wd.c revision 1.471
1 /* $NetBSD: wd.c,v 1.471 2025/02/17 18:45:01 jakllsch Exp $ */ 2 3 /* 4 * Copyright (c) 1998, 2001 Manuel Bouyer. All rights reserved. 5 * 6 * Redistribution and use in source and binary forms, with or without 7 * modification, are permitted provided that the following conditions 8 * are met: 9 * 1. Redistributions of source code must retain the above copyright 10 * notice, this list of conditions and the following disclaimer. 11 * 2. Redistributions in binary form must reproduce the above copyright 12 * notice, this list of conditions and the following disclaimer in the 13 * documentation and/or other materials provided with the distribution. 14 * 15 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR 16 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 17 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. 18 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, 19 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT 20 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 21 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 22 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 23 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF 24 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 25 */ 26 27 /*- 28 * Copyright (c) 1998, 2003, 2004 The NetBSD Foundation, Inc. 29 * All rights reserved. 30 * 31 * This code is derived from software contributed to The NetBSD Foundation 32 * by Charles M. Hannum and by Onno van der Linden. 33 * 34 * Redistribution and use in source and binary forms, with or without 35 * modification, are permitted provided that the following conditions 36 * are met: 37 * 1. Redistributions of source code must retain the above copyright 38 * notice, this list of conditions and the following disclaimer. 39 * 2. Redistributions in binary form must reproduce the above copyright 40 * notice, this list of conditions and the following disclaimer in the 41 * documentation and/or other materials provided with the distribution. 42 * 43 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 44 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 45 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 46 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 47 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 48 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 49 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 50 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 51 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 52 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 53 * POSSIBILITY OF SUCH DAMAGE. 54 */ 55 56 #include <sys/cdefs.h> 57 __KERNEL_RCSID(0, "$NetBSD: wd.c,v 1.471 2025/02/17 18:45:01 jakllsch Exp $"); 58 59 #include "opt_ata.h" 60 #include "opt_wd.h" 61 62 #include <sys/param.h> 63 #include <sys/systm.h> 64 #include <sys/kernel.h> 65 #include <sys/conf.h> 66 #include <sys/file.h> 67 #include <sys/stat.h> 68 #include <sys/ioctl.h> 69 #include <sys/buf.h> 70 #include <sys/bufq.h> 71 #include <sys/uio.h> 72 #include <sys/device.h> 73 #include <sys/disklabel.h> 74 #include <sys/disk.h> 75 #include <sys/syslog.h> 76 #include <sys/proc.h> 77 #include <sys/reboot.h> 78 #include <sys/vnode.h> 79 #include <sys/rndsource.h> 80 81 #include <sys/intr.h> 82 #include <sys/bus.h> 83 84 #include <dev/ata/atareg.h> 85 #include <dev/ata/atavar.h> 86 #include <dev/ata/wdvar.h> 87 #include <dev/ic/wdcreg.h> 88 #include <sys/ataio.h> 89 #include "locators.h" 90 91 #include <prop/proplib.h> 92 93 #define WDIORETRIES_SINGLE 4 /* number of retries for single-sector */ 94 #define WDIORETRIES 5 /* number of retries before giving up */ 95 #define RECOVERYTIME hz/2 /* time to wait before retrying a cmd */ 96 97 #define WDUNIT(dev) DISKUNIT(dev) 98 #define WDPART(dev) DISKPART(dev) 99 #define WDMINOR(unit, part) DISKMINOR(unit, part) 100 #define MAKEWDDEV(maj, unit, part) MAKEDISKDEV(maj, unit, part) 101 102 #define WDLABELDEV(dev) (MAKEWDDEV(major(dev), WDUNIT(dev), RAW_PART)) 103 104 #define DEBUG_FUNCS 0x08 105 #define DEBUG_PROBE 0x10 106 #define DEBUG_DETACH 0x20 107 #define DEBUG_XFERS 0x40 108 #ifdef ATADEBUG 109 #ifndef ATADEBUG_WD_MASK 110 #define ATADEBUG_WD_MASK 0x0 111 #endif 112 int wdcdebug_wd_mask = ATADEBUG_WD_MASK; 113 #define ATADEBUG_PRINT(args, level) \ 114 if (wdcdebug_wd_mask & (level)) \ 115 printf args 116 #else 117 #define ATADEBUG_PRINT(args, level) 118 #endif 119 120 static int wdprobe(device_t, cfdata_t, void *); 121 static void wdattach(device_t, device_t, void *); 122 static int wddetach(device_t, int); 123 static void wdperror(const struct wd_softc *, struct ata_xfer *); 124 125 static void wdminphys(struct buf *); 126 127 static int wd_firstopen(device_t, dev_t, int, int); 128 static int wd_lastclose(device_t); 129 static bool wd_suspend(device_t, const pmf_qual_t *); 130 static int wd_standby(struct wd_softc *, int); 131 132 CFATTACH_DECL3_NEW(wd, sizeof(struct wd_softc), 133 wdprobe, wdattach, wddetach, NULL, NULL, NULL, DVF_DETACH_SHUTDOWN); 134 135 extern struct cfdriver wd_cd; 136 137 static dev_type_open(wdopen); 138 static dev_type_close(wdclose); 139 static dev_type_read(wdread); 140 static dev_type_write(wdwrite); 141 static dev_type_ioctl(wdioctl); 142 static dev_type_strategy(wdstrategy); 143 static dev_type_dump(wddump); 144 static dev_type_size(wdsize); 145 static dev_type_discard(wddiscard); 146 147 const struct bdevsw wd_bdevsw = { 148 .d_open = wdopen, 149 .d_close = wdclose, 150 .d_strategy = wdstrategy, 151 .d_ioctl = wdioctl, 152 .d_dump = wddump, 153 .d_psize = wdsize, 154 .d_discard = wddiscard, 155 .d_cfdriver = &wd_cd, 156 .d_devtounit = disklabel_dev_unit, 157 .d_flag = D_DISK 158 }; 159 160 const struct cdevsw wd_cdevsw = { 161 .d_open = wdopen, 162 .d_close = wdclose, 163 .d_read = wdread, 164 .d_write = wdwrite, 165 .d_ioctl = wdioctl, 166 .d_stop = nostop, 167 .d_tty = notty, 168 .d_poll = nopoll, 169 .d_mmap = nommap, 170 .d_kqfilter = nokqfilter, 171 .d_discard = wddiscard, 172 .d_cfdriver = &wd_cd, 173 .d_devtounit = disklabel_dev_unit, 174 .d_flag = D_DISK 175 }; 176 177 /* #define WD_DUMP_NOT_TRUSTED if you just want to watch */ 178 static int wddoingadump = 0; 179 static int wddumprecalibrated = 0; 180 181 /* 182 * Glue necessary to hook WDCIOCCOMMAND into physio 183 */ 184 185 struct wd_ioctl { 186 LIST_ENTRY(wd_ioctl) wi_list; 187 struct buf wi_bp; 188 struct uio wi_uio; 189 struct iovec wi_iov; 190 atareq_t wi_atareq; 191 struct wd_softc *wi_softc; 192 }; 193 194 static struct wd_ioctl *wi_find(struct buf *); 195 static void wi_free(struct wd_ioctl *); 196 static struct wd_ioctl *wi_get(struct wd_softc *); 197 static void wdioctlstrategy(struct buf *); 198 199 static void wdrestart(void *); 200 static void wdstart1(struct wd_softc *, struct buf *, struct ata_xfer *); 201 static int wd_diskstart(device_t, struct buf *); 202 static int wd_dumpblocks(device_t, void *, daddr_t, int); 203 static void wd_iosize(device_t, int *); 204 static int wd_discard(device_t, off_t, off_t); 205 static void wdbioretry(void *); 206 static void wdbiorequeue(void *); 207 static void wddone(device_t, struct ata_xfer *); 208 static int wd_get_params(struct wd_softc *, struct ataparams *); 209 static void wd_set_geometry(struct wd_softc *); 210 static int wd_flushcache(struct wd_softc *, int); 211 static int wd_trim(struct wd_softc *, daddr_t, long); 212 static bool wd_shutdown(device_t, int); 213 214 static int wd_getcache(struct wd_softc *, int *); 215 static int wd_setcache(struct wd_softc *, int); 216 217 static void wd_sysctl_attach(struct wd_softc *); 218 static void wd_sysctl_detach(struct wd_softc *); 219 220 static const struct dkdriver wddkdriver = { 221 .d_open = wdopen, 222 .d_close = wdclose, 223 .d_strategy = wdstrategy, 224 .d_minphys = wdminphys, 225 .d_diskstart = wd_diskstart, 226 .d_dumpblocks = wd_dumpblocks, 227 .d_iosize = wd_iosize, 228 .d_firstopen = wd_firstopen, 229 .d_lastclose = wd_lastclose, 230 .d_discard = wd_discard 231 }; 232 233 #ifdef HAS_BAD144_HANDLING 234 static void bad144intern(struct wd_softc *); 235 #endif 236 237 #define WD_QUIRK_SPLIT_MOD15_WRITE 0x0001 /* must split certain writes */ 238 239 #define WD_QUIRK_FMT "\20\1SPLIT_MOD15_WRITE" 240 241 /* 242 * Quirk table for IDE drives. Put more-specific matches first, since 243 * a simple globing routine is used for matching. 244 */ 245 static const struct wd_quirk { 246 const char *wdq_match; /* inquiry pattern to match */ 247 int wdq_quirks; /* drive quirks */ 248 } wd_quirk_table[] = { 249 /* 250 * Some Seagate S-ATA drives have a PHY which can get confused 251 * with the way data is packetized by some S-ATA controllers. 252 * 253 * The work-around is to split in two any write transfer whose 254 * sector count % 15 == 1 (assuming 512 byte sectors). 255 * 256 * XXX This is an incomplete list. There are at least a couple 257 * XXX more model numbers. If you have trouble with such transfers 258 * XXX (8K is the most common) on Seagate S-ATA drives, please 259 * XXX notify thorpej (at) NetBSD.org. 260 * 261 * The ST360015AS has not yet been confirmed to have this 262 * issue, however, it is the only other drive in the 263 * Seagate Barracuda Serial ATA V family. 264 * 265 */ 266 { "ST3120023AS", WD_QUIRK_SPLIT_MOD15_WRITE }, 267 { "ST380023AS", WD_QUIRK_SPLIT_MOD15_WRITE }, 268 { "ST360015AS", WD_QUIRK_SPLIT_MOD15_WRITE }, 269 { NULL, 270 0 } 271 }; 272 273 static const struct wd_quirk * 274 wd_lookup_quirks(const char *name) 275 { 276 const struct wd_quirk *wdq; 277 const char *estr; 278 279 for (wdq = wd_quirk_table; wdq->wdq_match != NULL; wdq++) { 280 /* 281 * We only want exact matches (which include matches 282 * against globbing characters). 283 */ 284 if (pmatch(name, wdq->wdq_match, &estr) == 2) 285 return (wdq); 286 } 287 return (NULL); 288 } 289 290 static int 291 wdprobe(device_t parent, cfdata_t match, void *aux) 292 { 293 struct ata_device *adev = aux; 294 295 if (adev == NULL) 296 return 0; 297 if (adev->adev_bustype->bustype_type != SCSIPI_BUSTYPE_ATA) 298 return 0; 299 300 if (match->cf_loc[ATA_HLCF_DRIVE] != ATA_HLCF_DRIVE_DEFAULT && 301 match->cf_loc[ATA_HLCF_DRIVE] != adev->adev_drv_data->drive) 302 return 0; 303 return 1; 304 } 305 306 static void 307 wdattach(device_t parent, device_t self, void *aux) 308 { 309 struct wd_softc *wd = device_private(self); 310 struct dk_softc *dksc = &wd->sc_dksc; 311 struct ata_device *adev= aux; 312 int i, blank; 313 char tbuf[41],pbuf[9], c, *p, *q; 314 const struct wd_quirk *wdq; 315 int dtype = DKTYPE_UNKNOWN; 316 317 dksc->sc_dev = self; 318 319 ATADEBUG_PRINT(("wdattach\n"), DEBUG_FUNCS | DEBUG_PROBE); 320 mutex_init(&wd->sc_lock, MUTEX_DEFAULT, IPL_BIO); 321 #ifdef WD_SOFTBADSECT 322 SLIST_INIT(&wd->sc_bslist); 323 cv_init(&wd->sc_bslist_cv, "wdbadsect"); 324 #endif 325 wd->atabus = adev->adev_bustype; 326 wd->inflight = 0; 327 wd->drvp = adev->adev_drv_data; 328 329 wd->drvp->drv_openings = 1; 330 wd->drvp->drv_done = wddone; 331 wd->drvp->drv_softc = dksc->sc_dev; /* done in atabusconfig_thread() 332 but too late */ 333 334 SLIST_INIT(&wd->sc_retry_list); 335 SLIST_INIT(&wd->sc_requeue_list); 336 callout_init(&wd->sc_retry_callout, 0); /* XXX MPSAFE */ 337 callout_init(&wd->sc_requeue_callout, 0); /* XXX MPSAFE */ 338 callout_init(&wd->sc_restart_diskqueue, 0); /* XXX MPSAFE */ 339 340 aprint_naive("\n"); 341 aprint_normal("\n"); 342 343 /* read our drive info */ 344 if (wd_get_params(wd, &wd->sc_params) != 0) { 345 aprint_error_dev(self, "IDENTIFY failed\n"); 346 goto out; 347 } 348 349 for (blank = 0, p = wd->sc_params.atap_model, q = tbuf, i = 0; 350 i < sizeof(wd->sc_params.atap_model); i++) { 351 c = *p++; 352 if (c == '\0') 353 break; 354 if (c != ' ') { 355 if (blank) { 356 *q++ = ' '; 357 blank = 0; 358 } 359 *q++ = c; 360 } else 361 blank = 1; 362 } 363 *q++ = '\0'; 364 365 wd->sc_typename = kmem_asprintf("%s", tbuf); 366 aprint_normal_dev(self, "<%s>\n", wd->sc_typename); 367 368 wdq = wd_lookup_quirks(tbuf); 369 if (wdq != NULL) 370 wd->sc_quirks = wdq->wdq_quirks; 371 372 if (wd->sc_quirks != 0) { 373 char sbuf[sizeof(WD_QUIRK_FMT) + 64]; 374 snprintb(sbuf, sizeof(sbuf), WD_QUIRK_FMT, wd->sc_quirks); 375 aprint_normal_dev(self, "quirks %s\n", sbuf); 376 377 if (wd->sc_quirks & WD_QUIRK_SPLIT_MOD15_WRITE) { 378 aprint_error_dev(self, "drive corrupts write transfers with certain controllers, consider replacing\n"); 379 } 380 } 381 382 if ((wd->sc_params.atap_multi & 0xff) > 1) { 383 wd->drvp->multi = wd->sc_params.atap_multi & 0xff; 384 } else { 385 wd->drvp->multi = 1; 386 } 387 388 aprint_verbose_dev(self, "drive supports %d-sector PIO transfers,", 389 wd->drvp->multi); 390 391 /* 48-bit LBA addressing */ 392 if ((wd->sc_params.atap_cmd2_en & ATA_CMD2_LBA48) != 0) 393 wd->sc_flags |= WDF_LBA48; 394 395 /* Prior to ATA-4, LBA was optional. */ 396 if ((wd->sc_params.atap_capabilities1 & WDC_CAP_LBA) != 0) 397 wd->sc_flags |= WDF_LBA; 398 #if 0 399 /* ATA-4 requires LBA. */ 400 if (wd->sc_params.atap_ataversion != 0xffff && 401 wd->sc_params.atap_ataversion >= WDC_VER_ATA4) 402 wd->sc_flags |= WDF_LBA; 403 #endif 404 405 if ((wd->sc_flags & WDF_LBA48) != 0) { 406 aprint_verbose(" LBA48 addressing\n"); 407 wd->sc_capacity = 408 ((uint64_t) wd->sc_params.atap_max_lba[3] << 48) | 409 ((uint64_t) wd->sc_params.atap_max_lba[2] << 32) | 410 ((uint64_t) wd->sc_params.atap_max_lba[1] << 16) | 411 ((uint64_t) wd->sc_params.atap_max_lba[0] << 0); 412 wd->sc_capacity28 = 413 (wd->sc_params.atap_capacity[1] << 16) | 414 wd->sc_params.atap_capacity[0]; 415 /* 416 * Force LBA48 addressing for invalid numbers. 417 */ 418 if (wd->sc_capacity28 > 0xfffffff) 419 wd->sc_capacity28 = 0xfffffff; 420 } else if ((wd->sc_flags & WDF_LBA) != 0) { 421 aprint_verbose(" LBA addressing\n"); 422 wd->sc_capacity28 = 423 (wd->sc_params.atap_capacity[1] << 16) | 424 wd->sc_params.atap_capacity[0]; 425 /* 426 * Limit capacity to LBA28 numbers to avoid overflow. 427 */ 428 if (wd->sc_capacity28 > 0xfffffff) 429 wd->sc_capacity28 = 0xfffffff; 430 wd->sc_capacity = wd->sc_capacity28; 431 } else { 432 aprint_verbose(" chs addressing\n"); 433 wd->sc_capacity = 434 wd->sc_params.atap_cylinders * 435 wd->sc_params.atap_heads * 436 wd->sc_params.atap_sectors; 437 /* 438 * LBA28 size is ignored for CHS addressing. Use a reasonable 439 * value for debugging. The CHS values may be artificial and 440 * are mostly ignored. 441 */ 442 if (wd->sc_capacity < 0xfffffff) 443 wd->sc_capacity28 = wd->sc_capacity; 444 else 445 wd->sc_capacity28 = 0xfffffff; 446 } 447 if ((wd->sc_params.atap_secsz & ATA_SECSZ_VALID_MASK) == ATA_SECSZ_VALID 448 && ((wd->sc_params.atap_secsz & ATA_SECSZ_LLS) != 0)) { 449 wd->sc_blksize = 2ULL * 450 ((uint32_t)((wd->sc_params.atap_lls_secsz[1] << 16) | 451 wd->sc_params.atap_lls_secsz[0])); 452 } else { 453 wd->sc_blksize = 512; 454 } 455 wd->sc_sectoralign.dsa_firstaligned = 0; 456 wd->sc_sectoralign.dsa_alignment = 1; 457 if ((wd->sc_params.atap_secsz & ATA_SECSZ_VALID_MASK) == ATA_SECSZ_VALID 458 && ((wd->sc_params.atap_secsz & ATA_SECSZ_LPS) != 0)) { 459 wd->sc_sectoralign.dsa_alignment = 1 << 460 (wd->sc_params.atap_secsz & ATA_SECSZ_LPS_SZMSK); 461 if ((wd->sc_params.atap_logical_align & ATA_LA_VALID_MASK) == 462 ATA_LA_VALID) { 463 wd->sc_sectoralign.dsa_firstaligned = 464 (wd->sc_sectoralign.dsa_alignment - 465 (wd->sc_params.atap_logical_align & 466 ATA_LA_MASK)); 467 } 468 } 469 wd->sc_capacity512 = (wd->sc_capacity * wd->sc_blksize) / DEV_BSIZE; 470 format_bytes(pbuf, sizeof(pbuf), wd->sc_capacity * wd->sc_blksize); 471 aprint_normal_dev(self, "%s, %d cyl, %d head, %d sec, " 472 "%d bytes/sect x %llu sectors", 473 pbuf, 474 (wd->sc_flags & WDF_LBA) ? (int)(wd->sc_capacity / 475 (wd->sc_params.atap_heads * wd->sc_params.atap_sectors)) : 476 wd->sc_params.atap_cylinders, 477 wd->sc_params.atap_heads, wd->sc_params.atap_sectors, 478 wd->sc_blksize, (unsigned long long)wd->sc_capacity); 479 if (wd->sc_sectoralign.dsa_alignment != 1) { 480 aprint_normal(" (%d bytes/physsect", 481 wd->sc_sectoralign.dsa_alignment * wd->sc_blksize); 482 if (wd->sc_sectoralign.dsa_firstaligned != 0) { 483 aprint_normal("; first aligned sector: %jd", 484 (intmax_t)wd->sc_sectoralign.dsa_firstaligned); 485 } 486 aprint_normal(")"); 487 } 488 aprint_normal("\n"); 489 490 ATADEBUG_PRINT(("%s: atap_dmatiming_mimi=%d, atap_dmatiming_recom=%d\n", 491 device_xname(self), wd->sc_params.atap_dmatiming_mimi, 492 wd->sc_params.atap_dmatiming_recom), DEBUG_PROBE); 493 494 if (wd->sc_blksize <= 0 || !powerof2(wd->sc_blksize) || 495 wd->sc_blksize < DEV_BSIZE || wd->sc_blksize > MAXPHYS) { 496 aprint_normal_dev(self, "WARNING: block size %u " 497 "might not actually work\n", wd->sc_blksize); 498 } 499 500 if (strcmp(wd->sc_params.atap_model, "ST506") == 0) 501 dtype = DKTYPE_ST506; 502 else 503 dtype = DKTYPE_ESDI; 504 505 out: 506 /* 507 * Initialize and attach the disk structure. 508 */ 509 dk_init(dksc, self, dtype); 510 disk_init(&dksc->sc_dkdev, dksc->sc_xname, &wddkdriver); 511 512 /* Attach dk and disk subsystems */ 513 dk_attach(dksc); 514 disk_attach(&dksc->sc_dkdev); 515 wd_set_geometry(wd); 516 517 bufq_alloc(&dksc->sc_bufq, BUFQ_DISK_DEFAULT_STRAT, BUFQ_SORT_RAWBLOCK); 518 519 /* reference to label structure, used by ata code */ 520 wd->drvp->lp = dksc->sc_dkdev.dk_label; 521 522 /* Discover wedges on this disk. */ 523 dkwedge_discover(&dksc->sc_dkdev); 524 525 if (!pmf_device_register1(self, wd_suspend, NULL, wd_shutdown)) 526 aprint_error_dev(self, "couldn't establish power handler\n"); 527 528 wd_sysctl_attach(wd); 529 } 530 531 static bool 532 wd_suspend(device_t dv, const pmf_qual_t *qual) 533 { 534 struct wd_softc *sc = device_private(dv); 535 536 /* the adapter needs to be enabled */ 537 if (sc->atabus->ata_addref(sc->drvp)) 538 return true; /* no need to complain */ 539 540 wd_flushcache(sc, AT_WAIT); 541 wd_standby(sc, AT_WAIT); 542 543 sc->atabus->ata_delref(sc->drvp); 544 return true; 545 } 546 547 static int 548 wddetach(device_t self, int flags) 549 { 550 struct wd_softc *wd = device_private(self); 551 struct dk_softc *dksc = &wd->sc_dksc; 552 int bmaj, cmaj, i, mn, rc; 553 554 if ((rc = disk_begindetach(&dksc->sc_dkdev, wd_lastclose, self, flags)) != 0) 555 return rc; 556 557 /* locate the major number */ 558 bmaj = bdevsw_lookup_major(&wd_bdevsw); 559 cmaj = cdevsw_lookup_major(&wd_cdevsw); 560 561 /* Nuke the vnodes for any open instances. */ 562 for (i = 0; i < MAXPARTITIONS; i++) { 563 mn = WDMINOR(device_unit(self), i); 564 vdevgone(bmaj, mn, mn, VBLK); 565 vdevgone(cmaj, mn, mn, VCHR); 566 } 567 568 dk_drain(dksc); 569 570 /* Kill off any pending commands. */ 571 mutex_enter(&wd->sc_lock); 572 wd->atabus->ata_killpending(wd->drvp); 573 574 callout_halt(&wd->sc_retry_callout, &wd->sc_lock); 575 callout_destroy(&wd->sc_retry_callout); 576 callout_halt(&wd->sc_requeue_callout, &wd->sc_lock); 577 callout_destroy(&wd->sc_requeue_callout); 578 callout_halt(&wd->sc_restart_diskqueue, &wd->sc_lock); 579 callout_destroy(&wd->sc_restart_diskqueue); 580 581 mutex_exit(&wd->sc_lock); 582 583 bufq_free(dksc->sc_bufq); 584 585 /* Delete all of our wedges. */ 586 dkwedge_delall(&dksc->sc_dkdev); 587 588 if (flags & DETACH_POWEROFF) 589 wd_standby(wd, AT_POLL); 590 591 /* Detach from the disk list. */ 592 disk_detach(&dksc->sc_dkdev); 593 disk_destroy(&dksc->sc_dkdev); 594 595 dk_detach(dksc); 596 597 #ifdef WD_SOFTBADSECT 598 /* Clean out the bad sector list */ 599 while (!SLIST_EMPTY(&wd->sc_bslist)) { 600 struct disk_badsectors *dbs = SLIST_FIRST(&wd->sc_bslist); 601 SLIST_REMOVE_HEAD(&wd->sc_bslist, dbs_next); 602 kmem_free(dbs, sizeof(*dbs)); 603 } 604 wd->sc_bscount = 0; 605 #endif 606 if (wd->sc_typename != NULL) { 607 kmem_free(wd->sc_typename, strlen(wd->sc_typename) + 1); 608 wd->sc_typename = NULL; 609 } 610 611 pmf_device_deregister(self); 612 613 wd_sysctl_detach(wd); 614 615 #ifdef WD_SOFTBADSECT 616 KASSERT(SLIST_EMPTY(&wd->sc_bslist)); 617 cv_destroy(&wd->sc_bslist_cv); 618 #endif 619 620 mutex_destroy(&wd->sc_lock); 621 622 wd->drvp->drive_type = ATA_DRIVET_NONE; /* no drive any more here */ 623 wd->drvp->drive_flags = 0; 624 625 return (0); 626 } 627 628 /* 629 * Read/write routine for a buffer. Validates the arguments and schedules the 630 * transfer. Does not wait for the transfer to complete. 631 */ 632 static void 633 wdstrategy(struct buf *bp) 634 { 635 struct wd_softc *wd = 636 device_lookup_private(&wd_cd, WDUNIT(bp->b_dev)); 637 struct dk_softc *dksc = &wd->sc_dksc; 638 639 ATADEBUG_PRINT(("wdstrategy (%s)\n", dksc->sc_xname), 640 DEBUG_XFERS); 641 642 /* If device invalidated (e.g. media change, door open, 643 * device detachment), then error. 644 */ 645 if ((wd->sc_flags & WDF_LOADED) == 0 || 646 !device_is_enabled(dksc->sc_dev)) 647 goto err; 648 649 #ifdef WD_SOFTBADSECT 650 /* 651 * If the transfer about to be attempted contains only a block that 652 * is known to be bad then return an error for the transfer without 653 * even attempting to start a transfer up under the premis that we 654 * will just end up doing more retries for a transfer that will end 655 * up failing again. 656 */ 657 if (__predict_false(!SLIST_EMPTY(&wd->sc_bslist))) { 658 struct disklabel *lp = dksc->sc_dkdev.dk_label; 659 struct disk_badsectors *dbs; 660 daddr_t blkno, maxblk; 661 662 /* convert the block number to absolute */ 663 if (lp->d_secsize >= DEV_BSIZE) 664 blkno = bp->b_blkno / (lp->d_secsize / DEV_BSIZE); 665 else 666 blkno = bp->b_blkno * (DEV_BSIZE / lp->d_secsize); 667 if (WDPART(bp->b_dev) != RAW_PART) 668 blkno += lp->d_partitions[WDPART(bp->b_dev)].p_offset; 669 maxblk = blkno + (bp->b_bcount / wd->sc_blksize) - 1; 670 671 mutex_enter(&wd->sc_lock); 672 SLIST_FOREACH(dbs, &wd->sc_bslist, dbs_next) 673 if ((dbs->dbs_min <= bp->b_rawblkno && 674 bp->b_rawblkno <= dbs->dbs_max) || 675 (dbs->dbs_min <= maxblk && maxblk <= dbs->dbs_max)){ 676 mutex_exit(&wd->sc_lock); 677 goto err; 678 } 679 mutex_exit(&wd->sc_lock); 680 } 681 #endif 682 683 dk_strategy(dksc, bp); 684 return; 685 686 err: 687 bp->b_error = EIO; 688 bp->b_resid = bp->b_bcount; 689 biodone(bp); 690 } 691 692 static void 693 wdstart1(struct wd_softc *wd, struct buf *bp, struct ata_xfer *xfer) 694 { 695 struct dk_softc *dksc = &wd->sc_dksc; 696 const uint32_t secsize = dksc->sc_dkdev.dk_geom.dg_secsize; 697 698 KASSERT(bp == xfer->c_bio.bp || xfer->c_bio.bp == NULL); 699 KASSERT((xfer->c_flags & (C_WAITACT|C_FREE)) == 0); 700 KASSERT(mutex_owned(&wd->sc_lock)); 701 702 /* Reset state, so that retries don't use stale info */ 703 if (__predict_false(xfer->c_retries > 0)) { 704 xfer->c_flags = 0; 705 memset(&xfer->c_bio, 0, sizeof(xfer->c_bio)); 706 } 707 708 xfer->c_bio.blkno = bp->b_rawblkno; 709 xfer->c_bio.bcount = bp->b_bcount; 710 xfer->c_bio.databuf = bp->b_data; 711 xfer->c_bio.blkdone = 0; 712 xfer->c_bio.bp = bp; 713 714 /* Adjust blkno and bcount if xfer has been already partially done */ 715 if (__predict_false(xfer->c_skip > 0)) { 716 KASSERT(xfer->c_skip < xfer->c_bio.bcount); 717 KASSERT((xfer->c_skip % secsize) == 0); 718 xfer->c_bio.bcount -= xfer->c_skip; 719 xfer->c_bio.blkno += xfer->c_skip / secsize; 720 } 721 722 #ifdef WD_CHAOS_MONKEY 723 /* 724 * Override blkno to be over device capacity to trigger error, 725 * but only if it's read, to avoid trashing disk contents should 726 * the command be clipped, or otherwise misinterpreted, by the 727 * driver or controller. 728 */ 729 if (BUF_ISREAD(bp) && xfer->c_retries == 0 && wd->drv_chaos_freq > 0 && 730 (++wd->drv_chaos_cnt % wd->drv_chaos_freq) == 0) { 731 device_printf(dksc->sc_dev, "%s: chaos xfer %"PRIxPTR"\n", 732 __func__, (intptr_t)xfer & PAGE_MASK); 733 xfer->c_bio.blkno = 7777777 + wd->sc_capacity; 734 xfer->c_flags |= C_CHAOS; 735 } 736 #endif 737 738 /* 739 * If we're retrying, retry in single-sector mode. This will give us 740 * the sector number of the problem, and will eventually allow the 741 * transfer to succeed. If FUA is requested, we can't actually 742 * do this, as ATA_SINGLE is usually executed as PIO transfer by drivers 743 * which support it, and that isn't compatible with NCQ/FUA. 744 */ 745 if (xfer->c_retries >= WDIORETRIES_SINGLE && 746 (bp->b_flags & B_MEDIA_FUA) == 0) 747 xfer->c_bio.flags = ATA_SINGLE; 748 else 749 xfer->c_bio.flags = 0; 750 751 /* 752 * request LBA48 transfers when supported by the controller 753 * and needed by transfer offset or size. 754 */ 755 if (wd->sc_flags & WDF_LBA48 && 756 (((xfer->c_bio.blkno + xfer->c_bio.bcount / secsize) > 757 wd->sc_capacity28) || 758 ((xfer->c_bio.bcount / secsize) > 128))) 759 xfer->c_bio.flags |= ATA_LBA48; 760 761 /* 762 * If NCQ was negotiated, always use it for the first several attempts. 763 * Since device cancels all outstanding requests on error, downgrade 764 * to non-NCQ on retry, so that the retried transfer would not cause 765 * cascade failure for the other transfers if it fails again. 766 * If FUA was requested, we can't downgrade, as that would violate 767 * the semantics - FUA would not be honored. In that case, continue 768 * retrying with NCQ. 769 */ 770 if (WD_USE_NCQ(wd) && (xfer->c_retries < WDIORETRIES_SINGLE || 771 (bp->b_flags & B_MEDIA_FUA) != 0)) { 772 xfer->c_bio.flags |= ATA_LBA48; 773 xfer->c_flags |= C_NCQ; 774 775 if (WD_USE_NCQ_PRIO(wd) && 776 BIO_GETPRIO(bp) == BPRIO_TIMECRITICAL) 777 xfer->c_bio.flags |= ATA_PRIO_HIGH; 778 } 779 780 if (wd->sc_flags & WDF_LBA) 781 xfer->c_bio.flags |= ATA_LBA; 782 if (bp->b_flags & B_READ) { 783 xfer->c_bio.flags |= ATA_READ; 784 } else { 785 /* it's a write */ 786 wd->sc_flags |= WDF_DIRTY; 787 } 788 if (bp->b_flags & B_MEDIA_FUA) { 789 /* If not using NCQ, the command WRITE DMA FUA EXT is LBA48 */ 790 KASSERT((wd->sc_flags & WDF_LBA48) != 0); 791 if ((xfer->c_flags & C_NCQ) == 0) 792 xfer->c_bio.flags |= ATA_LBA48; 793 794 xfer->c_bio.flags |= ATA_FUA; 795 } 796 797 if (xfer->c_retries == 0) 798 wd->inflight++; 799 mutex_exit(&wd->sc_lock); 800 801 /* Queue the xfer */ 802 wd->atabus->ata_bio(wd->drvp, xfer); 803 804 mutex_enter(&wd->sc_lock); 805 } 806 807 static int 808 wd_diskstart(device_t dev, struct buf *bp) 809 { 810 struct wd_softc *wd = device_private(dev); 811 #ifdef ATADEBUG 812 struct dk_softc *dksc = &wd->sc_dksc; 813 #endif 814 struct ata_xfer *xfer; 815 struct ata_channel *chp; 816 unsigned openings; 817 int ticks; 818 819 mutex_enter(&wd->sc_lock); 820 821 chp = wd->drvp->chnl_softc; 822 823 ata_channel_lock(chp); 824 openings = ata_queue_openings(chp); 825 ata_channel_unlock(chp); 826 827 openings = uimin(openings, wd->drvp->drv_openings); 828 829 if (wd->inflight >= openings) { 830 /* 831 * pretend we run out of memory when the queue is full, 832 * so that the operation is retried after a minimal 833 * delay. 834 */ 835 xfer = NULL; 836 ticks = 1; 837 } else { 838 /* 839 * If there is no available memory, retry later. This 840 * happens very rarely and only under memory pressure, 841 * so wait relatively long before retry. 842 */ 843 xfer = ata_get_xfer(chp, false); 844 ticks = hz/2; 845 } 846 847 if (xfer == NULL) { 848 ATADEBUG_PRINT(("wd_diskstart %s no xfer\n", 849 dksc->sc_xname), DEBUG_XFERS); 850 851 /* 852 * The disk queue is pushed automatically when an I/O 853 * operation finishes or another one is queued. We 854 * need this extra timeout because an ATA channel 855 * might be shared by more than one disk queue and 856 * all queues need to be restarted when another slot 857 * becomes available. 858 */ 859 if (!callout_pending(&wd->sc_restart_diskqueue)) { 860 callout_reset(&wd->sc_restart_diskqueue, ticks, 861 wdrestart, dev); 862 } 863 864 mutex_exit(&wd->sc_lock); 865 return EAGAIN; 866 } 867 868 wdstart1(wd, bp, xfer); 869 870 mutex_exit(&wd->sc_lock); 871 872 return 0; 873 } 874 875 /* 876 * Queue a drive for I/O. 877 */ 878 static void 879 wdrestart(void *x) 880 { 881 device_t self = x; 882 struct wd_softc *wd = device_private(self); 883 struct dk_softc *dksc = &wd->sc_dksc; 884 885 ATADEBUG_PRINT(("wdstart %s\n", dksc->sc_xname), 886 DEBUG_XFERS); 887 888 if (!device_is_active(dksc->sc_dev)) 889 return; 890 891 dk_start(dksc, NULL); 892 } 893 894 static void 895 wddone(device_t self, struct ata_xfer *xfer) 896 { 897 struct wd_softc *wd = device_private(self); 898 struct dk_softc *dksc = &wd->sc_dksc; 899 const char *errmsg; 900 int do_perror = 0; 901 struct buf *bp; 902 903 ATADEBUG_PRINT(("wddone %s\n", dksc->sc_xname), 904 DEBUG_XFERS); 905 906 if (__predict_false(wddoingadump)) { 907 /* just drop it to the floor */ 908 ata_free_xfer(wd->drvp->chnl_softc, xfer); 909 return; 910 } 911 912 bp = xfer->c_bio.bp; 913 KASSERT(bp != NULL); 914 915 bp->b_resid = xfer->c_bio.bcount; 916 switch (xfer->c_bio.error) { 917 case ERR_DMA: 918 errmsg = "DMA error"; 919 goto retry; 920 case ERR_DF: 921 errmsg = "device fault"; 922 goto retry; 923 case TIMEOUT: 924 errmsg = "device timeout"; 925 goto retry; 926 case REQUEUE: 927 errmsg = "requeue"; 928 goto retry2; 929 case ERR_RESET: 930 errmsg = "channel reset"; 931 goto retry2; 932 case ERROR: 933 /* Don't care about media change bits */ 934 if (xfer->c_bio.r_error != 0 && 935 (xfer->c_bio.r_error & ~(WDCE_MC | WDCE_MCR)) == 0) 936 goto noerror; 937 errmsg = "error"; 938 do_perror = 1; 939 retry: /* Just reset and retry. Can we do more ? */ 940 if ((xfer->c_flags & C_RECOVERED) == 0) { 941 int wflags = (xfer->c_flags & C_POLL) ? AT_POLL : 0; 942 ata_channel_lock(wd->drvp->chnl_softc); 943 ata_thread_run(wd->drvp->chnl_softc, wflags, 944 ATACH_TH_DRIVE_RESET, wd->drvp->drive); 945 ata_channel_unlock(wd->drvp->chnl_softc); 946 } 947 retry2: 948 mutex_enter(&wd->sc_lock); 949 950 diskerr(bp, "wd", errmsg, LOG_PRINTF, 951 xfer->c_bio.blkdone, dksc->sc_dkdev.dk_label); 952 if (xfer->c_retries < WDIORETRIES) 953 printf(", xfer %"PRIxPTR", retry %d", 954 (intptr_t)xfer & PAGE_MASK, 955 xfer->c_retries); 956 printf("\n"); 957 if (do_perror) 958 wdperror(wd, xfer); 959 960 if (xfer->c_retries < WDIORETRIES) { 961 xfer->c_retries++; 962 963 /* Rerun ASAP if just requeued */ 964 if (xfer->c_bio.error == REQUEUE) { 965 SLIST_INSERT_HEAD(&wd->sc_requeue_list, xfer, 966 c_retrychain); 967 callout_reset(&wd->sc_requeue_callout, 968 1, wdbiorequeue, wd); 969 } else { 970 SLIST_INSERT_HEAD(&wd->sc_retry_list, xfer, 971 c_retrychain); 972 callout_reset(&wd->sc_retry_callout, 973 RECOVERYTIME, wdbioretry, wd); 974 } 975 976 mutex_exit(&wd->sc_lock); 977 return; 978 } 979 980 mutex_exit(&wd->sc_lock); 981 982 #ifdef WD_SOFTBADSECT 983 /* 984 * Not all errors indicate a failed block but those that do, 985 * put the block on the bad-block list for the device. Only 986 * do this for reads because the drive should do it for writes, 987 * itself, according to Manuel. 988 */ 989 if ((bp->b_flags & B_READ) && 990 ((wd->drvp->ata_vers >= 4 && xfer->c_bio.r_error & 64) || 991 (wd->drvp->ata_vers < 4 && xfer->c_bio.r_error & 192))) { 992 struct disk_badsectors *dbs; 993 994 dbs = kmem_zalloc(sizeof *dbs, KM_NOSLEEP); 995 if (dbs == NULL) { 996 device_printf(dksc->sc_dev, 997 "failed to add bad block to list\n"); 998 goto out; 999 } 1000 1001 dbs->dbs_min = bp->b_rawblkno; 1002 dbs->dbs_max = dbs->dbs_min + 1003 (bp->b_bcount /wd->sc_blksize) - 1; 1004 microtime(&dbs->dbs_failedat); 1005 1006 mutex_enter(&wd->sc_lock); 1007 SLIST_INSERT_HEAD(&wd->sc_bslist, dbs, dbs_next); 1008 wd->sc_bscount++; 1009 mutex_exit(&wd->sc_lock); 1010 } 1011 out: 1012 #endif 1013 bp->b_error = EIO; 1014 break; 1015 case NOERROR: 1016 #ifdef WD_CHAOS_MONKEY 1017 /* 1018 * For example Parallels AHCI emulation doesn't actually 1019 * return error for the invalid I/O, so just re-run 1020 * the request and do not panic. 1021 */ 1022 if (__predict_false(xfer->c_flags & C_CHAOS)) { 1023 xfer->c_bio.error = REQUEUE; 1024 errmsg = "chaos noerror"; 1025 goto retry2; 1026 } 1027 #endif 1028 1029 noerror: if ((xfer->c_bio.flags & ATA_CORR) || xfer->c_retries > 0) 1030 device_printf(dksc->sc_dev, 1031 "soft error (corrected) xfer %"PRIxPTR"\n", 1032 (intptr_t)xfer & PAGE_MASK); 1033 break; 1034 case ERR_NODEV: 1035 bp->b_error = EIO; 1036 break; 1037 } 1038 if (__predict_false(bp->b_error != 0) && bp->b_resid == 0) { 1039 /* 1040 * the disk or controller sometimes report a complete 1041 * xfer, when there has been an error. This is wrong, 1042 * assume nothing got transferred in this case 1043 */ 1044 bp->b_resid = bp->b_bcount; 1045 } 1046 1047 ata_free_xfer(wd->drvp->chnl_softc, xfer); 1048 1049 mutex_enter(&wd->sc_lock); 1050 wd->inflight--; 1051 mutex_exit(&wd->sc_lock); 1052 dk_done(dksc, bp); 1053 dk_start(dksc, NULL); 1054 } 1055 1056 static void 1057 wdbioretry(void *v) 1058 { 1059 struct wd_softc *wd = v; 1060 struct ata_xfer *xfer; 1061 1062 ATADEBUG_PRINT(("%s %s\n", __func__, wd->sc_dksc.sc_xname), 1063 DEBUG_XFERS); 1064 1065 mutex_enter(&wd->sc_lock); 1066 while ((xfer = SLIST_FIRST(&wd->sc_retry_list))) { 1067 SLIST_REMOVE_HEAD(&wd->sc_retry_list, c_retrychain); 1068 wdstart1(wd, xfer->c_bio.bp, xfer); 1069 } 1070 mutex_exit(&wd->sc_lock); 1071 } 1072 1073 static void 1074 wdbiorequeue(void *v) 1075 { 1076 struct wd_softc *wd = v; 1077 struct ata_xfer *xfer; 1078 1079 ATADEBUG_PRINT(("%s %s\n", __func__, wd->sc_dksc.sc_xname), 1080 DEBUG_XFERS); 1081 1082 mutex_enter(&wd->sc_lock); 1083 while ((xfer = SLIST_FIRST(&wd->sc_requeue_list))) { 1084 SLIST_REMOVE_HEAD(&wd->sc_requeue_list, c_retrychain); 1085 wdstart1(wd, xfer->c_bio.bp, xfer); 1086 } 1087 mutex_exit(&wd->sc_lock); 1088 } 1089 1090 static void 1091 wdminphys(struct buf *bp) 1092 { 1093 const struct wd_softc * const wd = 1094 device_lookup_private(&wd_cd, WDUNIT(bp->b_dev)); 1095 int maxsectors; 1096 1097 /* 1098 * The limit is actually 65536 for LBA48 and 256 for non-LBA48, 1099 * but that requires to set the count for the ATA command 1100 * to 0, which is somewhat error prone, so better stay safe. 1101 */ 1102 if (wd->sc_flags & WDF_LBA48) 1103 maxsectors = 65535; 1104 else 1105 maxsectors = 128; 1106 1107 if (bp->b_bcount > (wd->sc_blksize * maxsectors)) 1108 bp->b_bcount = (wd->sc_blksize * maxsectors); 1109 1110 minphys(bp); 1111 } 1112 1113 static void 1114 wd_iosize(device_t dev, int *count) 1115 { 1116 struct buf B; 1117 int bmaj; 1118 1119 bmaj = bdevsw_lookup_major(&wd_bdevsw); 1120 B.b_dev = MAKEWDDEV(bmaj,device_unit(dev),RAW_PART); 1121 B.b_bcount = *count; 1122 1123 wdminphys(&B); 1124 1125 *count = B.b_bcount; 1126 } 1127 1128 static int 1129 wdread(dev_t dev, struct uio *uio, int flags) 1130 { 1131 1132 ATADEBUG_PRINT(("wdread\n"), DEBUG_XFERS); 1133 return (physio(wdstrategy, NULL, dev, B_READ, wdminphys, uio)); 1134 } 1135 1136 static int 1137 wdwrite(dev_t dev, struct uio *uio, int flags) 1138 { 1139 1140 ATADEBUG_PRINT(("wdwrite\n"), DEBUG_XFERS); 1141 return (physio(wdstrategy, NULL, dev, B_WRITE, wdminphys, uio)); 1142 } 1143 1144 static int 1145 wdopen(dev_t dev, int flag, int fmt, struct lwp *l) 1146 { 1147 struct wd_softc *wd; 1148 struct dk_softc *dksc; 1149 int unit, part, error; 1150 1151 ATADEBUG_PRINT(("wdopen\n"), DEBUG_FUNCS); 1152 unit = WDUNIT(dev); 1153 wd = device_lookup_private(&wd_cd, unit); 1154 if (wd == NULL) 1155 return (ENXIO); 1156 dksc = &wd->sc_dksc; 1157 1158 if (! device_is_active(dksc->sc_dev)) 1159 return (ENODEV); 1160 1161 part = WDPART(dev); 1162 1163 if (wd->sc_capacity == 0) 1164 return (ENODEV); 1165 1166 /* 1167 * If any partition is open, but the disk has been invalidated, 1168 * disallow further opens. 1169 */ 1170 if ((wd->sc_flags & (WDF_OPEN | WDF_LOADED)) == WDF_OPEN) { 1171 if (part != RAW_PART || fmt != S_IFCHR) 1172 return EIO; 1173 } 1174 1175 error = dk_open(dksc, dev, flag, fmt, l); 1176 1177 return error; 1178 } 1179 1180 /* 1181 * Serialized by caller 1182 */ 1183 static int 1184 wd_firstopen(device_t self, dev_t dev, int flag, int fmt) 1185 { 1186 struct wd_softc *wd = device_private(self); 1187 struct dk_softc *dksc = &wd->sc_dksc; 1188 int error; 1189 1190 error = wd->atabus->ata_addref(wd->drvp); 1191 if (error) 1192 return error; 1193 1194 if ((wd->sc_flags & WDF_LOADED) == 0) { 1195 int param_error; 1196 1197 /* Load the physical device parameters. */ 1198 param_error = wd_get_params(wd, &wd->sc_params); 1199 if (param_error != 0) { 1200 aprint_error_dev(dksc->sc_dev, "IDENTIFY failed\n"); 1201 error = EIO; 1202 goto bad; 1203 } 1204 wd_set_geometry(wd); 1205 wd->sc_flags |= WDF_LOADED; 1206 } 1207 1208 wd->sc_flags |= WDF_OPEN; 1209 return 0; 1210 1211 bad: 1212 wd->atabus->ata_delref(wd->drvp); 1213 return error; 1214 } 1215 1216 /* 1217 * Caller must hold wd->sc_dk.dk_openlock. 1218 */ 1219 static int 1220 wd_lastclose(device_t self) 1221 { 1222 struct wd_softc *wd = device_private(self); 1223 1224 KASSERTMSG(bufq_peek(wd->sc_dksc.sc_bufq) == NULL, "bufq not empty"); 1225 1226 if (wd->sc_flags & WDF_DIRTY) 1227 wd_flushcache(wd, AT_WAIT); 1228 1229 wd->atabus->ata_delref(wd->drvp); 1230 wd->sc_flags &= ~WDF_OPEN; 1231 1232 return 0; 1233 } 1234 1235 static int 1236 wdclose(dev_t dev, int flag, int fmt, struct lwp *l) 1237 { 1238 struct wd_softc *wd; 1239 struct dk_softc *dksc; 1240 int unit; 1241 1242 unit = WDUNIT(dev); 1243 wd = device_lookup_private(&wd_cd, unit); 1244 dksc = &wd->sc_dksc; 1245 1246 return dk_close(dksc, dev, flag, fmt, l); 1247 } 1248 1249 void 1250 wdperror(const struct wd_softc *wd, struct ata_xfer *xfer) 1251 { 1252 static const char *const errstr0_3[] = {"address mark not found", 1253 "track 0 not found", "aborted command", "media change requested", 1254 "id not found", "media changed", "uncorrectable data error", 1255 "bad block detected"}; 1256 static const char *const errstr4_5[] = { 1257 "obsolete (address mark not found)", 1258 "no media/write protected", "aborted command", 1259 "media change requested", "id not found", "media changed", 1260 "uncorrectable data error", "interface CRC error"}; 1261 const char *const *errstr; 1262 int i; 1263 const char *sep = ""; 1264 1265 const struct dk_softc *dksc = &wd->sc_dksc; 1266 const char *devname = dksc->sc_xname; 1267 struct ata_drive_datas *drvp = wd->drvp; 1268 int errno = xfer->c_bio.r_error; 1269 1270 if (drvp->ata_vers >= 4) 1271 errstr = errstr4_5; 1272 else 1273 errstr = errstr0_3; 1274 1275 printf("%s: (", devname); 1276 1277 if (errno == 0) 1278 printf("error not notified"); 1279 1280 for (i = 0; i < 8; i++) { 1281 if (errno & (1 << i)) { 1282 printf("%s%s", sep, errstr[i]); 1283 sep = ", "; 1284 } 1285 } 1286 printf(")\n"); 1287 } 1288 1289 int 1290 wdioctl(dev_t dev, u_long cmd, void *addr, int flag, struct lwp *l) 1291 { 1292 struct wd_softc *wd = 1293 device_lookup_private(&wd_cd, WDUNIT(dev)); 1294 struct dk_softc *dksc = &wd->sc_dksc; 1295 1296 ATADEBUG_PRINT(("wdioctl\n"), DEBUG_FUNCS); 1297 1298 if ((wd->sc_flags & WDF_LOADED) == 0) 1299 return EIO; 1300 1301 switch (cmd) { 1302 #ifdef HAS_BAD144_HANDLING 1303 case DIOCSBAD: 1304 if ((flag & FWRITE) == 0) 1305 return EBADF; 1306 dksc->sc_dkdev.dk_cpulabel->bad = *(struct dkbad *)addr; 1307 dksc->sc_dkdev.dk_label->d_flags |= D_BADSECT; 1308 bad144intern(wd); 1309 return 0; 1310 #endif 1311 #ifdef WD_SOFTBADSECT 1312 case DIOCBSLIST: { 1313 uint32_t count, missing, skip; 1314 struct disk_badsecinfo dbsi; 1315 struct disk_badsectors *dbs, dbsbuf; 1316 size_t available; 1317 uint8_t *laddr; 1318 int error; 1319 1320 dbsi = *(struct disk_badsecinfo *)addr; 1321 missing = wd->sc_bscount; 1322 count = 0; 1323 available = dbsi.dbsi_bufsize; 1324 skip = dbsi.dbsi_skip; 1325 laddr = (uint8_t *)dbsi.dbsi_buffer; 1326 1327 /* 1328 * We start this loop with the expectation that all of the 1329 * entries will be missed and decrement this counter each 1330 * time we either skip over one (already copied out) or 1331 * we actually copy it back to user space. The structs 1332 * holding the bad sector information are copied directly 1333 * back to user space whilst the summary is returned via 1334 * the struct passed in via the ioctl. 1335 */ 1336 error = 0; 1337 mutex_enter(&wd->sc_lock); 1338 wd->sc_bslist_inuse++; 1339 SLIST_FOREACH(dbs, &wd->sc_bslist, dbs_next) { 1340 if (skip > 0) { 1341 missing--; 1342 skip--; 1343 continue; 1344 } 1345 if (available < sizeof(*dbs)) 1346 break; 1347 available -= sizeof(*dbs); 1348 memset(&dbsbuf, 0, sizeof(dbsbuf)); 1349 dbsbuf.dbs_min = dbs->dbs_min; 1350 dbsbuf.dbs_max = dbs->dbs_max; 1351 dbsbuf.dbs_failedat = dbs->dbs_failedat; 1352 mutex_exit(&wd->sc_lock); 1353 error = copyout(&dbsbuf, laddr, sizeof(dbsbuf)); 1354 mutex_enter(&wd->sc_lock); 1355 if (error) 1356 break; 1357 laddr += sizeof(*dbs); 1358 missing--; 1359 count++; 1360 } 1361 if (--wd->sc_bslist_inuse == 0) 1362 cv_broadcast(&wd->sc_bslist_cv); 1363 mutex_exit(&wd->sc_lock); 1364 dbsi.dbsi_left = missing; 1365 dbsi.dbsi_copied = count; 1366 *(struct disk_badsecinfo *)addr = dbsi; 1367 1368 /* 1369 * If we copied anything out, ignore error and return 1370 * success -- can't back it out. 1371 */ 1372 return count ? 0 : error; 1373 } 1374 1375 case DIOCBSFLUSH: { 1376 int error; 1377 1378 /* Clean out the bad sector list */ 1379 mutex_enter(&wd->sc_lock); 1380 while (wd->sc_bslist_inuse) { 1381 error = cv_wait_sig(&wd->sc_bslist_cv, &wd->sc_lock); 1382 if (error) { 1383 mutex_exit(&wd->sc_lock); 1384 return error; 1385 } 1386 } 1387 while (!SLIST_EMPTY(&wd->sc_bslist)) { 1388 struct disk_badsectors *dbs = 1389 SLIST_FIRST(&wd->sc_bslist); 1390 SLIST_REMOVE_HEAD(&wd->sc_bslist, dbs_next); 1391 mutex_exit(&wd->sc_lock); 1392 kmem_free(dbs, sizeof(*dbs)); 1393 mutex_enter(&wd->sc_lock); 1394 } 1395 mutex_exit(&wd->sc_lock); 1396 wd->sc_bscount = 0; 1397 return 0; 1398 } 1399 #endif 1400 1401 #ifdef notyet 1402 case DIOCWFORMAT: 1403 if ((flag & FWRITE) == 0) 1404 return EBADF; 1405 { 1406 register struct format_op *fop; 1407 struct iovec aiov; 1408 struct uio auio; 1409 int error1; 1410 1411 fop = (struct format_op *)addr; 1412 aiov.iov_base = fop->df_buf; 1413 aiov.iov_len = fop->df_count; 1414 auio.uio_iov = &aiov; 1415 auio.uio_iovcnt = 1; 1416 auio.uio_resid = fop->df_count; 1417 auio.uio_offset = 1418 fop->df_startblk * wd->sc_dk.dk_label->d_secsize; 1419 auio.uio_vmspace = l->l_proc->p_vmspace; 1420 error1 = physio(wdformat, NULL, dev, B_WRITE, wdminphys, 1421 &auio); 1422 fop->df_count -= auio.uio_resid; 1423 fop->df_reg[0] = wdc->sc_status; 1424 fop->df_reg[1] = wdc->sc_error; 1425 return error1; 1426 } 1427 #endif 1428 case DIOCGCACHE: 1429 return wd_getcache(wd, (int *)addr); 1430 1431 case DIOCSCACHE: 1432 return wd_setcache(wd, *(int *)addr); 1433 1434 case DIOCCACHESYNC: 1435 return wd_flushcache(wd, AT_WAIT); 1436 1437 case ATAIOCCOMMAND: 1438 /* 1439 * Make sure this command is (relatively) safe first 1440 */ 1441 if ((((atareq_t *) addr)->flags & ATACMD_READ) == 0 && 1442 (flag & FWRITE) == 0) 1443 return (EBADF); 1444 { 1445 struct wd_ioctl *wi; 1446 atareq_t *atareq = (atareq_t *) addr; 1447 int error1; 1448 1449 wi = wi_get(wd); 1450 wi->wi_atareq = *atareq; 1451 1452 if (atareq->datalen && atareq->flags & 1453 (ATACMD_READ | ATACMD_WRITE)) { 1454 void *tbuf; 1455 if (atareq->datalen < DEV_BSIZE 1456 && atareq->command == WDCC_IDENTIFY) { 1457 tbuf = kmem_zalloc(DEV_BSIZE, KM_SLEEP); 1458 wi->wi_iov.iov_base = tbuf; 1459 wi->wi_iov.iov_len = DEV_BSIZE; 1460 UIO_SETUP_SYSSPACE(&wi->wi_uio); 1461 } else { 1462 tbuf = NULL; 1463 wi->wi_iov.iov_base = atareq->databuf; 1464 wi->wi_iov.iov_len = atareq->datalen; 1465 wi->wi_uio.uio_vmspace = l->l_proc->p_vmspace; 1466 } 1467 wi->wi_uio.uio_iov = &wi->wi_iov; 1468 wi->wi_uio.uio_iovcnt = 1; 1469 wi->wi_uio.uio_resid = atareq->datalen; 1470 wi->wi_uio.uio_offset = 0; 1471 wi->wi_uio.uio_rw = 1472 (atareq->flags & ATACMD_READ) ? B_READ : B_WRITE; 1473 error1 = physio(wdioctlstrategy, &wi->wi_bp, dev, 1474 (atareq->flags & ATACMD_READ) ? B_READ : B_WRITE, 1475 wdminphys, &wi->wi_uio); 1476 if (tbuf != NULL && error1 == 0) { 1477 error1 = copyout(tbuf, atareq->databuf, 1478 atareq->datalen); 1479 kmem_free(tbuf, DEV_BSIZE); 1480 } 1481 } else { 1482 /* No need to call physio if we don't have any 1483 user data */ 1484 wi->wi_bp.b_flags = 0; 1485 wi->wi_bp.b_data = 0; 1486 wi->wi_bp.b_bcount = 0; 1487 wi->wi_bp.b_dev = dev; 1488 wi->wi_bp.b_proc = l->l_proc; 1489 wdioctlstrategy(&wi->wi_bp); 1490 error1 = wi->wi_bp.b_error; 1491 } 1492 *atareq = wi->wi_atareq; 1493 wi_free(wi); 1494 return(error1); 1495 } 1496 1497 case DIOCGSECTORALIGN: { 1498 struct disk_sectoralign *dsa = addr; 1499 int part = WDPART(dev); 1500 1501 *dsa = wd->sc_sectoralign; 1502 if (part != RAW_PART) { 1503 struct disklabel *lp = dksc->sc_dkdev.dk_label; 1504 daddr_t offset = lp->d_partitions[part].p_offset; 1505 uint32_t r = offset % dsa->dsa_alignment; 1506 1507 if (r < dsa->dsa_firstaligned) 1508 dsa->dsa_firstaligned = dsa->dsa_firstaligned 1509 - r; 1510 else 1511 dsa->dsa_firstaligned = (dsa->dsa_firstaligned 1512 + dsa->dsa_alignment) - r; 1513 } 1514 1515 return 0; 1516 } 1517 1518 default: 1519 return dk_ioctl(dksc, dev, cmd, addr, flag, l); 1520 } 1521 1522 #ifdef DIAGNOSTIC 1523 panic("wdioctl: impossible"); 1524 #endif 1525 } 1526 1527 static int 1528 wd_discard(device_t dev, off_t pos, off_t len) 1529 { 1530 struct wd_softc *wd = device_private(dev); 1531 daddr_t bno; 1532 long size, done; 1533 long maxatonce, amount; 1534 int result; 1535 1536 if (!(wd->sc_params.atap_ata_major & WDC_VER_ATA7) 1537 || !(wd->sc_params.support_dsm & ATA_SUPPORT_DSM_TRIM)) { 1538 /* not supported; ignore request */ 1539 ATADEBUG_PRINT(("wddiscard (unsupported)\n"), DEBUG_FUNCS); 1540 return 0; 1541 } 1542 maxatonce = 0xffff; /*wd->sc_params.max_dsm_blocks*/ 1543 1544 ATADEBUG_PRINT(("wddiscard\n"), DEBUG_FUNCS); 1545 1546 if ((wd->sc_flags & WDF_LOADED) == 0) 1547 return EIO; 1548 1549 /* round the start up and the end down */ 1550 bno = (pos + wd->sc_blksize - 1) / wd->sc_blksize; 1551 size = ((pos + len) / wd->sc_blksize) - bno; 1552 1553 done = 0; 1554 while (done < size) { 1555 amount = size - done; 1556 if (amount > maxatonce) { 1557 amount = maxatonce; 1558 } 1559 result = wd_trim(wd, bno + done, amount); 1560 if (result) { 1561 return result; 1562 } 1563 done += amount; 1564 } 1565 return 0; 1566 } 1567 1568 static int 1569 wddiscard(dev_t dev, off_t pos, off_t len) 1570 { 1571 struct wd_softc *wd; 1572 struct dk_softc *dksc; 1573 int unit; 1574 1575 unit = WDUNIT(dev); 1576 wd = device_lookup_private(&wd_cd, unit); 1577 dksc = &wd->sc_dksc; 1578 1579 return dk_discard(dksc, dev, pos, len); 1580 } 1581 1582 #ifdef B_FORMAT 1583 int 1584 wdformat(struct buf *bp) 1585 { 1586 1587 bp->b_flags |= B_FORMAT; 1588 return wdstrategy(bp); 1589 } 1590 #endif 1591 1592 int 1593 wdsize(dev_t dev) 1594 { 1595 struct wd_softc *wd; 1596 struct dk_softc *dksc; 1597 int unit; 1598 1599 ATADEBUG_PRINT(("wdsize\n"), DEBUG_FUNCS); 1600 1601 unit = WDUNIT(dev); 1602 wd = device_lookup_private(&wd_cd, unit); 1603 if (wd == NULL) 1604 return (-1); 1605 dksc = &wd->sc_dksc; 1606 1607 if (!device_is_active(dksc->sc_dev)) 1608 return (-1); 1609 1610 return dk_size(dksc, dev); 1611 } 1612 1613 /* 1614 * Dump core after a system crash. 1615 */ 1616 static int 1617 wddump(dev_t dev, daddr_t blkno, void *va, size_t size) 1618 { 1619 struct wd_softc *wd; 1620 struct dk_softc *dksc; 1621 int unit; 1622 1623 /* Check if recursive dump; if so, punt. */ 1624 if (wddoingadump) 1625 return EFAULT; 1626 wddoingadump = 1; 1627 1628 unit = WDUNIT(dev); 1629 wd = device_lookup_private(&wd_cd, unit); 1630 if (wd == NULL) 1631 return (ENXIO); 1632 dksc = &wd->sc_dksc; 1633 1634 return dk_dump(dksc, dev, blkno, va, size, 0); 1635 } 1636 1637 static int 1638 wd_dumpblocks(device_t dev, void *va, daddr_t blkno, int nblk) 1639 { 1640 struct wd_softc *wd = device_private(dev); 1641 struct dk_softc *dksc = &wd->sc_dksc; 1642 struct disk_geom *dg = &dksc->sc_dkdev.dk_geom; 1643 struct ata_xfer *xfer = &wd->dump_xfer; 1644 int err; 1645 1646 /* Recalibrate, if first dump transfer. */ 1647 if (wddumprecalibrated == 0) { 1648 wddumprecalibrated = 1; 1649 ata_channel_lock(wd->drvp->chnl_softc); 1650 /* This will directly execute the reset due to AT_POLL */ 1651 ata_thread_run(wd->drvp->chnl_softc, AT_POLL, 1652 ATACH_TH_DRIVE_RESET, wd->drvp->drive); 1653 1654 wd->drvp->state = RESET; 1655 ata_channel_unlock(wd->drvp->chnl_softc); 1656 } 1657 1658 memset(xfer, 0, sizeof(*xfer)); 1659 xfer->c_flags |= C_PRIVATE_ALLOC | C_SKIP_QUEUE; 1660 1661 xfer->c_bio.blkno = blkno; 1662 xfer->c_bio.flags = ATA_POLL; 1663 if (wd->sc_flags & WDF_LBA48 && 1664 (xfer->c_bio.blkno + nblk) > wd->sc_capacity28) 1665 xfer->c_bio.flags |= ATA_LBA48; 1666 if (wd->sc_flags & WDF_LBA) 1667 xfer->c_bio.flags |= ATA_LBA; 1668 xfer->c_bio.bcount = nblk * dg->dg_secsize; 1669 xfer->c_bio.databuf = va; 1670 #ifndef WD_DUMP_NOT_TRUSTED 1671 /* This will poll until the bio is complete */ 1672 wd->atabus->ata_bio(wd->drvp, xfer); 1673 1674 switch(err = xfer->c_bio.error) { 1675 case TIMEOUT: 1676 printf("wddump: device timed out"); 1677 err = EIO; 1678 break; 1679 case ERR_DF: 1680 printf("wddump: drive fault"); 1681 err = EIO; 1682 break; 1683 case ERR_DMA: 1684 printf("wddump: DMA error"); 1685 err = EIO; 1686 break; 1687 case ERROR: 1688 printf("wddump: "); 1689 wdperror(wd, xfer); 1690 err = EIO; 1691 break; 1692 case NOERROR: 1693 err = 0; 1694 break; 1695 default: 1696 panic("wddump: unknown error type %x", err); 1697 } 1698 1699 if (err != 0) { 1700 printf("\n"); 1701 return err; 1702 } 1703 #else /* WD_DUMP_NOT_TRUSTED */ 1704 /* Let's just talk about this first... */ 1705 printf("wd%d: dump addr 0x%x, cylin %d, head %d, sector %d\n", 1706 unit, va, cylin, head, sector); 1707 delay(500 * 1000); /* half a second */ 1708 #endif 1709 1710 wddoingadump = 0; 1711 return 0; 1712 } 1713 1714 #ifdef HAS_BAD144_HANDLING 1715 /* 1716 * Internalize the bad sector table. 1717 */ 1718 void 1719 bad144intern(struct wd_softc *wd) 1720 { 1721 struct dk_softc *dksc = &wd->sc_dksc; 1722 struct dkbad *bt = &dksc->sc_dkdev.dk_cpulabel->bad; 1723 struct disklabel *lp = dksc->sc_dkdev.dk_label; 1724 int i = 0; 1725 1726 ATADEBUG_PRINT(("bad144intern\n"), DEBUG_XFERS); 1727 1728 for (; i < NBT_BAD; i++) { 1729 if (bt->bt_bad[i].bt_cyl == 0xffff) 1730 break; 1731 wd->drvp->badsect[i] = 1732 bt->bt_bad[i].bt_cyl * lp->d_secpercyl + 1733 (bt->bt_bad[i].bt_trksec >> 8) * lp->d_nsectors + 1734 (bt->bt_bad[i].bt_trksec & 0xff); 1735 } 1736 for (; i < NBT_BAD+1; i++) 1737 wd->drvp->badsect[i] = -1; 1738 } 1739 #endif 1740 1741 static void 1742 wd_set_geometry(struct wd_softc *wd) 1743 { 1744 struct dk_softc *dksc = &wd->sc_dksc; 1745 struct disk_geom *dg = &dksc->sc_dkdev.dk_geom; 1746 1747 memset(dg, 0, sizeof(*dg)); 1748 1749 dg->dg_secperunit = wd->sc_capacity; 1750 dg->dg_secsize = wd->sc_blksize; 1751 dg->dg_nsectors = wd->sc_params.atap_sectors; 1752 dg->dg_ntracks = wd->sc_params.atap_heads; 1753 if ((wd->sc_flags & WDF_LBA) == 0) 1754 dg->dg_ncylinders = wd->sc_params.atap_cylinders; 1755 1756 disk_set_info(dksc->sc_dev, &dksc->sc_dkdev, wd->sc_typename); 1757 } 1758 1759 int 1760 wd_get_params(struct wd_softc *wd, struct ataparams *params) 1761 { 1762 int retry = 0; 1763 struct ata_channel *chp = wd->drvp->chnl_softc; 1764 const int flags = AT_WAIT; 1765 1766 again: 1767 switch (wd->atabus->ata_get_params(wd->drvp, flags, params)) { 1768 case CMD_AGAIN: 1769 return 1; 1770 case CMD_ERR: 1771 if (retry == 0) { 1772 retry++; 1773 ata_channel_lock(chp); 1774 (*wd->atabus->ata_reset_drive)(wd->drvp, flags, NULL); 1775 ata_channel_unlock(chp); 1776 goto again; 1777 } 1778 1779 if (wd->drvp->drive_type != ATA_DRIVET_OLD) 1780 return 1; 1781 /* 1782 * We `know' there's a drive here; just assume it's old. 1783 * This geometry is only used to read the MBR and print a 1784 * (false) attach message. 1785 */ 1786 strncpy(params->atap_model, "ST506", 1787 sizeof params->atap_model); 1788 params->atap_config = ATA_CFG_FIXED; 1789 params->atap_cylinders = 1024; 1790 params->atap_heads = 8; 1791 params->atap_sectors = 17; 1792 params->atap_multi = 1; 1793 params->atap_capabilities1 = params->atap_capabilities2 = 0; 1794 wd->drvp->ata_vers = -1; /* Mark it as pre-ATA */ 1795 /* FALLTHROUGH */ 1796 case CMD_OK: 1797 return 0; 1798 default: 1799 panic("wd_get_params: bad return code from ata_get_params"); 1800 /* NOTREACHED */ 1801 } 1802 } 1803 1804 int 1805 wd_getcache(struct wd_softc *wd, int *bitsp) 1806 { 1807 struct ataparams params; 1808 1809 if (wd_get_params(wd, ¶ms) != 0) 1810 return EIO; 1811 if (params.atap_cmd_set1 == 0x0000 || 1812 params.atap_cmd_set1 == 0xffff || 1813 (params.atap_cmd_set1 & WDC_CMD1_CACHE) == 0) { 1814 *bitsp = 0; 1815 return 0; 1816 } 1817 *bitsp = DKCACHE_WCHANGE | DKCACHE_READ; 1818 if (params.atap_cmd1_en & WDC_CMD1_CACHE) 1819 *bitsp |= DKCACHE_WRITE; 1820 1821 if (WD_USE_NCQ(wd) || (wd->drvp->drive_flags & ATA_DRIVE_WFUA)) 1822 *bitsp |= DKCACHE_FUA; 1823 1824 return 0; 1825 } 1826 1827 1828 static int 1829 wd_check_error(const struct dk_softc *dksc, const struct ata_xfer *xfer, 1830 const char *func) 1831 { 1832 static const char at_errbits[] = "\20\10ERROR\11TIMEOU\12DF"; 1833 1834 int flags = xfer->c_ata_c.flags; 1835 1836 if ((flags & AT_ERROR) != 0 && xfer->c_ata_c.r_error == WDCE_ABRT) { 1837 /* command not supported */ 1838 aprint_debug_dev(dksc->sc_dev, "%s: not supported\n", func); 1839 return ENODEV; 1840 } 1841 if (flags & (AT_ERROR | AT_TIMEOU | AT_DF)) { 1842 char sbuf[sizeof(at_errbits) + 64]; 1843 snprintb(sbuf, sizeof(sbuf), at_errbits, flags); 1844 device_printf(dksc->sc_dev, "%s: status=%s\n", func, sbuf); 1845 return EIO; 1846 } 1847 return 0; 1848 } 1849 1850 int 1851 wd_setcache(struct wd_softc *wd, int bits) 1852 { 1853 struct dk_softc *dksc = &wd->sc_dksc; 1854 struct ataparams params; 1855 struct ata_xfer *xfer; 1856 int error; 1857 1858 if (wd_get_params(wd, ¶ms) != 0) 1859 return EIO; 1860 1861 if (params.atap_cmd_set1 == 0x0000 || 1862 params.atap_cmd_set1 == 0xffff || 1863 (params.atap_cmd_set1 & WDC_CMD1_CACHE) == 0) 1864 return EOPNOTSUPP; 1865 1866 if ((bits & DKCACHE_READ) == 0 || 1867 (bits & DKCACHE_SAVE) != 0) 1868 return EOPNOTSUPP; 1869 1870 xfer = ata_get_xfer(wd->drvp->chnl_softc, true); 1871 1872 xfer->c_ata_c.r_command = SET_FEATURES; 1873 xfer->c_ata_c.r_st_bmask = 0; 1874 xfer->c_ata_c.r_st_pmask = 0; 1875 xfer->c_ata_c.timeout = 30000; /* 30s timeout */ 1876 xfer->c_ata_c.flags = AT_WAIT; 1877 if (bits & DKCACHE_WRITE) 1878 xfer->c_ata_c.r_features = WDSF_WRITE_CACHE_EN; 1879 else 1880 xfer->c_ata_c.r_features = WDSF_WRITE_CACHE_DS; 1881 1882 wd->atabus->ata_exec_command(wd->drvp, xfer); 1883 ata_wait_cmd(wd->drvp->chnl_softc, xfer); 1884 1885 error = wd_check_error(dksc, xfer, __func__); 1886 ata_free_xfer(wd->drvp->chnl_softc, xfer); 1887 return error; 1888 } 1889 1890 static int 1891 wd_standby(struct wd_softc *wd, int flags) 1892 { 1893 struct dk_softc *dksc = &wd->sc_dksc; 1894 struct ata_xfer *xfer; 1895 int error; 1896 1897 aprint_debug_dev(dksc->sc_dev, "standby immediate\n"); 1898 xfer = ata_get_xfer(wd->drvp->chnl_softc, true); 1899 1900 xfer->c_ata_c.r_command = WDCC_STANDBY_IMMED; 1901 xfer->c_ata_c.r_st_bmask = WDCS_DRDY; 1902 xfer->c_ata_c.r_st_pmask = WDCS_DRDY; 1903 xfer->c_ata_c.flags = flags; 1904 xfer->c_ata_c.timeout = 30000; /* 30s timeout */ 1905 1906 wd->atabus->ata_exec_command(wd->drvp, xfer); 1907 ata_wait_cmd(wd->drvp->chnl_softc, xfer); 1908 1909 error = wd_check_error(dksc, xfer, __func__); 1910 ata_free_xfer(wd->drvp->chnl_softc, xfer); 1911 return error; 1912 } 1913 1914 int 1915 wd_flushcache(struct wd_softc *wd, int flags) 1916 { 1917 struct dk_softc *dksc = &wd->sc_dksc; 1918 struct ata_xfer *xfer; 1919 int error; 1920 1921 /* 1922 * WDCC_FLUSHCACHE is here since ATA-4, but some drives report 1923 * only ATA-2 and still support it. 1924 */ 1925 if (wd->drvp->ata_vers < 4 && 1926 ((wd->sc_params.atap_cmd_set2 & WDC_CMD2_FC) == 0 || 1927 wd->sc_params.atap_cmd_set2 == 0xffff)) 1928 return ENODEV; 1929 1930 xfer = ata_get_xfer(wd->drvp->chnl_softc, true); 1931 1932 if ((wd->sc_params.atap_cmd2_en & ATA_CMD2_LBA48) != 0 && 1933 (wd->sc_params.atap_cmd2_en & ATA_CMD2_FCE) != 0) { 1934 xfer->c_ata_c.r_command = WDCC_FLUSHCACHE_EXT; 1935 flags |= AT_LBA48; 1936 } else 1937 xfer->c_ata_c.r_command = WDCC_FLUSHCACHE; 1938 xfer->c_ata_c.r_st_bmask = WDCS_DRDY; 1939 xfer->c_ata_c.r_st_pmask = WDCS_DRDY; 1940 xfer->c_ata_c.flags = flags | AT_READREG; 1941 xfer->c_ata_c.timeout = 300000; /* 5m timeout */ 1942 1943 wd->atabus->ata_exec_command(wd->drvp, xfer); 1944 ata_wait_cmd(wd->drvp->chnl_softc, xfer); 1945 1946 error = wd_check_error(dksc, xfer, __func__); 1947 wd->sc_flags &= ~WDF_DIRTY; 1948 ata_free_xfer(wd->drvp->chnl_softc, xfer); 1949 return error; 1950 } 1951 1952 /* 1953 * Execute TRIM command, assumes sleep context. 1954 */ 1955 static int 1956 wd_trim(struct wd_softc *wd, daddr_t bno, long size) 1957 { 1958 struct dk_softc *dksc = &wd->sc_dksc; 1959 struct ata_xfer *xfer; 1960 int error; 1961 unsigned char *req; 1962 1963 xfer = ata_get_xfer(wd->drvp->chnl_softc, true); 1964 1965 req = kmem_zalloc(512, KM_SLEEP); 1966 req[0] = bno & 0xff; 1967 req[1] = (bno >> 8) & 0xff; 1968 req[2] = (bno >> 16) & 0xff; 1969 req[3] = (bno >> 24) & 0xff; 1970 req[4] = (bno >> 32) & 0xff; 1971 req[5] = (bno >> 40) & 0xff; 1972 req[6] = size & 0xff; 1973 req[7] = (size >> 8) & 0xff; 1974 1975 /* 1976 * XXX We could possibly use NCQ TRIM, which supports executing 1977 * this command concurrently. It would need some investigation, some 1978 * early or not so early disk firmware caused data loss with NCQ TRIM. 1979 * atastart() et.al would need to be adjusted to allow and support 1980 * running several non-I/O ATA commands in parallel. 1981 */ 1982 1983 xfer->c_ata_c.r_command = ATA_DATA_SET_MANAGEMENT; 1984 xfer->c_ata_c.r_count = 1; 1985 xfer->c_ata_c.r_features = ATA_SUPPORT_DSM_TRIM; 1986 xfer->c_ata_c.r_st_bmask = WDCS_DRDY; 1987 xfer->c_ata_c.r_st_pmask = WDCS_DRDY; 1988 xfer->c_ata_c.timeout = 30000; /* 30s timeout */ 1989 xfer->c_ata_c.data = req; 1990 xfer->c_ata_c.bcount = 512; 1991 xfer->c_ata_c.flags |= AT_WRITE | AT_WAIT; 1992 1993 wd->atabus->ata_exec_command(wd->drvp, xfer); 1994 ata_wait_cmd(wd->drvp->chnl_softc, xfer); 1995 1996 kmem_free(req, 512); 1997 error = wd_check_error(dksc, xfer, __func__); 1998 ata_free_xfer(wd->drvp->chnl_softc, xfer); 1999 return error; 2000 } 2001 2002 bool 2003 wd_shutdown(device_t dev, int how) 2004 { 2005 struct wd_softc *wd = device_private(dev); 2006 2007 /* the adapter needs to be enabled */ 2008 if (wd->atabus->ata_addref(wd->drvp)) 2009 return true; /* no need to complain */ 2010 2011 wd_flushcache(wd, AT_POLL); 2012 if ((how & RB_POWERDOWN) == RB_POWERDOWN) 2013 wd_standby(wd, AT_POLL); 2014 return true; 2015 } 2016 2017 /* 2018 * Allocate space for a ioctl queue structure. Mostly taken from 2019 * scsipi_ioctl.c 2020 */ 2021 struct wd_ioctl * 2022 wi_get(struct wd_softc *wd) 2023 { 2024 struct wd_ioctl *wi; 2025 2026 wi = kmem_zalloc(sizeof(struct wd_ioctl), KM_SLEEP); 2027 wi->wi_softc = wd; 2028 buf_init(&wi->wi_bp); 2029 2030 return (wi); 2031 } 2032 2033 /* 2034 * Free an ioctl structure and remove it from our list 2035 */ 2036 2037 void 2038 wi_free(struct wd_ioctl *wi) 2039 { 2040 buf_destroy(&wi->wi_bp); 2041 kmem_free(wi, sizeof(*wi)); 2042 } 2043 2044 /* 2045 * Find a wd_ioctl structure based on the struct buf. 2046 */ 2047 2048 struct wd_ioctl * 2049 wi_find(struct buf *bp) 2050 { 2051 return container_of(bp, struct wd_ioctl, wi_bp); 2052 } 2053 2054 static uint 2055 wi_sector_size(const struct wd_ioctl * const wi) 2056 { 2057 switch (wi->wi_atareq.command) { 2058 case WDCC_READ: 2059 case WDCC_WRITE: 2060 case WDCC_READMULTI: 2061 case WDCC_WRITEMULTI: 2062 case WDCC_READDMA: 2063 case WDCC_WRITEDMA: 2064 case WDCC_READ_EXT: 2065 case WDCC_WRITE_EXT: 2066 case WDCC_READMULTI_EXT: 2067 case WDCC_WRITEMULTI_EXT: 2068 case WDCC_READDMA_EXT: 2069 case WDCC_WRITEDMA_EXT: 2070 case WDCC_READ_FPDMA_QUEUED: 2071 case WDCC_WRITE_FPDMA_QUEUED: 2072 return wi->wi_softc->sc_blksize; 2073 default: 2074 return 512; 2075 } 2076 } 2077 2078 /* 2079 * Ioctl pseudo strategy routine 2080 * 2081 * This is mostly stolen from scsipi_ioctl.c:scsistrategy(). What 2082 * happens here is: 2083 * 2084 * - wdioctl() queues a wd_ioctl structure. 2085 * 2086 * - wdioctl() calls physio/wdioctlstrategy based on whether or not 2087 * user space I/O is required. If physio() is called, physio() eventually 2088 * calls wdioctlstrategy(). 2089 * 2090 * - In either case, wdioctlstrategy() calls wd->atabus->ata_exec_command() 2091 * to perform the actual command 2092 * 2093 * The reason for the use of the pseudo strategy routine is because 2094 * when doing I/O to/from user space, physio _really_ wants to be in 2095 * the loop. We could put the entire buffer into the ioctl request 2096 * structure, but that won't scale if we want to do things like download 2097 * microcode. 2098 */ 2099 2100 void 2101 wdioctlstrategy(struct buf *bp) 2102 { 2103 struct wd_ioctl *wi; 2104 struct ata_xfer *xfer; 2105 int error = 0; 2106 2107 wi = wi_find(bp); 2108 if (wi == NULL) { 2109 printf("wdioctlstrategy: " 2110 "No matching ioctl request found in queue\n"); 2111 error = EINVAL; 2112 goto out2; 2113 } 2114 2115 xfer = ata_get_xfer(wi->wi_softc->drvp->chnl_softc, true); 2116 2117 /* 2118 * Abort if physio broke up the transfer 2119 */ 2120 2121 if (bp->b_bcount != wi->wi_atareq.datalen) { 2122 printf("physio split wd ioctl request... cannot proceed\n"); 2123 error = EIO; 2124 goto out; 2125 } 2126 2127 /* 2128 * Abort if we didn't get a buffer size that was a multiple of 2129 * our sector size (or overflows CHS/LBA28 sector count) 2130 */ 2131 2132 if ((bp->b_bcount % wi_sector_size(wi)) != 0 || 2133 (bp->b_bcount / wi_sector_size(wi)) >= 2134 (1 << NBBY)) { 2135 error = EINVAL; 2136 goto out; 2137 } 2138 2139 /* 2140 * Make sure a timeout was supplied in the ioctl request 2141 */ 2142 2143 if (wi->wi_atareq.timeout == 0) { 2144 error = EINVAL; 2145 goto out; 2146 } 2147 2148 if (wi->wi_atareq.flags & ATACMD_READ) 2149 xfer->c_ata_c.flags |= AT_READ; 2150 else if (wi->wi_atareq.flags & ATACMD_WRITE) 2151 xfer->c_ata_c.flags |= AT_WRITE; 2152 2153 if (wi->wi_atareq.flags & ATACMD_READREG) 2154 xfer->c_ata_c.flags |= AT_READREG; 2155 2156 if ((wi->wi_atareq.flags & ATACMD_LBA) != 0) 2157 xfer->c_ata_c.flags |= AT_LBA; 2158 2159 xfer->c_ata_c.flags |= AT_WAIT; 2160 2161 xfer->c_ata_c.timeout = wi->wi_atareq.timeout; 2162 xfer->c_ata_c.r_command = wi->wi_atareq.command; 2163 xfer->c_ata_c.r_lba = ((wi->wi_atareq.head & 0x0f) << 24) | 2164 (wi->wi_atareq.cylinder << 8) | 2165 wi->wi_atareq.sec_num; 2166 xfer->c_ata_c.r_count = wi->wi_atareq.sec_count; 2167 xfer->c_ata_c.r_features = wi->wi_atareq.features; 2168 xfer->c_ata_c.r_st_bmask = WDCS_DRDY; 2169 xfer->c_ata_c.r_st_pmask = WDCS_DRDY; 2170 xfer->c_ata_c.data = wi->wi_bp.b_data; 2171 xfer->c_ata_c.bcount = wi->wi_bp.b_bcount; 2172 2173 wi->wi_softc->atabus->ata_exec_command(wi->wi_softc->drvp, xfer); 2174 ata_wait_cmd(wi->wi_softc->drvp->chnl_softc, xfer); 2175 2176 if (xfer->c_ata_c.flags & (AT_ERROR | AT_TIMEOU | AT_DF)) { 2177 if (xfer->c_ata_c.flags & AT_ERROR) { 2178 wi->wi_atareq.retsts = ATACMD_ERROR; 2179 wi->wi_atareq.error = xfer->c_ata_c.r_error; 2180 } else if (xfer->c_ata_c.flags & AT_DF) 2181 wi->wi_atareq.retsts = ATACMD_DF; 2182 else 2183 wi->wi_atareq.retsts = ATACMD_TIMEOUT; 2184 } else { 2185 wi->wi_atareq.retsts = ATACMD_OK; 2186 if (wi->wi_atareq.flags & ATACMD_READREG) { 2187 wi->wi_atareq.command = xfer->c_ata_c.r_status; 2188 wi->wi_atareq.features = xfer->c_ata_c.r_error; 2189 wi->wi_atareq.sec_count = xfer->c_ata_c.r_count; 2190 wi->wi_atareq.sec_num = xfer->c_ata_c.r_lba & 0xff; 2191 wi->wi_atareq.head = (xfer->c_ata_c.r_device & 0xf0) | 2192 ((xfer->c_ata_c.r_lba >> 24) & 0x0f); 2193 wi->wi_atareq.cylinder = 2194 (xfer->c_ata_c.r_lba >> 8) & 0xffff; 2195 wi->wi_atareq.error = xfer->c_ata_c.r_error; 2196 } 2197 } 2198 2199 out: 2200 ata_free_xfer(wi->wi_softc->drvp->chnl_softc, xfer); 2201 out2: 2202 bp->b_error = error; 2203 if (error) 2204 bp->b_resid = bp->b_bcount; 2205 biodone(bp); 2206 } 2207 2208 static void 2209 wd_sysctl_attach(struct wd_softc *wd) 2210 { 2211 struct dk_softc *dksc = &wd->sc_dksc; 2212 const struct sysctlnode *node; 2213 int error; 2214 2215 /* sysctl set-up */ 2216 if (sysctl_createv(&wd->nodelog, 0, NULL, &node, 2217 0, CTLTYPE_NODE, dksc->sc_xname, 2218 SYSCTL_DESCR("wd driver settings"), 2219 NULL, 0, NULL, 0, 2220 CTL_HW, CTL_CREATE, CTL_EOL) != 0) { 2221 aprint_error_dev(dksc->sc_dev, 2222 "could not create %s.%s sysctl node\n", 2223 "hw", dksc->sc_xname); 2224 return; 2225 } 2226 2227 wd->drv_ncq = true; 2228 if ((error = sysctl_createv(&wd->nodelog, 0, NULL, NULL, 2229 CTLFLAG_READWRITE, CTLTYPE_BOOL, "use_ncq", 2230 SYSCTL_DESCR("use NCQ if supported"), 2231 NULL, 0, &wd->drv_ncq, 0, 2232 CTL_HW, node->sysctl_num, CTL_CREATE, CTL_EOL)) 2233 != 0) { 2234 aprint_error_dev(dksc->sc_dev, 2235 "could not create %s.%s.use_ncq sysctl - error %d\n", 2236 "hw", dksc->sc_xname, error); 2237 return; 2238 } 2239 2240 wd->drv_ncq_prio = false; 2241 if ((error = sysctl_createv(&wd->nodelog, 0, NULL, NULL, 2242 CTLFLAG_READWRITE, CTLTYPE_BOOL, "use_ncq_prio", 2243 SYSCTL_DESCR("use NCQ PRIORITY if supported"), 2244 NULL, 0, &wd->drv_ncq_prio, 0, 2245 CTL_HW, node->sysctl_num, CTL_CREATE, CTL_EOL)) 2246 != 0) { 2247 aprint_error_dev(dksc->sc_dev, 2248 "could not create %s.%s.use_ncq_prio sysctl - error %d\n", 2249 "hw", dksc->sc_xname, error); 2250 return; 2251 } 2252 2253 #ifdef WD_CHAOS_MONKEY 2254 wd->drv_chaos_freq = 0; 2255 if ((error = sysctl_createv(&wd->nodelog, 0, NULL, NULL, 2256 CTLFLAG_READWRITE, CTLTYPE_INT, "chaos_freq", 2257 SYSCTL_DESCR("simulated bio read error rate"), 2258 NULL, 0, &wd->drv_chaos_freq, 0, 2259 CTL_HW, node->sysctl_num, CTL_CREATE, CTL_EOL)) 2260 != 0) { 2261 aprint_error_dev(dksc->sc_dev, 2262 "could not create %s.%s.chaos_freq sysctl - error %d\n", 2263 "hw", dksc->sc_xname, error); 2264 return; 2265 } 2266 2267 wd->drv_chaos_cnt = 0; 2268 if ((error = sysctl_createv(&wd->nodelog, 0, NULL, NULL, 2269 CTLFLAG_READONLY, CTLTYPE_INT, "chaos_cnt", 2270 SYSCTL_DESCR("number of processed bio reads"), 2271 NULL, 0, &wd->drv_chaos_cnt, 0, 2272 CTL_HW, node->sysctl_num, CTL_CREATE, CTL_EOL)) 2273 != 0) { 2274 aprint_error_dev(dksc->sc_dev, 2275 "could not create %s.%s.chaos_cnt sysctl - error %d\n", 2276 "hw", dksc->sc_xname, error); 2277 return; 2278 } 2279 #endif 2280 2281 } 2282 2283 static void 2284 wd_sysctl_detach(struct wd_softc *wd) 2285 { 2286 sysctl_teardown(&wd->nodelog); 2287 } 2288 2289 #ifdef ATADEBUG 2290 int wddebug(void); 2291 2292 int 2293 wddebug(void) 2294 { 2295 struct wd_softc *wd; 2296 struct dk_softc *dksc; 2297 int unit; 2298 2299 for (unit = 0; unit <= 3; unit++) { 2300 wd = device_lookup_private(&wd_cd, unit); 2301 if (wd == NULL) 2302 continue; 2303 dksc = &wd->sc_dksc; 2304 printf("%s fl %x bufq %p:\n", 2305 dksc->sc_xname, wd->sc_flags, bufq_peek(dksc->sc_bufq)); 2306 2307 atachannel_debug(wd->drvp->chnl_softc); 2308 } 2309 return 0; 2310 } 2311 #endif /* ATADEBUG */ 2312
Indexes created Sat Sep 27 22:09:54 GMT 2025