b2e2719646255463 21487Author: Alex Deucher <alex@samba.(none)> 21488Date: Thu Mar 13 18:38:26 2008 -0400 21489 21490 R3xx: more mask work 21491 21492commit 2bf0236c03538ace3ce6d0e68f0829fc47d1385b 21493Author: Alex Deucher <alex@samba.(none)> 21494Date: Thu Mar 13 18:32:00 2008 -0400 21495 21496 R3xx: enable composite for non-mask cases 21497 21498commit 74286ba41302107d2fc626fee2181f7c4bc18164 21499Author: Alex Deucher <alex@samba.(none)> 21500Date: Thu Mar 13 18:25:32 2008 -0400 21501 21502 R3xx: add basic mask support 21503 21504commit a2bbe10d866567911b68f222b4758624bfe9bf84 21505Author: Alex Deucher <alex@samba.(none)> 21506Date: Thu Mar 13 18:16:53 2008 -0400 21507 21508 R300: setup source selects and output swizzling 21509 21510commit b9974ecce7d1932595226004858b08a7a6b188dc 21511Author: Alex Deucher <alex@samba.(none)> 21512Date: Thu Mar 13 17:35:38 2008 -0400 21513 21514 R3xx: set the texture id and add some register info 21515 21516commit 0ef700b7da5e554a0d0d166f3fde85ff45c9eb1f 21517Author: Alex Deucher <alex@samba.(none)> 21518Date: Thu Mar 13 17:02:25 2008 -0400 21519 21520 R3xx/R5xx: enable blending 21521 21522commit b35c09a597c93a1d9f06ef0091c96822b0653f98 21523Author: Dave Airlie <airlied@redhat.com> 21524Date: Thu Mar 13 18:42:29 2008 +1000 21525 21526 xv: fixup XAA on r500 textured video 21527 21528 the XAA area should never end up tiled. This may break with nooffscreen pixmaps 21529 21530commit d4446461c3630caff166456c351ace34f57cc119 21531Author: Matt Turner <mattst88@gmail.com> 21532Date: Tue Mar 11 21:20:53 2008 -0400 21533 21534 Properly fix uninitialized variables warnings 21535 21536 According to commit 9fd13e6773371c82b9799a5bda7c96ffa5cafe8c to 21537 xf86-video-intel by Kristian Høgsberg, there is a better way to fix the 21538 possibly initialized variables warnings. This patch will use Kristian's fix. 21539 21540commit 20adfd7390d9b1f100e0c4a14f175377b8335c82 21541Author: Alex Deucher <alex@cube.(none)> 21542Date: Tue Mar 11 20:09:35 2008 -0400 21543 21544 RADEON: enable output attributes that require a modeset immediately 21545 21546 This should fix bug 14915 21547 21548commit 53ba7f5771b0b53fb0d3bc29d64bdd3813756d10 21549Author: Alex Deucher <alex@cube.(none)> 21550Date: Tue Mar 11 19:12:40 2008 -0400 21551 21552 RADEON: fix vblank interrupts after VT switch or suspend/resume 21553 21554commit e946c097f0438afbea6f3dd37ee39d67d415708c 21555Author: Matt Turner <mattst88@gmail.com> 21556Date: Tue Mar 11 19:07:58 2008 -0400 21557 21558 [PATCH] Fix a few warnings 21559 21560commit 8e160508520c0a24ca90aad182f296461ca0d9b6 21561Author: Alex Deucher <alex@cube.(none)> 21562Date: Tue Mar 11 18:11:13 2008 -0400 21563 21564 DCE3: add support for PCIEPHY (untested) 21565 21566commit fbded88a2925f9f049936dad0736721e7b84a6ee 21567Author: Alex Deucher <alex@cube.(none)> 21568Date: Tue Mar 11 14:10:31 2008 -0400 21569 21570 ATOM: remove some cruft 21571 21572commit 3263f6e4a410281d620c288a92bb4521f7b6fc06 21573Author: Alex Deucher <alex@cube.(none)> 21574Date: Tue Mar 11 14:05:48 2008 -0400 21575 21576 DCE3: enable DPMS on DIG ports 21577 21578commit eb90e235b58c94f3d4d75394725ab2fe246a42ff 21579Author: Alex Deucher <alex@cube.(none)> 21580Date: Tue Mar 11 13:53:54 2008 -0400 21581 21582 DCE3: adjust PLL for DCE3 chips 21583 21584 this fixes stability issues on digital outputs and certain modes. 21585 21586commit 552615ccc5360baafb8bb41698c1ca27816fd4b2 21587Author: Alex Deucher <alex@cube.(none)> 21588Date: Tue Mar 11 13:38:29 2008 -0400 21589 21590 ATOMBIOS: enable load detection by default on both DACs 21591 21592 Load detection is reliable with atom, so enable it by default 21593 on both DACA and DACB, rather than just DACA. 21594 21595commit 78b10487cf222c96f8944ba25e2ea970506b3535 21596Author: Alex Deucher <alex@cube.(none)> 21597Date: Tue Mar 11 13:16:00 2008 -0400 21598 21599 DCE3: add output attribute to enable/disable coherent mode 21600 21601 Enabled by default. The TMDS transmitter can be programmed 21602 slightly differently depending on the chips in the panel. If you 21603 have problems with tmds on a particular panel, try disabling it. 21604 21605commit d20be31c46fbec623af4c3628a7c603ceacf500f 21606Author: Alex Deucher <alex@botch2.(none)> 21607Date: Mon Mar 10 21:05:43 2008 -0400 21608 21609 RV550: MC setup is like RV515 not RV530 21610 21611commit 38606b08b68842fbcc81c233009c1117269f3be9 21612Author: Matthieu Herrb <matthieu@bluenote.herrb.net> 21613Date: Sat Mar 8 23:22:59 2008 +0100 21614 21615 Makefile.am: nuke RCS Id 21616 21617commit 9d710ee1a44cf2f3a948fbdbe17ef09521cbe744 21618Author: Alex Deucher <alex@cube.(none)> 21619Date: Fri Mar 7 15:09:14 2008 -0500 21620 21621 AVIVO: clean up some unused variables 21622 21623commit c28c30c9f3d7bfebfd56a5c982c96f0090982054 21624Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 21625Date: Fri Mar 7 14:10:49 2008 -0500 21626 21627 RADEON: Fix crash in last commit 21628 21629commit c3a3635865e380c784a226c8ead069d4716d6b75 21630Author: Dave Airlie <airlied@redhat.com> 21631Date: Thu Mar 6 20:17:45 2008 -0500 21632 21633 RADEON: fix tiling/interlaced interaction with randr 1.2 21634 21635commit df1b94dc4eb1f35b636dbf2ec0ab1c2da9937c0d 21636Author: Alex Deucher <alex@botch2.(none)> 21637Date: Thu Mar 6 19:22:08 2008 -0500 21638 21639 DCE3: Ignore outputs with DIN connectors for now 21640 21641commit cb2dc19387c7b6494c47c76d683cf38a48700768 21642Author: Alex Deucher <alex@cube.(none)> 21643Date: Thu Mar 6 18:33:12 2008 -0500 21644 21645 AVIVO: fix typo from a previous commit 21646 21647 Leave tv dpms hook disabled or you may get bad interactions 21648 with the shared DAC 21649 21650commit 77355de48057e5e7e0d5b3f3cf5a7a92220a53b1 21651Author: Alex Deucher <alex@cube.(none)> 21652Date: Thu Mar 6 17:46:00 2008 -0500 21653 21654 AVIVO: don't add outputs for invalid connectors 21655 21656commit 600dbe080997a01ceaf6be86723189d518bc1281 21657Merge: 594743a99 5b7875d0c 21658Author: Alex Deucher <alex@cube.(none)> 21659Date: Thu Mar 6 17:31:37 2008 -0500 21660 21661 Merge branch 'master' of ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati 21662 21663commit 594743a99811a8b0f391412892414fcd158eeb56 21664Author: Alex Deucher <alex@cube.(none)> 21665Date: Thu Mar 6 17:30:21 2008 -0500 21666 21667 AVIVO: fix up memsize detection for IGP chips 21668 21669commit 5b7875d0cbfbdbcd1515c4e942d30de298b49dff 21670Author: Doug Chapman <doug.chapman@hp.com> 21671Date: Thu Mar 6 14:31:06 2008 -0500 21672 21673 Bug #14826: Fix a bogus check around xf86SetOperatingState. 21674 21675commit 651fe23f4c650ed91843dec48db24e18e8b91219 21676Merge: 3de2dc88c 766f464df 21677Author: Adam Jackson <ajax@redhat.com> 21678Date: Thu Mar 6 14:28:27 2008 -0500 21679 21680 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/driver/xf86-video-ati 21681 21682commit 41171c25cd235bafad26bcbabced16ead4b8c54b 21683Author: Alex Deucher <alex@cube.(none)> 21684Date: Thu Mar 6 14:05:18 2008 -0500 21685 21686 DCE3.0: add support for crtc memreq table 21687 21688commit 766f464dfdfccadef23e4232f2bce5db22195513 21689Author: Alex Deucher <alex@cube.(none)> 21690Date: Thu Mar 6 13:35:43 2008 -0500 21691 21692 RADEON: take 2 on proper pragma pack support for bsds 21693 21694 See bug 14594. Based on suggestion by Henry Zhao 21695 21696commit a842ce9ca6494e724a7828cead9b61c9ef02b6aa 21697Author: Alex Deucher <alex@cube.(none)> 21698Date: Thu Mar 6 12:32:18 2008 -0500 21699 21700 DCE3.0: Minor fixups 21701 21702commit 8a1ba374033591c725a78923aa30829e4de2a5ae 21703Author: Alex Deucher <alex@cube.(none)> 21704Date: Thu Mar 6 09:53:51 2008 -0500 21705 21706 RADEON: option to override TVDAC adj values from bios with driver defaults 21707 21708 If you have a washed out image on the tv dac, try this option. 21709 Option "DefaultTVDACAdj" "TRUE" 21710 21711commit 0ed48f8f651a28e189f9fee8c6b593da0178d21c 21712Author: Alex Deucher <alex@cube.(none)> 21713Date: Wed Mar 5 18:41:01 2008 -0500 21714 21715 AVIVO: Initial support for DCE 3.0 using atombios 21716 21717 DACs are working well, DIG support (DVI, HDMI, LVDS, etc.) 21718 still has some issues. 21719 21720commit 2901e99f1942842856cd39c1dcc8b22f3cf7d9e3 21721Author: Alex Deucher <alex@botch2.(none)> 21722Date: Wed Mar 5 10:40:06 2008 -0500 21723 21724 RADEON: fix fetching of dac2 adj values from newer bios tables 21725 21726commit 74eb981287d76836327830bd51272f605a07e0cc 21727Author: Alex Deucher <alex@botch2.(none)> 21728Date: Mon Mar 3 12:02:44 2008 -0500 21729 21730 ATOMBIOS: fix atombios parser support on *bsd 21731 21732 bsd requires a different pragma pack than Linux. 21733 See bug 14594. 21734 21735commit f7769ea86e265f347eb58c517ccb5ef8b35eec27 21736Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 21737Date: Sun Mar 2 14:49:21 2008 -0500 21738 21739 [PATCH] Ensure symbols used by other modules are visible. 21740 21741 The xf86-video-ati drivers are one of the cases where LoaderSymbol is 21742 widely used in some obscure ways. This patch fixes the problem, and 21743 allows compiling with -fvisibility=hidden. 21744 21745commit a4398ac3ad77216f2c8aa628425bef5f2912a0a9 21746Author: Alex Deucher <alex@cube.(none)> 21747Date: Sat Mar 1 18:52:26 2008 -0500 21748 21749 RS6xx: change isIGP checks to CHIP_FAMILY_RS690 21750 21751 these paths are only relevant on RS6xx chips 21752 21753commit 67d4d04836c05293b844bc505f303cfb04c0f8a4 21754Author: Alex Deucher <alex@cube.(none)> 21755Date: Sat Mar 1 18:33:18 2008 -0500 21756 21757 RADEON: use xf86SetDesiredModes() in screeninit and enterVT 21758 21759 this should restore the proper output state on VT switches 21760 21761commit be0858a84fbdf74c0b844f462933a221d48c707d 21762Author: Maciej Cencora <m.cencora@gmail.com> 21763Date: Sat Mar 1 18:11:58 2008 -0500 21764 21765 RADEON: remove driver rec copies of mc info, use save rec directly 21766 21767 info->mc_* were used and the immediately copied into info->ModeReg 21768 ones. Just use the ModeReg copies directly. 21769 21770commit dd8ee1b444f4b973a1e0fadca5f943f2162b5e94 21771Author: Alex Deucher <alex@cube.(none)> 21772Date: Sat Mar 1 16:23:51 2008 -0500 21773 21774 RADEON: memmap rework 1 21775 21776 Don't restore memmap regs on every mode switch. 21777 Just do memmap save/restore/setup on server start and VT switch. 21778 21779commit 1f6a23000001f3d1c21b5c04f94714a8caa7aa8b 21780Author: Alex Deucher <alex@cube.(none)> 21781Date: Sat Mar 1 15:53:42 2008 -0500 21782 21783 RADEON: only restore legacy dac regs on legacy radeons 21784 21785commit dee6cef8e62d0651c00319e03eea92940fd24aa4 21786Author: Alex Deucher <alex@botch2.(none)> 21787Date: Sat Mar 1 14:39:32 2008 -0500 21788 21789 RS4xx: enable exa render accel and textured video 21790 21791 RS6xx paths seem to work fine on RS4xx 21792 21793commit 129f737efe4e8d1a368e7db4b063bdcd9339cb09 21794Author: Alex Deucher <alex@cube.(none)> 21795Date: Sat Mar 1 14:32:30 2008 -0500 21796 21797 AVIVO: save/restore regs by block 21798 21799 Save/Restore the entire block for each output. 21800 This should fix VT switch problems. 21801 21802commit b069aadaa63a95d7a71b5cfbab83577b49501094 21803Author: Alex Deucher <alex@botch2.(none)> 21804Date: Fri Feb 29 22:36:02 2008 -0500 21805 21806 AVIVO: LVDS panels need dithering enabled 21807 21808 Fixes bug 14760 21809 21810commit fe87bdee815372b4b4d7d4c705e34681625b90f2 21811Author: Alex Deucher <alex@botch2.(none)> 21812Date: Fri Feb 29 13:10:13 2008 -0500 21813 21814 AVIVO: disable pageflipping on avivo chips until we have proper drm support 21815 21816commit fb3678c7f511d539a51cd090cb8b5041d7d2ba26 21817Author: Alex Deucher <alex@botch2.(none)> 21818Date: Fri Feb 29 13:01:21 2008 -0500 21819 21820 R5xx: fix register count when sending fragment program for textured video 21821 21822commit a66d37d1a896ec934989592457c2beff8e6f1639 21823Author: Alex Deucher <alex@botch2.(none)> 21824Date: Fri Feb 29 04:07:05 2008 -0500 21825 21826 fix off-by-one in last commit 21827 21828commit e56062960be0c8d3947861dd5e0691fce6516b99 21829Author: Alex Deucher <alex@cube.(none)> 21830Date: Thu Feb 28 19:16:39 2008 -0500 21831 21832 AVIVO: save/restore scaler regs 21833 21834commit ae1c39a9b3e666404d0931679c9078c2e125a8bc 21835Author: Alex Deucher <alex@cube.(none)> 21836Date: Thu Feb 28 18:53:55 2008 -0500 21837 21838 RS6xx: rework output parsing 21839 21840 Turns out it's not as complex as I originially thought. 21841 IGP chips just have non-standard GPIO entires for DDC. 21842 21843commit d8d6c9fe4ae7e1ab67dd041a251e901d97c29ed6 21844Author: Alex Deucher <alex@botch2.(none)> 21845Date: Thu Feb 28 17:01:14 2008 -0500 21846 21847 RS6xx: fix typos in previous commit 21848 21849 Noted by Maciej Cencora on IRC 21850 21851commit 46547ae8bdbc5c10f1fd028b95ec4c5c31a5b318 21852Author: Alex Deucher <alex@cube.(none)> 21853Date: Thu Feb 28 14:29:30 2008 -0500 21854 21855 AVIVO: disable dithering on DFPs 21856 21857 This should fix the color banding some people have noticed. 21858 Also save/restore DDIA regs on RS6xx 21859 21860commit 72a53d6f20ac29b3baddb7d8af04f19b76d2e04f 21861Author: Michel Dänzer <michel@tungstengraphics.com> 21862Date: Thu Feb 28 17:38:04 2008 +0100 21863 21864 Handle EXA coordinate limits more cleverly. 21865 21866 Generally set the 2D engine limits, and only enforce the 3D engine limits in the 21867 CheckComposite hook. This should still prevent useless migration of pixmaps the 21868 3D engine can't handle but allows for basic acceleration of bigger ones. 21869 21870 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=14708 . 21871 21872commit 5249f450a2487475a95531603cc8668db2c21c33 21873Author: Michel Dänzer <michel@tungstengraphics.com> 21874Date: Thu Feb 28 12:23:58 2008 +0100 21875 21876 Fix 16 bit packed YUV XVideo playback on big endian systems with DRI disabled. 21877 21878 http://bugs.freedesktop.org/show_bug.cgi?id=14668 21879 21880commit e40d75fd8b2aece9dae8076fac822a4a83025fb2 21881Author: Alex Deucher <alex@samba.(none)> 21882Date: Wed Feb 27 22:53:10 2008 -0500 21883 21884 R500: fragment program clean up and magic number conversion 21885 21886commit 140dadba36b2191f0e18e41dd987785abd5f55d2 21887Author: Alex Deucher <alex@samba.(none)> 21888Date: Wed Feb 27 22:21:12 2008 -0500 21889 21890 R300: fix up magic numbers in fragment program 21891 21892commit e521476bb5e2dfabc93747e43eb911a8a101357e 21893Author: Alex Deucher <alex@botch2.(none)> 21894Date: Wed Feb 27 21:26:55 2008 -0500 21895 21896 R300/R400: bump up the clip limits for textured video 21897 21898 This allows up to 2560x2560 (hw limit) 21899 21900commit 10db46f11d7e1c055c9ad6034c65ad163dad17dc 21901Author: Alex Deucher <alex@samba.(none)> 21902Date: Wed Feb 27 15:28:50 2008 -0500 21903 21904 AVIVO: make sure we select the right LUT for each crtc 21905 21906commit ea944f38dcfd871b27345698afea1cb986ecb049 21907Author: Alex Deucher <alex@samba.(none)> 21908Date: Wed Feb 27 14:37:52 2008 -0500 21909 21910 R300+: update RADEONCP_REFRESH() to reflect new location of scissor regs 21911 21912commit b865faf95666e2172c3eec143f77fe9c524e4983 21913Author: Alex Deucher <alex@samba.(none)> 21914Date: Wed Feb 27 14:05:44 2008 -0500 21915 21916 R100/R200: move r100/r200 specific 3D setup into appropriate blocks 21917 21918 R3xx+ doesn't have these regs. 21919 21920commit ee485ee0fc550414cd39cdb8fd2769fe0bc996a8 21921Author: George Sapountzis <gsap7@yahoo.gr> 21922Date: Wed Feb 27 18:41:31 2008 +0200 21923 21924 radeon: pass distcheck 21925 21926commit ff110c781299fb729e93572ba3d630ca59e8127d 21927Author: George Sapountzis <gsap7@yahoo.gr> 21928Date: Wed Feb 27 18:41:17 2008 +0200 21929 21930 add warning for r128 21931 21932commit b4a344f36332a12a51c9e896d93e9cb0672d31ef 21933Author: George Sapountzis <gsap7@yahoo.gr> 21934Date: Wed Feb 27 18:40:50 2008 +0200 21935 21936 add warning for mach64 21937 21938commit 81ce4097c12ab64a52d0992f981f588f5a297b7b 21939Author: George Sapountzis <gsap7@yahoo.gr> 21940Date: Wed Feb 27 18:40:23 2008 +0200 21941 21942 drop r128 21943 21944commit 873ff2033f1d0b84c91c48e829aa945f41b490b7 21945Author: George Sapountzis <gsap7@yahoo.gr> 21946Date: Wed Feb 27 18:40:09 2008 +0200 21947 21948 drop r128 (build system) 21949 21950commit 2b8b2676e7b2fd2952f5307a934ef0a11006f838 21951Author: George Sapountzis <gsap7@yahoo.gr> 21952Date: Wed Feb 27 18:39:49 2008 +0200 21953 21954 drop mach64 21955 21956commit 05bbbec11a04e7f6321b92c9dbec4ad4c0d130ff 21957Author: George Sapountzis <gsap7@yahoo.gr> 21958Date: Wed Feb 27 18:39:32 2008 +0200 21959 21960 drop mach64 (build system) 21961 21962commit 68888189cf8d460ef6f8f2f1431a6ffe9fcd8134 21963Author: Alex Deucher <alex@samba.(none)> 21964Date: Wed Feb 27 03:09:28 2008 -0500 21965 21966 RADEON: fix textured video uploads 21967 21968commit 7cfad216a3f1486be194c972b9b6bcf070358586 21969Author: Alex Deucher <alex@samba.(none)> 21970Date: Wed Feb 27 02:13:10 2008 -0500 21971 21972 RADEON: fix clipping for textured video 21973 21974commit a2dca1d68d751def34ef3c6f836574173737bf76 21975Author: Alex Deucher <alex@samba.(none)> 21976Date: Tue Feb 26 16:38:29 2008 -0500 21977 21978 RADEON: Convert textured video to use pipelined uploads 21979 21980commit b4fa1ce9d2da04d94521a82d3c2e95f0fe985ccc 21981Author: Alex Deucher <alex@samba.(none)> 21982Date: Tue Feb 26 14:45:25 2008 -0500 21983 21984 R300/R500: fix up clipping 21985 21986 Based on patch from Peter Zubaj. 21987 21988commit 44e527a117ab0a363135ff066c7f7e0c12e3dc89 21989Author: Alex Deucher <alex@samba.(none)> 21990Date: Tue Feb 26 14:33:24 2008 -0500 21991 21992 R300: fix cordinate clamping in render code 21993 21994 Based on Peter's fix for textured video 21995 21996commit 00ec17ad53d7ad43f19c9b723794ac1b8ef86826 21997Author: Peter Zubaj <pzubaj at marticonet.sk> 21998Date: Tue Feb 26 14:26:14 2008 -0500 21999 22000 R300: Fix clamping and pitch for textured video 22001 22002commit 4207793f9b6c891cb17ba715223d2002668885e3 22003Author: Alex Deucher <alex@samba.(none)> 22004Date: Tue Feb 26 13:16:22 2008 -0500 22005 22006 R300/R500: set the number graphics pipes properly 22007 22008 This should fix the checkerboard issues on r300/r350 cards. 22009 22010commit 823d8bf49e4c0a8c839354179942c5a1944c178d 22011Author: Alex Deucher <alex@samba.(none)> 22012Date: Tue Feb 26 12:25:14 2008 -0500 22013 22014 R300/R500: clean up magic numbers in textured video 22015 22016commit 8ea75b268f11794f4a7e7bac52cb256490ed3fd2 22017Author: Dave Airlie <airlied@linux.ie> 22018Date: Tue Feb 26 16:29:19 2008 +1000 22019 22020 regs: fix spelling properly 22021 22022commit f2816064a6c2c4c35ccba74b9aa80547e25c012e 22023Author: Dave Airlie <airlied@linux.ie> 22024Date: Tue Feb 26 16:28:24 2008 +1000 22025 22026 regs: fix spelling mistake 22027 22028 pointed out by plaes on irc 22029 22030commit 9d2ca30b90607085578dde1f314db663bd5f82ec 22031Author: Alex Deucher <alex@samba.(none)> 22032Date: Mon Feb 25 17:34:00 2008 -0500 22033 22034 R300/R500: clean up magic numbers in render code 22035 22036commit 153ad6fcf704cbf9f811d9986cd4baf04e82c9d2 22037Author: Dave Airlie <airlied@linux.ie> 22038Date: Mon Feb 25 07:10:48 2008 +1000 22039 22040 fixup check for EXA composite pointed out by Alan Swanson 22041 22042commit 85e470e64f629de72e361c77770e2e29998d1bf4 22043Merge: 27ddb39b1 1b84c76f2 22044Author: Alex Deucher <alex@samba.(none)> 22045Date: Sun Feb 24 05:37:22 2008 -0500 22046 22047 Merge master and fix conflicts 22048 22049 Merge branch 'master' of ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati 22050 22051 Conflicts: 22052 22053 src/radeon_commonfuncs.c 22054 22055commit 27ddb39b12a0b54e099fd5274c4c91f08e2d2822 22056Author: Alex Deucher <alex@samba.(none)> 22057Date: Sun Feb 24 05:30:11 2008 -0500 22058 22059 R300: clean up magic numbers in RADEONInit3DEngine 22060 22061commit d4c20f33ad6a1f88615cd7e09ad3638896873f9e 22062Author: Alex Deucher <alex@samba.(none)> 22063Date: Sun Feb 24 04:46:10 2008 -0500 22064 22065 R300: replace magic numbers in cache flush 22066 22067commit e52f1c8d2647b81d891ec0728dd582941a76c83f 22068Author: Alex Deucher <alex@samba.(none)> 22069Date: Sun Feb 24 04:43:18 2008 -0500 22070 22071 R300: fill in some more 3D bitfields 22072 22073commit 1b84c76f27c8d24cb42beae26abf000721901c1c 22074Author: Dave Airlie <airlied@redhat.com> 22075Date: Sun Feb 24 19:20:36 2008 +1100 22076 22077 rs690: initial textured video support 22078 22079commit 68158124366db883a901e960fe5f05f8df5efa42 22080Author: Dave Airlie <airlied@redhat.com> 22081Date: Tue Feb 19 19:51:18 2008 +1100 22082 22083 rs690: initial rs690 rotate 22084 22085 suffers same problem as r500 with clipping 22086 22087commit e614bb6965588bf09dcb87f5e08e67120ec9847f 22088Author: Dave Airlie <airlied@redhat.com> 22089Date: Sun Feb 24 01:46:05 2008 -0500 22090 22091 r500: convert fragprog to use register values 22092 22093commit 6ce9ee47c75620b2e5d211c5d59d17271a6a7b19 22094Author: Dave Airlie <airlied@redhat.com> 22095Date: Sat Feb 23 22:49:34 2008 -0500 22096 22097 r500: add textured video Xv adapter support 22098 22099commit 9aaf8b33b22b6ba112869558ae54e021b9487ad2 22100Author: Dave Airlie <airlied@redhat.com> 22101Date: Sat Feb 23 22:16:25 2008 -0500 22102 22103 r500: initial rotate support - not fully working yet. 22104 22105 Just an example of how to setup and run the r500 3D engine for rotation. 22106 this rotates for me but I get some strange clipping on the bottom of my screen 22107 22108commit 05dc3e4fc19d056ce99a7b110665adab2ca1ea21 22109Author: Adam Jackson <ajax@redhat.com> 22110Date: Sat Feb 23 20:29:51 2008 -0500 22111 22112 Clarify R500 US setup. 22113 22114commit 080606ad528972623e1ed1124d8efe7705a73446 22115Author: Adam Jackson <ajax@redhat.com> 22116Date: Sat Feb 23 20:21:17 2008 -0500 22117 22118 Add the R500 US index/data pair. 22119 22120commit edfb3b6bbf0ee17ace8e6ba704a6f54e249fec63 22121Author: Alex Deucher <alex@samba.(none)> 22122Date: Sat Feb 23 19:59:33 2008 -0500 22123 22124 RADEON: no textured video yet on XPRESS chips 22125 22126 Still need to sort out the VAP and PVS stuff 22127 22128commit 4146bfe5d00e40a86d17826fac50d04b2469621d 22129Author: Alex Deucher <alex@samba.(none)> 22130Date: Sat Feb 23 19:21:52 2008 -0500 22131 22132 R500: fix typo in new r5xx regs 22133 22134commit d9be9f34b0d3313e7b22b2a8bb0a8924ad3116bf 22135Author: Alex Deucher <alex@samba.(none)> 22136Date: Sat Feb 23 19:06:30 2008 -0500 22137 22138 RADEON: add textured video support for r1xx-r4xx radeons 22139 22140 Based on the kdrive ati video code by Eric Anholt. 22141 R3xx/R4xx still have some clipping issues in certain situations 22142 22143commit 9dc4acad79196e9d5d94dd710773bfa83456d47f 22144Author: Alex Deucher <alex@cube.(none)> 22145Date: Sat Feb 23 18:29:00 2008 -0500 22146 22147 RS6xx: gpio entry for DDIA varies depending on the number of DFP ports 22148 22149commit ed0a93edf28155308e7ab9d8705581bb38455ea0 22150Author: Adam Jackson <ajax@redhat.com> 22151Date: Sat Feb 23 15:02:17 2008 -0500 22152 22153 Fix R500_US_CONFIG. 22154 22155commit bc2bd6f841b51aeed3b6b4a47dbe758c200bc5a6 22156Author: Adam Jackson <ajax@redhat.com> 22157Date: Sat Feb 23 14:34:18 2008 -0500 22158 22159 Add R500 unified shader register block. 22160 22161commit a7b5c3bb74fc4de5e38a75ac31656445ce823464 22162Author: Alex Deucher <alex@cube.(none)> 22163Date: Fri Feb 22 19:35:11 2008 -0500 22164 22165 RS6xx: fix DDC on DDIA output (usually HDMI port) 22166 22167commit 3327a681e21101cc6f6e162f4e29f9937b69ccc3 22168Author: Alex Deucher <alex@cube.(none)> 22169Date: Fri Feb 22 17:05:56 2008 -0500 22170 22171 ATOM: properly set up DDIA output on RS6xx boards 22172 22173commit 1d0e9ab8b9451101b1b91943546f6c5833c21b3f 22174Author: Michel Dänzer <michel@tungstengraphics.com> 22175Date: Wed Feb 20 10:21:49 2008 +0100 22176 22177 radeon: Fix typo flagged by gcc -Wall. 22178 22179commit b5bd442b60dbc72fe4c1e928ab864aeb0fd7a3cb 22180Author: Alex Deucher <alex@botch2.(none)> 22181Date: Tue Feb 19 20:47:40 2008 -0500 22182 22183 R100: fix render accel for transforms 22184 22185 Not sure why we had a separate broken path for r100 vertex 22186 submission. 22187 22188commit a0a73208a21546ac120fb9a463261836c9ea7b55 22189Author: Alex Deucher <alex@botch2.(none)> 22190Date: Tue Feb 19 20:11:19 2008 -0500 22191 22192 RADEON: restore clock gating and CP clock errata on VT switch 22193 22194 This may help people with hangs on resume 22195 22196commit b77e2aff7453a9f370beba37ca3c25b92b3f97ff 22197Author: Alex Deucher <alex@botch2.(none)> 22198Date: Tue Feb 19 19:55:41 2008 -0500 22199 22200 RADEON: fix DDC types 5 and 6 22201 22202commit af82172a82f2bdf96e571def659a1c70f92dfdbf 22203Author: Alex Deucher <alex@botch2.(none)> 22204Date: Tue Feb 19 19:39:35 2008 -0500 22205 22206 RADEON: update man page with supported chips 22207 22208commit 1302d1b019c8091986aaf9c86b25c1f36a037788 22209Author: Alex Deucher <alex@botch2.(none)> 22210Date: Tue Feb 19 00:34:31 2008 -0500 22211 22212 Bump for 6.8 release 22213 22214commit 830d29e8b280c11e798fb5bf5278bf24ecbac7bf 22215Author: Alex Deucher <alex@cube.(none)> 22216Date: Mon Feb 18 21:17:12 2008 -0500 22217 22218 ATOM: fix RS6xx connector table parsing and fix some warnings 22219 22220commit 03aa4cc6d6e8c715a1c1d677cc1845223505b358 22221Author: Roland Scheidegger <sroland@tungstengraphics.com> 22222Date: Mon Feb 18 20:19:58 2008 -0500 22223 22224 bring back to life planar-to-packed conversion for rs4xx 22225 22226 Here's a patch to bring back the code for converting planar yuv to packed yuv, 22227 if a RS400 family chip is used (though I've no idea if they really all fail 22228 with planar yuv). 22229 22230 fixes bug 12744 22231 22232commit 66b0b3f43bb714bc007169d5ba6dabe34cfc008b 22233Author: George Sapountzis <gsap7@yahoo.gr> 22234Date: Mon Feb 18 10:33:59 2008 +0200 22235 22236 r128/radeon: probably correct fix for non pci access code 22237 22238commit d055b9e800ae50d08cca9db75fc666ce1da9ab52 22239Author: Dave Airlie <airlied@linux.ie> 22240Date: Sat Feb 16 09:13:43 2008 +1000 22241 22242 r128/radeon: hopeful fix for non pci access code 22243 22244commit c773bc6a314327da29e21e4ebac6fa7f3e98a0a7 22245Author: Dave Airlie <airlied@linux.ie> 22246Date: Sat Feb 16 09:05:07 2008 +1000 22247 22248 r128/radeon: fix build without pciaccess 22249 22250commit 690a52da5248f47a587a878d05fce9784957970b 22251Author: Dave Airlie <airlied@linux.ie> 22252Date: Sat Feb 16 08:33:36 2008 +1000 22253 22254 mach64: fix non pciaccess build 22255 22256commit fc85188fd95bf78b7f965cdde3e22b644c74ff81 22257Author: George Sapountzis <gsap7@yahoo.gr> 22258Date: Fri Feb 15 18:36:10 2008 +0200 22259 22260 ati: convert to pci probe 22261 22262 add pciids for each subdriver, make no use of the match_data functionality. 22263 22264 thanks to Alex Deucher for reviewing and testing. 22265 22266commit 665bd7e2f61cac3e029bbad5024034e5136deec1 22267Author: George Sapountzis <gsap7@yahoo.gr> 22268Date: Fri Feb 15 18:34:56 2008 +0200 22269 22270 ati wrapper: translate Device lines 22271 22272commit a596c1618f72179a45289a50a1f9e89462ce9667 22273Author: George Sapountzis <gsap7@yahoo.gr> 22274Date: Fri Feb 15 18:34:31 2008 +0200 22275 22276 r128: do not compile in PciChipsets twice 22277 22278commit 99cd8ff9a7e15fc2b4e55f8bc020f584173a8c2d 22279Author: George Sapountzis <gsap7@yahoo.gr> 22280Date: Fri Feb 15 18:33:51 2008 +0200 22281 22282 ati: drop duplicate pci-id defines 22283 22284commit 146b01b51069dc227d0b125babb3f6957c9b9de2 22285Author: George Sapountzis <gsap7@yahoo.gr> 22286Date: Fri Feb 15 18:33:29 2008 +0200 22287 22288 mach64: clean probe a little 22289 22290 do not report I/O ports now, they are reported later. 22291 22292commit f47d461331a032f9bdcf6f63336e848778cec6cc 22293Author: George Sapountzis <gsap7@yahoo.gr> 22294Date: Fri Feb 15 18:32:56 2008 +0200 22295 22296 mach64: minor cosmetic, I2C type 22297 22298commit 2c66f2e812195167df9ca113044d46deece776ac 22299Author: Dave Airlie <airlied@clockmaker.usersys.redhat.com> 22300Date: Fri Feb 15 10:59:30 2008 +1000 22301 22302 make distcheck pass 22303 22304commit ed87f367ddab7366f84917b43b31d76df4ce1110 22305Author: Alex Deucher <alex@botch2.(none)> 22306Date: Wed Feb 13 12:53:46 2008 -0500 22307 22308 RADEON: disable LVDS if panel size is 0x0 22309 22310 if we can't get the panel size from the bios or edid 22311 or a user specified option, assume it's not connected. 22312 22313commit 422d7f441fdbb318d22d44db661ac9bd61387bd2 22314Author: Alex Deucher <alex@cube.(none)> 22315Date: Tue Feb 12 15:35:46 2008 -0500 22316 22317 R6xx: when both crtcs are in use make sure they are both enabled 22318 22319 sometimes setting a mode on one crtc can cuase a blank screen on the other. 22320 make sure they are both enabled if they should be. 22321 22322commit 860f5af75274cb236f536e1da09da6bd9a579530 22323Author: Alex Deucher <alex@cube.(none)> 22324Date: Tue Feb 12 14:46:49 2008 -0500 22325 22326 R6xx: fix up use of bios scratch regs to reflect the new offsets 22327 22328commit 32f2119b43a0faf6069d8cc0816f0d9f7914c07f 22329Author: Alex Deucher <alex@cube.(none)> 22330Date: Tue Feb 12 14:20:02 2008 -0500 22331 22332 R6xx: bios scratch regs moved. 22333 22334commit 8d64be6ebd7f50d4bcb587afeee8252c1367dc77 22335Author: Alex Deucher <alex@samba.(none)> 22336Date: Tue Feb 12 12:53:09 2008 -0500 22337 22338 RADEON: make sure we always set up the FP crtc shadow registers for crtc0 22339 22340 The behavior changed when I added rmx center mode support. In cases where 22341 crtc0 drives a DAC this can lead to a blank screen. 22342 22343commit 810d192ee046077a894e0fb5f2dfd6a7c0130766 22344Author: Alex Deucher <alex@cube.(none)> 22345Date: Mon Feb 11 19:23:01 2008 -0500 22346 22347 R6xx: add missing objects 22348 22349commit 85043439426e534e561259ce98bebdd8508b36a9 22350Author: Alex Deucher <alex@botch2.(none)> 22351Date: Mon Feb 11 16:36:58 2008 -0500 22352 22353 R6xx: make sure we set up the HDP base properly 22354 22355commit 9ab5d2ec7c583c74f364d7cfbb54bcd2cd8ae2f5 22356Author: Alex Deucher <alex@botch2.(none)> 22357Date: Mon Feb 11 15:26:51 2008 -0500 22358 22359 RADEON: always restore crtc1 before crtc0 when using both crtcs 22360 22361 In some rare cases restoring crtc0 first will result in a blank screen 22362 on crtc1. If you are having issues with a blank screen on crtc1 22363 that used to work on 6.6.3 or before, this should help. 22364 22365commit e33edca75bd9df0aa19a33e74c38a6d02610befd 22366Author: Alex Deucher <alex@botch2.(none)> 22367Date: Mon Feb 11 00:33:12 2008 -0500 22368 22369 RADEON: Implement proper (hopefully) BIOS/DRIVER interaction for ATOM bios 22370 22371 Tested on atom-based Desktop cards. It'd nice to get some testing 22372 on atom-based laptops. 22373 22374commit 6524e33435a786f7de0064cdd1b04c1120d21593 22375Author: Alex Deucher <alex@botch2.(none)> 22376Date: Sun Feb 10 18:52:52 2008 -0500 22377 22378 RADEON: Implement proper (hopefully) BIOS/DRIVER interaction for COM bios 22379 22380 Tested on my M10-based laptop. 22381 22382commit 8606c1bd175893c77e8c758246a1aed166c8be51 22383Author: George Sapountzis <gsap7@yahoo.gr> 22384Date: Fri Feb 8 19:07:03 2008 +0200 22385 22386 mach64: factor out BIOS panel info parsing 22387 22388commit 9f33218c80f5a6d6d9464aa3db8ae25a4759f351 22389Author: George Sapountzis <gsap7@yahoo.gr> 22390Date: Fri Feb 8 19:06:41 2008 +0200 22391 22392 mach64: minor cosmetic, LCD panel id 22393 22394commit 8cd5a465a03834b4b2f3a19f8d36fc9d2e18c6d4 22395Author: George Sapountzis <gsap7@yahoo.gr> 22396Date: Fri Feb 8 19:06:27 2008 +0200 22397 22398 mach64: minor cosmetic, DAC type 22399 22400commit 7f4db96123fdcba9874226352802d702c2a498bd 22401Author: Dave Airlie <airlied@redhat.com> 22402Date: Fri Feb 8 10:47:10 2008 +1000 22403 22404 add rv670 pciids 22405 22406commit e8899b9978291c62a65f468c92f340f65ad5479d 22407Author: Alex Deucher <alex@botch2.(none)> 22408Date: Thu Feb 7 19:27:38 2008 -0500 22409 22410 R6xx: fix ddc after my i2c rework 22411 22412 Seems r6xx does something different for its i2c table, 22413 revert to the old behavior for now. 22414 22415commit 435de6c4e46ff2bebd4cee58888a66b936cd3fdf 22416Author: Alex Deucher <alex@samba.(none)> 22417Date: Thu Feb 7 19:14:13 2008 -0500 22418 22419 RADEON: sync up with latest upstream versions 22420 22421 atombios.h 22422 ObjectID.h 22423 22424commit 692789a293970f70b88ccb6adcf0676d8b495ae2 22425Author: George Sapountzis <gsap7@yahoo.gr> 22426Date: Thu Feb 7 18:03:37 2008 +0200 22427 22428 mach64: factor out BIOS multimedia parsing 22429 22430commit 933328ffd6d1d872a18d3de8624c4df845a64588 22431Author: George Sapountzis <gsap7@yahoo.gr> 22432Date: Thu Feb 7 18:03:03 2008 +0200 22433 22434 mach64: complement hint for sparc 22435 22436commit 956c8c81f3ff434930a0cb17b027b2f8e4eeabb2 22437Author: George Sapountzis <gsap7@yahoo.gr> 22438Date: Thu Feb 7 18:02:31 2008 +0200 22439 22440 mach64: consolidate refclk #2 22441 22442commit dce4cc26a8e2bf53805ec63763243f3ff6b4a6d3 22443Author: George Sapountzis <gsap7@yahoo.gr> 22444Date: Thu Feb 7 18:02:17 2008 +0200 22445 22446 mach64: consolidate refclk #1 22447 22448commit f7ed807f0d82a7446ebc4acdd4e94df44a675f19 22449Author: George Sapountzis <gsap7@yahoo.gr> 22450Date: Thu Feb 7 18:01:59 2008 +0200 22451 22452 mach64: cosmetic 22453 22454commit cda1cd198f33c26ef1b51532a2126468369743b8 22455Author: George Sapountzis <gsap7@yahoo.gr> 22456Date: Thu Feb 7 18:01:33 2008 +0200 22457 22458 mach64: factor out BIOS clock parsing 22459 22460commit 73ff279469be9c7cbf9f533b85fcb553694ff413 22461Author: George Sapountzis <gsap7@yahoo.gr> 22462Date: Thu Feb 7 18:00:55 2008 +0200 22463 22464 mach64: BIOSBase is no longer used 22465 22466commit caea326cc6f1932bb299f451be013651a5749ea7 22467Author: Dave Airlie <airlied@linux.ie> 22468Date: Wed Feb 6 06:36:13 2008 +1000 22469 22470 r300: move fragprog setup code to prepare composite for now 22471 22472commit 470cd6a401c6a3e8fea981a8fe97c28be3cfb81d 22473Author: Dave Airlie <airlied@linux.ie> 22474Date: Wed Feb 6 06:04:13 2008 +1000 22475 22476 r300: remove r300 specific vertex emission 22477 22478 Set the vertex and fragment engine to expect the mask coords. 22479 22480commit f65e8dfac23adfa199026765fe3a1ea08cf4da67 22481Author: Alex Deucher <alex@cube.(none)> 22482Date: Sun Feb 3 00:09:59 2008 -0500 22483 22484 RADEON: rework i2c handling 22485 22486 Split out clk, data, and lock regs and masks. some cards use different 22487 regs and masks for each. For cards with ATOMBIOS, use the i2c bios 22488 table to grab the i2c data. 22489 22490commit a38a903debc0a50dbc73f59dc2741bbea76d2bd9 22491Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 22492Date: Fri Feb 1 02:59:07 2008 -0500 22493 22494 RADEON: don't restore LVDS_PLL_CNTL for now 22495 22496 seems to cause problems with resume for some users. 22497 this needs further investigation. 22498 see bug 12596 22499 22500commit 5d7bea2b62c86868baf1960665a40249a43bafc5 22501Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 22502Date: Fri Feb 1 02:53:22 2008 -0500 22503 22504 RADEON: remove redundant RADEONDisableDisplays() 22505 22506 use RADEONBlank() instead 22507 22508commit 73b437ce232c94c0067a0d2f70538b6e1e8c07a7 22509Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 22510Date: Fri Feb 1 02:47:06 2008 -0500 22511 22512 RADEON: remove unused "aspect" scaler option 22513 22514commit f1fb9e4daa29bc379f653f847254db1496b625fd 22515Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 22516Date: Fri Feb 1 02:39:18 2008 -0500 22517 22518 RADEON: Implement "center" mode for RMX on legacy radeons 22519 22520commit bcd590103e04bfdb4f12413beacebf344f07e88e 22521Author: Alex Deucher <alex@botch2.(none)> 22522Date: Tue Jan 29 12:12:54 2008 -0500 22523 22524 RADEON: update man page to reflect AGP 1x default revert 22525 22526commit 09348a83d06ba9d3129499c4daedd44a68771530 22527Author: Alex Deucher <alex@botch2.(none)> 22528Date: Tue Jan 29 12:09:24 2008 -0500 22529 22530 Seems the default is more reliable... we can't win. 22531 22532 Revert "radeon: Default to 1x again with non-v3 AGP cards." 22533 22534 This reverts commit b653e5a628bfa4dfb168e96f93f41eb910f409fb. 22535 22536commit 0c26806245381b925b73ff9f3b747e3dcf0ebd6f 22537Author: Alex Deucher <alex@botch2.(none)> 22538Date: Tue Jan 29 10:26:48 2008 -0500 22539 22540 RADEON: Add new RV380 pci id 22541 22542 bug 14289 22543 22544commit ce77ed78a877023da72dbe51609aef9a07e250b1 22545Author: Maciej Cencora <m.cencora@gmail.com> 22546Date: Mon Jan 28 19:02:56 2008 -0500 22547 22548 RS690: Implement MC idle check 22549 22550commit b7de4ff52cfbdd85ee65000613632e21b92af24e 22551Author: George Sapountzis <gsap7@yahoo.gr> 22552Date: Sat Jan 26 19:28:05 2008 +0200 22553 22554 mach64: add hint for sparc and minor cosmetic. 22555 22556commit 09d713aa0ed6367b4457420b3c2832fe8eca9b00 22557Author: Alex Deucher <alex@botch2.(none)> 22558Date: Wed Jan 23 18:05:41 2008 -0500 22559 22560 Revert "RADEON: adjust PAL mode hstart" 22561 22562 This reverts commit 719a9a376e34d99032af75e3f7b002670ccb816b. 22563 22564 This breaks TV out on some cards. 22565 22566commit 719a9a376e34d99032af75e3f7b002670ccb816b 22567Author: Alex Deucher <alex@botch2.(none)> 22568Date: Sun Jan 20 18:40:53 2008 -0500 22569 22570 RADEON: adjust PAL mode hstart 22571 22572 Most people seem to get a more aligned picture with this setting 22573 22574commit b2db8657fb888cff6d64c6dcb182caac389776ce 22575Author: Alex Deucher <alex@botch2.(none)> 22576Date: Sun Jan 20 18:33:22 2008 -0500 22577 22578 RADEON: re-work i2c for DDC 22579 22580 Unify the radeon/avivo paths and grab the data/clk masks from bios 22581 if available 22582 22583commit a43003b24022a833e604f41b2873c0350b34181c 22584Author: Alex Deucher <alex@botch2.(none)> 22585Date: Sat Jan 19 18:49:53 2008 -0500 22586 22587 RADEON: get dac2 adj values from the bios tables 22588 22589commit d4596c52ac9994be26e9ec2d7d57b3892c34abdb 22590Author: Alex Deucher <alex@botch2.(none)> 22591Date: Sat Jan 19 17:17:26 2008 -0500 22592 22593 RADEON: grab pll_in_min/pll_in_max from bios tables if available 22594 22595commit 9a0947c812d0d38d1bca6a91140ac210831a6cb4 22596Author: Alex Deucher <alex@botch2.(none)> 22597Date: Sat Jan 19 13:57:50 2008 -0500 22598 22599 ATOM: Use LVDS edid from bios if available 22600 22601commit 7238258c12def8ef273e5362f716d165f720c5a5 22602Author: Kusanagi Kouichi <slash@ma.neweb.ne.jp> 22603Date: Sat Jan 19 15:04:21 2008 +0100 22604 22605 radeon: Partial fix for XVideo RGB image distortions. 22606 22607commit 32be3cf9d6c34e60ff8c3d6cfe9f73f1869c50e4 22608Author: Brice Goglin <Brice.Goglin@ens-lyon.org> 22609Date: Fri Jan 18 14:42:14 2008 -0500 22610 22611 RADEON: print the name of the output when printing the EDID 22612 22613commit 4ba9430ee97dbce8f77db8de6ce9b753a75e453d 22614Author: George Sapountzis <gsap7@yahoo.gr> 22615Date: Fri Jan 18 18:18:18 2008 +0200 22616 22617 mach64: workaround for corruption at upper-left 22618 22619 commit possible workaround, it's reported multiple times ... 22620 22621commit 12c00111b68c9cf4872a424258c6f8b7247aac47 22622Author: George Sapountzis <gsap7@yahoo.gr> 22623Date: Fri Jan 18 18:17:40 2008 +0200 22624 22625 use stand-alone drivers for each chip family. 22626 22627 Do not load the ati wrapper when the user specifies the sub-driver name in 22628 the Driver line of xorg.conf. Also, for -configure cause the wrapper to fail 22629 and let each sub-driver speak for themselves. 22630 22631commit 24c7d134cd450f9e2cca85e4a2fc3253d250be04 22632Author: Alex Deucher <alex@botch2.(none)> 22633Date: Fri Jan 18 09:50:38 2008 -0500 22634 22635 RADEON: make sure EXA Composite is actually disabled on XPRESS chips. 22636 22637commit eaf425436008092abe81208321a2b3b6698a5d79 22638Author: Dave Airlie <airlied@redhat.com> 22639Date: Fri Jan 18 20:11:57 2008 +1000 22640 22641 fixup register 6594 save/restore 22642 22643commit 3de2dc88cf26ff5932f11cecdf975777b8aa2a4a 22644Author: Adam Jackson <ajax@redhat.com> 22645Date: Wed Jan 16 14:55:05 2008 -0500 22646 22647 Bump CRTC size limits on AVIVO chips so 30" displays work without tweaking. 22648 22649 Note that the CRTC size limits we're using right now are _not_ the 22650 hardware limits, they're just heuristics until we can resize the front 22651 buffer properly. 22652 22653commit 2a89a31481d71a56a9930073cf99d3ae7b4290e1 22654Author: Alex Deucher <alex@samba.(none)> 22655Date: Thu Jan 17 15:08:17 2008 -0500 22656 22657 RADEON: use radeon_output->Flags for tracking RMX rather than mode->Flags 22658 22659commit 495e3119250ffb48489debbaabe560d23753cc43 22660Author: Alex Deucher <alex@samba.(none)> 22661Date: Thu Jan 17 14:56:19 2008 -0500 22662 22663 AVIVO: Add support for RMX 22664 22665 Both centered and expansion modes are supported. Select 22666 using output attributes. 22667 22668commit 6bd510a211f25d52e74791e4a429cd2218ced541 22669Author: Alex Deucher <alex@samba.(none)> 22670Date: Wed Jan 16 18:09:49 2008 -0500 22671 22672 RADEON: add a message about render accel on newer cards 22673 22674commit 85bf3439fe2579aec48f5cd8d65a9d51b1ae8535 22675Author: Alex Deucher <alex@samba.(none)> 22676Date: Wed Jan 16 17:52:06 2008 -0500 22677 22678 R300: only init3d on r3xx 22679 22680commit dbb2ca471dfbff245b30c5055871dee0dc0e3d15 22681Author: Alex Deucher <alex@samba.(none)> 22682Date: Wed Jan 16 17:10:02 2008 -0500 22683 22684 R300: only enable render accel on non-IGP r3xx/r4xx chips for now 22685 22686commit 3c72b100bcfacee600644669b586e86cfd32754e 22687Author: Alex Deucher <alex@samba.(none)> 22688Date: Wed Jan 16 16:55:42 2008 -0500 22689 22690 R300: First pass at render accel 22691 22692 This first pass is pretty limited. All it currently supports 22693 is transforms for rotation. No blending yet. 22694 22695 Based on inital implementation from Wolke Liu with 22696 additional lock-up fixes by Dave Airlie. 22697 22698commit 2ba3562d2af911fdd90881049599e239d27260bc 22699Author: George Sapountzis <gsap7@yahoo.gr> 22700Date: Sat Jan 12 17:11:59 2008 +0200 22701 22702 ati wrapper: xf86PciInfo.h is enough 22703 22704commit c2caeb11a97dad5379d70881c5c0fd834a8c3d54 22705Author: George Sapountzis <gsap7@yahoo.gr> 22706Date: Sat Jan 12 16:18:34 2008 +0200 22707 22708 ati wrapper: add DriverRec's and use them 22709 22710commit 19e1b180fec6f83a474e125465bc60111c0f43e0 22711Author: George Sapountzis <gsap7@yahoo.gr> 22712Date: Sat Jan 12 16:18:07 2008 +0200 22713 22714 mach64: load for both "ati" and "mach64" as driver names 22715 22716 similar to r128/radeon 22717 22718commit 92f54400d5450b29b3a76d5ecc927cf0d73e156e 22719Author: George Sapountzis <gsap7@yahoo.gr> 22720Date: Sat Jan 12 16:17:47 2008 +0200 22721 22722 mach64: add version (need not match with ati) 22723 22724 similar to r128/radeon 22725 22726commit 311ec7b6c54a50a4b8a5a445f7283da2b0b2e0f5 22727Author: George Sapountzis <gsap7@yahoo.gr> 22728Date: Sat Jan 12 16:17:21 2008 +0200 22729 22730 atimisc: rename to mach64 22731 22732commit 5244e235262290eab8a3546f449295c12ed8f6c7 22733Author: Alex Deucher <alex@samba.(none)> 22734Date: Mon Jan 14 16:11:09 2008 -0500 22735 22736 ATOMBIOS: disable the scaler by default on avivo cards 22737 22738 Fixes bug 14001 22739 22740commit 729da30c80d6545b75c8faea44754634f477cc09 22741Merge: 000741e25 625a885a9 22742Author: Alex Deucher <alex@samba.(none)> 22743Date: Mon Jan 14 10:05:01 2008 -0500 22744 22745 Merge branch 'master' of ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati 22746 22747commit 625a885a964488da2a35065eb597a17ee57b22a9 22748Author: George Sapountzis <gsap7@yahoo.gr> 22749Date: Wed Jan 9 17:39:56 2008 +0200 22750 22751 ati wrapper: use pci ids 22752 22753commit 48865200ca9f1daf87e52620b9f8175f88dd886f 22754Author: George Sapountzis <gsap7@yahoo.gr> 22755Date: Wed Jan 9 19:01:04 2008 +0200 22756 22757 radeon: remove stray include 22758 22759commit a5a6b873353611cb7e46e5e375f039957c7051a7 22760Author: George Sapountzis <gsap7@yahoo.gr> 22761Date: Wed Jan 9 13:03:32 2008 +0200 22762 22763 radeon: remove stray _X_EXPORT 22764 22765commit 80c9974b6cdc0147d777df6990b3f3aacd87969d 22766Author: George Sapountzis <gsap7@yahoo.gr> 22767Date: Tue Jan 8 18:15:47 2008 +0200 22768 22769 r128: pci-rework conversion 22770 22771 compile-tested only 22772 22773commit cd4b39403d74f314e1c2cfa4cf0e4fe542891dc3 22774Author: George Sapountzis <gsap7@yahoo.gr> 22775Date: Sat Mar 10 23:12:15 2007 +0200 22776 22777 Drop symbol lists from r128, radeon, theatre. 22778 22779 compile-tested only 22780 22781commit 63b4b8213fabc5a57f897b60b6eaa9f78d86e6ff 22782Author: George Sapountzis <gsap7@yahoo.gr> 22783Date: Wed Mar 21 20:24:22 2007 +0200 22784 22785 [mach64] Drop symbol lists. 22786 22787 egrep LOADER\|SymLists\|Symbols src/*.[hc] 22788 22789 not needed after server commit bca9364f3f4a2376edbcf57a34f704ce28be21ba, i.e. 22790 xserver 1.2 22791 22792 compile-tested only 22793 22794commit 000741e250e54122b0adc91694eb4bfa320a70fb 22795Author: Alex Deucher <alex@samba.(none)> 22796Date: Thu Jan 10 14:49:48 2008 -0500 22797 22798 RADEON: clean up output handling 22799 22800commit 10e7636c02478b8ffe183bb0c46229ca0d6584e1 22801Author: Kristian Høgsberg <krh@bitplanet.net> 22802Date: Wed Jan 9 12:47:39 2008 -0500 22803 22804 RADEON: fix crtc routing on r4xx cards when using atom to init DVO chip 22805 22806commit 3af671f5963810dbfd63abc9889b1d46b68f404c 22807Author: Alex Deucher <alex@botch2.(none)> 22808Date: Wed Jan 9 11:30:25 2008 -0500 22809 22810 RADEON: restore FP2 regs before external encoders 22811 22812 This may fix krh's dvi problem 22813 22814commit 2a54c6bb09ade2ec8f998dfc1624017029d47fa3 22815Author: Alex Deucher <alex@botch2.(none)> 22816Date: Tue Jan 8 18:43:54 2008 -0500 22817 22818 RADEON: Make default output actually work... 22819 22820commit fa3e2055225c27e25465fc46786da1b7574fd3cc 22821Author: Alex Deucher <alex@botch2.(none)> 22822Date: Mon Jan 7 01:13:09 2008 -0500 22823 22824 RADEON: add default outputs if no connected devices 22825 22826 If no connected devices found at server startup, default 22827 to something so the server comes up. LVDS on mobility chips, 22828 DAC or TMDS on others. 22829 22830commit d972cc9237eb90b49b11f8d2bdc5b61f628911dc 22831Author: Alex Deucher <alex@botch2.(none)> 22832Date: Sat Jan 5 17:19:06 2008 -0500 22833 22834 RADEON: Fix TVStandard option 22835 22836commit 45656b9d5a426053da2a094de39c2690c0c6f799 22837Author: Alex Deucher <alex@botch2.(none)> 22838Date: Sat Jan 5 12:00:55 2008 -0500 22839 22840 R128: Like powerpc, don't use VGA by default on sparc 22841 22842commit b8e8db4675d07e45782de0d7c67ee0fd85eaedb3 22843Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 22844Date: Fri Jan 4 20:16:19 2008 -0500 22845 22846 RADEON: fix tvdac load detection at server start up 22847 22848commit 3ba7f393d0669df36848715799de8affc10a5534 22849Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 22850Date: Fri Jan 4 20:00:48 2008 -0500 22851 22852 RADEON: further fixup for pScrn->pScreen issue 22853 22854 the previous fix seems to cause the driver to hang on 22855 some cards. 22856 22857commit a0de9c0844f9e066e0f02e8cd8045bdd278e6494 22858Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 22859Date: Fri Jan 4 19:48:30 2008 -0500 22860 22861 RADEON: improve ntsc image centering 22862 22863commit 03d2f25801c8a8ec15030f06008df112d07c1a2d 22864Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 22865Date: Fri Jan 4 18:19:54 2008 -0500 22866 22867 RADEON: fix compile 22868 22869commit 308848783ed9ae27aed7c7de6ee813d375ef495e 22870Author: Dave Airlie <airlied@linux.ie> 22871Date: Sat Jan 5 09:11:55 2008 +1000 22872 22873 i2c: a line mux of 0 is valid 22874 22875commit 94d5a432f72801f821d1c4ce952baba17552659d 22876Author: Dave Airlie <airlied@linux.ie> 22877Date: Fri Jan 4 17:16:54 2008 +1000 22878 22879 atombios: i2c lines are all done with sw so we can accept any of them. 22880 22881 hch reported this working on his laptop with 0x7e60 as EDID for LVDS 22882 22883commit 2e4473b63d65801ae8ac5a8391de232b2201d958 22884Author: Alex Deucher <alex@samba.(none)> 22885Date: Thu Jan 3 15:47:50 2008 -0500 22886 22887 RADEON: fix crash when setting rotation in the config file 22888 22889 xf86CrtcRotate() accesses pScrn->pScreen which is not set 22890 during ScreenInit(). This should also be fixed in the server. 22891 See bug 12129 22892 22893commit c652208861bffca94f06b7f67688ce220e050bfb 22894Author: Michel Dänzer <michel@tungstengraphics.com> 22895Date: Thu Jan 3 17:54:58 2008 +0100 22896 22897 radeon: Adapt manpage to reality wrt AGP options. 22898 22899commit ab451e4b7a5423d61b57cf0646599267d8504af4 22900Author: Michel Dänzer <michel@tungstengraphics.com> 22901Date: Thu Jan 3 17:52:39 2008 +0100 22902 22903 radeon: Miscellaneous warning fixes. 22904 22905commit 394c52273328e90518568b694ee79dc1a8dab651 22906Author: Dave Airlie <airlied@linux.ie> 22907Date: Thu Jan 3 18:56:16 2008 +1000 22908 22909 r500: tvout avoid doing dpms here it makes my tv mode go all crappy 22910 22911 need to investigate further 22912 22913commit 1c647279f021d01e110980727b7c7dd7efae1642 22914Author: Dave Airlie <airlied@linux.ie> 22915Date: Thu Jan 3 11:55:28 2008 +1000 22916 22917 r600: change devices list depending on connector 22918 22919commit f36db6e10d32a68b32d20ae4ad02cfc0bfd1c9c3 22920Author: Dave Airlie <airlied@linux.ie> 22921Date: Thu Jan 3 11:27:47 2008 +1000 22922 22923 r600: fixup crash on unknown output type 22924 22925 not sure this is 100% the correct approach 22926 22927commit 1accfdd590828e95e0d68a576c8ee05a06a86e43 22928Author: Alex Deucher <alex@samba.(none)> 22929Date: Wed Jan 2 19:48:28 2008 -0500 22930 22931 RADEON: various avivo tv-out related clean-ups 22932 22933commit ce34090c758ac91171cb6adb9b8a36e4adbf99cf 22934Merge: 2180f04b6 30cab1dbe 22935Author: Alex Deucher <alex@samba.(none)> 22936Date: Wed Jan 2 16:41:36 2008 -0500 22937 22938 Merge branch 'master' of git+ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati 22939 22940commit 2180f04b6fb94a638f6274cb4455d5688b324dbc 22941Author: Alex Deucher <alex@samba.(none)> 22942Date: Wed Jan 2 16:41:06 2008 -0500 22943 22944 RADEON: preliminary support for ATOM RMX 22945 22946 not functional yet. 22947 22948commit dab4dc285154d40303aadaa849b85f8e251e578e 22949Author: Alex Deucher <alex@samba.(none)> 22950Date: Wed Jan 2 16:27:19 2008 -0500 22951 22952 RADEON: add support for ATOM component video table 22953 22954 Component output is still not working. 22955 22956commit 30cab1dbebb7bdb925f2fe9f91b84183312bbbfd 22957Author: Alex Deucher <alex@botch2.(none)> 22958Date: Wed Jan 2 15:41:01 2008 -0500 22959 22960 RADEON: Make sure all old IGP chips have HasSingleDac set 22961 22962 fix the csv file and re-gen the headers. 22963 22964commit e8e585651215b011e3ad07c59d0eab9107ccd8c6 22965Author: Dave Airlie <airlied@redhat.com> 22966Date: Wed Jan 2 10:14:46 2008 +1000 22967 22968 PLL/r600: tweak pll to pick first one found instead of keeping going 22969 22970commit 14aa4060ad27ecb3d40b2b17ee4cf7cc55a121dd 22971Author: Dave Airlie <airlied@redhat.com> 22972Date: Wed Jan 2 09:49:44 2008 +1000 22973 22974 r600: fix tv-out output naming 22975 22976commit f65374f5e15bfd391a1838a686cd87d3bab8043d 22977Author: Maciej Cencora <m.cencora@gmail.com> 22978Date: Mon Dec 31 09:44:34 2007 +1000 22979 22980 atombios: initial rs690 patches 22981 22982commit a674f683e6699c30664d9cd468a64de991c3fd7e 22983Author: Dave Airlie <airlied@linux.ie> 22984Date: Sun Dec 30 17:43:17 2007 +1000 22985 22986 atombios: enable TV-out detection code - tv out works for me with this 22987 22988commit aa7c28cbd943bb525698515d444cb5097880e364 22989Author: Dave Airlie <airlied@linux.ie> 22990Date: Sun Dec 30 17:40:37 2007 +1000 22991 22992 atombios: enable support for using tv timings 22993 22994 enable support for the atombios tv timings selection by programming the crtc 22995 with the tv timings if a tv is detected on the output 22996 22997commit 0bc3fd595a73e12a424571697d164a09a6a4c072 22998Author: Dave Airlie <airlied@linux.ie> 22999Date: Sun Dec 30 16:39:58 2007 +1000 23000 23001 atombios: add support for reading tv standard from atombios 23002 23003 fix typo in atombios header file 23004 23005commit bfa22d676a6f333503104041f62222f4de9bb7d8 23006Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23007Date: Wed Dec 26 12:29:47 2007 -0500 23008 23009 RADEON: fix typo that broke tv load detection 23010 23011commit 0c99554c6fab1192f5e8595469c21b5f33e1eb4f 23012Author: David Miller <davem@davemloft.net> 23013Date: Wed Dec 26 02:19:12 2007 -0500 23014 23015 [RADEON]: Add missing break in SCLK calculation. 23016 23017commit 6e0d5cc1c62fbfc1962fa0d6333f0c0a8c6327bd 23018Author: David Miller <davem@davemloft.net> 23019Date: Wed Dec 26 02:17:34 2007 -0500 23020 23021 [RADEON]: Like powerpc, don't use VGA by default on sparc. 23022 23023commit 2b1fae668ddabbc72e5fc31365302ea722174df1 23024Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23025Date: Wed Dec 26 02:04:06 2007 -0500 23026 23027 RADEON: fix PAL tv-out 23028 23029 Many thanks to Andrew Randrianasulu for providing me with 23030 pll reg dumps 23031 23032commit d736eb5732da573162c70712dc4e8b0114986702 23033Author: Alex Deucher <alex@botch2.(none)> 23034Date: Mon Dec 24 12:34:15 2007 -0500 23035 23036 RADEON: default "IgnoreLidStatus" to TRUE 23037 23038 Seems there are lots of busted ACPI lid status and people 23039 starting X with the lid closed. 23040 23041commit 653da558148cc601bc1f80253e92ef98c75ef37a 23042Author: Alex Deucher <alex@botch2.(none)> 23043Date: Mon Dec 24 01:11:56 2007 -0500 23044 23045 RADEON: restore crtc regs before VGA regs. 23046 23047 It seems some radeons don't restore text console properly if 23048 the crtc regs are restored after the VGA regs. 23049 Thanks to Sverre Froyen for helping me track this down 23050 23051commit ad3325f672a241449ca239c8ee3a24b6d7703d76 23052Author: Alex Deucher <alex@botch2.(none)> 23053Date: Sun Dec 23 17:18:42 2007 -0500 23054 23055 RADEON: Add "IgnoreLidStatus" option 23056 23057 Generally, users that start X with the laptop lid closed 23058 want to use one or more external monitors rather than the 23059 internal panel and do not want the internal panel to be on 23060 by default. Others, it seems, want to always have the 23061 internal panel on, regardless of the lid. I can't win. 23062 Enable this option to force the latter. 23063 23064commit 20eedf348a527e1e0a5450bc22d7564895034a66 23065Author: Alex Deucher <alex@botch2.(none)> 23066Date: Sun Dec 23 11:27:55 2007 -0500 23067 23068 RADEON: fix pll input setup on mac cards 23069 23070 the function was exiting before the complete setup was finished. 23071 23072commit 4f2e833e8ebaba3ad85ec5314fff8fa05395b679 23073Author: Arkadiusz Miskiewicz <arekm@maven.pl> 23074Date: Fri Dec 21 18:56:34 2007 -0500 23075 23076 configure.ac fixes 23077 23078commit 2b6e8e2b8f74e94560de89693ecbc7260536591e 23079Author: Arkadiusz Miskiewicz <arekm@maven.pl> 23080Date: Fri Dec 21 18:10:17 2007 -0500 23081 23082 RADEON: various cleanups 23083 23084commit 4c6f60e3b19ac55ab1255c79df03b1df5950864e 23085Author: Alex Deucher <alex@botch2.(none)> 23086Date: Fri Dec 21 17:33:04 2007 -0500 23087 23088 RADEON: clean up prototypes 23089 23090commit 8c761afdcb9baf1649b93449692fb9ab67bc2c80 23091Author: Alex Deucher <alex@botch2.(none)> 23092Date: Fri Dec 21 16:24:49 2007 -0500 23093 23094 RADEON: more re-org 23095 23096 move save/restore routines into legacy_crtc/output 23097 23098commit 910773d3a6c717f9d4762ea7b9ee6c3ae061781e 23099Author: Alex Deucher <alex@botch2.(none)> 23100Date: Fri Dec 21 15:53:15 2007 -0500 23101 23102 RADEON: more re-org. move XAA Mem init to radeon_accel.c 23103 23104commit 0631a23bd103f9b74e525da2c41304eab60c6f17 23105Author: Alex Deucher <alex@botch2.(none)> 23106Date: Fri Dec 21 15:40:18 2007 -0500 23107 23108 RADEON: fix rn50 reversion from last merge 23109 23110commit 5b917797a13c6caa80028d1842a284598e874288 23111Author: Alex Deucher <alex@botch2.(none)> 23112Date: Fri Dec 21 15:30:20 2007 -0500 23113 23114 RADEON: remove no longer used radeon_display.c 23115 23116commit bf14aa5f88fc3b4e69d71db5b23248b8bb2018d0 23117Author: Alex Deucher <alex@botch2.(none)> 23118Date: Fri Dec 21 15:29:47 2007 -0500 23119 23120 RADEON: re-org legacy crtc/output code into separate files 23121 23122commit be7f8fd338f5af8b632f16a83db41e15d00af469 23123Author: Dave Airlie <airlied@linux.ie> 23124Date: Fri Dec 21 10:03:49 2007 +1000 23125 23126 fix mode bandwidth configure check 23127 23128commit 3f9b597dedc45379b0bc0b631f3f924c403bca48 23129Author: Dave Airlie <airlied@linux.ie> 23130Date: Fri Dec 21 09:55:42 2007 +1000 23131 23132 fixup clip test include handling 23133 23134commit 9a5b501332c0a1f10af20845af48c9ddd2ce26a0 23135Author: Dave Airlie <airlied@linux.ie> 23136Date: Fri Dec 21 09:45:55 2007 +1000 23137 23138 set ddc line correctly post-merge 23139 23140commit 3c31b96afa20913ad947e68fe0c3a662e5eafbdd 23141Merge: eb99c3c5c f5e8c1850 23142Author: Dave Airlie <airlied@linux.ie> 23143Date: Fri Dec 21 09:36:22 2007 +1000 23144 23145 Merge remote branch 'origin/atombios-support' 23146 23147 Conflicts: 23148 23149 src/radeon_display.c 23150 src/radeon_driver.c 23151 23152commit eb99c3c5c9a2249cb84920f0f225e525fc3a4144 23153Author: Alex Deucher <alex@botch2.(none)> 23154Date: Thu Dec 20 18:14:38 2007 -0500 23155 23156 Bump for RC release 23157 23158commit 8d49ff1da917b7f8240267953ef6ce4ff04daecb 23159Author: Alex Deucher <alex@botch2.(none)> 23160Date: Thu Dec 20 01:09:57 2007 -0500 23161 23162 RADEON: check for xf86_crtc_clip_video_helper() in configure.ac 23163 23164 use xf86_crtc_clip_video_helper() from the server if available. 23165 23166commit fb7a4e24f2da3561ef81371ca4013a4f13806e91 23167Author: Adam Jackson <ajax@redhat.com> 23168Date: Wed Dec 19 19:15:19 2007 -0500 23169 23170 Fix RN50 mode filtering. 23171 23172 The old code would attempt to limit the maximum pixel size of the screen 23173 by limiting the maximum PLL frequency. This ends up confusing the PLL 23174 computation code since sometimes your maximum freq can be lower than your 23175 minimum freq. More to the point it's just wrong, maximum PLL frequency 23176 isn't the same thing as maximum pixel clock, and even that isn't the same 23177 thing as maximum scanout pixels per second. 23178 23179 The correct thing to do is filter by the mode's effective memory bandwidth. 23180 23181commit c1b0b69cc50516c3b7e881b0eb46cb3cd2e9dce6 23182Merge: 0e6634870 ce4fa1ced 23183Author: Alex Deucher <alex@botch2.(none)> 23184Date: Wed Dec 19 11:54:50 2007 -0500 23185 23186 Merge branch 'master' of ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati 23187 23188commit 0e6634870d1ab38ee8c83f6bda1ba60364997853 23189Author: Alex Deucher <alex@botch2.(none)> 23190Date: Wed Dec 19 11:54:27 2007 -0500 23191 23192 RADEON: skip empty connectors when creating outputs 23193 23194commit 6afbf718c151dc3c5c59bd3136b58a93a114d798 23195Author: Alex Deucher <alex@botch2.(none)> 23196Date: Wed Dec 19 11:48:38 2007 -0500 23197 23198 RADEON: add support for legacy radeons with DVI and no connector table 23199 23200commit f5e8c185001e62e744310667c2d1bd3fe6542a62 23201Author: Dave Airlie <airlied@clockmaker.usersys.redhat.com> 23202Date: Wed Dec 19 10:38:58 2007 +1000 23203 23204 more endian related fixage 23205 23206commit 98b247066d00db66abe91f518cd93b5c4da4cfb4 23207Author: Dave Airlie <airlied@clockmaker.usersys.redhat.com> 23208Date: Wed Dec 19 10:25:41 2007 +1000 23209 23210 fix big endian build since zaphod fixups 23211 23212commit ce4fa1cedec0cf56b9979dfaa12a8d3a7c643df4 23213Author: Arkadiusz Miskiewicz <arekm@maven.pl> 23214Date: Tue Dec 18 15:34:14 2007 -0500 23215 23216 RADEON: fix fd leak in lid detect code 23217 23218commit 20568f66f9a9a60a33bd9a69ccc14a891c656836 23219Author: Arkadiusz Miskiewicz <arekm@maven.pl> 23220Date: Tue Dec 18 15:32:10 2007 -0500 23221 23222 RADEON: more cleanups and warning fixes 23223 23224commit 1496194200adbcb044ec3977367a0908262e389c 23225Author: Arkadiusz Miskiewicz <arekm@maven.pl> 23226Date: Tue Dec 18 15:29:53 2007 -0500 23227 23228 RADEON: driver cleanups, warning fixes 23229 23230commit bd7206fa120495037e3fea0c920d0031d7715bf6 23231Author: Alex Deucher <alex@botch2.(none)> 23232Date: Tue Dec 18 03:03:11 2007 -0500 23233 23234 RADEON: fix another merge error 23235 23236 this broken legacy radeons 23237 23238commit 65a3ac7530e11bb7d818a988fd0cf1dde7688fa4 23239Author: Alex Deucher <alex@samba.(none)> 23240Date: Tue Dec 18 00:15:38 2007 -0500 23241 23242 RADEON: more PLL tweaks 23243 23244commit d93a0e10b8bc6e3797a3cf6c1e28ca413a7c38e4 23245Author: Alex Deucher <alex@samba.(none)> 23246Date: Mon Dec 17 20:32:45 2007 -0500 23247 23248 RADEON: post div tweaks for legacy radeon 23249 23250commit 03b8b49f6f502c45552b018fd8c44d366b2d576f 23251Author: Alex Deucher <alex@samba.(none)> 23252Date: Mon Dec 17 20:20:04 2007 -0500 23253 23254 RADEON: fix typo from merge 23255 23256commit 19b9d3708852b7efe2b05249c8359dadb924dd94 23257Merge: cf685f37e 29706ca58 23258Author: Alex Deucher <alex@samba.(none)> 23259Date: Mon Dec 17 20:07:32 2007 -0500 23260 23261 Merge branch 'atombios-support' of git+ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati into atombios-support 23262 23263commit cf685f37ec874f0aacd09e7c4eb0402c6daec1b0 23264Merge: 2a134af01 44d07c4cc 23265Author: Alex Deucher <alex@samba.(none)> 23266Date: Mon Dec 17 20:07:07 2007 -0500 23267 23268 Merge branch 'master' of git+ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati into atombios-support 23269 23270 merge master and fix conflicts 23271 23272commit 29706ca585ebd9b7b3521521a17016322e9ecccb 23273Author: Dave Airlie <airlied@linux.ie> 23274Date: Tue Dec 18 10:55:38 2007 +1000 23275 23276 fixup shadow setup on !r600 23277 23278commit 44d07c4ccce9acb5bd21a17acb082e91f7225764 23279Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23280Date: Mon Dec 17 18:56:12 2007 -0500 23281 23282 RADEON: typo from last commit 23283 23284commit 4da3782239921eb377216d4de4a9cc5bb55e0e8a 23285Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23286Date: Mon Dec 17 18:51:31 2007 -0500 23287 23288 RADEON: add output enable masks 23289 23290 add output enable masks for outputs that drive 23291 more than one connector. Make sure we don't turn off 23292 an output that's driving another connector. 23293 23294commit 5c5d2d19b2b032a06dd333b4ecc029aac342fb93 23295Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23296Date: Mon Dec 17 18:15:55 2007 -0500 23297 23298 RADEON: whitespace clean-ups 23299 23300commit 9f1d8220315c8894a17f2cc328025dc682b0c6e0 23301Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23302Date: Mon Dec 17 18:04:05 2007 -0500 23303 23304 RADEON: more PLL fixes 23305 23306 - reduce the calculation accuracy 23307 - certain LVDS panels seem to only like certain ref_divs 23308 - add pll flags to handle special cases 23309 - adjust the pll limits on legacy cards 23310 23311commit 2a134af01bc85de758ad83a85b23d1bba4a1a0f5 23312Author: Dave Airlie <airlied@redhat.com> 23313Date: Mon Dec 17 15:00:36 2007 +1000 23314 23315 r600: add shadow support to r600 driver to at least make 2d useable 23316 23317commit 614414611a9f246cbc74f579a79987fff97cf571 23318Author: Dave Airlie <airlied@redhat.com> 23319Date: Mon Dec 17 11:10:14 2007 +1000 23320 23321 radeon: cleanup pitch calculation and make r600 work again 23322 23323commit 79a375dbc7f323e2f551490a35f44ec36bed877c 23324Author: George Wu <geo@flood.OCF.Berkeley.EDU> 23325Date: Mon Dec 17 10:55:36 2007 +1000 23326 23327 r600: might as well fix VT for R600 23328 23329commit bc213ee723a45f2c396b4ed211a50f7642349973 23330Author: Alex Deucher <alex@samba.(none)> 23331Date: Sun Dec 16 14:54:00 2007 -0500 23332 23333 RADEON: fix sclock printout 23334 23335commit 4747c1f3cd4167b6a51d4864a297719ea48b9346 23336Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23337Date: Sun Dec 16 14:07:29 2007 -0500 23338 23339 RADEON: Make sure LVDS_EN bit is set when enabling LVDS 23340 23341commit a9817b2cf436a536dbc43ad77abc3bdcc53d346d 23342Author: Alex Deucher <alex@samba.(none)> 23343Date: Sat Dec 15 20:51:53 2007 -0500 23344 23345 RADEON: clean up units in PLL calculation 23346 23347commit b653e5a628bfa4dfb168e96f93f41eb910f409fb 23348Author: Michel Dänzer <michel@tungstengraphics.com> 23349Date: Sat Dec 15 00:50:10 2007 +0100 23350 23351 radeon: Default to 1x again with non-v3 AGP cards. 23352 23353 Seems more reliable in general than what was set up by firmware - fingers 23354 crossed... 23355 23356commit 6229825fa5d6715569098afbb21a40f7a2e7e6be 23357Author: Michel Dänzer <michel@tungstengraphics.com> 23358Date: Sat Dec 15 00:48:26 2007 +0100 23359 23360 radeon: Warning fixes. 23361 23362 The lid detection code probably wouldn't work on other non-x86 platforms 23363 though... 23364 23365commit 818ccf0fd4b5879171c5f20526d5a58638f8fde5 23366Author: Fredrik Höglund <fredrik@kde.org> 23367Date: Fri Dec 14 23:56:12 2007 +0100 23368 23369 RADEON: Fix the vertex coordinates for transformed pictures 23370 23371 This partially fixes transformed pictures on R100/R200 based 23372 cards. The texture still doesn't appear to be clamped correctly, 23373 but since that doesn't matter for rotations at perpendicular 23374 angles, I'm committing this now so randr rotation and reflection 23375 will work properly. 23376 23377commit 3cfbcf4cafbdfdb33411d16e51fb1f77cd0f93dd 23378Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23379Date: Fri Dec 14 17:11:00 2007 -0500 23380 23381 RADEON: Fix PLL set up on certain notebooks 23382 23383 Some LVDS panels require specific PLL dividers as 23384 specified in the bios tables. Make sure to use them 23385 if the output is LVDS. 23386 23387commit a84d446fd301d456bcea8f7abdc52e5a30776412 23388Author: Alex Deucher <alex@botch2.(none)> 23389Date: Fri Dec 14 02:17:14 2007 -0500 23390 23391 RADEON: select fb_div0 for LVDS on RV410 (x700) mobility 23392 23393 Fixes bug 8038 23394 I wonder if desktop RV410 need a similar fix. 23395 If your x700 laptop panel has problems after this let me know. 23396 23397commit b3eed3d87f76779b5a62a3115f99a31484dc38e0 23398Author: Alex Deucher <alex@samba.(none)> 23399Date: Fri Dec 14 00:20:10 2007 -0500 23400 23401 RADEON: fix typo in previous cursor fix 23402 23403commit 814c6c48aebba2e45ce257289b922cd7e92caf2a 23404Author: Alex Deucher <alex@samba.(none)> 23405Date: Thu Dec 13 18:45:09 2007 -0500 23406 23407 RADEON: rework PLL calculation 23408 23409 - Take into account the limits from the bios tables 23410 - Unify the PLL calculation between legacy and avivo chips 23411 23412commit f5ac34983411e4c4f41ab1817dce582830f398fd 23413Merge: f2b2e0804 6ccf5b33d 23414Author: Alex Deucher <alex@samba.(none)> 23415Date: Wed Dec 12 22:37:44 2007 -0500 23416 23417 Merge branch 'master' of git+ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati into atombios-support 23418 23419 merge and fix conflicts 23420 23421commit f2b2e0804183b52d9d3f56ad85b3552ece76c544 23422Author: Alex Deucher <alex@samba.(none)> 23423Date: Wed Dec 12 22:18:37 2007 -0500 23424 23425 RADEON: fix rotation on avivo chips 23426 23427 There are still some issues, but this is better than before. 23428 23429commit 6ccf5b33d27218ae1c45ab93c122438ed536d8ba 23430Author: Alex Deucher <alex@botch2.(none)> 23431Date: Wed Dec 12 20:12:06 2007 -0500 23432 23433 RADEON: only enable vblanks if we want them 23434 23435 should fix bug 13610 23436 23437commit 1668f2056f56370f1b5681c13f1e14904e301216 23438Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23439Date: Wed Dec 12 19:39:08 2007 -0500 23440 23441 RADEON: use /proc/acpi to determine lid status 23442 23443 Linux only 23444 23445commit 33a39947f7f79533cd90007a17d57b20126642c6 23446Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23447Date: Wed Dec 12 18:50:18 2007 -0500 23448 23449 RADEON: fix cursors when using rotation 23450 23451 allocate separate cursor buffers for each crtc 23452 23453commit 9e5efdecd12092031a4aebce58747cb4a6f48f28 23454Author: Arkadiusz Miskiewicz <arekm@maven.pl> 23455Date: Tue Dec 11 23:53:03 2007 -0500 23456 23457 sparse fixes and cleanups from arekm 23458 23459commit 372bf41818fdafc6a9d2914aee3a8e359f668f02 23460Author: Alex Deucher <alex@botch2.(none)> 23461Date: Tue Dec 11 14:04:58 2007 -0500 23462 23463 RADEON: handle HMDI properly (untested) and fix some merge leftovers 23464 23465commit 3c22ad977c25d5ca2811821fcac6bb8ecd79994a 23466Merge: c9a0cee97 f3d2ec3a5 23467Author: Alex Deucher <alex@botch2.(none)> 23468Date: Tue Dec 11 13:11:15 2007 -0500 23469 23470 Merge branch 'master' of ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati into atombios-support 23471 23472 fix conflicts 23473 23474commit f3d2ec3a5ae61215c792018320158750e7aa937c 23475Author: Alex Deucher <alex@botch2.(none)> 23476Date: Tue Dec 11 11:57:27 2007 -0500 23477 23478 RADEON: rewrite PLL computation 23479 23480 Algorithm adapted from BeOS radeon driver with some tweaks by me. 23481 Actually calulate and use the reference divider rather than using the bios default. 23482 Also, always calculate the PLL, rather than falling back to bios dividers. 23483 This should fix bugs 12913, 13590, 13533, possibly others. 23484 23485commit c9a0cee97ca69e8fe1e1937c7670fa903214cded 23486Author: Dave Airlie <airlied@linux.ie> 23487Date: Tue Dec 11 06:03:46 2007 +1000 23488 23489 more zaphod fixes - some other work maybe needed 23490 23491commit 9b125312ab6edc585e4f5931a6a6de81e13b6acc 23492Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23493Date: Mon Dec 10 13:53:15 2007 -0500 23494 23495 RADEON: only update crtc values when RMX is active 23496 23497commit 3a161e1b5d80361e318ced8da5c19e797749d693 23498Author: Alex Deucher <alex@botch2.(none)> 23499Date: Mon Dec 10 00:57:26 2007 -0500 23500 23501 RADEON: bios PLL cleanup 23502 23503commit 731830297f2fc4a416882aacfb0d9b5f8ed32520 23504Author: Dave Airlie <airlied@linux.ie> 23505Date: Mon Dec 10 15:50:38 2007 +1000 23506 23507 fixup some warnings 23508 23509commit 2818e2b02ca90c9dfa50905b5311b2ae83ac3b0c 23510Author: Dave Airlie <airlied@linux.ie> 23511Date: Mon Dec 10 15:43:52 2007 +1000 23512 23513 add more to configure.ac for using out-of-tree mode src 23514 23515commit cc3c36100986f9d8060bc2d433373d4806f8e730 23516Author: Dave Airlie <airlied@linux.ie> 23517Date: Mon Dec 10 15:25:56 2007 +1000 23518 23519 add support for building against legacy servers similiar to Intel codepaths 23520 23521commit 9c278cb7fa7f18d13bde053fd75221cfba9da377 23522Merge: 6451ea2dc cc167b9bb 23523Author: Dave Airlie <airlied@linux.ie> 23524Date: Mon Dec 10 15:18:03 2007 +1000 23525 23526 Merge branch 'zaphod-lolz' of git://git.freedesktop.org/git/xorg/driver/xf86-video-ati into atombios-support 23527 23528 Conflicts: 23529 23530 src/radeon.h 23531 src/radeon_crtc.c 23532 src/radeon_driver.c 23533 src/radeon_output.c 23534 23535commit 6451ea2dcc4fac762442f699935864f4a8d445f7 23536Merge: 0d89556bf 5896ca409 23537Author: Dave Airlie <airlied@linux.ie> 23538Date: Mon Dec 10 15:08:42 2007 +1000 23539 23540 Merge branch 'master' into atombios-support 23541 23542commit 5896ca4097d439f59f90f397939132c061c3c364 23543Author: LisaWu <liswu@ati.com> 23544Date: Fri Dec 7 09:45:05 2007 +0100 23545 23546 radeon: Use %u instead of %d for unsigned value. 23547 23548commit df44f8380268c27d3978c4e91d736f093322b8b8 23549Author: Michel Dänzer <michel@tungstengraphics.com> 23550Date: Fri Dec 7 09:41:47 2007 +0100 23551 23552 radeon: Use gettimeofday instead of xf86getsecs. 23553 23554commit cc167b9bb7f1c3b8579e51e7bc2fca2f8eba6bd1 23555Author: Dave Airlie <airlied@redhat.com> 23556Date: Fri Dec 7 15:41:36 2007 +1000 23557 23558 disable tiling for zaphod heads 23559 23560commit 2ce8d192533a8c795714c5a9fb308ec74db40287 23561Author: Dave Airlie <airlied@redhat.com> 23562Date: Fri Dec 7 15:35:21 2007 +1000 23563 23564 don't add fboffset to info->FB it already is mapped at the offset 23565 23566commit 0dcd926d3092100854b3e362d6659d4950508aeb 23567Author: Dave Airlie <airlied@redhat.com> 23568Date: Fri Dec 7 14:45:04 2007 +1000 23569 23570 radeon: bring back zaphod all is forgiven. 23571 23572 You've whined, you've cried, you've nagged, and you're guilt trippin has 23573 made me do it... It actually wasn't as hard as I thought it would be. 23574 23575 Still not perfect, couple of things to fix yet 23576 23577commit bb5ede557bf32a42eef158ff0fbcfe1c6ede098a 23578Author: Dave Airlie <airlied@redhat.com> 23579Date: Fri Dec 7 14:30:32 2007 +1000 23580 23581 radeon: move savedreg/modereg into entity instead of info 23582 23583commit 64ab1cdf343a9a69e7e9e64f0bba77c54a94e9d0 23584Author: James Cloos <cloos@jhcloos.com> 23585Date: Thu Dec 6 15:51:12 2007 -0500 23586 23587 Add missing PHONY line for automatic ChangeLog generation 23588 23589commit 0d89556bfa41a3acbd6afe85b062e3a21f2ca057 23590Author: Dave Airlie <airlied@ppcg5.localdomain> 23591Date: Thu Dec 6 19:23:06 2007 +1100 23592 23593 powerpc: build fixes from last merge 23594 23595commit dbe3d2608ecc9896db9c23b3a347b50748c51e13 23596Merge: 48e31cdaa 21ed43539 23597Author: Dave Airlie <airlied@redhat.com> 23598Date: Thu Dec 6 14:22:03 2007 +1000 23599 23600 Merge branch 'master' into atombios-support 23601 23602 Conflicts: 23603 23604 src/radeon_output.c 23605 23606commit 21ed435398e4a398dd8a0a5d7c1d4cc45e916332 23607Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23608Date: Tue Dec 4 17:08:58 2007 -0500 23609 23610 RADEON: add MacModel imac-g5-isight for iMac G5 iSight 23611 23612 Thanks to Étienne Bersac for helping to figure this out. 23613 23614commit 54bfd522405d9fdfb69d3a59e111ac3d63483dbb 23615Author: Étienne Bersac <bersace03@laposte.net> 23616Date: Tue Dec 4 14:22:42 2007 -0500 23617 23618 RADEON: fix typo 23619 23620commit 5022d006cfc06ca0395981526b2c2c94c6878567 23621Author: Michel Dänzer <michel@tungstengraphics.com> 23622Date: Sun Dec 2 17:27:33 2007 +0100 23623 23624 radeon: Further XVideo fixes. 23625 23626 * Make sure pitch constraints are always met for DMA upload blits. 23627 * RGB24 is not affected by endianness. 23628 23629commit 6ed55b70b23dfdc7b41103ea59c1df2bda5e41e6 23630Author: Kusanagi Kouichi <slash@ma.neweb.ne.jp> 23631Date: Sun Dec 2 17:18:46 2007 +0100 23632 23633 radeon: Fix crash with XVideo 24bit RGB images. 23634 23635 See https://bugs.freedesktop.org/show_bug.cgi?id=13274 . 23636 23637commit a697b590899bb7704ec4d7ae9a9c3cbbfcaef382 23638Author: Michel Dänzer <michel@tungstengraphics.com> 23639Date: Sun Dec 2 17:11:20 2007 +0100 23640 23641 Fix build against xserver master. 23642 23643 (DE)ALLOCATE_LOCAL are gone. 23644 23645commit 48e31cdaa0caa21573879af5b9267773fe89176a 23646Author: George Wu <geo@ocf.berkeley.edu> 23647Date: Sun Dec 2 15:25:09 2007 +1000 23648 23649 RADEON/R600: small code cleanup 23650 23651commit 00b4480aa2c5d7f751e34fc964f431b90b14c8d2 23652Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 23653Date: Sat Dec 1 14:18:40 2007 -0500 23654 23655 RADEON: add options for force TV out as detected and to set TV standard 23656 23657 Also fix a typo in internal tv-out parsing 23658 23659commit 1e029fef5fe264f2ced445b80bf6070abcb84b82 23660Author: Alex Deucher <alex@samba.(none)> 23661Date: Sat Dec 1 00:58:51 2007 -0500 23662 23663 RADEON: move GPIO lookup to a separate function 23664 23665commit dcbef1ba9dfcf35c28e058832a55adf00afb472e 23666Author: Alex Deucher <alex@samba.(none)> 23667Date: Sat Dec 1 00:35:25 2007 -0500 23668 23669 RADEON: fix typo in previous commit 23670 23671 check gpio table revision before connector table revision 23672 23673commit fdce0598a2228c48c84deae1d7bebb2d7b3e979b 23674Author: Alex Deucher <alex@samba.(none)> 23675Date: Sat Dec 1 00:15:34 2007 -0500 23676 23677 RADEON: convert atombios connector table parsing to use ATOM structs 23678 23679 convert and add hpd info 23680 23681commit e3d7de9cc956aec5f940ad6db09e826b3a69523a 23682Author: Alex Deucher <alex@botch2.(none)> 23683Date: Fri Nov 30 20:14:42 2007 -0500 23684 23685 RADEON: remove unused cruft from last atom import 23686 23687commit d5d83411e8a884154d671aad440524507cce313e 23688Author: Alex Deucher <alex@botch2.(none)> 23689Date: Fri Nov 30 20:11:42 2007 -0500 23690 23691 RADEON: save/restore avivo crtc cursor control 23692 23693 this should prevent the cursor from showing up on in text 23694 mode or vesafb etc. after running the driver. 23695 23696commit af0196f7bf0d1d5d211391149c18935d64ed2b06 23697Merge: d9858a2b3 0175b7998 23698Author: Alex Deucher <alex@botch2.(none)> 23699Date: Fri Nov 30 16:40:28 2007 -0500 23700 23701 Merge branch 'master' of ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati into atombios-support 23702 23703 merge master and fix conflicts 23704 23705commit d9858a2b3744b99003cfb9f31b743a2d31b322e9 23706Author: Dave Airlie <airlied@linux.ie> 23707Date: Sat Dec 1 06:49:59 2007 +1000 23708 23709 radeon: add in pll spread spectrum workaround 23710 23711commit 0175b79987ef4d7b0ce8238c3bdde989e504516a 23712Author: Alex Deucher <alex@botch2.(none)> 23713Date: Fri Nov 30 15:37:42 2007 -0500 23714 23715 RADEON: rework MacModel option 23716 23717 this brings in some previous research from Michel Dänzer, 23718 Sjoerd Simons, and myself. Hopefully, the driver will pick 23719 the correct MacModel in more cases. This also changes the 23720 default connector table for desktop Macs to dual DVI rather 23721 than DVI+VGA as that seems to be the case more often than not. 23722 External TMDS chips are handled separately now as well. 23723 Eventually we should add an option to allow the user to specify 23724 what external TMDS chip they need, but we don't have enough info 23725 yet, so we'll rely on OF to init the external chip in most cases 23726 for now. 23727 23728commit e1945f1f25a34310bd58ce128c8ff27ecc985618 23729Merge: b368b0f22 df7777bff 23730Author: Alex Deucher <alex@botch2.(none)> 23731Date: Fri Nov 30 14:30:55 2007 -0500 23732 23733 Merge branch 'atombios-support' of ssh://agd5f@git.freedesktop.org/git/xorg/driver/xf86-video-ati into atombios-support 23734 23735commit b368b0f22cd1d7ef9b4c65d82929c76f3b82d573 23736Author: Alex Deucher <alex@botch2.(none)> 23737Date: Fri Nov 30 14:29:27 2007 -0500 23738 23739 RADEON: disable atom pll set for r4xx cards 23740 23741 the clocks do not get set correctly in all cases. this needs 23742 further investigation. 23743 23744commit 5af15739571c09260750bcfd3620e16fd7fec862 23745Author: Alex Deucher <alex@botch2.(none)> 23746Date: Fri Nov 30 14:24:30 2007 -0500 23747 23748 RADEON: small cleanup of pll code 23749 23750commit df7777bff40c1feabcc12d2148ad6ac5213efbb3 23751Author: George Wu <geo@ocf.berkeley.edu> 23752Date: Fri Nov 30 17:49:33 2007 +1000 23753 23754 Add LVTMA PWRSEQ registers to fix VT switching for LVDS 23755 23756commit 9840a0fd4fc8c980533fcd4a02c55cd0d5634b6d 23757Author: Alex Deucher <alex@botch2.(none)> 23758Date: Thu Nov 29 13:27:37 2007 -0500 23759 23760 RADEON: add MacModel "mini-internal" for minis with internal TMDS 23761 23762 Some macs (minis and powerbooks) use internal tmds, others use external tmds 23763 and not just for dual-link TMDS, it shows up with single-link as well. 23764 Unforunately, there doesn't seem to be any good way to figure it out. 23765 23766commit 017c939cf0a2b12fbdc1681cc70c28b23ae3b397 23767Author: Alex Deucher <alex@samba.(none)> 23768Date: Thu Nov 29 02:52:14 2007 -0500 23769 23770 RADEON: implement CLUT adjust support 23771 23772commit 9963b0fe01feb6dd0cb555b874a48f6fa3b255cb 23773Author: Alex Deucher <alex@samba.(none)> 23774Date: Thu Nov 29 00:46:23 2007 -0500 23775 23776 RADEON: fix cursor offset on avivo chips 23777 23778commit 6c56e3d7655b17e93e8823aefe34b05291104695 23779Author: Alex Deucher <alex@botch2.(none)> 23780Date: Tue Nov 27 15:27:36 2007 -0500 23781 23782 RADEON: switch r4xx to atombios load detection 23783 23784 works great 23785 23786commit 7561242e5b79bc2798ca3aace2b79e1a36949488 23787Author: Alex Deucher <alex@botch2.(none)> 23788Date: Tue Nov 27 14:50:36 2007 -0500 23789 23790 RADEON: re-org load detection for legacy chips 23791 23792commit febdcc2dccd42acbcd68ae630b7811cae5c58e8a 23793Author: Dave Airlie <airlied@linux.ie> 23794Date: Wed Nov 28 05:10:57 2007 +1000 23795 23796 legacy: fix fb/agp read/writes 23797 23798commit bb8545146959b748994be055d5b3de66ec66c8b2 23799Author: Alex Deucher <alex@samba.(none)> 23800Date: Mon Nov 26 17:34:51 2007 -0500 23801 23802 RADEON: first pass at TV/Component video 23803 23804 Untested and not likely to work just yet. 23805 23806commit e2bde646b864dca9056d9ecfe23a0d905647ea9a 23807Author: Alex Deucher <alex@samba.(none)> 23808Date: Mon Nov 26 14:35:57 2007 -0500 23809 23810 RADEON: move crtc output source selection into atombios_output.c 23811 23812 The function fits better as an output function and should now 23813 work with clones as well. 23814 23815commit 16c9be4107678a2a58d3418b7f1cc94d695ca8d6 23816Author: Alex Deucher <alex@samba.(none)> 23817Date: Mon Nov 26 14:20:54 2007 -0500 23818 23819 RADEON: add default connector table for avivo chips 23820 23821commit 6f080d00e6f4f84d5e0d6b4eff302bf42c230e81 23822Author: Arkadiusz Miskiewicz <arekm@maven.pl> 23823Date: Mon Nov 26 12:43:30 2007 -0500 23824 23825 RADEON: fix backlight control on some laptops 23826 23827 It seems the bios scratch regs are involved in backlight control 23828 on some laptops. This patch fixes the problematic laptops and doesn't 23829 seem to break the previous bios lid and output control fixes. 23830 23831commit dcf22aed87366f4625fb5042cb84fecccd9ceece 23832Author: Alex Deucher <alex@botch2.(none)> 23833Date: Mon Nov 26 11:10:03 2007 -0500 23834 23835 RADEON: only return status unknown for XPRESS chips 23836 23837 this seems to cause more issues than it attempted to fix 23838 so limit it to XPRESS chips for now. 23839 23840commit 206e280f02324641b4fe5a1986e26adf0e021fd4 23841Author: Alex Deucher <alex@botch2.(none)> 23842Date: Mon Nov 26 09:39:27 2007 -0500 23843 23844 RADEON: fix typo in man page 23845 23846commit 4e792db655dc92d2864db36b4d8f6714908de8e8 23847Author: Dave Airlie <airlied@redhat.com> 23848Date: Fri Nov 23 15:44:44 2007 +1000 23849 23850 r500: set default minimum pll freq 23851 23852commit a13b4c69c105c096dd05e6de2c5c154c9a8bcc71 23853Author: Dave Airlie <airlied@redhat.com> 23854Date: Fri Nov 23 15:25:06 2007 +1000 23855 23856 r5xx: cleanup pll code.. 23857 23858 Clean the PLL code to use the radeon pll structs. 23859 23860commit 5d792b5673bbf4784eb0ec059221e9b57232a122 23861Author: Dave Airlie <airlied@redhat.com> 23862Date: Fri Nov 23 15:03:13 2007 +1000 23863 23864 radeon: fix up memory mapping issues for vt switch 23865 23866commit 558a2ef266c1ca517c7fb464b0ccfef83238c913 23867Author: Dave Airlie <airlied@redhat.com> 23868Date: Fri Nov 23 14:39:32 2007 +1000 23869 23870 fix silly spaces 23871 23872commit 1bda4424a7031de437acfca9c41d4b3668e36051 23873Author: Dave Airlie <airlied@redhat.com> 23874Date: Fri Nov 23 14:39:19 2007 +1000 23875 23876 r600: add hi agp address for mc 23877 23878commit 133234c31a294f24a3968a576aad2bb8b89d0f6a 23879Author: Dave Airlie <airlied@redhat.com> 23880Date: Fri Nov 23 14:15:18 2007 +1000 23881 23882 atombios: use values from object header 23883 23884commit dbf6eae7e7a4bd1bc60fefdc7ab6276ed3f097c4 23885Author: Dave Airlie <airlied@redhat.com> 23886Date: Fri Nov 23 11:55:05 2007 +1000 23887 23888 atombios: add initial object header parsing for r600 cards 23889 23890commit 6b103915c11fc79d2efc43c44fc2a00c3bc64ede 23891Author: Dave Airlie <airlied@redhat.com> 23892Date: Fri Nov 23 09:24:20 2007 +1000 23893 23894 r500: make vt switch work again for me 23895 23896commit d24208276aad7669feeed527dced60c76d39eae6 23897Author: Dave Airlie <airlied@redhat.com> 23898Date: Fri Nov 23 08:50:25 2007 +1000 23899 23900 avivo: fix typo in register saving 23901 23902commit 197a62704742a4a19736c2637ac92d1dc5ab34ed 23903Author: Adam Jackson <ajackson@redhat.com> 23904Date: Thu Nov 22 20:26:23 2007 +1000 23905 23906 radeon: fix openoffice/render bug on r100 chips 23907 23908commit 64010fc4eae8359c01e430f64252931242c91435 23909Author: Dave Airlie <airlied@linux.ie> 23910Date: Thu Nov 22 20:25:31 2007 +1000 23911 23912 Revert "Disable RENDER acceleration by default on some RV200 chips." 23913 23914 This reverts commit 145da701bf4fb9c0ad9f95620b20914ae0126852. 23915 23916 pull in fix from ajax next commit 23917 23918commit 145da701bf4fb9c0ad9f95620b20914ae0126852 23919Author: Stefan Dirsch <sndirsch@suse.de> 23920Date: Thu Nov 22 08:38:09 2007 +0100 23921 23922 Disable RENDER acceleration by default on some RV200 chips. 23923 23924 Novell Bug #341699: Render acceleration is known to be broken 23925 on at least "Radeon 7500 QW" and "Radeon Mobility M7 LW". 23926 23927commit e810c3ae9908cd57e95b1b091cded87cbfc12fdc 23928Author: Roland Scheidegger <sroland@tungstengraphics.com> 23929Date: Thu Nov 22 02:37:55 2007 +0100 23930 23931 really do not set up surface regs for depth buf on r100-class igps (bug #13080) 23932 23933 fix the if condition testing for these chips... 23934 23935commit 0cc7e94849c1525750fabd04cf58f2dee88372e0 23936Author: Alex Deucher <alex@botch2.(none)> 23937Date: Wed Nov 21 17:06:17 2007 -0500 23938 23939 RADEON: reorder crtc/pll setup 23940 23941commit d56bde98efceaa8344e62f8e98db90c4bb642331 23942Author: Alex Deucher <alex@botch2.(none)> 23943Date: Wed Nov 21 17:03:39 2007 -0500 23944 23945 RADEON: fix crtc to output routing 23946 23947 Thanks to AMD for the information 23948 23949commit a12390c832abe423def60e39cd5a9118e5910339 23950Merge: d5317922f e74dca194 23951Author: Alex Deucher <alex@botch2.(none)> 23952Date: Wed Nov 21 02:24:48 2007 -0500 23953 23954 Merge branch 'atombios-support' of git://git.freedesktop.org/git/xorg/driver/xf86-video-ati into atombios-support 23955 23956commit d5317922f29a57b6c4127826a2fc126c5fd7c117 23957Author: Alex Deucher <alex@botch2.(none)> 23958Date: Wed Nov 21 02:23:37 2007 -0500 23959 23960 RADEON: attempt to fix crtc to output routing 23961 23962 The output routing seems to be based on the output ids from the bios 23963 connector tables and the connected status in the bios scratch regs. 23964 I don't fully understand this yet, but this seems to work 23965 for the most part, however changing modes can sometimes 23966 lead to a blanked head. This can be remedied by forcing 23967 a dpms off cycle. 23968 23969commit 81ce299bffd75540925b4c8234adf11226147165 23970Author: Alex Deucher <alex@botch2.(none)> 23971Date: Wed Nov 21 01:35:44 2007 -0500 23972 23973 RADEON: provide clearer debugging info 23974 23975commit e74dca19416b13f97db9d1fc06299b988057d6a4 23976Author: Dave Airlie <airlied@redhat.com> 23977Date: Wed Nov 21 16:24:25 2007 +1000 23978 23979 re-enable mobility chips 23980 23981commit 9c5b813dd6b3492cbc9833bc59792a5cec457e51 23982Author: Alex Deucher <alex@botch2.(none)> 23983Date: Wed Nov 21 01:22:42 2007 -0500 23984 23985 RADEON: major re-org and clean up of atom output control 23986 23987 - use radeon_output->devices to determine output 23988 - clean up and simplify dpms and mode set 23989 23990commit 7634cb6b96f938bc6615eb2c49ae75aaefd04cce 23991Author: Alex Deucher <alex@botch2.(none)> 23992Date: Wed Nov 21 00:10:14 2007 -0500 23993 23994 RADEON: make naming consistent and remove some cruft 23995 23996commit 908748343fc9a6cdc38af0fc028c63a82766da3f 23997Author: Alex Deucher <alex@botch2.(none)> 23998Date: Wed Nov 21 00:05:42 2007 -0500 23999 24000 RADEON: store devices ids from bios for each driver output 24001 24002commit 3975da2ea8cb628f7f66c3f26c5dfa181cd1c532 24003Merge: e283aa332 295ce277b 24004Author: Alex Deucher <alex@botch2.(none)> 24005Date: Tue Nov 20 23:52:29 2007 -0500 24006 24007 Merge branch 'atombios-support' of git://git.freedesktop.org/git/xorg/driver/xf86-video-ati into atombios-support 24008 24009commit e283aa332adf0134243a4fa3d14263719cd8a3fd 24010Author: Alex Deucher <alex@botch2.(none)> 24011Date: Tue Nov 20 23:35:46 2007 -0500 24012 24013 RADEON: add LVDS atom support 24014 24015commit e4bc3e1e7bb45571367d41b5328ff2590810b0f9 24016Author: Alex Deucher <alex@botch2.(none)> 24017Date: Tue Nov 20 18:09:29 2007 -0500 24018 24019 RADEON: enable/disable the right TMDS controller 24020 24021commit 3e47683ffaa44a89cda9bcddf530643befb27efa 24022Author: Alex Deucher <alex@botch2.(none)> 24023Date: Tue Nov 20 18:01:15 2007 -0500 24024 24025 RADEON: fixup for bios tables with wrong connector types 24026 24027commit 7412952eb1d1e9857cdab8417f7305f676900827 24028Author: Alex Deucher <alex@botch2.(none)> 24029Date: Tue Nov 20 18:00:12 2007 -0500 24030 24031 RADEON: switch to using ATOM defines for bios device table 24032 24033commit 295ce277bb0a44b9539b3dba575e7aff279dc2d0 24034Author: Dave Airlie <airlied@redhat.com> 24035Date: Wed Nov 21 08:56:40 2007 +1000 24036 24037 add missing files for make dist 24038 24039 noted by ndim on #radeonhd 24040 24041commit 20083b0695987b25e442ecbdec24f3cb6f1ac2ae 24042Author: Dave Airlie <airlied@redhat.com> 24043Date: Wed Nov 21 08:53:44 2007 +1000 24044 24045 LVDS on r500/r600 needs some work disable mobile chips for now 24046 24047commit e4b8a4479ddea9b083b3a763dc0b9302e7b9a82a 24048Author: Dave Airlie <airlied@redhat.com> 24049Date: Wed Nov 21 08:06:12 2007 +1000 24050 24051 r600: add memory controller regs from AMD 24052 24053commit aa88da974b97ea1e9bbb47b3494543575c09d912 24054Author: Dave Airlie <airlied@redhat.com> 24055Date: Wed Nov 21 08:01:35 2007 +1000 24056 24057 rs690 is !> r600 24058 24059commit c8872603454e6a4ffed9fc7d9adc2c364a429608 24060Author: Dave Airlie <airlied@clockmaker.usersys.redhat.com> 24061Date: Tue Nov 20 22:33:39 2007 +1000 24062 24063 radeon: restructure pci ids to avoid effort later 24064 24065 This uses a single file with all the pciids and parameters for radeon 24066 family and setup in it. I don't run the perl script at build time to avoid 24067 a perl dependency on build but adding pci ids should be done via the csv file 24068 with openoffice or gnumeric if possible. 24069 24070commit a5b34c2f1f7d5346c4489cb30e07291d1217026c 24071Author: Dave Airlie <airlied@redhat.com> 24072Date: Tue Nov 20 16:37:29 2007 +1000 24073 24074 r600: use standard memory controller setup paths 24075 24076commit 4a523da5221d53f2efa49da2326500e9b0b9f14d 24077Author: Dave Airlie <airlied@redhat.com> 24078Date: Tue Nov 20 15:31:11 2007 +1000 24079 24080 r600: get r600 to work non-accelerated. 24081 24082 DDC still not working yet 24083 24084commit 80023441ba46882bc810ff3790c7148059f155f5 24085Author: Dave Airlie <airlied@redhat.com> 24086Date: Tue Nov 20 14:10:23 2007 +1000 24087 24088 r600: block r600 startup due to lack of memory controller info 24089 24090commit f6fbbacc17bf9b1073d3e993b225987fd9173182 24091Author: Dave Airlie <airlied@redhat.com> 24092Date: Tue Nov 20 13:41:55 2007 +1000 24093 24094 atombios: add warnings for r500 and r600 24095 24096commit 45a8b083c123b820c008f04ab857a64a8facec14 24097Author: Dave Airlie <airlied@redhat.com> 24098Date: Tue Nov 20 13:37:00 2007 +1000 24099 24100 atombios: add all r5xx and r6xx pci ids 24101 24102commit 5d023e2c3c2ab44ea57ffadc9607025d602c376c 24103Merge: 0d1e0c780 c88726034 24104Author: Dave Airlie <airlied@redhat.com> 24105Date: Tue Nov 20 13:02:43 2007 +1000 24106 24107 Merge branch 'master' into atombios-support 24108 24109 Conflicts: 24110 24111 src/radeon_chipset.h 24112 src/radeon_driver.c 24113 src/radeon_probe.c 24114 24115commit 0d1e0c7805b3d8e56ccb49465e6b144afb7bdc51 24116Author: Dave Airlie <airlied@redhat.com> 24117Date: Tue Nov 20 09:08:04 2007 +1000 24118 24119 r5xx: add 71c5 for macbook pro 24120 24121commit d5909b30595c103bb5f42cd1704330f944bba49c 24122Author: Dave Airlie <airlied@redhat.com> 24123Date: Tue Nov 20 08:15:58 2007 +1000 24124 24125 r5xx: cleanups after last merge 24126 24127commit fe2f7a09050fb7a345a1f52239f8f3c4f1053891 24128Merge: 744c8cb6c 49055d8af 24129Author: Dave Airlie <airlied@redhat.com> 24130Date: Tue Nov 20 08:04:32 2007 +1000 24131 24132 Merge branch 'master' into agd-atom-merge 24133 24134 Conflicts: 24135 24136 src/radeon_cursor.c 24137 src/radeon_output.c 24138 24139commit 744c8cb6c293fcaa687566f52901644e699baace 24140Merge: e258fbe41 e530af79a 24141Author: Dave Airlie <airlied@redhat.com> 24142Date: Tue Nov 20 07:56:33 2007 +1000 24143 24144 Merge branch 'agd-atom' of ssh://people.freedesktop.org/~agd5f/xf86-video-ati-atom into agd-atom 24145 24146commit e258fbe411d255a1044b61d7ff738aee3fb5b7f4 24147Author: Dave Airlie <airlied@redhat.com> 24148Date: Mon Nov 19 16:35:05 2007 +1000 24149 24150 makes 2-headed cursor work 24151 24152commit e530af79adf51b3e95a0eca676c915a34dcbf4a7 24153Merge: 69e197f2c 52aba8d73 24154Author: Alex Deucher <alex@botch2.(none)> 24155Date: Mon Nov 19 00:59:30 2007 -0500 24156 24157 Merge branch 'agd-atom' of /home/alex/git/airlied/xf86-video-ati2 into agd-atom 24158 24159commit 69e197f2c8002aacf2587754c8d3bd63c88f85b1 24160Merge: 5e8940fa6 862dcabfe 24161Author: Alex Deucher <alex@botch2.(none)> 24162Date: Mon Nov 19 00:57:34 2007 -0500 24163 24164 Merge branch 'agd-atom' of /home/alex/git/airlied/xf86-video-ati2 into agd-atom 24165 24166commit fca47ad083449f4cf9063dd970cdcebea6a7f110 24167Author: Dave Airlie <airlied@redhat.com> 24168Date: Mon Nov 19 15:53:40 2007 +1000 24169 24170 add z3ro's pciids 24171 24172commit 5e8940fa6e33d09091aa4bcf04b0f9e79596e1b8 24173Author: Alex Deucher <alex@botch2.(none)> 24174Date: Mon Nov 19 00:52:38 2007 -0500 24175 24176 fix logic in connector table check for TVs and switch counter to symbolic names 24177 24178commit c19123fd9483758eb6b286c3dffcb6d79d5b1ee5 24179Author: Dave Airlie <airlied@redhat.com> 24180Date: Mon Nov 19 15:46:58 2007 +1000 24181 24182 add firegl card on ajaxs machine 24183 24184commit f02f340e466a415b4e01648ca1e323f4ce125885 24185Author: Alex Deucher <alex@botch2.(none)> 24186Date: Mon Nov 19 00:39:19 2007 -0500 24187 24188 Don't assign a gpio for TV 24189 24190commit 52aba8d73189ba959f19c0437499d5e7a8829827 24191Merge: 862dcabfe 5e8940fa6 24192Author: Dave Airlie <airlied@redhat.com> 24193Date: Mon Nov 19 15:30:46 2007 +1000 24194 24195 Merge branch 'agd-atom' of ssh://people.freedesktop.org/~agd5f/xf86-video-ati-atom into agd-atom 24196 24197commit 862dcabfe0c10751d815e5cdd7436c10c2c2db10 24198Author: Dave Airlie <airlied@redhat.com> 24199Date: Mon Nov 19 15:30:08 2007 +1000 24200 24201 r520: nail i2c enable/disable issue 24202 24203commit f3dd7f413b670eeb6b8639f6677d72050ad5fe04 24204Author: Alex Deucher <alex@botch2.(none)> 24205Date: Mon Nov 19 00:19:39 2007 -0500 24206 24207 Don't detect TV out for now 24208 24209commit 94de0e22d7229ca71e18e1e849d8545d9ca7bafe 24210Author: Dave Airlie <airlied@redhat.com> 24211Date: Mon Nov 19 14:47:33 2007 +1000 24212 24213 i2c: fix bus enable stuff 24214 24215commit 3f1fc7eef13ea02fa5119e9b51d499841b801f2d 24216Author: Alex Deucher <alex@botch2.(none)> 24217Date: Mon Nov 19 00:02:14 2007 -0500 24218 24219 CRTs/DFPs may share a DVI port, but TV and CV don't 24220 24221commit 8f84c5ad4c4af14612ea68fe6f24d0d527f00acc 24222Author: Alex Deucher <alex@botch2.(none)> 24223Date: Sun Nov 18 23:43:06 2007 -0500 24224 24225 fix typo in loop 24226 24227commit 384cd8f52c89d089c6559e2eedbae45641fcd14e 24228Merge: f3f0e4ec9 234b60730 24229Author: Dave Airlie <airlied@redhat.com> 24230Date: Mon Nov 19 14:02:55 2007 +1000 24231 24232 Merge branch 'agd-atom' of ../xf86-video-ati into agd-atom 24233 24234commit f3f0e4ec92c935c89ddb2f4241fe4335a521b439 24235Author: Alex Deucher <alex@botch2.(none)> 24236Date: Sun Nov 18 23:14:01 2007 -0500 24237 24238 RADEON: unify connectortype handling 24239 24240commit 234b6073054ac7630e82781683e666b94b2f12de 24241Author: Dave Airlie <airlied@redhat.com> 24242Date: Mon Nov 19 14:02:09 2007 +1000 24243 24244 restore avivo memory map registers at correct places 24245 24246commit 459a30ba511fe2fa8051380a9741fcfd9bb401ef 24247Author: Dave Airlie <airlied@redhat.com> 24248Date: Mon Nov 19 13:44:38 2007 +1000 24249 24250 fix type for r520 agp code 24251 24252commit 760af92412ef0d5cc44e52e7cec11fd80c4aaaeb 24253Author: Alex Deucher <alex@botch2.(none)> 24254Date: Sun Nov 18 22:34:59 2007 -0500 24255 24256 RADEON: unify DDC line handling 24257 24258commit e73bf6290da20dd61798ace775999ce1cb550934 24259Author: Dave Airlie <airlied@redhat.com> 24260Date: Mon Nov 19 13:32:16 2007 +1000 24261 24262 add x1900xt support 24263 24264commit 2e37937bacd624d616b91c41006c113791ebe98d 24265Author: Alex Deucher <alex@botch2.(none)> 24266Date: Sun Nov 18 20:18:50 2007 -0500 24267 24268 RADEON: step one in output rework 24269 24270 re-organize the output type 24271 24272commit 679e7a2e0d1b213524b8109193483bc9840fb116 24273Author: Alex Deucher <alex@botch2.(none)> 24274Date: Sun Nov 18 17:56:51 2007 -0500 24275 24276 Few fixes from the last commit. 24277 24278 Update parser works fine on r4xx. 24279 24280commit 1cd7cc3e6758ab1012f3ced6e958a1517f45557f 24281Author: Alex Deucher <alex@botch2.(none)> 24282Date: Sun Nov 18 17:44:36 2007 -0500 24283 24284 WIP: new atom code comples. 24285 24286 I commented out the object parsing for the time being as 24287 using it will require some thought as to new output 24288 related data structures. 24289 24290commit b155fa872ee4ca5d801e942aee6e619cef104f35 24291Author: Alex Deucher <alex@botch2.(none)> 24292Date: Sat Nov 17 00:34:56 2007 -0500 24293 24294 WIP: more new ATOM integration work 24295 24296commit 67db114d97abed7a607467e5d67c7b4ffa2c347e 24297Merge: 7d06a8791 ea1534659 24298Author: Alex Deucher <alex@botch2.(none)> 24299Date: Fri Nov 16 14:29:53 2007 -0500 24300 24301 Merge branch 'agd-atom' of /home/alex/git/airlied/xf86-video-ati2 into agd-atom 24302 24303commit ea1534659de87d3d75eb20d808d039cff22cb537 24304Author: Dave Airlie <airlied@linux.ie> 24305Date: Fri Nov 16 18:46:02 2007 +1000 24306 24307 avivo: fixup some i2c stuff 24308 24309commit 7d06a8791839ce6b22e2449646832b79cebf1b21 24310Author: Alex Deucher <alex@botch2.(none)> 24311Date: Fri Nov 16 02:43:00 2007 -0500 24312 24313 WIP: sync up with the latest ATOM bios code in rhd 24314 24315 doesn't compile ATM 24316 24317commit 3614d80ceb9a7b3615b0baab3cf2dd34ed4ab464 24318Author: Dave Airlie <airlied@linux.ie> 24319Date: Fri Nov 16 17:22:39 2007 +1000 24320 24321 add missing hpd register 24322 24323commit 9a2715fda97ac0ebcb45650a416e0652aab575b8 24324Author: Dave Airlie <airlied@linux.ie> 24325Date: Fri Nov 16 17:22:25 2007 +1000 24326 24327 make i2c unlock/lock registers for gpios 24328 24329commit 20dc549fff9a4137c93ebed449d05e0c437b6bc1 24330Author: Dave Airlie <airlied@linux.ie> 24331Date: Fri Nov 16 17:01:33 2007 +1000 24332 24333 avivo i2c: consolidate the avivo i2c code 24334 24335commit 3e62730f79a13883a65a568bc821bc56055a4ab7 24336Author: Dave Airlie <airlied@linux.ie> 24337Date: Fri Nov 16 15:19:00 2007 +1000 24338 24339 atombios: fixup warnings in atombios files 24340 24341commit cca7af3c4910983f7f090792986fcbfa0dc97cfb 24342Author: Dave Airlie <airlied@linux.ie> 24343Date: Fri Nov 16 15:04:01 2007 +1000 24344 24345 remove avivo_reg.h 24346 24347commit d39eb2077c6b2fc094ccd952772528eb9428c587 24348Author: Dave Airlie <airlied@linux.ie> 24349Date: Fri Nov 16 15:00:50 2007 +1000 24350 24351 radeon: rename a large section of avivo regs to documented names 24352 24353commit b7774c28dde72a205a40be78003df72eabfb9b1f 24354Author: Dave Airlie <airlied@linux.ie> 24355Date: Fri Nov 16 14:48:36 2007 +1000 24356 24357 Add copyright headers 24358 24359commit 3cfcd2164b400bd0d1cb4ede8eeb01abba9d75c8 24360Merge: efac14e66 718bfd3b6 24361Author: Alex Deucher <alex@botch2.(none)> 24362Date: Thu Nov 15 23:25:39 2007 -0500 24363 24364 Merge branch 'agd-atom' of /home/alex/git/airlied/xf86-video-ati2 into agd-atom 24365 24366commit efac14e669a0c6184f8848191eb49ffb21934ee1 24367Author: Dave Airlie <airlied@linux.ie> 24368Date: Thu Nov 15 23:17:25 2007 -0500 24369 24370 r5xx: fix typo for crtc offset 24371 24372commit e6db621c37ff615be286462f000d67a662c5c331 24373Author: Alex Deucher <alex@botch2.(none)> 24374Date: Thu Nov 15 23:15:56 2007 -0500 24375 24376 fix INMC() and OUTMC() on !AVIVO chips 24377 24378 WR_EN is bit 8 so don't use OUTREG8. 24379 24380commit 52ba3fdd1ce05983fabedff234cfaf4c60fba38d 24381Author: Alex Deucher <alex@botch2.(none)> 24382Date: Thu Nov 15 23:12:30 2007 -0500 24383 24384 atombios_dac_detect() takes care of primary vs tv dac itself 24385 24386commit 49055d8aff91ff12186feaf5343c8fd2f96bcba0 24387Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24388Date: Thu Nov 15 22:56:09 2007 -0500 24389 24390 RADEON: set proper defaults for tv dac BGADJ/DACADJ 24391 24392 we should get these values from the bios tables, but for now use 24393 some reasonable defaults. This should fix the washed out color 24394 problems on bugs 1082 and 12844. 24395 24396commit 821acf38b716ab87c3d07263d6e4a139fe54803f 24397Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24398Date: Thu Nov 15 22:28:42 2007 -0500 24399 24400 RADEON: Make sure we set the MT properly for connected status unknown 24401 24402commit a94123f33ec6584fbdfc4b9ecd543d1357de8814 24403Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24404Date: Thu Nov 15 22:19:54 2007 -0500 24405 24406 Revert "Portability fix from netbsd" 24407 24408 This reverts commit c9264aa53bf1470ad9104d1e7c4a8ce13c49c270. 24409 This breaks damage support. See bug 13244 24410 24411commit 718bfd3b61879172eee819fdab7080d5d4c0a756 24412Author: Dave Airlie <airlied@linux.ie> 24413Date: Fri Nov 16 10:37:04 2007 +1000 24414 24415 r5xx: fix typo for crtc offset 24416 24417commit e9d721c31372db045550f9562534b28f16121bb9 24418Author: Roland Scheidegger <sroland@tungstengraphics.com> 24419Date: Tue Nov 13 23:42:42 2007 +0100 24420 24421 ignore sometime bogus agp_mode bit from chip (bug #13190) 24422 24423 bit is wrong on at least X700 cards with rialto pcie-agp bridge chip, 24424 should be safe to use just the bit from the bridge hopefully to make 24425 agp setup work on these cards and not adversely affect others. 24426 24427commit b865886d00907899297ae864358eb26d9980975d 24428Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24429Date: Sun Nov 11 20:38:41 2007 -0500 24430 24431 bump for RC release 24432 24433commit 7aeb35e5ad1aed6e78a3d8565fbfbfe66232ab45 24434Author: Alex Deucher <alex@botch2.(none)> 24435Date: Sun Nov 11 19:29:30 2007 -0500 24436 24437 fix from last commit 24438 24439commit 2ea95900547165e86ad3f8a41ce3331a05bad60e 24440Author: Alex Deucher <alex@botch2.(none)> 24441Date: Sun Nov 11 18:43:03 2007 -0500 24442 24443 Add full parsing support for atom bios connector table 24444 24445commit 7ce730828c293f0810dfdc554df48dfd76e35c49 24446Author: Alex Deucher <alex@botch2.(none)> 24447Date: Sun Nov 11 12:37:01 2007 -0500 24448 24449 rework crtc output source setup 24450 24451commit d61b6c78aa7810a2f9b9e2d9d95aab4295de80ce 24452Author: Alex Deucher <alex@botch2.(none)> 24453Date: Sun Nov 11 11:58:17 2007 -0500 24454 24455 make sure i2c bus exists before using it 24456 24457commit 342e3e207efda42ba679731c30dfb9d5e9d5643f 24458Author: Alex Deucher <alex@botch2.(none)> 24459Date: Fri Nov 9 17:11:43 2007 -0500 24460 24461 combine outputs based on id 24462 24463commit 8078c299d5941460243944d55051547c1a4d3791 24464Author: Alex Deucher <alex@botch2.(none)> 24465Date: Fri Nov 9 16:35:08 2007 -0500 24466 24467 use atom to program plls on r4xx 24468 24469commit 5febe2c96642f61d006abe6e8081e69d5b95adc0 24470Author: Alex Deucher <alex@botch2.(none)> 24471Date: Fri Nov 9 16:24:56 2007 -0500 24472 24473 turn off vga control when using ext modes 24474 24475commit 83f170c32c08c74a9e3466ffa0e0a0606c74427b 24476Author: Alex Deucher <alex@botch2.(none)> 24477Date: Thu Nov 8 19:33:13 2007 -0500 24478 24479 add pci ids for X1550PRO 24480 24481commit 5cdcaba0f6e9de2d15cfcc109ab97d9fd423e3bf 24482Author: Alex Deucher <alex@botch2.(none)> 24483Date: Thu Nov 8 19:28:03 2007 -0500 24484 24485 make sure to assign gpio 24486 24487commit 2dcb852778301b9284a2b4906dcf64f95ed638b7 24488Author: Alex Deucher <alex@botch2.(none)> 24489Date: Thu Nov 8 18:39:23 2007 -0500 24490 24491 pull in another of Dave's fixes 24492 24493commit 96273016a0bbdfa4d3a4e6275a3b09eeeadaa534 24494Author: Alex Deucher <alex@botch2.(none)> 24495Date: Wed Nov 7 01:22:30 2007 -0500 24496 24497 fix and move crtc source set up atombios_output.c 24498 24499 it's really more output related. 24500 24501commit 5c495c81cc3bcd4a38d06954243ed3bdc85bdc07 24502Author: Alex Deucher <alex@botch2.(none)> 24503Date: Wed Nov 7 01:04:11 2007 -0500 24504 24505 add support for initing external tmds via ATOM 24506 24507commit 5c13d9355280e6de44ebbf8de7ea89a6b91c7388 24508Author: Alex Deucher <alex@botch2.(none)> 24509Date: Wed Nov 7 00:17:28 2007 -0500 24510 24511 add avivo output stuff 24512 24513commit 68e7f5c67e2e9d2162b469ce31f452f3f89756b5 24514Author: Alex Deucher <alex@botch2.(none)> 24515Date: Tue Nov 6 23:43:29 2007 -0500 24516 24517 more avivo updates 24518 24519commit 303562dfb57e13c027b2aa9289d54e547c829ff1 24520Author: Alex Deucher <alex@botch2.(none)> 24521Date: Tue Nov 6 23:06:46 2007 -0500 24522 24523 add additional connector types 24524 24525commit 0d3e0735f710cb7b9505e4330997aa332f73c102 24526Author: Alex Deucher <alex@botch2.(none)> 24527Date: Tue Nov 6 22:59:25 2007 -0500 24528 24529 First round of avivo support 24530 24531commit 0abfe3150ce3eed4db93ccc2975bd4622dfa54a7 24532Author: Alex Deucher <alex@botch2.(none)> 24533Date: Tue Nov 6 18:47:00 2007 -0500 24534 24535 Add atombios files 24536 24537commit 20f01950e42babc308b4470df6a3c6628c932003 24538Author: Alex Deucher <alex@botch2.(none)> 24539Date: Tue Nov 6 18:04:43 2007 -0500 24540 24541 for r4xx ATOM cards, just use ATOM for PLL 24542 24543 while crtc timing and pll seem to work fine, output setup 24544 and routing don't seem to work too reliably with atom. 24545 AMD claims ATOM was still pretty new with r4xx so 24546 it's probably better to stick with direct programming for 24547 some things. 24548 24549commit 37af2f6b52c8f217b205486742a6b46a864da8ed 24550Author: Michel Dänzer <michel@tungstengraphics.com> 24551Date: Mon Nov 5 10:58:25 2007 +0100 24552 24553 radeon: Use indirect register access for (un)hiding the cursor. 24554 24555 This avoids hangs that occur when touching these registers while the card is 24556 busy with some cards. 24557 24558commit 5659e2f1b08e44d0c4f21a403e91b4801ef78f62 24559Author: Michel Dänzer <michel@tungstengraphics.com> 24560Date: Mon Nov 5 10:32:04 2007 +0100 24561 24562 radeon: Fix build warning on powerpc. 24563 24564commit d41d73f5f9d8baee7137e94da2c7852ea0412180 24565Author: Michel Dänzer <michel@tungstengraphics.com> 24566Date: Mon Nov 5 10:30:23 2007 +0100 24567 24568 Revert "radeon: Reinstate sync in radeon_crtc_show/hide_cursor." 24569 24570 This reverts commit 0241cac643fa1c08a45ea44f5c670b290e760ad8. 24571 24572 Taking the DRI lock in paths called from the SIGIO handler isn't safe. For 24573 example, the SIGIO handler may be called while the lock is held by an AIGLX 24574 context, resulting in deadlock. See e.g. 24575 https://bugs.freedesktop.org/show_bug.cgi?id=13005 . 24576 24577commit 78a3eabff382e8ebe33df2039076fb083bcc361b 24578Author: Alex Deucher <alex@botch2.(none)> 24579Date: Sun Nov 4 14:11:26 2007 -0500 24580 24581 WIP: get ATOM crtc stuff working on r4xx 24582 24583commit 5e9ebd8e496b72b051053d637c63b2956b7861d3 24584Author: Alex Deucher <alex@botch2.(none)> 24585Date: Sat Nov 3 20:46:17 2007 -0400 24586 24587 Initial support for r4xx 24588 24589 - add r3xx/r4xx MC access macros and functions 24590 24591commit c106075ccb81ca2ee4894743e676fd37653c8dce 24592Author: Alex Deucher <alex@botch2.(none)> 24593Date: Sat Nov 3 18:29:20 2007 -0400 24594 24595 More of Dave's ATOM init code. 24596 24597commit e18f5d61806b445ad77d93e258fbce9422b52bb6 24598Author: Alex Deucher <alex@botch2.(none)> 24599Date: Sat Nov 3 18:20:55 2007 -0400 24600 24601 Initial integration of Atom code and some of Dave's code. 24602 24603 ATOM builds, but it's not hooked up yet. 24604 24605commit aaafc6a8f33d4736c37c43d3fd5e43085a2590fb 24606Author: Brice Goglin <bgoglin@debian.org> 24607Date: Sun Oct 28 21:10:32 2007 +0100 24608 24609 Fix typo in 'disposing cursor' 24610 24611 Bug #12984. 24612 24613commit a5c55c418319c1315ad73c305ef48b44fe1d9580 24614Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24615Date: Sat Oct 27 19:46:11 2007 -0400 24616 24617 RADEON: disable FP DETECT for DVI. 24618 24619 These bits don't seem to be reliable or routed right 24620 in all cases. 24621 24622commit 49cf7cb3db36ce7734f7e314a040a240191d6477 24623Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24624Date: Thu Oct 25 21:23:09 2007 -0400 24625 24626 RADEON: Fix IgnoreEDID option 24627 24628commit 70bbc0cfe699588f8ae4123efca0e1baffc91000 24629Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24630Date: Thu Oct 25 20:56:11 2007 -0400 24631 24632 RADEON: clean up mode handling and improve reliability of DDC 24633 24634 - Reduce the number of times we do DDC and don't 24635 skip the magical GPIO dance when doing DDC. 24636 - If no DDC modes are found, add the screen modes 24637 24638commit 1a4f590c2ac071f0c348f38e9f26fa6093722453 24639Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24640Date: Thu Oct 25 19:25:36 2007 -0400 24641 24642 RADEON: document DefaultTMDSPLL option 24643 24644commit 2d26e0d0bcf25e210dc7771e76638dcfafcc2356 24645Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24646Date: Thu Oct 25 19:18:18 2007 -0400 24647 24648 RADEON: Add driver option to force TV DAC load detection on 24649 24650commit a9306b7986467c0794c91625ad2ece597b718462 24651Author: Alex Deucher <alex@botch2.(none)> 24652Date: Thu Oct 18 20:06:58 2007 -0400 24653 24654 RADEON: fix possible crash in radeon_crtc_mode_set() 24655 24656 when radeon_crtc_mode_set() is called during ScreenInit(), 24657 we don't have pScrn->pScreen yet. 24658 24659commit 115b4b65ef679a46d90b7fc8ac22ace37038b3fc 24660Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24661Date: Wed Oct 17 20:25:47 2007 -0400 24662 24663 RADEON: Remove LVDSBiosNativeMode Option 24664 24665 Shouldn't be needed any longer as I've finally sorted out 24666 the LVDS issues due to the crtc setup. 24667 24668commit f2eb5c1cc69a4f7b0754ec6b2efde4fa1650ab6d 24669Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24670Date: Tue Oct 16 18:54:28 2007 -0400 24671 24672 RADEON: finish fixing LVDS/RMX 24673 24674 For panels, both the mode and crtc values need to be updated based 24675 on the native mode timing. This shoudl fix any remaining problems 24676 with RMX. 24677 24678commit 93561688087493fccb7ac387d728feec61212180 24679Author: Michel Dänzer <michel@tungstengraphics.com> 24680Date: Tue Oct 16 10:57:06 2007 +0200 24681 24682 radeon: Handle 64 bit PCI resource bases better. 24683 24684commit 08945fc8a14ba57e32350e85daa3354619e9e72e 24685Author: Alex Deucher <alex@botch2.(none)> 24686Date: Mon Oct 15 21:12:46 2007 -0400 24687 24688 RADEON: allow the user to set any tv standard 24689 24690 allow the user to set any tv standard rather than 24691 limiting it to what the bios says is supported. 24692 24693commit 571548b2b95ce289f26a77c11deb6266f0b18516 24694Author: Alex Deucher <alex@botch2.(none)> 24695Date: Mon Oct 15 20:33:56 2007 -0400 24696 24697 RADEON: use native timing for RMX 24698 24699 RMX should work again with native timings after the previous 24700 patch. 24701 24702commit 2f87bff293a343b40c1be096933a5ae126632468 24703Author: Alex Deucher <alex@botch2.(none)> 24704Date: Mon Oct 15 20:06:28 2007 -0400 24705 24706 RADEON: Fix subtle change in crtc reg init 24707 24708 At some point we changed how hsync_wid and vsync_wid were clipped. 24709 Previously we used the field size as a mask when building the 24710 register. This got changed to setting the value to the field 24711 size if it was greater. this probably explains number stange 24712 mode bugs. 24713 24714commit fd5bb7bb5e968127b87102320eccc1222f205e5a 24715Author: Michel Dänzer <michel@tungstengraphics.com> 24716Date: Sat Oct 13 18:56:43 2007 +0200 24717 24718 radeon: Improve detection of default value for Option "MacModel" on Linux. 24719 24720 * Detect all PowerBooks with dual link DVI (according to developer.apple.com) 24721 and both Mac Mini G4 models. Other PowerBooks and iBooks should be covered 24722 by the 'detected as' line in /proc/cpuinfo. 24723 * Give specific instructions for overriding and reporting incorrect detection. 24724 * Only perform detection when Option "MacModel" isn't present with a valid 24725 value. 24726 * Close /proc/cpuinfo file handle when done with it. 24727 * Coding style cleanups. 24728 24729commit fbf121add5584049627f07345fc502b2aefc88ce 24730Author: Eugen Dedu <Eugen.Dedu@pu-pm.univ-fcomte.fr> 24731Date: Sat Oct 13 16:51:58 2007 +0200 24732 24733 radeon: Derive default value of Option "MacModel" from /proc/cpuinfo on Linux. 24734 24735commit 5db3afaa1fdb69d382ac769ef40191a4b964d28e 24736Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24737Date: Thu Oct 11 20:25:20 2007 -0400 24738 24739 RADEON: return status unknown for flaky chip/connector combinations 24740 24741 This should at least get something on the screen. 24742 XPRESS chips, I'm looking in your general direction... 24743 24744commit 1148d332f52f3780897aae3fcd5d6a67687c42cd 24745Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24746Date: Thu Oct 11 19:09:45 2007 -0400 24747 24748 RADEON: fix potential crash in mode handling 24749 24750commit 905b19df33d6f912336b65895e97775ffec07728 24751Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24752Date: Thu Oct 11 18:55:08 2007 -0400 24753 24754 RADEON: small cleanup from last commit 24755 24756commit 7afd04c1e4ffa6e4e5ba08ae90ba002237dc282b 24757Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24758Date: Thu Oct 11 00:16:45 2007 -0400 24759 24760 RADEON: tell the bios not to muck with the hardware while the driver is active 24761 24762 by toggling the appropriate bios scratch regs you can tell 24763 the bios not the touch the hw while the driver is active. 24764 This should prevent the bios from scrambling the hardware 24765 when users open the lid or toggle bios hotkeys. 24766 24767 fixes bug 12567 24768 24769commit 1b231d28fdda5cdc44bb9d2075d4edfd8f17e21f 24770Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24771Date: Wed Oct 10 21:54:46 2007 -0400 24772 24773 RADEON: attempt to detect lid status on legacy bios 24774 24775 I'm not sure how reliable this is. The lid closed 24776 bit only seems to be set if the lid is closed at boot 24777 so LVDS will default to on if you restart X with the 24778 lid closed after boot. 24779 24780commit 47ef583b993607bc3a372dd125c6e5228c4ec7bf 24781Author: Alex Deucher <alex@botch2.(none)> 24782Date: Wed Oct 10 19:35:31 2007 -0400 24783 24784 RADEON: more XPRESS connector table hacks 24785 24786 Seems XPRESS desktop chips with DVI-D list a proprietary 24787 connector type rather than DVI-D. try and do the right thing. 24788 Also, XPRESS chips seem to include connector type 8 and ddc 24789 type 8 which are unknown at this time. 24790 24791commit c9264aa53bf1470ad9104d1e7c4a8ce13c49c270 24792Author: Joerg Sonnenberger <joerg@britannica.bec.de> 24793Date: Tue Oct 9 23:18:29 2007 -0400 24794 24795 Portability fix from netbsd 24796 24797commit 80eee856938756e1222526b6c39cee8b5252b409 24798Author: Matthieu Herb <> 24799Date: Tue Oct 9 16:17:50 2007 -0400 24800 24801 RADEON: fix console restore on netbsd 24802 24803 Include the mode restore bugfix from monolithic Xorg, that is derived 24804 from the version in xsrc which in turn was provided by Matthieu Herb 24805 over 3 years ago on the XFree86 lists. Suggested by various 24806 developers, hold-back due to the working state in xorg-server 1.1.1. 24807 Tracing down the exact change showed that the changed default color 24808 depth made this issue a lot more prominent again. Discussed with Eric 24809 Anholt. 24810 24811commit 547fef4c6382f8a4951c086ee531b804cba31075 24812Author: Lars Weiler <pylon@gentoo.org> 24813Date: Tue Oct 9 16:10:34 2007 -0400 24814 24815 include stdint.h to fix the build on some platforms 24816 24817 fixes bug 12733 24818 24819commit ddec5783103045084a66a34d71326cc6e4a3f598 24820Author: George Sapountzis <gsap7@yahoo.gr> 24821Date: Tue Oct 9 13:27:27 2007 +0300 24822 24823 theatre compiles with pci-rework. 24824 24825commit 051435610a66735fd455bbb526fa294fcfe8c0b6 24826Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24827Date: Sun Oct 7 19:39:47 2007 -0400 24828 24829 RADEON: still more LVDS fixes 24830 24831 Seems some laptops need the native mode from the bios for 24832 LVDS while others seem to prefer a CVT mode. Add an option 24833 to pick the preferred mode. The default it to use the bios 24834 table timing. 24835 24836commit 0b03a73b7dcb4aa192c42f2a4c842d324c358122 24837Author: Dave Airlie <airlied@linux.ie> 24838Date: Sat Oct 6 20:11:19 2007 +1000 24839 24840 radeon: add support for DDC type 6 24841 24842 thanks to arekm on xorg-devel for testing/feedback 24843 24844commit 83d7506a603889b6c18736cff9b2cd3296c95c84 24845Author: Alex Deucher <alex@botch2.(none)> 24846Date: Fri Oct 5 20:25:04 2007 -0400 24847 24848 Bump for RC release 24849 24850commit fe66cc4e02a52e1d014ae5de192c65e8abd7f58a 24851Author: Tormod Volden <bugzi06.fdo.tormod@xoxy.net> 24852Date: Fri Oct 5 20:10:55 2007 -0400 24853 24854 RADEON: rename to PCI_CHIP_RS485_5975 everywhere 24855 24856 RS482_5975 is actually RS485. see bug 12048 24857 24858commit 314e8fdbe5369e4c1786d176d03d0d2a046f25d5 24859Author: Tormod Volden <bugzi06.fdo.tormod@xoxy.net> 24860Date: Fri Oct 5 20:05:54 2007 -0400 24861 24862 RADEON: Disable DRI by default for RS482 cards 24863 24864 fixes bug 12048 24865 24866commit abbafe6db7c0d24efc61a05950d4214c00d3f043 24867Author: Michel Dänzer <michel@tungstengraphics.com> 24868Date: Thu Oct 4 11:33:09 2007 +0200 24869 24870 radeon: Fix compile warnings. 24871 24872commit 1110957ba703475847b8e72023c6edcbd1068aef 24873Author: Michel Dänzer <michel@tungstengraphics.com> 24874Date: Thu Oct 4 11:32:48 2007 +0200 24875 24876 radeon: Fix for pci-rework. 24877 24878 Surprisingly easy, thanks to George's pci-rework changes. 24879 24880commit 7addf41885ec5658f531624a9c24ea5bd7d22d19 24881Merge: 3a958ba13 78c6bd305 24882Author: Michel Dänzer <michel@tungstengraphics.com> 24883Date: Thu Oct 4 11:06:08 2007 +0200 24884 24885 Merge remote branch 'origin/pci-rework' 24886 24887commit 3a958ba136c3fae5a4ddd56373ac7cd47046f10e 24888Author: Alex Deucher <alex@botch2.(none)> 24889Date: Wed Oct 3 22:59:14 2007 -0400 24890 24891 RADEON: remove some cruft 24892 24893commit b6bda79f72df5e5bf9c6b71fa3298e765da506bd 24894Author: Alex Deucher <alex@botch2.(none)> 24895Date: Wed Oct 3 22:56:27 2007 -0400 24896 24897 RADEON: more fixes for single crtc chips 24898 24899commit 5f5c4e6ad61c45c24f1443b91b4bc5375efdebc0 24900Author: Alex Deucher <alex@botch2.(none)> 24901Date: Wed Oct 3 22:46:21 2007 -0400 24902 24903 RADEON: remove RADEONSaveMode() 24904 24905 It's only called by RADEONSave(), so just move its 24906 contents into RADEONSave(). this mirrors RADEONRestore(). 24907 24908commit bfede412b3a3cd11769a580b167c528734146096 24909Author: Alex Deucher <alex@botch2.(none)> 24910Date: Wed Oct 3 22:19:16 2007 -0400 24911 24912 RADEON: minor fixes for external TMDS 24913 24914commit 597dffce9bdc200003d0be880235258386a0bdd7 24915Author: Alex Deucher <alex@botch2.(none)> 24916Date: Wed Oct 3 21:43:08 2007 -0400 24917 24918 RADEON: final fix for RMX/LVDS 24919 24920 It seems that on quite a few laptops the LVDS H/V timing 24921 from the bios tables are way off or just doesn't work period. 24922 Either we are using it wrong or we need some additional checks 24923 when we parse it. Only the dot clock seems to really matter, 24924 so use cvt modes and update the dotclock. This seems to work 24925 correctly in all cases. 24926 24927commit cc0c2d8e61600652b1f9cb3dc49db2ef62b1e40d 24928Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24929Date: Wed Oct 3 00:59:37 2007 -0400 24930 24931 RADEON: Finally sort out LVDS modes 24932 24933 - The panel timing from the bios tables is for the crtc, 24934 not the native panel mode, so add cvt modes for the panel 24935 and fix up the crtc values in mode_fixup(). 24936 - rename and reconfigure RADEONValidateFPModes() to what it 24937 really does: add screen modes. I suppose for backwards 24938 compatibility we ought to add the screen modes to every 24939 reasonable output, but everyone should really use xrandr 24940 or an output monitor section to add custom modes. Hopefully 24941 this will go away at some point. 24942 24943commit 0ca184c3c35032df39ea7ce5d2d4aba1a97b6426 24944Author: Alex Deucher <alex@botch2.(none)> 24945Date: Tue Oct 2 19:38:18 2007 -0400 24946 24947 RADEON: fix RADEONInitDispBandwidth() on single crtc cards 24948 24949 Should fix bug 12490 24950 24951commit a5fb2a5d54cdd562a97881cdea1a4af35b483aab 24952Author: Dave Airlie <airlied@linux.ie> 24953Date: Mon Oct 1 08:48:27 2007 +1000 24954 24955 radeon: fix powerpc build 24956 24957commit 22519fde1e002f28d6036d448fcd18452d00f1bb 24958Author: Alex Deucher <alex@botch2.(none)> 24959Date: Sun Sep 30 13:11:20 2007 -0400 24960 24961 RADEON: add support for ext tmds table and ext tmds chip init 24962 24963 This probably won't work on all chips as the various gpio lines 24964 seem to need special magic to to actually talk to the i2c slave 24965 chips. 24966 24967commit dcc376e2d2a13329dd03f1bc4b471329757a6f5f 24968Author: Alex Deucher <alex@botch2.(none)> 24969Date: Wed Sep 26 23:27:38 2007 -0400 24970 24971 RADEON: remove cruft 24972 24973commit 99ceaefa18c6e07b55106cca0ea8996fa73667be 24974Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 24975Date: Sun Sep 30 12:02:54 2007 -0400 24976 24977 RADEON: more work on LVDS mode validation and fixups 24978 24979 Hopefully this will fix up the weird mode issues with LVDS and 24980 native and RMX modes, but I can't seem to get quite the right 24981 combo to fix everyone. 24982 24983commit d808781d48adf01e80b5bb476bae2d2f599030f1 24984Author: Dave Airlie <airlied@redhat.com> 24985Date: Fri Sep 28 14:50:42 2007 +1000 24986 24987 rc410: disable DRI by default due to it not working 24988 24989commit 4d9f944e40c6b9ac7e80428dff49c6e83fda1d8f 24990Author: Michel Dänzer <michel@tungstengraphics.com> 24991Date: Thu Sep 27 17:50:15 2007 +0200 24992 24993 radeon: Fix build on little endian platforms. 24994 24995 This was accidentally broken when moving away from using pixmap data pointers 24996 directly. 24997 24998commit 5902e913010a2fb0c17efe01f9aade73f9646f16 24999Author: Michel Dänzer <michel@tungstengraphics.com> 25000Date: Thu Sep 27 12:53:34 2007 +0200 25001 25002 radeon: EXA pitch fixes. 25003 25004 The combined pitch/offset registers only support pitches up to 16320 bytes. 25005 25006 With EXA >= 2.3, set the maxPitchBytes field accordingly. With older versions, 25007 limit maxX such that the pitch of 32bpp pixmaps doesn't exceed the limit. 25008 25009 Also check the limit in RADEONGetOffsetPitch just in case. 25010 25011 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=12551 . 25012 25013commit ca728973bce504e03bf6228d18287906c35eaa3f 25014Author: Michel Dänzer <michel@tungstengraphics.com> 25015Date: Thu Sep 27 11:14:21 2007 +0200 25016 25017 radeon: Don't use pixmap data pointer directly with EXA. 25018 25019 Fixes segfault with current xserver master without AccelDFS. 25020 25021commit 21593d04d222b05dbba9abd31eaa3bfb91d999b6 25022Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25023Date: Mon Sep 24 23:01:20 2007 -0400 25024 25025 RADEON: more clean up of mode code 25026 25027 also print out LVDS mode info from bios for both 25028 legacy and atom bioses 25029 25030commit 22cccb99d762567757d3fd38795f71e943f1faf9 25031Author: Alex Deucher <alex@botch2.(none)> 25032Date: Mon Sep 24 00:10:01 2007 -0400 25033 25034 bump for RC release 25035 25036commit fad9e7b00de051f632000eb2304ac23a524f7c8e 25037Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25038Date: Sun Sep 23 23:43:26 2007 -0400 25039 25040 RADEON: Rework LVDS mode validation 25041 25042 The old validation code wasn't really well suited to randr. 25043 This fixes several issues: 25044 - missing display size for panels with edid 25045 - broken duplicate modes 25046 25047commit bf9674b9df351108e069d037ec10869a4154a881 25048Author: Alex Deucher <alex@botch2.(none)> 25049Date: Sun Sep 23 15:01:34 2007 -0400 25050 25051 RADEON: properly set default RMX types 25052 25053commit b27135bce8b41d69290613b440a338b0a7fe0200 25054Author: Alex Deucher <alex@botch2.(none)> 25055Date: Sun Sep 23 14:44:38 2007 -0400 25056 25057 RADEON: fix up LVDS handling for r3xx and newer 25058 25059commit d5cf268be2c9b3e320d7101f5213f7d5609b6308 25060Author: Alex Deucher <alex@botch2.(none)> 25061Date: Sat Sep 22 18:31:51 2007 -0400 25062 25063 RADEON: Fix crash when Xv window is outside of either crtc 25064 25065commit 8aca0ca6be7122e0ece20985a1862553a1b225c0 25066Author: Alex Deucher <alex@botch2.(none)> 25067Date: Sat Sep 22 17:56:50 2007 -0400 25068 25069 Revert "RADEON: fix crash when Xv window is outside of either crtc" 25070 25071 This reverts commit 9109e62e3be7f96b41b534ab517fdf1baf458806. 25072 25073 This breaks ABI. better fix to come. 25074 25075commit d6e1d9be1cd5786a24ed4bb8e45fa7e04df56622 25076Author: Alex Deucher <alex@botch2.(none)> 25077Date: Sat Sep 22 16:14:06 2007 -0400 25078 25079 RADEON: Don't make the entity as shareable 25080 25081 This should prevent people from mistakenly trying to run 25082 zaphod mode 25083 25084commit 81114af1cdddc0b10d076f2e38c7a00c1223cc48 25085Author: Alex Deucher <alex@botch2.(none)> 25086Date: Sat Sep 22 15:51:23 2007 -0400 25087 25088 RADEON: preliminary support for mac mini 25089 25090 Option "MacModel" "mini" 25091 may not be 100% correct yet 25092 25093commit 6c482e453bc8156886294d0c1b8f3f1b3dcf4b36 25094Author: Alex Deucher <alex@botch2.(none)> 25095Date: Sat Sep 22 15:11:20 2007 -0400 25096 25097 RADEON: Fix RMX on LVDS 25098 25099 LVDS + RMX doesn't seem to like having the crtc values tweaked. 25100 25101commit 9109e62e3be7f96b41b534ab517fdf1baf458806 25102Author: Michel Dänzer <daenzer@debian.org> 25103Date: Sat Sep 22 14:19:10 2007 -0400 25104 25105 RADEON: fix crash when Xv window is outside of either crtc 25106 25107 Should fix bug 12175 25108 25109commit d2ce4a5003ce1291ea2327b2c00a0b24408fe26c 25110Author: Alex Deucher <alex@botch2.(none)> 25111Date: Sat Sep 22 10:41:55 2007 -0400 25112 25113 RADEON: RMX updates 25114 25115 - add option to turn off RMX 25116 - turn off RMX by default on DVI 25117 - add infrastructure to support more RMX modes 25118 25119commit f95b9ab729376083bf0d12987ee260ec2aba721e 25120Author: Alex Deucher <alex@botch2.(none)> 25121Date: Fri Sep 21 00:12:37 2007 -0400 25122 25123 RADEON: remove more old cruft 25124 25125commit 5a6f74103f0ec0d451d0e2573442efe5922848af 25126Author: Maciej Cencora <m.cencora@gmail.com> 25127Date: Thu Sep 20 23:56:08 2007 -0400 25128 25129 RADEON: fix video in on RV380 (tested on X600 VIVO) 25130 25131commit c72a365386e19f9257db041d44b09ad499cc9f6a 25132Author: Alex Deucher <alex@botch2.(none)> 25133Date: Thu Sep 20 23:49:57 2007 -0400 25134 25135 RADEON: fix up dvo support (still no external chip init) 25136 25137commit 5e4d98470b6412a686883c554e7eb7badbe78c4d 25138Author: Alex Deucher <alex@botch2.(none)> 25139Date: Thu Sep 20 23:22:48 2007 -0400 25140 25141 RADEON: round 3 on the PLLs. should fix the LVDS issues 25142 25143commit c5e2a2f09af807006c7ea493a8e90ff77abe207c 25144Author: Alex Deucher <alex@botch2.(none)> 25145Date: Wed Sep 19 19:58:28 2007 -0400 25146 25147 bump for RC release 25148 25149commit 46ff78b9f010ca24178d4363761be00eb3ecb632 25150Author: Alex Deucher <alex@botch2.(none)> 25151Date: Wed Sep 19 19:49:58 2007 -0400 25152 25153 RADEON: add default connector setup for single crtc chips 25154 25155commit 509ca0cb0e1cde905b47db2bbac6f2a58523b279 25156Author: Alex Deucher <alex@botch2.(none)> 25157Date: Wed Sep 19 19:41:17 2007 -0400 25158 25159 RADEON: fix external TMDS parsing in legacy bios connector table 25160 25161commit 5a59547ef7e986c9613023d941c3354cd476faf1 25162Author: Alex Deucher <alex@botch2.(none)> 25163Date: Wed Sep 19 19:28:09 2007 -0400 25164 25165 Revert "RADEON: initial pass at external TMDS support" 25166 25167 This reverts commit 4000a710c93dd2d82891e4082bc7fa922ba9c5f4. 25168 25169 This needs to be reworked and needs more soak time. so revert 25170 for now. 25171 25172commit dc333884c540d536bffe51a5ebfaf1822af6a91c 25173Author: Alex Deucher <alex@botch2.(none)> 25174Date: Tue Sep 18 22:23:12 2007 -0400 25175 25176 RADEON: fix plls again 25177 25178 previous fix seemed to break other chips. Lets see how 25179 this goes. 25180 25181commit 271e541088f455d1cfedff87e88c9a4fdbdbe424 25182Author: Alex Deucher <alex@botch2.(none)> 25183Date: Tue Sep 18 22:12:31 2007 -0400 25184 25185 RADEON: Remove more mergedfb cruft 25186 25187commit 8ae69c496eba701e744cca0605f73242673f7b3f 25188Author: Alex Deucher <alex@botch2.(none)> 25189Date: Tue Sep 18 20:12:21 2007 -0400 25190 25191 RADEON: adjust pll restore 25192 25193 some chips seem to be pickier than others. fixes bug 12467 25194 25195commit 8a84dc8ee260716026fa661a6dced2000621efa2 25196Author: Michel Dänzer <michel@tungstengraphics.com> 25197Date: Tue Sep 18 11:53:39 2007 +0200 25198 25199 radeon: Remove declaration of unused variable. 25200 25201commit 2fbbd9c5f0ec27e72780e8b054bc59502f1e5651 25202Author: Alex Deucher <alex@botch2.(none)> 25203Date: Tue Sep 18 00:45:56 2007 -0400 25204 25205 RADEON: note that MergedFB support has been replaced with randr 1.2 25206 25207commit 81b708dbd35c29bf2d67e8c655af97b60e61498b 25208Author: Alex Deucher <alex@botch2.(none)> 25209Date: Mon Sep 17 23:54:03 2007 -0400 25210 25211 RADEON: turn off load detection on tv dac by default 25212 25213 it's just too unreliable and even when it is most people 25214 don't want it. enable it via output attribute or force 25215 the outputs on with xrandr 25216 25217commit c3ab53159d42157985a25e9125eaa5362273b73d 25218Author: Alex Deucher <alex@botch2.(none)> 25219Date: Mon Sep 17 23:28:51 2007 -0400 25220 25221 RADEON: XPRESS connector tables are crap 25222 25223 besides specifying the pimary DAC (which they don't have), 25224 most seem to specify DDC_CRT2 for DDC, which does not seem 25225 to ever work. Some have reported success with DDC_MONID so 25226 lets give that a try. 25227 25228commit 1fafa970fa78ddea29a8804b6cd0aea52e4fa2e1 25229Author: Alex Deucher <alex@botch2.(none)> 25230Date: Mon Sep 17 23:10:13 2007 -0400 25231 25232 RADEON: ignore edid digital bit on analog ports 25233 25234 Some monitors have this bit set even on analog ports 25235 25236commit 4000a710c93dd2d82891e4082bc7fa922ba9c5f4 25237Author: Alex Deucher <alex@botch2.(none)> 25238Date: Mon Sep 17 23:04:17 2007 -0400 25239 25240 RADEON: initial pass at external TMDS support 25241 25242 - Based on Dave Airlie's initial work and the dvo support 25243 in the intel driver. 25244 - Only sil164 is supported at the moment. 25245 - Once we get some testing and such, we ought to move the dvo 25246 drivers out of the drivers so they can be shared among all drivers 25247 - Doesn't seem to work on my card (r4xx ATOM) card ATM 25248 - Legacy bios table programming sequence not implemented yet 25249 25250commit 38515d402555eaa61c686d42973e59f659b07466 25251Author: Alex Deucher <alex@botch2.(none)> 25252Date: Sat Sep 15 11:39:47 2007 -0400 25253 25254 RADEON: don't restore bios scratch regs 25255 25256 Leave them as programmed by the bios 25257 fixes bug 12424 25258 25259commit 78c6bd305f2ea70629d1dda3c0c48c067451f5fb 25260Author: George Sapountzis <gsap7@yahoo.gr> 25261Date: Mon Sep 10 19:00:16 2007 +0300 25262 25263 [mach64] Revert to multiple mappings per BAR. 25264 25265 libpciaccess added subrange mappings and mtrr handling. 25266 25267 compile-tested only. 25268 25269commit 0241cac643fa1c08a45ea44f5c670b290e760ad8 25270Author: Michel Dänzer <michel@tungstengraphics.com> 25271Date: Sat Sep 8 00:22:40 2007 +0200 25272 25273 radeon: Reinstate sync in radeon_crtc_show/hide_cursor. 25274 25275 Make sure the DRI lock is held though, as these can be called asynchronously. 25276 25277 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=12245 . 25278 25279commit 49933e2f7d590811f2bc8c0d51a09f3b7f14845e 25280Author: Alex Deucher <alex@botch2.(none)> 25281Date: Sat Sep 1 12:23:07 2007 -0400 25282 25283 RADEON: automatically disable tiling if requested virtual desktop exceeds surface limits 25284 25285 Also, tweak default desktop sizes and add some informational messages 25286 25287commit d49a60bc939c9f3e9d57c23e44263f5bb52f09fb 25288Author: Sjoerd Simons <sjoerd@luon.net> 25289Date: Thu Aug 30 11:40:13 2007 +0200 25290 25291 radeon: Fix Option "MacModel". 25292 25293commit 2b860726fff058623e0e8e1b0aca092c246875ce 25294Author: Michel Dänzer <michel@tungstengraphics.com> 25295Date: Thu Aug 30 11:38:28 2007 +0200 25296 25297 radeon: Don't build currently unused RADEONSavePalette. 25298 25299commit 12187a6aa93049c002a4171344d03c713f7f3c5d 25300Author: Alex Deucher <alex@botch2.(none)> 25301Date: Wed Aug 29 23:11:30 2007 -0400 25302 25303 RADEON: Add quirk and connector tables for apple laptops 25304 25305 As far as I can tell there are three apple laptop variants: 25306 ibook - LVDS, TVDAC drives TV or VGA via dongle 25307 powerbook-duallink - LVDS, TV, External TMDS/Primary DAC 25308 powerbook - LVDS, TV, Internal TMDS/Primary DAC 25309 use Option "MacModel" "<string>" 25310 to enable the appropriate quirks where string is one of the above 25311 25312 We can't yet init the external TMDS directly, but if OF inits it, 25313 it should work. This should also fix bug 9955. 25314 25315 Please test! 25316 25317commit 61c1fdaa8553581944f78a11e6f9aa76163a468a 25318Author: Alex Deucher <alex@botch2.(none)> 25319Date: Tue Aug 28 23:47:19 2007 -0400 25320 25321 RADEON: add option to force tmds pll to default table 25322 25323 Also rework the tmds pll output attribute handling a bit 25324 25325commit 0d9087bc60bb95c770b899cfed29699c02bdac49 25326Author: Alex Deucher <alex@botch2.(none)> 25327Date: Tue Aug 28 23:08:20 2007 -0400 25328 25329 RADEON: remove some cruft forgotten in a previous commit 25330 25331commit 6ff0645ecfe65727e8ef5d5e6215b4e03078e1a5 25332Author: George Sapountzis <gsap7@yahoo.gr> 25333Date: Tue Aug 28 15:37:11 2007 +0300 25334 25335 [mach64] Convert to pci-rework, keeping source-code compatibility. 25336 25337 It still uses the old probe method though, this is due to the ati wrapper. 25338 25339commit 7b38d9a1209f87255e5bb0aefe46a363ce4fb6ef 25340Merge: 2e3d43af1 673f79972 25341Author: George Sapountzis <gsap7@yahoo.gr> 25342Date: Tue Aug 28 16:01:12 2007 +0300 25343 25344 Merge branch 'master' into pci-rework 25345 25346 Conflicts: 25347 25348 src/ati.c 25349 25350commit 673f799729824f4439dd5f681f75dd5aab50947f 25351Author: Alex Deucher <alex@botch2.(none)> 25352Date: Tue Aug 28 00:42:30 2007 -0400 25353 25354 RADEON: Update tv attributes immediately 25355 25356commit ad6f7ad1b2ccae0bc0a416b9b0ca22709c9d5199 25357Author: Alex Deucher <alex@botch2.(none)> 25358Date: Tue Aug 28 00:08:41 2007 -0400 25359 25360 RADEON: remove the "default" tv_standard option 25361 25362commit 17e0f9e6cbfdb115034d327bd34d46339fd632b7 25363Author: Alex Deucher <alex@botch2.(none)> 25364Date: Mon Aug 27 23:59:03 2007 -0400 25365 25366 RADEON: enable load detection for tvdac if output count for tvdac < 2 25367 25368commit 42839fb5a8584196e7b18375bff6c426ed0347d9 25369Author: Alex Deucher <alex@botch2.(none)> 25370Date: Mon Aug 27 23:44:13 2007 -0400 25371 25372 RADEON: make load detection an output attribute for analog outputs 25373 25374 Since TV/VGA/DVI-I can share the TV DAC, we often get false detection 25375 of all inputs that share that DAC. Make load detection an output 25376 attribute. Enabled by default on primary dac and on cards where 25377 tv dac is (usually) dedicated to tv (non-IGP mobilities). 25378 25379commit a5a1055d64ab4fa16bfb03a412ae6c4fe69ff65d 25380Author: Alex Deucher <alex@botch2.(none)> 25381Date: Mon Aug 27 22:42:22 2007 -0400 25382 25383 RADEON: make tmds pll an output attribute 25384 25385 sometimes the bios tmds plls are busted for certain 25386 monitors. sometimes the dirver tables are. Let the user 25387 pick at run time. 25388 25389commit a12e4aa01bf1c5723c3c791ff9bdc26eef21d5ea 25390Author: Alex Deucher <alex@botch2.(none)> 25391Date: Sun Aug 26 18:51:29 2007 -0400 25392 25393 Bump for new release 25394 25395commit d43596e5f5d7c60f96b57bc3e743a9b40eb7109d 25396Author: Alex Deucher <alex@botch2.(none)> 25397Date: Sun Aug 26 18:07:50 2007 -0400 25398 25399 RADEON: Fix rotation. works now 25400 25401commit 47eb3327c258bb0cfd9a1d5677624b9988a39057 25402Author: Alex Deucher <alex@botch2.(none)> 25403Date: Sun Aug 26 15:43:22 2007 -0400 25404 25405 RADEON: minor tweak to tv out 25406 25407commit f36720377737210c985b196d9a988efdd767f1c7 25408Author: Alex Deucher <alex@samba.(none)> 25409Date: Sun Aug 26 14:13:06 2007 -0400 25410 25411 RADEON: fix depth 16 palette for real this time 25412 25413commit f2b13f1457bf860b075310d3962254be0ed7bea3 25414Author: Alex Deucher <alex@botch2.(none)> 25415Date: Sun Aug 26 13:27:19 2007 -0400 25416 25417 RADEON: Only update pixclks_cntl when updating tv routing 25418 25419 No need to re-set all of the pll2 stuff 25420 25421commit 5d044b9f74c7aa7e12f2822896fed881e2ca9d19 25422Author: Alex Deucher <alex@botch2.(none)> 25423Date: Sat Aug 25 21:03:08 2007 -0400 25424 25425 RADEON: fix crtc clipping for Xv 25426 25427commit 3fd2d22a02812d5f86cdc1c9503f48362b0c362b 25428Author: Alex Deucher <alex@botch2.(none)> 25429Date: Sat Aug 25 17:37:35 2007 -0400 25430 25431 RADEON: remove fbdev option 25432 25433 FBDev support is currently broken, and it not really compatible 25434 with randr 25435 25436commit 3469e1aa08792890fa6a5c72da52a1992a0b382c 25437Author: Alex Deucher <alex@botch2.(none)> 25438Date: Fri Aug 24 20:42:13 2007 -0400 25439 25440 RADEON: add extra green data in depth 16 25441 25442 Apparently some radeons need this? 25443 25444commit 71f650d1bc432514516f7ac64a5e8a54c5227881 25445Author: Michel Dänzer <michel@tungstengraphics.com> 25446Date: Fri Aug 24 09:21:39 2007 +0200 25447 25448 Require xorg-server >= 1.3 for RandR 1.2. 25449 25450commit d7230939f523610c57f92bdfc72966bdbc6f1070 25451Author: Michel Dänzer <michel@tungstengraphics.com> 25452Date: Fri Aug 24 09:21:14 2007 +0200 25453 25454 64 bit warning fixes. 25455 25456 For printf vs. CARD32, use %u or %x and and a cast to unsigned. 25457 25458commit 91c45fedfd155a153dcd2c3f3e30986bfbd44e6f 25459Author: Dave Airlie <airlied@nx6125b.(none)> 25460Date: Fri Aug 24 15:05:01 2007 +1000 25461 25462 radeon: don't disable dac if either tv or vga is using it 25463 25464 On my rs480 I had to vt switch to get hotplug VGA working due to the tv-out 25465 code turning off the dac when the vga code was actually using it. 25466 25467commit 056ca6bb5adf974290693b55de6cd6880d2132d1 25468Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25469Date: Thu Aug 23 20:19:47 2007 -0400 25470 25471 RADEON: NONE to None to match other port info 25472 25473commit 13fd53286bdda2c55683bdb5f63e7d345f6c63ef 25474Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25475Date: Thu Aug 23 20:17:51 2007 -0400 25476 25477 RADEON: set (hopefully) reasonable default max desktop sizes 25478 25479 Based on the amount of vram. We really need ttm... 25480 25481commit 53bad86ca48a9b6529c1f0989ee568d9d48841c6 25482Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25483Date: Thu Aug 23 19:24:59 2007 -0400 25484 25485 RADEON: clean up the logic in crtc_mode_set() 25486 25487commit 4712dedea225e9e07177aebda2ffc6290d1f53c7 25488Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25489Date: Thu Aug 23 18:51:34 2007 -0400 25490 25491 RADEON: Always assume LVDS is connected 25492 25493 Not all bioses seem to set the right scratch bits. If we have 25494 LVDS (via bios table or otherwise) assume it's connected. 25495 25496commit b0f170c5f736ecba1a5899d602c4173fe9b9b1fa 25497Author: Michel Dänzer <michel@tungstengraphics.com> 25498Date: Thu Aug 23 12:55:40 2007 +0200 25499 25500 radeon: Remove unnecessary #include <time.h>. 25501 25502commit ac54c0e4360099697755d14b1030def73d8235b0 25503Author: Michel Dänzer <michel@tungstengraphics.com> 25504Date: Wed Aug 22 14:33:59 2007 +0200 25505 25506 radeon: Warning fixes. 25507 25508commit de26e406f52b3b13f03eee2b8023924ec6406f0a 25509Author: Alon Ziv <alonz@nolaviz.org> 25510Date: Mon Jul 30 22:47:59 2007 +0300 25511 25512 radeon: Sane handling of timeouts in WaitForVerticalSync(2). 25513 25514 RADEONWaitForVerticalSync() and RADEONWaitForVerticalSync2() need to wait 25515 for a timeout specified in milliseconds; looping around usleep() causes 25516 the timeout to be unnecessarily long, as the OS may sleep longer than 25517 requested (on Linux the minimum actual sleep value may be several ms). 25518 25519 The new logic uses gettimeofday() in the loop to see when the (absolute) 25520 timeout has arrived. 25521 25522 Signed-off-by: Alon Ziv <alonz@nolaviz.org> 25523 25524commit c66e5de26ae93caa368213f3cce139aacec955d2 25525Author: Sascha Sommer <saschasommer@freenet.de> 25526Date: Thu Aug 23 12:11:51 2007 +0200 25527 25528 radeon: Round down RMX stretch ratios. 25529 25530 Fixes issues with RMX scaling, see 25531 https://bugs.freedesktop.org/show_bug.cgi?id=8983 . 25532 25533commit 633c1fff10a3be4c9f48c1995e330d60bf6abbb2 25534Author: Michel Dänzer <michel@tungstengraphics.com> 25535Date: Thu Aug 23 12:11:41 2007 +0200 25536 25537 radeon: Sync pages when enabling page flipping with EXA as well. 25538 25539 Exclude the DRI window(s) though to avoid scribbling over 3D rendering. 25540 25541commit 4f8010ce22043c0f8d60c0f49d270ce98c9d2466 25542Author: Michel Dänzer <michel@tungstengraphics.com> 25543Date: Thu Aug 23 12:11:12 2007 +0200 25544 25545 radeon: Don't synchronize DRI windows between pages when possible. 25546 25547commit 7b527054a7c81d1d1dbc79d41b9e53064dab68cb 25548Author: Michel Dänzer <michel@tungstengraphics.com> 25549Date: Thu Aug 23 12:10:33 2007 +0200 25550 25551 radeon: Restore memmap registers even if only AGP location changed. 25552 25553commit d7ba9f001c0ab645984526afd0e64d1c6a6d654a 25554Author: Michel Dänzer <michel@tungstengraphics.com> 25555Date: Thu Aug 23 11:39:54 2007 +0200 25556 25557 radeon: Change a test to info->IsIGP from several IGP families. 25558 25559commit 8c7c22e22b6076abc80e4e1aaa8d1f4cf2f3ed14 25560Author: Michel Dänzer <michel@tungstengraphics.com> 25561Date: Thu Aug 23 11:38:17 2007 +0200 25562 25563 radeon: Wait for pending overlay flip to finish before emitting new one. 25564 25565commit 92fa7cc00688d7bfc1fb72e645ac30c6d92669c6 25566Author: Michel Dänzer <michel@tungstengraphics.com> 25567Date: Thu Aug 23 11:38:16 2007 +0200 25568 25569 radeon: Don't call RADEONDRIRefreshArea when the damaged region is empty. 25570 25571commit 5cb20c2dc5eca9d7d7d78e9924ea1b90076e7253 25572Author: Dave Airlie <airlied@linux.ie> 25573Date: Thu Aug 23 18:46:41 2007 +1000 25574 25575 radeon: bug 11899 + debian 435040 - test for usefbdev before calling int10 25576 (cherry picked from commit 3334c247540b113f9c4a5ce1879d8648a45a959a) 25577 25578commit 9d6261c024c697a211da4b6300420153eb72a264 25579Author: Roland Bär <roland@verifysoft.de> 25580Date: Thu Aug 23 18:33:34 2007 +1000 25581 25582 radeon: bug 11860 fix some mem leaks 25583 25584commit 7bc1f862bc5f992f213143fbafef52459ba7db4a 25585Author: Roland Bär <roland@verifysoft.de> 25586Date: Thu Aug 23 18:37:35 2007 +1000 25587 25588 radeon: bug 11861 - dead code removal in radeon_video.c 25589 (cherry picked from commit 8e3a6f83016cd8c4cfd43ceee4cbf0a8dc018b2a) 25590 25591commit 9d38c8aa1a7d6fb1af41ee8abdb4a95f94843538 25592Author: Dave Airlie <airlied@clockmaker.usersys.redhat.com> 25593Date: Thu Aug 23 20:10:24 2007 +1000 25594 25595 radeon: cleanup some warnings 25596 25597commit e4c8969b48a6c8dcc4e7f9852479d24a0204fc0d 25598Author: Dave Airlie <airlied@redhat.com> 25599Date: Thu Aug 23 19:56:21 2007 +1000 25600 25601 updated release numbering for randr 1.2 25602 25603commit 81a8093f2ffdce59d4e8a44f65f3d5c771fd6425 25604Merge: c08e6ec9e 53a67e319 25605Author: Dave Airlie <airlied@redhat.com> 25606Date: Thu Aug 23 19:51:47 2007 +1000 25607 25608 Merge remote branch 'origin/randr-1.2' 25609 25610commit c08e6ec9e7ac06caee53689b0ec50ef6a7a0ba37 25611Merge: 5793e8753 a0d0fcd3b 25612Author: Dave Airlie <airlied@clockmaker.usersys.redhat.com> 25613Date: Thu Aug 23 19:42:19 2007 +1000 25614 25615 Merge branch 'master' into randr-merge 25616 25617 Conflicts: 25618 25619 src/radeon.h 25620 src/radeon_bios.c 25621 src/radeon_display.c 25622 src/radeon_dri.c 25623 src/radeon_driver.c 25624 src/radeon_modes.c 25625 src/radeon_probe.h 25626 src/radeon_video.c 25627 25628commit 53a67e31904bec9a3aa1bd24de8034dcafea1d2a 25629Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25630Date: Tue Aug 21 21:20:41 2007 -0400 25631 25632 RADEON: Fix color problem on pre-R3xx chips tv-out 25633 25634commit e9719e8e02eef46717ae9b4d8c7998466dac30cb 25635Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25636Date: Tue Aug 21 21:17:20 2007 -0400 25637 25638 RADEON: more tv out fixes and clean up 25639 25640commit 36c22a49580d86a6518b67f31a78bd53d39491af 25641Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25642Date: Tue Aug 21 20:28:39 2007 -0400 25643 25644 RADEON: fix tv-out on R3xx 25645 25646 R3xx apparently needs the tv clock forced on. 25647 25648commit 9470bd67731059f26859ed5f0bea3ade09e2c80c 25649Author: Alex Deucher <alex@botch2.com> 25650Date: Mon Aug 20 20:54:06 2007 -0400 25651 25652 RADEON: Add DefaultConnectorTable option 25653 25654 This option skips the parsing the BIOS connector table 25655 and falls back to chip specific defaults. 25656 Also remove man page section for the now gone bioshotkeys 25657 option. 25658 25659commit a0d0fcd3bab765b4db25e04884fd8a342abb9c66 25660Author: iLisa Wu <liswu@ati.com> 25661Date: Mon Aug 20 15:44:07 2007 +0200 25662 25663 Fix crash with no valid mode in xorg.conf's modelist and empty Virtual 25664 25665 If the resolution defined in xorg.conf failed to find a matching mode in the 25666 supported modelist, and no virtual desktop dimensions are defined the xorg.conf 25667 either, virtual X and Y dimension will be set to 0 which will cause Xserver 25668 crash. 25669 25670 (Novell bugzilla #296856, closed) 25671 25672commit b275febdb0918e8cebdffbb433b0eeb3ff4d3746 25673Author: Alex Deucher <alex@samba.(none)> 25674Date: Sun Aug 19 20:55:32 2007 -0400 25675 25676 RADEON: turn off TVCLK when blanking tv encoder 25677 25678commit a90d675832ddb02c81ace010ccbf02619b70edac 25679Author: Alex Deucher <alex@botch2.com> 25680Date: Thu Aug 16 21:55:14 2007 -0400 25681 25682 RADEON: fix Xv clipping and overlay sourcing 25683 25684 - Basically just copied from the intel driver. I'm planning to push 25685 this to the server soon, but add it now to get things working 25686 and to provide compat for older servers. 25687 25688 - Overlay crtc source control attribute now called XV_CRTC 25689 The old attribute XV_SWITCHCRT has been removed. If anyone cares, 25690 we can add it back as an alias to XV_CRTC 25691 XV_CRTC: -1 auto, 0 crtc0, 1 crtc1 25692 25693commit 5793e8753d11432bf95c7c6dd80c811e16aba058 25694Author: Alex Deucher <alex@botch2.com> 25695Date: Wed Aug 15 19:26:36 2007 -0400 25696 25697 RADEON: Remove RADEONRestoreMode() 25698 25699 Since we no longer use it to write modes all it did 25700 was restore some of the regs for a console restore. 25701 Just move the relevant bits into RADEONRestore() and 25702 remove it. 25703 25704commit 6f011aaabaf18d66ffc255ad76aaf938b2396302 25705Author: Lisa Wu <Lisa.Wu@amd.com> 25706Date: Wed Aug 15 19:17:51 2007 -0400 25707 25708 RADEON: fix console restore on r3xx and r4xx 25709 25710 When restoring the console we need to: 25711 1) we need to restore DAC registers after all other registers are 25712 restored and CRTCs are enabled. 25713 2) we need to enable CRTC2 registers before CRTC1 registers 25714 25715commit c8dad98abb042c6abbbee18f9ae4db72084bc513 25716Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25717Date: Wed Aug 15 00:55:24 2007 -0400 25718 25719 RADEON: make sure crtc routing is correct in r300_detect_tv() 25720 25721 Also remove some unused variable 25722 25723commit f8cd74435f0072dbf5f6e83d67d2d5e1f4e82c91 25724Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25725Date: Tue Aug 14 23:50:57 2007 -0400 25726 25727 RADEON: minor cleanup 25728 25729commit 6f398cd07ea734dd66a8eac71b629e59123d75b8 25730Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25731Date: Tue Aug 14 23:42:32 2007 -0400 25732 25733 RADEON: Implement improved tv load detection for r300 25734 25735 The previous implementation resulted in false positives 25736 on occasion. This method works much more reliably. 25737 Based on beos code by Thomas Kurschel 25738 25739commit 366a1d4c240ac93622caff97b652696db99bf2e6 25740Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25741Date: Tue Aug 14 23:25:37 2007 -0400 25742 25743 RADEON: Turn off tv encoder in disableoutputs() 25744 25745commit 5b4a04c23e3f1ec2490418b111f417c16463c709 25746Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25747Date: Tue Aug 14 23:11:52 2007 -0400 25748 25749 RADEON: Make sure RMX is always programmed for crtc1 25750 25751 If the RMX registers are left as programmed by the bios 25752 this can lead to a blank screen when crtc1 is feeding a 25753 DAC. Fix found and reported by Lisa Wu <liswu@ati.com> 25754 Fixes bug 11985 25755 25756commit db2a828b2f21b92cd654b309d137204334975b89 25757Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 25758Date: Tue Aug 14 22:47:29 2007 -0400 25759 25760 RADEON: Add load detection 25761 25762 Based on the beos driver by Thomas Kurschel and the 25763 existing load detection code in this driver. 25764 25765commit 6fc3ddbbb6fbbee1f6076c776e5b46c0c772b6d4 25766Author: Alex Deucher <alex@botch2.com> 25767Date: Sun Aug 12 12:00:20 2007 -0400 25768 25769 RADEON: order the VGA and DVI ports correctly 25770 25771commit b7738d2ef82e1759adf78e5db1291f8739b4166f 25772Author: Alex Deucher <alex@botch2.com> 25773Date: Sat Aug 11 17:50:42 2007 -0400 25774 25775 RADEON: fix connector setup when there's no bios tables 25776 25777commit d0895f67e327bb268fd59fcfd8fc22678d804f57 25778Author: Luc Verhaegen <libv@skynet.be> 25779Date: Thu Aug 9 12:11:40 2007 +0200 25780 25781 AGPFastWrite risk reduction. 25782 25783 Actively warn the user in the log about the effects of AGPFastWrite and 25784 sanitise AGPFastWrite handling while we're here. 25785 25786commit c01000bc684f6a23a38a52f0808182cefaa686bb 25787Author: Alex Deucher <alex@botch2.com> 25788Date: Wed Aug 8 00:07:16 2007 -0400 25789 25790 RADEON: fix typo in comment 25791 25792commit 5c3f49e651c36f3bd14fa29359e24825d8f7f77f 25793Author: Lisa Wu <liswu@ati.com> 25794Date: Wed Aug 8 00:05:47 2007 -0400 25795 25796 RADEON: make sure RADEON_CRTC2_CRT2_ON bit state is properly accounted for 25797 25798 - when the TV DAC is used for crtc1 make sure to keep this bit set when 25799 initializing crtc2. 25800 - fixes bug 11894 25801 25802commit 08fe7ad00fba523775e95b5e6295fe23a5119d60 25803Author: Alex Deucher <alex@botch2.com> 25804Date: Tue Aug 7 23:16:05 2007 -0400 25805 25806 RADEON: fix ConnectorTable option after connector table re-work 25807 25808 We may want to add the possiblity of more connectors to this option 25809 25810commit f56b90a60393ed187c0e39e149b2a3a32331ea52 25811Author: Alex Deucher <alex@botch2.com> 25812Date: Tue Aug 7 23:08:02 2007 -0400 25813 25814 RADEON: Make sure the default TV standard is supported 25815 25816commit a88a0e77eec514325b4d07bf7be9fb04e6f8e244 25817Author: Alex Deucher <alex@botch2.com> 25818Date: Tue Aug 7 00:43:20 2007 -0400 25819 25820 RADEON: minor fix to legacy bios connector table 25821 25822commit 3752808d6c08a9727370ef8d79088e787791e131 25823Author: Alex Deucher <alex@botch2.com> 25824Date: Sun Aug 5 16:13:39 2007 -0400 25825 25826 RADEON: Always set MonType in radeon_detect() 25827 25828 - Always set the the MonType in radeon detect even if no 25829 monitor is detected so users can still force disconnected 25830 outputs on. For DVI-I users will have to set the 25831 dvi_monitor_type attribute to force digital or analog, 25832 everything else should just work. 25833 25834commit b4ec3e436afb5bc99ec755cbd96eee9a1ec492a5 25835Author: Alex Deucher <alex@botch2.com> 25836Date: Sun Aug 5 15:37:04 2007 -0400 25837 25838 RADEON: add tv out properties 25839 25840commit 1fc2a1120e7c05938e2bd72d3c7837ecff8bc9da 25841Author: Alex Deucher <alex@botch2.com> 25842Date: Sun Aug 5 14:41:57 2007 -0400 25843 25844 RADEON: switch DVI-I monitor type attribute to string 25845 25846commit 7d8eb3751d74bd8a1fb9fa2d2fcb9c4c895f6ba4 25847Author: Alex Deucher <alex@botch2.com> 25848Date: Sun Aug 5 14:31:18 2007 -0400 25849 25850 RADEON: make backlight attribute lower case 25851 25852commit e71bb88bc9052af8866fb6945dbc06dbb0e6d1c0 25853Author: Alex Deucher <alex@botch2.com> 25854Date: Sun Aug 5 14:26:15 2007 -0400 25855 25856 RADEON: switch RMX attribute to string type 25857 25858 Still not actually hooked up. 25859 25860commit 284323135ec04635dfa9cabd5790a35b953abca2 25861Author: Alex Deucher <alex@botch2.com> 25862Date: Sun Aug 5 04:35:24 2007 -0400 25863 25864 RADEON: typo 25865 25866commit d1abdad167aa24ac970c69422435df443c82ebd6 25867Author: Alex Deucher <alex@botch2.com> 25868Date: Sun Aug 5 03:45:02 2007 -0400 25869 25870 RADEON: fixes 25871 25872 - fix output ordering 25873 - set tv-out to return un-connected for now in radeon_detect() 25874 25875commit 0cca25d8d6a0cb0d29b68e6cd9c699d1390aede1 25876Author: Alex Deucher <alex@botch2.com> 25877Date: Sun Aug 5 03:19:24 2007 -0400 25878 25879 RADEON: Fix DVI-I support in ATOM bios connector table parsing 25880 25881commit ba5496ae7973786802962bf649dd91c219531749 25882Author: Alex Deucher <alex@botch2.com> 25883Date: Sun Aug 5 02:27:32 2007 -0400 25884 25885 RADEON: refactor output init to handle multiple DVI or VGA 25886 25887 - refactor output init to handle multiple DVI or VGA with the new 25888 bios table parsing 25889 25890commit 2ec22783ddf4c522df9e5fd1b2003854486d7a2b 25891Author: Alex Deucher <alex@botch2.com> 25892Date: Sun Aug 5 01:39:35 2007 -0400 25893 25894 RADEON: attempt to do the right thing for standards other than PAL or NTSC 25895 25896commit cc8e1d95f1b90a259beea4e8cc4d7e29af660919 25897Author: Alex Deucher <alex@botch2.com> 25898Date: Sun Aug 5 01:14:36 2007 -0400 25899 25900 RADEON: Major rework of BIOS table parsing 25901 25902 - greatly simplify ATOM and legacy connector table parsing 25903 - use bios tables to detect LVDS and TV outputs 25904 - add support for TV table parsing (legacy only) 25905 25906commit 1de52d91ff3a04b9b587b858e1e5be40d3a7fd0a 25907Author: Dave Airlie <airlied@linux.ie> 25908Date: Sat Aug 4 17:58:58 2007 +1000 25909 25910 update configure.ac for 6.6.193 release 25911 25912commit 371001c0433db1d17e468f3ea99ea57f922145a3 25913Author: Dave Airlie <airlied@linux.ie> 25914Date: Sat Aug 4 17:51:19 2007 +1000 25915 25916 radeon: remove unused variables 25917 25918commit 90946c53d7a4b23d03270ad6da0450759a11de3d 25919Author: Dave Airlie <airlied@linux.ie> 25920Date: Sat Aug 4 17:51:06 2007 +1000 25921 25922 ati: add -Wall for gcc 25923 25924 As per luc's "suggestion" in Novell bug. 25925 25926 This may generate warnings on 64-bit until we get rid of CARD32 25927 25928commit e30a145934df8f6a7f71290d6c75e4239f9d52f7 25929Author: Dave Airlie <airlied@linux.ie> 25930Date: Sat Aug 4 17:44:46 2007 +1000 25931 25932 radeon: cleanup some pieces of the dpms/blank register programming 25933 25934commit a156db5e8b037ed12a448f70045453baf9d0c504 25935Author: Luc Verhaegen <libv@skynet.be> 25936Date: Sat Aug 4 17:37:18 2007 +1000 25937 25938 Clean up PortInfo to CRTC mapping 25939 25940 Also sanitise blanking and DPMS functions 25941 25942 Fixes from Novell Bug 264720, and fd.o 10772 25943 25944commit b61a49f2a5401560f85e11bcdd005287433cad12 25945Author: Alex Deucher <alex@botch2.com> 25946Date: Fri Aug 3 19:27:59 2007 -0400 25947 25948 RADEON: Change indexing of TV constants table in preparation for standard re-work 25949 25950commit 288fa627274cb399059262d4f8bd844fc220a042 25951Author: Alex Deucher <alex@botch2.com> 25952Date: Thu Aug 2 02:37:16 2007 -0400 25953 25954 RADEON: avoid a divide by 0 and only save tv out regs if the chip has them 25955 25956commit b66a1bc7994b33d349c1519761e431959311c85f 25957Author: Alex Deucher <alex@botch2.com> 25958Date: Thu Aug 2 02:11:20 2007 -0400 25959 25960 RADEON: fix tv-out enable/disable 25961 25962commit d86592c8d5ce45d81d8a726c263e870e94fbcf11 25963Author: Alex Deucher <alex@botch2.com> 25964Date: Thu Aug 2 00:50:51 2007 -0400 25965 25966 RADEON: add missing break 25967 25968commit 971feb34843225030fff05b3f9d3801534fbf2d4 25969Author: Alex Deucher <alex@botch2.com> 25970Date: Thu Aug 2 00:50:04 2007 -0400 25971 25972 RADEON: move tv dac enable to enabledisplay() 25973 25974commit 98d7e00437bea78e03180eb30ff30de3455d9d1a 25975Author: Alex Deucher <alex@botch2.com> 25976Date: Thu Aug 2 00:20:50 2007 -0400 25977 25978 RADEON: limit tv modes to the only one we can program at the moment 25979 25980commit 5c549c1d42f7bbc556942af13aff2661fae856f2 25981Author: Alex Deucher <alex@botch2.com> 25982Date: Wed Aug 1 23:45:07 2007 -0400 25983 25984 RADEON: remove unused elements 25985 25986commit b03978028fd975eb6946503d3a56a49c5a67f339 25987Author: Alex Deucher <alex@botch2.com> 25988Date: Wed Aug 1 22:39:16 2007 -0400 25989 25990 RADEON: convert hard coded tv out values to calculations 25991 25992commit cf54222f1fa37366b2c2b39c82f8afc02f32e63c 25993Author: Alex Deucher <alex@botch2.com> 25994Date: Tue Jul 31 02:01:49 2007 -0400 25995 25996 RADEON: more fixes... 25997 25998commit 22d460d3ad991223aa1fbd7e5edeb45e36c65dc0 25999Author: Alex Deucher <alex@botch2.com> 26000Date: Tue Jul 31 01:28:05 2007 -0400 26001 26002 RADEON: fix name of tv output 26003 26004commit 4822a2b837334f408f962646ab5ea4f8b0335ac9 26005Author: Alex Deucher <alex@botch2.com> 26006Date: Tue Jul 31 01:18:40 2007 -0400 26007 26008 RADEON: tv-out fixes. works now. tested on rv350. 26009 26010 VT siwtch is busted, and xrandr doesn't play nice yet. 26011 uncomment code in radeon_output.c to test. 26012 26013commit fe494c9db2995bb8ce7a028ecf9626e0cb0cf506 26014Author: Alex Deucher <alex@botch2.com> 26015Date: Sun Jul 29 15:26:34 2007 -0400 26016 26017 RADEON: add info about tv out code and authorship 26018 26019commit 6b9b7a7bdc290d07de9b226691ec8025af8db896 26020Author: Alex Deucher <alex@botch2.com> 26021Date: Sun Jul 29 15:23:14 2007 -0400 26022 26023 RADEON: Initial pass at integrated tv out support 26024 26025 Based on the GATOS tv-out support by Federico Ulivi <fulivi@lycos.com> 26026 and information from ati with substantial rework by myself. 26027 26028 Code is not actually hooked up yet. 26029 26030commit 8d043db1817d94edeb72ab208dfea60026715d48 26031Author: Alex Deucher <alex@botch2.com> 26032Date: Wed Jul 25 20:37:58 2007 -0400 26033 26034 RADEON: Compute PLL VCO gain 26035 26036commit 62f06d89da3f7160d5e4df8d7ce6fe1a94e9d07c 26037Author: Alex Deucher <alex@botch2.com> 26038Date: Wed Jul 25 20:22:25 2007 -0400 26039 26040 RADEON: write out saved vclk and pixclk values 26041 26042commit 165a07cbbfcd94e3d1fac434b8fada8d29428a09 26043Author: Brice Goglin <Brice.Goglin@ens-lyon.org> 26044Date: Wed Jul 25 19:01:02 2007 +0200 26045 26046 Minor fixes in the manpages 26047 26048 - ati.man 26049 . add references to radeon and r128 manpages in SEE ALSO 26050 (Xavier Bestel in Debian bug #386001). 26051 . remove the manpage suffix from atimisc reference in the text 26052 since there is no such manpage for now. 26053 - radeon.man 26054 . typo in "specifying" (reported by A. Costa in Debian bug #432059). 26055 . replace a non-ascii character with the corresponding groff escape 26056 sequence so that it works whatever the locale (Julien Cristau). 26057 26058commit 9cc3ab8320162f371bba15dc131f23c5de2013fc 26059Author: Alex Deucher <alex@botch2.com> 26060Date: Wed Jul 25 02:24:14 2007 -0400 26061 26062 RADEON: hack around console restore hang on r4xx 26063 26064 r4xx seems to hang when unblanking the crtc(s) when 26065 restoring the console. This needs to be investigated 26066 further. 26067 26068commit 193d58b0ea9de3e326017e50b88e2f3726854506 26069Author: Alex Deucher <alex@botch2.com> 26070Date: Wed Jul 25 01:27:58 2007 -0400 26071 26072 RADEON: fix randr on r4xx 26073 26074 Seems the way we parse the ATOM connector table results 26075 in reversed connectors. 26076 26077commit 262e32bd9ea4123b116362b33b5798753b2568fe 26078Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26079Date: Thu Jul 19 23:48:57 2007 -0400 26080 26081 RADEON: fix palette when depth != 24 26082 26083commit 18b00b47a483e7854727e99126808ddf361e7a4a 26084Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26085Date: Thu Jul 19 19:01:42 2007 -0400 26086 26087 R128: don't clip modes to panel on laptops when only using the CRT port 26088 26089 fixes bug 5832 26090 26091commit 882fe7631586b0a7919f808588a2ea4fb555f7e8 26092Author: Brice Goglin <brice.goglin@ens-lyon.org> 26093Date: Thu Jul 19 18:44:27 2007 -0400 26094 26095 MACH64: avoid crash in Xv code 26096 26097 fixes bug 11054 26098 26099commit 7b3e22e96cfc98621bd20fa76317e6d8f7242165 26100Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26101Date: Thu Jul 5 20:37:10 2007 -0400 26102 26103 RADEON: set default values for crtc_offset_cntls 26104 26105commit 2a35ccfb8ffee9f0ef6f8d4f0eeb80a471543caa 26106Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26107Date: Thu Jul 5 20:04:04 2007 -0400 26108 26109 RADEON: Fix tiling on r1xx and r2xx 26110 26111 spotted by Andrew Randrianasulu <randrik@mail.ru> 26112 fixes bug 11357 26113 26114commit ac9cbaf32176cf144bc694ac879e7c9e2920f762 26115Author: Michel Dänzer <michel@tungstengraphics.com> 26116Date: Mon Jul 2 10:06:02 2007 +0200 26117 26118 radeon: Fully zero-initialize info->CRT2pScrn->monitor. 26119 26120 This prevents modes from getting rejected due to the uninitialized maxPixClock 26121 value. 26122 26123 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=9816 . 26124 26125commit bdcae622100c81a4d9a53938542b64908bacd195 26126Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26127Date: Fri Jun 29 01:15:57 2007 -0400 26128 26129 RADEON: save crtc on/off state for console restore 26130 26131 - prevents a possible hang if console is only using 1 crtc 26132 26133commit 21be0d3a1f4700572c6425800596785d6850626f 26134Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26135Date: Fri Jun 29 00:38:50 2007 -0400 26136 26137 RADEON: simplify console restore 26138 26139 - still need to track crtcs for blank/unblank when restoring text console 26140 26141commit d9bf28b55d98fea2d285c9c46362aaf2175d0a46 26142Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26143Date: Thu Jun 28 23:52:28 2007 -0400 26144 26145 RADEON: factor out surface_cntl init into one function 26146 26147commit 9f193985627be8e6ea1418a424e825ddbc4957b2 26148Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26149Date: Thu Jun 28 23:43:13 2007 -0400 26150 26151 RADEON: move crtc base setups to new functions 26152 26153commit 0f361e9e80a29d287fa42436c32c657e3c102539 26154Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26155Date: Thu Jun 28 23:08:07 2007 -0400 26156 26157 RADEON: fix corruption after 3D apps run 26158 26159commit 1d4630067d293d26284d5fe230debef3913ec6cf 26160Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26161Date: Thu Jun 28 22:24:04 2007 -0400 26162 26163 RADEON: set info->CurrentLayout.displayWidth in RADEONSetPitch() 26164 26165commit 2754d1ba01fc3367019487e0c0f59d74c950aaa7 26166Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26167Date: Thu Jun 28 22:15:22 2007 -0400 26168 26169 RADEON: Switch cursor back to memcpy() 26170 26171commit 7901bcafa92dccd319ddb5de4627d806a39f15f9 26172Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26173Date: Mon Jun 25 18:39:14 2007 -0400 26174 26175 RADEON: make sure we unblank in RADEONRestore() 26176 26177 reported by Andrew Randrianasulu (bugs 11357, 11146) 26178 26179commit 6a3599d6155d073a3116c4b13bdf04b44bb9e087 26180Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26181Date: Mon Jun 25 18:20:54 2007 -0400 26182 26183 RADEON: make sure we set the stride properly 26184 26185commit 771c37fcac8104b894120cc3fc0154b31b0318d1 26186Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26187Date: Mon Jun 25 17:53:26 2007 -0400 26188 26189 RADEON: attempt to fix cursor on big endian (only tested on x86) 26190 26191commit 5b5b90c2cea7e36895354f5872acd3fc769653f9 26192Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26193Date: Mon Jun 25 16:27:54 2007 -0400 26194 26195 RADEON: fix possible segfault on mobility chips with MM tables 26196 26197 spotted by Stefan Buehler 26198 26199commit 01bd5eb713500d3e3d4351865d460c8d1e476454 26200Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26201Date: Mon Jun 25 16:16:18 2007 -0400 26202 26203 RADEON: fix possible segfault on mobility chips with MM tables 26204 26205 reported by Stefan Buehler 26206 26207commit 52cc1dc1491559a9055f3ba6dd54064bb382ad86 26208Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26209Date: Mon Jun 25 15:49:19 2007 -0400 26210 26211 RADEON: tiling and FB size fixes 26212 26213 - fix tiling on older radeons 26214 - allow the user to specify FB size using Virtual line in config 26215 Current default is 1600x1200 26216 - reset ecp_div for Xv in randr crtc mode set 26217 26218commit 3bb46c03fddd3bd79bdab887366aeeced0290a3a 26219Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26220Date: Fri Jun 22 01:12:38 2007 -0400 26221 26222 RADEON: bump crtc max to 1200 so 1600x1200 will work out of the box 26223 26224 - a better fix will be forthcoming... 26225 26226commit b9fac5695bf8f123b391e3b162c4865258a685dd 26227Author: Andrew Randrianasulu <randrik@mail.ru> 26228Date: Fri Jun 22 01:01:14 2007 -0400 26229 26230 RADEON: Fix VT switch hangs 26231 26232 - re-order DRI resume functions in EnterVT() 26233 - fixes bug 11287 26234 26235commit 92e65d5e0d6817ff4c9a08020a0a9b3a8c3c98b0 26236Author: Andrew Randrianasulu <randrik@mail.ru> 26237Date: Fri Jun 22 00:55:00 2007 -0400 26238 26239 RADEON: Fix VT switch hangs 26240 26241 - reorder RADEONDRISetVBlankInterrupt() and RADEONDRIResume() 26242 - see bug 11287 26243 26244commit ec1fd2c388ec2c4033644776ee8588405a3d25f1 26245Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26246Date: Tue Jun 19 00:39:59 2007 -0400 26247 26248 RADEON: Split FP reg restore to only touch regs relevant to the output 26249 26250 - only restore output specific regs 26251 - restore fp2_gen_cntl on R200 DAC restore as R200 DAC2 is connected via 26252 DVO 26253 26254commit 0f5a92667281ca3f80218bc67d4031125a5e4793 26255Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26256Date: Mon Jun 18 23:35:21 2007 -0400 26257 26258 RADEON: only touch LVDS on mobility chips 26259 26260commit b72ff160f908bf3aa9f64705377e92d80360a4f7 26261Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26262Date: Mon Jun 18 21:51:45 2007 -0400 26263 26264 RADEON: only touch LVDS on mobility chips 26265 26266commit 80313621ffa8c409ae63fc1b28c15fc4abdc3a7b 26267Author: Henry Zhao <henryz@localhost.localdomain> 26268Date: Fri Jun 15 17:00:05 2007 -0700 26269 26270 Update CRT2pScrn->monitor->Last to reflect the 26271 last mode of CRT2pScrn. See bug 11278. 26272 26273commit 5868af53204e8fc210735e68e891e15085285629 26274Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26275Date: Tue Jun 12 23:53:48 2007 -0400 26276 26277 RADEON: yet another LVDS... 26278 26279commit f19a6f7ee5bf4ec632e7813359f167599c08e823 26280Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26281Date: Tue Jun 12 23:44:41 2007 -0400 26282 26283 RADEON: additional LVDS off fix (missed in previous commit) 26284 26285commit 4f2a1ba9c2fc20042d67132f986a86a9783a245e 26286Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26287Date: Tue Jun 12 23:41:24 2007 -0400 26288 26289 RADEON: additional LVDS off fix (missed in previous commit) 26290 26291commit 9fc416ce08cc5d74faa7b184e80ce43ed2f4fbee 26292Author: Michel Dänzer <michel@tungstengraphics.com> 26293Date: Tue Jun 12 23:35:09 2007 -0400 26294 26295 radeon: Fix panel size detection from registers with stretched mode programmed. 26296 26297commit c4045cbc5e6975520efef422ce9df2a7d99a7ff4 26298Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26299Date: Tue Jun 12 23:32:18 2007 -0400 26300 26301 RADEON: make sure to clear RADEON_LVDS_ON when turning LVDS off 26302 26303 - fixes bug 3483 26304 26305commit 090888d3baf61a70640d28e99afdce20952cdea0 26306Author: Tormod Volden <bugzi06.fdo.tormod@xoxy.net> 26307Date: Tue Jun 12 23:30:18 2007 -0400 26308 26309 RADEON: limit PanelPwrDly to 2000 ms on ATOM bios as per legacy bios 26310 26311 - fixes bug 11238 26312 26313commit 44748a826b52924a6e050215d8d62755e0ac7fb2 26314Author: Tormod Volden <bugzi06.fdo.tormod@xoxy.net> 26315Date: Tue Jun 12 23:23:49 2007 -0400 26316 26317 RADEON: limit PanelPwrDly to 2000 ms on ATOM bios as per legacy bios 26318 26319 - fixes bug 11238 26320 26321commit c292fc64499ff4cc135c07deda99cf4169f8fef4 26322Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26323Date: Tue Jun 12 23:20:18 2007 -0400 26324 26325 RADEON: make sure RADEON_LVDS_ON is cleared when turning LVDS off 26326 26327 - fixes bug 3483 26328 26329commit c6a3286d6f6ddda89115d98d45665dadd78bf41d 26330Author: Michel Dänzer <michel@tungstengraphics.com> 26331Date: Mon Jun 11 09:39:38 2007 +0200 26332 26333 radeon: Fix panel size detection from registers with stretched mode programmed. 26334 26335commit 7a6b3c0f77c69019268f585c51c3dcbcc99014d4 26336Author: Michel Dänzer <michel@tungstengraphics.com> 26337Date: Mon Jun 11 09:39:38 2007 +0200 26338 26339 radeon: Don't loop indefinitely if no mode matches detected panel size. 26340 26341commit f31fd9ce598841c505a0b5ed32bf124f49fea332 26342Author: Michel Dänzer <michel@tungstengraphics.com> 26343Date: Mon Jun 11 09:39:38 2007 +0200 26344 26345 radeon: Fix some more ErrorFs when setting mode. 26346 26347commit 9dd9f09a9021653e5590d27bd7aa0e7af85416ad 26348Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26349Date: Thu Jun 7 15:32:31 2007 -0400 26350 26351 RADEON: fix some issues with bios table init code 26352 26353 - fix some issues with bios table init code 26354 - re-org preinit to better handle bios table init of cards 26355 26356commit 9ad311eeb0ac2b70a862dbb9de278154ba9142ec 26357Author: Tilman Sauerbeck <tilman@code-monkey.de> 26358Date: Thu Jun 7 11:03:02 2007 +0200 26359 26360 Fixed 'make dist'. 26361 26362commit 2527f2b69aa7dffa3ba4359c45955e3185bdf0e6 26363Author: Alex Deucher <alex@botch2.com> 26364Date: Wed Jun 6 00:08:45 2007 -0400 26365 26366 RADEON: implement support for posting cards based on x86 bios tables (untested) 26367 26368 This is based on the netbsd radeonfb driver by Garrett D'Amore. 26369 The code is not hooked up yet, but should allow you to post cards 26370 without needing to execute bios code. This should be useful on 26371 non-x86 platforms, for posting secndary cards, and for suspend/resume. 26372 Works on legacy bioses only (no ATOM support). 26373 26374commit e67d1420bf65055ecb6fdfe6b1b1f53aae83854a 26375Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26376Date: Tue Jun 5 22:31:50 2007 -0400 26377 26378 RADEON: fixup RADEONDisplayVideo() to better handle the crtc being used 26379 26380commit 7886405308e3288d5c86b6f2c7dbfa8ff865139c 26381Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26382Date: Mon Jun 4 21:16:56 2007 -0400 26383 26384 RADEON: fix indenting 26385 26386commit f54ad565a69deb52547fd04e123f56fc1294cd0a 26387Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26388Date: Mon Jun 4 21:13:16 2007 -0400 26389 26390 RADEON: no need to go through all the crtcs, we've got what we need 26391 26392commit 687879bff716ad01f9f158860deb8ba770faab99 26393Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26394Date: Mon Jun 4 21:08:40 2007 -0400 26395 26396 RADEON: add support for RMX on DVI, make sure RMX is only enabled on crtc1 26397 26398commit e6161e472ff266f69547704a61040228a8704b06 26399Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26400Date: Mon Jun 4 20:32:55 2007 -0400 26401 26402 RADEON: remove some cruft 26403 26404commit be0ce38232ea4f5679c5829ab663939144c6e617 26405Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26406Date: Mon Jun 4 20:22:15 2007 -0400 26407 26408 RADEON: number the outputs if there are more than one of the same 26409 26410commit 8ec617f6493dd0aea5d11f92e3d58c3feef8c8fd 26411Author: Dave Airlie <airlied@linux.ie> 26412Date: Sun Jun 3 17:32:43 2007 +1000 26413 26414 radeon: disable irqs at server start until 3D app starts 26415 26416commit 2d40fa55e8d7a1cfb204d66ca4a4d95a3b13d5b5 26417Author: Dave Airlie <airlied@linux.ie> 26418Date: Sun Jun 3 17:32:43 2007 +1000 26419 26420 radeon: disable irqs at server start until 3D app starts 26421 26422commit dcb64a4d3947e5a9fbda4b72e29a5b6102370f07 26423Author: Dave Airlie <airlied@linux.ie> 26424Date: Sun Jun 3 17:10:49 2007 +1000 26425 26426 radeon: disable vbl interrupts when no 3d is running on a new enough drm 26427 26428commit 73d8e3ec8536b4777490b7ba457566f02233811f 26429Author: Dave Airlie <airlied@linux.ie> 26430Date: Sun Jun 3 17:07:59 2007 +1000 26431 26432 randr-1.2 increase dri minor version for mesa to know we can do vbl on both 26433 26434commit 63f0d4ed0a98830ecbe18c6e4174689111a59b68 26435Author: Dave Airlie <airlied@linux.ie> 26436Date: Sun Jun 3 16:49:14 2007 +1000 26437 26438 randr-1.2: add support for vblank on both heads 26439 26440 Also disable vbl when 3D isn't running 26441 26442commit d7775c1b38b5bdc439a27ec2c3c3a03a5b24cf57 26443Author: Dave Airlie <airlied@linux.ie> 26444Date: Sun Jun 3 12:46:31 2007 +1000 26445 26446 randr-1.2: make native mode preferred 26447 26448 If we have a native mode from the BIOS make it preferred mode 26449 26450commit 5a9516fe4f17854acd4fbf3a8eadf5139081dbd4 26451Author: Paul TBBle Hampson <Paul.Hampson@pobox.com> 26452Date: Sat Jun 2 14:28:18 2007 -0400 26453 26454 RADEONProbePLLParameters sets pll->reference_div, and 26455 RADEONGetPanelInfoFromReg uses it. 26456 26457 This ensures that the former is called before the latter. 26458 (this should fix randr on ppc) 26459 26460commit 802804461ebdca9a951e7e562ec68fd08d8eae01 26461Author: Matthieu Herrb <matthieu@deville.herrb.com> 26462Date: Thu May 31 22:23:57 2007 -0600 26463 26464 Fix build whithout XF86DRI 26465 26466commit 0e0946e0aa2527794e07473199e851bbfbc47cb1 26467Author: Michel Dänzer <michel@tungstengraphics.com> 26468Date: Wed May 30 18:56:53 2007 +0200 26469 26470 radeon: Lots of warning fixes. 26471 26472 Move code where it's used, remove unused variables, etc. 26473 26474commit 31c1be420d5277dd15505bd73e6144827a0580cd 26475Author: Dave Airlie <airlied@nx6125b.(none)> 26476Date: Wed May 30 17:49:01 2007 +1000 26477 26478 remove these syncs, at least on rs480 it doesn't break 26479 26480 probably requires testing on other r300 based cards, with the syncs in 26481 we hang when moving the cursor into the second CRTC. 26482 26483commit 7fc02657c4d740941fbda5a8823cf45de3eca3f8 26484Merge: 800bf5327 4c61c0ee9 26485Author: Dave Airlie <airlied@nx6125b.(none)> 26486Date: Wed May 30 17:27:22 2007 +1000 26487 26488 Merge branch 'origin' into randr-1.2-test 26489 26490 Conflicts: 26491 26492 src/radeon_cursor.c 26493 src/radeon_display.c 26494 src/radeon_driver.c 26495 26496commit 800bf53279e2c2bf854682bbfd6fa16d03afed00 26497Author: Matthieu Herrb <matthieu@deville.herrb.com> 26498Date: Tue May 29 23:31:13 2007 -0600 26499 26500 Fix build without XF86DRI 26501 26502commit 4c61c0ee91a2ffeefce30972a584486f1df1d1ae 26503Author: Matthieu Herrb <matthieu@deville.herrb.com> 26504Date: Tue May 29 21:35:35 2007 -0600 26505 26506 Fix build without XF86DRI 26507 26508commit 5337e7bd0069a3f2c4ab22b21a19471427ad3d81 26509Author: Dave Airlie <airlied@nx6125b.(none)> 26510Date: Wed May 30 08:10:44 2007 +1000 26511 26512 radeon: add bios quirk for nx6125 monid 26513 26514commit bff809dc8ed07ac39e9b576a87916486a5e37156 26515Author: Dave Airlie <airlied@nx6125b.(none)> 26516Date: Wed May 30 08:02:26 2007 +1000 26517 26518 rs480: more unknown regs 26519 26520 Hardcode the values from a working fglrx run, this works for me now 26521 26522 I've no idea what it might do for anyone else 26523 26524commit 104105fee5c3945d3f210e6a4cb73ab492c61543 26525Author: Dave Airlie <airlied@nx6125b.(none)> 26526Date: Tue May 29 19:09:33 2007 +1000 26527 26528 rs480: make second crtc work with magic number in magic register. 26529 26530 I've no idea why or what this does. 26531 26532commit 5aa603bcabbb077dec169c48438c2e2ebe1195d7 26533Author: Dave Airlie <airlied@nx6125b.(none)> 26534Date: Tue May 29 07:23:24 2007 +1000 26535 26536 rs480: only has single dac 26537 26538commit c52322354fe64725733842b3356798c50e7735d5 26539Merge: dd6a966e8 975da595f 26540Author: Dave Airlie <airlied@nx6125b.(none)> 26541Date: Tue May 29 07:21:48 2007 +1000 26542 26543 Merge branch 'origin' 26544 26545commit c2637a01f9aca4032262c66ade305f5fe2c54294 26546Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26547Date: Sat May 26 18:26:35 2007 -0400 26548 26549 RADEON: remove some dead code from the last commit 26550 26551commit bbb769c4107bfcae682e46d026e54cbfb67d62cd 26552Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26553Date: Sat May 26 18:25:06 2007 -0400 26554 26555 RADEON: more rotation work (still not there) 26556 26557 - once again borrowed heavily from intel 26558 26559commit a69f90be9384244744fabfa76469ede9cd26ac98 26560Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26561Date: Sat May 26 17:54:48 2007 -0400 26562 26563 RADEON: first pass at rotation (not working yet) 26564 26565 - based heavily on intel 26566 26567commit 3b619d88a65a5801c85d120f2d39704194433f10 26568Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26569Date: Sat May 26 17:06:18 2007 -0400 26570 26571 RADEON: clean up and add comment regarding clones 26572 26573commit 089ae4afd23322bf602d340c531c36a4c8b461ed 26574Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26575Date: Sat May 26 15:49:11 2007 -0400 26576 26577 RADEON: Switch SetOutputType() to use names rather than numbers 26578 26579commit c4cef0c9481257a744ac99dfc7beb988ce51a8ee 26580Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26581Date: Sat May 26 13:12:37 2007 -0400 26582 26583 RADEON: make sure we have the lock when SYNCing 26584 26585commit 899c62e987d5c7524817ba85675e39a8a28e0232 26586Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26587Date: Sat May 26 12:54:01 2007 -0400 26588 26589 RADEON: Make sure LVDS and FP2 routing info gets written 26590 26591commit 18857184ffa6847815d349c020b003f8401e36ee 26592Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26593Date: Sat May 26 12:27:00 2007 -0400 26594 26595 RADEON: fix handling of DRI lock 26596 26597commit 3a61453efb4f04492cef823b6dd1273b55c6a785 26598Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26599Date: Sat May 26 01:38:09 2007 -0400 26600 26601 RADEON: implement backlight control for LVDS 26602 26603 This code is currently disabled as I'm not sure which 26604 laptops actually use this method for backlight control. 26605 My laptop seems to use another method as adjusting the 26606 backlight level doesn't seem to touch LVDS_GEN_CNTL. 26607 Maybe just macs? 26608 26609commit ceec3f62257bafe4771e75d3c4f1d2a517d7acf8 26610Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26611Date: Fri May 25 23:29:51 2007 -0400 26612 26613 RADEON: implement auto/analog/digital output property for DVI-I (untested) 26614 26615commit e3e9c608651e7cbb9851489274815f47d2a8dbbb 26616Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26617Date: Fri May 25 19:26:24 2007 -0400 26618 26619 RADEON: more clean of last commits 26620 26621commit a7d2c9c0f6155c4e65a2e73f3832c0d8ca7af5fc 26622Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26623Date: Fri May 25 19:09:11 2007 -0400 26624 26625 RADEON: fixup last commit and add support for ext TMDS detect 26626 26627commit e50ca35ce2ea79dadb38ce14e459eed836452ff7 26628Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26629Date: Fri May 25 18:56:32 2007 -0400 26630 26631 RADEON: clean up non-DDC probe, add detect for DVI 26632 26633commit abb9b57e7b4162ab6c21fd1e809d24f13f7e1ea1 26634Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26635Date: Wed May 23 21:37:58 2007 -0400 26636 26637 RADEON: clean up some logic 26638 26639commit 679236defd28bea8874004377436347950ff7bec 26640Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26641Date: Wed May 23 21:18:40 2007 -0400 26642 26643 RADEON: add output properties (not functional yet) 26644 26645commit 975da595f032c145ad74079ff8edeaead779dc7b 26646Author: Michel Dänzer <michel@tungstengraphics.com> 26647Date: Tue May 22 10:56:47 2007 +0200 26648 26649 radeon: Provide new DRI texOffsetStart hook when available with EXA. 26650 26651commit e38cad5633c974467f6417051eda8d487ac3686e 26652Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26653Date: Tue May 22 00:09:46 2007 -0400 26654 26655 RADEON: Only add valid connectors 26656 26657commit 8275151baac22c34149cef0b7d922771d24abc3e 26658Author: Michel Dänzer <michel@tungstengraphics.com> 26659Date: Mon May 21 10:25:48 2007 +0200 26660 26661 radeon: HW cursor cleanup. 26662 26663 Don't needlessly turn the HW cursor on/off in RADEONLoadCursor*(). 26664 26665 Besides cleaning up the code, this semms to avoid some HW cursor related 3D 26666 lockups, see https://bugs.freedesktop.org/show_bug.cgi?id=10815 . My best 26667 guess is that this is because the engine is now always idled before touching 26668 the CRTC registers. 26669 26670commit 137e3fc1899078af0f72303ab0a4e6cf35804a7b 26671Author: Michel Dänzer <michel@tungstengraphics.com> 26672Date: Mon May 21 10:25:48 2007 +0200 26673 26674 radeon: Suppress debugging output by default. 26675 26676 It can be enabled at runtime by increasing the log verbosity level. 26677 26678 Also change the prefix from (**) to (II) to make grepping the log file for 26679 defaults overridden by xorg.conf more useful again. 26680 26681 Turn some MC related debugging output into normal informational output as it's 26682 useful for recognizing corner cases that can cause stability issues. 26683 26684commit aed193a47a939451d9a6d05b02653b1d73e1d523 26685Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26686Date: Sun May 20 21:13:59 2007 -0400 26687 26688 RADEON: fix crtc1 (un)blanking code after last commit 26689 26690 - crtc1 MC control is on CRTC_GEN_CNTL 26691 - fix indenting 26692 26693commit 870c8043068a0f44b53d1148371b2cc1e3970a7b 26694Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26695Date: Sun May 20 21:03:10 2007 -0400 26696 26697 RADEON: several updates 26698 26699 - move output init from InitCrtc() functions to the output mode_set() 26700 - take the crtc off the MC when blanking 26701 - move EnableDisplay() to output commit() 26702 - clean up some dead code 26703 - don't enable dacs in initcrtc() functions, this is taken care of in EnableDisplay() 26704 26705commit 09bfc8ed000f95ede5b73f2bad69edc1a4d9bac6 26706Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26707Date: Sun May 20 18:06:22 2007 -0400 26708 26709 update to 6.6.192 for rc release 26710 26711commit 764cb73e8dec4040cdd418d249fc504399fca3ee 26712Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26713Date: Sun May 20 17:26:26 2007 -0400 26714 26715 Fix regular/"xinerama"/zaphod dualhead mode 26716 26717 - logic in RADEONUnblank() was wrong 26718 - Calling RADEONSetupConnectors() on second instance screwed up the port info 26719 - still seem to be HW cursor issues with zaphod mode 26720 26721commit 9a147fef8e0e2ede2a0008c4ecfbd9b00c8dc5f6 26722Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26723Date: Fri May 18 23:45:11 2007 -0400 26724 26725 RADEON: if connector is VGA set TMDS to none 26726 26727commit bbd6faff4c6acb48970d774375c8a61861405f96 26728Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26729Date: Fri May 18 23:30:19 2007 -0400 26730 26731 RADEON: Change default LVDS i2c line for powerpc 26732 26733commit 7940ec364f3cbf02ba64b92c74cbaad4555baf38 26734Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26735Date: Fri May 18 00:23:08 2007 -0400 26736 26737 RADEON: cleanup 26738 26739commit f71bfde7352ef858c1041037d7dc77c237e315a4 26740Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26741Date: Fri May 18 00:12:03 2007 -0400 26742 26743 RADEON: more cleanup 26744 26745commit f711e266d6927dec648f6ff26f15c6f48643f78c 26746Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26747Date: Thu May 17 23:59:07 2007 -0400 26748 26749 RADEON: turn off all outputs in screeninit(). We'll turn on the ones we want later 26750 26751commit bfc1c372d7475b7fa6bffb8a79dec1dc2f11ec59 26752Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26753Date: Thu May 17 23:45:29 2007 -0400 26754 26755 RADEON: More cleanup 26756 26757commit c60b3bc9e3e7463bdb42e2478be8cc3f22c63c68 26758Author: Paul TBBle Hampson <Paul.Hampson@pobox.com> 26759Date: Thu May 17 23:26:54 2007 -0400 26760 26761 Fix for infinite loop in RADEONGetLVDSInfo 26762 26763commit 11289f8206cab1a94bd64a3938cf9af50f19497e 26764Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26765Date: Tue May 15 00:48:31 2007 -0400 26766 26767 RADEON: remove some debugging code 26768 26769commit fba1a11e287ebf04cf311645d31299896bad7283 26770Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26771Date: Tue May 15 00:44:29 2007 -0400 26772 26773 RADEON: Move DAC regs to their own Save() function 26774 26775commit 9ed00e959fcdab2739a7e64ca1e303a6faf28f15 26776Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26777Date: Tue May 15 00:36:05 2007 -0400 26778 26779 RADEON: Lots of small fixes 26780 26781 - remove extra crtc2_base set 26782 - remove some debugging code 26783 - clean up RestoreMode() 26784 - make sure a DDC reg gets assigned for LCD_DDC 26785 - make sure we adjust the right frame in AdjustFrame() 26786 - rename RADEONMapControllers() to RADEONPrintPortMap() to 26787 reflect what it actually does now 26788 - make i2c bus name match the DDC port 26789 - remove or comment out un-needed code 26790 26791commit 5106c8fe5f74865cac8b00937739a4efde9fc254 26792Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26793Date: Sun May 13 21:48:26 2007 -0400 26794 26795 RADEON: Further cleanup 26796 26797commit 224a73e41e9be344d5644203e7ebd5a3a8272604 26798Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26799Date: Sun May 13 21:41:59 2007 -0400 26800 26801 RADEON: remove remnants of "old" multi-head support 26802 26803 The old screen based multi-head code is broken at this point 26804 and I have no intention of fixing it, so to clean things up 26805 I've removed it. 26806 26807commit 9c2f20a83a5fed14225f4c3ebcd8ca41e9d4bd48 26808Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26809Date: Sun May 13 21:13:23 2007 -0400 26810 26811 RADEON: additional cleanups 26812 26813commit e187321ab8dd58d2b2fe92c062d070ba4820a2bf 26814Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26815Date: Sun May 13 19:49:10 2007 -0400 26816 26817 RADEON: add new ConnectorTable option and re-add PanelSize option 26818 26819commit e60a7bcaf5611ad4706b1d1442ce4cae49145f42 26820Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26821Date: Sun May 13 18:14:29 2007 -0400 26822 26823 RADEON: remove mergedfb, etc. cruft from man page 26824 26825commit 51caa5dccc4a2e14d1b41bb7868ad20468dd4cfe 26826Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26827Date: Sun May 13 18:10:02 2007 -0400 26828 26829 RADEON: remove old mergedfb and dualhead options that are no longer used 26830 26831commit c5da9d4040cb08598d171d20f84d3f6c20a033e0 26832Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26833Date: Sun May 13 18:05:38 2007 -0400 26834 26835 RADEON: remove more dead code 26836 26837commit e776fa9ecc53d63f916a5447a76fb2bb3b824167 26838Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26839Date: Sun May 13 15:43:39 2007 -0400 26840 26841 RADEON: re-org randr code 26842 26843 - move crtc stuff to radeon_crtc.c 26844 - move output stuff to radeon_output.c 26845 26846commit 58ce388452b7bc790c438d75c9cf4a0f69f0d7b2 26847Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26848Date: Sun May 13 15:05:01 2007 -0400 26849 26850 RADEON: Remove dead code 26851 26852commit 66e8e6c8348d007930730e90295588efe8108844 26853Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26854Date: Sun May 13 14:25:03 2007 -0400 26855 26856 RADEON: several fixes 26857 26858 - give better DDC names 26859 - disable DAC if connector is DVI-D 26860 - fix indentation in ProbeDDCModes() 26861 26862commit 1c16c2ce9c5b02b03d23da965127d82eea4c4039 26863Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26864Date: Sun May 13 13:54:04 2007 -0400 26865 26866 RADEON: fix VT switch 26867 26868commit 7c66e903368f77ecc4d3bd1c9f08d2adbd85e83c 26869Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26870Date: Sun May 13 13:37:59 2007 -0400 26871 26872 RADEON: move crtc offset handling into init/save/restore functions 26873 26874commit aec078eb0740651fba8ec602e8239bd679efc8ad 26875Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26876Date: Sun May 13 11:57:57 2007 -0400 26877 26878 RADEON: fix up DDCConnected() 26879 26880commit 0550c37ecc434b8075fb3c367d100ff27625bb64 26881Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26882Date: Sun May 13 11:44:50 2007 -0400 26883 26884 RADEON: don't need to pass pScrn to EnableDisplay() 26885 26886commit 117220527de9fd3158f600645bcfcaf46847f45f 26887Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26888Date: Sun May 13 11:37:35 2007 -0400 26889 26890 RADEON: remove hardcoded output limit and PortInfo stuff 26891 26892commit 7e5c29961ac2a9e9dbe5d6d2d73d11cd018d62b5 26893Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26894Date: Fri May 11 18:00:40 2007 +0200 26895 26896 RADEON: Fix RMX after the last commit 26897 26898commit ab5603edd8fc3ef0560bdfb6a6d9c6af2a49d1e5 26899Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26900Date: Fri May 11 17:34:35 2007 +0200 26901 26902 RADEON: Move LVDS, TMDS, DAC properties to the output rec 26903 26904commit 94eb0681de0641e490f06486468617a727fefe86 26905Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26906Date: Fri May 11 14:39:48 2007 +0200 26907 26908 RADEON: switch output dpms to use RADEONEnableDisplay() 26909 26910commit 673ede5578d5d9caf2adf0445fe1e684b034eea5 26911Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26912Date: Fri May 11 14:28:17 2007 +0200 26913 26914 RADEON: re-arrange output mode setting 26915 26916commit 0070a7d787adaae99f7bc2659be4b0f49f439db5 26917Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26918Date: Fri May 11 13:06:15 2007 +0200 26919 26920 RADEON: more re-org 26921 26922 - move crtc mode setting around 26923 - add dri lock/unlock to crtc lock/unlock calls 26924 26925commit bba456232ac9a6218aa7fbd504d6093fa72860cf 26926Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26927Date: Thu May 10 15:33:51 2007 +0200 26928 26929 RADEON: fix VT switch 26930 26931commit 1779a12a947401e5c6bcf784b47e9b3c80d37204 26932Author: Jesse Barnes <jbarnes@jbarnes-mobile.amr.corp.intel.com> 26933Date: Thu May 10 06:22:35 2007 -0700 26934 26935 Add cscope files to .gitignore 26936 26937commit 0cb23277666db3b30438c6f88840d861e04df414 26938Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26939Date: Thu May 10 15:20:56 2007 +0200 26940 26941 RADEON: randr driver re-org checkpoint 26942 26943 - split the mode setting per-crtc 26944 - reduce start up flicker 26945 26946commit 33c370b1d8350945f80ac12097d3e91243a400f2 26947Author: Jesse Barnes <jbarnes@jbarnes-mobile.amr.corp.intel.com> 26948Date: Wed May 9 16:16:39 2007 -0700 26949 26950 RADEON: 26951 - fix an ugly modesetting bug: if we happened to set the mode on 26952 CRTC1 before CRTC2, CRTC2's RestoreMode function would clobber 26953 CRTC1's CRTC_OFFSET register since we never updated 26954 ModeReg.crtc_offset... so make AdjustFrame use ModeReg and pull 26955 the call to it up before RestoreMode, seems to work ok here. 26956 26957commit 6263248a0044777a352e4ee7380b4b8f9afd091b 26958Author: Jesse Barnes <jbarnes@jbarnes-mobile.amr.corp.intel.com> 26959Date: Wed May 9 14:52:00 2007 -0700 26960 26961 RADEON: 26962 - use fixup_mode hook to set RADEON_USE_RMX flag so panel scaling 26963 works 26964 - use valid_mode hook to prune invalid default modes from list 26965 - use adjusted_mode in crtc_mode_set (using adjusted_mode from 26966 fixup hook) 26967 26968commit d2497009e395800fbde5777465f3087a54b94418 26969Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26970Date: Wed May 9 22:23:45 2007 +0200 26971 26972 randr checkpoint 26973 26974 - server still hangs if you start with external monitor connected 26975 - RMX not working 26976 - more than 2 outputs now possible (untested) 26977 26978commit 4488f0737d5268168eab41440b7a3b5732efb15e 26979Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26980Date: Wed May 9 16:41:13 2007 +0200 26981 26982 more randr re-work 26983 26984 - remove RADEONQueryConnectedDisplays(); randr takes care of this now 26985 - print edid after randr detection 26986 26987commit 4d992386e2ab9d8c50f0484445564325dfb42930 26988Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26989Date: Wed May 9 16:12:21 2007 +0200 26990 26991 enable all outputs on each crtc on mode restore 26992 26993commit 2618cf2aa8ed76411b943eb90c95869814c2f151 26994Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 26995Date: Wed May 9 15:48:40 2007 +0200 26996 26997 More re-org to allow more than 2 outputs 26998 26999 - Move radeon output crtc map into crtc_set_mode 27000 - in modeinit, set up all outputs attached to crtc 27001 27002commit 61b9e79cbeee6f735a4c82ec8a802aee85d8b890 27003Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 27004Date: Wed May 9 15:11:23 2007 +0200 27005 27006 start to clean up MapControllers 27007 27008commit 6a724dd798c1a7b461672993c02be83a7bccded6 27009Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 27010Date: Wed May 9 14:48:17 2007 +0200 27011 27012 minor cleanup 27013 27014commit 03860fed24b4f76cc7f1f4210ec8f8040fa04777 27015Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 27016Date: Tue May 8 19:27:48 2007 +0200 27017 27018 remove some old mergedfb cruft 27019 27020commit 76670f665ebec7cdf40a04bf9379cb3ad4417507 27021Merge: 83f81ed5e a3ee42207 27022Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 27023Date: Tue May 8 18:41:25 2007 +0200 27024 27025 Merge branch 'master' into randr-1.2 and fix conflicts 27026 27027commit 83f81ed5e3c33c94c80500316c37a7cbfc51f41f 27028Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 27029Date: Mon May 7 22:54:16 2007 +0200 27030 27031 RADEON: Fix cursor handling for randr 27032 27033commit 5fc21cd34436919300018d8d4850fc67db284eb3 27034Author: Jesse Barnes <jbarnes@jbarnes-mobile.amr.corp.intel.com> 27035Date: Mon May 7 13:31:20 2007 -0700 27036 27037 - fix randr 1.2 on pre-RV350 chips by enabling DAC_MACRO_CNTL writes 27038 - set dac_cntl on non-primary crtcs 27039 - set XCRT_CNT_EN in CRTC_EXT_CNTL just because 27040 - fix warnings in calls to xf86PrintModeline (wants screen index not 27041 pScrnInfo) 27042 27043commit 99b3df154317f0209618e532282a3e7ad091c00f 27044Author: Alex Deucher <alex@t41p.hsd1.va.comcast.net> 27045Date: Mon May 7 19:02:35 2007 +0200 27046 27047 RADEON - update randr cursor handling, LVDS setup fix 27048 27049 - quick fix for the cursor handling to update to the latest 27050 server bits 27051 - make sure connector type is CONNECTOR_PROPRIETARY for LVDS. 27052 27053commit a3ee42207aab77d93655a82fdcb32be38268b85f 27054Author: Dave Airlie <airlied@linux.ie> 27055Date: Wed Apr 25 09:52:22 2007 +1000 27056 27057 radeon: another attempt at fixing the mergedfb refresh rate 27058 27059 This attempts to keep it inside the 32-bit limit when multiplying things out 27060 later in the randr tree. 27061 27062 Let me know if I screwed this up.. 27063 27064commit 3828237200fc16d4d32664fb8358950c213d4897 27065Author: Dave Airlie <airlied@nx6125b.(none)> 27066Date: Sun Apr 22 11:36:00 2007 +1000 27067 27068 radeon: add support for DDC on some laptop chipsets 27069 27070 I noticed fglrx has DDC for the panel in the rs480 laptop, however radeon 27071 didn't pick it up, so I valgrinded fglrx and spotted 0x1a0/0x1a4 accesses 27072 I actually noticed this before from the BIOS but never figured it out. 27073 27074 So now I get DDC from the LCD on this laptop. 27075 27076commit dd6a966e862b774a8e8b9e1a085309219673efad 27077Author: Dave Airlie <airlied@nx6125b.(none)> 27078Date: Sun Apr 22 11:36:00 2007 +1000 27079 27080 radeon: add support for DDC on some laptop chipsets 27081 27082 I noticed fglrx has DDC for the panel in the rs480 laptop, however radeon 27083 didn't pick it up, so I valgrinded fglrx and spotted 0x1a0/0x1a4 accesses 27084 I actually noticed this before from the BIOS but never figured it out. 27085 27086 So now I get DDC from the LCD on this laptop. 27087 27088commit c81ed9bd7b37c9d02141d10f6c7bad3d0c57032f 27089Author: Dave Airlie <airlied@linux.ie> 27090Date: Sat Apr 21 18:58:40 2007 +1000 27091 27092 radeon: fix build on older server 27093 27094commit 406eec71116a58d42288a7f1c809a92d5bda7350 27095Author: Dave Airlie <airlied@linux.ie> 27096Date: Sat Apr 21 18:56:28 2007 +1000 27097 27098 radeon: fix build since patches for IBM don't actually build 27099 27100commit ad119960095b4b64f4c6793f65950c9967ce4989 27101Author: Matthias Hopf <mhopf@suse.de> 27102Date: Thu Apr 19 11:54:46 2007 +0200 27103 27104 Disable RMX for IBM Lewis server. 27105 27106 Due to the hardware layout RMX ddc_mode has to be set. 27107 If ddc_mode is set, RADEONValdiateFPModes() shouldn't be called. 27108 Bugzilla #10620 (3). 27109 27110commit 16ef77df4ebaf5ea13baa82972aaf98e71ac32ee 27111Author: Matthias Hopf <mhopf@suse.de> 27112Date: Wed Apr 18 17:36:15 2007 +0200 27113 27114 Set sync polarity restriction flags even for non-"digital separate" monitors. 27115 27116 According to Lisa Wu, this is correct regarding the VESA EEDID standard. 27117 Bugzilla #10620 (2), original patch by Lisa Wu @ATI 27118 27119commit 0abce69f0d826a7ca1a41d963cd4730b6e01c145 27120Author: Matthias Hopf <mhopf@suse.de> 27121Date: Wed Apr 18 17:32:52 2007 +0200 27122 27123 Fix inconsistent use of Mode lists. 27124 27125 Some scans used to only check every second entry, some stopped at the entry 27126 before the last entry. 27127 Bugzilla #10620 (1), original patch by Lisa Wu @ATI 27128 27129commit aea801cf9a5ce519a53d6fffd9a3a2e526ec79ea 27130Author: Matthias Hopf <mhopf@suse.de> 27131Date: Fri Apr 13 16:16:05 2007 +0200 27132 27133 Fix crash if MergedFB and secondary head not found 27134 27135 If the secondary head isn't found (Monitor unplugged etc.) but MergedFB 27136 is configured, the driver segfaults because it tries to access the mode 27137 list private structures, which are not filled in. 27138 27139commit 07ddffb32e6293c77b32c94b87ec468caef3d6f5 27140Author: Matthias Hopf <mhopf@suse.de> 27141Date: Wed Apr 11 14:36:51 2007 +0200 27142 27143 Fixed typo in mode list generation. 27144 27145commit 4effa67ea75736a31b9e78a7b35acf74b067c43e 27146Author: Dave Airlie <airlied@linux.ie> 27147Date: Mon Apr 9 22:08:31 2007 +1000 27148 27149 radeon: add support for enabling direct rendering on RS480 27150 27151 Thanks to Matthew Garrett and Ubuntu for the hw loan to get this working. 27152 27153 Still no 3D driver support but at least you should get CP acceleration for 27154 2D now. 27155 27156commit 6b25a4c48796e022a093f3072574ffe9709ecaf4 27157Author: Michel Dänzer <michel@tungstengraphics.com> 27158Date: Wed Apr 4 10:47:29 2007 +0200 27159 27160 radeon: Link nearest modes by default for clone mode. 27161 27162 This makes sure all modes of both CRTCs will be available by default with 27163 MergedFB. 27164 27165commit 3c892f163ec1fa9be6e733aab091c9b718f41efc 27166Author: Michel Dänzer <michel@tungstengraphics.com> 27167Date: Wed Apr 4 09:52:37 2007 +0200 27168 27169 radeon: Always sort modes when adding to list. 27170 27171 This makes sure mode lists will always be sorted from larger to smaller. 27172 27173commit 3a8190ccc79969925257e7b980b78d79053d208d 27174Author: Michel Dänzer <michel@tungstengraphics.com> 27175Date: Wed Apr 4 09:47:07 2007 +0200 27176 27177 radeon: Don't shrink virtual size based on secondary modes. 27178 27179commit 9b1e97284ce185d358ca756a235d2cee346fa53f 27180Author: Henry Zhao <henryz@localhost.localdomain> 27181Date: Sat Mar 31 23:01:52 2007 -0800 27182 27183 10205: Radeon driver's own mode validation code does not work properly 27184 27185commit 9c2dcd19be8fc2cc29e637d1e9748e66196e3900 27186Author: Henry Zhao <henryz@localhost.localdomain> 27187Date: Sat Mar 31 20:10:03 2007 -0800 27188 27189 9337: EDID modes do not participate in validation for CRT monitor 27190 27191commit 1acd6d6fa42acec07fb11aeb189f492ddb021cb4 27192Author: Michel Dänzer <michel@tungstengraphics.com> 27193Date: Fri Mar 30 11:06:10 2007 +0200 27194 27195 radeon: Guard some MergedFB specific code with info->MergedFB tests. 27196 27197 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=10442 . 27198 27199commit 1a71106c0e4fe5f650239dc694163fdf52d33663 27200Author: Michel Dänzer <michel@tungstengraphics.com> 27201Date: Tue Mar 27 10:13:21 2007 +0200 27202 27203 radeon: Fix typo. 27204 27205commit f87e57d4d773a019d1cc8a10425c57480430f6a4 27206Author: Dave Airlie <airlied@linux.ie> 27207Date: Tue Mar 27 18:08:54 2007 +1000 27208 27209 radeon: fix up crtc debug dereference problem 27210 27211commit 66b4a571a4e7960da6807d3f30955aa08e89ccc6 27212Author: Dave Airlie <airlied@linux.ie> 27213Date: Tue Mar 27 17:00:37 2007 +1000 27214 27215 update number to 6.6.191 for rc release 27216 27217commit fca30a6b581cb6c1466ab1bc316df8fed5d82b60 27218Author: Alex Deucher <alex@botch2.com> 27219Date: Mon Mar 26 23:26:51 2007 -0400 27220 27221 fix NULL dereference 27222 27223 when IsSecondary is true, crtc1 is NULL 27224 Noticed by Sverre Froyen. 27225 27226commit 97d8d1ed10d069343f2b3172ba64ca421821a602 27227Author: George Sapountzis <gsap7@yahoo.gr> 27228Date: Sat Mar 24 20:02:12 2007 +0200 27229 27230 Move atichip.c from ati to atimisc. 27231 27232commit aa8f5b02ebc9be60df48722588261627d6a457e8 27233Author: George Sapountzis <gsap7@yahoo.gr> 27234Date: Sat Mar 24 19:53:02 2007 +0200 27235 27236 [mach64] Use Mach64Chipsets[] instead of ATIChipNames[]. 27237 27238 atimisc is PCI-only now, we can get the chip name with xf86TokenToString(). 27239 27240commit 39e896a1e688ea2d2d21f88c1c5d34c5810aac1c 27241Author: George Sapountzis <gsap7@yahoo.gr> 27242Date: Sat Mar 24 19:47:18 2007 +0200 27243 27244 [mach64] Set pATI->ChipRevision correctly, instead of overriding. 27245 27246 There is no need to override pATI->ChipRevision for GX/CX, as it is only 27247 reported with a printf. 27248 27249commit f046a910ca117279fbabc6281b2e23439ec9ea4e 27250Author: George Sapountzis <gsap7@yahoo.gr> 27251Date: Fri Mar 23 22:19:17 2007 +0200 27252 27253 Drop probing by driver name from "Device" section. 27254 27255 atimisc is PCI-only now, so we only need to scan the PCI space. 27256 27257commit 9cd175d9cd4ed710fccb303664c77519ecaf1e21 27258Author: George Sapountzis <gsap7@yahoo.gr> 27259Date: Fri Mar 23 22:12:48 2007 +0200 27260 27261 Fold FillIn() back to Probe(). 27262 27263commit d7a8cd0e476034796fc38e25a28cd28d05ea4a13 27264Author: George Sapountzis <gsap7@yahoo.gr> 27265Date: Fri Mar 23 22:10:03 2007 +0200 27266 27267 Drop now unneeded _X_EXPORT's. 27268 27269commit 58626d8b78f26f0d9c480033d3c3a12e541342b1 27270Author: George Sapountzis <gsap7@yahoo.gr> 27271Date: Fri Mar 23 21:30:19 2007 +0200 27272 27273 Move {atimach64,r128,radeon}_probe.c from ati to subdrivers. 27274 27275 Subdrivers are now loaded from the wrapper Probe function rather than at screen 27276 creation time. 27277 27278 The wrapper Identify callback only prints chip families now, chip lists are 27279 printed when a subdriver is loaded. This also avoids duplication of subdriver 27280 Identify callbacks. 27281 27282 Unknown radeons should still get a list of known radeon and then probe fails... 27283 27284 Probe for atimisc last to avoid needless loading in most cases (r128, radeon). 27285 27286commit 2e3d43af1e5077cc61dd8668551a6291368d9ed2 27287Author: George Sapountzis <gsap7@yahoo.gr> 27288Date: Mon Mar 19 19:05:04 2007 +0200 27289 27290 [mach64] Compute MMIO address once for each case (PIO, MMIO). 27291 27292 For an MMIO-only driver, the extra MMIO address probing is not needed since we 27293 already compute the MMIO address during probe and do not make any BIOS calls. 27294 27295 For a PIO driver, this extra probing could even result in a wrong MMIO address 27296 since the checking is done by reading through PIO, not MMIO. The MMIO address 27297 is later corrected but having the extra probe there is useless. 27298 27299commit e7defc0e930f1b2d83623a769b2bfdb05c66a5fe 27300Author: George Sapountzis <gsap7@yahoo.gr> 27301Date: Mon Mar 19 19:04:10 2007 +0200 27302 27303 [mach64] Move setting of PIO and MMIO addresses up, after BIOS init. 27304 27305 Partially revert "Minor refactoring of pATI->Block0Base computation." 27306 27307 This parially reverts commit 2c8b33b761e4411451aea1eea3a89d629aeea40d. 27308 27309commit 3dfdba195f1ff794ab00585ab80e7ea6ac92fcf4 27310Author: George Sapountzis <gsap7@yahoo.gr> 27311Date: Sun Mar 18 16:40:07 2007 +0200 27312 27313 [mach64] Make mapping code friendlier to pci-rework source-code compatibility. 27314 27315commit 321db6d87c9110654ab386e5ec270116eca1e04f 27316Author: George Sapountzis <gsap7@yahoo.gr> 27317Date: Sun Mar 18 16:15:20 2007 +0200 27318 27319 [mach64] Add macros for pci-rework source-code compatibilty. 27320 27321commit dfd07b6e99020d1db43d7ce0cae4423d8c6b1f05 27322Author: George Sapountzis <gsap7@yahoo.gr> 27323Date: Sun Mar 18 16:38:26 2007 +0200 27324 27325 [mach64] Drop checking that BAR base + size fits in 32 bit. 27326 27327 I can't see how these tests could fail on 32-bit. OTOH they are potentially 27328 problematic with 64-bit PCI addresses. 27329 27330commit e674338a98c50800637b8ebc01adf3aec2a3eb38 27331Author: George Sapountzis <gsap7@yahoo.gr> 27332Date: Sun Mar 18 15:50:41 2007 +0200 27333 27334 [mach64] Single mapping per BAR. 27335 27336 - We reset write-combining of the FB when MMIO registers fall in the linear 27337 aperture. 27338 27339 - We set MMIOInLinear in all three aperture mappings since the mapping code 27340 depends on it. 27341 27342 - We map the linear aperture in its entirety for MMIOInLinear also. 27343 This simplifies mapping of HW cursor and the LE aperture on BE archs, as 27344 they always fall in the linear aperture. 27345 27346 - I assumed that BAR[2] is page-aligned (we know it's 4K-aligned for sure). 27347 27348commit e7969de8ec572d8801806f9525a9a081a58a03f1 27349Author: George Sapountzis <gsap7@yahoo.gr> 27350Date: Sun Mar 18 15:35:25 2007 +0200 27351 27352 [mach64] Drop poking the VGA MMIO aperture. 27353 27354 We will not be able to handle this when we do mappings for PCI BARs only and 27355 it should be handled with a CPIO driver anyway. 27356 27357commit 166c760a86165330175023e07c4b2bd6891633c5 27358Author: Michel Dänzer <michel@tungstengraphics.com> 27359Date: Tue Mar 20 09:16:02 2007 +0100 27360 27361 Fix advertised minimum minor version of the DRI module. 27362 27363 We don't automatically require bumped minor versions. 27364 27365commit 1bdd376dbd57de8925244f0808f974d6d8cff39d 27366Author: Michel Dänzer <michel@tungstengraphics.com> 27367Date: Tue Mar 20 09:14:39 2007 +0100 27368 27369 radeon: Only sync to hardware when really necessary with EXA. 27370 27371 In particular, don't sync again after accelerated DownloadFromScreen, which 27372 syncs implicitly. This avoids calling into the kernel when it's not necessary, 27373 which can be relevant in some situations. 27374 27375commit 3cfa3a5c8daf03aaad6fc30d275709f6eb717d29 27376Author: Michel Dänzer <michel@tungstengraphics.com> 27377Date: Tue Mar 20 09:13:24 2007 +0100 27378 27379 radeon: Unify code to release the CP. 27380 27381commit e174d8df8c801fad95e5f79cff69187c200bee6e 27382Author: Michel Dänzer <michel@tungstengraphics.com> 27383Date: Tue Mar 20 09:12:51 2007 +0100 27384 27385 radeon: Minor BlockHandler cleanups. 27386 27387 Don't flush indirect buffer in BlockHandler; it's done in LeaveServer. 27388 27389 Also set the EXA engine mode to unknown only at the end. 27390 27391commit 113fb4b61e709a9b54fc2ef73efce06011e771c1 27392Author: Michel Dänzer <michel@tungstengraphics.com> 27393Date: Tue Mar 20 09:12:09 2007 +0100 27394 27395 radeon: Disable CP line acceleration on RV280s. 27396 27397 There have been several reports of stability issues with things like the 27398 xscreensaver hack hypercube. 27399 27400commit 4651d00b183cb498879d605c4b93cd3a0c63cb33 27401Author: Michel Dänzer <michel@tungstengraphics.com> 27402Date: Tue Mar 20 09:10:19 2007 +0100 27403 27404 radeon: Make sure 3D clients will re-upload textures to video RAM after LeaveVT. 27405 27406 Walk the SAREA texList and bump the age of every active object, so their owners 27407 will consider them kicked out when they grab the HW lock next time. 27408 27409commit 2d2fb54ba370c1df9ef5102e83c17a7ff5c55403 27410Author: Michel Dänzer <michel@tungstengraphics.com> 27411Date: Thu Mar 15 17:26:56 2007 +0100 27412 27413 atimisc: Fix strict aliasing violation flagged by -Wall. 27414 27415commit cfb82631fb683a0dc9bbc8f2d214646a8f1bca0b 27416Author: Michel Dänzer <michel@tungstengraphics.com> 27417Date: Thu Mar 15 17:12:43 2007 +0100 27418 27419 radeon: Fix build warnings. 27420 27421commit 50a524445ed1ea3c4fab56de772921ba86a21d35 27422Author: George Sapountzis <gsap7@yahoo.gr> 27423Date: Thu Mar 8 17:43:48 2007 +0200 27424 27425 Move atioption.c from ati to atimisc. 27426 27427commit 52684241ba1ffb393cb655ecc09cbd0cef5b1717 27428Author: George Sapountzis <gsap7@yahoo.gr> 27429Date: Thu Mar 8 07:07:02 2007 +0200 27430 27431 Move atiprobe.c from ati to atimisc. 27432 27433commit 2a1cd107a593630001799d6cd9e72c64222553b2 27434Author: Jesse Barnes <jbarnes@jbarnes-mobile.amr.corp.intel.com> 27435Date: Tue Mar 6 14:35:50 2007 -0800 27436 27437 Add prepare/commit hooks to output and crtc func table 27438 27439commit 703c6fc0142ffc600285c13fe6dafecf988c0a1d 27440Author: Jesse Barnes <jbarnes@jbarnes-mobile.amr.corp.intel.com> 27441Date: Tue Mar 6 10:12:42 2007 -0800 27442 27443 Update for new CRTC resize hooks. 27444 27445commit 1e270b411cf691c69194932b864ee2c9db292263 27446Author: George Sapountzis <gsap7@yahoo.gr> 27447Date: Sun Mar 4 18:21:49 2007 +0200 27448 27449 [mach64] Bug 5586: overlay scaler limited to 720 pixels on ATI Rage Pro. 27450 27451commit 54f9b8c1d01ecb97e2d4d247918f0efc86ed6037 27452Author: George Sapountzis <gsap7@yahoo.gr> 27453Date: Sun Mar 4 18:16:40 2007 +0200 27454 27455 [mach64] Xv: use single surface and encoding structs. 27456 27457commit e1e55b533d3ae528c8da37dcb77ed906d05697fb 27458Author: Dave Airlie <airlied@linux.ie> 27459Date: Sun Mar 4 19:11:34 2007 +1100 27460 27461 radeon: add option to set pci aperture size from config file 27462 27463 This requires a drm > 1.26 to work 27464 27465commit 58ee31d015cf8bec0edca62a46faec0b3505be8c 27466Author: Dave Airlie <airlied@linux.ie> 27467Date: Sun Feb 25 23:29:09 2007 +1100 27468 27469 fix typo 27470 27471commit 3cfe94d5438961b869766dacbcd13fde8b770ca2 27472Merge: 31c018ca4 3b43788c4 27473Author: Dave Airlie <airlied@linux.ie> 27474Date: Sun Feb 25 23:27:19 2007 +1100 27475 27476 Merge branch 'radeon-randrv12-v4' 27477 27478 Conflicts: 27479 27480 src/radeon_display.c 27481 src/radeon_driver.c 27482 27483commit 3b43788c45f51ad2d3e8e64383c412f4ddd97207 27484Author: Dave Airlie <airlied@linux.ie> 27485Date: Sun Feb 25 23:17:31 2007 +1100 27486 27487 cleanup radeon code against master server 27488 27489commit 24c6fa7cfac5602ba9e6e2f331bcac52fab258e5 27490Author: Dave Airlie <airlied@linux.ie> 27491Date: Sun Feb 25 21:43:54 2007 +1100 27492 27493 make radeon randr build against master server 27494 27495commit 4374895d393bf170134ccffc188521fe515d5a77 27496Author: George Sapountzis <gsap7@yahoo.gr> 27497Date: Sat Feb 24 21:25:13 2007 +0200 27498 27499 [mach64] Consolidate adjustments of mode timings, part 3. 27500 27501 Cosmetic. 27502 27503commit 8b9a4ec8af90b2f9f0d0a3a7f6627061df7e01b3 27504Author: George Sapountzis <gsap7@yahoo.gr> 27505Date: Sat Feb 24 21:22:49 2007 +0200 27506 27507 [mach64] Consolidate adjustments of mode timings, part 2. 27508 27509 Factor out to seperate function. 27510 27511commit bb226c9e7218d2a65c056fe74cb7eece0550670f 27512Author: George Sapountzis <gsap7@yahoo.gr> 27513Date: Sat Feb 24 21:20:53 2007 +0200 27514 27515 [mach64] Consolidate adjustments of mode timings, part 1. 27516 27517 Move LCD block from atimode.c to atimach64.c . 27518 27519commit 31c018ca4a18ce426b29006f103f56eee7f985fa 27520Author: Roland Scheidegger <sroland@tungstengraphics.com> 27521Date: Wed Feb 21 13:47:37 2007 +0100 27522 27523 add missing part of last commit... 27524 27525commit df07fa14da73e92d1a6ee0173468ab5c075d1cbb 27526Author: Roland Scheidegger <sroland@tungstengraphics.com> 27527Date: Wed Feb 21 13:05:42 2007 +0100 27528 27529 fix alignment issues with planar yuv and a bug with packed uyvy 27530 27531 respect that all source planar yuv planes are already dword aligned. 27532 Some attempts to fix up odd widths and odd heights (which are a bit strange 27533 for 4:2:0 formats). They still don't quite work 100% correctly (at the borders) 27534 but neither do they with packed yuv formats. 27535 While here, fix totally broken packed UYVY format by inserting missing break... 27536 Both bugs reported by Felipe Contreras. 27537 27538commit dfcb431adfbbaaee0d262d32735585555a0cbde4 27539Author: Dave Airlie <airlied@linux.ie> 27540Date: Tue Feb 20 21:20:28 2007 +1100 27541 27542 radeon: move some i2c registers out into common register file 27543 27544 These regs are more useful than just multimedia 27545 27546commit 7a25512415bb40b772491bb4a773dfe02ac8b71a 27547Author: Alex Deucher <alex@botch2.com> 27548Date: Fri Feb 16 00:44:11 2007 -0500 27549 27550 Add "DRI" option 27551 27552 This option allows you to disable the DRI per card. It also 27553 removes the "RN50Force3D" option as it is now covered by this 27554 option. RN50 users should set this to TRUE if they want to force 27555 the DRI on. 27556 27557commit fa30ec6d5cd9bf4eb1a960592ca7311175219e4b 27558Author: George Sapountzis <gsap7@yahoo.gr> 27559Date: Tue Feb 13 15:35:32 2007 +0200 27560 27561 [mach64] PreInit: maxPitch, minor cosmetic. 27562 27563commit 1777dcc1956d910073e13322767d9a3de41a949b 27564Author: George Sapountzis <gsap7@yahoo.gr> 27565Date: Tue Feb 13 15:03:40 2007 +0200 27566 27567 [mach64] PreInit: pitchInc is local. 27568 27569commit 9cdf4fcc3a93a4e20244286392dc31a0b8e6a10a 27570Author: George Sapountzis <gsap7@yahoo.gr> 27571Date: Tue Feb 13 11:53:42 2007 +0200 27572 27573 [mach64] PreInit: Use goto bail. 27574 27575 Match ATILock - ATIUnlock. 27576 27577commit 482b85144fee42b6543dcc97a75899e363e05e9e 27578Author: George Sapountzis <gsap7@yahoo.gr> 27579Date: Mon Feb 12 14:27:37 2007 +0200 27580 27581 [mach64] Enable RENDER acceleration on the Pro variants. 27582 27583commit 63248f0b4308a4487cda3aa22daa36e3e0d38d14 27584Author: Dave Airlie <airlied@linux.ie> 27585Date: Mon Feb 12 19:37:36 2007 +1100 27586 27587 fix LVDS by moving bios reading around 27588 27589commit 6a25f620d40cbb063de94aa6b5267296cd5a0670 27590Author: David Airlie <airlied@asimov.stargames.com.au> 27591Date: Mon Feb 12 15:59:43 2007 +1100 27592 27593 remove assignments to pOutput before it is created 27594 27595commit 9234d8045c5fefbd1a781209409e55a13e3e5370 27596Author: David Airlie <airlied@asimov.stargames.com.au> 27597Date: Mon Feb 5 11:53:18 2007 +1100 27598 27599 remove all the locking glxgears runs fine over screen changes without it 27600 27601commit eda9fb15e8bc042661a91a3d8c921006dfb3ddd9 27602Author: George Sapountzis <gsap7@yahoo.gr> 27603Date: Tue Nov 21 04:02:41 2006 +0200 27604 27605 Consolidate atixv.c w/ atimach64xv.c, part 2. 27606 27607 Move {Initialize,Close}XVideo from atixv.c to atimach64xv.c . 27608 27609commit 67b240ddd4a6549a23885afc723b6efa7c52db4c 27610Author: George Sapountzis <gsap7@yahoo.gr> 27611Date: Tue Oct 31 16:44:10 2006 +0200 27612 27613 Consolidate atixv.c w/ atimach64xv.c, part 1. 27614 27615 Merge ATIXVInitializeAdaptor() in ATIMach64XVInitialiseAdaptor(). 27616 27617commit 4dac7896549392f7378cad9620618cec28f41a16 27618Author: George Sapountzis <gsap7@yahoo.gr> 27619Date: Sun Oct 29 03:55:39 2006 +0200 27620 27621 Consolidate aticursor.c w/ atimach64cursor.c 27622 27623commit 23b12b9ff55224e8c3c45eb58ea5fbf2f747d82e 27624Author: George Sapountzis <gsap7@yahoo.gr> 27625Date: Tue Oct 31 16:10:07 2006 +0200 27626 27627 Consolidate atiaccel.c w/ atimach64accel.c, part 2. 27628 27629 Merge ATIInitializeAcceleration() in ATIMach64AccelInit(). 27630 27631commit bc3b222b57d7edf7bdce00a53e8abd9047c314a4 27632Author: George Sapountzis <gsap7@yahoo.gr> 27633Date: Sat Oct 28 19:29:12 2006 +0300 27634 27635 Consolidate atiaccel.c w/ atimach64accel.c, part 1. 27636 27637 Move ATIResizeOffscreenLinear() from atiaccel.c to atimach64xv.c and make it 27638 static. It is only used by XV memory management functions. 27639 27640commit d876889add41c8052d4f3feaa21ddf384f5a7054 27641Author: George Sapountzis <gsap7@yahoo.gr> 27642Date: Tue Nov 21 04:47:06 2006 +0200 27643 27644 Drop non-PCI support from atividmem.c 27645 27646 - we require a PCI device (pVideo != NULL) 27647 - map the VGA framebuffer with xf86MapDomainMemory() 27648 27649commit 2c8b33b761e4411451aea1eea3a89d629aeea40d 27650Author: George Sapountzis <gsap7@yahoo.gr> 27651Date: Tue Nov 21 04:51:02 2006 +0200 27652 27653 Minor refactoring of pATI->Block0Base computation. 27654 27655commit a659b7235332e19659c56cc9d7af362bd8885a29 27656Author: George Sapountzis <gsap7@yahoo.gr> 27657Date: Sun Feb 4 11:28:54 2007 +0200 27658 27659 Cleanup checks for pATI->PCIInfo. 27660 27661 Fixes a bug where shadowfb would not be enabled by default in the non-accel 27662 case because pATI->PCIInfo is now set after ATIProcessOptions() is called. 27663 27664commit 9d77aabdff919360f0c9333105436c31f1f5749a 27665Merge: ff8ea19fc 57822be75 27666Author: George Sapountzis <gsap7@yahoo.gr> 27667Date: Mon Feb 5 19:16:51 2007 +0200 27668 27669 Merge branch 'mach64-pci-1' 27670 27671commit d7317604c843e21c1df048a9253ed55a1957cccd 27672Author: David Airlie <airlied@linux.ie> 27673Date: Fri Feb 2 14:54:00 2007 +1100 27674 27675 get rid of mergetype and displaytype from radeon.h 27676 27677commit 15c68ac19ff9971ae02cf52036ba36ddea1a6759 27678Author: David Airlie <airlied@linux.ie> 27679Date: Fri Feb 2 14:30:16 2007 +1100 27680 27681 rename some variables to output from connector 27682 27683commit 72ef23ef7dd159d827e8e122482a58928d532268 27684Author: David Airlie <airlied@linux.ie> 27685Date: Fri Feb 2 14:26:29 2007 +1100 27686 27687 fixup mode probing a bit neater 27688 27689commit 96acf6b2b242454345cc4b9cfc7ca07e0b597b43 27690Author: David Airlie <airlied@linux.ie> 27691Date: Fri Feb 2 14:26:19 2007 +1100 27692 27693 bring over setproperty from intel code 27694 27695commit 76bc53f9b153880730ab61dcd2b6e4e7717e4058 27696Author: David Airlie <airlied@linux.ie> 27697Date: Fri Feb 2 14:15:03 2007 +1100 27698 27699 move i2c buses into outputs 27700 27701commit 47fb9ce657f018177a35b449a4d716dc03be9327 27702Author: David Airlie <airlied@linux.ie> 27703Date: Fri Feb 2 11:45:40 2007 +1100 27704 27705 fixup callers to crtcsetmode fixes my cursors 27706 27707commit 720730b1b0de632488d3b9818210ec5e9c7f07ae 27708Author: David Airlie <airlied@linux.ie> 27709Date: Thu Feb 1 16:43:38 2007 +1100 27710 27711 add locks for crtcs and some missing lines from intel update 27712 27713commit 6748732658850ea506f623a3622aa7135513ffd0 27714Author: David Airlie <airlied@linux.ie> 27715Date: Thu Feb 1 16:26:31 2007 +1100 27716 27717 update to latest intel codebase modulo using their mode set 27718 27719 and it breaks my cursor 27720 27721commit a77f08298dc7e097025e3f7f92e3665c0ef30095 27722Author: David Airlie <airlied@linux.ie> 27723Date: Mon Jan 29 16:12:24 2007 +1100 27724 27725 radeon: always reset cursor regs 27726 27727commit ff8ea19fcdce099732f9359e53cd62b9a04bfa6d 27728Author: Alex Deucher <alex@botch2.hsd1.va.comcast.net> 27729Date: Tue Jan 30 21:11:09 2007 -0500 27730 27731 add option to force the DRI on for RN50 chips 27732 27733 Allow user to force the DRI on for RN50 chips. 27734 3D is not guaranteed to work on these chips, 27735 however in some cases it does. 27736 fixes bug 9802. 27737 27738commit 57822be75740f339445f2375d44632560f4bbe57 27739Author: George Sapountzis <gsap7@yahoo.gr> 27740Date: Tue Nov 21 03:08:26 2006 +0200 27741 27742 Drop pATI->Chipset, no longer used. 27743 27744commit 8b7c88e9ed549777413136c0f2ccc61747b010d8 27745Author: George Sapountzis <gsap7@yahoo.gr> 27746Date: Thu Dec 28 14:46:10 2006 +0200 27747 27748 Add option "probe_sparse" to force probing sparse I/O bases. 27749 27750commit 3654b786f1ad1f4e6bfb2fff5ff057254958f7c0 27751Author: George Sapountzis <gsap7@yahoo.gr> 27752Date: Wed Dec 27 17:59:14 2006 +0200 27753 27754 Sane probe, based on work by Luc Verhaegen <libv@skynet.be>. 27755 27756commit 0117c9bde0d0bbb487bc259cf55e8066ebd63f6a 27757Author: George Sapountzis <gsap7@yahoo.gr> 27758Date: Wed Dec 27 17:49:45 2006 +0200 27759 27760 Drop checking of sparse I/O bases conflicts 27761 27762 This will be handled with an option somehow. 27763 27764commit 87cf2f1874ca8a7cc492d2e2b6918ba69a8aa713 27765Author: George Sapountzis <gsap7@yahoo.gr> 27766Date: Mon Aug 14 17:55:14 2006 +0300 27767 27768 Minor refactoring of module (un)loading for TV_OUT. 27769 27770commit 0b31333c9147af21c50493872779da12a3668e5d 27771Author: George Sapountzis <gsap7@yahoo.gr> 27772Date: Mon Nov 20 22:08:49 2006 +0200 27773 27774 Drop support for non-programmable clocks. 27775 27776 From README.ati: 27777 27778 Clocks for supported programmable clock generators: 27779 27780 The driver currently supports all programmable clock generators known 27781 to exist on Mach64 adapters. 27782 27783 Clocks for unsupported programmable clock generators: 27784 27785 This case is unlikely to occur, but is documented for the sake of 27786 completeness. 27787 27788 Thus: 27789 27790 - check for (pATI->ProgrammableClock > ATI_CLOCK_FIXED) && 27791 (pATI->ProgrammableClock < ATI_CLOCK_MAX) 27792 27793 - drop "probe_clocks" option 27794 27795 - pATIHW->ClockUnmap is no longer used 27796 27797 - pATIHW->ClockMap is only used with NewHW.crtc which is always ATI_CRTC_MACH64 27798 and has the identity map, so drop it 27799 27800 - (pATI->ProgrammableClock != ATI_CLOCK_INTERNAL) => (pATI->depth <= 8) 27801 27802commit c9a9ea5ebd1c27881c1fac19842cfa3af55c126e 27803Author: George Sapountzis <gsap7@yahoo.gr> 27804Date: Mon Nov 20 15:09:41 2006 +0200 27805 27806 Banked memory is no longer needed. 27807 27808 - (pATI->BankInfo.BankSize = 0) in all cases, cull pATI->BankInfo 27809 27810 - only keep the minimal pATIHW.SetBank interface for save/restore 27811 27812 - clean ATISwap() a little, (NewHW.crtc != ATI_CRTC_VGA) 27813 27814 - (UseSmallApertures == TRUE) <=> pATI->VGAAdapter 27815 27816commit 383e963e275d351ea3631c352f5795340162d69f 27817Author: George Sapountzis <gsap7@yahoo.gr> 27818Date: Sun Jan 28 00:44:31 2007 +0200 27819 27820 Always use the linear aperture. 27821 27822 - drop (pATI->OptionLinear == FALSE) 27823 27824 - AcceleratorVideoRAM is always set, i.e. VGAVideoRAM is not used 27825 27826 - pATI->LinearBase is always set 27827 27828 - xf86LinearVidMem() is now checked in atipreinit() for both CPIO and MMIO 27829 27830commit 9b126f45db27496c1ad16db65a61fe641a018983 27831Author: George Sapountzis <gsap7@yahoo.gr> 27832Date: Mon Nov 20 20:38:59 2006 +0200 27833 27834 We always use the accelerator CRTC within the xserver. 27835 27836 - cull (pATI->NewHW.crtc != ATI_CRTC_MACH64). 27837 27838commit 0abcb2b2c84a50e64bb6aa9e94760659ecf33add 27839Author: George Sapountzis <gsap7@yahoo.gr> 27840Date: Sun Jan 28 00:19:15 2007 +0200 27841 27842 Clean pATI->Adapter and pATI->VGAAdapter. 27843 27844 - cull (pATI->Adapter != ATI_ADAPTER_MACH64) 27845 27846 - treat pATI->VGAAdapter as Bool 27847 27848commit 817b4b5338dba4d8eb862f1f79ef0edb75aad7f4 27849Author: George Sapountzis <gsap7@yahoo.gr> 27850Date: Wed Dec 27 21:16:18 2006 +0200 27851 27852 Keep PCI mach64 only, drop: 27853 27854 - Chip < ATI_CHIP_88800GXC 27855 - Chipset != ATI_CHIPSET_ATI 27856 - Adapter != ATI_ADAPTER_MACH64 27857 27858 - depth < 8 27859 27860commit d350860e29f043e98bfb1da74b26280f1755ab6f 27861Author: George Sapountzis <gsap7@yahoo.gr> 27862Date: Wed Dec 27 20:56:45 2006 +0200 27863 27864 Keep PCI mach64 only (atipreinit.c, atimode.c), drop: 27865 27866 - Chip < ATI_CHIP_88800GXC 27867 - Chipset != ATI_CHIPSET_ATI 27868 - Adapter != ATI_ADAPTER_MACH64 27869 27870 - depth < 8 27871 27872 atimode.c only: 27873 27874 - NewHW.crtc != ATI_CRTC_MACH64 27875 27876 This allows to drop VGACalculate(), VGAWonderCalculate() cruft early. 27877 27878commit 5fa77f2d122e7267911a15235338d8d3f9eece2e 27879Author: George Sapountzis <gsap7@yahoo.gr> 27880Date: Wed Dec 27 14:50:48 2006 +0200 27881 27882 Drop probing for VGA, VGAWonder, 8514, Mach32. 27883 27884 - ChipHasSUBSYS_CNTL 27885 - Coprocessor 27886 - SharedAccelerator 27887 - SharedVGA <=> (VGAAdapter != ATI_ADAPTER_NONE) 27888 27889commit 1b2a9e1735da63bbb8cb41939668f789cee22579 27890Author: George Sapountzis <gsap7@yahoo.gr> 27891Date: Mon Nov 20 02:35:56 2006 +0200 27892 27893 Drop AVOID_NON_PCI. 27894 27895 This was not set anyway, because configure.ac would compute ATIMISC_NON_PCI and 27896 then test ATI_AVOID_NON_PCI to set AVOID_NON_PCI... 27897 27898commit 2fb84dd2fc1a663c7b9d77e706189439ffe08381 27899Author: Alex Deucher <alex@botch2.hsd1.va.comcast.net> 27900Date: Thu Jan 25 02:01:42 2007 -0500 27901 27902 I think we can say r300 support is not experimental 27903 27904commit e155659f4b006edff31f1fda88b32bf608dbeece 27905Author: Alex Deucher <alex@botch2.hsd1.va.comcast.net> 27906Date: Thu Jan 25 01:59:27 2007 -0500 27907 27908 update man page. fixes bug 6457 27909 27910 Most cards support 2D and 3D now 27911 27912commit fa12fe1cc90dd745f3eea35a07d4f3efd652e7b4 27913Author: Dave Airlie <airlied@linux.ie> 27914Date: Tue Jan 23 19:37:51 2007 +1100 27915 27916 remove unused restore palette function 27917 27918commit 4d3649b84a3325043c38cece4499de0095cebd71 27919Author: Dave Airlie <airlied@linux.ie> 27920Date: Tue Jan 23 19:37:21 2007 +1100 27921 27922 remove isused and isactive 27923 27924commit 9149e763865598c307cbefc753ff6ebdeeaf32ae 27925Author: David Airlie <airlied@asimov.stargames.com.au> 27926Date: Tue Jan 23 15:20:11 2007 +1100 27927 27928 radeon: fixup problem with cursor not going off 27929 27930 Also fixup secondary dac detect to return unknown for now 27931 27932commit 2a13a3b641d9acf4f50472e1c8ba07633c3b78d6 27933Author: David Airlie <airlied@asimov.stargames.com.au> 27934Date: Tue Jan 23 10:36:06 2007 +1100 27935 27936 fixup init disp bandwidth 27937 27938 This probably break old-style dualhead badly 27939 27940commit cbd84bed13582e82f8b2e84aa152602474c09cd4 27941Author: Dave Airlie <airlied@linux.ie> 27942Date: Mon Jan 22 22:19:46 2007 +1100 27943 27944 removed even more mergedfb remains 27945 27946commit 6296882135b183425a219efb75374d3b172de370 27947Author: Dave Airlie <airlied@linux.ie> 27948Date: Mon Jan 22 22:14:48 2007 +1100 27949 27950 remove more mergedfb relics 27951 27952commit 26e1bab2c7a4ad52710746265b3da495b54dab36 27953Author: Dave Airlie <airlied@linux.ie> 27954Date: Mon Jan 22 22:11:55 2007 +1100 27955 27956 radeon: fixup some of the video code after randr 27957 27958 This may not be complete, I need to test on dual-head system later 27959 27960commit 9008e1caa45e4a18e6f3289c7b17036730fe578a 27961Author: Dave Airlie <airlied@linux.ie> 27962Date: Mon Jan 22 21:16:47 2007 +1100 27963 27964 remove unused merged code 27965 27966commit e29a32d711553fcb4ca9928122ac285fe0b0c1a7 27967Author: Dave Airlie <airlied@linux.ie> 27968Date: Mon Jan 22 21:09:53 2007 +1100 27969 27970 add back overlay CRTC chooser this may not work yet 27971 27972commit aa9d04ba94a3cd60b248231da517e2817591fc69 27973Author: Dave Airlie <airlied@linux.ie> 27974Date: Sun Jan 21 20:29:56 2007 +1100 27975 27976 remove alot more mergedfb support 27977 27978commit e8be0056e8ff666f63a294691661f5dab289203e 27979Author: Dave Airlie <airlied@linux.ie> 27980Date: Thu Jan 18 16:41:25 2007 +1100 27981 27982 remove mergedfb 27983 27984commit a648050a3cc60f92b1ca0b3d707aadf93d076d91 27985Author: Dave Airlie <airlied@linux.ie> 27986Date: Wed Jan 17 17:28:07 2007 +1100 27987 27988 move some cursor code around for show/hide 27989 27990commit d2ecfb507282726122bb8b0d17fd3637d0ae7d46 27991Author: Dave Airlie <airlied@linux.ie> 27992Date: Wed Jan 17 17:16:27 2007 +1100 27993 27994 make dri work again 27995 27996commit 369f7c85ceff983defb7657b80ec9cd3e5440b07 27997Author: Dave Airlie <airlied@linux.ie> 27998Date: Wed Jan 17 11:14:09 2007 +1100 27999 28000 comment out unblanking on VT switch 28001 28002commit bdb66a2042f02c4b57bd3c0181a00b39fcbdb232 28003Author: Dave Airlie <airlied@linux.ie> 28004Date: Sun Jan 14 21:07:05 2007 +1100 28005 28006 radeon: get randr-1.2 mode switching mostly working on my laptop 28007 28008 The main change is to fix the dac macro + cntl register writes for rv350 and 28009 above, this still has a problem with resetting the same mode after connect 28010 disconnect cycle, need to talk to keithp 28011 28012commit aa0e7337815d4daca4df0671768621330b759011 28013Author: Dave Airlie <airlied@linux.ie> 28014Date: Sun Jan 14 19:17:32 2007 +1100 28015 28016 use more randr-ish names for variables 28017 28018commit 9d37f23aea43b74a7ec640e6b03617c8392e2572 28019Author: Dave Airlie <airlied@linux.ie> 28020Date: Sun Jan 14 19:10:11 2007 +1100 28021 28022 radeon: move blank/unblank to use randr 28023 28024commit 6bd4fe42789c38e7e804826715214ce6badcca6a 28025Author: Dave Airlie <airlied@linux.ie> 28026Date: Sun Jan 14 18:38:59 2007 +1100 28027 28028 disable dpms on/off as it was writing regs before we wanted to 28029 28030commit 3949288ed26a91e180b178f4796f2f7e1bdc8ed6 28031Author: Dave Airlie <airlied@linux.ie> 28032Date: Sun Jan 14 16:26:39 2007 +1100 28033 28034 radeon: destroy and fixup LVDS crtc 28035 28036commit f5ffb9fb3bdd7cec020240c5831ae92ee744972c 28037Author: Timo Jyrinki <timo.jyrinki@hut.fi> 28038Date: Fri Jan 12 23:08:50 2007 +1100 28039 28040 ati: fix detection of some newer radeons using ati wrapper 28041 28042 This is mainly due to the cards having a different resource 1. 28043 Fixes 6796 28044 28045commit 1e690c57ad1dfaeb760ef6e799faa6398e9a0551 28046Author: Dave Airlie <airlied@linux.ie> 28047Date: Fri Jan 12 23:03:10 2007 +1100 28048 28049 always create a second CRTC I think this should solve any problems 28050 28051commit 25b36a4c56422ad8e25a4a1c55055c2f062213ca 28052Author: Dave Airlie <airlied@linux.ie> 28053Date: Fri Jan 5 16:13:16 2007 +1100 28054 28055 minor randr1.2 fixups 28056 28057commit abe8791e4f9fa3e88273897b351387cd33822734 28058Author: Dave Airlie <airlied@linux.ie> 28059Date: Fri Jan 5 15:09:43 2007 +1100 28060 28061 hook up DPMS through xf86 28062 28063commit 8a9c68c2234b91ed38555f892afdad30b9e4b455 28064Author: Dave Airlie <airlied@linux.ie> 28065Date: Fri Jan 5 15:09:15 2007 +1100 28066 28067 set binding for new heads 28068 28069commit d6a2b8aeb5a97ee907fd45b574bc6e4ab4b3aede 28070Author: Dave Airlie <airlied@linux.ie> 28071Date: Fri Jan 5 09:33:54 2007 +1100 28072 28073 fix LUT 28074 28075commit ef1d36e56dec1fec37cee2dfd9cb5bf8ce2c485c 28076Author: Roland Scheidegger <sroland@tungstengraphics.com> 28077Date: Wed Jan 3 15:56:23 2007 +0100 28078 28079 fix uninitialized mode regs (bug 9495) 28080 28081 Since the reorganization of the mode setting code, the mode registers relying 28082 on state already set (by bios) were not read, thus clearing out all bits the 28083 driver does not touch. At the very least, this could lead to completely 28084 nonfunctional to misbehaving dvi output (see bug 9495). Fix this by using the 28085 SavedReg values, which also makes it more obvious that those are bits which 28086 were not set by the driver previously, but come from register readback. 28087 (cherry picked from 9506f7015a1e442f4ca0bd3bfae555ec7e8a5f37 commit) 28088 28089commit 55aa832157bdebcba2d58896777942d108c352b0 28090Author: Dave Airlie <airlied@linux.ie> 28091Date: Thu Jan 4 17:57:31 2007 +1100 28092 28093 fix cursor handling 28094 28095commit 4f8a7cafdc77e98dc44f9eced876560b1ee01117 28096Author: Dave Airlie <airlied@linux.ie> 28097Date: Thu Jan 4 16:23:39 2007 +1100 28098 28099 hook up crtc modesetting 28100 28101commit 52f749c8a613ee316044abe82156ee270412ced8 28102Author: Dave Airlie <airlied@linux.ie> 28103Date: Thu Jan 4 16:23:26 2007 +1100 28104 28105 fix typo 28106 28107commit d7ff61c6822cbede7f5b59b411048d33dbae9ee4 28108Author: Dave Airlie <airlied@linux.ie> 28109Date: Thu Jan 4 13:38:26 2007 +1100 28110 28111 hook up detection of output 28112 28113commit 50439d7e9f4f6f9933bacd59f8bb3e655a35dbc2 28114Author: Dave Airlie <airlied@linux.ie> 28115Date: Thu Jan 4 13:21:57 2007 +1100 28116 28117 fixup some function calls 28118 28119commit e067bfee92f2b0877108355619b2fb9188a9d15a 28120Author: Dave Airlie <airlied@linux.ie> 28121Date: Thu Jan 4 13:21:45 2007 +1100 28122 28123 add more files for back compat 28124 28125commit f1f34627ffbe2136ac3e023c01c0430412919ded 28126Author: Dave Airlie <airlied@linux.ie> 28127Date: Thu Jan 4 12:50:03 2007 +1100 28128 28129 make restore mode non-static 28130 28131commit 249ec67c296b34d0c7c1d78602628d2f7ce6a045 28132Author: Dave Airlie <airlied@linux.ie> 28133Date: Thu Jan 4 12:43:03 2007 +1100 28134 28135 more minor fixes to get radeon up to speed 28136 28137commit a43c1d55f5f855d9e6ae939dd4eec1c607b6d514 28138Author: Dave Airlie <airlied@linux.ie> 28139Date: Thu Jan 4 12:29:51 2007 +1100 28140 28141 bring radeon randr code inline with intel randr code 28142 28143commit c54a31565e3ea50f1af7b81b1bf0c0e437099f47 28144Author: Dave Airlie <airlied@linux.ie> 28145Date: Thu Jan 4 10:39:01 2007 +1100 28146 28147 radeon: fixup fastwrites to be always off 28148 28149 My 8500 in i845 doesn't work with fastwrites even setup by the firmware. 28150 28151commit dda2bbfdc2e0da6e8b9cb8d1bf78b91f1bb8131f 28152Author: Dave Airlie <airlied@linux.ie> 28153Date: Thu Jan 4 10:29:54 2007 +1100 28154 28155 radeon: don't touch Controller[1] on single CRTC cards. 28156 28157 This might fix bug 9371 28158 28159commit 9506f7015a1e442f4ca0bd3bfae555ec7e8a5f37 28160Author: Roland Scheidegger <sroland@tungstengraphics.com> 28161Date: Wed Jan 3 15:56:23 2007 +0100 28162 28163 fix uninitialized mode regs (bug 9495) 28164 28165 Since the reorganization of the mode setting code, the mode registers relying 28166 on state already set (by bios) were not read, thus clearing out all bits the 28167 driver does not touch. At the very least, this could lead to completely 28168 nonfunctional to misbehaving dvi output (see bug 9495). Fix this by using the 28169 SavedReg values, which also makes it more obvious that those are bits which 28170 were not set by the driver previously, but come from register readback. 28171 28172commit d5c6306385cbc7b13a132ae72ed04e04e24d9886 28173Author: George Sapountzis <gsap7@yahoo.gr> 28174Date: Tue Dec 19 09:45:24 2006 +0200 28175 28176 Use portable XORG_RELEASE_VERSION macro. 28177 28178commit 97df0a32b14e34f22dce9cd3b375c1ac5fb1c5eb 28179Author: Michel Dänzer <michel@tungstengraphics.com> 28180Date: Tue Dec 19 11:28:21 2006 +0100 28181 28182 radeon: Advertise 2D engine limits to EXA when we're not using the 3D engine. 28183 28184 Also round up to the maximum width and height, as that's what EXA compares. 28185 28186commit 295823d0879a5b574bb79843a6acd43adb9259e5 28187Author: Michel Dänzer <michel@tungstengraphics.com> 28188Date: Tue Dec 19 11:19:20 2006 +0100 28189 28190 radeon: Default to leaving AGP transfer mode and fast writes unchanged. 28191 28192 Based on the assumption that firmware should have set up the card and host 28193 bridge appropriately for these settings, this may actually be safer, at least 28194 for the transfer rate; leaving fast writes enabled is hopefully safe as well, 28195 it certainly is on my sytem. 28196 28197 See https://bugs.freedesktop.org/show_bug.cgi?id=9284 . 28198 28199commit 81f3b4070b70483d6da4f7587e52a64ac69c8ca0 28200Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 28201Date: Mon Dec 18 14:06:15 2006 -0500 28202 28203 Revert "Convert callers of LookupWindow() to dixLookupWindow()." 28204 28205 This reverts commit 48ff33a1770f3684cd50184db8f1944a456d9974. 28206 28207commit 87592ffb717da1f0a1767a38918d16d60953599c 28208Author: Dave Airlie <airlied@linux.ie> 28209Date: Sun Dec 17 20:10:58 2006 +1100 28210 28211 add setdpi call 28212 28213commit 41cd50487bc85708a3f791dfa70bf85223d91a65 28214Author: Dave Airlie <airlied@linux.ie> 28215Date: Sun Dec 17 20:08:07 2006 +1100 28216 28217 add displayWidth set 28218 28219commit 555e0f988c571aa47bc62ec6d9612a71f1a1f59b 28220Author: Dave Airlie <airlied@linux.ie> 28221Date: Sun Dec 17 19:59:19 2006 +1100 28222 28223 add randr pre init 28224 28225commit 3e2c72d7ac80ed874bcf2887a5253dac1c9bb02c 28226Author: Dave Airlie <airlied@linux.ie> 28227Date: Sun Dec 17 19:49:08 2006 +1100 28228 28229 fixup even more remnants of old code 28230 28231commit 4356b031886e00b3ed5ac1dbceeadebd7fc29fc2 28232Author: Dave Airlie <airlied@linux.ie> 28233Date: Sun Dec 17 19:47:45 2006 +1100 28234 28235 fix remnant of old code 28236 28237commit 17d39502a8dbea395051007aebffdf23644e9ee1 28238Author: Dave Airlie <airlied@linux.ie> 28239Date: Sun Dec 17 19:44:39 2006 +1100 28240 28241 use getddcmodes properly 28242 28243commit bf0b364cbd8682f297e2d110ece1f72e7151340d 28244Author: Dave Airlie <airlied@linux.ie> 28245Date: Sun Dec 17 17:31:08 2006 +1100 28246 28247 fix up output namings 28248 28249commit 4962a5430844114de864418d803f3182b90792ff 28250Author: Dave Airlie <airlied@linux.ie> 28251Date: Sun Dec 17 16:58:07 2006 +1100 28252 28253 remove some code that moved into server 28254 28255commit b47013f41249516ae82e22bd75caf573da2cc13c 28256Author: Dave Airlie <airlied@linux.ie> 28257Date: Sat Dec 16 12:02:37 2006 +1100 28258 28259 hook up mode detect to new randr code 28260 28261commit 48ff33a1770f3684cd50184db8f1944a456d9974 28262Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 28263Date: Fri Dec 15 17:43:39 2006 -0500 28264 28265 Convert callers of LookupWindow() to dixLookupWindow(). 28266 28267commit 62a5e7565b66d7b7d65609c034f34b55bd266617 28268Author: Dave Airlie <airlied@linux.ie> 28269Date: Fri Dec 15 19:48:42 2006 +1100 28270 28271 bring over latest code from Intel git repo. 28272 28273 this code doesn't work yet just realigns with intel work 28274 28275commit 29124400c7f193317d41d8cfd748371a239cfea1 28276Author: Dave Airlie <airlied@linux.ie> 28277Date: Sun Dec 10 18:00:17 2006 +1100 28278 28279 hook up randr crtc setting 28280 28281commit 120c854f185c1e7711cf0dee19303fdb8894d49d 28282Author: Roland Scheidegger <sroland@tungstengraphics.com> 28283Date: Fri Dec 8 02:51:52 2006 +0100 28284 28285 radeon: avoid unnecessary OUTPLL/INPLL calls when displaying video 28286 28287 It is not necessary to always emit a OUTPLL/INPLL pair when we display 28288 a video frame. On some chips there are erratas for which the workarounds 28289 cause a 10ms delay by those calls. This is related to #5876 though those 28290 affected may suffer from other slowness issues too. 28291 28292commit ccd37b3e88cdcfe03b7c707b6082339eb8f11e6b 28293Author: Michel Dänzer <michel@tungstengraphics.com> 28294Date: Thu Dec 7 18:41:48 2006 +0100 28295 28296 radeon: Move Xv option handling code into RADEONPreInitXv(). 28297 28298 This unclutters RADEONPreInit() somewhat, but more importantly moves comparison 28299 against info->ChipFamily after that's initialized. 28300 28301commit 02be8c04a268eaa71bdfe891331758715fef4e71 28302Author: Michel Dänzer <michel@tungstengraphics.com> 28303Date: Thu Dec 7 18:31:58 2006 +0100 28304 28305 radeon: Make cache flushing for R300 series less invasive. 28306 28307 Instead of calling the DRM CP idle ioctl, just emit the cache flush commands 28308 into the CP stream. 28309 28310commit be46f7b6f7a729695a0affe2328077666283af12 28311Author: Michel Dänzer <michel@tungstengraphics.com> 28312Date: Thu Dec 7 16:27:10 2006 +0100 28313 28314 Fix radeon compile warnings. 28315 28316commit 59c5b4f69c97fb233144677ab18095f88aed006b 28317Author: Daniel Stone <daniel@fooishbar.org> 28318Date: Wed Dec 6 18:48:55 2006 +0200 28319 28320 Makefile.am: re-add ChangeLog to CLEANFILES 28321 28322commit 7611f14dc697b75cd7bb6493bb65b3a7ead379b8 28323Author: Daniel Stone <daniel@fooishbar.org> 28324Date: Wed Dec 6 18:47:33 2006 +0200 28325 28326 Makefile.am: fix changelog hook 28327 Make changelog hook safe for all situations. 28328 28329commit f2a9d2d61addc4e3d5e128d8078a04c9a136be6a 28330Author: Roland Scheidegger <sroland@tungstengraphics.com> 28331Date: Wed Dec 6 00:16:41 2006 +0100 28332 28333 radeon: fix up packed yuv broken by planar yuv patch 28334 28335 fix the forgotten leftuv value for packed yuv which is 28336 needed to get correct uv starting pixel (fixes broken clipping / 28337 non-null src start pixel as tvtime uses) 28338 28339commit d2aa57a4bdd4a0deb2c659bb07f3f57d1116f0d1 28340Author: Roland Scheidegger <sroland@tungstengraphics.com> 28341Date: Tue Dec 5 01:30:13 2006 +0100 28342 28343 radeon: use overlay scaler native planar yuv capability. 28344 28345 Radeons can do planar yuv just fine, there is no need to convert all data 28346 to packed yuv manually. This saves some cpu cycles as well as some 28347 (graphic card) ram. 28348 28349commit 5d5fa1b86e5179b061f0db47fe0227d1b84c37f8 28350Author: Dave Airlie <airlied@linux.ie> 28351Date: Mon Dec 4 18:53:33 2006 +1100 28352 28353 update radeon driver to Intel driver xf86Crtc interfaces 28354 28355commit 51d1cf19e71dd5de47f2c6467f4a1685eefd9e1e 28356Author: Dave Airlie <airlied@linux.ie> 28357Date: Sun Dec 3 16:30:01 2006 +1100 28358 28359 Add radeon randr 1.2 initial attempt at support 28360 28361 This doesn't do a huge amount yet 28362 28363commit eb17c9aed2144701ad7bd1042b2905446e4d708a 28364Author: airlied <airlied@optimus.localdomain> 28365Date: Sat Nov 18 15:48:02 2006 +1100 28366 28367 move radeon over to not using X's mode management like the intel driver. 28368 28369 this builds and at least seems to pick a mode.. might be near to hooking 28370 up randr1.2 soon 28371 28372commit 941b5120916ebff69ec0ee4ca7a46105d5306e5d 28373Author: airlied <airlied@optimus.localdomain> 28374Date: Sat Nov 18 10:41:12 2006 +1100 28375 28376 make radeon modes use generic code 28377 28378commit 608427283ea59cdfe3d85e1cf32719bc3809dc16 28379Author: airlied <airlied@optimus.localdomain> 28380Date: Sat Nov 18 10:35:31 2006 +1100 28381 28382 add more functions from i830 driver 28383 28384 This adds more generic functions from the i830 driver to the radeon 28385 28386commit d5b8cafc185b7d9fb909cb18a08615f81d89eaf8 28387Author: airlied <airlied@optimus.localdomain> 28388Date: Sat Nov 18 10:22:38 2006 +1100 28389 28390 add generic mode handling code from Intel driver 28391 28392commit 9cfa82e1670ad85746926995972a535ddf03ee07 28393Author: Dave Airlie <airlied@linux.ie> 28394Date: Sun Dec 3 15:56:38 2006 +1100 28395 28396 move monitor detection into single function 28397 28398commit 032a00d155ecdcdcab3ca0d047f552fe6b338a97 28399Author: airlied <airlied@optimus.localdomain> 28400Date: Sat Nov 18 10:23:12 2006 +1100 28401 28402 make DDC function more generic 28403 28404commit 2632ae24b3b50c69567e92611ce7c2cd2ee98baa 28405Author: Dave Airlie <airlied@linux.ie> 28406Date: Sun Dec 3 15:22:38 2006 +1100 28407 28408 re-program bios 4 scratch so we know if the LCD panel is there post startup 28409 28410commit ddec94e8f2f5746bad0e998d15303e771803d22b 28411Author: Dave Airlie <airlied@linux.ie> 28412Date: Sun Dec 3 11:07:23 2006 +1100 28413 28414 move crtc to connector mapping to a connector to crtc mapping. 28415 28416 Still not allowing any clone modes, but heading in the correct direction 28417 I hope... there is a chance this will regress something from superpatch.. 28418 28419commit 218cb7f8db35d0ce8f9fc4ece72bd5ec3f52106f 28420Author: Dave Airlie <airlied@linux.ie> 28421Date: Sun Dec 3 09:53:35 2006 +1100 28422 28423 radeon: dynamically allocate connectors 28424 28425commit f22fd02bc7f7ef67e33c9bd93eae2d7488669b5e 28426Author: Dave Airlie <airlied@linux.ie> 28427Date: Sat Dec 2 19:48:00 2006 +1100 28428 28429 make CRTCs malloc'ed rather than part of Ent 28430 28431 This is leading towards randr-1.2 believe me :-) 28432 28433commit 13efd1f75fbd29a4d1f69d7d5a5bb2b8d90ad820 28434Author: Dave Airlie <airlied@linux.ie> 28435Date: Sat Dec 2 19:16:41 2006 +1100 28436 28437 move HasCRTC2 to pRADEONEnt from info 28438 28439commit 3933b86390eb2f4198d490acd4914fb470c40f47 28440Author: Michel Dänzer <michel@tungstengraphics.com> 28441Date: Fri Dec 1 18:00:17 2006 +0100 28442 28443 Make log file output clear about origin of used overlay scaler buffer width. 28444 28445commit 563dfb52f5b24d2f765fae0394e599c11dcf0b5b 28446Author: Michel Dänzer <michel@tungstengraphics.com> 28447Date: Fri Dec 1 17:56:19 2006 +0100 28448 28449 Increase default overlay scaler buffer width for RV350. 28450 28451 Works with 1920x1080 video on my M10. 28452 28453commit 984ea3d9b7c27d8183ca94ad9d7905f03613fe10 28454Author: Michel Dänzer <michel@tungstengraphics.com> 28455Date: Fri Dec 1 17:42:23 2006 +0100 28456 28457 Fix build warning. 28458 28459commit 9f5ea3981449f29ff204eb154166e8fc813205fa 28460Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 28461Date: Wed Nov 29 16:46:57 2006 +0100 28462 28463 radeon: bug #1462, predownscale to make HD video work. 28464 28465 Use the overlay scaler's predownscale capability to make videos with large 28466 horizontal resolution work if it exceeds the scaler buffer width. Make the 28467 scaler buffer width user-configurable since we don't know it for all chips, 28468 and using predownscaling may otherwise reduce quality even if it wouldn't 28469 be needed. This should fix bug #1462. 28470 28471commit 35a5eaf5adf103ef57f8d4590dae25b0c50bd780 28472Author: airlied <airlied@optimus.localdomain> 28473Date: Tue Nov 28 17:31:05 2006 +1100 28474 28475 fixup EXA after previous commit 28476 28477commit ba047ae863aac869595a31ea94f102084636fd29 28478Author: airlied <airlied@optimus.localdomain> 28479Date: Tue Nov 28 17:26:10 2006 +1100 28480 28481 radeon: initial move framebuffer away from zero prep. 28482 28483 In theory the driver should be able to handle the front buffer not at VRAM 0 28484 In practice it didn't.. this is cleanup for at least XAA parts of the driver 28485 to allow for the frontbuffer to move. It has to re-organise a large part of ScreenInit so things happen in the correct order otherwise some things get setup in-correctly. (not sure EXA with fb not at 0 works yet...) 28486 28487commit 0d42cab2d889070ab6de05990e2b19724595f7af 28488Author: Alex Deucher <alex@t41p.(none)> 28489Date: Wed Nov 22 13:44:49 2006 -0500 28490 28491 clean up previous VIP commit. 28492 28493commit 8e0f57571228729cc40ed59266484ecc43819f44 28494Author: Alex Deucher <alex@t41p.(none)> 28495Date: Wed Nov 22 13:38:01 2006 -0500 28496 28497 Don't mess with the VIP bus on mobility chips. Seems to cause fan 28498 problems on asus laptops (see bug 7463). We don't support video-in 28499 on any laptops at the moment anyway. 28500 28501commit e4850610596a08082a528c38186268b72012e78c 28502Author: Alain Péteut <alain.peteut@gmx.ch> 28503Date: Wed Nov 22 12:54:55 2006 -0500 28504 28505 enable i2c for PCI_CHIP_RV250_If. allows AIW fuctionality to work. 28506 patch and testing from Alain Péteut 28507 28508commit ff38c2c9cf0c6acbdaf53b681ec721487cf411f8 28509Author: Michel Dänzer <michel@tungstengraphics.com> 28510Date: Sun Nov 19 17:11:46 2006 +0100 28511 28512 radeon: Bug #8989: Fix page flipping artifacts with XAA. 28513 28514 We still have to force-sync the pages on enabling page flipping with XAA as the 28515 second page may have been clobbered by the offscreen pixmap cache since they 28516 were last synced. 28517 28518commit 2dd0892817090709b2fb7ffba920595ada00fcad 28519Author: Michel Dänzer <michel@tungstengraphics.com> 28520Date: Sun Nov 19 17:09:11 2006 +0100 28521 28522 Fix compile warnings. 28523 28524commit 8c1b9269aaac746989bdb9be02915722bcf6fec0 28525Author: airlied <airlied@optimus.localdomain> 28526Date: Fri Nov 17 18:47:50 2006 +1100 28527 28528 forgot SetPitch function is called 28529 28530commit 3a0338bb487739a1740ddfb4d36d53cfa7f798c6 28531Author: airlied <airlied@optimus.localdomain> 28532Date: Fri Nov 17 18:37:00 2006 +1100 28533 28534 split mode handling code into separate modes source file 28535 28536commit 13e0a95564220d42ad8060b43b0aaaba3002157a 28537Author: David Airlie <airlied@asimov.stargames.com.au> 28538Date: Mon Nov 13 09:35:18 2006 +1100 28539 28540 fix up issue with bindings and mergedfb 28541 28542commit 63444770ff30afeb134befed5bb3a5c28b7bc4e9 28543Author: George Sapountzis <gsap7@yahoo.gr> 28544Date: Thu Nov 9 22:01:28 2006 +0200 28545 28546 [mach64] Prefer double-buffered visuals for GLX. 28547 28548 So that 3D apps apropriately fail in DRI... 28549 28550commit 0411f68d71c75cda38ad46ad5c0f5cc3cda64464 28551Author: George Sapountzis <gsap7@yahoo.gr> 28552Date: Thu Nov 9 22:00:56 2006 +0200 28553 28554 [mach64] Use asynchronous DMA mode by default. 28555 28556commit 8dd668db05e8f1282fa0e29a91e38c7cbdc97aa1 28557Author: Dave Airlie <airlied@linux.ie> 28558Date: Sun Nov 12 16:33:21 2006 +1100 28559 28560 radeon: abstract init display bandwidth from mergedfb 28561 28562commit 45a9f434b77ad892efcdba267a935135b972dd42 28563Author: Dave Airlie <airlied@linux.ie> 28564Date: Sun Nov 12 16:12:56 2006 +1100 28565 28566 add radeon Init2 function, to abstract some mergedfb mode setting 28567 28568commit d0828804aec0852afaca44ee41c4792edc40dbb4 28569Author: Dave Airlie <airlied@linux.ie> 28570Date: Sun Nov 12 15:59:02 2006 +1100 28571 28572 add initial controller binding, remove uses of mergedfb test 28573 28574commit 78061ca460c9d6fd338ccf273bbfa152f471622e 28575Author: Dave Airlie <airlied@linux.ie> 28576Date: Mon Oct 23 16:49:59 2006 +1000 28577 28578 radeon: separate static controller setup from non-static 28579 28580commit f917f65e8d2b13d8b9bfc4320f4a935f5a5c4bff 28581Author: George Sapountzis <gsap7@yahoo.gr> 28582Date: Thu Nov 2 21:16:53 2006 +0200 28583 28584 Officially require libdrm 2.2.0 28585 28586 Note that this is only for the headers at compile-time. The driver still 28587 uses libdrm 1.2.0 from Xserver at run-time. 28588 28589commit ca8c148942371c9445f32985625ce38eabc016e1 28590Author: George Sapountzis <gsap7@yahoo.gr> 28591Date: Thu Oct 26 15:40:46 2006 +0300 28592 28593 Handle .99 in version patch level. 28594 28595commit 23cec754158c94b43225b2070fc0a0b003e7bd25 28596Author: George Sapountzis <gsap7@yahoo.gr> 28597Date: Thu Oct 26 15:39:49 2006 +0300 28598 28599 [mach64] EXA: reserve DRI buffers only when DRI is active. 28600 28601commit 9cd78af1d06f753e7bd7e1acaff25398bde45f6e 28602Author: Dave Airlie <airlied@linux.ie> 28603Date: Mon Oct 23 16:52:07 2006 +1000 28604 28605 radeon: set preferred modes 28606 28607commit 914a5f0117ccb76292bdd81ea4a2a79d33319550 28608Author: Michel Dänzer <michel@tungstengraphics.com> 28609Date: Mon Oct 23 10:13:24 2006 +0200 28610 28611 radeon: Pitch doesn't matter when height == 1. 28612 28613commit 2bcb51d66edaa944379cf8c8ca1ba91fffdc20a8 28614Author: Michel Daenzer <michel@tungstengraphics.com> 28615Date: Sun Oct 22 17:26:28 2006 +0200 28616 28617 Bug #6756: Attempt to fix repeat picture acceleration. 28618 28619 Always use normalized texture coordinates on R200 and fall back if a POT 28620 texture pitch doesn't match the HW's implicit pitch. 28621 28622commit dabffb8335027b60ca1fc554423e196dfb9acd6d 28623Author: Henry Zhao <henry.zhao@sun.com> 28624Date: Tue Oct 17 21:17:47 2006 -0400 28625 28626 improved results for mergedfb xrandr refresh rates - bug 6966 28627 28628commit 831fc4a009841f68fe161b6e0db5b212d7d8079b 28629Author: Henry Zhao <henry.zhao@sun.com> 28630Date: Sun Oct 15 17:12:06 2006 -0400 28631 28632 fix bug 6966 (Henry Zhao, Shem Multinymous) 28633 28634commit 5425be996692425e57884e5a4c159fc79510e89f 28635Author: Alex Deucher <alex@t41p.(none)> 28636Date: Sun Oct 15 16:49:24 2006 -0400 28637 28638 clean up magic number 28639 28640commit 457178f418983041ea59df703a17513a47d17044 28641Author: Alex Deucher <alex@t41p.(none)> 28642Date: Sun Oct 15 16:40:08 2006 -0400 28643 28644 Disable all ouputs on initial mode set. The ones we want will get 28645 enabled by RADEONEnableDisplay(). 28646 28647commit 3fb3f410931a13542fdd4c51c4c9ecada7b37d76 28648Author: Alex Deucher <alex@t41p.(none)> 28649Date: Sun Oct 15 13:29:48 2006 -0400 28650 28651 Rework the output init code. Divide into separate functions 28652 for each output. 28653 28654commit 45e13f3c8786e75276cce85de4af653713d9b4d8 28655Author: Alex Deucher <alex@t41p.(none)> 28656Date: Sun Oct 15 13:18:01 2006 -0400 28657 28658 Remove paneloff option. This is already covered by monitorlayout and 28659 shouldn't be needed with xrandr++. 28660 Also remove commented out Bioshotkey code. 28661 28662commit 4abb5b7afea8d37e95461335e52b16168e555dab 28663Author: Alex Deucher <alex@t41p.(none)> 28664Date: Sun Oct 15 13:10:22 2006 -0400 28665 28666 Remove Bioshotkeys option. Leave the bios as configured at boot. 28667 Most distros enable this option by defalt anyway, and with xrandr++ 28668 it won't be an issue anyway. 28669 28670commit 2933140b5f1f8a6e270460d8918a14943bd0f174 28671Author: Michel Dänzer <michel@tungstengraphics.com> 28672Date: Sun Oct 15 17:05:02 2006 +0200 28673 28674 radeon: Use xf86XVFillKeyHelperDrawable() when available. 28675 28676commit 4edbfb9eb952f9b3f5f597749192c17baf15cdce 28677Author: Michel Dänzer <michel@tungstengraphics.com> 28678Date: Sun Oct 15 17:03:13 2006 +0200 28679 28680 radeon: Track screen damage all the time for page flipping. 28681 28682 This way we don't ever have to forcedly synchronize the pages so we hopefully 28683 never clobber one page's contents when we shouldn't. 28684 28685commit c9c2ac2b42398f5bfad909f3289e2545bb7b3888 28686Author: Michel Dänzer <michel@tungstengraphics.com> 28687Date: Sat Oct 14 15:49:22 2006 +0200 28688 28689 radeon: Don't try to synchronize pages when the CP isn't running. 28690 28691 Also move the DamageEmpty() call into RADEONDRIRefreshArea(). 28692 28693commit c19915a401ccf69046bed03c8d5067f2e72a094e 28694Author: Dave Airlie <airlied@linux.ie> 28695Date: Fri Oct 6 09:55:40 2006 +1000 28696 28697 always turn on DAC for scanning the CRT connect. 28698 28699 Ubuntu 63994 - myself and BenH discussed this as a good idea on irc 28700 28701commit b8547aa89f93a7fefbefb1c0d2743999930ac785 28702Author: Dave Airlie <airlied@linux.ie> 28703Date: Wed Oct 4 01:44:24 2006 +1000 28704 28705 update version number to 6.6.99 in preparation for a 6.7 branch at some stage 28706 28707commit 23a6f97e08fd49e1cae03cd97cae67a5f06b7634 28708Author: Alex Deucher <alex@samba.(none)> 28709Date: Tue Oct 3 13:48:09 2006 -0400 28710 28711 FP timing regs required for both internal and external TMDS 28712 28713commit 829be884702043fe4c9e349488d2585bf0561ae9 28714Author: Alex Deucher <alex@samba.(none)> 28715Date: Tue Oct 3 13:41:21 2006 -0400 28716 28717 remove bogus registers 28718 28719commit cd900580dbc8928ebc28514162b4c1dbfe9fd5ae 28720Author: George Sapountzis <gsap7@yahoo.gr> 28721Date: Tue Oct 3 00:32:22 2006 +0300 28722 28723 [mach64] Fix build for !defined(USE_EXA). 28724 28725commit 22860a3c0685a25f982983654303ad02aedc02ec 28726Author: George Sapountzis <gsap7@yahoo.gr> 28727Date: Sun Jul 16 19:23:56 2006 +0300 28728 28729 Bug 6242: [mach64] Use private DMA buffers. 28730 28731 Map the DMA buffers read-only. This eliminates a security problem where a 28732 client can alter the contents of the DMA buffer after submission to the DRM. 28733 28734commit cb53fe7c1e45937746e43437ae6adb0355306ae9 28735Author: Alex Deucher <alex@t41p.(none)> 28736Date: Sun Oct 1 18:46:05 2006 -0400 28737 28738 fix LVDS plus internal TMDS 28739 28740commit 145c716de6d42778e447581578f3e4db13c9a653 28741Author: David Woodhouse <dwmw2@infradead.org> 28742Date: Sun Oct 1 18:22:14 2006 +0200 28743 28744 Fix radeon vs fbdev pitch calculation. 28745 28746 Fetch the correct displayWidth from fbdevhw after setting the mode. 28747 28748commit d43ad88fa3913437f6987ab5ab46a38f0cb555a8 28749Author: Michel Dänzer <michel@tungstengraphics.com> 28750Date: Sat Sep 30 19:25:20 2006 +0200 28751 28752 radeon: Add support for page flipping with EXA. 28753 28754 Also use the damage layer directly instead of via shadowfb and blit dirty 28755 rectangles to the second page in LeaveServer in order to try and improve the 28756 tradeoff between performance and correctness. 28757 28758commit 146ea328b7962f56c6ab47fbe2adc7561b36189c 28759Author: Michel Dänzer <michel@tungstengraphics.com> 28760Date: Sat Sep 30 19:22:26 2006 +0200 28761 28762 radeon: Fix function name typo which could probably lead to an unresolved symbol. 28763 28764 Also remove declaration of an unused variable. 28765 28766commit 6671c1b01bf29d8f1cacf9306ef658b967d8a3cf 28767Author: Michel Dänzer <michel@tungstengraphics.com> 28768Date: Sat Sep 30 18:29:45 2006 +0200 28769 28770 radeon: Always call RADEONGetAccessibleVRAM(). 28771 28772 This fixes issues with Option "UseFBDev" and the new memmap code. 28773 28774 OTOH, there is potential for this to break IGPs, but that's what Bugzilla and 28775 git-bisect are for... 28776 28777commit c979c7f87fc449cc9631820101edff82c3013e66 28778Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 28779Date: Sat Sep 30 15:14:42 2006 +0200 28780 28781 radeon: Check ROM more thoroughly before treating it as an x86 BIOS. 28782 28783 This prevents the ROMs on some Mac cards getting treated as x86 BIOSen, 28784 resulting in garbage values. 28785 28786commit 37c114acbabe35881c39642e293519e2635907aa 28787Author: Michel Dänzer <michel@tungstengraphics.com> 28788Date: Sat Sep 30 15:11:42 2006 +0200 28789 28790 Remove some dead code. 28791 28792 Fix two compile warnings. 28793 28794commit 6d7b0763173f56050411d7bdf9a583b9bfc050c4 28795Author: Adam Jackson <ajax@benzedrine.nwnk.net> 28796Date: Fri Sep 29 18:50:44 2006 -0400 28797 28798 Undo unintended configure.ac change. 28799 28800 Dammit, git, grow a sensible UI already. 28801 28802commit f96ddd32d3a9e64dc0adec5c50386d6db5b182ee 28803Author: Ronald Wahl <rwahl@gmx.de> 28804Date: Fri Sep 29 17:51:42 2006 -0400 28805 28806 Hook up DFP power to DPMS events. 28807 28808commit f6b42149c5a3fee663575fe6bbe78b62695bdbc8 28809Author: Alex Deucher <alex@samba.(none)> 28810Date: Fri Sep 29 00:44:36 2006 -0400 28811 28812 tvdac handling was reversed between r200 and other chips 28813 28814commit c0660495080719c052d6393ede707755929102cd 28815Author: Dave Airlie <airlied@linux.ie> 28816Date: Wed Sep 27 18:12:52 2006 +1000 28817 28818 radeon: check DDC info for both connectors not just the pScrn one. 28819 28820 Otherwise we disable DDC before we've checked the CRT if LVDS is in use 28821 28822commit 5b98545618141e86164f3ce66469f9e16b1cf6d4 28823Author: Dave Airlie <airlied@linux.ie> 28824Date: Wed Sep 27 18:10:30 2006 +1000 28825 28826 radeon: don't move connectors around if LVDS connector is connector 1 28827 28828 On my laptop this allow me to plug or not plug a monitor without affecting 28829 my lvds 28830 28831commit 74c725bd86348b5bf9bba0c50e8a6a7885d7ff7b 28832Author: Dave Airlie <airlied@linux.ie> 28833Date: Mon Sep 25 11:35:32 2006 +1000 28834 28835 radeon: cleanup some bits from superpatch 28836 28837commit 9827afaa40720a58e2fc029c0bf92ad2fe223d11 28838Merge: 10b4b46c1 ad8259c81 28839Author: Dave Airlie <airlied@linux.ie> 28840Date: Mon Sep 25 11:25:17 2006 +1000 28841 28842 Merge branch 'radeon-sp' 28843 28844commit ad8259c814629e741ed5567923f40879cc0c7051 28845Author: Dave Airlie <airlied@linux.ie> 28846Date: Mon Sep 25 11:25:03 2006 +1000 28847 28848 radeon : add back tv-dac register write 28849 28850commit 10b4b46c16ff3748856b732f9a39de40ba197112 28851Author: Dave Airlie <airlied@linux.ie> 28852Date: Mon Sep 25 09:08:47 2006 +1000 28853 28854 remove all CVS annotations from the ATI driver files 28855 28856commit 976dec04786e9b050945c290840b4cdc978c0b23 28857Author: Dave Airlie <airlied@linux.ie> 28858Date: Mon Sep 25 09:00:21 2006 +1000 28859 28860 radeon: put back in reverseddc code 28861 28862 BenH claims it is needed, agd5f can fix it up later if it isn't. 28863 28864commit 980fb2f6bd7641c8c57769b0c67e3561903e89a9 28865Author: Dave Airlie <airlied@linux.ie> 28866Date: Sat Sep 23 08:31:56 2006 +1000 28867 28868 radeon: add some missing bits of superpatch for crtc registers 28869 28870commit 1ba4f36821e60289cad937abbb0edb273c88436c 28871Author: Dave Airlie <airlied@linux.ie> 28872Date: Sat Sep 23 08:25:42 2006 +1000 28873 28874 radeon: finish radeon_driver.c import for register changes 28875 28876commit 547543bbefe605a453bfa5ae6d063ae02c5f040e 28877Author: Alex Deucher <alexdeucher@gmail.com> 28878Date: Sat Sep 23 08:21:59 2006 +1000 28879 28880 radeon: re-organise FP and CRTC register setting routines 28881 28882commit fd978140bcb7670f28c684c06c2b6c611c26bef4 28883Author: Dave Airlie <airlied@linux.ie> 28884Date: Sat Sep 23 08:09:57 2006 +1000 28885 28886 radeon: initial CRTC register setting alignment 28887 28888commit 406f4911e7ce821002c111d6bcdec35f5b56c943 28889Author: Dave Airlie <airlied@linux.ie> 28890Date: Sat Sep 23 07:59:24 2006 +1000 28891 28892 radeon: setup the radeon init function to use secondary 28893 28894commit 452f0f8079d65679905ed5178a256534ef0db0e4 28895Author: Dave Airlie <airlied@linux.ie> 28896Date: Sat Sep 23 07:41:57 2006 +1000 28897 28898 radeon: bring over final set of changes for radeon_display.c from susperpatch 28899 28900 This brings over Alex's final set of changes... 28901 28902 i'm unsure about the disappearance of reverseddc - alex to review 28903 28904commit a8b9aa69259a0706d0688d2656ece5a292799242 28905Author: Dave Airlie <airlied@linux.ie> 28906Date: Sat Sep 23 07:38:25 2006 +1000 28907 28908 radeon: align tabs and make display detection like superpatch 28909 28910commit 8d4ae98c528a2309dc3c545893dbba2e7f76916c 28911Author: Dave Airlie <airlied@linux.ie> 28912Date: Sat Sep 23 07:26:44 2006 +1000 28913 28914 radeon: align some output with superpatch 28915 28916commit 65b54847c5e79b2ec3cb1a08b265229443c9deb9 28917Author: Dave Airlie <airlied@linux.ie> 28918Date: Sat Sep 23 07:22:31 2006 +1000 28919 28920 radeon: change function name to Displays rather than Monitors 28921 28922commit 7e16d9fff96dcdff8c25dbbe2a51be0023b7985a 28923Author: Dave Airlie <airlied@linux.ie> 28924Date: Sat Sep 23 07:21:15 2006 +1000 28925 28926 radeon: make GetConnectedMonitors a static void 28927 28928commit 6f63b3dc8b3e2299f5f02abc1c565176d2d46cee 28929Author: Dave Airlie <airlied@linux.ie> 28930Date: Sat Sep 23 07:19:43 2006 +1000 28931 28932 radeon: some minor alignments and cleanups for info-> 28933 28934commit c9f3ced3b1c194cacf1ab9436ee6c2d4efd3a8e9 28935Author: Alex Deucher <alexdeucher@gmail.com> 28936Date: Sat Sep 23 07:16:10 2006 +1000 28937 28938 radeon: add reverse display option to switch displays on non-mobility cards 28939 28940commit e752443c540d9a94522050c61ef7c33d912fece5 28941Author: Dave Airlie <airlied@linux.ie> 28942Date: Fri Sep 22 07:10:10 2006 +1000 28943 28944 radeon: flesh out mapcontrollers 28945 28946commit 7a0e4eb47655c4701fc3a4b33b6c8e9f378ccc9e 28947Author: Dave Airlie <airlied@linux.ie> 28948Date: Fri Sep 22 07:05:56 2006 +1000 28949 28950 radeon: add some info to the map controllers 28951 28952commit 75bf69c868be6164d0041a32f52244f1ef83e050 28953Author: Dave Airlie <airlied@linux.ie> 28954Date: Fri Sep 22 07:03:52 2006 +1000 28955 28956 radeon: add simple map controllers call 28957 28958commit 8ba5abd1329347c7fbf4b4d09c95da0d23b4884a 28959Author: Dave Airlie <airlied@linux.ie> 28960Date: Fri Sep 22 06:58:41 2006 +1000 28961 28962 radeon: move comment 28963 28964commit e8d77b7c36fa791a2350a6f4b15728b15d8f3de0 28965Author: Dave Airlie <airlied@linux.ie> 28966Date: Fri Sep 22 06:55:51 2006 +1000 28967 28968 radeon: refactor blank/unblank 28969 28970commit 8534723f5710e2916b4a22efe22a6b7365cc8db6 28971Author: Alex Deucher <alexdeucher@gmail.com> 28972Date: Fri Sep 22 06:36:03 2006 +1000 28973 28974 radeon: make DVI_I default connector 28975 28976commit d7fc79fd9f094ac462de0883af7e6030cb6195a6 28977Author: Dave Airlie <airlied@linux.ie> 28978Date: Fri Sep 22 06:35:34 2006 +1000 28979 28980 radeon: add enable display function 28981 28982commit 4a54886d510f26b29d27e5c9a73647554291b1a6 28983Merge: 70264d645 9fa176c79 28984Author: Dave Airlie <airlied@linux.ie> 28985Date: Fri Sep 22 06:00:04 2006 +1000 28986 28987 Merge branch 'master' into radeon-spm1 28988 28989 Conflicts: 28990 28991 src/radeon_driver.c 28992 28993commit 9fa176c7989030e7340cc9b85d0b6a7b34303091 28994Author: Alex Deucher <alex@samba.(none)> 28995Date: Tue Sep 19 13:58:17 2006 -0400 28996 28997 Provide all resolutions (instead only the largest one) in MergedFB clone 28998 mode without config file (Henry Zhao) 28999 29000commit e742aeb28c7d9d6e75932c408bcc7c44af52e303 29001Author: Alex Deucher <alex@samba.(none)> 29002Date: Tue Sep 19 13:45:33 2006 -0400 29003 29004 fix large negative refresh rates with mergedfb (Henry Zhao) 29005 fixes bug 6966 29006 29007commit 9c0c805704f1df1b66adc2a6b9313597ae5b86cc 29008Author: Dave Airlie <airlied@linux.ie> 29009Date: Mon Sep 18 19:41:21 2006 +1000 29010 29011 radeon: fix rn50 memmap 29012 29013 This should fixup the rn50/m6/m7 memory map problem that has been reported 29014 since BenH's memory mapping changes 29015 29016commit 70264d645724718a48ca229fedc04e33681b7d31 29017Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29018Date: Mon Sep 18 09:56:25 2006 +1000 29019 29020 radeon; put back old TV DAC code for now 29021 29022commit 53eb843ad8ec5a292dff0844330ecfa9b7b2a0cb 29023Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29024Date: Mon Sep 18 07:55:21 2006 +1000 29025 29026 radeon: move code around in InitCrtcRegisters 29027 29028commit bcae4352079109a9dd16117bf9ec19dd215180bf 29029Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29030Date: Mon Sep 18 07:33:48 2006 +1000 29031 29032 radeon: cleanup DPMS set on/off code to use new controller structure 29033 29034commit 87622f56b3dd7e2d2f00ad71e5b569eac31fb2b1 29035Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29036Date: Mon Sep 18 07:15:54 2006 +1000 29037 29038 radeon: cleanup DacPowerSet 29039 29040commit a417fa081926b2f84e2cbdc355cc7462d1ebdfad 29041Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29042Date: Mon Sep 18 07:15:41 2006 +1000 29043 29044 radeon: fixup InitFPRegisters function call to not pass info 29045 29046commit 8bff57a938e92dfc8729e40f7aa9ad9a96262131 29047Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29048Date: Mon Sep 18 07:02:35 2006 +1000 29049 29050 radeon: copy of overupdated bios hotkeys code from Alex 29051 29052commit 76352c684d5e840c9f7ac0781e3ecb32951b32db 29053Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29054Date: Mon Sep 18 06:54:28 2006 +1000 29055 29056 radeon: minor cleanup of masking for crtc registers 29057 29058commit a7b97697ed13ac25844890439ffba31524e03a16 29059Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29060Date: Mon Sep 18 06:53:23 2006 +1000 29061 29062 radeon: initialise controller structures 29063 29064commit 362ee3778764633cb4c9388aa329c5b715885038 29065Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29066Date: Mon Sep 18 06:41:14 2006 +1000 29067 29068 radeon: move InitFPRegisters up to align with superpatch 29069 29070commit 11fb9a14741746c752b0001c1b1d39a54e69657a 29071Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29072Date: Mon Sep 18 06:35:50 2006 +1000 29073 29074 radeon: add tv dac adj lookup 29075 29076commit a15ac59c4d1a15380910be2a7438111b09161c81 29077Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29078Date: Mon Sep 18 06:35:28 2006 +1000 29079 29080 radeon: re-add ATI patch for rn50 modes 29081 29082commit d80ea2d67b63ea2f2ce0277952af186b69b19e7d 29083Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29084Date: Mon Sep 18 06:30:30 2006 +1000 29085 29086 radeon: re-add DAC power on/off code around detect 29087 29088commit 0a93936f9f048b8837f1c765fea7ccb61f072697 29089Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29090Date: Mon Sep 18 06:25:15 2006 +1000 29091 29092 radeon: restore secondary head behaviour 29093 29094commit 6da5a0bfdedb2062ec835f45514c22917abdcf4c 29095Author: Alex Deucher <alexdeucher@gmail.com> 29096Date: Mon Sep 18 06:16:58 2006 +1000 29097 29098 radeon: move to using Alex's controller structure 29099 29100commit cd1bc1d1febf0a79d835c7dc8b353123278dec51 29101Author: Alex Deucher <alexdeucher@gmail.com> 29102Date: Mon Sep 18 02:44:09 2006 +1000 29103 29104 radeon: add save/restore of some new registers 29105 29106 move some registers about 29107 29108commit 9b5c4cf8a98b0c20bb821f5003495481fbd0c10c 29109Author: Alex Deucher <alexdeucher@gmail.com> 29110Date: Mon Sep 18 02:28:32 2006 +1000 29111 29112 radeon: add pre init controller function 29113 29114commit 311e776c4429485d29aaf11d350cab8e6fe7564b 29115Author: Alex Deucher <alexdeucher@gmail.com> 29116Date: Mon Sep 18 02:21:49 2006 +1000 29117 29118 radeon: add missing tv-out registers to register list 29119 29120commit 65cb77c79c2c5cb536494200d4fba6815d7e771c 29121Author: Alex Deucher <alexdeucher@gmail.com> 29122Date: Mon Sep 18 02:21:29 2006 +1000 29123 29124 radeon: restructure files along lines of Alex's superpatch 29125 29126commit 25fa71ca0a24190cc8be2e80769443e0e09303b3 29127Author: Henry Zhao <henryz@ati.com> 29128Date: Mon Sep 18 03:15:54 2006 +1000 29129 29130 Power up DAC around check for CRT connectedness 29131 29132commit 39fcac431eba60adf2d4e357cfc811dbae36a4e1 29133Author: Henry Zhao <henryz@ati.com> 29134Date: Mon Sep 18 03:09:43 2006 +1000 29135 29136 Bump up clock rate for ES1000 to allow higher resolutions 29137 29138commit 7d9e2b450f0fa42553156f57572a6dab70a0ea13 29139Author: Dave Airlie <airlied@starflyer-ubuntu.(none)> 29140Date: Tue Sep 12 07:04:23 2006 +1000 29141 29142 radeon: this stops mergedfb working on my ATI M7 development board 29143 29144 I've no idea what this is for, it came from an ATI codedrop and it may 29145 fix some laptop or production system, but it looks to me more like a dirty 29146 hack than a true fix, as I have an M7 which I need to specify CRT for the 29147 second head and it breaks mergedfb for me. 29148 29149commit ad1778eb1eb59de14ec309b99131931cbccf64e9 29150Author: Michel Dänzer <michel@tungstengraphics.com> 29151Date: Sun Sep 10 23:32:44 2006 +0200 29152 29153 Revert "* [ Bug 6966 ] large negative refresh rates reported" 29154 29155 This reverts caaed927a07ffbac68b08246185ef93c1e7bb98c commit. 29156 29157 See https://bugs.freedesktop.org/show_bug.cgi?id=8137 . 29158 29159commit 6e92981b024b59c2237f949321f6b2591dada519 29160Author: Tilman Sauerbeck <tilman@code-monkey.de> 29161Date: Sun Sep 3 21:11:10 2006 +0200 29162 29163 Removed the warning re. the r300 DRI driver status. 29164 29165commit caaed927a07ffbac68b08246185ef93c1e7bb98c 29166Author: Henry Zhao <henryz@localhost.localdomain> 29167Date: Thu Aug 31 13:27:45 2006 -0700 29168 29169 * [ Bug 6966 ] large negative refresh rates reported 29170 * [ Bug 6309 ] radeon/r128 fails to read hsync/vsync rates when rang 29171 * descriptor is missing 29172 * [ Bug 5473 ] Blank screen with Radeon Mobility X700 (Acer Ferrari 4005) 29173 * Reboot should not be needed for external CRT to function when connected 29174 later on Ferrari 4000 29175 * Modes need to be sorted, and sorted correctly 29176 * Provide all resolutions (instead only the largest one) in MergedFB clone 29177 mode without config file 29178 * Bump up clock rate for ES1000 to allow higher resolution 29179 29180commit 6d2028b84ff2b82c849b74776258f8e52c952e36 29181Author: Dave Airlie <airlied@linux.ie> 29182Date: Fri Aug 25 10:05:44 2006 +1000 29183 29184 and of course it takes two commits to do a simple commit 29185 29186commit e90c6feddbcd719fda3bbc3d48fd55045e9d5adc 29187Author: Dave Airlie <airlied@linux.ie> 29188Date: Fri Aug 25 09:59:58 2006 +1000 29189 29190 add changelog hook from nv driver 29191 29192commit 1e06212f792d1ed1f7545e16824b86fbd34fa284 29193Author: Adam Jackson <ajax@benzedrine.nwnk.net> 29194Date: Thu Aug 24 19:52:30 2006 -0400 29195 29196 Enable CPIO in the atimisc driver on ia64 platforms. 29197 29198 This fixes many mach64s, but potentially breaks mach32. Mach64 is widespread 29199 on ia64 machines but PCI mach32's are quite rare on any platform. 29200 29201commit dfa47b09b5b8d7eeaa3f8ae629ebb3c7efb51511 29202Author: Adam Jackson <ajax@benzedrine.nwnk.net> 29203Date: Thu Aug 24 19:50:46 2006 -0400 29204 29205 Disable DRI on xpress 200{,m} since we know it's broken. 29206 29207commit 1dcd9b1f75ba2307beb42556aa7f2f945171d5fb 29208Author: Adam Jackson <ajax@benzedrine.nwnk.net> 29209Date: Thu Aug 24 19:42:26 2006 -0400 29210 29211 Disable DRI on RN50 in a more pleasant way. 29212 29213 Rather than publish a non-existant driver name for RN50, just don't bother 29214 initializing it in the first place. 29215 29216commit 64b18e246586b1043a12f381a9d4c834464ed9b5 29217Author: Dave Airlie <airlied@linux.ie> 29218Date: Fri Aug 25 09:40:04 2006 +1000 29219 29220 update version numbers for release 29221 29222commit be229d97e1f0282f2099c7ace7cb257a2738a58f 29223Author: Dave Airlie <airlied@linux.ie> 29224Date: Fri Aug 25 09:39:46 2006 +1000 29225 29226 make colortiling default to on all cards 29227 29228commit 750ed51bc3d5119e2d2a91a05ae5a971dcd3b88a 29229Author: George Sapountzis <gsap7@yahoo.gr> 29230Date: Mon Aug 7 17:09:02 2006 +0300 29231 29232 [mach64] RENDER support: drop caching DP_WRITE_MASK. 29233 29234 Caching DP_WRITE_MASK causes artifacts, e.g. with evas_xrender_x11_test. 29235 29236commit b2beea2fa8949874d3d57fb9b43fe85cc08a8bff 29237Author: George Sapountzis <gsap7@yahoo.gr> 29238Date: Sat Aug 5 18:36:24 2006 +0300 29239 29240 [mach64] RENDER support: save/restore, cache texture registers. 29241 29242commit 43aaed99950640c3695b3c2b91faabf00c6338a7 29243Author: George Sapountzis <gsap7@yahoo.gr> 29244Date: Sat Aug 12 22:12:07 2006 +0300 29245 29246 [mach64] RENDER support (bug #6877). 29247 29248 RENDER accleration is disabled by default. It is strongly recommended that the 29249 patch from bug #6772 and the 'exa-damagetrack branch' are merged in the xserver 29250 before enabling RENDER acceleration. To enable RENDER acceleration, add the 29251 following in xorg.conf: 29252 29253 Section "Device" 29254 [...] 29255 29256 Option "AccelMethod" "exa" 29257 Option "RenderAccel" "true" 29258 EndSection 29259 29260commit d113f40fd97408661193895ff3aa82262f100143 29261Author: Dave Airlie <airlied@linux.ie> 29262Date: Sat Aug 12 17:16:33 2006 +1000 29263 29264 radeon: make some local data static 29265 29266 This is inspired by a patch in XFree86 from David Dawes, it isn't the 29267 same as we have different gamma curves. 29268 29269commit a1da2b5e0a38feae01107676a76bdc5c185ad34b 29270Author: George Sapountzis <gsap7@yahoo.gr> 29271Date: Sat Aug 5 03:18:34 2006 +0300 29272 29273 [mach64] Halve max w/h for hardware-accelerated operations. 29274 29275 EXA hits these limits for some operations on offscreen pixmaps and GTPRO seems 29276 to support the reduced limits. 29277 29278commit a66c8d2f9855caa10142b39f155dd8232b1f2389 29279Author: George Sapountzis <gsap7@yahoo.gr> 29280Date: Sat Aug 5 03:16:56 2006 +0300 29281 29282 [mach64] EXA: add memcpy-based UTS/DFS. 29283 29284 EXA hits more optimized paths when it does not have to fallback because of 29285 missing UTS/DFS. 29286 29287commit 0fb8503d60cf646c91a806f9ed322f9ceff18baa 29288Author: George Sapountzis <gsap7@yahoo.gr> 29289Date: Sat Aug 5 03:16:15 2006 +0300 29290 29291 [mach64] XAA: minor refactoring of memory manager setup. 29292 29293 - DRI: factor out to separate function and add comments on layout 29294 - no DRI: factor out to separate function, maxScanlines is always equal to 29295 ATIMach64MaxY 29296 29297commit e203d86643d5d70bf18248712d05b72b79aee705 29298Author: George Sapountzis <gsap7@yahoo.gr> 29299Date: Sat Aug 5 03:15:42 2006 +0300 29300 29301 [mach64] EXA support. 29302 29303commit dc1e289a611a17090e6dc7ae8a8d3f26d20df4eb 29304Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 29305Date: Sat Jul 29 01:05:35 2006 +0300 29306 29307 Bug #6623: Fix I2C bus Mach64 initialisation failure. 29308 29309commit b377f403e0cd5253402ef3945e5944d1a2763a94 29310Author: Marc Aurele La France <tsi@xfree86.org> 29311Date: Sat Jul 29 01:03:34 2006 +0300 29312 29313 Fix warnings for Option:1 bit fields. 29314 29315commit 91b8b0ca41ad0b9659f0982a05148cab8558e9d5 29316Author: Marc Aurele La France <tsi@xfree86.org> 29317Date: Sat Jul 29 01:02:48 2006 +0300 29318 29319 Bug #807: copy throttling. 29320 29321 Fix corruption associated with the engine randomly not waiting for a copy 29322 operation to commit its results. 29323 29324commit 7800407a2db4b7307192bf3b43fffb9c4748b9c0 29325Author: Marc Aurele La France <tsi@xfree86.org> 29326Date: Sat Jul 29 01:00:56 2006 +0300 29327 29328 Bug #807: invalidate read-back cache. 29329 29330 Fix corruption when scaling 1-pixel wide pixmaps, in which case the first CPU 29331 read of the framebuffer will return stale data. 29332 29333 This bug always manifests itself in mach64 EXA (to be committed soon), with 29334 this patch mach64 EXA passes rendercheck. 29335 29336commit 49b11c540d37152eee4907ab30353ede01e4fb56 29337Author: George Sapountzis <gsap7@yahoo.gr> 29338Date: Sat Jul 29 00:52:28 2006 +0300 29339 29340 Bug #807: split out TestRegisterCaching block for DP regs. 29341 29342commit 290c5aaa580428bc4748a47e4b9bcc22b90fc8d4 29343Author: George Sapountzis <gsap7@yahoo.gr> 29344Date: Sat Jul 29 00:51:53 2006 +0300 29345 29346 Bug #807: split out TestRegisterCaching block for DP regs. 29347 29348commit fba8c839b47e2c8d6a6d65950c3431ff5b870aa0 29349Author: Michel Dänzer <michel@tungstengraphics.com> 29350Date: Sun Jul 23 18:51:47 2006 +0200 29351 29352 Bug #6819: Fix broken rendering with *8b8g8r8 source formats. 29353 29354 Use the corresponding texture formats with R200 family. 29355 29356 Unfortunately, it looks like the R100 family can't accelerate these source 29357 formats, so we have to fall back for them. 29358 29359commit 60ebe97e2d5d27c2bf4213e201cef74503809cb6 29360Author: Tilman Sauerbeck <tilman@code-monkey.de> 29361Date: Mon Jul 3 19:37:28 2006 +0200 29362 29363 Move variable declarations before statements to be C89 compatible 29364 29365commit 6900ff3c5e574881b1470e2363b200e99ab8bf3e 29366Author: Alex Williamson <alex.williamson@hp.com> 29367Date: Mon Jul 3 13:09:47 2006 +0200 29368 29369 Bug #6872: Disable digital output while reading ROM with Radeon 7000. 29370 29371 This is a workaround for ATI Radeon 7000 erratum F1. 29372 29373commit 920ac1f56fad6f48a30e65836ce054815f270738 29374Author: Michel Dänzer <michel@tungstengraphics.com> 29375Date: Mon Jul 3 01:21:18 2006 +0200 29376 29377 Do even less cache flushing when the previous engine mode is known. 29378 29379 OTOH, flush everything when the previous engine mode is unknown, and mark the 29380 engine mode as unknown in a couple more cases. 29381 29382commit dd0dd4bab76907b4b1b6d814f24e73e57d71f587 29383Author: Eric Anholt <anholt@freebsd.org> 29384Date: Mon Jul 3 01:10:10 2006 +0200 29385 29386 Bug #6762: Remove gratuitous cache flushing. 29387 29388commit 22f165f290f57e638d198f24a37f62fc16b72bae 29389Author: Michel Dänzer <michel@tungstengraphics.com> 29390Date: Mon Jul 3 00:39:45 2006 +0200 29391 29392 Build and warning fixes for !defined(USE_XAA). 29393 29394commit ea5105230c7c3519730b8d3c250e9ba9a4e4374a 29395Author: Michel Dänzer <michel@tungstengraphics.com> 29396Date: Mon Jul 3 00:03:22 2006 +0200 29397 29398 Reserve HW cursor memory statically instead of from EXA offscreen. 29399 29400 As a result, quite a bit of code can be removed that attempted to deal with the 29401 EXA offscreen memory disappearing, with varying success. Because the HW cursor 29402 memory is now always immediately after the front buffer and before the back 29403 buffer, this also fixes bug #6808. 29404 29405 As the HW cursor memory is now reserved statically with EXA and XAA, change 29406 RADEONUseHWCursor() to check info->cursor instead of info->cursor_offset, for 29407 which 0 might become valid with HW cursor enabled in the future. 29408 29409 Change info->cursor_offset from unsigned long to CARD32 as the corresponding 29410 register is 32 bits wide, and to avoid issues with printf. 29411 29412commit 66586f905857ea5bf37cb5d34949cd450a9ed067 29413Author: Michel Dänzer <michel@tungstengraphics.com> 29414Date: Sun Jul 2 23:38:20 2006 +0200 29415 29416 Clean up some redundant code. 29417 29418commit 1159c85861dc507ebf323aca82e32fcc67726420 29419Author: Michel Dänzer <michel@tungstengraphics.com> 29420Date: Sun Jul 2 18:24:28 2006 +0200 29421 29422 Bug #6548: Fix playback of interlaced video with more than 1023 lines. 29423 29424 This will still break with 2048 lines... 29425 29426commit 11bedb7c845fc6074383890d5525315aebe05f2c 29427Author: Michel Dänzer <michel@tungstengraphics.com> 29428Date: Sun Jul 2 17:03:23 2006 +0200 29429 29430 Bug #6939: Disable RENDER acceleration by default on RN50. 29431 29432 Based on patch from Weixing Zhang, but still allows enabling it with Option 29433 "RenderAccel". 29434 29435commit 4a0c22e32901a510c7aad070be25c31519c284e3 29436Author: Weixing Zhang <anzhang@ati.com> 29437Date: Sun Jul 2 16:55:25 2006 +0200 29438 29439 Bug #6939: Disable 3D acceleration on RN50 by setting invalid ClientDriverName. 29440 29441commit 9c4a1e3300d30b23806988aa9328fc6ef92e2663 29442Author: Frank Gevaerts <frank@gevaerts.be> 29443Date: Thu Jun 29 18:21:22 2006 +0200 29444 29445 Fix build for !defined(USE_EXA). 29446 29447commit ee0544b670567491b7770f469a64a6cd85519ba1 29448Author: Michel Dänzer <michel@tungstengraphics.com> 29449Date: Thu Jun 29 18:19:58 2006 +0200 29450 29451 Remove unused variables. 29452 29453commit 15c556f5a8c9069d3dd00ce139d46c39b7298951 29454Author: Michel Dänzer <michel@tungstengraphics.com> 29455Date: Thu Jun 29 18:18:21 2006 +0200 29456 29457 Don't declare unused variables on __powerpc__ and __alpha__. 29458 29459commit 055ac2836ec0afd51cf554d043a59068c36a07d3 29460Author: Michel Dänzer <michel@tungstengraphics.com> 29461Date: Thu Jun 29 18:14:45 2006 +0200 29462 29463 Fix strict aliasing violation flagged by gcc -Wall. 29464 29465commit 4b1904017caa976c138594a86e75feaf470e72b5 29466Author: Dave Airlie <airlied@linux.ie> 29467Date: Mon Jun 26 20:34:34 2006 +1000 29468 29469 radeon: force CP and VIP clocks on some r300 and rv100 chips. 29470 29471 fglrx appears to do this on r300 and Jerome Glisse has spent most of his 29472 life tracking down what caused the 9800 to lockup. This is my attempt to 29473 fix this. Please report if it works for you. 29474 29475 Signed-off-by: Dave Airlie <airlied@linux.ie> 29476 29477commit c5d8c8db3e75976687bfb364f0a7b235c7ae8793 29478Author: Michel Dänzer <michel@tungstengraphics.com> 29479Date: Sat Jun 24 16:01:19 2006 +0200 29480 29481 Bug #7274: Fix corruption with 'small' accelerated DownloadFromScreen transfers. 29482 29483commit f25ba3dc0e7850f0b1b78f54c0e7211bf2e81b3a 29484Author: Tilman Sauerbeck <tilman@code-monkey.de> 29485Date: Fri Jun 16 11:38:18 2006 +0200 29486 29487 Remove obsolete FIXME. 29488 29489commit 764013ec61b9043f1487513658e374569f7154a9 29490Author: Tilman Sauerbeck <tilman@code-monkey.de> 29491Date: Fri Jun 16 10:56:02 2006 +0200 29492 29493 Use AC_CHECK_HEADER() to check for exa.h 29494 29495commit 3c701d49b3eb7adb82267cec53163e475913aca3 29496Author: Tilman Sauerbeck <tilman@code-monkey.de> 29497Date: Fri Jun 16 09:59:29 2006 +0200 29498 29499 Use AC_MSG_RESULT() instead of echo for check results. 29500 29501commit dbef62615f38aebf4cdc579e3b26a7e7c719abc4 29502Author: Tilman Sauerbeck <tilman@code-monkey.de> 29503Date: Fri Jun 16 08:50:49 2006 +0200 29504 29505 conditionally declare pPCI and PciReg, too 29506 29507commit a01c01787cfbe3706ae7747cde1657abd79ecdaa 29508Author: Michel Dänzer <michel@tungstengraphics.com> 29509Date: Wed Jun 21 09:38:47 2006 +0200 29510 29511 Bug #7283: Build fixes for !defined(XF86DRI). 29512 29513 Thanks to Matthieu Herrb for pointing out some of these. 29514 29515commit 79d1c573133e64ddf4446b6bb0eaab4a5a6c73ab 29516Author: Michel Dänzer <michel@tungstengraphics.com> 29517Date: Sun Jun 18 17:39:09 2006 +0200 29518 29519 Implement accelerated EXA DownloadFromScreen hook. 29520 29521 x11perf -getimage numbers only increase by about 20-30% on my PowerBook with an 29522 M9, but by about 100 times(!) with a PCIe X550. I suspect the former could 29523 perform better with PCI as opposed to AGP transfers, which would also remove 29524 the need to disable this by default with AGP. 29525 29526commit 2bcdc7cfd013bd457d35f927390e5cc3544f30f1 29527Author: Michel Dänzer <michel@tungstengraphics.com> 29528Date: Sun Jun 18 16:01:00 2006 +0200 29529 29530 Remove #include <fbdevhw.h>, not used here. 29531 29532commit 25a5ccb63a9e3074b4c8eaf48844877d84ff76bb 29533Author: Michel Dänzer <michel@tungstengraphics.com> 29534Date: Sun Jun 18 15:26:55 2006 +0200 29535 29536 Prefer visuals with stencil when it's hardware acclerated. 29537 29538 Also make the formatting of the if statements consistent. 29539 29540commit 8fcc436f19f9c1371295af2d0c45608fb43c584b 29541Author: Michel Dänzer <michel@tungstengraphics.com> 29542Date: Sun Jun 18 15:16:08 2006 +0200 29543 29544 Take part of the blame for the radeon driver. 29545 29546commit c8366d7b457565a06acdf6b40c5448c7f13f1641 29547Author: Michel Dänzer <michel@tungstengraphics.com> 29548Date: Sun Jun 18 14:24:14 2006 +0200 29549 29550 Add Option "DepthBits". 29551 29552 This allows getting higher depth buffer precision and stencil hardware 29553 acceleration in depth 16 or saving some video RAM at the cost of these features 29554 in depth 24. 29555 29556commit 23a21d00db6f3b88b54c44648f968018a03677d3 29557Author: Tilman Sauerbeck <tilman@code-monkey.de> 29558Date: Fri Jun 16 16:37:07 2006 +1000 29559 29560 [PATCH] move the declaration of pMach64 into the correct ifndef statement 29561 Signed-off-by: Tilman Sauerbeck <tilman@code-monkey.de> 29562 Signed-off-by: Dave Airlie <airlied@linux.ie> 29563 29564commit c6b33db9d2c751e34c2152b0bb7e0e84a812589f 29565Author: Dave Airlie <airlied@linux.ie> 29566Date: Fri Jun 16 16:22:03 2006 +1000 29567 29568 ati: prepare for release 29569 29570 Update some versions for release 29571 29572commit 9c534b5330534fc6cf392936a9b200a5128c787c 29573Author: Michel Dänzer <michel@tungstengraphics.com> 29574Date: Thu Jun 15 16:09:55 2006 +0200 29575 29576 Some more ignores. 29577 29578commit 3d68698647fb17a6f5ae9a225f552cb0130fa946 29579Author: Michel Dänzer <michel@tungstengraphics.com> 29580Date: Thu Jun 15 16:07:15 2006 +0200 29581 29582 Fix semantic glitches pointed out by gcc -Wall. 29583 29584 The radeon driver builds warning-free with gcc -Wall here now. 29585 29586commit ab2b48e0df328fd08a2412911126b3184a601aea 29587Author: Michel Dänzer <michel@tungstengraphics.com> 29588Date: Thu Jun 15 16:06:13 2006 +0200 29589 29590 Remove another unused variable. 29591 29592commit 56bd5a370472663af5607cd327f757367b238db0 29593Author: Michel Dänzer <michel@tungstengraphics.com> 29594Date: Thu Jun 15 15:35:05 2006 +0200 29595 29596 Fix some more cosmetic warnings. 29597 29598commit ec93e08b24f7c0e583970d08bfe18ad142e6b047 29599Merge: 051c49d0f 6ca986242 29600Author: Michel Dänzer <michel@tungstengraphics.com> 29601Date: Thu Jun 15 15:03:27 2006 +0200 29602 29603 Merge branch 'master' of ssh+git://git.freedesktop.org/git/xorg/driver/xf86-video-ati 29604 29605commit 6ca986242c80bcfe7ca2d435a772bbdf034892c4 29606Author: Dave Airlie <airlied@linux.ie> 29607Date: Thu Jun 15 21:24:46 2006 +1000 29608 29609 remove unused variable 29610 29611commit 9a5f5bc0d3f5e62b79355993ca6246382394daee 29612Author: Dave Airlie <airlied@linux.ie> 29613Date: Thu Jun 15 21:24:16 2006 +1000 29614 29615 remove all printf specifier warnings on Linux 29616 29617commit 051c49d0f002c72d1d8f15f6df264d8b093d4e95 29618Merge: a840afc75 dfac8191d 29619Author: Michel Dänzer <michel@tungstengraphics.com> 29620Date: Thu Jun 15 13:10:44 2006 +0200 29621 29622 Merge branch 'master' of ssh+git://git.freedesktop.org/git/xorg/driver/xf86-video-ati 29623 29624commit dfac8191dbda29ef7f6ce33d7356fee89c17d59b 29625Author: Dave Airlie <airlied@linux.ie> 29626Date: Thu Jun 15 21:05:16 2006 +1000 29627 29628 fix last ati*.[ch] format specifiers 29629 29630commit 33525540f3a9d6760c6edcfe765d2a004a6b894f 29631Author: Dave Airlie <airlied@linux.ie> 29632Date: Thu Jun 15 21:03:01 2006 +1000 29633 29634 fix more printf formats 29635 29636commit 6bcf6f29af24b6725e097dd72af8e8b9f6d0676c 29637Author: Dave Airlie <airlied@linux.ie> 29638Date: Thu Jun 15 20:59:33 2006 +1000 29639 29640 some more unused variable cleanups 29641 29642commit a840afc75a0453b4694e3f987ebdbddc6844a169 29643Merge: c90c49da3 d511ef486 29644Author: Michel Dänzer <michel@tungstengraphics.com> 29645Date: Thu Jun 15 12:46:58 2006 +0200 29646 29647 Merge branch 'master' of ssh+git://git.freedesktop.org/git/xorg/driver/xf86-video-ati 29648 29649commit d511ef486a15b582e5d1708a17a00701f5cf1ccd 29650Author: Dave Airlie <airlied@linux.ie> 29651Date: Thu Jun 15 20:39:22 2006 +1000 29652 29653 ati/r128: add includes to get C function decls 29654 29655 Also fix a void pointer cast 29656 29657commit f46c3bebe821ee0031770d98d04bd406d1abb9ed 29658Author: Dave Airlie <airlied@linux.ie> 29659Date: Thu Jun 15 20:34:56 2006 +1000 29660 29661 fix some more unused variables and bad parameter passing 29662 29663commit 6c3f81683abab8ff6b8f2fcaf01a2a1a63af5ef6 29664Author: Dave Airlie <airlied@linux.ie> 29665Date: Thu Jun 15 20:25:00 2006 +1000 29666 29667 ati: fix -Wall in atidri.c 29668 29669 atidri.c has some parameter passing bugs, fixed those and cleaned other 29670 warnings. 29671 29672commit 58c6aac0669101481ef6b8217e5cede4573ece42 29673Author: Dave Airlie <airlied@linux.ie> 29674Date: Thu Jun 15 19:58:11 2006 +1000 29675 29676 theatre: fix all -Wall warnings in theatre* 29677 29678 This fixes all the warnings, it doesn't clean up the horrible mess that is 29679 theatre*.[ch]. 29680 29681commit c90c49da3844f93d2d29070b15bb7def72b98b0f 29682Author: Michel Dänzer <michel@tungstengraphics.com> 29683Date: Thu Jun 15 10:29:46 2006 +0200 29684 29685 Collapse .gitignore files into single toplevel file, sorted with LC_ALL=C. 29686 29687 Suggested by Keith Packard. 29688 29689commit b7a3e4c9b33f72ae447f99c0c765f24fce988142 29690Author: Michel Dänzer <michel@tungstengraphics.com> 29691Date: Thu Jun 15 09:45:12 2006 +0200 29692 29693 Fix some implicit function declarations. 29694 29695commit 556ddc9c41020271e92389a3900d097ec443c625 29696Author: Michel Dänzer <michel@tungstengraphics.com> 29697Date: Thu Jun 15 09:30:43 2006 +0200 29698 29699 .cvsignore -> .gitignore 29700 29701commit 4708327969a21fc171dcd80a1b506028cc8e4f90 29702Author: Dave Airlie <airlied@linux.ie> 29703Date: Thu Jun 15 07:37:03 2006 +1000 29704 29705 ati: remove Changelog from ATI driver. 29706 29707commit ecbec8379c24bd6c16ced674d5e6affcd3985e46 29708Author: Michel Daenzer <michel@daenzer.net> 29709Date: Thu Jun 8 17:29:12 2006 +0000 29710 29711 Bug #5678: Fix setting of transfer mode for AGPv3 devices (Tilman 29712 Sauerbeck). 29713 29714commit 874ff105bc63c4fc242fbabaec04fde2cb81fdd4 29715Author: Daniel Stone <daniel@fooishbar.org> 29716Date: Sat Jun 3 10:50:33 2006 +0000 29717 29718 Bug #1889: Stop the CCE before an engine reset. (Conn) 29719 29720commit f9db5595fe4d62c54b3a501b2af64b4e2b7035cd 29721Author: Michel Daenzer <michel@daenzer.net> 29722Date: Sun May 21 14:33:34 2006 +0000 29723 29724 Derive ATI_VERSION_{MAJOR,MINOR,PATCH} from package version in order to 29725 hopefully confuse bug triagers a little less. 29726 29727commit 25ba85cab8b3d831da646fef19f01769410fc958 29728Author: Alex Deucher <agd5f@yahoo.com> 29729Date: Wed May 10 02:46:37 2006 +0000 29730 29731 - remove obsolete option "OverlayOnCRT2" (bug 6488) 29732 29733commit ede75fe2b1850a4d9f03c724b7c1d1ed65e60f2f 29734Author: Michel Daenzer <michel@daenzer.net> 29735Date: Tue May 9 14:40:23 2006 +0000 29736 29737 Drop support for radeon DRM <= 1.7. The 2.4 kernel has version 1.7, and 29738 various problems up to system hangs on X server startup have been 29739 reported with that and DRI enabled. 29740 29741commit bc93a46c4fe499f23b9c0ff52e2e36b4c587f7d6 29742Author: Michel Daenzer <michel@daenzer.net> 29743Date: Mon May 8 09:29:34 2006 +0000 29744 29745 Bugzilla #4640 <https://bugs.freedesktop.org/show_bug.cgi?id=4640> Patch 29746 #5273 <https://bugs.freedesktop.org/attachment.cgi?id=5273>: 29747 Add cases for FOURCC_RGBA32 and FOURCC_RGB24. This prevents clients from 29748 receiving incorrect data sizes with these formats. (Jan Schmidt) 29749 29750commit 2b0cdd9448a24ea067b0d78f319b99c1041df2e0 29751Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 29752Date: Wed May 3 04:25:27 2006 +0000 29753 29754 Enable decoding of version 4 TMDS PLL config tables, they actually work, 29755 the problem in the comment only concerned pre-production cards that Hui 29756 tested apparently. Also set back the default value to what it was 29757 initially, fixes some more cards that don't have a proper table (Mac 29758 Mini for example) 29759 29760commit 36c04b88416905b9ae305583e8e34fefdbfafc50 29761Author: Michel Daenzer <michel@daenzer.net> 29762Date: Mon May 1 17:49:41 2006 +0000 29763 29764 Bugzilla #6755 <https://bugs.freedesktop.org/show_bug.cgi?id=6755> Patch 29765 #5536 <https://bugs.freedesktop.org/attachment.cgi?id=5536>: 29766 Change HostDataBlit interface to take dst_offset_pitch and coordinates 29767 instead of just a destination pointer, as the latter is not sufficient 29768 with tiling. Also, use HW clipping to avoid overwriting destination 29769 data outside of the specified width. 29770 Adapt to new HostDataBlit interface. This fixes corruption with 29771 UploadToScreen to the front buffer (from exaPutImage). 29772 29773commit b90d3710e67caad0686efab491c2b76ae877920c 29774Author: Michel Daenzer <michel@daenzer.net> 29775Date: Sat Apr 29 21:30:23 2006 +0000 29776 29777 Bugzilla #1992 <https://bugs.freedesktop.org/show_bug.cgi?id=1992> Patch 29778 #4988 <https://bugs.freedesktop.org/attachment.cgi?id=4988>: Fix Option 29779 "DisplayPriority" "HIGH" for R300 family. 29780 29781commit 30f3ce5c5ba205a23748def11de019ee727897f3 29782Author: Michel Daenzer <michel@daenzer.net> 29783Date: Sat Apr 29 20:34:57 2006 +0000 29784 29785 Properly de-initialize EXA driver data, so EXA can be initialized again in 29786 the next generation. 29787 29788commit d705429c230c9cc1645ae540a51dba62cae9f23e 29789Author: Michel Daenzer <michel@daenzer.net> 29790Date: Sat Apr 29 20:18:48 2006 +0000 29791 29792 Fix tests for (un)forcing HW cursor. The broken tests caused problems when 29793 switching to console with 3D clients running. 29794 29795commit 6142cb079d4a5844e4863c800baf4ae19c636af5 29796Author: Michel Daenzer <michel@daenzer.net> 29797Date: Sat Apr 29 20:12:38 2006 +0000 29798 29799 Prefer double-buffered visuals. As a consequence, the root window's visual 29800 is double-buffered by default, which is desirable in a number of 29801 situations, in particular when running a GLX compositing manager with 29802 AIGLX. 29803 29804commit eb841974d041b19461b7fb726687df8938b2e39b 29805Author: Eric Anholt <anholt@freebsd.org> 29806Date: Fri Apr 28 02:55:08 2006 +0000 29807 29808 Bug #6761: Fix font rendering in non-DRI mode by replacing RB2D_DSTCACHE_* 29809 with RB3D_DSTCACHE_*. The RB2D versions are read-only mirrors of the 29810 RB3D. This is masked when DRI is enabled because the DRM uses the right 29811 registers in its flushing. 29812 29813commit 85e2ee6de6fd347aad66785ec7119f92013cd994 29814Author: Dave Airlie <airlied@linux.ie> 29815Date: Wed Apr 26 08:38:47 2006 +0000 29816 29817 Make the ATI driver use the PCI BAR size to determine the MMIO register 29818 aperture size. Otherwise it tries to overmap things if they are up near 29819 the end of memory. 29820 29821commit 13b19edf5accb7427436d3f0b1c9a9efb4fe98fa 29822Author: Eric Anholt <anholt@freebsd.org> 29823Date: Wed Apr 26 07:26:15 2006 +0000 29824 29825 Add support for some component-alpha operations (InReverse, OutReverse) 29826 necessary for testing upcoming general component alpha acceleration 29827 assistance in EXA. Tested with rendercheck on an r100. 29828 29829commit 44a45bcb82d49349f4fdfae3e4b8f24f98d99b46 29830Author: Eric Anholt <anholt@freebsd.org> 29831Date: Tue Apr 25 23:31:56 2006 +0000 29832 29833 Use a CPPFLAGS with XORG_CFLAGS so that the exa test can actually find its 29834 headers and not fail. Also correct some nearby spelling. 29835 29836commit cc1825b503ce7e785cbb4611fd177477a1081b6e 29837Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 29838Date: Mon Apr 24 07:44:52 2006 +0000 29839 29840 Fix enabling of external VGA output on some cards with dual entity setup 29841 instead of MergedFB. This was broken due to bogus manipulations of a 29842 register between the 2 entities. This fix is a big "violent" in that it 29843 always enables the analog output regardless of what is plugged. It 29844 should work though (at the expense of some additional power 29845 consumption). A proper fix should be easy to do once Alex gets his 29846 output mapping rework in. 29847 29848commit a66145bc9d487f7e19237cf292f12750cb959f61 29849Author: Dave Airlie <airlied@linux.ie> 29850Date: Fri Apr 21 00:38:44 2006 +0000 29851 29852 Bugzilla 6672: fix interrupts on radeon since memmap changes 29853 29854commit f1f057df9aaba42986d9e20afe6f609a90318333 29855Author: Dave Airlie <airlied@linux.ie> 29856Date: Fri Apr 21 00:34:43 2006 +0000 29857 29858 Add support for backing up the PCIE GART table on VT switch. This makes 29859 suspend/resume work a lot better on PCIE cards. 29860 29861commit 31396df3b0101571bb85a4a66b67f5a8b0c992a9 29862Author: Adam Jackson <ajax@nwnk.net> 29863Date: Sat Apr 8 01:19:17 2006 +0000 29864 29865 Bump to 6.6.0 and release. This module's versioning is completely screwed, 29866 so until we split out mach64/r128/radeon please just increment the 29867 package number whenever you make significant changes. 29868 29869commit d706f494ed22afd369991e8c75a2d4f5aed15305 29870Author: Aaron Plattner <aplattner@nvidia.com> 29871Date: Fri Apr 7 23:06:02 2006 +0000 29872 29873 Add a DrawablePtr argument to the XV functions to pave the way for 29874 redirected video. 29875 29876commit f94685aec26b65f60c00a4b1bcd8f26bd80eb6e4 29877Author: Eric Anholt <anholt@freebsd.org> 29878Date: Tue Apr 4 07:42:36 2006 +0000 29879 29880 Bug #5632: Fix up ati.4 to explain that it is a wrapper driver for the 29881 other three. While here, touch up radeon.4, and add an authors list 29882 gleaned from checking copyright headers. 29883 29884commit 58c9b1d84f514755945e7a0d25fd5530af34ce95 29885Author: Eric Anholt <anholt@freebsd.org> 29886Date: Sun Apr 2 21:45:12 2006 +0000 29887 29888 Add missing .cvsignore lines. 29889 29890commit f1ce6170ef21e1f9769337c3138cd8d65c33e40a 29891Author: Alex Deucher <agd5f@yahoo.com> 29892Date: Sat Apr 1 23:02:40 2006 +0000 29893 29894 - Fix dpi when switching from clone to dualhead with MergedFB. 29895 - Add ConstantDPI option to force a particlar dpi across mode changes Both 29896 based on Thomas Winischhofer's sis code. 29897 29898commit 607f18cfbfc2a7b5509f8b567bf510fce31b361e 29899Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 29900Date: Mon Mar 27 06:12:57 2006 +0000 29901 29902 Fix a regression of my latest fix that broke color tiling on some setups. 29903 RADEONRestoreMemMapRegisters() must not clear CRTC_OFFSET_CNTL and 29904 friends unconditionally but only when there is a map change or it will 29905 clear them after the DRI updates the AGP location, thus turning tiling 29906 back off on the CRTC. Also remove a duplicate call to 29907 RADEONInitCommonRegisters(). 29908 29909commit bf1c4471137f10b7de9d79519ca2240058997e18 29910Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 29911Date: Thu Mar 23 18:04:17 2006 +0000 29912 29913 commit the actual changes instead of just the Changelog... 29914 29915commit aa5fff9bae4818bbf99e8d9caa8f2ce9905294ca 29916Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 29917Date: Thu Mar 23 17:59:25 2006 +0000 29918 29919 Use programmable tap coefficients for (hopefully) better video image 29920 quality when downscaling. Remove unnecessary code programming the tap 29921 coefficients twice. Disable some code which is supposed to improve 29922 filtering quality but causes very visible image quality problems at 29923 least with rv250 when upscaling. Fix potentially broken FOURCC_RGB16T 29924 format (untested if it was broken, untested if it works now...). Change 29925 some magic numbers to (new) macro defines. 29926 29927commit 17c52ff15ebe2cf220d5988ee13b599c41799886 29928Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 29929Date: Thu Mar 23 01:37:15 2006 +0000 29930 29931 Call RADEONRestoreCommonRegisters() before RADEONRestoreCrtc2Registers() 29932 instead of after. I can't find any good reason why we did it backward 29933 until now and it definitely causes problems as the workaround for VT 29934 switch in RADEONRestoreCommonRegisters() will break output to the 29935 secondary head in various situations. Also does an unrelated minor 29936 update to a comment. 29937 29938commit 673e4d7501861066fdd0874919a4524b820b944b 29939Author: Kristian Høgsberg <krh@redhat.com> 29940Date: Thu Mar 23 00:06:56 2006 +0000 29941 29942 Add atipciids.h. 29943 29944commit 5dcb4152dca6ac1fe9342f44eb48c0198b15772e 29945Author: Kristian Høgsberg <krh@redhat.com> 29946Date: Wed Mar 22 22:30:14 2006 +0000 29947 29948 Convert use of xf86fopen() and other xf86 wrapped libc symbols to use libc 29949 symbols directly. The xf86* versions aren't supposed to be used 29950 directly. 29951 Drop libc wrapper; don't include xf86_ansic.h and add includes now missing. 29952 29953commit 21acdf371e3cf913c5affbd3e86641cfb115311c 29954Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 29955Date: Sat Mar 18 00:08:24 2006 +0000 29956 29957 Clear the offsets when updating the memory map instead of when enabling the 29958 CRTCs and try to make that code more reliable. Doesn't disable the 29959 CRTCs beforehand. 29960 29961commit 323ecb92e40d71c5ef994b41b6d8dedba6dd6203 29962Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 29963Date: Fri Mar 17 03:00:53 2006 +0000 29964 29965 Fix various small cosmetic issues. Change a driver message, get the order 29966 right for requesting drm versions, replace the use of some numbers with 29967 the respective macro defines in radeon_video.c, and add some more macro 29968 defines. None of that really matters. 29969 29970commit dc552edf191e14249421f8b27146874f16b80c1a 29971Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 29972Date: Thu Mar 16 21:53:58 2006 +0000 29973 29974 Add missing return statement (the driver would not start randomly), add 29975 RV410 to the list of new generation cards for the memory map setup and 29976 fix a bogus message. 29977 29978commit 901409f47cc990721c1648b1c08e23450e6a4452 29979Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 29980Date: Thu Mar 16 04:32:22 2006 +0000 29981 29982 Fix incorrect CRTC2_OFFSET values when using old-style dual head (not 29983 MergedFB). Fixes Xinerama for me. 29984 29985commit 4d6656062129da0489eb4bc898871379ba891d8f 29986Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 29987Date: Thu Mar 16 03:09:34 2006 +0000 29988 29989 Add various workarounds that seem to fix some remaning lockup scenarios I'm 29990 experiencing with the driver when setting the memory map. Some of the 29991 magic delays are a bit dodgy but they seem to work, I suppose I can't 29992 do better now without help from ATI. Also removed some really too noisy 29993 debug messages. 29994 29995commit 36799e3ea7ddde5fce528e8bc17ea942933a9111 29996Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 29997Date: Wed Mar 15 04:03:37 2006 +0000 29998 29999 Fix various issues with DRI & server recycle by putting DRICloseSreen in 30000 the normal CloseScreen callback chain at a spot that matches the call 30001 to DRIFinishScreenInit in order to unwrap at the right time. 30002 30003commit 50923b6a9325592edc8b02b17646f96b9b8c4a01 30004Author: Luc Verhaegen <libv@skynet.be> 30005Date: Wed Mar 15 00:46:33 2006 +0000 30006 30007 Add some autoconf to keep this driver from building EXA when a slightly 30008 older server with older exa is installed. 30009 The way in which this problem was treated by the actually related 30010 developers explains a lot about this driver. 30011 30012commit 74f4ed4275866475b3c182fae962df2fc0946b85 30013Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 30014Date: Mon Mar 13 02:57:17 2006 +0000 30015 30016 Fix typo in the function setting the color space transformation factors for 30017 old radeon 30018 30019commit 65625c6c2053792cd14b54c904bf46f360ef1565 30020Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 30021Date: Mon Mar 13 00:09:44 2006 +0000 30022 30023 Earliest fix introduced a bug due to another chicken & egg problem in card 30024 detection. Now split PreInitConfig into PreInitChipType and PreInitVRAM 30025 so that PreInitDRI can be done just in between. Restores proper AGPx4 30026 functionality (when enabled) 30027 30028commit 3a290115af41b0ecb94eb201d7e7abb42ebbfec6 30029Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 30030Date: Sun Mar 12 06:53:27 2006 +0000 30031 30032 Stop using xf86PciInfo.h, instead use a local copy of the PCI IDs we need 30033 in atipciids.h so we can update the ATI driver independently of the 30034 server when new chips are added 30035 30036commit d6b8f0a5eeb332563ec3ec7082f03be1797d26d8 30037Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30038Date: Sat Mar 11 04:19:47 2006 +0000 30039 30040 Add #ifdef XF86DRI guards around references to info->allowPageFlip and 30041 info->depthMoves, since they are only defined if building with DRI. 30042 30043commit 69fe90bc501df9bd9e74b894bfca79a4025725bd 30044Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 30045Date: Fri Mar 10 13:00:49 2006 +0000 30046 30047 Use "FBTexPercent" option for XAA too (only for increasing memory reserved 30048 for textures). 30049 30050commit 8033d7423267fc1b3725df6a077dd910b586daaf 30051Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 30052Date: Thu Mar 9 23:26:27 2006 +0000 30053 30054 More memory map fixes !!! This time, we add a way to get the DRM kernel 30055 module version early, in order to properly limit the framebuffer size 30056 when a module that doesn't support the new map is loaded. That will 30057 improve backward compatibility. I also made the minimum required kernel 30058 module version 1.3 which hopefully shouldn't be a problem for anybody, 30059 and cleaned up the ordering in which we test for DRM version to enable 30060 features 30061 30062commit c637939d3c4ce5cdddbc5200d5ffbca4c3927050 30063Author: Michel Daenzer <michel@daenzer.net> 30064Date: Thu Mar 9 15:41:16 2006 +0000 30065 30066 RN50: Skip modes that exceed memory bandwidth. 30067 Bugzilla #5766 <https://bugs.freedesktop.org/show_bug.cgi?id=5766> Patch 30068 #4636 <https://bugs.freedesktop.org/attachment.cgi?id=4636> 30069 - Acknowledge that RN50 only has one CRTC, and use this to distinguish it 30070 from RV100. 30071 - Fix detection of RN50 memory type and bus width. 30072 - Model RN50 memory bandwidth limits by capping the pixel clock range based 30073 on memory clock, bpp and memory bus width. (ATI Technologies Inc.) 30074 30075commit 890b95a3ddad7634f3aea635cb236b221e902acc 30076Author: Eric Anholt <anholt@freebsd.org> 30077Date: Thu Mar 9 06:06:25 2006 +0000 30078 30079 Update drivers that support EXA for the new EXA ABI. This consists of 30080 moving all the accel and card members into the driver ptr, filling in 30081 the exa_major/ exa_minor fields, and always using LoadSubModule so we 30082 can check the module version up front. Only tested on ATI. 30083 30084commit e19ccfa18fee046c54d7a9401e4e0098b2300598 30085Author: Michel Daenzer <michel@daenzer.net> 30086Date: Fri Mar 3 16:41:41 2006 +0000 30087 30088 Add Option "FBTexPercent" to override the amount of video RAM reserved for 30089 OpenGL textures with EXA. 30090 30091commit ca00035eccb6e3495b27f8e64f2d016d8d2a9ce2 30092Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 30093Date: Thu Mar 2 02:05:17 2006 +0000 30094 30095 Fix the fix ... I didn't interpret PciInfo->size properly and forgot that 30096 the function works in Kb not bytes... Ooops. 30097 30098commit b2145aea36bb035bff048366c607b967d70fff49 30099Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 30100Date: Wed Mar 1 21:35:14 2006 +0000 30101 30102 Extend the alignement workaround to post-rv280 chips as well (thanks Hui) 30103 and also limit the CPU accessible memory to the size of the PCI BAR 30104 size (yeah, it was passed to the driver, I just got blind for a while 30105 it seems) 30106 30107commit 163581a82fe9395672c31b41cd67aa44158f5201 30108Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 30109Date: Tue Feb 28 23:34:03 2006 +0000 30110 30111 Workaround for rv280 bug which needs the framebuffer memory mapping to be 30112 aligned to its own size. Also add back limit of 128Mb of mapped fb 30113 space for now until I fully sort out what's up with dual function 30114 cards. I would really need to access the PCI BAR sizes here but that 30115 information seem to be missing from the PCI info passed to the driver. 30116 30117commit 5ff8fb2bea1fe7d14db53c19ff3a3a56668ec417 30118Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 30119Date: Tue Feb 28 00:52:53 2006 +0000 30120 30121 Add the newly added (and tons of previously forgotten ones) pci ids to the 30122 ati wrapper too (probably not required for it to work as everything 30123 should default to radeon anyway there). 30124 30125commit 1574b41a0f894f9b6e30d3ee94e596695a4c64f0 30126Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 30127Date: Sun Feb 26 10:01:32 2006 +0000 30128 30129 Fix page flipping with XAA. The new code ended up calling ShadowFBInit() 30130 before XAAInit(), which breaks ShadowFB. 30131 30132commit 1f43a584a684af8078631425886cce236500e0e6 30133Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 30134Date: Sat Feb 25 01:57:05 2006 +0000 30135 30136 Add pci ids known to exist (see #4284 for instance). There are still 30137 entries which probably don't really exist (cancelled cards and such), 30138 leave them as-is. Fix the name of some entries, mostly based on the 30139 .inf file of the newest catalyst driver. Use own family id for rv410 30140 and rs400, though there is no different code (yet?). 30141 30142commit 5c141bb15d1163e04c012a0cdf0699d534f0be37 30143Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 30144Date: Thu Feb 16 23:27:44 2006 +0000 30145 30146 Memory map and misc fixes. This completely reworks the way the driver 30147 detects available & accessible video memory and initializes the card 30148 internal memory map. Along the way, I re-ordered some of the stuffs in 30149 pre-init in order to remove some dodgy codepath, fixed various init, 30150 exit and vt switch issues, and overall made the driver more robust. The 30151 full memory map fix requires an updated DRM. The X driver no longer 30152 tries to position the AGP aperture, this becomes entirely DRM 30153 responsibility. 30154 30155commit 8c6e5a6eb50d371fd4ec700c26aa628cb5f45731 30156Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 30157Date: Wed Feb 15 23:36:55 2006 +0000 30158 30159 The engine setup would trigger a bogus line write before the proper 30160 addresses are setup, thus causing the card to try to bus master over 30161 the system bus to some stale location (usually 0 but could depend on 30162 whatever was used before X). Remove that useless init bit, and always 30163 setup some good enough engine source & destination pointers in case we 30164 have some other spurrious engine activity going on. 30165 30166commit d4ec0223029d8fe03a836a5a6e9a7cc8897b4c90 30167Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 30168Date: Mon Feb 13 05:30:13 2006 +0000 30169 30170 Fix EXA acceleration of textures with byteswap or tiling From John Clemens 30171 <john@deater.net> 30172 30173commit 79c60ebbc34aaf0ac0b0723898d5b63015547178 30174Author: Eric Anholt <anholt@freebsd.org> 30175Date: Tue Feb 7 04:39:59 2006 +0000 30176 30177 Fix a copy'n'paste-o in the EXA render implementation that resulted in 30178 [ax]8b8g8r8 pict formats not being supported on r200, but no other 30179 issues. 30180 30181commit 8d74f3d49e73626f5d658ba6586a7c1487dae141 30182Author: Daniel Stone <daniel@fooishbar.org> 30183Date: Thu Jan 19 14:20:37 2006 +0000 30184 30185 Bug #5656: Set all Radeon XPRESS 200 R[CS]4xx-based chips to have IsIGP = 30186 1, fixing modesetting issues. 30187 30188commit e490f4b86e6bba70f7c884fca3cf41d159323689 30189Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30190Date: Fri Jan 13 01:37:43 2006 +0000 30191 30192 XF86_VERSION_CURRENT -> XORG_VERSION_CURRENT 30193 30194commit ac296d9137b02753bc79a9930d168855adea9c95 30195Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30196Date: Fri Jan 13 01:02:01 2006 +0000 30197 30198 When failing to parse MetaModes, reset CRT2pScrn to NULL after free'ing it 30199 so we don't crash when trying to use it later (such as in 30200 RADEONFreeRec). 30201 30202commit dd98ce78407c9eddd7adaad1d73d6910e6949b87 30203Author: Adam Jackson <ajax@nwnk.net> 30204Date: Sun Jan 8 19:02:16 2006 +0000 30205 30206 Bug #5523: Make radeon compile without USE_XAA. (Tilman Sauerbeck) 30207 30208commit a99429b2bc6407ff2a094e6ee30ce2032f734966 30209Author: Kevin E Martin <kem@kem.org> 30210Date: Wed Dec 21 02:29:59 2005 +0000 30211 30212 Update package version for X11R7 release. 30213 30214commit 96e225b950c74e64f3f788be8a8489c5075ea55a 30215Author: Adam Jackson <ajax@nwnk.net> 30216Date: Mon Dec 19 16:25:52 2005 +0000 30217 30218 Stub COPYING files 30219 30220commit 6bea785ea0bdf01673a6ddc45765ae7372fdd371 30221Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30222Date: Mon Dec 19 09:07:50 2005 +0000 30223 30224 Typo fixes, mailing list & url updates, and other changes to prepare for 30225 X11R6.9 & 7.0 releases. 30226 30227commit 07d5cf8fc37eb72dbb344081ff07d59dbf568e9c 30228Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30229Date: Thu Dec 15 23:15:37 2005 +0000 30230 30231 Bugzilla #5238 <https://bugs.freedesktop.org/show_bug.cgi?id=5238> Patch 30232 #3978 <https://bugs.freedesktop.org/attachment.cgi?id=3978> r128 man 30233 page dualhead note update (Alex Deucher) 30234 30235commit 6db5cbfbf6d405d9871892d1255dd7442e4539e9 30236Author: Kevin E Martin <kem@kem.org> 30237Date: Thu Dec 15 00:21:32 2005 +0000 30238 30239 Update package version number for final X11R7 release candidate. Bump 30240 driver version number. 30241 30242commit 726fd3a81dfbd1b0c8e4ede104a48523daecab92 30243Author: Adam Jackson <ajax@nwnk.net> 30244Date: Wed Dec 14 19:41:28 2005 +0000 30245 30246 Bug #1760: Fix ati wrapper multihead to be dlloader friendly. 30247 30248commit 328a2d74dabe32ce35e879799412666feedb0eef 30249Author: Adam Jackson <ajax@nwnk.net> 30250Date: Fri Dec 9 18:25:02 2005 +0000 30251 30252 Bug #3869: Don\'t walk off the end of an array in connector detection. 30253 (Eric Anholt) 30254 30255commit 03961a05438e916b6725f4e547f23a84ead13c5e 30256Author: Adam Jackson <ajax@nwnk.net> 30257Date: Fri Dec 9 18:15:31 2005 +0000 30258 30259 Bug #1106: Fix memory allocation for Render. (Michel Daenzer) 30260 30261commit 091e5c584124f9b849c0f46802b2f0d8b0cc2388 30262Author: Kevin E Martin <kem@kem.org> 30263Date: Thu Dec 8 17:55:18 2005 +0000 30264 30265 Add configure options to allow hard-coded paths to be changed. 30266 30267commit e38e2df15c9a4b08dee6386a225e98827bb35aad 30268Author: Kevin E Martin <kem@kem.org> 30269Date: Thu Dec 8 17:54:40 2005 +0000 30270 30271 Allow hard-coded paths to be configurable. 30272 30273commit b8203604bf9bd3813668313a1c8c17040b9d93ce 30274Author: Kevin E Martin <kem@kem.org> 30275Date: Tue Dec 6 22:48:33 2005 +0000 30276 30277 Change *man_SOURCES ==> *man_PRE to fix autotools warnings. 30278 30279commit a86920e812bfd40ec9facfc234220e83066dd263 30280Author: Kevin E Martin <kem@kem.org> 30281Date: Sat Dec 3 05:49:32 2005 +0000 30282 30283 Update package version number for X11R7 RC3 release. 30284 30285commit 010adc8988846e6f76f111a9fd32eca021d12fbc 30286Author: Kevin E Martin <kem@kem.org> 30287Date: Sat Dec 3 05:45:17 2005 +0000 30288 30289 Bump version numbers for X11R6.9/X11R7 RC3. 30290 30291commit a5efa6724e7fd8318e262060b084630632dfcb44 30292Author: Kevin E Martin <kem@kem.org> 30293Date: Fri Dec 2 02:16:07 2005 +0000 30294 30295 Remove extraneous AC_MSG_RESULT. 30296 30297commit feeef3cba1f4736a96daf3002ec902ced2874198 30298Author: Adam Jackson <ajax@nwnk.net> 30299Date: Wed Nov 30 18:51:22 2005 +0000 30300 30301 Bump libdrm dep to 2.0. 30302 30303commit 7ca2b3fa30fedd07951f7c6e73ed4d2a6f6606ea 30304Author: Adam Jackson <ajax@nwnk.net> 30305Date: Tue Nov 29 23:29:58 2005 +0000 30306 30307 Only build dlloader modules by default. 30308 30309commit e167e5151e32d54919e96cbfdcf932e16b1cab2a 30310Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30311Date: Mon Nov 28 22:04:08 2005 +0000 30312 30313 Change *mandir targets to use new *_MAN_DIR variables set by xorg-macros.m4 30314 update to fix bug #5167 (Linux prefers *.1x man pages in man1 subdir) 30315 30316commit 0015b1a007bed61aaf389f8f927e4ae0568d7da7 30317Author: Eric Anholt <anholt@freebsd.org> 30318Date: Mon Nov 21 10:49:09 2005 +0000 30319 30320 Add .cvsignores for drivers. 30321 30322commit 00b8f6ccfbb5163c47e9def0e4ecc6ac7be7e8af 30323Author: Eric Anholt <anholt@freebsd.org> 30324Date: Fri Nov 11 11:05:42 2005 +0000 30325 30326 Bug #1028: Don't (incorrectly) mark the 24/8 z/stencil case as slow, which 30327 resulted in whining at runtime about the visual not being supported. 30328 30329commit 33326847ccc9cf4cd6a23a257e5392e348e6dd22 30330Author: Kevin E Martin <kem@kem.org> 30331Date: Wed Nov 9 21:31:20 2005 +0000 30332 30333 Bump version numbers for X11R6.9/X11R7 RC2. 30334 30335commit 68e11246bd6f6a641807230834dce4e7fd6911af 30336Author: Kevin E Martin <kem@kem.org> 30337Date: Wed Nov 9 21:15:14 2005 +0000 30338 30339 Update package version number for X11R7 RC2 release. 30340 30341commit 1582d3b34643babcbb794d2df8d7c9a755d569c9 30342Author: Matthias Hopf <mhopf@suse.de> 30343Date: Wed Nov 9 17:42:57 2005 +0000 30344 30345 Bug #4998: Add ';' and ',' as valid separators for MetaModes as well. 30346 30347commit 2d22d78d157e07e6ec6a3e2d03fc17f102f6e00e 30348Author: Kevin E Martin <kem@kem.org> 30349Date: Wed Nov 9 07:30:03 2005 +0000 30350 30351 Fix DoLoadableServer NO build. 30352 30353commit 999f67763a5ea9fef7b7b576262f0da079efcfee 30354Author: Michel Daenzer <michel@daenzer.net> 30355Date: Tue Nov 8 16:30:48 2005 +0000 30356 30357 bugzilla #988 (https://bugs.freedesktop.org/show_bug.cgi?id=988) Fix typo 30358 which may or may not have had a negative impact on stability with R300 30359 class cards. 30360 30361commit 5aa401efa0b8d81da418eafeb857142a33f85da9 30362Author: Michel Daenzer <michel@daenzer.net> 30363Date: Tue Nov 8 14:05:00 2005 +0000 30364 30365 bugzilla #4494 (https://bugs.freedesktop.org/show_bug.cgi?id=4494) Call 30366 RADEONDGAInit() even if EXA is enabled; the problematic DGA 30367 acceleration hooks are already disabled in that case. 30368 This fixes DGA issues such as DGA mouse (typically used by games in 30369 fullscreen mode) not working. 30370 Fix suggested by Eric Anholt. 30371 30372commit 345180131a7b2752773dc2a3ee1221ad1e05c815 30373Author: Michel Daenzer <michel@daenzer.net> 30374Date: Tue Nov 8 13:35:23 2005 +0000 30375 30376 bugzilla #4916 (https://bugs.freedesktop.org/show_bug.cgi?id=4916) Fix 30377 segfault with MergedFB and EXA enabled and HW cursor temporarily 30378 disabled. 30379 30380commit 23911421ffa7bdd2e191a72c41bf1f26c92c1a67 30381Author: Michel Daenzer <michel@daenzer.net> 30382Date: Tue Nov 8 09:44:41 2005 +0000 30383 30384 bugzilla #4951 (https://bugs.freedesktop.org/show_bug.cgi?id=4951) 30385 attachment #3689 (http://bugs.freedesktop.org/attachment.cgi?id=3689) 30386 Fix HW cursor getting silently disabled with EXA. 30387 30388commit e1259c0117a190cf4c6e7be07d5aa83e03afb3a0 30389Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 30390Date: Mon Nov 7 19:28:40 2005 +0000 30391 30392 Replaced the variadic macros(gcc) by macros according to C99 standard 30393 30394commit 67bdaae04ab37f38a6d0c8dc4b6328c3ba3b3393 30395Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30396Date: Sun Nov 6 07:41:29 2005 +0000 30397 30398 Sun bug #6325220: Xorg dumps core in dual monitor configuration with radeon 30399 driver if CRT2Position not specified in xorg.conf 30400 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6325220> 30401 (Stuart Krietman - Sun Microsystems) 30402 30403commit 733ebf6e4a7c7b2f1a861077d43da2a853e4c0d7 30404Author: Kevin E Martin <kem@kem.org> 30405Date: Tue Nov 1 15:08:52 2005 +0000 30406 30407 Update pkgcheck depedencies to work with separate build roots. 30408 30409commit dd0eb76337bfdaec8f456380a6ced9e0487e190f 30410Author: Adam Jackson <ajax@nwnk.net> 30411Date: Fri Oct 21 18:40:18 2005 +0000 30412 30413 Bug #4829: Remove some useless DRI symbol checks. 30414 30415commit e6c1859e3a7c84b882e0c30d0783255ed1258d9e 30416Author: Donnie Berkholz <spyderous@gentoo.org> 30417Date: Thu Oct 20 23:01:58 2005 +0000 30418 30419 Bug #4831 <https://bugs.freedesktop.org/show_bug.cgi?id=4831>. Check for 30420 x86_64 in addition to amd64 in host_cpu. 30421 30422commit 33460fb393c9591babcc43b30a73eef3971765c2 30423Author: Kevin E Martin <kem@kem.org> 30424Date: Wed Oct 19 02:48:02 2005 +0000 30425 30426 Update package version number for RC1 release. 30427 30428commit 77d5c3549db4f6a7ff9f969f885d7a4e58ab2b16 30429Author: Kevin E Martin <kem@kem.org> 30430Date: Tue Oct 18 15:46:37 2005 +0000 30431 30432 Do the following to make the drivers pass distcheck: 30433 - Only define EXTRA_DIST when there are actually man pages to be processed 30434 - Add missing man pages to keyboard and ati drivers EXTRA_DIST 30435 30436commit 35e717078c840414b0d6a0e37b3ffed0a7b102aa 30437Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30438Date: Tue Oct 18 00:01:52 2005 +0000 30439 30440 Use @DRIVER_MAN_SUFFIX@ instead of $(DRIVER_MAN_SUFFIX) in macro 30441 substitutions to work better with BSD make 30442 30443commit ff6e5b0412fe347575992c83bb7cb47b6f96f390 30444Author: Adam Jackson <ajax@nwnk.net> 30445Date: Mon Oct 17 22:57:32 2005 +0000 30446 30447 More 1.7 braindamage: define EXTRA_DIST in terms of @DRIVER_NAME@ instead 30448 of indirectly 30449 30450commit a2e7c55850d383e3af948d445b1645d8a47811cf 30451Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30452Date: Mon Oct 17 00:09:01 2005 +0000 30453 30454 Use sed & cpp to substitute variables in driver man pages 30455 30456commit 524fb1440760a14c53fb3f238aad6c3cdb7e6a08 30457Author: Dave Airlie <airlied@linux.ie> 30458Date: Sun Oct 9 10:36:50 2005 +0000 30459 30460 Secure the Radeon PCIE GART tables in framebuffer memory by not allowing 30461 the clients to get access to a secure area of framebuffer. 30462 30463commit e526211b86ec5343002978debfcf78f74a10c69e 30464Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 30465Date: Tue Oct 4 20:01:03 2005 +0000 30466 30467 Check in generated README files 30468 30469commit 431d0e144ae4c294e1f6110fc3db497ce9bd9cfd 30470Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 30471Date: Mon Oct 3 21:22:07 2005 +0000 30472 30473 - For all drivers that have a <driver>.sgml file, add code in their build 30474 system to build the README file at make dist time 30475 - in util/macros/xorg-macros.m4, add a new XORG_CHECK_LINUXDOC macro that 30476 will check if the required tools and files exist, and if so set a 30477 conditional. 30478 - util/modular/symlink.sh 30479 - Link all the <driver>.sgml to xf86-video-<driver>/README.sgml 30480 - Add all the README.<driver> to the list of excluded files 30481 - xc/programs/Xserver/hw/xfree86/doc/sgml/SiS.sgml: Various changes to make 30482 it spew less warnings when the text file is built. 30483 30484commit 9147c20e5ff662ac511a19e7623b406dffee2242 30485Author: Adam Jackson <ajax@nwnk.net> 30486Date: Mon Oct 3 06:14:58 2005 +0000 30487 30488 Bug #4001: Don't crash when PanelSize is unset. (Alan Coopersmith) 30489 30490commit 7e2090e3ebf0bd699d66d336353c097d658a86ae 30491Author: Matthieu Herrb <matthieu.herrb@laas.fr> 30492Date: Sat Oct 1 20:42:06 2005 +0000 30493 30494 missed one missing return value. 30495 30496commit a843e4281134de1545d538cb8970179b1d6977f6 30497Author: Matthieu Herrb <matthieu.herrb@laas.fr> 30498Date: Sat Oct 1 17:09:48 2005 +0000 30499 30500 add missing return value on success in RADEONSetupMemXAA(). 30501 30502commit 98f7d043f894c73fecaf14440c1a24b6c2fc4b06 30503Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 30504Date: Wed Sep 28 15:10:38 2005 +0000 30505 30506 Make ati driver distcheck 30507 radeon_commonfuncs.c \ 30508 radeon_exa_render.c \ 30509 radeon_exa_funcs.c \ 30510 30511commit e7b1af5d1677e53820dd1c0926cc8b5a4284901e 30512Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 30513Date: Tue Sep 27 23:55:15 2005 +0000 30514 30515 Move the RADEONEnablePageFlip/RADEONChangeSurfaces calls back to the end of 30516 the function (the code above will cause pixmap cache to be freed and 30517 copied around, and these functions must be called after that otherwise 30518 pixmap damage will occur). May be related to #4440, #4456 (unclear if 30519 these setups involve 3d apps) 30520 30521commit bc3c26fe365347c427e04385c1df95a5eec14f28 30522Author: Eric Anholt <anholt@freebsd.org> 30523Date: Tue Sep 27 23:34:11 2005 +0000 30524 30525 Fix Radeon MMIO Render acceleration with EXA by writing out floating- point 30526 coordinates rather than integers. 30527 30528commit 58dfb95c446b73fbc34784f0894cf89152dab043 30529Author: Adam Jackson <ajax@nwnk.net> 30530Date: Sat Sep 24 18:45:45 2005 +0000 30531 30532 Disable the {Open,Close}FullScreen DRI protocol. Remove empty FullScreen 30533 stubs from drivers, comment the non-empty ones. 30534 30535commit 65b5f0017e29903565aff0b7527d955ed016c9a6 30536Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30537Date: Tue Sep 20 16:53:34 2005 +0000 30538 30539 Only include radeon_dri.h & radeon_sarea.h if XF86DRI is defined. 30540 Include "compiler.h" to define __inline__ on non-gcc compilers 30541 30542commit 0b7441b19a1a5df91aca6b1a02944afb636bc523 30543Author: Eric Anholt <anholt@freebsd.org> 30544Date: Sun Sep 18 02:32:23 2005 +0000 30545 30546 Break EXA ABI while we still can. Add coordinates to the UploadToScreen 30547 hook so we can upload a subset of a pixmap, and convert the current 30548 drivers to respect that. Use this support to directly UploadToScreen in 30549 exaGlyphs, providing a 47.4% +/-2.4% decrease in wall time for ls -lR 30550 programs/Xserver in an antialiased gnome-terminal on an M6 (n=3, caches 30551 hot). I would have bumped major version, only I can't tell what the 30552 EXA_VERSION_* is supposed to be doing as opposed to the module version. 30553 30554commit 64e7db2698ee48af92594c16d72e33c16372c9d7 30555Author: Eric Anholt <anholt@freebsd.org> 30556Date: Sat Sep 17 20:02:01 2005 +0000 30557 30558 - Don't try to upload 0 byte-per-pixel (PICT_a1) data using 30559 RADEONHostDataBlit. 30560 - Disable the shortcut for switching from 3d to 3d in radeon_exa.c. It 30561 appears that we do need the cache flush here, thought it's not clear 30562 why. Disable the 2d to 2d shortcut while here, since I'm unsure of what 30563 we're doing. Exposed by the following bit: 30564 - Bug #4485: Add a new routine, exaGlyphs, to handle font drawing. Glyphs 30565 were being accumulated in from non-migratable scratch pixmaps, causing 30566 the destination pixmap to move towards screen but the migration 30567 necessary for source never to happen, leading to abysmal performance. 30568 Instead, copy the scratch glyph data into a real pixmap first, then 30569 composite from that into the destination, allowing for migration. time 30570 ls -lR from programs/Xserver showed 26.9% (+/- 6.3%) decrease in wall 30571 time (n=3). 30572 - Create exaDrawableUse* wrapping exaPixmapUse*, but which are aware of 30573 windows needing backing store. Makes migration code prettier, and 30574 ensures that composited windows will be migrated as normal when we turn 30575 off cw for EXA. (issue brought up by keithp) 30576 30577commit 3c71198a088e71a105e274f176c8ceb503615d46 30578Author: Dave Airlie <airlied@linux.ie> 30579Date: Sat Sep 17 07:47:51 2005 +0000 30580 30581 Switch radeon driver to using a card type rather than having a IsPCI flag, 30582 allow for PCI/AGP/PCIE cards. Set PCI gart allocation to only happen on 30583 PCIE cards. Add Radeon PCIE card detection using capability bits. 30584 30585commit 10cf3523f236a017a1ca7e5ee57e02176b924059 30586Author: Eric Anholt <anholt@freebsd.org> 30587Date: Wed Sep 14 22:06:41 2005 +0000 30588 30589 Bug #2573: Remove the disabling of automatic cache flushing on non-r300 30590 cards. It has been implicated in hanging at least r300s and some r100s, 30591 and seemed to be the cause of some bad rendering for me. 30592 30593commit d68894a463a3bb08c97e91ff6fc4b65b29183ff6 30594Author: Eric Anholt <anholt@freebsd.org> 30595Date: Tue Sep 13 22:04:12 2005 +0000 30596 30597 Remove the hardcoded USE_EXA/USE_XAA defines now that they've been 30598 hardcoded in configure.ac (config.h). 30599 30600commit baa9b5b6e19cf9794bf810e3c582bf1d7f808663 30601Author: Dave Airlie <airlied@linux.ie> 30602Date: Mon Sep 12 09:27:07 2005 +0000 30603 30604 Fix Radeon EXA for when RENDER isn't defined 30605 30606commit 8c89610389197dc94d3bceaf44d9d2ea91848f1b 30607Author: Dave Airlie <airlied@linux.ie> 30608Date: Mon Sep 12 09:17:42 2005 +0000 30609 30610 Move GART offset/size outside XAA 30611 30612commit eba44d9747adde8385f2a242dee5a8e73ce5a205 30613Author: Daniel Stone <daniel@fooishbar.org> 30614Date: Mon Sep 12 09:04:36 2005 +0000 30615 30616 Define USE_XAA and USE_EXA for the time being. 30617 30618commit a255e7b0984890daad98b74ace0c572965fe61f5 30619Author: Eric Anholt <anholt@freebsd.org> 30620Date: Sun Sep 11 21:00:44 2005 +0000 30621 30622 Attempt to fix up the modular pieces for EXA support. (untested) 30623 30624commit 3e0d9c945a6a71cc476d27341ef18618529a91a8 30625Author: Eric Anholt <anholt@freebsd.org> 30626Date: Sun Sep 11 20:58:53 2005 +0000 30627 30628 Add support for EXA to the radeon driver. Building EXA and XAA support is 30629 controlled at compile time, plus the runtime option of Option 30630 "AccelMethod" "EXA" or "XAA". The XAA support appears to remain as 30631 before, while the EXA pieces need just a little more polishing. Notable 30632 features: 30633 - Render acceleration working on Radeon 100 and 200-series with DRI on. 30634 - DRI works with EXA Notable issues: 30635 - DGA disabled in the EXA case. 30636 - Backbuffer moves disabled in the EXA case. 30637 - No textured XVideo. 30638 - MMIO render acceleration is close but still has some issues. 30639 - Memory pressure while using Composite is really troublesome with DRI on. 30640 This patch is based on an initial patch by Zack Rusin, with significant 30641 work by Benjamin Herrenschmidt and myself. 30642 30643commit fd62082b68ac3aadd8ffc441352d75d88334904e 30644Author: Dave Airlie <airlied@linux.ie> 30645Date: Sun Sep 11 08:51:38 2005 +0000 30646 30647 Add support for allocating PCI GART table in framebuffer memory. This is 30648 needed to support PCIE Radeons using a new DRM. 30649 30650commit cb63f8d9c0563fb0eff28e2be6d4adf5666540d2 30651Author: Dave Airlie <airlied@linux.ie> 30652Date: Tue Sep 6 10:03:19 2005 +0000 30653 30654 Handle VIP timeouts more gracefully -- impose a hard limit of 10ms on 30655 waiting for VIP_BUSY so we don't lock up hard, spinning. 30656 30657commit dbd83cf18e1bf5e7e31e1ac150227bd27e8b1739 30658Author: Adam Jackson <ajax@nwnk.net> 30659Date: Sun Aug 28 20:51:14 2005 +0000 30660 30661 Bug #4148: Bump required DRM version for r300 to 1.17. 30662 30663commit ce0589ff3b9fcf8c6bab64da4dba40ba540a6a03 30664Author: Adam Jackson <ajax@nwnk.net> 30665Date: Sun Aug 28 18:10:34 2005 +0000 30666 30667 Bug #1109: Fix VGA init on Rage Mobility 7500 (Marc Le France) 30668 30669commit 250c8732a54924eeb459880f2905f8e59d038e60 30670Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 30671Date: Sun Aug 28 18:00:23 2005 +0000 30672 30673 Modified the licens type from GPL to a X/MIT one 30674 30675commit 94c0196652c3cfa0558f61b55f5df9bf50f9ac49 30676Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30677Date: Fri Aug 26 20:04:01 2005 +0000 30678 30679 #include "extensions/dpms.h" -> <X11/extensions/dpms.h> 30680 30681commit 107e42cabe1cc41fe86b0218c3229b5f34b2df84 30682Author: Daniel Stone <daniel@fooishbar.org> 30683Date: Fri Aug 26 07:00:24 2005 +0000 30684 30685 Fix 'ugly' warning in r128. (Branden Robinson, Debian #069) 30686 30687commit d3f5911c85ccfaaf50c05ce80d56dac13fd15701 30688Author: Daniel Stone <daniel@fooishbar.org> 30689Date: Fri Aug 26 05:30:52 2005 +0000 30690 30691 Only use vgahw symbols if we're building vgahw for this architecture. TODO: 30692 Fix this for the modular tree (maybe an unconditional yes?). Define 30693 that we're using a DPMS server and include DPMS headers. (Debian #024b) 30694 30695commit 879213455740d2d653d38c712275f07610e4dc9c 30696Author: Alex Deucher <agd5f@yahoo.com> 30697Date: Fri Aug 26 02:11:24 2005 +0000 30698 30699 - fix pitch increment (Aapo Tahkola, Roland Scheidegger) 30700 30701commit 6519cc71e4387ef979de90be517145611511ff1a 30702Author: Daniel Stone <daniel@fooishbar.org> 30703Date: Thu Aug 25 03:34:58 2005 +0000 30704 30705 Fix bug in Radeon connector table parsing -- parse all entries, not just 30706 the first. (Ben Herrenschmidt) 30707 30708commit 8734a5d0bafb60f5e38b2cee10587f01fc016379 30709Author: Daniel Stone <daniel@fooishbar.org> 30710Date: Thu Aug 25 03:07:50 2005 +0000 30711 30712 Blacklist Radeon M300 from multimedia I2C code. Handle I2C timeouts more 30713 gracefully -- impose a hard limit of 10ms on waiting for I2C_GO so we 30714 don't lock up hard, spinning. 30715 30716commit 578564b4f7fce1f02d056c2b12ffb7803d268829 30717Author: Adam Jackson <ajax@nwnk.net> 30718Date: Mon Aug 22 03:05:52 2005 +0000 30719 30720 Explicit license statement. 30721 30722commit c9005c91bf6aaf4728a6606284f5d24e0b4a8282 30723Author: Daniel Stone <daniel@fooishbar.org> 30724Date: Thu Aug 18 09:03:46 2005 +0000 30725 30726 Update autogen.sh to one that does objdir != srcdir. 30727 30728commit 508a7421a1386b5b783a867d860b858a55f216d3 30729Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 30730Date: Mon Aug 15 07:30:05 2005 +0000 30731 30732 Egbert's 64bit fixes for mixed 32/64bit clients 30733 30734commit 204290412fbe8b4ac1a0f4f97bc7aec2847a1df7 30735Author: Daniel Stone <daniel@fooishbar.org> 30736Date: Sun Aug 14 16:14:51 2005 +0000 30737 30738 Make DGA, CPIO, non-PCI, and TV-Out support detected per-platform, using 30739 the same semantics as in drivers/ati/Imakefile. 30740 30741commit adb093bdf3cea8108a40bcd72f09924b416de0c1 30742Author: Alex Deucher <agd5f@yahoo.com> 30743Date: Mon Aug 8 23:42:36 2005 +0000 30744 30745 - radeon XPRESS chips are actually RV3xx based 30746 30747commit 9c11018beb186b4b303524685fb62629b9ca863e 30748Author: Vladimir Dergachev <volodya@mindspring.com> 30749Date: Sun Aug 7 14:46:55 2005 +0000 30750 30751 Modified: 30752 programs/Xserver/hw/xfree86/drivers/ati/radeon_driver.c 30753 Update warning about DRI support on R300 and later cards. 30754 30755commit a28b2c76c7df2fe65a93eed34bc85c2cf5afb314 30756Author: Vladimir Dergachev <volodya@mindspring.com> 30757Date: Sun Aug 7 14:40:58 2005 +0000 30758 30759 Modified: 30760 programs/Xserver/hw/xfree86/drivers/ati/radeon_driver.c 30761 Properly configure fbLocation and related registers for R300 cards. (Dave 30762 Airlie) 30763 30764commit 692fb0f8b51abeccaf70b51fc1defc62132b15e9 30765Author: Alex Deucher <agd5f@yahoo.com> 30766Date: Sat Aug 6 17:23:16 2005 +0000 30767 30768 - missing line in mergedfb option parsing. Spotted by Christoph Bauer 30769 30770commit 14e77a39320a7a074b5c2eeeea9750b469abc978 30771Author: Ian Romanick <idr@us.ibm.com> 30772Date: Fri Aug 5 23:04:32 2005 +0000 30773 30774 I have resolved this by making all the "Direct rendering disabled" messages 30775 X_WARNING and all the "Direct rendering enabled" messages X_INFO. 30776 Ideally, we need to factor some of this stuff out to common code. There 30777 is a *LOT* of duplication among the *_dri.c files under 30778 programs/Xserver/hw/xfree86/drivers. 30779 I'm trying to clean up as much low hanging fruit as I can and got tired of 30780 waiting for ajax. :P 30781 30782commit 8e88d36f78c13025847310b411402f1d46a2e0cd 30783Author: Adam Jackson <ajax@nwnk.net> 30784Date: Sun Jul 31 17:19:27 2005 +0000 30785 30786 Post-merge fixups: 30787 - Disable building the gamma driver 30788 - Fix numerous Imakefiles to get header paths and subdirs right 30789 - Symlink client-side GLX lib source from Mesa 30790 - Add new shader subdirs to the Mesa core build 30791 - Tweak the DDX drivers to not ask for "xf86dri.h"; they don't need it, 30792 since they don't generate protocol, and it was breaking the build. 30793 30794commit fb9f712fd1044feb8ce6a99aabaeb5966123e268 30795Author: Kevin E Martin <kem@kem.org> 30796Date: Fri Jul 29 21:22:42 2005 +0000 30797 30798 Various changes preparing packages for RC0: 30799 - Verify and update package version numbers as needed 30800 - Implement versioning scheme 30801 - Change bug address to point to bugzilla bug entry form 30802 - Disable loadable i18n in libX11 by default (use --enable-loadable-i18n to 30803 reenable it) 30804 - Fix makedepend to use pkgconfig and pass distcheck 30805 - Update build script to build macros first 30806 - Update modular Xorg version 30807 30808commit ee45e1fb5da793fa14770b28fd662e69d52b2c6a 30809Author: Michel Daenzer <michel@daenzer.net> 30810Date: Fri Jul 29 19:45:14 2005 +0000 30811 30812 bugzilla #3911 (https://bugs.freedesktop.org/show_bug.cgi?id=3911) 30813 attachment #3191 (http://bugs.freedesktop.org/attachment.cgi?id=3191) 30814 Disable bus mastering while updating MC_FB_LOCATION and friends to 30815 prevent the X server from hanging on startup every now and then under 30816 some circumstances. (ATI Technologies Inc.) 30817 30818commit 44a66ec5a01c3b4dfa046a66b80a3b7e9dfff537 30819Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 30820Date: Fri Jul 15 15:02:47 2005 +0000 30821 30822 Install theathre drivers in @moduledir@/multimedia 30823 30824commit dbcf104689ca368523715eb5945bcd800985799b 30825Author: Kevin E Martin <kem@kem.org> 30826Date: Fri Jul 15 03:39:58 2005 +0000 30827 30828 Fix filename typo 30829 30830commit 2256fabf1cd69c950639f236f61e59e68cc8e6fa 30831Author: Kevin E Martin <kem@kem.org> 30832Date: Thu Jul 14 22:30:47 2005 +0000 30833 30834 Fix check for headers. Use {XORG,DRI}_CFLAGS subst instead of replacing 30835 CFLAGS. 30836 30837commit 520317863ffb365eca17c96b4d545adb554bbbd6 30838Author: Kevin E Martin <kem@kem.org> 30839Date: Thu Jul 14 19:35:48 2005 +0000 30840 30841 Make compiling with DRI support configurable/autodetecable. Autodetect DRI 30842 sdk headers. Expand help string for xorg-module-dir option. 30843 30844commit ab5264838ae34bb29791bc08d5a4f853a995faf1 30845Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 30846Date: Thu Jul 14 17:10:01 2005 +0000 30847 30848 fuck you cvs 30849 30850commit a608c4422665ce1e933483b7c2a25ab2daae5930 30851Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 30852Date: Thu Jul 14 16:03:36 2005 +0000 30853 30854 Try adding build system for ATI driver again. Who knows, maybe CVS works 30855 today. 30856 30857commit d715561d28d82591adf5c7504068c1818eb2245c 30858Author: Adam Jackson <ajax@nwnk.net> 30859Date: Mon Jul 11 02:29:45 2005 +0000 30860 30861 Prep for modular builds by adding guarded #include "config.h" everywhere. 30862 30863commit 5b37b695019b7a4522a96a63f1f6b035160e6a25 30864Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 30865Date: Mon Jul 4 23:44:59 2005 +0000 30866 30867 Bug #2604: Disable workaround for horrible flickering when scrolling 30868 vertically in a virtual screen when color tiling and page flip are both 30869 enabled, as it potentially does more harm than good. The flickering 30870 still needs fixing probably though. 30871 30872commit 08b886850f115f404187acd619aa2c2ca18ae3c7 30873Author: Daniel Stone <daniel@fooishbar.org> 30874Date: Fri Jul 1 22:43:11 2005 +0000 30875 30876 Change all misc.h and os.h references to <X11/foo.h>. 30877 30878commit 1c7c3372776d6d116facabbf055f27af3e96f59f 30879Author: Alex Deucher <agd5f@yahoo.com> 30880Date: Fri Jul 1 04:16:42 2005 +0000 30881 30882 - fix depth tiling for r3/4xx chips (Aapo Tahkola) 30883 - adjust limits for tiled surfaces on r3/4xx 30884 30885commit e8c2574557950bfa150806432d566cd38c47158f 30886Author: Alex Deucher <agd5f@yahoo.com> 30887Date: Fri Jul 1 02:56:04 2005 +0000 30888 30889 - Fix Support for Philips FM1236/F tuner on ATI AIW 9600 XT (Jeff Smith) 30890 Bug 3401 30891 30892commit d258e1c1566d005693260230712c9d78f1447dba 30893Author: Alex Deucher <agd5f@yahoo.com> 30894Date: Thu Jun 30 04:46:11 2005 +0000 30895 30896 - change back part of a comment I seem to have accidently changed. 30897 30898commit 6d9c02d58eae1966e8fa8d7d305849c6e57434ad 30899Author: Alex Deucher <agd5f@yahoo.com> 30900Date: Thu Jun 30 02:02:10 2005 +0000 30901 30902 - small mergedfb fix from Thomas Winischhofer 30903 30904commit 85365b791542cab221c37c9b3b4593ea0c6c9e2d 30905Author: Alex Deucher <agd5f@yahoo.com> 30906Date: Thu Jun 30 01:40:17 2005 +0000 30907 30908 - add support for colortiling on r3/4xx hardware (disabled by default). In 30909 order for pageflipping to work on r3/4xx we will need to update the the 30910 sarea and drm so that the crtc1/2 xytiling regs get updated. Initially 30911 worked out by Aapo Tahkola. 30912 30913commit 3ec99c9cd821aeacdd57a9bbc20c4e4ec4711790 30914Author: Alex Deucher <agd5f@yahoo.com> 30915Date: Thu Jun 30 00:56:01 2005 +0000 30916 30917 - Increase the priority of display 0/1 in the memory controller for r3/4xx 30918 hardware when displaypriority is set to HIGH. Fixes display problems in 30919 high res modes. Originally reported by Aapo Tahkola. 30920 30921commit a7523e6c0be7a600963e9de10cf0060d16117bf6 30922Author: Alex Deucher <agd5f@yahoo.com> 30923Date: Thu Jun 30 00:33:36 2005 +0000 30924 30925 - fix some OUTREGs that should be OUTPLLs in dynamicclocks code 30926 30927commit c7e1d31c6aec3a99061c77c4d6f57e2e874e37f1 30928Author: Adam Jackson <ajax@nwnk.net> 30929Date: Sun Jun 26 04:38:12 2005 +0000 30930 30931 Bug #3628: Fix video driver submodule visibility. 30932 30933commit 1beca962ecee5b5c51eb2fb8451fcc2a2ed4b97c 30934Author: Adam Jackson <ajax@nwnk.net> 30935Date: Sat Jun 25 21:16:53 2005 +0000 30936 30937 Bug #3626: _X_EXPORT tags for video and input drivers. 30938 30939commit 9a4895c4c2fd9614d83adfe3967f4f46dc4dfff6 30940Author: Eric Anholt <anholt@freebsd.org> 30941Date: Wed Jun 22 22:50:49 2005 +0000 30942 30943 Add missing detection of the PCI ID for 0x5b62, Radeon X600 (Vladimir 30944 Kushnir). 30945 30946commit 43e3cc7c21160f4536ed860f97d02545c267dfbc 30947Author: Alex Deucher <agd5f@yahoo.com> 30948Date: Wed Jun 15 23:40:50 2005 +0000 30949 30950 Sync up radeon mergedfb with Thomas' latest sis mergedfb changes: 30951 - non-rectangular mergedfb desktops 30952 - screen offsets 30953 30954commit b67f9f5c875efaa41bded7ed2f38b2a5fbd17597 30955Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30956Date: Sat Jun 4 20:26:28 2005 +0000 30957 30958 Sync with 6.8.2 branch: Bug #591 30959 (https://bugs.freedesktop.org/show_bug.cgi?id=591) attachment #308 30960 (https://bugs.freedesktop.org/attachment.cgi?id=308): Fix video timing 30961 problems with Sony PCG-C1VN Vaio Picturebook notebook && co. Patch by 30962 Marc Aurele La France 30963 30964commit daae70d9083e9c15358f160f0df3176e344f1cb9 30965Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 30966Date: Sat Jun 4 20:04:36 2005 +0000 30967 30968 Sync with 6.8.2 branch: Bugzilla #1306 30969 (https://bugs.freedesktop.org/show_bug.cgi?id=1306) attachment #1291 30970 (https://bugs.freedesktop.org/attachment.cgi?id=1291): Fix SEGV in 30971 "radeon" driver. Patch by Kevin E. Martin <kem@freedesktop.org> 30972 30973commit c27c193f192b117232dbe39575d3d5ab7c6a1b1e 30974Author: Felix Kuehling <fxkuehl@gmx.de> 30975Date: Fri May 27 22:33:23 2005 +0000 30976 30977 Fix for bugzilla https://bugs.freedesktop.org/show_bug.cgi?id=3414 (patch 30978 https://bugs.freedesktop.org/attachment.cgi?id=2782): Don't use RMX if 30979 we are Dell Server. 30980 30981commit 3b674815e0a6d19b0719e71261c0909b92173aed 30982Author: Adam Jackson <ajax@nwnk.net> 30983Date: Tue May 24 23:26:50 2005 +0000 30984 30985 Typo fix from bug #3163 commit 30986 30987commit 0845d1db801ae74688464bd4c41a7e40e4b50186 30988Author: Egbert Eich <eich@suse.de> 30989Date: Mon May 23 09:50:20 2005 +0000 30990 30991 Explicitely disable DPMS power states during close screen. Fix macro that 30992 takes a value and a mask of bits to modify to be behave evquvalently if 30993 the bits that should remain unchanged are set or unset in the value 30994 (Bugzilla #3369). 30995 30996commit 74b86b036def72e5180a46871b8452ba9eea9bb1 30997Author: Adam Jackson <ajax@nwnk.net> 30998Date: Wed May 18 17:47:35 2005 +0000 30999 31000 Bug #3163: Create new DRIINFO_*_VERSION macros to indicate the version 31001 number of the DRIInfoRec, disambiguating it from the XF86DRI protocol 31002 version number. Modify DRIQueryVersion to return the libdri version 31003 number, which all DDXes were requesting implicitly. Fix the DDXes to 31004 check for the DRIINFO version they were compiled against. 31005 31006commit 0f4a8fc6aa5157a8b48beb49f30511665cfa4b63 31007Author: Matthieu Herrb <matthieu.herrb@laas.fr> 31008Date: Wed Apr 20 21:07:44 2005 +0000 31009 31010 programs/Xserver/hw/xfree86/drivers/ati/theatre.h Build fix for 31011 DoLoadableServer = NO systems. 31012 31013commit 3355d12f832909b05e8b022fad784c9f7af7341f 31014Author: Daniel Stone <daniel@fooishbar.org> 31015Date: Wed Apr 20 12:25:22 2005 +0000 31016 31017 Fix includes right throughout the Xserver tree: 31018 change "foo.h" to <X11/foo.h> for core headers, e.g. X.h, Xpoll.h; 31019 change "foo.h", "extensions/foo.h" and "X11/foo.h" to 31020 <X11/extensions/foo.h> for extension headers, e.g. Xv.h; 31021 change "foo.[ch]" to <X11/Xtrans/foo.[ch]> for Xtrans files. 31022 31023commit 31cb4758eae4703d4de31a0b291847cefa855fc3 31024Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 31025Date: Sun Apr 17 23:28:09 2005 +0000 31026 31027 Added fifo_read and fifo_write in tge generic bus definition 31028 31029commit 2b363e02c7c122f07e6fc406fb8ee52d97b55cad 31030Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 31031Date: Sun Apr 17 23:26:36 2005 +0000 31032 31033 Added new options for the ATI driver: RageTheatreMicrocPath and 31034 RageTheatreMicrocType 31035 31036commit 6286efeba564c63ec198c091e4bba22027586901 31037Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 31038Date: Sun Apr 17 23:23:45 2005 +0000 31039 31040 Addeded support for FI1216ME (PAL) version of the tuner, detection of 31041 UDA1380 sound processor 31042 31043commit 9a63c216e90622cd96d4555155df06dd66c3235b 31044Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 31045Date: Sun Apr 17 23:18:26 2005 +0000 31046 31047 Use the theatre_detect to determine the chip ID and then load the 31048 coresponding theatre module 31049 31050commit 247057762813f12f171b1fa05f31c5ace4c245a7 31051Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 31052Date: Sun Apr 17 23:15:31 2005 +0000 31053 31054 Added functions that access the FIFO on RT200 31055 31056commit bcccf9259784f21d84d52c526aa68df0787118b4 31057Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 31058Date: Sun Apr 17 23:13:45 2005 +0000 31059 31060 Added RT200 registers 31061 31062commit da9d27f8c8cccf2f40a244b25ffe0d3162c41a63 31063Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 31064Date: Sun Apr 17 23:09:28 2005 +0000 31065 31066 This is the theatre chip detection module 31067 31068commit e4623bbce161c41ccac65210fd2539211e8af715 31069Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 31070Date: Sun Apr 17 23:07:54 2005 +0000 31071 31072 Now the detection of the theatre chip is in other module called 31073 theatre_detect 31074 31075commit 405b4dcb067346fce7fd9f6d8a6507a58c740ac7 31076Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 31077Date: Sun Apr 17 23:06:17 2005 +0000 31078 31079 Added the RageTheatre200 video demodulator support 31080 31081commit 310eb24a45b145efbd47196a1c3ec750cf004154 31082Author: Daniel Stone <daniel@fooishbar.org> 31083Date: Mon Apr 4 23:07:08 2005 +0000 31084 31085 Release SDA/SDL I2C lines from an asserted-low state after a DDC probe, 31086 which caused Apple Studio Display monitors to shut off after a second 31087 or so (Ben Herrenschmidt). 31088 31089commit a416667f980a8679352be897af2fb3215da2d000 31090Author: Alex Deucher <agd5f@yahoo.com> 31091Date: Thu Mar 31 23:18:10 2005 +0000 31092 31093 - Add new Radeon pci ids (ATI devrel), bug 2827 31094 31095commit 2e7365ffb767099faf3f191dc65341dbb3ef6247 31096Author: Vladimir Dergachev <volodya@mindspring.com> 31097Date: Thu Mar 31 18:26:05 2005 +0000 31098 31099 Modified: 31100 programs/Xserver/hw/xfree86/drivers/ati/radeon_video.c 31101 Patch from Antti Ajanki - provide VBI capture support for PAL and SECAM 31102 (You need GATOS km module to actually get at the data, the change 31103 merely configures the hardware to handle it correctly). 31104 31105commit 9a3c1c8e4cba2885b707db0f855e11dc6e640b49 31106Author: Michel Daenzer <michel@daenzer.net> 31107Date: Tue Mar 29 06:14:34 2005 +0000 31108 31109 bugzilla #2844 (https://bugs.freedesktop.org/show_bug.cgi?id=2844) 31110 attachment #2230 (http://bugs.freedesktop.org/attachment.cgi?id=2230) 31111 Wait for idle before uploading HW cursor data to the framebuffer to 31112 avoid lockups when moving the mouse on some machines. 31113 31114commit 97863f112850be5206ea555297354c1c104fcdbe 31115Author: Kevin E Martin <kem@kem.org> 31116Date: Tue Mar 29 03:49:04 2005 +0000 31117 31118 bugzilla #2840 (https://bugs.freedesktop.org/show_bug.cgi?id=2840) 31119 attachment #2227 (https://bugs.freedesktop.org/attachment.cgi?id=2227) 31120 Fix build problems introduced by previous Radeon PLL errata 31121 workarounds. 31122 31123commit 04cbb8d174cc5e3c7ecdd4e171170213ecb3c7ed 31124Author: Dave Airlie <airlied@linux.ie> 31125Date: Sat Mar 26 00:53:01 2005 +0000 31126 31127 bugzilla #2057 (https://bugs.freedesktop.org/show_bug.cgi?id=2057) 31128 attachment #1516 (https://bugs.freedesktop.org/attachment.cgi?id=1516) 31129 Add TVOUT Support for Mach64 (Leif Delgass, fixed up for Xorg by me). 31130 31131commit 5be4bf9000bdf58584a10a6b8e285d0f173304fa 31132Author: Alex Deucher <agd5f@yahoo.com> 31133Date: Thu Mar 24 06:45:52 2005 +0000 31134 31135 - radeon PLL access errata workarounds (Benjamin Herrenschmidt, acked by 31136 Hui Yu) 31137 31138commit 34af3da80304a13b6407dcc0f4db54d6c8bb7a41 31139Author: Michel Daenzer <michel@daenzer.net> 31140Date: Fri Mar 18 22:36:47 2005 +0000 31141 31142 Add support for production version of ATI RN50/ES1000. (ATI Technologies 31143 Inc.) 31144 31145commit b56a5f96048b920a7751318dde211e881f53b2bb 31146Author: Vladimir Dergachev <volodya@mindspring.com> 31147Date: Thu Mar 17 03:36:25 2005 +0000 31148 31149 Modified: 31150 programs/Xserver/hw/xfree86/drivers/ati/radeon_mergedfb.c Back out 31151 RadeonWaitForIdleMMIO that I put in earlier. 31152 From discussion on dri-devel we do not need to be as careful with INREG() 31153 access on Radeons as we are on mach64 chips. 31154 The fix is backed out due to: 31155 it breaks DRI on SMP systems (due to absense of the lock in mouse handling 31156 functions) 31157 I am no longer able to reproduce the lockup I saw before 31158 31159commit 01a9c1157cdfbc5f74b035383e060b8419979563 31160Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 31161Date: Tue Mar 8 23:41:37 2005 +0000 31162 31163 Fix accidentally removed depth pitch adjustment if displayWidth is not a 31164 multiple of 32 pixels (bug #788) 31165 31166commit 9b994b40710103b263364153429ebdb9ea3f58ed 31167Author: Vladimir Dergachev <volodya@mindspring.com> 31168Date: Fri Mar 4 22:21:40 2005 +0000 31169 31170 programs/Xserver/hw/xfree86/drivers/ati/radeon_dri.c Let through all chips 31171 R300 and above - presumably the filtering should take place at the DRM 31172 driver level. Correct cut'n'paste position misprint. Big thanks for 31173 Michel Danzer for pointing out both improvements. 31174 31175commit 7d37a9f1f9fffe98067c3b5bc4749cc799fb0c77 31176Author: Vladimir Dergachev <volodya@mindspring.com> 31177Date: Fri Mar 4 05:48:31 2005 +0000 31178 31179 programs/Xserver/hw/xfree86/drivers/ati/radeon_dri.c Turns out Radeon R420 31180 chips work fine with R300 microcode. Thanks go to Rune Petersen for 31181 careful investigation. 31182 31183commit f1edbb7852d85f660188498f578a0a697c2094fe 31184Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 31185Date: Tue Feb 22 00:39:15 2005 +0000 31186 31187 Fix building radeon driver with BuildXF86DRI set to NO, the breakage was 31188 caused by the tiling support (bug #2481) 31189 31190commit b853dbb50fb7c46fb5a70f349a3f264f0bb94e8c 31191Author: Vladimir Dergachev <volodya@mindspring.com> 31192Date: Sat Feb 19 16:00:12 2005 +0000 31193 31194 Modified: 31195 programs/Xserver/hw/xfree86/drivers/ati/radeon_video.c 31196 Fix an INREG without WaitForIdleMMIO. This one should be unlikely to 31197 trigger, but we should still do this in the name of correctness. 31198 31199commit 84290849b4aee336e9db0779d93fc2d7ed8881b1 31200Author: Vladimir Dergachev <volodya@mindspring.com> 31201Date: Sat Feb 19 00:57:23 2005 +0000 31202 31203 Modified: 31204 programs/Xserver/hw/xfree86/drivers/ati/radeon_mergedfb.c 31205 Add a call to RADEONWaitForIdleMMIO() within RADEONChooseCursorCRTC() as 31206 this function uses OUTREGP() which, internally, calls INREG(). 31207 Thus (due to SilkenMouse) INREG() could be called while CP engine was busy. 31208 I have observed this lockup on RV350 + R300 3d driver and have no idea 31209 why no issues were reported for R200 cards. Perhaps, few of them have 31210 enough memory to run merged fb and 3d simultaneously ? 31211 This fix is the right thing to do, however, we might also need to put fifo 31212 management calls into other cursor handling functions (in particular 31213 those that write cursor image directly to video memory). 31214 31215commit 60070063e2130af0b9d5d34eed7adbae0c0fa67b 31216Author: Michel Daenzer <michel@daenzer.net> 31217Date: Fri Feb 18 19:55:35 2005 +0000 31218 31219 Bug #2576: Add support for ATI RN50/ES1000. (ATI Technologies Inc.) 31220 31221commit 32032cdc78097840b7e2685fdec11bc0816aa0ac 31222Author: Adam Jackson <ajax@nwnk.net> 31223Date: Fri Feb 18 16:42:11 2005 +0000 31224 31225 Bug #2567: Fix reversed memset arguments. (David Krause) 31226 31227commit d5fadd6d86ae2a2f8c1bb55e36ab3d6d49649b46 31228Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 31229Date: Wed Feb 9 02:22:39 2005 +0000 31230 31231 do not enable color tiling on r300 chips and newer, since it doesn't work 31232 (yet) (Bugzilla #2504) 31233 31234commit 5ff35300e66a81ecc17c4e9396eb6d158f5c8e1b 31235Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 31236Date: Wed Feb 2 21:35:30 2005 +0000 31237 31238 fix error with color tiling and hw scrolling, it looks like rv100 doesn't 31239 need the special treatment, after all (reported by Stephane Marchesin) 31240 31241commit 7a9031683914bb1d17d223634e5dbf2e18955651 31242Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 31243Date: Mon Jan 31 20:10:28 2005 +0000 31244 31245 change option EnableColorTiling to ColorTiling 31246 31247commit a2b123f7672ea678290798786ecb408a86d8c683 31248Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 31249Date: Wed Jan 26 18:23:41 2005 +0000 31250 31251 (Stephane Marchesin, me) Add support for color (framebuffer) tiling. 31252 Convert all XAA accel code to use explicit dst and src offset pitch 31253 control to make it work. Use the new drm surface ioctls to set up the 31254 tiled surfaces (if dri is enabled). 31255 31256commit e0df1e41724cd009fe7c833d2f9555f5d87e3202 31257Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 31258Date: Wed Jan 26 17:39:56 2005 +0000 31259 31260 fix issues with mergedfb + pageflip (fix up sarea frame values) 31261 31262commit aeb01807263c47f49abaf00031d8505e8ebda82c 31263Author: Egbert Eich <eich@suse.de> 31264Date: Mon Jan 24 15:40:21 2005 +0000 31265 31266 Don't exclude 'RADEONInfoPtr info = ...' declaration even when not building 31267 with DRI support (Bugzilla #2370) 31268 31269commit c6c30cfa354eb456638f10771a7cb02271696af2 31270Author: Vladimir Dergachev <volodya@mindspring.com> 31271Date: Thu Jan 13 17:35:49 2005 +0000 31272 31273 Modified: 31274 programs/Xserver/hw/xfree86/drivers/ati/radeon_driver.c 31275 Some of the newer Radeons (R420 for example) have MM_TABLE's with size 0x33 31276 (instead of 0x0c that we know). Furthermore, they lockup during access 31277 of I2C bus ! 31278 Add a check to validate MM_TABLE. Note: the check is not more stringent as 31279 other card manufacturers (like Asus or Tekram) might make cards with 31280 slightly different MM_TABLE. 31281 31282commit 414c035a965b9db15656058c0e9cfeb3acf0158f 31283Author: Alex Deucher <agd5f@yahoo.com> 31284Date: Fri Dec 24 18:26:19 2004 +0000 31285 31286 - Fix MCLK / SCLK calculation (Benjamin Herrenschmidt) bug 2144 31287 - formatting cleanup 31288 31289commit 7c7b278883676412f5709cf812587b1408628133 31290Author: Alex Deucher <agd5f@yahoo.com> 31291Date: Sun Dec 19 19:49:33 2004 +0000 31292 31293 - Add support for Dualhead on M3/M4 (bug 1760) 31294 31295commit 08a45118b53129f3b9c30004404b906c7d7690be 31296Author: Alex Deucher <agd5f@yahoo.com> 31297Date: Sun Dec 19 19:29:10 2004 +0000 31298 31299 - Make dualhead ram allocation cleaner and related messages clearer 31300 31301commit 97cfd7b1489a9510cd233d6dc7154c210f32b9f5 31302Author: Alex Deucher <agd5f@yahoo.com> 31303Date: Sun Dec 19 06:56:37 2004 +0000 31304 31305 - Add DRI resume support (Micha Feigin) bug 2063 31306 31307commit 5b8577ae3200d967b4316156f16028cbcfd21983 31308Author: Alex Deucher <agd5f@yahoo.com> 31309Date: Sun Dec 19 06:11:06 2004 +0000 31310 31311 - fix warnings (Ronny V. Vindenes) 31312 31313commit 447124a663a24a7fffb0cd241d18dd896a43098b 31314Author: Alex Deucher <agd5f@yahoo.com> 31315Date: Sat Dec 18 21:37:48 2004 +0000 31316 31317 - fix warning and potential error in radeon_driver.c 31318 31319commit ac6166158a58b377e7383950c169956f78b62093 31320Author: Alex Deucher <agd5f@yahoo.com> 31321Date: Fri Dec 17 19:19:32 2004 +0000 31322 31323 - Add VGAAccess option to r128 driver. Based on Benh's radeon patch. r128 31324 should now work on ppcs without usefbdev. To use this on ppc make sure 31325 you set "usefbdev" to false. (bug 2089) 31326 - Apply Benjamin Herrenschmidt's radeon patches (bug 2064) 31327 31328commit ac860ed6bbed84f6fc39a9acf53ee87646cfc8dc 31329Author: Vladimir Dergachev <volodya@mindspring.com> 31330Date: Fri Dec 17 16:50:36 2004 +0000 31331 31332 Modified: 31333 programs/Xserver/hw/xfree86/drivers/ati/radeon_accel.c 31334 programs/Xserver/hw/xfree86/drivers/ati/radeon_dri.c 31335 Move DMA robustness fix into radeon_dri.c::RADEONEnterServer() as per 31336 suggestion by Michel D�nzer. 31337 I could not trigger a lockup, even with r300_demo (possibly it has code 31338 that flushes cache inside ?), so this must be good enough.. 31339 31340commit 3ba5b525c4d6390184fdf79b28c954d5caadc6a4 31341Author: Vladimir Dergachev <volodya@mindspring.com> 31342Date: Thu Dec 16 16:26:56 2004 +0000 31343 31344 Modified: 31345 programs/Xserver/hw/xfree86/drivers/ati/radeon_accel.c 31346 Add DMA robustness fix from Nicolai Haehnle. 31347 TODO: move this fix into DRM driver or other parts of Xserver when R300 31348 support is more stable. 31349 This fix affects only R300 and later variants and results in small slowdown 31350 when using CP acceleration in comparison to the previous version of the 31351 driver. 31352 We are still a lot faster in CP mode than using MMIO register access. 31353 31354commit c7b5f4a9eb49c0edd690c972be1015c98408ffaf 31355Author: Adam Jackson <ajax@nwnk.net> 31356Date: Wed Dec 15 19:35:33 2004 +0000 31357 31358 Bug #2008: Fix gcc4 compile problem. 31359 31360commit 8c1969f22fba2e1c9ed34d5ca77bee3a9f577640 31361Author: Vladimir Dergachev <volodya@mindspring.com> 31362Date: Sun Dec 12 20:42:29 2004 +0000 31363 31364 Modified: 31365 programs/Xserver/hw/xfree86/drivers/ati/radeon_render.c 31366 programs/Xserver/hw/xfree86/drivers/ati/radeon_accel.c 31367 Make DMA blitting code more robust. 31368 31369commit 8a9077bc70fd800983f138a224b95301992a2296 31370Author: Vladimir Dergachev <volodya@mindspring.com> 31371Date: Sun Dec 12 19:28:37 2004 +0000 31372 31373 Modified: 31374 programs/Xserver/hw/xfree86/drivers/ati/radeon_driver.c 31375 programs/Xserver/hw/xfree86/drivers/ati/radeon_dri.c 31376 Enable DRM support permanently, allowing its use for any CHIP_FAMILY > 31377 R300. Fix bug in naming Mesa3d driver (it was defaulting to r200) 31378 31379commit 845df6dd96785b59c44c670009c4fcdb20e147ab 31380Author: Alex Deucher <agd5f@yahoo.com> 31381Date: Sun Dec 12 17:29:45 2004 +0000 31382 31383 - missing diff from agp 8x support 31384 31385commit 36b92808e4ff1311f91543388c18cbe8e9eb4aab 31386Author: Vladimir Dergachev <volodya@mindspring.com> 31387Date: Sun Dec 12 16:05:35 2004 +0000 31388 31389 Modified: 31390 programs/Xserver/hw/xfree86/drivers/ati/radeon_accel.c 31391 programs/Xserver/hw/xfree86/drivers/ati/radeon_driver.c 31392 programs/Xserver/hw/xfree86/drivers/ati/radeon_dri.c 31393 programs/Xserver/hw/xfree86/drivers/ati/radeon_common.h 31394 programs/Xserver/hw/xfree86/drivers/ati/radeon_version.h Add R300 (and 31395 later) DRM support, protected by X_R300_DRM option. 31396 Note: at the moment this is for 2d support only, if enabled running 31397 glxgears locks up my machine, but DMAForXv=yes works fine. 31398 31399commit 68ea67ce98b8f3ad4c17f987e0e2b92f79a96fbf 31400Author: Alex Deucher <agd5f@yahoo.com> 31401Date: Sun Dec 12 02:00:48 2004 +0000 31402 31403 - add AGP 8x support (Hui Yu, Michel Daenzer) bug 1819 31404 31405commit 93675d288d5ebb35d4724ebbfc455a703e25a311 31406Author: Alex Deucher <agd5f@yahoo.com> 31407Date: Sun Dec 12 00:14:32 2004 +0000 31408 31409 - only mess with Dynamicclocks on mobility chips (Benh, me) bug 1912 31410 - fix display detection (Hui Yu) bug 1559 31411 - Add new option "BIOSHotkeys" to allow the bios to control output 31412 switching (off by default) bug 2000 31413 - update the savage man page to document the "CrtOnly" option 31414 31415commit ed217003127028483fd2956851d438977169d154 31416Author: Daniel Stone <daniel@fooishbar.org> 31417Date: Mon Dec 6 13:27:56 2004 +0000 31418 31419 Apply Debian patch to re-POST via VBE if driver thinks it has no video 31420 memory (a symptom of incompletely coming up from D3 power state), which 31421 fixes some Dell laptops (closes: #1890). 31422 31423commit 2b8ab42b05b32710fa4f71a25e2f84192e637fbd 31424Author: Michel Daenzer <michel@daenzer.net> 31425Date: Mon Dec 6 06:27:53 2004 +0000 31426 31427 When direct rendering is enabled, use hostdata blits to transfer data from 31428 system memory to video RAM, which should reduce CPU usage especially 31429 with larger videos. Can be disabled via Option "DMAForXv" if there 31430 should be any stability issues, but it's been stable for me during a 31431 week of testing. Based on a patch by Nikolaus Meine 31432 <meine@tnt.uni-hannover.de>. Probably fix endianness issues in some 31433 newer XVideo code, untested. 31434 Also use hostdata blits to transfer RENDER image data to video RAM to avoid 31435 idling the accelerator engine. Increases RENDER performance 31436 significantly for me. 31437 These changes were only tested on an M9 in a Titanium PowerBook but should 31438 work with all Radeons where direct rendering is supported. 31439 31440commit 960d15ae3a44efe7c02a3d6d0acbabe63e07be74 31441Author: Daniel Stone <daniel@fooishbar.org> 31442Date: Sat Dec 4 22:34:54 2004 +0000 31443 31444 Disable interrupt handling after the DRM handler has been removed, 31445 preventing lockups (closes: #1886). 31446 Allow pitches with a granularity of 32, rather than 64, because some 31447 clamshell iBooks have 800-byte pitches. Change Xv allocation 31448 granularity to 8 bytes, rather than 16 bytes, also (closes: #1888). 31449 31450commit 4706dfb7804a46e3cc5aafc3644f5a1134ab3d19 31451Author: Daniel Stone <daniel@fooishbar.org> 31452Date: Sat Dec 4 22:24:06 2004 +0000 31453 31454 Retry CCE idle commands until the (higher) timeout is reached, and properly 31455 stop the engine if we totally fail -- gets rid of serious performance 31456 issues and lockups with r128 DRI. 31457 31458commit bce508a288d6b95708dbbfc6b235e2d95b6b96ad 31459Author: Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk> 31460Date: Sat Dec 4 00:42:59 2004 +0000 31461 31462 Encoding of numerous files changed to UTF-8 31463 31464commit e9c9fd279ed43cc0d779266efcea69a67541c794 31465Author: Egbert Eich <eich@suse.de> 31466Date: Tue Nov 30 08:38:44 2004 +0000 31467 31468 Make Xorg the default server to install. 31469 Avoid PIO access on IA64. Some IA64 machine check if legacy ports outside 31470 the VGA range are accessed. The ATi driver however does this to probe 31471 for ISA Mach8/32/64. Since no IA64 has ISA slots this restriction 31472 should not be relevant to the user. 31473 Avoid recursive calls of xf86scanpci(). This function normally detects that 31474 it has been called before by checking if the PCI structure is filled 31475 out. So far if this was not the case (because PCI probing has failed 31476 for some reason) the function is traversed again. With the chipset 31477 specific PCI bus probing this can lead to an endless recursive loop as 31478 the post-probing code calls xf86scanpci() from within this function. 31479 The OS specific PCI code for Linux worked only if bus 0 was populated as it 31480 checked for the presence of /proc/bus/pci/00. Fixed to check for 31481 /proc/bus/pci/<bus_to_look_for> instead. 31482 31483commit fffbcb761c178e025e3d55d428a803b1284481e0 31484Author: Matthieu Herrb <matthieu.herrb@laas.fr> 31485Date: Sat Nov 27 15:14:15 2004 +0000 31486 31487 xc/programs/Xserver/hw/xfree86/driver/ati/radeon.man 31488 xc/programs/Xserver/hw/xfree86/driver/voodoo/voodo.man 31489 xc/programs/Xserver/hw/xfree86/driver/wsfb/wsfb.man XF86Config -> 31490 __xconfigfile__ 31491 31492commit ad6347d187f9afa053c755aa3a2da9d821916cb2 31493Author: Daniel Stone <daniel@fooishbar.org> 31494Date: Tue Nov 23 21:27:43 2004 +0000 31495 31496 Use fp_horz_stretch instead of fp_vert_stretch for horizontal stretch 31497 factor calculation in panel size derivation in the Radeon driver 31498 (closes: #1881). 31499 31500commit ac054e427c6ce8cd5b89459dd3f16ca20e6b2d6a 31501Author: Vladimir Dergachev <volodya@mindspring.com> 31502Date: Sun Oct 24 18:17:36 2004 +0000 31503 31504 Add registers to facilitate work with stereo 3d. 31505 31506commit 445ed4b9b61347c866ee67dc0cb110904d667b84 31507Author: Alex Deucher <agd5f@yahoo.com> 31508Date: Fri Oct 22 03:51:58 2004 +0000 31509 31510 - make the overlay work on crtc2 again 31511 - make PutVideo() MergedFB aware 31512 31513commit 54cffc3fe273cf3c9ed2a8aef833d4bd337a8a09 31514Author: Egbert Eich <eich@suse.de> 31515Date: Mon Oct 18 15:19:44 2004 +0000 31516 31517 Fixing one of the infamous 'control reaches end of non void function 31518 problems'. 31519 31520commit 7a97d363327e0b2c8646d961e00d05f99766b54f 31521Author: Egbert Eich <eich@suse.de> 31522Date: Mon Oct 18 14:21:45 2004 +0000 31523 31524 Made handling of DevelDrivers for x86-64 more conformant to other 31525 platforms. 31526 Compress all font encodings (Stefan Dirsch). 31527 Fixed warnings. 31528 Turn on forwarding XNSpotLocation event to XIM server in OffTheSpot and 31529 Root mode (bugzilla #1580, James Su). 31530 Added another compose key combination for the Euro symbol (Stefan Dirsch). 31531 Added support for UTF-8 in ja_JP, ko_KR and zh_TW locales (Mike Fabian). 31532 Changed default encoding for ru from KOI8-R to ISO8859-5 (Mike Fabian). 31533 This is the encoding that is also used by glibc. We may break other 31534 libcs - lets see who complains. 31535 Added explanation for DESTDIR to install to a different directory than /. 31536 Added some early bailouts to atiprobe if PCI structure pointer is NULL to 31537 prevent sig11. 31538 XV support doesn't depend on 2D acceleration any more. This patch removes 31539 this limitation from the NSC driver. This is a patch that I have 31540 committed to XFree86 a while ago but never ported over to X.Org. 31541 Matthieu Herrb contributed some build fixes. 31542 Fixing SetDPMSTimers() so that DPMS*Time == 0 disables the appropriate 31543 timer. This takes advantage of the fact that TimerSet() with a timeout 31544 argument 0 behaves like TimerCanel(). 31545 Use /dev/xconsole (named pipe) or devpts for system logger (Werner Fink). 31546 Create missing links for backward compatibility to XFree86 (Stefan Dirsch). 31547 Changed comment to mention xorg. 31548 Changed cursor for the 'kill' action from XC_draped_box to XC_Pirate. If 31549 you don't like it we can change it back (original author unkown). 31550 Added 'pic' to the man page rendering command pipeline (Werner Fink). 31551 Added missing return value (Stefan Dirsch, Roland Mainz) 31552 31553commit c7ab340fecb94a89c25d6bc7f686b7c97cd3067f 31554Author: Alex Deucher <agd5f@yahoo.com> 31555Date: Sun Oct 17 22:45:57 2004 +0000 31556 31557 - Fix DFP detection in r128 driver (bug 1386, Klaus Kusche) 31558 - Fix EmulateWheelInertia defaults (bug 1115, Andrew Pimlott) 31559 31560commit ace6e1b7933cb415b8efc41dffa335938f9c89b5 31561Author: Vladimir Dergachev <volodya@mindspring.com> 31562Date: Sun Oct 17 17:57:33 2004 +0000 31563 31564 Modified: 31565 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_driver.c On my Inspiron 31566 8600 both internal and external ports are marked DAC_PRIMARY in BIOS. 31567 So be extra careful - only swap when the first port is not DAC_PRIMARY 31568 31569commit 22405e81b812c239b5d8653b402b5f846e6e10b1 31570Author: Alex Deucher <agd5f@yahoo.com> 31571Date: Sun Oct 17 16:52:02 2004 +0000 31572 31573 - Add Thomas' latest sis MergedFB enhancements to the radeon driver 31574 31575commit 186efa31b2cb7285717678bb1c4df4cd9e182395 31576Author: Vladimir Dergachev <volodya@mindspring.com> 31577Date: Wed Oct 13 23:02:42 2004 +0000 31578 31579 Modified: 31580 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_video.c 31581 xc/programs/Xserver/hw/xfree86/drivers/i2c/fi1236.h Squash annoying warning 31582 about fi1236_dump_status 31583 31584commit 43b24861e36023613a99627c57ac44721ed1a548 31585Author: Egbert Eich <eich@suse.de> 31586Date: Tue Oct 12 19:13:43 2004 +0000 31587 31588 Set fbdev mode as the default mode on PPC (Olaf Hering). 31589 Added support for IBM space saver keyboard (Stefan Dirsch). Added support 31590 for Cherry CyMotion Master XPress (Marcus Schaefer). 31591 Change order of SetDisplayDevice(), HWRestore(), UnbindGART() and 31592 RestoreBIOSMemSize() to be exactly opposite to the Save procedure in 31593 EnterVT() (Matthias Hopf, Alan Hourihane). 31594 Fix text mode restauration by removing the assumption that the register 31595 which determines which head is programmed is set. to the active head by 31596 the BIOS (Mark Vojkovich). 31597 When I wrote the resource code 5 years ago I made some assumptions which 31598 turned out to be false: I've assumed that the bus number of the PCI 31599 hostbridge would be the PCI bus the bridge links to. This is not 31600 correct. Fixing this assumption is not easy. However I hope that the 31601 attached patch will make the system work regardless as it 'ignores' 31602 host bridges from which the target bus is not known. This should not 31603 matter at all as we really don't care about host bridges (unless we 31604 have bridge specific code which retrieves information about the 31605 bridge). 31606 Fixed server crash on reset when a structure allocated in PreInit() was 31607 freed on CloseScreen(). 31608 Fixed ring buffer lock ups that happened because the structure that 31609 contained ringbuffer data was not zeroed after allocation. 31610 Fixed numerous warnings due to signed unsigned comparisons. 31611 programs/Xserver/hw/xfree86/drivers/nv/nv_driver.c: 31612 (NVBacklightEnable): Changed the order in which the sequencer registers 31613 and the backlight control registers are written. The sequencer control 31614 register need to be written first otherwise DPMS screen blanking 31615 produces vertical bars on a mobile device. lib/Xau/Imakefile: Build 31616 libXau static library PIC so it can be linked into toolkits that would 31617 like to wrap its functionality. 31618 31619commit 515b7c4b748f70c17aee149a33ee1b41632daa1a 31620Author: Alex Deucher <agd5f@yahoo.com> 31621Date: Mon Oct 11 23:20:26 2004 +0000 31622 31623 - fix resume on radeons (bug 1220) 31624 - clean up some old cruft in the savage i2c code 31625 31626commit 07e1aaee862c4835601c704e14fcb12e28f4e78e 31627Author: Egbert Eich <eich@suse.de> 31628Date: Mon Oct 11 09:58:04 2004 +0000 31629 31630 Improving DPMS handling on VT swich and server termination/abort: previous 31631 version called the driver directly and too late. 31632 Unblank secondary screen explicitely. Don't rely on the value read during 31633 register save as the BIOS have blanked the secondary head. 31634 Checking if server isn't switched away before calling sync. Sanity check 31635 for possible bugs in aother areas of the code. 31636 Fixing default amount of of allocated video memory from AGP for i810: Use 31637 16MB if less than 192MB are installed else use 24MB (Matthias Hopf). 31638 31639commit a8c77480e7f4ddf8c45bea3ecec7b6564c0d6452 31640Author: Vladimir Dergachev <volodya@mindspring.com> 31641Date: Fri Oct 8 00:56:05 2004 +0000 31642 31643 Modified: 31644 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_video.c Remove redundant 31645 and incorrect line. 31646 31647commit ec831d4fba23dc5f4d0a1450aa63efe5ff5937a2 31648Author: Vladimir Dergachev <volodya@mindspring.com> 31649Date: Tue Oct 5 01:16:01 2004 +0000 31650 31651 Modified: 31652 xc/programs/Xserver/hw/xfree86/drivers/ati/theatre.h Patch from Matthieu 31653 Herrb - fix compilation with DoLoadableServer=NO 31654 31655commit f1dadde8f642b332a2dcd2e2a5779f25fec7a3de 31656Author: Adam Jackson <ajax@nwnk.net> 31657Date: Mon Oct 4 21:52:47 2004 +0000 31658 31659 Correct radeon man page, IGP is supported for 3D but R400 is not. 31660 31661commit e3dba9841521466629828501c4fdccbba3d6af89 31662Author: Vladimir Dergachev <volodya@mindspring.com> 31663Date: Mon Oct 4 15:24:34 2004 +0000 31664 31665 Modified: 31666 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_video.c Do not probe for 31667 Rage Theatre if no multimedia table is present and no parameters have 31668 been specified. 31669 31670commit 4aadb784928ebbd60dce172c9a9a80e8d84e943f 31671Author: Vladimir Dergachev <volodya@mindspring.com> 31672Date: Mon Oct 4 14:53:11 2004 +0000 31673 31674 Modified: 31675 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_video.c This is really 31676 embarassing - I found a part of code that had <cr> at the end of the 31677 lines. I am certain I never saw it before - talk about code rot ! 31678 31679commit db5a27081d9d3ab4cc1481a7d110d11103b3de19 31680Author: Vladimir Dergachev <volodya@mindspring.com> 31681Date: Sun Oct 3 23:54:40 2004 +0000 31682 31683 Modified: 31684 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_mm_i2c.c Remove debugging 31685 statement that cause loading of DLL modules to fail. 31686 31687commit 37a8d8a53f12f829783011c1e077fefb5729e618 31688Author: Vladimir Dergachev <volodya@mindspring.com> 31689Date: Sun Oct 3 14:41:45 2004 +0000 31690 31691 Modified: 31692 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_video.c Fix compilation 31693 with gcc 3.4.x (patch by Ronny V. Vindenes) 31694 31695commit 6dbcd503fe21c6dec27c2b97adc0f39ba0a7ab2a 31696Author: Vladimir Dergachev <volodya@mindspring.com> 31697Date: Sun Oct 3 03:07:17 2004 +0000 31698 31699 Modified: 31700 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_video.[c,h] 31701 xc/programs/Xserver/hw/xfree86/drivers/ati/theatre.c 31702 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_reg.h Port the rest of 31703 GATOS Radeon-specific code. Remove "checkpoint" noise from logs. Test 31704 with AIW Radeon 7500 31705 31706commit 7cb911291d459f08f54444d09081d3378fe73b8b 31707Author: Vladimir Dergachev <volodya@mindspring.com> 31708Date: Sun Oct 3 00:01:14 2004 +0000 31709 31710 Modified: 31711 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon.h 31712 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_driver.c 31713 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_video.[c,h] 31714 xc/programs/Xserver/hw/xfree86/drivers/ati/theatre.c Detect and initialize 31715 Rage Theatre (RT100) 31716 31717commit 6c1996174b9f7afc8f4e33c35e4864b31696d4d6 31718Author: Vladimir Dergachev <volodya@mindspring.com> 31719Date: Sat Oct 2 20:55:42 2004 +0000 31720 31721 Modified: 31722 xc/programs/Xserver/hw/xfree86/drivers/ati/Imakefile 31723 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_reg.h 31724 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_video.[c,h] 31725 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_vip.c Add VIP bus access 31726 code, hook it into initialization system. 31727 31728commit 136649a1c2065624a9cf17386054c6595aa6d85e 31729Author: Vladimir Dergachev <volodya@mindspring.com> 31730Date: Sat Oct 2 20:35:04 2004 +0000 31731 31732 Modified: 31733 xc/programs/Xserver/hw/xfree86/drivers/ati/theatre* 31734 xc/programs/Xserver/hw/xfree86/drivers/ati/generic_bus.h Add Rage Theatre 31735 (RT100) module code and hook it into the build system 31736 31737commit 7a17b1c294fa8c7af0b1cdea0618cb6060479356 31738Author: Vladimir Dergachev <volodya@mindspring.com> 31739Date: Sat Oct 2 01:59:13 2004 +0000 31740 31741 Modified: 31742 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon.h 31743 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_video.c 31744 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_video.h 31745 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_mm_i2c.c Add code to 31746 access and initialize multimedia i2c bus. Hook it up. 31747 31748commit e0123e24b6895dde7b385528da24c4e6e9d5a4cb 31749Author: Vladimir Dergachev <volodya@mindspring.com> 31750Date: Sat Oct 2 01:36:43 2004 +0000 31751 31752 Add more multimedia registers. 31753 31754commit 576f53be4f0fa54bde7167872b7508b98e5c9454 31755Author: Vladimir Dergachev <volodya@mindspring.com> 31756Date: Sat Oct 2 00:43:08 2004 +0000 31757 31758 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_driver.c 31759 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon.h Add TV-in specific 31760 options. Add TV-in specific fields to radeon per-screen structure, add 31761 code to initialize them. Preserve MM_TABLE (instead of whole VBIOS as 31762 was done in GATOS drivers) 31763 31764commit 9a9ba82172428b09312fdd6e11da04b21c8529fc 31765Author: Adam Jackson <ajax@nwnk.net> 31766Date: Wed Sep 22 04:38:03 2004 +0000 31767 31768 Bug #1252, #1253, #1255, #1256: Various typo fixes from Dave Jones. 31769 31770commit 700b0dc5148d0170c9223765d208ccef447242df 31771Author: Adam Jackson <ajax@nwnk.net> 31772Date: Tue Sep 21 18:10:34 2004 +0000 31773 31774 Convert the ATI drivers to ANSI function prototypes. 31775 31776commit 17e5f3c1169066be1004306d12b5398cf01cbd35 31777Author: Eric Anholt <anholt@freebsd.org> 31778Date: Fri Sep 17 05:16:45 2004 +0000 31779 31780 Bugzilla #1170: Fix an overcounting of DMA in BEGIN_RING that wasn't 31781 harmful. Bugzilla #1169: Add counting for 31782 BEGIN_RING/OUT_RING/ADVANCE_RING to avoid these sorts of problems in 31783 the future. 31784 31785commit 29bbee6b8a8dd1b883e07547691ff8e1289459ce 31786Author: Adam Jackson <ajax@nwnk.net> 31787Date: Fri Sep 17 03:04:50 2004 +0000 31788 31789 Bug #1192: Remove cfb support from drivers where its use is an option. 31790 Delete xf24_32bpp, as s3virge was the last user. Fix up some comments 31791 to refer to fb rather than cfb. 31792 31793commit cec06d4126f5f31d14ace73798c71ec75e544c45 31794Author: Alex Deucher <agd5f@yahoo.com> 31795Date: Thu Sep 16 22:15:05 2004 +0000 31796 31797 Update comment 31798 31799commit dbeb8d28ac3469f4a7d7d8dae979ae876d4cdcba 31800Author: Kevin E Martin <kem@kem.org> 31801Date: Fri Sep 3 23:26:19 2004 +0000 31802 31803 Updated release notes for bug #999. 31804 Update Radeon man page to note that RenderAccel is now enabled by default 31805 on certain chips. 31806 31807commit c829f667c8a79bbdd40e1f924d57638e7bd425dc 31808Author: Michel Daenzer <michel@daenzer.net> 31809Date: Mon Aug 30 04:31:50 2004 +0000 31810 31811 Hotfix for Eric Anholt's fix for bug #1156 : Clear the 31812 RADEON_NONSURF_AP0_SWP_16BPP bit for 4 bpp data, otherwise it breaks at 31813 depth 16. 31814 31815commit 677246740b5b538e73a7fbde661ec6a97319430b 31816Author: Adam Jackson <ajax@nwnk.net> 31817Date: Fri Aug 27 20:31:26 2004 +0000 31818 31819 Back out my last change, dupe. 31820 31821commit a8eaa15264ff0db1596d653e91260c9a39a8b5ee 31822Author: Adam Jackson <ajax@nwnk.net> 31823Date: Fri Aug 27 18:32:40 2004 +0000 31824 31825 Bug #1061 (documentation-only update): Document the RenderAccel option in 31826 the radeon manpage. 31827 31828commit 360dab062f895c19e1784146e6684f75a20c5b24 31829Author: Eric Anholt <anholt@freebsd.org> 31830Date: Fri Aug 27 03:36:28 2004 +0000 31831 31832 Bug #1156: Fix the build on PPC by splitting the byte-order-setting code 31833 into a separate routine. While here, fix the byte swapping in the 32bpp 31834 case. (Michel Daenzer) 31835 31836commit 9b19bbb4d37b79c97365aa492cd93931d492b3a8 31837Author: Eric Anholt <anholt@freebsd.org> 31838Date: Fri Aug 27 03:23:47 2004 +0000 31839 31840 Bug #1157: Fix Radeon line acceleration in the DMA case (Michel Daenzer). 31841 31842commit 838488b6b46359334e5e781589dbee3832909628 31843Author: Kevin E Martin <kem@kem.org> 31844Date: Fri Aug 27 02:38:56 2004 +0000 31845 31846 Fix Radeon compilation without Xinerama (Bug #1179, Simon Schubert). 31847 31848commit 5f54d4f45e2adc6b15e98152db88743c243aa128 31849Author: Kevin E Martin <kem@kem.org> 31850Date: Wed Aug 25 00:30:41 2004 +0000 31851 31852 Fix hangs during startup with Radeon driver (Bug #1007, Sjoerd Simons). 31853 Fix i810 driver startup on Sony VAIO laptop (Bug #1084, Dave Airlie). 31854 Fix several library build problem (Bug #1147). 31855 Fix build problems (Bugs #1148 and #1149). 31856 31857commit dde880e876f455f1fd66efbcc055e177886397c9 31858Author: Eric Anholt <anholt@freebsd.org> 31859Date: Tue Aug 17 20:22:55 2004 +0000 31860 31861 Reenable Radeon Render acceleration now that interactions with 3D clients 31862 should be resolved. 31863 31864commit 279efef503fdbd442a9bd00f0073d9f6006d9e61 31865Author: Eric Anholt <anholt@freebsd.org> 31866Date: Sun Aug 15 00:44:57 2004 +0000 31867 31868 Fix unresolved symbol warnings in r128 driver (Adam Jackson) 31869 31870commit 955f2b89f94aaa20b5eafd045b01e3eeef3d29ad 31871Author: Kevin E Martin <kem@kem.org> 31872Date: Thu Aug 12 05:00:22 2004 +0000 31873 31874 Fix context switching on Radeons and build problem when XF86DRI is not 31875 defined (Bug #922, Eric Anholt). 31876 31877commit 8f3ea53f26493a5f02410297a587d1d95c067094 31878Author: Adam Jackson <ajax@nwnk.net> 31879Date: Thu Aug 12 02:23:46 2004 +0000 31880 31881 oops, these didn't added somehow. 31882 31883commit 6ec74414bd33e9d8dfd5119058d15f3a6206e9d9 31884Author: Adam Jackson <ajax@nwnk.net> 31885Date: Thu Aug 12 01:03:06 2004 +0000 31886 31887 Factor out R{128,ADEON}Chipsets into their own headers so they can be 31888 sanely included in both the ati core and the r128/radeon submodules; 31889 r128 and radeon can now be directly loaded under dlloader. Teach imstt 31890 about the dlloader- frendly way to talk to fbdevhw. (Partial for bugs 31891 #393 and #400.) 31892 31893commit a5b1e4dd7adacd17b63582e477df3a585efb249a 31894Author: Adam Jackson <ajax@nwnk.net> 31895Date: Wed Aug 11 01:20:28 2004 +0000 31896 31897 Bug #978: Silence more unresolved symbol warnings from the ati driver. 31898 Patch from Stefan Dirsch. 31899 31900commit 1356884132491010a2e41eb2b5be4a227b9b8138 31901Author: Kevin E Martin <kem@kem.org> 31902Date: Wed Aug 11 01:14:14 2004 +0000 31903 31904 Fix Radeon build problem with #define DoLoadableServer NO 31905 31906commit f5c93cb015e334975c2797a23b68798372a284c9 31907Author: Adam Jackson <ajax@nwnk.net> 31908Date: Tue Aug 10 04:01:42 2004 +0000 31909 31910 Bug #978: Silence unresolved symbol warnings from the ati driver. Reported 31911 by Michel Daenzer. 31912 31913commit a2da1caf5848a78b10afd99811fd1a5677af6786 31914Author: Kevin E Martin <kem@kem.org> 31915Date: Mon Aug 9 22:37:22 2004 +0000 31916 31917 Fix install problem on platforms not using xorg.cf/xfree86.cf (Bug #339, 31918 Harold L. Hunt II, Alexander Gottwald). 31919 Fix crash when using X core font in zh_CN.UTF-8 locale (Bug #368, Yu Shao, 31920 David Dawes). 31921 Fix glXMakeCurrent(Dpy, None, NULL) crash (Bug #719, Adam Jackson). 31922 HP-PA build fix (Bug #828, Guy Martin, Paul Anderson). 31923 Fix SDK build for GATOS and Wacom driver (Bug #829, Bryan Stine). 31924 Fix attempt to read video ROM before enabling it (Bug #843, Ivan 31925 Kokshaysky, Mike A. Harris). 31926 Fix detection of primary adapter (Bug #843, Ivan Kokshaysky, Mike A. 31927 Harris). 31928 Clarify xset man page description of how to use the keyboard repeat rate 31929 settings (Bug #846, Mike A. Harris). 31930 Fix problem where print-screen key would get remapped to sys-req in certain 31931 keymaps, which broke GNOME printscreen functionality (Bug #847, Owen 31932 Taylor). 31933 Fix several render problems: 31934 - MMIO mode support 31935 - Hang on IGP chips 31936 - VT switching hang 31937 - 3D render corruption (Bug #922, Hui Yu). 31938 31939commit 3171664c6a02424736864204b731249097fa7aae 31940Author: Kevin E Martin <kem@kem.org> 31941Date: Fri Aug 6 20:07:35 2004 +0000 31942 31943 Disable Render acceleration in Radeon driver temporarily until it no longer 31944 conflicts with the 3D code. 31945 31946commit 27ef7adf3c1f6c3ca79b42b468d08b8542348cdd 31947Author: Michel Daenzer <michel@daenzer.net> 31948Date: Wed Aug 4 13:17:31 2004 +0000 31949 31950 Use info->ModeReg.surface_cntl to restore byte swapping for the framebuffer 31951 aperture on big endian machines, remove superfluous local variables and 31952 register reads. 31953 Adapt framebuffer aperture byte swapping to texture format before copying 31954 data to offscreen area and restore it afterwards on big endian 31955 machines, fixes Render acceleration there. 31956 reviewed by: Hui Yu <hyu@ati.com>, Kevin E. Martin <kem@freedesktop.org> 31957 31958commit 2efdc9f847fa9ff3096667b16df84817fc65605d 31959Author: Eric Anholt <anholt@freebsd.org> 31960Date: Wed Aug 4 10:05:37 2004 +0000 31961 31962 - Add two new XAA hooks, SetupForCPUToScreenTexture2 and 31963 SetupForCPUToScreenAlphaTexture2. These add a dstFormat argument after 31964 the previous format argument, which the driver needs to use to properly 31965 set up the destination format. Two new arrays are added for the list of 31966 destination formats supported that correspond to the previous format 31967 arrays for sources. 31968 - Make Render acceleration only occur when the new hook for that 31969 acceleration type is supplied and the dst format list is set, along 31970 with the src format list being set. Without knowing the destination 31971 format, the Render acceleration couldn't properly support all the 31972 destinations it might encounter. 31973 - Bump XAA module minor version. 31974 - Update the Radeon Render acceleration to use the new hooks when the XAA 31975 module is sufficiently new. Fix a bug in the src/dst alpha booleans for 31976 ops, and use them to set blend_cntl to support destinations without 31977 alpha. Add missing PICT_a1r5g5b5 texture format, and add list 31978 terminator. (!) 31979 31980commit 1e16c66551eed750330bb74a92208c45a7112cd0 31981Author: Matthieu Herrb <matthieu.herrb@laas.fr> 31982Date: Tue Aug 3 09:42:20 2004 +0000 31983 31984 programs/Xserver/hw/xfree86/drivers/ati/radeon_merged_fb.c Add missing 31985 argument to xf86DrvMsg(). 31986 31987commit fd440a1e2110e05d116207800c123c281fbf5e19 31988Author: Matthieu Herrb <matthieu.herrb@laas.fr> 31989Date: Tue Aug 3 09:38:03 2004 +0000 31990 31991 programs/Xserver/hw/xfree86/drivers/ati/r128_driver.c 31992 programs/Xserver/hw/xfree86/drivers/ati/r128_probe.h 31993 programs/Xserver/hw/xfree86/drivers/ati/radeon_driver.c 31994 programs/Xserver/hw/xfree86/drivers/ati/radeon_probe.h Fix Constification 31995 problems. 31996 31997commit dd0f86c4cb046c79a65144f767e97fd99e7f61c9 31998Author: Kevin E Martin <kem@kem.org> 31999Date: Sun Aug 1 23:22:01 2004 +0000 32000 32001 Fix Radeon crash on server exit or generation 32002 32003commit c64f90b673f5c487dfe059f1636b04a1c2ab7dc2 32004Author: Adam Jackson <ajax@nwnk.net> 32005Date: Sat Jul 31 01:21:18 2004 +0000 32006 32007 Change several LoaderSymbol calls introduced by the bug #400 patch to 32008 *Weak() resolver functions. 32009 32010commit b091b4b074cd61c9067ba95ef016f7a29a38c38d 32011Author: Hui YU <hyu@ati.com> 32012Date: Fri Jul 30 22:20:21 2004 +0000 32013 32014 Support for New radeon chips: R420/M18, R423, RV370/M22, RV380/M24, RS300. 32015 Add special handlings for DELL triple-head server (RV100). Misc. bug 32016 fixes for flat panel, host aperture, etc (Bug #946) 32017 32018commit d04f61bfaf6952e832d52e1cd6014e86435aebd9 32019Author: Adam Jackson <ajax@nwnk.net> 32020Date: Fri Jul 30 20:30:51 2004 +0000 32021 32022 Bug #400 (partial): Driver fixes for the dlloader. When using dlloader, all 32023 framebuffer formats except cfb and the overlay modes should work, and 32024 r128 and radeon need to be loaded from the ati driver (both issues to 32025 be fixed soon). Tested on i740, s3virge, mach64, tdfx, vesa, and vga 32026 drivers. elfloader users shouldn't be affected. 32027 32028commit 6dcb78bb818c3c1c7fe406bee5379f9bef023f4b 32029Author: Alex Deucher <agd5f@yahoo.com> 32030Date: Wed Jul 28 02:35:30 2004 +0000 32031 32032 - add gamma correction support to the radeon video overlay 32033 32034commit 26029dc9378bbd95835d397402bfc9e6b4f1a478 32035Author: Alex Deucher <agd5f@yahoo.com> 32036Date: Mon Jul 26 23:14:37 2004 +0000 32037 32038 - Add Radeon DynamicClocks option 32039 - Add small fixes and clean ups from ati's last code drop (typo_fixes, 32040 remove_fudge, laptop, xvfix) 32041 - fix possible segfault in mga_dri.c (Ryan Underwood) 32042 - Add Xv support to pre-nm2160 neomagic chipsets 32043 32044commit 9f251b16e702da486f824307e72a6f80c6987b5d 32045Author: Adam Jackson <ajax@nwnk.net> 32046Date: Mon Jul 26 19:14:58 2004 +0000 32047 32048 Bug #848: Clean up gatos notice messages. 32049 32050commit 6f1e01ecdb11913fd768b9918f516b8379b7858f 32051Author: Eric Anholt <anholt@freebsd.org> 32052Date: Fri Jul 23 05:36:14 2004 +0000 32053 32054 - Add a new option, BuildDevelDRIDrivers, which controls whether 32055 DevelDRIDrivers is non-empty. DevelDRIDrivers contains those drivers 32056 aren't built by default either due to being unusable on that 32057 architecture (but buildable for testing) or due to security concerns, 32058 as in the mach64 and savage cases. 32059 - Merge the Mach64 DDX's DRI support, under the XF86DRI_DEVEL define, which 32060 is set only when BuildDevelDRIDrivers is set. 32061 - Fix up of the sets of DRI drivers built for various architectures. 32062 - Fix build of unichrome DRI driver, which is now enabled on x86. 32063 32064commit b68322f467b7e524092af15e33ab872abb99b6a2 32065Author: Eric Anholt <anholt@freebsd.org> 32066Date: Thu Jul 22 07:03:04 2004 +0000 32067 32068 Merge and resolve conflicts from DRI/DRM/Mesa 20040721 imports. 32069 32070commit 57de5e285d6271fc4b2bc8a951a98be6308d649d 32071Author: Eric Anholt <anholt@freebsd.org> 32072Date: Fri Jul 16 04:38:54 2004 +0000 32073 32074 Bugzilla #803,804: Fix build in the !XF86DRI case after DRI and Radeon 32075 Render acceleration merges. 32076 32077commit 9b0ec803bd588d9af778bfd307ee6d958e4a1e73 32078Author: Adam Jackson <ajax@nwnk.net> 32079Date: Sat Jul 10 14:22:35 2004 +0000 32080 32081 Bugzilla #849: Fix DRI pagesize assumptions in radeon and r128 drivers. 32082 from mharris@, reviewed and applied to DRI by airlied@. 32083 32084commit ccd95bfb291a0c70272fab6822ef379b83dc298d 32085Author: Egbert Eich <eich@suse.de> 32086Date: Tue Jul 6 14:37:47 2004 +0000 32087 32088 Separated Intel drivers from default DriDrivers to avoid building them on 32089 IA64 (Egbert Eich). 32090 Fixed wrong function prototype (Egbert Eich). 32091 Don't test for generic VGA on IA64 (Egbert Eich). 32092 Fixed a segfault when accessing a structure before verifying the pointer 32093 exists (Egbert Eich). 32094 Added a showcache option for debugging (Egbert Eich). 32095 Increase default video RAM size to 16MB when DRI is enabled and more than 32096 128MB are available (Egbert Eich). Fixed lockups during mode switch. 32097 Problem was introduced when attempting to copy the behavior during 32098 LeaveVT()/EnterVT() but but forgetting to call I810DRILeave() before 32099 I810DRIEnter(). The entire DRILeave()/Enter() scenario has been 32100 commented out as it didn't seem to be necessary (Egbert Eich). 32101 Fix TweakMemorySize() (tested with i855/i865) (Egbert Eich). 32102 increased MAX_DEVICES to 128 (Egbert Eich). 32103 Use OS provided PCI config space access as default method (Egbert Eich). 32104 Added support for Linux 2.6 proc file format. 32105 Fixed unaligned accesses to pieces of the VBE info block. VESA did not 32106 align elements to size (Egbert Eich). 32107 32108commit 6bdf3d36ea3086432af5157540fde4e0aa0c44dc 32109Author: Eric Anholt <anholt@freebsd.org> 32110Date: Sun Jun 20 07:34:55 2004 +0000 32111 32112 Add file missed in last commit: Bug #748: 32113 - Add Render acceleration for Radeon 100 and 200-series cards, enabled by 32114 default. Notable performance gains include 27fps in cairogears instead 32115 of 6fps (compared to 50 with glitz), and my disks are now the 32116 bottleneck in an ls -lR in gnome-terminal. Only supported in DRI mode 32117 because the MMIO submission hangs the card so far, but the code is left 32118 in because it may be supportable soon. 32119 - Add xorg.conf options to disable Render acceleration and to force the 32120 subpixel order in the server (Hui Yu). 32121 Many thanks to Hui Yu at ATI for the code this was based off of, Michel 32122 Daenzer for bugfixes and suggestions, and LinuxFund for sponsoring the 32123 work. 32124 32125commit 8c1b37420ea890f76edb97aed61ee427a871cd3f 32126Author: Eric Anholt <anholt@freebsd.org> 32127Date: Sun Jun 20 07:26:08 2004 +0000 32128 32129 Bug #748: 32130 - Add Render acceleration for Radeon 100 and 200-series cards, enabled by 32131 default. Notable performance gains include 27fps in cairogears instead 32132 of 6fps (compared to 50 with glitz), and my disks are now the 32133 bottleneck in an ls -lR in gnome-terminal. Only supported in DRI mode 32134 because the MMIO submission hangs the card so far, but the code is left 32135 in because it may be supportable soon. 32136 - Add xorg.conf options to disable Render acceleration and to force the 32137 subpixel order in the server (Hui Yu). 32138 Many thanks to Hui Yu at ATI for the code this was based off of, Michel 32139 Daenzer for bugfixes and suggestions, and LinuxFund for sponsoring the 32140 work. 32141 32142commit 6ecf374d500afe6da494dfdd6566396ec65b6d6a 32143Author: Eric Anholt <anholt@freebsd.org> 32144Date: Wed Jun 16 09:43:59 2004 +0000 32145 32146 Merge DRI-trunk-20040613 changes in programs/Xserver/hw/xfree86/drivers, 32147 with the following notes: 32148 - Savage and Mach64 (= ati/ati*.[ch] changes) DRI not merged due to 32149 insecurity. 32150 - VIA driver converted to new drmContext and drmHandle names. 32151 - Radeon driver merge conflicted in many places, and MergedFB at least 32152 could probably use some checking at this point. 32153 32154commit bea8085e04136b0ef513c17bb65c54069ec531e1 32155Author: Eric Anholt <anholt@freebsd.org> 32156Date: Wed Jun 16 09:25:58 2004 +0000 32157 32158 DRI trunk-20040613 import 32159 32160commit 7b588cda922992a8ee2d04853ba1533cf43592a3 32161Author: Eric Anholt <anholt@freebsd.org> 32162Date: Wed Jun 16 09:25:58 2004 +0000 32163 32164 Initial revision 32165 32166commit 6cbb5bbff43d65a762d432659793333868837f0b 32167Author: Eric Anholt <anholt@freebsd.org> 32168Date: Wed Jun 16 09:22:56 2004 +0000 32169 32170 DRI XFree86-4_3_99_12-merge import 32171 32172commit 3586c3c7955365a2af0b8a286ea0d0488ddfd137 32173Author: Egbert Eich <eich@suse.de> 32174Date: Fri Apr 23 19:26:46 2004 +0000 32175 32176 Merging XORG-CURRENT into trunk 32177 32178commit 538768e0f472026325be9dd37dc4bfe7ea7d3b75 32179Author: Egbert Eich <eich@suse.de> 32180Date: Sun Mar 14 08:33:20 2004 +0000 32181 32182 Importing vendor version xf86-4_4_99_1 on Sun Mar 14 00:26:39 PST 2004 32183 32184commit 5f179e796d1cd310790d06eebb474ad67114ec84 32185Author: Egbert Eich <eich@suse.de> 32186Date: Wed Mar 3 12:12:18 2004 +0000 32187 32188 Importing vendor version xf86-4_4_0 on Wed Mar 3 04:09:24 PST 2004 32189 32190commit 493c44580191beff63575c4032f772fbbf1b06eb 32191Author: Egbert Eich <eich@suse.de> 32192Date: Thu Feb 26 13:35:52 2004 +0000 32193 32194 readding XFree86's cvs IDs 32195 32196commit 942a47fd3c5e8df85642672c01072bf97285bc30 32197Author: Egbert Eich <eich@suse.de> 32198Date: Thu Feb 26 09:23:18 2004 +0000 32199 32200 Importing vendor version xf86-4_3_99_903 on Wed Feb 26 01:21:00 PST 2004 32201 32202commit dda79b5ee3eb49f924a3213592d93ba981c99394 32203Author: Kaleb Keithley <kaleb@freedesktop.org> 32204Date: Mon Feb 23 20:35:02 2004 +0000 32205 32206 Import most of XFree86 4.4RC3. This import excludes files which have the 32207 new license. If we want to, later we can import 4.4RC3 again and pick 32208 up the files that have the new license, but for now the vendor branch 32209 is "pure." 32210 32211commit 08104f18ac497561c969694ff709da6c361ccea7 32212Author: Egbert Eich <eich@suse.de> 32213Date: Thu Jan 29 08:08:34 2004 +0000 32214 32215 Importing vendor version xf86-012804-2330 on Thu Jan 29 00:06:33 PST 2004 32216 32217commit 162d76b9c48ed17d6e645057db58c9f13436f25f 32218Author: Kaleb Keithley <kaleb@freedesktop.org> 32219Date: Fri Dec 19 20:55:06 2003 +0000 32220 32221 XFree86 4.3.99.902 (RC 2) 32222 32223commit 195772ec74f7a9b7d5bf294b3bc2c3b960e08a31 32224Author: Kaleb Keithley <kaleb@freedesktop.org> 32225Date: Thu Dec 4 22:03:16 2003 +0000 32226 32227 XFree86 4.3.99.901 (RC 1) 32228 32229commit 73105e438e90f950249cb265ba99001d4db3dc60 32230Author: Kaleb Keithley <kaleb@freedesktop.org> 32231Date: Tue Nov 25 19:28:36 2003 +0000 32232 32233 XFree86 4.3.99.16 Bring the tree up to date for the Cygwin folks 32234 32235commit 0a53864f9581196604d04fd28b4e94a9b0b73d18 32236Author: Kaleb Keithley <kaleb@freedesktop.org> 32237Date: Tue Nov 25 19:28:36 2003 +0000 32238 32239 Initial revision 32240 32241commit c56aaef79028261e0a3be0c86ffd3f16b6bdd9ba 32242Author: Kaleb Keithley <kaleb@freedesktop.org> 32243Date: Fri Nov 14 16:48:55 2003 +0000 32244 32245 XFree86 4.3.0.1 32246 32247commit d09f463a5d1ce73e0b65d5276fbcca393fa2da46 32248Author: Kaleb Keithley <kaleb@freedesktop.org> 32249Date: Fri Nov 14 16:48:55 2003 +0000 32250 32251 Initial revision 32252 32253commit d9af6dc32652502d84ea8da5d57a5ab45429c4ad 32254Author: Kaleb Keithley <kaleb@freedesktop.org> 32255Date: Fri Nov 14 15:54:49 2003 +0000 32256 32257 R6.6 is the Xorg base-line 32258