1 /* $NetBSD: subr_disk.c,v 1.97 2009/05/20 03:26:21 dyoung Exp $ */ 2 3 /*- 4 * Copyright (c) 1996, 1997, 1999, 2000, 2009 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Jason R. Thorpe of the Numerical Aerospace Simulation Facility, 9 * NASA Ames Research Center. 10 * 11 * Redistribution and use in source and binary forms, with or without 12 * modification, are permitted provided that the following conditions 13 * are met: 14 * 1. Redistributions of source code must retain the above copyright 15 * notice, this list of conditions and the following disclaimer. 16 * 2. Redistributions in binary form must reproduce the above copyright 17 * notice, this list of conditions and the following disclaimer in the 18 * documentation and/or other materials provided with the distribution. 19 * 20 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 21 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 22 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 23 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 24 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 25 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 26 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 27 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 28 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 29 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 30 * POSSIBILITY OF SUCH DAMAGE. 31 */ 32 33 /* 34 * Copyright (c) 1982, 1986, 1988, 1993 35 * The Regents of the University of California. All rights reserved. 36 * (c) UNIX System Laboratories, Inc. 37 * All or some portions of this file are derived from material licensed 38 * to the University of California by American Telephone and Telegraph 39 * Co. or Unix System Laboratories, Inc. and are reproduced herein with 40 * the permission of UNIX System Laboratories, Inc. 41 * 42 * Redistribution and use in source and binary forms, with or without 43 * modification, are permitted provided that the following conditions 44 * are met: 45 * 1. Redistributions of source code must retain the above copyright 46 * notice, this list of conditions and the following disclaimer. 47 * 2. Redistributions in binary form must reproduce the above copyright 48 * notice, this list of conditions and the following disclaimer in the 49 * documentation and/or other materials provided with the distribution. 50 * 3. Neither the name of the University nor the names of its contributors 51 * may be used to endorse or promote products derived from this software 52 * without specific prior written permission. 53 * 54 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 55 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 56 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 57 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 58 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 59 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 60 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 61 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 62 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 63 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 64 * SUCH DAMAGE. 65 * 66 * @(#)ufs_disksubr.c 8.5 (Berkeley) 1/21/94 67 */ 68 69 #include <sys/cdefs.h> 70 __KERNEL_RCSID(0, "$NetBSD: subr_disk.c,v 1.97 2009/05/20 03:26:21 dyoung Exp $"); 71 72 #include <sys/param.h> 73 #include <sys/kernel.h> 74 #include <sys/kmem.h> 75 #include <sys/buf.h> 76 #include <sys/syslog.h> 77 #include <sys/disklabel.h> 78 #include <sys/disk.h> 79 #include <sys/sysctl.h> 80 #include <lib/libkern/libkern.h> 81 82 /* 83 * Compute checksum for disk label. 84 */ 85 u_int 86 dkcksum(struct disklabel *lp) 87 { 88 return dkcksum_sized(lp, lp->d_npartitions); 89 } 90 91 u_int 92 dkcksum_sized(struct disklabel *lp, size_t npartitions) 93 { 94 u_short *start, *end; 95 u_short sum = 0; 96 97 start = (u_short *)lp; 98 end = (u_short *)&lp->d_partitions[npartitions]; 99 while (start < end) 100 sum ^= *start++; 101 return (sum); 102 } 103 104 /* 105 * Disk error is the preface to plaintive error messages 106 * about failing disk transfers. It prints messages of the form 107 108 hp0g: hard error reading fsbn 12345 of 12344-12347 (hp0 bn %d cn %d tn %d sn %d) 109 110 * if the offset of the error in the transfer and a disk label 111 * are both available. blkdone should be -1 if the position of the error 112 * is unknown; the disklabel pointer may be null from drivers that have not 113 * been converted to use them. The message is printed with printf 114 * if pri is LOG_PRINTF, otherwise it uses log at the specified priority. 115 * The message should be completed (with at least a newline) with printf 116 * or addlog, respectively. There is no trailing space. 117 */ 118 #ifndef PRIdaddr 119 #define PRIdaddr PRId64 120 #endif 121 void 122 diskerr(const struct buf *bp, const char *dname, const char *what, int pri, 123 int blkdone, const struct disklabel *lp) 124 { 125 int unit = DISKUNIT(bp->b_dev), part = DISKPART(bp->b_dev); 126 void (*pr)(const char *, ...); 127 char partname = 'a' + part; 128 daddr_t sn; 129 130 if (/*CONSTCOND*/0) 131 /* Compiler will error this is the format is wrong... */ 132 printf("%" PRIdaddr, bp->b_blkno); 133 134 if (pri != LOG_PRINTF) { 135 static const char fmt[] = ""; 136 log(pri, fmt); 137 pr = addlog; 138 } else 139 pr = printf; 140 (*pr)("%s%d%c: %s %sing fsbn ", dname, unit, partname, what, 141 bp->b_flags & B_READ ? "read" : "writ"); 142 sn = bp->b_blkno; 143 if (bp->b_bcount <= DEV_BSIZE) 144 (*pr)("%" PRIdaddr, sn); 145 else { 146 if (blkdone >= 0) { 147 sn += blkdone; 148 (*pr)("%" PRIdaddr " of ", sn); 149 } 150 (*pr)("%" PRIdaddr "-%" PRIdaddr "", bp->b_blkno, 151 bp->b_blkno + (bp->b_bcount - 1) / DEV_BSIZE); 152 } 153 if (lp && (blkdone >= 0 || bp->b_bcount <= lp->d_secsize)) { 154 sn += lp->d_partitions[part].p_offset; 155 (*pr)(" (%s%d bn %" PRIdaddr "; cn %" PRIdaddr "", 156 dname, unit, sn, sn / lp->d_secpercyl); 157 sn %= lp->d_secpercyl; 158 (*pr)(" tn %" PRIdaddr " sn %" PRIdaddr ")", 159 sn / lp->d_nsectors, sn % lp->d_nsectors); 160 } 161 } 162 163 /* 164 * Searches the iostatlist for the disk corresponding to the 165 * name provided. 166 */ 167 struct disk * 168 disk_find(const char *name) 169 { 170 struct io_stats *stat; 171 172 stat = iostat_find(name); 173 174 if ((stat != NULL) && (stat->io_type == IOSTAT_DISK)) 175 return stat->io_parent; 176 177 return (NULL); 178 } 179 180 void 181 disk_init(struct disk *diskp, const char *name, const struct dkdriver *driver) 182 { 183 184 /* 185 * Initialize the wedge-related locks and other fields. 186 */ 187 mutex_init(&diskp->dk_rawlock, MUTEX_DEFAULT, IPL_NONE); 188 mutex_init(&diskp->dk_openlock, MUTEX_DEFAULT, IPL_NONE); 189 LIST_INIT(&diskp->dk_wedges); 190 diskp->dk_nwedges = 0; 191 diskp->dk_labelsector = LABELSECTOR; 192 disk_blocksize(diskp, DEV_BSIZE); 193 diskp->dk_name = name; 194 diskp->dk_driver = driver; 195 } 196 197 /* 198 * Attach a disk. 199 */ 200 void 201 disk_attach(struct disk *diskp) 202 { 203 204 /* 205 * Allocate and initialize the disklabel structures. 206 */ 207 diskp->dk_label = kmem_zalloc(sizeof(struct disklabel), KM_SLEEP); 208 diskp->dk_cpulabel = kmem_zalloc(sizeof(struct cpu_disklabel), 209 KM_SLEEP); 210 if ((diskp->dk_label == NULL) || (diskp->dk_cpulabel == NULL)) 211 panic("disk_attach: can't allocate storage for disklabel"); 212 213 /* 214 * Set up the stats collection. 215 */ 216 diskp->dk_stats = iostat_alloc(IOSTAT_DISK, diskp, diskp->dk_name); 217 } 218 219 int 220 disk_begindetach(struct disk *dk, int (*lastclose)(device_t), 221 device_t self, int flags) 222 { 223 int rc; 224 225 rc = 0; 226 mutex_enter(&dk->dk_openlock); 227 if (dk->dk_openmask == 0) 228 ; /* nothing to do */ 229 else if ((flags & DETACH_FORCE) == 0) 230 rc = EBUSY; 231 else if (lastclose != NULL) 232 rc = (*lastclose)(self); 233 mutex_exit(&dk->dk_openlock); 234 235 return rc; 236 } 237 238 /* 239 * Detach a disk. 240 */ 241 void 242 disk_detach(struct disk *diskp) 243 { 244 245 /* 246 * Remove from the drivelist. 247 */ 248 iostat_free(diskp->dk_stats); 249 250 /* 251 * Release the disk-info dictionary. 252 */ 253 if (diskp->dk_info) { 254 prop_object_release(diskp->dk_info); 255 diskp->dk_info = NULL; 256 } 257 258 /* 259 * Free the space used by the disklabel structures. 260 */ 261 kmem_free(diskp->dk_label, sizeof(*diskp->dk_label)); 262 kmem_free(diskp->dk_cpulabel, sizeof(*diskp->dk_cpulabel)); 263 } 264 265 void 266 disk_destroy(struct disk *diskp) 267 { 268 269 mutex_destroy(&diskp->dk_openlock); 270 mutex_destroy(&diskp->dk_rawlock); 271 } 272 273 /* 274 * Mark the disk as busy for metrics collection. 275 */ 276 void 277 disk_busy(struct disk *diskp) 278 { 279 280 iostat_busy(diskp->dk_stats); 281 } 282 283 /* 284 * Finished disk operations, gather metrics. 285 */ 286 void 287 disk_unbusy(struct disk *diskp, long bcount, int read) 288 { 289 290 iostat_unbusy(diskp->dk_stats, bcount, read); 291 } 292 293 /* 294 * Return true if disk has an I/O operation in flight. 295 */ 296 bool 297 disk_isbusy(struct disk *diskp) 298 { 299 300 return iostat_isbusy(diskp->dk_stats); 301 } 302 303 /* 304 * Set the physical blocksize of a disk, in bytes. 305 * Only necessary if blocksize != DEV_BSIZE. 306 */ 307 void 308 disk_blocksize(struct disk *diskp, int blocksize) 309 { 310 311 diskp->dk_blkshift = DK_BSIZE2BLKSHIFT(blocksize); 312 diskp->dk_byteshift = DK_BSIZE2BYTESHIFT(blocksize); 313 } 314 315 /* 316 * Bounds checking against the media size, used for the raw partition. 317 * The sector size passed in should currently always be DEV_BSIZE, 318 * and the media size the size of the device in DEV_BSIZE sectors. 319 */ 320 int 321 bounds_check_with_mediasize(struct buf *bp, int secsize, uint64_t mediasize) 322 { 323 int64_t sz; 324 325 sz = howmany(bp->b_bcount, secsize); 326 327 if (bp->b_blkno + sz > mediasize) { 328 sz = mediasize - bp->b_blkno; 329 if (sz == 0) { 330 /* If exactly at end of disk, return EOF. */ 331 bp->b_resid = bp->b_bcount; 332 return 0; 333 } 334 if (sz < 0) { 335 /* If past end of disk, return EINVAL. */ 336 bp->b_error = EINVAL; 337 return 0; 338 } 339 /* Otherwise, truncate request. */ 340 bp->b_bcount = sz << DEV_BSHIFT; 341 } 342 343 return 1; 344 } 345 346 /* 347 * Determine the size of the transfer, and make sure it is 348 * within the boundaries of the partition. Adjust transfer 349 * if needed, and signal errors or early completion. 350 */ 351 int 352 bounds_check_with_label(struct disk *dk, struct buf *bp, int wlabel) 353 { 354 struct disklabel *lp = dk->dk_label; 355 struct partition *p = lp->d_partitions + DISKPART(bp->b_dev); 356 uint64_t p_size, p_offset, labelsector; 357 int64_t sz; 358 359 /* Protect against division by zero. XXX: Should never happen?!?! */ 360 if (lp->d_secpercyl == 0) { 361 bp->b_error = EINVAL; 362 return -1; 363 } 364 365 p_size = p->p_size << dk->dk_blkshift; 366 p_offset = p->p_offset << dk->dk_blkshift; 367 #if RAW_PART == 3 368 labelsector = lp->d_partitions[2].p_offset; 369 #else 370 labelsector = lp->d_partitions[RAW_PART].p_offset; 371 #endif 372 labelsector = (labelsector + dk->dk_labelsector) << dk->dk_blkshift; 373 374 sz = howmany(bp->b_bcount, DEV_BSIZE); 375 if ((bp->b_blkno + sz) > p_size) { 376 sz = p_size - bp->b_blkno; 377 if (sz == 0) { 378 /* If exactly at end of disk, return EOF. */ 379 bp->b_resid = bp->b_bcount; 380 return 0; 381 } 382 if (sz < 0) { 383 /* If past end of disk, return EINVAL. */ 384 bp->b_error = EINVAL; 385 return -1; 386 } 387 /* Otherwise, truncate request. */ 388 bp->b_bcount = sz << DEV_BSHIFT; 389 } 390 391 /* Overwriting disk label? */ 392 if (bp->b_blkno + p_offset <= labelsector && 393 bp->b_blkno + p_offset + sz > labelsector && 394 (bp->b_flags & B_READ) == 0 && !wlabel) { 395 bp->b_error = EROFS; 396 return -1; 397 } 398 399 /* calculate cylinder for disksort to order transfers with */ 400 bp->b_cylinder = (bp->b_blkno + p->p_offset) / 401 (lp->d_secsize / DEV_BSIZE) / lp->d_secpercyl; 402 return 1; 403 } 404 405 int 406 disk_read_sectors(void (*strat)(struct buf *), const struct disklabel *lp, 407 struct buf *bp, unsigned int sector, int count) 408 { 409 bp->b_blkno = sector; 410 bp->b_bcount = count * lp->d_secsize; 411 bp->b_flags = (bp->b_flags & ~B_WRITE) | B_READ; 412 bp->b_oflags &= ~BO_DONE; 413 bp->b_cylinder = sector / lp->d_secpercyl; 414 (*strat)(bp); 415 return biowait(bp); 416 } 417 418 const char * 419 convertdisklabel(struct disklabel *lp, void (*strat)(struct buf *), 420 struct buf *bp, uint32_t secperunit) 421 { 422 struct partition rp, *altp, *p; 423 int geom_ok; 424 425 memset(&rp, 0, sizeof(rp)); 426 rp.p_size = secperunit; 427 rp.p_fstype = FS_UNUSED; 428 429 /* If we can seek to d_secperunit - 1, believe the disk geometry. */ 430 if (secperunit != 0 && 431 disk_read_sectors(strat, lp, bp, secperunit - 1, 1) == 0) 432 geom_ok = 1; 433 else 434 geom_ok = 0; 435 436 #if 0 437 printf("%s: secperunit (%" PRIu32 ") %s\n", __func__, 438 secperunit, geom_ok ? "ok" : "not ok"); 439 #endif 440 441 p = &lp->d_partitions[RAW_PART]; 442 if (RAW_PART == 'c' - 'a') 443 altp = &lp->d_partitions['d' - 'a']; 444 else 445 altp = &lp->d_partitions['c' - 'a']; 446 447 if (lp->d_npartitions > RAW_PART && p->p_offset == 0 && p->p_size != 0) 448 ; /* already a raw partition */ 449 else if (lp->d_npartitions > MAX('c', 'd') - 'a' && 450 altp->p_offset == 0 && altp->p_size != 0) { 451 /* alternate partition ('c' or 'd') is suitable for raw slot, 452 * swap with 'd' or 'c'. 453 */ 454 rp = *p; 455 *p = *altp; 456 *altp = rp; 457 } else if (lp->d_npartitions <= RAW_PART && 458 lp->d_npartitions > 'c' - 'a') { 459 /* No raw partition is present, but the alternate is present. 460 * Copy alternate to raw partition. 461 */ 462 lp->d_npartitions = RAW_PART + 1; 463 *p = *altp; 464 } else if (!geom_ok) 465 return "no raw partition and disk reports bad geometry"; 466 else if (lp->d_npartitions <= RAW_PART) { 467 memset(&lp->d_partitions[lp->d_npartitions], 0, 468 sizeof(struct partition) * (RAW_PART - lp->d_npartitions)); 469 *p = rp; 470 lp->d_npartitions = RAW_PART + 1; 471 } else if (lp->d_npartitions < MAXPARTITIONS) { 472 memmove(p + 1, p, 473 sizeof(struct partition) * (lp->d_npartitions - RAW_PART)); 474 *p = rp; 475 lp->d_npartitions++; 476 } else 477 return "no raw partition and partition table is full"; 478 return NULL; 479 } 480 481 /* 482 * disk_ioctl -- 483 * Generic disk ioctl handling. 484 */ 485 int 486 disk_ioctl(struct disk *diskp, u_long cmd, void *data, int flag, 487 struct lwp *l) 488 { 489 int error; 490 491 switch (cmd) { 492 case DIOCGDISKINFO: 493 { 494 struct plistref *pref = (struct plistref *) data; 495 496 if (diskp->dk_info == NULL) 497 error = ENOTSUP; 498 else 499 error = prop_dictionary_copyout_ioctl(pref, cmd, 500 diskp->dk_info); 501 break; 502 } 503 504 default: 505 error = EPASSTHROUGH; 506 } 507 508 return (error); 509 } 510