ChangeLog revision f395c03e
1commit d269962724b57e21142c9e1c6631a45a46d87b47
2Author: Matt Turner <mattst88@gmail.com>
3Date:   Tue Jan 17 14:44:51 2017 -0800
4
5    xf86-video-siliconmotion 1.7.9
6    
7    Signed-off-by: Matt Turner <mattst88@gmail.com>
8
9commit e67f3949648c20e16ac756ae28ea4a2b4a1d0f4c
10Author: Adam Jackson <ajax@redhat.com>
11Date:   Tue Jul 19 10:03:56 2016 -0400
12
13    Adapt Block/WakeupHandler signature for ABI 23
14    
15    Signed-off-by: Adam Jackson <ajax@redhat.com>
16
17commit 647e5a93d9387e3bc8d2c8569e3a5f996acfb169
18Author: Thomas Klausner <wiz@NetBSD.org>
19Date:   Mon Sep 14 18:36:32 2015 +0200
20
21    Use X11/Xarch.h for endian detection.
22    
23    Reviewed-by: Connor Behan <connor.behan@gmail.com>
24
25commit a4e406c20ab0590214f779c2319ac7b92594a606
26Author: Connor Behan <connor.behan@gmail.com>
27Date:   Sun Aug 23 11:46:37 2015 -0400
28
29    Check for drawables before compositing
30    
31    This should prevent a crash when something like Cairo tries to composite
32    a solid picture.
33    
34    Signed-off-by: Connor Behan <connor.behan@gmail.com>
35
36commit 70d3d3a384651e5d22840a0de8644eb8b10c5775
37Author: Julien Cristau <jcristau@debian.org>
38Date:   Sat May 9 16:56:51 2015 +0200
39
40    Bump to 1.7.8
41    
42    Signed-off-by: Julien Cristau <jcristau@debian.org>
43
44commit b22c590a546b2ea5f256e7654c4a6b165b320e80
45Author: Julien Cristau <jcristau@debian.org>
46Date:   Fri May 8 13:40:46 2015 +0200
47
48    Fix a couple const-ness warnings
49    
50    smi_accel.c: In function 'SMI_GEReset':
51    smi_accel.c:51:1: warning: passing argument 4 of 'SMI_GEReset' discards 'const' qualifier from pointer target type
52      WaitIdle();
53     ^
54    smi_accel.c:38:1: note: expected 'char *' but argument is of type 'const char *'
55     SMI_GEReset(ScrnInfoPtr pScrn, int from_timeout, int line, char *file)
56     ^
57    [...]
58    smi_driver.c: In function 'SMI_DetectPanelSize':
59    smi_driver.c:1011:12: warning: assignment discards 'const' qualifier from pointer target type
60         if ((s = xf86GetOptValString(pSmi->Options, OPTION_PANEL_SIZE)) != NULL) {
61                ^
62    
63    Signed-off-by: Julien Cristau <jcristau@debian.org>
64
65commit ca1c4c19d688d215b52f4e7e2a634c0cbcfff3c1
66Author: Julien Cristau <jcristau@debian.org>
67Date:   Fri May 8 13:41:01 2015 +0200
68
69    Fix SwitchMode and AdjustFrame hooks
70    
71    smi_driver.c: In function 'SMI_Probe':
72    smi_driver.c:360:24: warning: assignment from incompatible pointer type
73       pScrn->SwitchMode    = SMI_SwitchMode;
74                            ^
75    smi_driver.c:361:24: warning: assignment from incompatible pointer type
76       pScrn->AdjustFrame   = SMI_AdjustFrame;
77                            ^
78    
79    Signed-off-by: Julien Cristau <jcristau@debian.org>
80
81commit d06454e89384957fa2b77ff35b35108dc657a510
82Author: Julien Cristau <jcristau@debian.org>
83Date:   Fri May 8 13:27:56 2015 +0200
84
85    Fix BlockHandler declaration
86    
87    smi_video.c: In function 'SMI_SetupVideo':
88    smi_video.c:940:24: warning: assignment from incompatible pointer type
89         pSmi->BlockHandler = pScreen->BlockHandler;
90                            ^
91    
92    Signed-off-by: Julien Cristau <jcristau@debian.org>
93
94commit 6073828547bdff86d147ab54accb12bc4230d6aa
95Author: Julien Cristau <jcristau@debian.org>
96Date:   Fri May 8 13:26:00 2015 +0200
97
98    Fix shadow compiler warning
99    
100    smi_video.c: In function 'SetAttrSAA7111':
101    smi_video.c:795:6: warning: declaration of 'i' shadows a parameter [-Wshadow]
102      int i;
103          ^
104    smi_video.c:723:39: warning: shadowed declaration is here [-Wshadow]
105     SetAttrSAA7111(ScrnInfoPtr pScrn, int i, int value)
106                                           ^
107    
108    Signed-off-by: Julien Cristau <jcristau@debian.org>
109
110commit c31d7f853d7469085f96f1e37923c260884c611c
111Author: Adam Jackson <ajax@redhat.com>
112Date:   Tue Sep 30 10:34:07 2014 -0400
113
114    Remove dependency on xf86PciInfo.h
115    
116    Signed-off-by: Adam Jackson <ajax@redhat.com>
117
118commit 08d459d2b548ce89264f45c7018f1cda9f08d795
119Author: Adam Jackson <ajax@redhat.com>
120Date:   Tue Sep 30 10:31:52 2014 -0400
121
122    Fix a typo
123    
124    Signed-off-by: Adam Jackson <ajax@redhat.com>
125
126commit b0b287209cfba5c64c7584b1a82ed2a2f1eab7e6
127Author: Adam Jackson <ajax@redhat.com>
128Date:   Tue Sep 30 10:28:48 2014 -0400
129
130    Fix build against xserver 1.17
131    
132    Signed-off-by: Adam Jackson <ajax@redhat.com>
133
134commit 9b563415326e02f0b89f716c29b2fc22f393fb96
135Author: Adam Jackson <ajax@redhat.com>
136Date:   Wed May 21 14:01:02 2014 -0400
137
138    Use own thunk function instead of vgaHW*Weak
139    
140    I plan to remove the Weak functions from a future server.
141    
142    Signed-off-by: Adam Jackson <ajax@redhat.com>
143
144commit f19d7e463c30f1364e82e8c9f87b8a8407d53680
145Author: Adam Jackson <ajax@redhat.com>
146Date:   Wed Jan 9 22:59:39 2013 -0500
147
148    Remove miInitializeBackingStore()
149    
150    Signed-off-by: Adam Jackson <ajax@redhat.com>
151
152commit 2beb25bd41508ab5e2af39c696485fcb5f0ca7d6
153Author: Dave Airlie <airlied@redhat.com>
154Date:   Tue Jul 17 16:05:09 2012 +1000
155
156    xf86-video-siliconmotion: bump to version 1.7.7
157    
158    Signed-off-by: Dave Airlie <airlied@redhat.com>
159
160commit 4b4b8e9c047bd34c516ca8ab2b15fd1c541d05d5
161Author: Dave Airlie <airlied@redhat.com>
162Date:   Tue Jul 17 14:53:30 2012 +1000
163
164    smi: convert to new X server screen conversion apis.
165    
166    Signed-off-by: Dave Airlie <airlied@redhat.com>
167
168commit 1aad7b44e3e141bd8d3e589b7cd410e791050bd0
169Author: Dave Airlie <airlied@redhat.com>
170Date:   Tue Jul 17 02:01:07 2012 +0100
171
172    smi: port to compat-api for new server.
173    
174    Signed-off-by: Dave Airlie <airlied@redhat.com>
175
176commit 891ea02395079b1480faf701156185ad01405e6d
177Author: Dave Airlie <airlied@redhat.com>
178Date:   Tue Jul 17 01:52:45 2012 +0100
179
180    smi: fix build against XAA-less server
181
182commit e4660be7c12de53cac57c0b7d4e792a071b4fcfc
183Author: Alan Coopersmith <alan.coopersmith@oracle.com>
184Date:   Fri Mar 23 19:49:14 2012 -0700
185
186    xf86-video-siliconmotion 1.7.6
187    
188    Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
189
190commit 63afc35a662bc2dc460f6ff6fab65a8462aa5e76
191Author: Matt Turner <mattst88@gmail.com>
192Date:   Thu Feb 9 17:14:25 2012 -0500
193
194    Add &component=Driver/siliconmotion to Bugzilla link
195    
196    Signed-off-by: Matt Turner <mattst88@gmail.com>
197
198commit 12695ed1edd8c590252e76d530bcecaa0036e109
199Author: Matt Turner <mattst88@gmail.com>
200Date:   Thu Feb 9 16:21:59 2012 -0500
201
202    Don't check for randrproto or renderproto
203    
204    The driver supports neither randr or render.
205    
206    Signed-off-by: Matt Turner <mattst88@gmail.com>
207
208commit 0e7e9cd639196cb42b0a79b2671f6bacc558035f
209Author: Adam Jackson <ajax@redhat.com>
210Date:   Mon Dec 19 17:26:34 2011 -0500
211
212    Fix for new vgaHW ABI
213    
214    Signed-off-by: Adam Jackson <ajax@redhat.com>
215
216commit 208a703776d6dfbd01babbe2f220a7198dea4f5c
217Author: Adam Jackson <ajax@redhat.com>
218Date:   Mon Dec 19 17:25:39 2011 -0500
219
220    Make failure to XAA non-fatal
221    
222    No shadowfb support in this driver yet.
223    
224    Signed-off-by: Adam Jackson <ajax@redhat.com>
225
226commit 7d9c1a49b87df15cdb78c529db554ea82c77033f
227Author: Adam Jackson <ajax@redhat.com>
228Date:   Thu Nov 17 15:53:51 2011 -0500
229
230    Check ABI major not encoded ABI
231    
232    Signed-off-by: Adam Jackson <ajax@redhat.com>
233
234commit 037b839c84ad6dcbe9aade420cf0d3bf06f23119
235Author: Adam Jackson <ajax@redhat.com>
236Date:   Wed Nov 16 15:03:25 2011 -0500
237
238    Adapt to missing PIOOffset in videoabi 12
239    
240    Signed-off-by: Adam Jackson <ajax@redhat.com>
241
242commit 087226bf4e442b83985078f5df481eb91f4f65ed
243Author: Julien Cristau <jcristau@debian.org>
244Date:   Sun Feb 27 17:12:58 2011 +0100
245
246    Bump to 1.7.5
247
248commit b20cfa3d4f531b612a28ea88e5f5a6b1df183ea0
249Author: Julien Cristau <jcristau@debian.org>
250Date:   Sun Feb 27 15:49:12 2011 +0100
251
252    Replace calls to xalloc/xcalloc/xfree with malloc/calloc/free
253    
254    Also don't check for NULL before calling free().
255    
256    Signed-off-by: Julien Cristau <jcristau@debian.org>
257
258commit 312540ac447acaf80db7e341aa75f90f3e821438
259Author: Julien Cristau <jcristau@debian.org>
260Date:   Sun Feb 27 15:52:33 2011 +0100
261
262    Fix compiler warning on 64bit
263    
264    src/smi_driver.c: In function ‘SMI_MapMem’:
265    src/smi_driver.c:1498: warning: format ‘%08lX’ expects type ‘long unsigned int’, but argument 6 has type ‘CARD32’
266    
267    Signed-off-by: Julien Cristau <jcristau@debian.org>
268
269commit 6008906f0e328180b17e7b604b97dfdfd4280508
270Author: Julien Cristau <jcristau@debian.org>
271Date:   Sun Feb 27 15:46:13 2011 +0100
272
273    Fix build with xserver 1.10
274    
275    ScrnInfo->pixmapPrivate is gone
276    
277    Signed-off-by: Julien Cristau <jcristau@debian.org>
278
279commit 903aac1db82e17954b74581019ebb946d5b43fe3
280Author: Alan Coopersmith <alan.coopersmith@oracle.com>
281Date:   Sat Oct 30 09:41:07 2010 -0700
282
283    Sun's copyrights now belong to Oracle
284    
285    Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
286
287commit 71e61276ac62dfee15b275568084f8b8a55fc661
288Author: Gaetan Nadon <memsize@videotron.ca>
289Date:   Wed Jul 21 16:49:04 2010 -0400
290
291    config: add comments for main statements
292
293commit 5cdb9326ab19da691f4a417def47fd0c869cefb7
294Author: Gaetan Nadon <memsize@videotron.ca>
295Date:   Wed Jul 21 16:07:00 2010 -0400
296
297    config: replace deprecated use of AC_OUTPUT with AC_CONFIG_FILES
298    
299    Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
300
301commit 987c55bc8f69551eea917fba94827ac31c04a49f
302Author: Gaetan Nadon <memsize@videotron.ca>
303Date:   Wed Jul 21 14:37:41 2010 -0400
304
305    config: replace deprecated AC_HELP_STRING with AS_HELP_STRING
306    
307    Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
308
309commit f2569300c0f1b5714b4c4f09d07a16e9434bee00
310Author: Gaetan Nadon <memsize@videotron.ca>
311Date:   Wed Jul 21 14:05:23 2010 -0400
312
313    config: replace deprecated AM_CONFIG_HEADER with AC_CONFIG_HEADERS
314    
315    Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
316
317commit 6b1d1833b38e1d21760174493e4f8f0af4214574
318Author: Gaetan Nadon <memsize@videotron.ca>
319Date:   Wed Jul 21 09:27:42 2010 -0400
320
321    config: complete AC_INIT m4 quoting
322    
323    Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
324
325commit dc67a8e8fe220697c3dcbca63c9ce0608a75888c
326Author: Gaetan Nadon <memsize@videotron.ca>
327Date:   Tue Jul 20 20:24:42 2010 -0400
328
329    config: remove unrequired AC_HEADER_STDC
330    
331    Autoconf says:
332    "This macro is obsolescent, as current systems have conforming
333    header files. New programs need not use this macro".
334    
335    Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
336
337commit 64995ca7f1287493e0616cc6469fdc2465900325
338Author: Gaetan Nadon <memsize@videotron.ca>
339Date:   Tue Jul 20 19:41:31 2010 -0400
340
341    config: remove AC_PROG_CC as it overrides AC_PROG_C_C99
342    
343    XORG_STRICT_OPTION from XORG_DEFAULT_OPTIONS calls
344    AC_PROG_C_C99. This sets gcc with -std=gnu99.
345    If AC_PROG_CC macro is called afterwards, it resets CC to gcc.
346    
347    Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
348
349commit 9520e4e64fa758eef024feda09cb6435e41db705
350Author: Gaetan Nadon <memsize@videotron.ca>
351Date:   Tue Jul 20 18:45:19 2010 -0400
352
353    config: update AC_PREREQ statement to 2.60
354    
355    Unrelated to the previous patches, the new value simply reflects
356    the reality that the minimum level for autoconf to configure
357    all x.org modules is 2.60 dated June 2006.
358    
359    ftp://ftp.gnu.org/gnu/autoconf/autoconf-2.60.tar.gz
360    
361    Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
362
363commit 53cddc508ea0026db04f04772c401010d62d6af3
364Author: Gaetan Nadon <memsize@videotron.ca>
365Date:   Tue Jul 20 16:15:30 2010 -0400
366
367    config: upgrade to util-macros 1.8 for additional man page support
368    
369    Use MAN_SUBST now supplied in XORG_MANPAGE_SECTIONS
370    The value of MAN_SUBST is the same for all X.Org packages.
371
372commit 8087bc23a7662f1426a8bd6febf0d5d3844e6107
373Author: Francisco Jerez <currojerez@riseup.net>
374Date:   Wed Jun 16 14:59:37 2010 +0200
375
376    xv: Make sure we have a CRTC before setting up the overlay.
377    
378    Fixes a segfault when the destination area is off screen.
379    
380    Signed-off-by: Francisco Jerez <currojerez@riseup.net>
381    Tested-by: Krzysztof Halasa <khc@pm.waw.pl>
382
383commit 2de1f7ae0b4441c1501568358acf2b870bb9dcc6
384Author: Alan Coopersmith <alan.coopersmith@oracle.com>
385Date:   Tue Apr 27 15:11:35 2010 -0700
386
387    xf86-video-siliconmotion 1.7.4
388    
389    Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
390
391commit 2dd7cd7508b459638a640905f8793ea85f41af82
392Author: Alan Coopersmith <alan.coopersmith@oracle.com>
393Date:   Tue Apr 27 15:11:02 2010 -0700
394
395    Require xorg-macros 1.4 for XORG_INSTALL
396    
397    Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
398
399commit b6d8a1adec9f780826090fb923c54f16fd7f9ce7
400Author: Alan Coopersmith <alan.coopersmith@oracle.com>
401Date:   Tue Apr 27 15:08:36 2010 -0700
402
403    Use $(AM_V_GEN) & $(SED) to generate man pages
404    
405    Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
406
407commit 7510112f30e33cfb9650ee939564f73ff9633bea
408Author: Gaetan Nadon <memsize@videotron.ca>
409Date:   Thu Feb 11 10:08:07 2010 -0500
410
411    config: move CWARNFLAGS from configure.ac to Makefile.am
412    
413    Compiler warning flags should be explicitly set in the makefile
414    rather than being merged with other packages compiler flags.
415    
416    Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
417
418commit 3f2aaabcdfe339908338dff42aabc1fef47e8694
419Author: Cedric Cellier <rixed@happyleptic.org>
420Date:   Mon Feb 8 11:44:28 2010 +0100
421
422    Fix lack of precision in video resizing. #26443
423    
424    On some videos the last displayed line was wrong. This can
425    be fixed using LynxEM+ VPR68. Code borrowed from siliconmotion's
426    in-house driver.
427    Also fix a typo.
428    
429    Signed-off-by: Cedric Cellier <rixed@happyleptic.org>
430    Signed-off-by: Francisco Jerez <currojerez@riseup.net>
431
432commit 6bf8f8d7e13ccbb773765c4fcdcaf6d1d2993122
433Author: Alan Coopersmith <alan.coopersmith@sun.com>
434Date:   Fri Jan 15 13:50:43 2010 -0800
435
436    Update Sun license notices to current X.Org standard form
437    
438    Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
439
440commit 6650644b51947c0c13e71b0fa01f795e997ab750
441Author: Gaetan Nadon <memsize@videotron.ca>
442Date:   Tue Dec 15 22:01:02 2009 -0500
443
444    configure.ac: remove unused sdkdir=$(pkg-config...) statement
445    
446    The sdkdir variable isn't use, so remove the statement.
447    
448    Acked-by: Dan Nicholson <dbn.lists@gmail.com>
449    
450    Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
451
452commit e8e1d7d22aee2b6aa7caecf910b616a8e39ac061
453Author: Gaetan Nadon <memsize@videotron.ca>
454Date:   Mon Nov 23 09:25:05 2009 -0500
455
456    Makefile.am: add ChangeLog and INSTALL on MAINTAINERCLEANFILES
457    
458    Now that the INSTALL file is generated.
459    Allows running make maintainer-clean.
460
461commit 6acfadf1c541626ce445f5d71438c845f196e37f
462Author: Gaetan Nadon <memsize@videotron.ca>
463Date:   Wed Oct 28 14:41:41 2009 -0400
464
465    INSTALL, NEWS, README or AUTHORS files are missing/incorrect #24206
466    
467    Automake 'foreign' option is specified in configure.ac.
468    Remove from Makefile.am
469
470commit 638e15adec65c00cf3be68249c8372befe6a12a1
471Author: Gaetan Nadon <memsize@videotron.ca>
472Date:   Wed Oct 28 14:09:09 2009 -0400
473
474    INSTALL, NEWS, README or AUTHORS files are missing/incorrect #24206
475    
476    Add missing INSTALL file. Use standard GNU file on building tarball
477    README may have been updated
478    Remove AUTHORS file as it is empty and no content available yet.
479    Remove NEWS file as it is empty and no content available yet.
480
481commit dd3ce955349f7545d9aa2aacbb018811d7b9faa3
482Author: Gaetan Nadon <memsize@videotron.ca>
483Date:   Mon Oct 26 12:54:22 2009 -0400
484
485    Several driver modules do not have a ChangeLog target in Makefile.am #23814
486    
487    The git generated ChangeLog replaces the hand written one.
488    Update configure.ac to xorg-macros level 1.3.
489    Use XORG_DEFAULT_OPTIONS which replaces four XORG_* macros
490    Update Makefile.am to add ChangeLog target if missing
491    Remove ChangeLog from EXTRA_DIST or *CLEAN variables
492    This is a pre-req for the INSTALL_CMD
493
494commit b48c3f38767cdcd71eccfa6a0f4a15501a39f83a
495Author: Gaetan Nadon <memsize@videotron.ca>
496Date:   Thu Oct 22 12:34:18 2009 -0400
497
498    .gitignore: use common defaults with custom section # 24239
499    
500    Using common defaults will reduce errors and maintenance.
501    Only the very small or inexistent custom section need periodic maintenance
502    when the structure of the component changes. Do not edit defaults.
503
504commit 8a4f000f22bbec70193415372cfb478d6e2164d8
505Author: Adam Jackson <ajax@redhat.com>
506Date:   Mon Nov 2 16:03:00 2009 -0500
507
508    Change clockRange to ClockRange, not ClockRanges
509    
510    ClockRanges is a silly type and I want rid of it, and the one extra
511    field it provides that's not in ClockRange, we're not using.
512    
513    Signed-off-by: Adam Jackson <ajax@redhat.com>
514
515commit 52f40e7569bbcc9bce8f5d885b0de2868398e8df
516Author: Dave Airlie <airlied@redhat.com>
517Date:   Wed Aug 5 11:09:26 2009 +1000
518
519    fix build after headers moved
520
521commit 9d9bfe0349693b283e5dc1ced6fa882f10619253
522Author: Dave Airlie <airlied@redhat.com>
523Date:   Wed Aug 5 11:08:15 2009 +1000
524
525    update for 1.7.3
526
527commit 115b564f78b038baef2d14c037785850efb4e9ca
528Author: Dave Airlie <airlied@redhat.com>
529Date:   Tue Jul 28 15:22:41 2009 +1000
530
531    siliconmotion: change to using ABI version check
532
533commit 70f2463faa6d21517d97aa460871e1b5f3fa6505
534Author: Dave Airlie <airlied@redhat.com>
535Date:   Tue Jul 28 13:32:36 2009 +1000
536
537    siliconmotion: update for resources/RAC API removal
538
539commit 841df790b5b6700e9a1d6d41ee04b0b913329084
540Author: Peter Hutterer <peter.hutterer@who-t.net>
541Date:   Thu Jul 16 11:55:05 2009 +1000
542
543    Update to xextproto 7.1 support.
544    
545    DPMS header was split into dpms.h (client) and dpmsconst.h (server). Drivers
546    need to include dpmsconst.h if xextproto 7.1 is available.
547    
548    Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
549
550commit d754e9a2954e56a26a94e58d87a450a825c250e6
551Author: Adam Jackson <ajax@redhat.com>
552Date:   Thu Jul 2 11:43:52 2009 -0400
553
554    siliconmotion 1.7.2
555
556commit cd9e9f25294abb739c1772e3a9a7a9c0a921ee82
557Author: Francisco Jerez <currojerez@riseup.net>
558Date:   Tue Jun 23 13:11:02 2009 +0200
559
560    Fix EXA rotation for SM502 at 32bpp (bug 22432).
561
562commit 97498c048c897e5753e61d3b4ab231025974d67c
563Author: Adam Jackson <ajax@redhat.com>
564Date:   Thu May 28 15:17:58 2009 -0400
565
566    Remove useless loader symbol lists.
567
568commit 536e5df957698251206326edc5a13e833f7c72b1
569Author: Francisco Jerez <currojerez@riseup.net>
570Date:   Wed May 27 01:23:36 2009 +0200
571
572    Program MCLK/MXCLK on startup for a secondary SM502.
573    
574    This should allow using a SM502 as secondary display
575    device (bug 21810).
576
577commit 6a370fa2b6b8fcbd556dd6f9bf92872e9bea23e8
578Author: Alan Coopersmith <alan.coopersmith@sun.com>
579Date:   Mon May 11 22:31:30 2009 -0700
580
581    Fill in COPYING file, add SubmittingPatches URL to README
582
583commit 962e0c4f08180ac51994aef8d6d042108d5841c1
584Author: Francisco Jerez <currojerez@riseup.net>
585Date:   Sun May 10 01:53:05 2009 +0200
586
587    Widen the pixel clock readjust interval for SM712.
588    
589    Set it to 3MHz so that the pixel clock frequency is overridden
590    when it's found to be 49MHz, which is reported to be unstable.
591
592commit 3cc1b9d4bcf1b7a756ea4426c93392ff6d82e9a6
593Author: Francisco Jerez <currojerez@riseup.net>
594Date:   Tue May 5 00:45:10 2009 +0200
595
596    Fail gracefully when the specified virtual screen doesn't fit in RAM.
597
598commit 32628d9884d577d7a672c172c12b7097be276700
599Author: Francisco Jerez <currojerez@riseup.net>
600Date:   Tue May 5 00:44:17 2009 +0200
601
602    Probe the amount of installed memory by trial and error on SM712.
603    
604    In some cases the BIOS hasn't filled in the "scratchpad registers"
605    (SR71) with the right amount of memory installed (e.g. MIPS
606    platform). There seems to be no other way to do it than to test it.
607    
608    This should fix bug 21528.
609
610commit 00921b014fa0b5358c22a769cf2450cbd4bdc8a5
611Author: Francisco Jerez <currojerez@riseup.net>
612Date:   Tue Apr 28 22:24:03 2009 +0200
613
614    Bump version to 1.7.1.
615
616commit abf1ba79f2bfe61f24cfa43deb0400d7c5f95bd5
617Author: Francisco Jerez <currojerez@riseup.net>
618Date:   Tue Apr 21 21:37:45 2009 +0200
619
620    Increase the maximum clock value to 200MHz on SM712.
621    
622    The default MCLK setting was higher than the clock limit, and it
623    failed.
624
625commit 41e5c49024d4e27a7be3da02017000a22b59016f
626Author: Francisco Jerez <currojerez@riseup.net>
627Date:   Tue Apr 21 21:36:10 2009 +0200
628
629    Don't attempt monitor detection on SM712.
630    
631    The databook says nothing about it, and it doesn't work.
632
633commit 6691678b5f04d52a1e08f9eaa6d511df9b7eeb4d
634Author: Jamie Lentin <jm@lentin.co.uk>
635Date:   Thu Mar 19 21:54:12 2009 +0000
636
637    Stop clearing of "VESA compliance power down mode" bit
638    
639    Clearing this bit causes an OQO 01+ w/SMI720 to power down the LCD,
640    leave it alone.
641    
642    Signed-off-by: Jamie Lentin <jm@lentin.co.uk>
643    Signed-off-by: Francisco Jerez <currojerez@riseup.net>
644
645commit 05480382c508bd27c1755405518ce9c339f0f80e
646Author: Matthieu Herrb <matthieu.herrb@laas.fr>
647Date:   Fri Mar 13 21:19:16 2009 +0100
648
649    Fix direct access to IO space on chipsets with no IOBase mapping.
650    
651    Signed-off-by: Francisco Jerez <currojerez@riseup.net>
652
653commit e62807140ede74e262ed70c846017dc3506b511e
654Author: Niels de Vos <niels.devos@wincor-nixdorf.com>
655Date:   Wed Feb 11 10:10:13 2009 +0100
656
657    siliconmotion: Fix disabling of debugging if SMI501_CLI_DEBUG is set to 0
658    
659    Due to checking if SMI501_CLI_DEBUG is defined, some debugging will
660    be enabled if SMI501_CLI_DEBUG is set to 0. A single #if should be
661    used instead. Some debugging code already does this.
662    
663    Signed-off-by: Niels de Vos <niels.devos@wincor-nixdorf.com>
664    Signed-off-by: Francisco Jerez <currojerez@riseup.net>
665
666commit 0b34c4a29fcf45ac24361b4691cca0fd99cc3a3f
667Author: Francisco Jerez <currojerez@riseup.net>
668Date:   Wed Feb 4 03:49:08 2009 +0100
669
670    Wait for vertical retrace before writing registers at SMILynx_CrtcDPMS_*
671
672commit 1d46cca27b0d38d5355b5f93dd0ed0fe4f250d3f
673Author: Francisco Jerez <currojerez@riseup.net>
674Date:   Tue Feb 3 22:15:45 2009 +0100
675
676    Drop the outdated configuration options documentation in README.
677
678commit 64f722c3121a8f28cfc5b532ddec0ac83095ac95
679Author: Francisco Jerez <currojerez@riseup.net>
680Date:   Tue Feb 3 22:17:33 2009 +0100
681
682    Minor corrections at the man page.
683
684commit 94d9c659b71c84db46fa279d9d8736e1589e155c
685Author: Francisco Jerez <currojerez@riseup.net>
686Date:   Tue Feb 3 22:18:14 2009 +0100
687
688    Set dualhead to on by default on SM72x chipsets.
689
690commit 4d98e9c6a6e357ed736cd312a957c1e0a7b1d36b
691Author: Francisco Jerez <currojerez@riseup.net>
692Date:   Wed Feb 4 01:56:20 2009 +0100
693
694    Dynamically switch virtual refresh mode.
695    
696    Probably this makes dualhead mode more useful because it makes
697    possible displaying video on the LCD as long as the CRT output is
698    disabled or cloned.
699
700commit 1c57cfa814504c33fcad51115d77f2d166fb7903
701Author: Alan Coopersmith <alan.coopersmith@sun.com>
702Date:   Fri Jan 30 21:00:17 2009 -0800
703
704    Add README with pointers to mailing list, bugzilla & git repos
705
706commit 2aee107d992cb58bfbb2f527cfa73ea9f6957998
707Author: Alan Coopersmith <alan.coopersmith@sun.com>
708Date:   Fri Jan 9 16:33:43 2009 -0800
709
710    Remove xorgconfig & xorgcfg from See Also list in man page
711
712commit fdc2c50682b240e5966e4e5a7b45552ee2043bfc
713Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
714Date:   Mon Dec 29 18:41:05 2008 -0200
715
716    Bump release to 1.7.0.
717    
718      This driver supports smi 50x chipsets, randr 1.2, exa, dual head,
719    etc.
720    
721      Special thanks to Teddy Wang <teddy.wang AT siliconmotion.com.cn>
722    for support and help in responding and/or triaging hardware related
723    questions.
724
725commit 97eb5da1040ca7882e60b86ffaf25b6c021a3df5
726Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
727Date:   Mon Dec 29 18:39:54 2008 -0200
728
729    Use util-macros XORG_CHANGELOG and XORG_CWARNFLAGS.
730    
731      Also correct only compilation warning about possibly
732    uninitialized variable.
733
734commit deb2f845dd370ba819d02cb21b8a481827497af6
735Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
736Date:   Mon Dec 29 18:35:22 2008 -0200
737
738    Simplification and redundancy removal in video code.
739    
740      FOURCC_YV12 and FOURCC_I420 handling also was buggy. First it was
741    doing a noop by swapping offset2 and offset3 values twice, and second,
742    swap is not required when using smi 501/502 CSC video.
743      Changed SMI_DisplayVideo0501_CSC() to not set static values to
744    registers in a possible loop, if there is clipping.
745
746commit 46741589529809c17aa1e9719492a4b623de6ddf
747Author: Francisco Jerez <currojerez@gmail.com>
748Date:   Sat Dec 27 00:52:05 2008 +0100
749
750    Some more quirks for the SM712.
751    
752    * Program the MCLK to 157MHz on startup.
753    * Adjust the requested pixel clock if it's near one of the known
754      stable frequencies.
755    * Prefer the clock alternative with post scalar turned on when the
756      denominator is even.
757
758commit 439adf4455c651926040eeeec52a092f14f3196b
759Author: Francisco Jerez <currojerez@gmail.com>
760Date:   Mon Dec 22 17:28:35 2008 +0100
761
762    Add support for clone mode on Lynx chipsets.
763
764commit 2b3fa385a6da4be5ad6719dd115834b96d1ea3e3
765Author: Francisco Jerez <currojerez@gmail.com>
766Date:   Sun Dec 21 19:37:14 2008 +0100
767
768    Add a CRTC/Output implementation using BIOS for modesetting.
769    
770    After the RandR1.2 implementation the "UseBIOS" option wasn't actually
771    programming the hardware through VESA BIOS, this brings back that
772    functionality.
773
774commit 75b1c309f1418d70e8cf112744ff95f372d4ef75
775Author: Francisco Jerez <currojerez@gmail.com>
776Date:   Sun Dec 21 16:58:29 2008 +0100
777
778    Fix SMI_CrtcShadowAllocate.
779
780commit 068de9c0a302aba21e4d6a27eb6d330375ff2d2f
781Author: Francisco Jerez <currojerez@gmail.com>
782Date:   Sun Dec 21 16:02:57 2008 +0100
783
784    Allocate crtc->funcs and output->funcs in the heap.
785
786commit 373440667636d4e3214028710e33ac89d8f18ff1
787Author: Francisco Jerez <currojerez@gmail.com>
788Date:   Sun Dec 21 13:51:01 2008 +0100
789
790    Cleanup the Lynx register saving/restoring code.
791    
792    Save some registers not previously tracked, and use pSmi->mode instead
793    of continuously reading the hardware state.
794
795commit 67a1e4b090d6647b6c45f1534c67d9ea8a223d44
796Author: Francisco Jerez <currojerez@gmail.com>
797Date:   Sat Dec 20 15:41:13 2008 +0100
798
799    Add some quirks for SM712 modesetting.
800
801commit 7f252345c0c6b260c5b37aff98a22679eb5c6c82
802Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
803Date:   Fri Dec 19 21:15:01 2008 -0200
804
805    Enable the SMI 501/502 command list interpreter in a "debug" build.
806    
807      To enable it, set SMI501_CLI_DEBUG to 1 in smi.h, and use
808    Option "AcellMethod "EXA"
809    in the Device section of /etc/X11/xorg.conf
810    This code is enabled mainly for debug purposes. To make if have an
811    actual performance gain (like when using a sm50x with a "low profile"
812    "main" processor") it should be required to actually do busy loops
813    in kernel mode (and hope the costs of context switch will pay it).
814    In kernel mode it is possible to wait for an interrupt being triggered
815    when the command list is processed, or when the 2d engine is idle.
816    
817      This commit should be functional, but, mainly due to debug messages,
818    should be significantly slower then a build with MI501_CLI_DEBUG
819    defined to 0.
820
821commit 6a91bcc111902c45cc75c865893848b7c6c0a0b1
822Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
823Date:   Thu Dec 18 20:47:24 2008 -0200
824
825    Add definitions for the SMI 501/502 "command list interpreter".
826    
827      This also changes some bit operations to use a "bitfield" equivalent
828    one, with named fields, that should make it easier to understand what
829    is being tested.
830    
831      The enum smi_cli_cmd_code in smi_501.h is code that was added to a
832    experimental smi_drm.h, but the hardware only supports basic 2d accel,
833    and to compensate for the extra overhead for maintaining a command
834    list (assuming it worked correctly) it would be required to have a
835    special handling, like calling an ioctl to do the "busy loop" in the
836    kernel (that is, should wait for an irq or a timeout).
837    
838      The problem is that even if waiting for a idle engine before crafting
839    a command, and waiting again after submitting the command, there would
840    be corruption on screen after some time. So, the "busy loop" in the
841    kernel would only be useful if still using direct writes to mmio
842    registers.
843
844commit 5a07709ae41600bc02205753c64d764f11838240
845Author: Francisco Jerez <currojerez@gmail.com>
846Date:   Tue Dec 16 02:39:49 2008 +0100
847
848    Fix crashes when switching VTs with EXA enabled.
849    
850    Call ModifyPixmapHeader on SMI_EnterVT to update the screen pixmap
851    address tracked by EXA.
852    
853    Set memoryBase in the ExaDriverRec to NULL when switching out: some
854    pixmap could get allocated near the former aperture address and it
855    could be erroneously considered as being in offscreen memory.
856
857commit 47bcfb505886d16ff864f26a9cfa5f43b5da710e
858Author: Francisco Jerez <currojerez@gmail.com>
859Date:   Sat Dec 13 19:13:50 2008 +0100
860
861    Enable linear memory mode on SMI_MapMmio.
862    
863    Set UseBIOS to off by default for SM720. Some improvements on the
864    debugging output.
865
866commit 6f972c9022fc3a315b60246a69399807443d2ca0
867Author: Francisco Jerez <currojerez@gmail.com>
868Date:   Thu Dec 11 16:21:17 2008 +0100
869
870    Fall back to UseBIOS off when VBEInit fails.
871
872commit 1fc340dcf0abd7fa0837fef2c711aaec4fa8bcb8
873Author: Francisco Jerez <currojerez@gmail.com>
874Date:   Thu Dec 11 16:18:50 2008 +0100
875
876    Fix XAA SolidFill with 32 bpp framebuffer.
877
878commit dfc4510e8f4adeed520be47621ad429bc2d18f7e
879Author: Francisco Jerez <currojerez@gmail.com>
880Date:   Thu Dec 4 15:35:22 2008 +0100
881
882    Some corrections on the Lynx modesetting code.
883    
884    This fixes 1280x1024 modes on the VGA output with dual head on.
885    Also, solve a build problem when using -DSMI_DEBUG, as pointed by
886    Richard Schwarting.
887
888commit 6e1d29d19b205eac85b2f83de126ba88501e9b74
889Author: Francisco Jerez <currojerez@gmail.com>
890Date:   Tue Dec 2 23:43:26 2008 +0100
891
892    Disable screen centering on mode initialization.
893
894commit 695a6066cfe7ad0e568edc596b8704bb9bf8754f
895Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
896Date:   Sat Dec 13 04:20:20 2008 -0200
897
898    Janitor/maintenance patches.
899    
900    o Define HAVE_XMODES in config.h.
901    o Correct the LEAVE() macro declaration when SMI_DEBUG is defined.
902    o Remove prototype for SMI_DGAInit() if HAVE_XMODES is not defined.
903    o Properly scale value read from sm501 CURRENT_CLOCK register, to
904      print an appropriate message on startup.
905
906commit 191cbd3f8419cc5b428f262607d7a8594e574f67
907Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
908Date:   Fri Dec 12 15:50:53 2008 -0200
909
910    Unless there is a xorg.conf option, don't modify mclck neither mxclck.
911    
912      The option set_mclck was renamed to just MCLK, and a new MXCLCK option
913    was added for smi501/502 chips.
914      Note that previously, on MSOC chips, the set_mclck option actually
915    changed mxclck, what is now done by the MXCLK option.
916      The manpage was also update to reflect the options usage.
917      A new SMI501_FindMemClock() function was added, as a possible
918    placeholder for choosing alternate clocks when programming MMIO 0x68.
919    But it should be noted that the smi 501 databook says:
920    Miscellaneous Timing
921    	Read/Write MMIO_base + 0x000068
922    	Power-on Default 0x00000000
923    	[...]
924    	5:4 Divider Output Frequency Selection for Second PLL.
925    		00: 336 MHz.
926    		01: 288 MHz.
927    		10: 240 MHz.
928    		11: 192 MHz.
929    while the 502 databook says:
930    	Miscellaneous Timing
931    	Read/Write MMIO_base + 0x000068
932    	Power-on Default 0b0000.00xx.0000.1001.0000.1001.0000.0000
933    	[...]
934    	5:4 Reserved These bits are reserved.
935    but still tells to refer to those bits.
936
937commit a2352ff23787f37530dff68e03028dd78925e4b4
938Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
939Date:   Fri Nov 28 20:23:11 2008 -0200
940
941    Properly restore console on the smi 501/502.
942    
943      The current kernel framebuffer module expects accelerations registers
944    in the default boot state, otherwise, it will display some corruption,
945    or not correctly clear the screen, among other side effects.
946      Now it should be restoring everything the kernel framebuffer relies
947    on not being modified.
948
949commit 5a65559e12452a47f9848c47cfa6e30cfb871f0b
950Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
951Date:   Fri Nov 28 19:49:20 2008 -0200
952
953    DGA is dependent on HAVE_XMODES.
954
955commit 06c78cee718a86b1521c1fc1075d08bb9ad257f1
956Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
957Date:   Tue Nov 25 19:57:05 2008 -0200
958
959    Remove smi_dga.c and polylines code from smi_xaa.c.
960    
961      DGA was also not being properly test to not enable the pSmi fields,
962    but since the dga implementation is very basic, it is better to just
963    remove it (if compiling the sources on pre 1.4 X Server, DGA will be
964    disabled).
965    
966      The polylines code in xmi_xaa.c was already if 0'ed for some time,
967    but it may be reinstantiated to accelerate (back) xaa rotated modes.
968    Currently only randr+exa accelerates rotated modes.
969
970commit 9570cfdc132b0acb1d2b70edc6a3a5a812ef9691
971Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
972Date:   Tue Nov 25 15:07:39 2008 -0200
973
974    Ensure the engine is idle before switching to a virtual terminal.
975    
976      At least in the smi501/502, it may even completely lock in some special
977    cases (like when there is a browser playing some youtube video).
978      The code in SMI_CloseScreen() was almost identical to SMI_LeaveVT(),
979    so it now just call SMI_Leave() if required.
980
981commit 26ce0b163d9cf44ec1bc1d1bd460887aa5b405a7
982Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
983Date:   Mon Nov 24 15:15:43 2008 -0200
984
985    Use a single debug macro for leaving functions.
986    
987      The LEAVE() macro, when debug is disabled, now is defined as:
988    <hash>define LEAVE(...)		return __VA_ARGS__
989    This avoids the requirement of a return statement after the LEAVE()
990    macro call, what can be confusing.
991
992commit 2057d7ae05ef269f75a4d53fe5c4914a10856cb1
993Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
994Date:   Wed Nov 19 16:22:29 2008 -0200
995
996    Declare DGA fields only when not using xf86DiDGAInit.
997
998commit c2e318eaec3ac281457acdca4ea10fa368f646d3
999Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1000Date:   Mon Nov 17 20:08:49 2008 -0200
1001
1002    Enable XAA and randr rotated modes.
1003    
1004      Use the same logic for offscreen memory management as used in
1005    smi_video.c. This also makes the functions SMI_AllocateMemory and
1006    SMI_FreeMemory no longer static in smi_video.c.
1007
1008commit ed6aba73259a1412e23a954d6a8dba439ad98a8a
1009Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1010Date:   Fri Nov 14 22:37:56 2008 -0200
1011
1012    Correct EXA+Randr+rotated mode.
1013    
1014      The smi 501/502 cannot rotate-blt more than 32 bytes at a time.
1015    Patch based on smi's sample smi_shadow.c.
1016      XAA + randr + rotated mode currently not supported.
1017
1018commit 85d118ee288fe7aa32db003bfa322ddf06944bf6
1019Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1020Date:   Fri Nov 14 22:04:17 2008 -0200
1021
1022    Don't explicitly call the cursor routines.
1023    
1024      On the MSOC, the hardware cursor was remaining on screen, when using
1025    EXA and a randr rotated mode.
1026
1027commit cd10af22240b9df7c264b15c483d8cc4505ee359
1028Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1029Date:   Fri Nov 14 20:36:09 2008 -0200
1030
1031    Only compile smi_dga.c when xf86DiDGAInit is not available.
1032    
1033      Besides reducing driver size, this should make DGA+EXA accelerated.
1034
1035commit 41ef793b0ed6014ed49bef011259f977833aef5d
1036Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1037Date:   Fri Nov 14 19:24:45 2008 -0200
1038
1039    Correct regressions in CSCVideo
1040    
1041      Commit 0870d46718fe4e01953efd63cec46c54140b20f9 broke the clipping
1042    code. That has been corrected in this patch. Anyway, the hardware color
1043    space conversion should not be hidden down in the driver code, instead
1044    it should be usable by external programs as well, instead of having
1045    some applications doing it by software.
1046
1047commit be226bfaad76e7b48e006fe55ebae09b044ba4b7
1048Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1049Date:   Fri Nov 14 16:56:46 2008 -0200
1050
1051    MSOC: Hide cursor on mode setup when using software cursor.
1052    
1053      The kernel framebuffer module uses the hw cursor for the console cursor,
1054    so the driver must ensure it is hidden when setting a video mode.
1055
1056commit 5b8583fc2e5626ed98877858158d9687d7c94469
1057Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1058Date:   Fri Nov 14 15:47:52 2008 -0200
1059
1060    Be more liberal about mode width validation.
1061    
1062      The code was inherited from first import of latest smi sources, but
1063    the driver actually programs a mode, so there is no need to check if
1064    there will be a table entry with register contents matching the requested
1065    mode.
1066
1067commit ffb4be763ced82158a3a1d2e06fb0e81815e057b
1068Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1069Date:   Thu Nov 13 20:01:18 2008 -0200
1070
1071    Don't restore alpha registers, as they are not being modified.
1072    
1073      Restoring the registers "without need" caused some instability, with
1074    random locks.
1075
1076commit 6b6da04d566ec5b9d723c9f28791c604f76526dd
1077Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1078Date:   Thu Nov 13 19:44:14 2008 -0200
1079
1080    Changed to use panel plane tl and tr to center modes smaller then panel size.
1081    
1082      The code is under "<hash>ifdef USE_PANEL_CENTER", as it is buggy, and
1083    regardless of value set to right and bottom, it will crop from 0 to
1084    mode-width/mode-height, and then display a lot of screen artifacts, due
1085    to improper programming. Either way, the CRT in clone mode will display
1086    correctly.
1087
1088commit d2709b1af22a06b24606ec8c01f39f1ca27ca8d3
1089Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1090Date:   Thu Nov 13 18:00:46 2008 -0200
1091
1092    Enable gamma correction on the MSOC.
1093
1094commit 0870d46718fe4e01953efd63cec46c54140b20f9
1095Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1096Date:   Tue Nov 11 19:14:36 2008 -0200
1097
1098    Correct a problem in CSC video on the smi 501/502.
1099    
1100      This significantly simplify the function SMI_DisplayVideo0501_CSC(),
1101    but the real cause of the problem that this patch corrects is that
1102    whenever MMIO 0x1000fc bit 24 (Vertical Linear Filter Control)  is
1103    set, the last line of the rendered video will display only noisy,
1104    regardless of video being rendered 1x1, or scaled up or down.
1105
1106commit 09644ad3da4cf4c2576532137fdc81156e9c06e6
1107Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1108Date:   Tue Nov 11 15:34:27 2008 -0200
1109
1110    Disable detection of crt for the smi 501/502.
1111    
1112      As described by siliconmotion, it does not work on the smi 502.
1113    Does it work on older revisions? Keeping <hash>ifdef'ed out for now
1114    as it hopefully will be corrected in a newer hardware revision.
1115
1116commit caac648932df69419b8f86c5793a6b33f5997099
1117Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1118Date:   Tue Nov 11 15:06:15 2008 -0200
1119
1120    Implement an argb cursor using the alpha layer.
1121    
1122      The code is <hash>ifdef'ed out by <hash>if SMI_CURSOR_ALPHA_PLANE
1123    because the smi 502 hardware is buggy, but it appears to have worked
1124    for some older hardware revisions (by looking at code available at
1125    ftp.siliconmotion.com.tw).
1126      Keeping it at least for now, as it serves as a simple way to reproduce
1127    the problems described in smi.h.
1128
1129commit 61dd360a30c8e530185bc71b134c287ab1cdde2c
1130Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1131Date:   Tue Nov 11 14:47:58 2008 -0200
1132
1133    Add support for the reading/writing the alpha layer registers.
1134
1135commit d018ea2272cc1938cc0236eeb210692c1d1fd1df
1136Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1137Date:   Tue Nov 11 14:35:36 2008 -0200
1138
1139    Correct wrong offset value for secondary hardware cursor.
1140    
1141      Hardware cursor is still disabled in dual head mode due to some problems,
1142    usually when having both panel and crt mapped to the same address, what
1143    appears to confuse the "modes" code.
1144
1145commit f97c57873cd7622c373a394548ed435c1975abbc
1146Author: Francisco Jerez <currojerez@gmail.com>
1147Date:   Sat Nov 8 21:40:49 2008 +0100
1148
1149    Remove unused fifo_* options.
1150
1151commit 418b28028e69d14e74bf673be56123e99103e5bf
1152Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1153Date:   Sat Nov 8 16:04:58 2008 -0200
1154
1155    Remove the ZoomOnLCD option and update manpage for PCIRetry.
1156    
1157      ZoomOnLCD controlled if Ctrl/Alt/+ or Ctrl/Alt/- could be used to
1158    reprogram the mode on the panel. The option was not used, and now, if
1159    different resolutions could be used, it should be advertised by randr.
1160    
1161      PCIBurst and PCIRetry options are now enabled by default.
1162    
1163      The fifo_xxx options need a review, as they appear to be associated
1164    with the older WaitQueue macro, that did ignore its argument. But these
1165    options are not used, and may be removed or reimplemented at some time.
1166
1167commit bbc38b2a14d748bb9d9e153fcdcb259e5b10bdc8
1168Author: Francisco Jerez <currojerez@gmail.com>
1169Date:   Fri Nov 7 16:57:50 2008 +0100
1170
1171    Update the man page
1172    
1173    Reflect changes in the configuration options, additional supported
1174    chipsets and RandR1.2 interface.
1175
1176commit 7f8b7535a9a905c1d6c1d89cd1dd4dd506025429
1177Author: Nathael Pajani <nathael.pajani@ed3l.fr>
1178Date:   Thu Mar 20 07:25:00 2008 -0800
1179
1180    Bit twelve on CPR00 bitfield is not bit eleven...
1181
1182commit 98969b1b96244085ff5b0c76d2a463c1e436bcfa
1183Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1184Date:   Thu Nov 6 16:35:15 2008 -0200
1185
1186    Use the CRT Monitor Detect MSOC register.
1187    
1188      Probably I misunderstood the usage of this register, as it will always
1189    show the same value, regardless of having a crt connected to the "vga"
1190    port or not. Only difference is that while the detect bit is set,
1191    the crt will be blank. This patch should be more of a placeholder for
1192    a possible correction, but it should not cause any side effects, unless
1193    the data field can be zero on a valid situation.
1194      In my test computer, this field, in base 2 is always 1000000010000100,
1195    what doesn't really look like data in rgb 8:8:8...
1196
1197commit 87a3cc0652666c1796fd56f00810834193f3eca0
1198Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1199Date:   Tue Nov 4 16:38:12 2008 -0200
1200
1201    Allow setting modes larger then panel for secondary output.
1202    
1203      Sample usage on sm502:
1204    $ DISPLAY=:0 xrandr --output VGA --newmode 1280x1024 109.00 1280 1368 1496 1712  1024 1027 1034 1063 -hsync +vsync
1205    $ DISPLAY=:0 xrandr --addmode VGA 1280x1024
1206    $ DISPLAY=:0 xrandr --output VGA --mode 1280x1024 --pos 768x176
1207      Last command assuming using XAA (with ``Virtual 2048 1024''), so that it
1208    will have panel in top-left, and crt in bottom-right.
1209
1210commit 1d7b9547e1d37601f14410857c22f256e5c74f14
1211Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1212Date:   Tue Nov 4 15:57:12 2008 -0200
1213
1214    Add CSCVideo option to smi 501/502.
1215    
1216      This code is an adaptation of SMI sample implementation.
1217      CSCVideo is an alternate way to render video, that should reduce memory
1218    bandwidth usage, and uses the color space conversion hardware to render
1219    video directly to the framebuffer memory.
1220      When using randr dual head mode, currently only XAA is supported. As it
1221    has a fixed format framebuffer (when using the Virtual xorg.conf option).
1222      This patch also ensures that pScrn->displayWidth, pScrn->virtualX and
1223    pScrn->virtualY are only changed by the driver when using EXA.
1224
1225commit 93308d067c93126dd0aba9b718d9e0ba28dde9f4
1226Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1227Date:   Thu Oct 30 19:07:40 2008 -0200
1228
1229    Make MSOC work with EXA offscreen manager.
1230
1231commit e5cd9e4640555b9a3d4fab205e4b77faaca89d89
1232Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1233Date:   Thu Oct 30 19:05:40 2008 -0200
1234
1235      Update memory detection for MSOC.
1236    
1237      Use "local memory" field of mmio 0x010 and not "system memory", as
1238    it uses system memory values whenever there is a choice (just sets
1239    bits to zero).
1240
1241commit 5acf7e22c256637ba7139733cdba1c59557dd7ff
1242Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1243Date:   Thu Oct 30 19:01:54 2008 -0200
1244
1245      Disable HW cursor in dualhead mode in smi 501/502
1246    
1247      Other chipsets are doing the same. But this apparently is to circumvent
1248    a problem in randr/cursor implementation, that appears to not work very
1249    well with hardware that supports two hw cursors, but not argb cursors.
1250
1251commit c756ed67cbcf5bdd40c300627802d8bcd3d95123
1252Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1253Date:   Wed Oct 29 19:17:54 2008 -0200
1254
1255      SMI501/502 cursor fixes.
1256    
1257      Use separate buffers for crt and panel hw cursors.
1258      Describe, and workaround a problem with Dualhead and hw cursor enabled.
1259    It still has the inconvenient of having a blinking X when exiting the server
1260    or switching to a VT; maybe should also save the image currently configured
1261    also (and restore on exit), or just disable hw cursor on dual head mode.
1262      Also revert change to not initialize video on dualhead. Better to have
1263    fully functional video only on panel, that can be resized, then non
1264    resizable video..
1265
1266commit 2671776999a5add36e1aeaad4080bfccedc2b1a5
1267Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1268Date:   Tue Oct 28 19:45:31 2008 -0200
1269
1270    Update msoc to use randr cursor routines
1271    
1272      pSmi->IsSecondary was not used, and there should not exist two
1273    instances of the driver managing the same card.
1274      Macro CHECK_SECONDARY() was dependant on a second instance of the driver
1275    for the same card, and also removed.
1276      Some rework should still be done in the cursor code, as it sometimes,
1277    when displaying different data, will leave the hw cursor visible, while
1278    using an argb cursor.
1279
1280commit 91b5613fb07fec0f319623ca7ec1efd67f04379b
1281Author: Francisco Jerez <currojerez@gmail.com>
1282Date:   Tue Oct 28 00:13:55 2008 +0100
1283
1284    Remove shadowfb based rotation support.
1285    
1286    Also done some clean up, like removing unused/redundant members of the
1287    screen private structure, and removing now unused configuration file
1288    options.
1289
1290commit 0faeddb208d8fade4737be4bcadb8814696bd7b4
1291Author: Francisco Jerez <currojerez@gmail.com>
1292Date:   Tue Oct 28 00:06:37 2008 +0100
1293
1294    Lynx hardware cursor code adapted to the CRTC interfaces.
1295
1296commit 5d151781515cf2e3af1322ce6ed5f8d871c05791
1297Author: Francisco Jerez <currojerez@gmail.com>
1298Date:   Mon Oct 27 23:37:30 2008 +0100
1299
1300    Changes in the video overlay clipping code.
1301    
1302    Clip the video to the CRTC viewport by using
1303    xf86_crtc_clip_video_helper, and allow XV in
1304    dual head mode.
1305
1306commit 486e92399f09b9d765d03184fe29a32b3bc4ecc5
1307Author: Francisco Jerez <currojerez@gmail.com>
1308Date:   Mon Oct 27 23:12:36 2008 +0100
1309
1310    Simple EXA Composite implementation.
1311    
1312    * Implementation of the PictOpSrc render operation with a
1313      source coordinate transformation through 2D engine rotate-BITBLTs.
1314    * Fix EXA Copy and Solid in 24 bpp packed color mode.
1315
1316commit c0447d33c82829248e642b3156fd9a3c0d0eb709
1317Author: Francisco Jerez <currojerez@gmail.com>
1318Date:   Mon Oct 27 22:39:52 2008 +0100
1319
1320    Some corrections in the CRTC code.
1321    
1322    * Make screen resizing behave better with XAA.
1323    * Move some local variable declarations to the top of
1324      the function definition to conform ANSI C.
1325    * Make the Lynx panel modesetting code to actually
1326      program the panel sync pulse width in dualhead mode.
1327    * Also, fix a crash when using libpciaccess
1328      (the memory wasn't being unmapped).
1329
1330commit f2c83671cccc42345bfc9b506365936bbb6b9ef0
1331Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1332Date:   Fri Oct 24 17:55:02 2008 -0200
1333
1334    Make Dualhead option functional.
1335    
1336      Slightly change clock selection code to start using 501 compatible values,
1337    before checking 502 values, if it is a 502.
1338      DPMS for the VGA/second output was being set with bits inverted, that is
1339    dpms-on was programmed as dpms-off and vice versa. This was one of the
1340    reasons of dual head not working.
1341      Corrected wrong vsync programming for the crt. Cut&paste/typo caused
1342    programming vsync with hsync values, and thus, never getting the crt
1343    to accept the mode being programmed.
1344      If adding:
1345      Option "Dualhead" "True"
1346    to xorg.conf, now you should be able to do things like:
1347    $ DISPLAY=:0.0 xrandr --output VGA --right-of LVDS
1348    $ DISPLAY=:0.0 xrandr --output VGA --below LVDS
1349    and so on.
1350      *Iff* there is some way to not have any limitations for video playback,
1351    this option should be made default or automatically configured.
1352
1353commit 1efe36ed5cf5b0931daa915ca3ce231b78168d87
1354Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1355Date:   Wed Oct 22 19:06:06 2008 -0200
1356
1357    Make UseFBDev option functional again.
1358    
1359      It is mean't to be a fallback option, that as long as kernel boots
1360    in framebuffer mode, the X Server should also work.
1361      Correct incorrect value being checked when printing state of Dualhead
1362    option.
1363      The M value when programming PLL3 is actually an 8 bits integer,
1364    so correct it and comments about it.
1365
1366commit dd083c26de400cee8c67977f46cf90d048b22ff9
1367Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1368Date:   Tue Oct 21 19:41:55 2008 -0200
1369
1370    Use existing "Dualhead" option in MSOC.
1371    
1372      If Dualhead is set, it will attempt to use setups that allow different
1373    outputs with different contents, instead of crt always cloning panel.
1374      Still not fully functional.
1375
1376commit 197b4231586b9ada1c590a01b8dd7a5fcda936e6
1377Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1378Date:   Tue Oct 21 14:22:01 2008 -0200
1379
1380    Crt interface corrections.
1381    
1382      Correct wrong clock calculation for the 501 clock setting, that is used
1383    for pre 502 panel interface and crt.
1384      Adds a few new programmable fields to MSOCRegRec:crt_display_ctl.
1385      Call SMI501_CrtcAdjustFrame() after adjusting a mode, instead of
1386    before changing registers.
1387
1388commit 71990e0beece6e60ffbffccfdf56e5a71641a120
1389Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1390Date:   Mon Oct 20 15:09:09 2008 -0200
1391
1392    Don't change M1XCLK unless option specified in xorg.conf.
1393    
1394      Also add more complete description of programmable clocks.
1395
1396commit 8b7ce66e26113ec5859566e7f3d0fd15d3e29eaa
1397Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1398Date:   Fri Oct 17 18:58:45 2008 -0300
1399
1400    Update sm502 pll3 programming.
1401    
1402      Also print the value of sm502 registers (should be a noop on sm501).
1403      Added the frequency field to the MSOCRegRec's misc_ctl, as it is
1404    a read/write registers, and changing bit 24 may change the logic
1405    of pll3 programming (needs clarification with SMI).
1406
1407commit de1b633f916f4fcaaa95b226a8622d37041c86d4
1408Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1409Date:   Thu Oct 16 15:41:44 2008 -0300
1410
1411    Revert/modify some RandR changes to reenable XAA.
1412    
1413      XAA was not initialized anymore in RandR patch, but it is still the
1414    default accel method. Also reenable offscreen fb manager.
1415      At least on the MSOC OEM sample (using X Server 1.4), needs some more
1416    work as EXA is crashing in a call to exaPrepareAccess() when switching
1417    to a terminal; this seems to be related to disabling exa offscreen
1418    pixmaps on the smi501.
1419
1420commit e020afa359434b4da06ef8509a3e504423f824d2
1421Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1422Date:   Thu Oct 16 15:11:36 2008 -0300
1423
1424    Correct incorrect pll3 calculation.
1425    
1426      Double value must be truncated when calculating the difference from the
1427    possible clock and requested clock.
1428
1429commit f686d1b4fdd11d3d969ff4db21128d7867c9044d
1430Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1431Date:   Tue Oct 14 19:05:31 2008 -0300
1432
1433    Extra MSOC tweaks for the RandR1.2 changes.
1434    
1435      The "prototype" computer no longer locks, and only "missing feature"
1436    from before the patch is that video and Xaa don't work (only with Exa).
1437      For the moment, just force crt as not connected, and set the proper
1438    registers to make the secondary output clone the panel.
1439
1440commit fc722212d297a5141245beb7c8ae396606a8f221
1441Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1442Date:   Tue Oct 14 15:46:52 2008 -0300
1443
1444    Remove dependency on xf86cvt.c.
1445    
1446      It actually was not added in a previous commit (due to using git reset
1447    to remake a commit, and not readding it), but now, changed only call
1448    xf86CVTMode if it is available in the X Server, otherwise, for older
1449    X Servers, a modeline should be specified in xorg.conf.
1450
1451commit 5adb3b502ef79827b89fb11926b6198c3a57d2e6
1452Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1453Date:   Tue Oct 14 14:48:42 2008 -0300
1454
1455    Minor corrections for smi501 for the randr1.2 integration.
1456    
1457      Also correct printing information about MCLK in the smi501/502, it
1458    should print M2CLK (aka M1CLK o the 502) as it is the clock that
1459    controls video engine/memory.
1460
1461commit 6e0837305533f940adb7a09f9b013077c0888f2d
1462Author: Francisco Jerez <currojerez@gmail.com>
1463Date:   Sat Oct 11 14:24:48 2008 +0200
1464
1465    RandR1.2 initial implementation (WIP)
1466    
1467    Moved most of the Lynx-specific code out of the main
1468    functions at smi_driver.c to some new files: smilynx_hw.c
1469    (With CRTC-independent code like global hardware initialization and
1470    mode saving/restoring), smilynx_crtc.c (With the CRTC-local
1471    procedures) and smilynx_output.c (Output power management, DDC and
1472    monitor detection, currently).
1473    
1474      Done something similar with the SMI501 code: split
1475    SMI501_ModeInit in three separate functions: SMI501_HWInit that does
1476    the global initialization, and the CRTC mode_set callbacks
1477    SMI501_ModeSet_crt and SMI501_ModeSet_lcd at smi501_crtc.c.
1478    
1479      The SMI501_ModeSet code is divided into SMI501_WriteMode_common,
1480    SMI501_WriteMode_lcd and SMI501_WriteMode_crt, each one updates a
1481    different register set in the hardware.
1482    
1483      Inside smi_crtc.c, there is mainly hardware independent
1484    code... Initial CRT controller allocation, shadows, rotation and
1485    framebuffer resizing code.
1486    
1487      The shadow code currently relies on EXA... I'm not sure what are the
1488    problems of the EXA implementation in MSOC. Does it work?
1489    
1490      I think it would be a good thing to get rid of XAA soon: The next
1491    thing I'll be working on will be EXA Composite... I hope it's possible
1492    to get EXA acceleration working at least as fast as the current
1493    XAA. (Although Teddy Wang confirmed me that the SMI720 DMA engine is
1494    broken...)
1495    
1496      So, the patch adds some features like new Lynx dualhead modesetting
1497    code, but it probably breaks some other things:
1498    * Video Overlay: I suppose it does work with EXA activated. It seems
1499    it is also possible to have simultaneous overlays in both CRTCs with the
1500    Lynx hardware (I have some code for this, but still work in progress).
1501    * Hardware Cursor (It needs more integration in the CRTC interfaces).
1502    * The old Shadow FB / rotation code, which we should probably drop.
1503    
1504      These are in my TODO list... Along with some more cleaning and other
1505    minor issues in the modesetting code.
1506    
1507      About the SMI501 RandR1.2 implementation... I suppose it's specially
1508    lacking per-output DPMS (The DPMS field in the System Control register
1509    only affects the CRT, doesn't it?).
1510
1511commit 605199b1fa80d5a65e11a38270f6d0974466d1dc
1512Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1513Date:   Thu Oct 9 18:51:50 2008 -0300
1514
1515    Add a PanelSize/60Hz CVT mode at driver initialization
1516    
1517      This driver is expected to be used with XServer 1.4 or newer, but a
1518    (probably temporary) copy of xf86cvt.c was also added to the build.
1519      This modeline automatically added implies it would not be required
1520    to specify one in xorg.conf, as it should not be uncommon "non standard"
1521    modes, like 800x480 or 1024x600.
1522
1523commit 55052509d75492f0936273686826c51fcbc831c1
1524Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1525Date:   Thu Oct 9 18:12:02 2008 -0300
1526
1527    Make the input frequency in SMI501_FindPLLClock a variable
1528    
1529      Instead of using a fixed 24Mhz input frequency, as "unofficially"
1530    specified by SMI, a guess of using a 12 multiplier was used, and it
1531    corrects the screen flicker problem as well as make a secondary output
1532    work again.
1533
1534commit 61f814d1c8ad9440bab89eda44e64b01ba0eb703
1535Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1536Date:   Wed Oct 8 18:36:11 2008 -0300
1537
1538    Rewrite WaitQueue and WaitIdle accell macros
1539    
1540      WaitQueue() did receive an unused parameter, that was removed. The
1541    parameter was unused since version 1.0 in XFree86 CVS, so probably
1542    nobody knows for what it was used.
1543      WaitIdle() now also replaces WaitIdleEmpty(), as an idle engine
1544    should also have an empty fifo.
1545      SMI_SubsequentScreenToScreenCopy() was changed to use WaitIdle()
1546    instead of WaitQueue() before submitting the commands to the engine.
1547    This fixes a nasty lockup when using XaaOffscreenPixmaps (what is
1548    desired/wanted to help improve performance), and the lockup in the
1549    MSOC was clearly due to an engine overrun, when sending commands
1550    faster then the engine could process.
1551
1552commit 1656fb6de5308ff586676e8d6a8aa8d4579ebf4c
1553Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1554Date:   Fri Oct 3 18:55:14 2008 -0300
1555
1556    Don't use the 1 multiplier on older chipsets.
1557    
1558      It is not in the specs, so don't use it on chipsets older then the 502.
1559      Some review should be done in SMI501_FindPLLClock() as when hardcoding it
1560    to use the clock selection code for older chips, it appears to choose a
1561    better refresh rate (no flicking at 1024x600).
1562
1563commit da789f720e2d772dff92adc10aa024a6a921fbbd
1564Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1565Date:   Fri Oct 3 16:30:23 2008 -0300
1566
1567    Simplify regsmi.h by removing most unused SMI501 defines.
1568    
1569      The values that should be used, should also have a more meaningful name,
1570    as something like:
1571    doesn't help much, and the fact that these "ports" can be accessed from
1572    several "bases" may be confusing also, as there is SCRBase, DCRBase,
1573    FPRBase, DPRBase, CPRBase (and IOBase). The 501 code should probably be
1574    changed to just use IOBase, as currently it uses SCRBase, that happens
1575    to be at pSmi->MapBase + 0x0000.
1576    
1577      This patch also completely transforms the WaitQueue macro in a noop
1578    for the MSOC. In SMI sample code it is almost the current code, just
1579    that the macro body is commented out. Hopefully this will be addressed
1580    soon.
1581
1582commit bacddd1d31a5f3de14477914958f27b6f081c125
1583Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1584Date:   Fri Oct 3 15:10:55 2008 -0300
1585
1586    Rename macro "bitfield" to "bits" and correct a wrong division.
1587
1588commit 22a7fd6711e2a40633981cd21fa3808c9246f13a
1589Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1590Date:   Thu Oct 2 16:11:59 2008 -0300
1591
1592    Remove the IN_SEQ and OUT_SEQ macros.
1593    
1594      These macros access vga registers, but don't make it clear in their name.
1595
1596commit 973f68bf0464c4d2aec2acf7c8e6db201782f4ec
1597Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1598Date:   Thu Oct 2 15:48:51 2008 -0300
1599
1600    Rewrite some macros to not have side effects in if/else nesting.
1601
1602commit 82e17c442c3969938435a8b46d2679995af64425
1603Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1604Date:   Thu Oct 2 15:25:41 2008 -0300
1605
1606    Add support for the extra divider in the alternate pixel clock setting.
1607    
1608      Also remove extra includes that were not used.
1609
1610commit b32769305275519f7f4d1e0871a71514636ca026
1611Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1612Date:   Wed Oct 1 18:27:27 2008 -0300
1613
1614    Correct clock programming for the SMI 501/502
1615    
1616      The new code is based on the file SM502Clock.pdf provide by SMI engineers
1617    (many thanks). And now it is expected to work correctly. The few remaining
1618    details that are unclear (for the extended 502 modesetting mode) should be
1619    resolved soon, and those are:
1620    *  MMIO:0x74:bit15 (PLL Output Divided by 2)
1621    	0: Disable.
1622    	1: Enable.
1623       (does this mean it can use a 12MHz clock instead of 24? or anything else?)
1624    *  Should bit 31 of "current clock" always be set when using the alternate
1625       mode setting for the 502? The alternate modesetting allows a very closer
1626       pixel clock programming option, usually with 0 difference from the
1627       xf86 mode.
1628    
1629      I will ask SMI to make SM502Clock.pdf also available in the ftp site, it
1630    is just 3 pages, or possibly even better, an update for the existing
1631    documentation.
1632
1633commit ea34e20852ee6007e2f0fcacd5068c151865be1e
1634Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1635Date:   Tue Sep 30 17:49:25 2008 -0300
1636
1637    Remove the macro field, and rename the detail structure to f.
1638    
1639      Now access to bitfields looks like:
1640    record->register.f.field
1641      instead of:
1642    field(record->register, field)
1643      what should be easier to read and understand.
1644
1645commit a91872f813a6299f32b9f833876c54afd0b53aaa
1646Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1647Date:   Tue Sep 30 17:33:36 2008 -0300
1648
1649    Update to match the SMI 502 chipset specs.
1650    
1651      The 502 is a superset of the 501, also recognized as PCI_CHIP_SMI501.
1652      The driver is only known to work with smi_501 chipset
1653    "SM501 VoyagerGX Rev. AA (rev c0)".
1654
1655commit e830572bca9b8eaf50a36a12d1fe31cdfec6280e
1656Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1657Date:   Fri Sep 26 19:43:45 2008 -0300
1658
1659    Don't always program CRT clock and registers.
1660    
1661      This causes instabilities, and there was also a bug where it would
1662    tell the crt to not select panel data (due to removing the "else").
1663
1664commit 314eecb6628840a942ff2e76d8c7402301feac77
1665Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1666Date:   Fri Sep 26 19:30:17 2008 -0300
1667
1668    Properly check pScrn->driverPrivate before deferencing it.
1669    
1670       Add missing divide by 1000 in smi_501.c, as well as a FIXME comment
1671    about a crash, and possible cause as well as possible ways to program
1672    other clocks.
1673
1674commit f18222ffcff8e78b3cfcd220cad06a1daa3fc920
1675Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1676Date:   Fri Sep 26 17:55:42 2008 -0300
1677
1678      Don't try to find the closest clock, just use highest one.
1679    
1680      Previous patch wasn't fully correct, just that the values were generating
1681    the same register dump as the smi 2.2.5 driver version.
1682      When correcting it to not mix Mhz and Khz in the math, the problems
1683    returned.... Anyway, it just works when selecting the highest clock, and
1684    the attached CRT will display the panel image. This is illogical, so maybe
1685    the specs are outdated... The previous code has been #ifdef'ed out in
1686    "CALC_CLOCK".
1687      Note that it still changes the boot default clock, from the default
1688    288Mhz to 336Mhz, and in this patch, it also always sets both, panel and
1689    crt clocks (what happens here, is that most likely, by doing this, it
1690    forces the hardware to "somehow" autoconfigure itself; will try to get
1691    some information from SMI about this, but this is the boot default,
1692    and what smi sources do...).
1693
1694commit bdba7c1397186cd15d519db55c2345db44990be4
1695Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1696Date:   Wed Sep 24 18:58:35 2008 -0300
1697
1698    Fix incorrect understanding of the pixel clock from specs.
1699    
1700      The clock that should be programmed is the bandwidth, and not the pixel
1701    clock itself. Based on analysis of register contents, after "importing"
1702    a slightly modified version of xf86ModeBandwidth() and properly converting
1703    values, now it properly programs both, the LCD and a "clone" CRT or just
1704    a secondary panel.
1705    
1706      Modes generated by either gtf or cvt correctly works.
1707    
1708      A (hopefully temporary) fallback was also added, and if the option
1709    "UseFBDev" is used, the driver will not attempt to set/modify the
1710    current video mode.
1711    
1712      TODO: Generate a modeline "on the fly", based on monitor configured
1713    specs, or some set of defaults, so that it should not be required to
1714    specify a modeline in xorg.conf.
1715
1716commit e64b3d63be2a4148c4d10ac160e2bd2f3e8c7fd3
1717Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1718Date:   Mon Sep 22 22:01:16 2008 -0300
1719
1720    Correct problems in clock setting.
1721    
1722      One clock must be changed at a time, first setting the pll value,
1723    then waiting for 16ms (one vsync), then setting the divider/shift
1724    values, and again waiting 16ms.
1725      Code was working after split of SMI501_ModeInit(), because some
1726    clocks were already using system boot default, but would most likely
1727    cause a crash when actually changing values (currently only p2_xxx
1728    or v2_xxx is changed, but it would cause problems on some system where
1729    the initial values don't match the ones set by the driver).
1730    
1731      TODO: Either don't change m2clk and mclk or ensure other values
1732    work correctly (don't change because the kernel should already
1733    have set those if a value other then the boot default should be
1734    used).
1735
1736commit 1f0af5ea50e3e3b2c9ed882f7ef2fb8871e780bd
1737Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1738Date:   Mon Sep 22 21:41:12 2008 -0300
1739
1740    Correct logic in sw cursor handling and add missing entries to .gitignore.
1741
1742commit 862cbd8c1845b54a38996b92e8e22d153f38cdd2
1743Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1744Date:   Mon Sep 22 19:09:12 2008 -0300
1745
1746    Split SMI501_ModeInit in two functions.
1747    
1748      This allows using alternate MSOCRegRec structures to set a mode.
1749    This should be useful to restore the initial mode.
1750
1751commit bcbf1854be851b8fa1ec3f384b11c312b0533bf3
1752Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1753Date:   Mon Sep 22 18:29:31 2008 -0300
1754
1755    Remove unused .cvsignore files.
1756
1757commit 1014ffe0a2c55119d6bfbcf632f3d4ebe77af212
1758Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1759Date:   Mon Sep 22 18:15:21 2008 -0300
1760
1761    Simplify hw cursor and sw cursor option handling.
1762    
1763      Also rename SMIRec field hwcursor to HwCursor, as there are few usages,
1764    and this way, match the case usage for other field names.
1765
1766commit 6b5c22d8680573c8a6b259d78ba3c8435514fde9
1767Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1768Date:   Mon Sep 22 17:05:21 2008 -0300
1769
1770    Rework/simplify debug macros.
1771    
1772      Instead of cut&paste of the name of the current function everywhere,
1773    just use cpp's __FUNCTION__ predefined macro.
1774      Create two macros to exit a function, named LEAVE() and RETURN().
1775    Functions returning void should call LEAVE() and then explicitly return
1776    for now. "Logged" function calls are indented, so a review was done to
1777    ensure functions with a ENTER() also have the proper exit macro.
1778      The DEBUG macro was changed to have variadic arguments, and this way
1779    it is no longer required to prefix arguments with VERBLEV, but now it
1780    also is not possible to use another "verbosity value", but it wasn't used
1781    in any of the DEBUG macro calls.
1782
1783commit cf7097c19b31671a53bc8161bd29b9f79f1d3d85
1784Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1785Date:   Mon Sep 22 14:25:04 2008 -0300
1786
1787    Fix a leak and minor cosmetic change.
1788    
1789      Leak was due to only releasing return value of xf86GetEntityInfo
1790    if the chipset was PCI_CHIP_SMI501.
1791      Cosmetic change was to log pci burst and pci retry default values
1792    as X_DEFAULT instead of X_PROBED.
1793
1794commit ee1e72a248b2df6522283c0dd3e5433d7f961c62
1795Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1796Date:   Fri Sep 19 18:06:20 2008 -0300
1797
1798    Kludge to not lock the SMI 501 when running at 8bpp.
1799
1800commit 9ae840a358959615020fa4adb76c1487231927f3
1801Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1802Date:   Fri Sep 19 18:00:16 2008 -0300
1803
1804    Add MSOC palette support to run at 8 bpp.
1805    
1806      Colors should be correctly set, but most output using Render is not
1807    properly drawn, i.e. anti aliased fonts, etc. Need to investigate
1808    further, but it should be a general problem, not exclusive to the
1809    driver.
1810
1811commit f4339a6a7aa077cb355dd23b77ea568ad2a481e3
1812Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1813Date:   Fri Sep 19 16:52:17 2008 -0300
1814
1815    Enable pci retry and pci burst by default.
1816    
1817      Rename fields pci_burst and NoPCIRetry to PCIBurst and PCIRetry,
1818    to match field names case usage, and updated logic for the removal
1819    of "No" from the pci retry option.
1820      Pci retry and pci burst greatly increase video playback speed,
1821    allowing smooth video play on the SMI 501.
1822
1823commit 782ef31a9273e33100ae6b4bd20a4d1d3f493aa4
1824Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1825Date:   Fri Sep 19 15:48:34 2008 -0300
1826
1827    Add initial exa support for SMI501.
1828    
1829      This just allows exa to run, with a minimal ammount of features avaiable.
1830
1831commit 5d9ab8195e692f78dd7e57e7d17d8b46f7f59221
1832Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1833Date:   Thu Sep 18 18:27:51 2008 -0300
1834
1835    Add a missing CHECK_SECONDARY macro call.
1836
1837commit 2876b55c3ed82b9be023acbfc4ddf993494c3c9b
1838Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1839Date:   Thu Sep 18 18:23:37 2008 -0300
1840
1841    Correct video offscreen memory allocation routines.
1842    
1843      It was always failing in the first allocation, if using XAA, what caused
1844    some players to not work correct. This is a major problem for the MSOC as
1845    it still only supports XAA.
1846      Also add a workaround for a hardware problem, as described in
1847    "SM501 Rev.AA engineering addendum v0.1.pdf"
1848    "A-27 PIXELS TOO BRIGHT ON VIDEO SHRINK".
1849
1850commit 79297e797de3497773539eb9f89e8594797ee512
1851Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1852Date:   Thu Sep 18 18:16:21 2008 -0300
1853
1854    Update for new smi_501 interfaces.
1855    
1856      This also adds some simplification for the driver code, by adding
1857    some function pointers to avoid the need to check the chipset everywhere.
1858      The new SMI_DetectMCLK function doesn't actually change pSmi->MCLK to
1859    avoid chances of it stoping working on non MSOC chipsets, as it
1860    previously not did set it, and for the MSOC, for the moment, default to
1861    whatever value is already set, i.e. doesn't reprogram it.
1862      Added new memory detection code for the MSOC, that better matches the
1863    documentation.
1864
1865commit 95e312b712a1e4a476ef31c5302faf77c22915a9
1866Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1867Date:   Thu Sep 18 18:10:54 2008 -0300
1868
1869      Complete rewrite of smi_501.c and smi_501.h.
1870    
1871      The previous version was dependant on kernel framebuffer, as it was
1872    just failing a test if not having a modeline defined, and in the test
1873    case, it is running at 1024x600.
1874      Now it properly programs the video hardware, and the procedure is
1875    expected to be very well documented.
1876
1877commit cf5132907e4709c872089fe2d79837b2de4e45b7
1878Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1879Date:   Tue Sep 2 18:16:41 2008 -0300
1880
1881    Rename global smi501 functions to have SMI501 prefix.
1882    
1883      Functions renamed were setDPMS -> SMI501_SetDPMS,
1884    regRead32 -> SMI501_Read32 and regWrite32 -> SMI501_Write32.
1885      Also moved SMI_SetDPMS to start of smi_501.c to have extern
1886    functions at start of the file.
1887
1888commit a2a090e10f031aed6dbf6d143d6ce8abdf7a2392
1889Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1890Date:   Tue Sep 2 17:55:16 2008 -0300
1891
1892    Correct all compiler warning messages.
1893    
1894      Code compiled with:
1895    gcc -Wall -Wbad-function-cast -Wdeclaration-after-statement \
1896        -Wmissing-prototypes -Wmissing-declarations -Wnested-externs \
1897        -fno-strict-aliasing -Wold-style-definition -Wpointer-arith \
1898        -Wstrict-prototypes
1899    
1900      This patch also makes most smi_501.c functions static, as they are not
1901    called from anywhere else.
1902      Also removed prototypes for non existent functions in smi_501.h.
1903
1904commit 2d5ba5f7310fa8e11f7349d116aa871a239ac35f
1905Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1906Date:   Tue Sep 2 16:32:31 2008 -0300
1907
1908    Correct xv video problems on MSOC.
1909    
1910      Set VIDEO_CLIP_TO_VIEWPORT as it is used in MSI sources, and also add
1911    proper code to detect panel size, as this information is used to manage
1912    offscreen memory (but shouldn't it be using pScrn->virtualY, etc?)
1913
1914commit e111c85d6cca3f2206b6bce119b284edccf0cd4f
1915Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1916Date:   Mon Sep 1 18:41:43 2008 -0300
1917
1918    Don't pretend this driver compiles on XFree86.
1919    
1920      Either remove check for XF86_VERSION_CURRENT or at least add the
1921    implementation of RegionsEqual() to smi_video.c.
1922
1923commit b23c5c5e434aaba2128abcae61a8564a62d2e7a1
1924Author: Francisco Jerez <currojerez@gmail.com>
1925Date:   Mon Sep 1 21:28:13 2008 +0200
1926
1927    Fix XAA, ShadowFB and VT switching for non-sm501 chipsets
1928    
1929    Signed-off-by: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1930
1931commit 6a01c87b7c75f698ff2b783f8d341526159ebfe2
1932Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1933Date:   Mon Sep 1 18:00:03 2008 -0300
1934
1935    Split SMI_MapMem in SMI_MapMem and SMI_MapMmio
1936    
1937      This is required because MSOC needs mmio to detect ammount of memory.
1938      The patch also creates a SMI_DetectMem() function, that should slightly
1939    simplify SMI_PreInit().
1940
1941commit c45ef4676812f72dcff53ba85e9fa95836704c75
1942Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1943Date:   Fri Aug 29 19:33:31 2008 -0300
1944
1945    Bump version to 1.6.1.
1946
1947commit 4eee032811d1487df9dcf94708b1bcafc7e63392
1948Author: Arnaud Patard <apatard@mandriva.com>
1949Date:   Fri Aug 29 19:33:31 2008 -0300
1950
1951    Correct a problem when handling i420 format.
1952    
1953    The problem was that the driver was converting i420->YV12 2 times
1954    instead of only once.
1955
1956commit 76c931e149917563485ae6ad2950aa28a531b0d1
1957Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1958Date:   Fri Aug 29 19:31:57 2008 -0300
1959
1960    Update MSOC video interface.
1961    
1962      This synchronizes smi_video.c with smi 2.2.5 sources.
1963
1964commit 0eb35f2065314807a604a44e571bbb898db23543
1965Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1966Date:   Fri Aug 29 19:20:54 2008 -0300
1967
1968    Update xaa and generic acceleration code for the MSOC.
1969    
1970      Currently, the support for EXA is not functional, neither is
1971    ShadowFB or bpp/depth other than 16.
1972
1973commit 375e35fc7df4a21c2e0bc933774cc2f5f29b89d1
1974Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1975Date:   Fri Aug 29 19:36:35 2008 -0300
1976
1977    MSOC doesn't access VGA registers or VBE/INT10
1978    
1979      This should be the biggest patch in the "merge". Done in a single chunk
1980    to have it compilable at all stages.
1981
1982commit 164d921465ae7b3a0fa2551358f7c0b5a18a425a
1983Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1984Date:   Fri Aug 29 18:47:14 2008 -0300
1985
1986    Add code to probe and recognize the SMI501 chipset
1987
1988commit db330497c3b757361ac9c3602b3a4ece0a401dd3
1989Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1990Date:   Fri Aug 29 15:58:16 2008 -0300
1991
1992    Change SILICONMOTION_NAME value
1993    
1994      xf86VDrvMsgVerb prints up to 14 characters prefix, where prefix has the
1995    format "%s(%d): " so, use name "SMI" instead of "Silicon Motion"
1996
1997commit 36309132aef067822a8901ce24fda817be58546d
1998Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
1999Date:   Fri Aug 29 15:52:36 2008 -0300
2000
2001    Add initial support and macros for the MSOC.
2002    
2003    This patch add the new files and basic required definitions.
2004
2005commit 96166eafc2b0e5db627aa4f8b0422f953837772b
2006Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
2007Date:   Fri Aug 29 14:48:44 2008 -0300
2008
2009    Fix build for removal of xf86Version.h
2010    
2011      This is the first patch of series to merge code from the version 2.2.5
2012    from SiliconMotion and code from
2013    http://cgit.freedesktop.org/~agd5f/xf86-video-smi501
2014    
2015      After the merge, work will continue, to correct some extra problems,
2016    and make it work with EXA at least (as only the Xorg version has
2017    EXA support).
2018
2019commit 682b9a1b97a82c7d084c63860a8c753cc45d9faf
2020Author: Francisco Jerez <currojerez@gmail.com>
2021Date:   Sat Aug 16 13:24:27 2008 +0200
2022
2023    Allow using XV and RandR rotation simultaneously.
2024    
2025     As it doesn't seem feasible to rotate the video, I did some modifications
2026    at SMI_PutImage and SMI_ClipVideo to make PutImage work (incorrectly) with
2027    rotation enabled. The image is displayed unrotated.
2028    
2029    Signed-off-by: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
2030
2031commit f6110b1ef5934b45812ee95e70b1e64f44386dda
2032Author: Francisco Jerez <currojerez@gmail.com>
2033Date:   Sat Aug 16 13:08:25 2008 +0200
2034
2035    Some fixes in the EXA UTS/DTS code.
2036    
2037     In DTS:
2038      * It uses the screen Bpp inestead of the pixmap Bpp...
2039        this gives some problems when using pixmaps with different depth
2040        to the screen.
2041     In UTS:
2042      * aligned_pitch was computed from src_pitch inestead of the
2043        pixmap width.
2044      * When writing the target coordinates to the DE registers, it does
2045        y*0xFFFF inestead of y & 0xFFFF.
2046      * I renamed source_pitch to src_pixelpitch as it is very confusing
2047        to have src_pitch and source_pitch.
2048      * It isn't necessary to call WaitQueue before copying each scanline,
2049        but it seems it almost doesn't affect performance.
2050    
2051    Signed-off-by: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
2052
2053commit 209097ba5b44a0ce0da7f1ea52150dcace2b5244
2054Author: Francisco Jerez <currojerez@gmail.com>
2055Date:   Sat Aug 16 13:00:31 2008 +0200
2056
2057    RandR rotation implemented.
2058    
2059    * I added the configuration file option "RandRRotation".
2060    * I replaced pSmi->ShadowPitch with pSmi->screenStride, it seems it
2061       makes more sense because the lower word of ShadowPitch may change
2062       independently.
2063    * I moved the SMI_DEDataFormat to smi_accel.c because it seems it is
2064      a piece of code repeated many times in the driver.
2065    * At some places, it is assumed the framebuffer is at FBOffset:
2066      when using a shadow framebuffer, FBOffset is the location of the
2067      on-screen framebuffer (0 should be used). This made e.g. EXA
2068      completly useless with ShadowFB enabled (it crashed).
2069    * In the FBManager initialization, I have replaced xf86InitFBManager
2070      with xf86InitFBManagerRegion to reserve some additional space as
2071      screen fb: it's unlikely to be the case, but a less efficient
2072      alignment in the rotated mode could make the rotated mode need more
2073      memory than the unrotated one.
2074      This is not a problem with EXA as the offscreen memory parameters
2075      can be easily modified when doing the rotation.
2076    * In SMI_RefreshArea it's assumed that some DE registers are already
2077      in some state, this is specially not true when using EXA.
2078    * SMI_ValidMode rejects a rotated mode with different dimensions
2079      than the panel. This seems to work now.
2080    
2081    Signed-off-by: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
2082
2083commit 45c6aedd6fe4e4d6abe58d1bb39ec01049404f09
2084Author: Francisco Jerez <currojerez@gmail.com>
2085Date:   Sat Aug 16 12:50:26 2008 +0200
2086
2087    Updates in SMI_EnterVT when remapping memory.
2088    
2089      Update the screen pixmap header with the new aperture address
2090    when remapping framebuffer memory in SMI_EnterVT.
2091      Update pSmi->EXADriverPtr->memoryBase too, in case EXA is
2092    being used.
2093    
2094    Signed-off-by: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
2095
2096commit 8feca790a6e92799019237ac69a4ef618cacfaae
2097Author: Francisco Jerez <currojerez@gmail.com>
2098Date:   Sat Aug 16 12:40:52 2008 +0200
2099
2100    Make the int10/VBE initialization depend on the UseBIOS configuration option.
2101    
2102    Signed-off-by: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
2103
2104commit 772e899e71efab208b844b09c75f9f49ff453386
2105Author: Francisco Jerez <currojerez@gmail.com>
2106Date:   Sat Aug 16 12:25:18 2008 +0200
2107
2108    Memory detection moved before memory mapping in SMI_PreInit.
2109    
2110    Take into account the VideoRam configuration option.
2111    
2112    Signed-off-by: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br>
2113
2114commit d28ed6aa0d28fdcf8f555b2bb8a58fe28d1b464b
2115Author: Adam Jackson <ajax@redhat.com>
2116Date:   Fri Aug 15 14:24:31 2008 -0400
2117
2118    Dead code removal.
2119
2120commit 7cfecd313aaea74ad5675c53b4d0099d6e9dec39
2121Author: Adam Jackson <ajax@redhat.com>
2122Date:   Wed Mar 19 15:26:14 2008 -0400
2123
2124    siliconmotion 1.6.0
2125
2126commit c0e80b14a8af7c9df44919b2df58030e59ecbf88
2127Author: Adam Jackson <ajax@redhat.com>
2128Date:   Wed Mar 19 15:25:23 2008 -0400
2129
2130    Fix distcheck
2131
2132commit e53d07525952e1281170db01c48d7a14a6983049
2133Author: Adam Jackson <ajax@redhat.com>
2134Date:   Wed Mar 19 15:24:11 2008 -0400
2135
2136    Death to RCS tags.
2137
2138commit 46e68bf899754f6f70a0f5a7b0d42c5e208aaa5a
2139Author: Dave Airlie <airlied@redhat.com>
2140Date:   Mon Mar 10 13:53:00 2008 +1000
2141
2142    smi: add pciaccess support
2143
2144commit 75182ca115cba8fcf31304aea88ca3d4b9aa66ac
2145Author: Matthieu Herrb <matthieu.herrb@laas.fr>
2146Date:   Sat Mar 8 23:49:58 2008 +0100
2147
2148    Makefile.am: nuke RCS Id
2149
2150commit ae556406ab6c4de410b68b3c8bcebb434b3efc9d
2151Author: Brice Goglin <bgoglin@debian.org>
2152Date:   Tue Aug 7 10:00:25 2007 +0200
2153
2154    Define SILICONMOTION_VERSION* using PACKAGE_VERSION*
2155
2156commit 1a803a8f91a931c00106f9d3d41cfa5d74c19f55
2157Author: Alex Deucher <alex@botch2.com>
2158Date:   Tue Mar 20 21:08:38 2007 -0400
2159
2160    minor fixup
2161
2162commit 503027d38dab759277c0a67746ac205a1dfa9277
2163Author: Dennis De Winter <dennis.de.winter@gmail.com>
2164Date:   Fri Mar 16 21:29:19 2007 -0400
2165
2166    Fix UTS issues with flash
2167    
2168    Also remove safety padding on memory manager calculation
2169
2170commit fe4114033f2853e2df8d37c490deba26bf0366e8
2171Author: Alex Deucher <alex@botch2.com>
2172Date:   Thu Mar 15 02:11:24 2007 -0400
2173
2174    hostdata blit UTS works, but SW is faster
2175    
2176    UTS is working, but SW is faster so leave it disabled
2177
2178commit 62442ce9e62b484e4f2067e3c52c153cd356be2c
2179Author: Alex Deucher <alex@botch2.com>
2180Date:   Wed Mar 14 20:13:28 2007 -0400
2181
2182    more clock fixes
2183
2184commit 6ff1241636e25d2c0f20bd14e92c8585f209aeb0
2185Author: Alex Deucher <alex@botch2.com>
2186Date:   Tue Mar 13 02:32:52 2007 -0400
2187
2188    bump to 1.5.1
2189
2190commit b53db3bc5ba7648d392dc630ef47e57fb8a0afb4
2191Author: Alex Deucher <alex@botch2.com>
2192Date:   Tue Mar 13 02:31:30 2007 -0400
2193
2194    fix clk calculation on older lynx chips
2195
2196commit 7d1392be291e99fd3dde2292b82423b5f78d3a21
2197Author: Alex Deucher <alex@botch2.com>
2198Date:   Tue Mar 13 00:44:29 2007 -0400
2199
2200    bump to 1.5.0
2201
2202commit 449fedec0e38289d10d75799bbf995b0e7189911
2203Author: Alex Deucher <alex@botch2.com>
2204Date:   Tue Mar 13 00:13:35 2007 -0400
2205
2206    fix Solid()
2207    
2208    Unforunately, for 32 bpp solid fills the HW ignores alpha.
2209
2210commit 6d40f87d6cddde2d88f25be686b6392449cdecf1
2211Author: Alex Deucher <alex@botch2.com>
2212Date:   Mon Mar 12 23:58:41 2007 -0400
2213
2214    remove spurious waitqueue() line
2215
2216commit 452cabd6e3870b832cb19c8ad8a1fce27eef9c48
2217Author: Alex Deucher <alex@botch2.com>
2218Date:   Mon Mar 12 01:11:16 2007 -0400
2219
2220    Update copyright.
2221
2222commit 55266778ac693582131acae80def32211e09f18f
2223Author: Alex Deucher <alex@botch2.com>
2224Date:   Mon Mar 12 01:02:53 2007 -0400
2225
2226    switch the engine to quick start mode
2227    
2228    Rather than writing to the engine command register for every
2229    operation, set quick start mode, and the operation executes
2230    automatically when the width register is written.
2231
2232commit 4b37b514907da4a567261569eaebcc5a35aff6db
2233Author: Alex Deucher <alex@botch2.com>
2234Date:   Mon Mar 12 00:12:23 2007 -0400
2235
2236    fixup waitqueue slots
2237
2238commit 493c93de288437501fd898800bce45300379ebc3
2239Author: Alex Deucher <alex@botch2.com>
2240Date:   Sun Mar 11 18:49:48 2007 -0400
2241
2242    more exa fixups, disable UTS for now.
2243
2244commit c1d6373624428c8c52e4d1d295b1d77ed75391a0
2245Author: Alex Deucher <alex@botch2.com>
2246Date:   Sun Mar 11 16:22:05 2007 -0400
2247
2248    engine likes to use blit command for solid fills
2249
2250commit 9db380c99decdb9ab7fbf481788a5b63a84cd6e0
2251Author: Alex Deucher <alex@botch2.com>
2252Date:   Sun Mar 11 15:53:29 2007 -0400
2253
2254    fix postscalar shift
2255
2256commit b6082e628289221864e96ebc41ce93d8efe2d2ce
2257Author: Alex Deucher <alex@botch2.com>
2258Date:   Sun Mar 11 14:09:45 2007 -0400
2259
2260    support 32 bpp
2261
2262commit 3ebbc2e55a1584e01949ec760501ac4a3c086b04
2263Author: Alex Deucher <alex@botch2.com>
2264Date:   Sun Mar 11 13:40:06 2007 -0400
2265
2266    fix up vclk1 setup
2267
2268commit 3e62ec8736ea9ea64a2c0c5c41df9c6d742397ca
2269Author: Alex Deucher <alex@botch2.com>
2270Date:   Tue Mar 6 02:42:52 2007 -0500
2271
2272    revert the last commit.
2273    
2274    FBBase already take the register offset into account.
2275
2276commit 7da75a99159f80f215ee409377128f7dcbc4064d
2277Author: Alex Deucher <alex@botch2.com>
2278Date:   Tue Mar 6 02:38:02 2007 -0500
2279
2280    fix memoryBase
2281    
2282    pSmi->FBOffset is the offset from the start of the framebuffer
2283    pScrn->fbOffset is the aperture offset where the framebuffer
2284    starts
2285
2286commit ba0b2492269ea81e91f7ce4fb8da618187a3997f
2287Author: Alex Deucher <alex@botch2.com>
2288Date:   Tue Mar 6 02:24:08 2007 -0500
2289
2290    first pass at UTS
2291
2292commit e650ca36da5cf7614e4f1a0a524eadb46b1cc21d
2293Author: Alex Deucher <alex@botch2.com>
2294Date:   Mon Mar 5 23:58:25 2007 -0500
2295
2296    fix up EXA code
2297
2298commit c5b23ba581a3a4395b528d4ed824d5df986527e4
2299Author: Dennis De Winter <de.winter.dennis@gmail.com>
2300Date:   Sun Mar 4 23:17:57 2007 -0500
2301
2302    Initial import of EXA support
2303    
2304    Based on Dennis' code from bug 8721 with some minor
2305    changes by me.
2306
2307commit cf83ef7685609cd9daea193b4c842d232edcf506
2308Author: Alex Deucher <alex@botch2.com>
2309Date:   Sun Mar 4 17:18:10 2007 -0500
2310
2311    actually calculate and use vclk1
2312
2313commit 3030e5a6afda5d05ddee5b586295a9e608b7a0de
2314Author: Alex Deucher <alex@botch2.com>
2315Date:   Sun Mar 4 02:45:55 2007 -0500
2316
2317    Whitespace and formatting changes
2318
2319commit 606971fc8a323b0b33bb43524941b4bae6a0c334
2320Author: Alan Coopersmith <alan.coopersmith@sun.com>
2321Date:   Tue Feb 27 13:04:31 2007 -0800
2322
2323    renamed: .cvsignore -> .gitignore
2324
2325commit 4b3d28435d2f046d48a38694d89166fe38b00597
2326Author: Alan Coopersmith <alan.coopersmith@sun.com>
2327Date:   Tue Feb 27 13:04:01 2007 -0800
2328
2329    Replace XFree86 references in man page
2330
2331commit 3ff1d1aaf93df90f3a96b87bd8447db44b1722f5
2332Author: Alex Deucher <alex@samba.(none)>
2333Date:   Wed Jan 10 21:58:47 2007 -0500
2334
2335    only idle the engine on mode switches
2336    
2337    on the initial mode set the engine doesn't idle
2338    for some users.
2339
2340commit 72c1d44182a49dbd502bbdc0b48f69a43ad5aed0
2341Author: Adam Jackson <ajax@benzedrine.nwnk.net>
2342Date:   Fri Oct 13 18:41:44 2006 -0400
2343
2344    Bump to 1.4.2
2345
2346commit 36bbaaf3f44321fc7bb8ff1b71ff1d7f87f3b19b
2347Author: Alex Deucher <alex@samba.(none)>
2348Date:   Sat Sep 16 15:48:54 2006 -0400
2349
2350    make siliconmotion build again.  somehow a bunch of defines in
2351    regsmi.h got lost?  commit 9a58bc9adbc982680dd8df27caeb9a5819f2adad looks
2352    like the culprit, but siliconmotion has built fine subsequent to that commit.
2353    Not sure exactly what's going on.
2354
2355commit c090b2d97dbedfc278a4691b70cf3f73f9799aa6
2356Author:  <alex@samba.(none)>
2357Date:   Thu Jul 20 23:35:30 2006 -0400
2358
2359    - add dualhead support for lynx chips (pretty limited due to severe hardware
2360    limitations -- I'll improve it a bit at some point)
2361    - small driver clean ups
2362
2363commit 845d45cd2066be53f6e2e53d5af6375045dac6e2
2364Author: Adam Jackson <ajax@nwnk.net>
2365Date:   Sat Apr 8 01:43:52 2006 +0000
2366
2367    Bump to 1.4.1 for Xv changes.
2368
2369commit 2951e947c56909c9d51bf3f937502d0f682b1d76
2370Author: Aaron Plattner <aplattner@nvidia.com>
2371Date:   Fri Apr 7 23:13:59 2006 +0000
2372
2373    Add a DrawablePtr argument to the XV functions to pave the way for
2374        redirected video.
2375
2376commit 9a316eaea700e3a356f922506bf2c79e00fd9d83
2377Author: Adam Jackson <ajax@nwnk.net>
2378Date:   Fri Apr 7 20:58:57 2006 +0000
2379
2380    Unlibcwrap. Bump server version requirement. Bump to 1.4.0.
2381
2382commit c394f38c9a998eebec134a9a77ced46b9c005b9d
2383Author: Luc Verhaegen <libv@skynet.be>
2384Date:   Wed Feb 8 16:52:34 2006 +0000
2385
2386    SiliconMotion: Clean up insane pScrn->clock usage.
2387
2388commit 2453cfe259dec826fceab3cff1b2c3803d3d8042
2389Author: Kevin E Martin <kem@kem.org>
2390Date:   Wed Dec 21 02:30:02 2005 +0000
2391
2392    Update package version for X11R7 release.
2393
2394commit e398882fba13cd081916d2ff04ead4f9023c9bfb
2395Author: Adam Jackson <ajax@nwnk.net>
2396Date:   Mon Dec 19 16:25:54 2005 +0000
2397
2398    Stub COPYING files
2399
2400commit 818baab410deef2b3a3b6653dd2134dea29e98de
2401Author: Kevin E Martin <kem@kem.org>
2402Date:   Thu Dec 15 00:24:19 2005 +0000
2403
2404    Update package version number for final X11R7 release candidate.
2405
2406commit 6c95cd7b0ac0c113c25bf985365f3ccb1d30d7c9
2407Author: Kevin E Martin <kem@kem.org>
2408Date:   Tue Dec 6 22:48:37 2005 +0000
2409
2410    Change *man_SOURCES ==> *man_PRE to fix autotools warnings.
2411
2412commit 10d5542f95e86463b96cc1fd31378473c2394bc0
2413Author: Kevin E Martin <kem@kem.org>
2414Date:   Sat Dec 3 05:49:35 2005 +0000
2415
2416    Update package version number for X11R7 RC3 release.
2417
2418commit 8180ea0cb0df275831b900836d9387bb2a730de4
2419Author: Kevin E Martin <kem@kem.org>
2420Date:   Fri Dec 2 02:16:10 2005 +0000
2421
2422    Remove extraneous AC_MSG_RESULT.
2423
2424commit 73e22f89e205a61c315ffaca63cc7f5ab1f2df5f
2425Author: Adam Jackson <ajax@nwnk.net>
2426Date:   Tue Nov 29 23:30:03 2005 +0000
2427
2428    Only build dlloader modules by default.
2429
2430commit 8b156b75c37fb642d6c680cc9fb5361fb0a8ff74
2431Author: Alan Coopersmith <Alan.Coopersmith@sun.com>
2432Date:   Mon Nov 28 22:04:10 2005 +0000
2433
2434    Change *mandir targets to use new *_MAN_DIR variables set by xorg-macros.m4
2435        update to fix bug #5167 (Linux prefers *.1x man pages in man1 subdir)
2436
2437commit 8ad5ff0aa74eabc5b4dca2e0422d36713415b1b1
2438Author: Eric Anholt <anholt@freebsd.org>
2439Date:   Mon Nov 21 10:49:18 2005 +0000
2440
2441    Add .cvsignores for drivers.
2442
2443commit 0d4f340f197b14dacc5c33d82f0094ee1f9bb2a8
2444Author: Kevin E Martin <kem@kem.org>
2445Date:   Wed Nov 9 21:15:20 2005 +0000
2446
2447    Update package version number for X11R7 RC2 release.
2448
2449commit 280644786854eea8e8dbe7873e6c79d18f8048f0
2450Author: Kevin E Martin <kem@kem.org>
2451Date:   Tue Nov 1 15:08:54 2005 +0000
2452
2453    Update pkgcheck depedencies to work with separate build roots.
2454
2455commit bcc4f2555d4e4fa7ba2cf022165d3dd8ef331cd9
2456Author: Kevin E Martin <kem@kem.org>
2457Date:   Wed Oct 19 02:48:04 2005 +0000
2458
2459    Update package version number for RC1 release.
2460
2461commit 2da8c99697874a012dccfc582561a0b89e111ecd
2462Author: Alan Coopersmith <Alan.Coopersmith@sun.com>
2463Date:   Tue Oct 18 00:01:54 2005 +0000
2464
2465    Use @DRIVER_MAN_SUFFIX@ instead of $(DRIVER_MAN_SUFFIX) in macro
2466        substitutions to work better with BSD make
2467
2468commit 9f9d3fcd3dca04b7d960d8777e8f8ba612fa1510
2469Author: Adam Jackson <ajax@nwnk.net>
2470Date:   Mon Oct 17 22:57:35 2005 +0000
2471
2472    More 1.7 braindamage: define EXTRA_DIST in terms of @DRIVER_NAME@ instead
2473        of indirectly
2474
2475commit 286bf9de5baebd75ca129e50af9dc5f66dbf4e4f
2476Author: Alan Coopersmith <Alan.Coopersmith@sun.com>
2477Date:   Mon Oct 17 00:09:05 2005 +0000
2478
2479    Use sed & cpp to substitute variables in driver man pages
2480
2481commit 0db428f930db99d70897ca90b9ec0e130f4e6635
2482Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk>
2483Date:   Tue Sep 27 22:04:08 2005 +0000
2484
2485    Add some more documentation files for siliconmotion driver
2486
2487commit bdc896858ed819dcb79bb7443bfa57d63c8e0fad
2488Author: Daniel Stone <daniel@fooishbar.org>
2489Date:   Thu Aug 18 09:03:48 2005 +0000
2490
2491    Update autogen.sh to one that does objdir != srcdir.
2492
2493commit a3add2e05412767c49ca1c0a1069cd5805ea09e9
2494Author: Kevin E Martin <kem@kem.org>
2495Date:   Fri Jul 29 21:22:44 2005 +0000
2496
2497    Various changes preparing packages for RC0:
2498    - Verify and update package version numbers as needed
2499    - Implement versioning scheme
2500    - Change bug address to point to bugzilla bug entry form
2501    - Disable loadable i18n in libX11 by default (use --enable-loadable-i18n to
2502        reenable it)
2503    - Fix makedepend to use pkgconfig and pass distcheck
2504    - Update build script to build macros first
2505    - Update modular Xorg version
2506
2507commit 5754df51d9b0f8bb8dac7b8c0072da929f1241f0
2508Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk>
2509Date:   Tue Jul 26 22:12:58 2005 +0000
2510
2511    Build system for siliconmotion
2512
2513commit 4b98eb4b30d694139d5e597c445d65d637ab7472
2514Author: Adam Jackson <ajax@nwnk.net>
2515Date:   Mon Jul 11 02:29:59 2005 +0000
2516
2517    Prep for modular builds by adding guarded #include "config.h" everywhere.
2518
2519commit 6389dd3dc1259865a4b5fcc6b6c7b3f4429593cf
2520Author: Daniel Stone <daniel@fooishbar.org>
2521Date:   Fri Jul 1 22:43:12 2005 +0000
2522
2523    Change all misc.h and os.h references to <X11/foo.h>.
2524
2525commit 1dc135becf5898950e04c84cc4d87870840140c1
2526Author: Alex Deucher <agd5f@yahoo.com>
2527Date:   Thu Jun 30 00:14:26 2005 +0000
2528
2529    - add 800x480 mode to smi driver (bug 3639)
2530
2531commit 0670b675c2133dd7c29f79bdf17d24d64181537d
2532Author: Adam Jackson <ajax@nwnk.net>
2533Date:   Sat Jun 25 21:16:56 2005 +0000
2534
2535    Bug #3626: _X_EXPORT tags for video and input drivers.
2536
2537commit 39112b558fdccd2b0a15b8e41d6215e3b0e6e7bd
2538Author: Daniel Stone <daniel@fooishbar.org>
2539Date:   Wed Apr 20 12:25:29 2005 +0000
2540
2541    Fix includes right throughout the Xserver tree:
2542    change "foo.h" to <X11/foo.h> for core headers, e.g. X.h, Xpoll.h;
2543    change "foo.h", "extensions/foo.h" and "X11/foo.h" to
2544        <X11/extensions/foo.h> for extension headers, e.g. Xv.h;
2545    change "foo.[ch]" to <X11/Xtrans/foo.[ch]> for Xtrans files.
2546
2547commit c34e24d83a0f84ad90b7cbaba78d54cc779f74d5
2548Author: Adam Jackson <ajax@nwnk.net>
2549Date:   Fri Sep 17 03:04:54 2004 +0000
2550
2551    Bug #1192: Remove cfb support from drivers where its use is an option.
2552        Delete xf24_32bpp, as s3virge was the last user. Fix up some comments
2553        to refer to fb rather than cfb.
2554
2555commit 61457616aea5d5d7d98e869d3b24e989c2404d6b
2556Author: Adam Jackson <ajax@nwnk.net>
2557Date:   Mon Aug 16 09:13:15 2004 +0000
2558
2559    Bug #1087: Make sure all the *Weak symbols are documented in the import
2560        list, so the loader knows not to complain about unresolved symbols when
2561        an optional module isn't loaded. This typically manifests as warnings
2562        about fbdevHW symbols when the user isn't using the framebuffer
2563        console.
2564
2565commit 74176c460a4f5206cbc7cd7b9f48d1b6cedae7cb
2566Author: Eric Anholt <anholt@freebsd.org>
2567Date:   Sat Jul 31 01:21:51 2004 +0000
2568
2569    Rename some COMPOSITE macros and enums to avoid conflicting with the
2570        COMPOSITE extension's new macro. These appear to only be used
2571        internally.
2572
2573commit 0f239a7f12e1c0a0b7ede424b0f1eb0fe189e9d6
2574Author: Adam Jackson <ajax@nwnk.net>
2575Date:   Sat Jul 31 01:21:19 2004 +0000
2576
2577    Change several LoaderSymbol calls introduced by the bug #400 patch to
2578    *Weak() resolver functions.
2579
2580commit 865ba2a4c1107bb82d22ed648f8a1d63c833d6fc
2581Author: Adam Jackson <ajax@nwnk.net>
2582Date:   Fri Jul 30 20:30:54 2004 +0000
2583
2584    Bug #400 (partial): Driver fixes for the dlloader. When using dlloader, all
2585        framebuffer formats except cfb and the overlay modes should work, and
2586        r128 and radeon need to be loaded from the ati driver (both issues to
2587        be fixed soon). Tested on i740, s3virge, mach64, tdfx, vesa, and vga
2588        drivers. elfloader users shouldn't be affected.
2589
2590commit 9a58bc9adbc982680dd8df27caeb9a5819f2adad
2591Author: Eric Anholt <anholt@freebsd.org>
2592Date:   Wed Jun 16 09:23:38 2004 +0000
2593
2594    DRI XFree86-4_3_99_12-merge import
2595
2596commit 1f957fb7e44762197b0fc5a3b7def25c7888929a
2597Author: Egbert Eich <eich@suse.de>
2598Date:   Wed May 26 16:24:09 2004 +0000
2599
2600    Fixing setting of physical framebuffer base for several drivers. C&T
2601        driver: Fixed setting of minimal clocks for HiQV chipsets. Neomagic
2602        driver: improved support for lowres double scan modes.
2603
2604commit 445708ff1249e96628dd6b11e20f83e568fb4ad1
2605Author: Egbert Eich <eich@suse.de>
2606Date:   Fri Apr 23 19:45:06 2004 +0000
2607
2608    Merging XORG-CURRENT into trunk
2609
2610commit 90141c8b6da8e6b57f2e4bae486a3d39c4069a8b
2611Author: Egbert Eich <eich@suse.de>
2612Date:   Sun Mar 14 08:33:31 2004 +0000
2613
2614    Importing vendor version xf86-4_4_99_1 on Sun Mar 14 00:26:39 PST 2004
2615
2616commit 1082b4e0760359234cb5f51b953bbffc65d594ec
2617Author: Egbert Eich <eich@suse.de>
2618Date:   Wed Mar 3 12:12:24 2004 +0000
2619
2620    Importing vendor version xf86-4_4_0 on Wed Mar 3 04:09:24 PST 2004
2621
2622commit b1c51d42211f715043a3252550ef2fde1a3065e5
2623Author: Egbert Eich <eich@suse.de>
2624Date:   Thu Feb 26 13:35:54 2004 +0000
2625
2626    readding XFree86's cvs IDs
2627
2628commit 2ae35696e96079b192c2cac4e092f6deeb300325
2629Author: Egbert Eich <eich@suse.de>
2630Date:   Thu Feb 26 09:23:23 2004 +0000
2631
2632    Importing vendor version xf86-4_3_99_903 on Wed Feb 26 01:21:00 PST 2004
2633
2634commit ac5b2461e620df7026eefda6a389127e7d41f67d
2635Author: Kaleb Keithley <kaleb@freedesktop.org>
2636Date:   Fri Dec 19 20:55:14 2003 +0000
2637
2638    XFree86 4.3.99.902 (RC 2)
2639
2640commit 14cd4af100e1deb1c7b63e9f19587e38ce1f08f2
2641Author: Kaleb Keithley <kaleb@freedesktop.org>
2642Date:   Tue Nov 25 19:28:39 2003 +0000
2643
2644    XFree86 4.3.99.16 Bring the tree up to date for the Cygwin folks
2645
2646commit 1684aa7a21a194b6e7dd518f3895c9b2dfce229a
2647Author: Kaleb Keithley <kaleb@freedesktop.org>
2648Date:   Fri Nov 14 16:48:55 2003 +0000
2649
2650    XFree86 4.3.0.1
2651
2652commit 78af703cb019a60cc93843efcd4889ccb15abd13
2653Author: Kaleb Keithley <kaleb@freedesktop.org>
2654Date:   Fri Nov 14 16:48:55 2003 +0000
2655
2656    Initial revision
2657