1 /* $NetBSD: fss.c,v 1.37 2007/11/10 18:53:57 rmind Exp $ */ 2 3 /*- 4 * Copyright (c) 2003 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Juergen Hannken-Illjes. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 3. All advertising materials mentioning features or use of this software 19 * must display the following acknowledgement: 20 * This product includes software developed by the NetBSD 21 * Foundation, Inc. and its contributors. 22 * 4. Neither the name of The NetBSD Foundation nor the names of its 23 * contributors may be used to endorse or promote products derived 24 * from this software without specific prior written permission. 25 * 26 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 27 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 28 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 29 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 30 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 31 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 32 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 33 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 34 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 35 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 36 * POSSIBILITY OF SUCH DAMAGE. 37 */ 38 39 /* 40 * File system snapshot disk driver. 41 * 42 * Block/character interface to the snapshot of a mounted file system. 43 */ 44 45 #include <sys/cdefs.h> 46 __KERNEL_RCSID(0, "$NetBSD: fss.c,v 1.37 2007/11/10 18:53:57 rmind Exp $"); 47 48 #include "fss.h" 49 50 #include <sys/param.h> 51 #include <sys/systm.h> 52 #include <sys/namei.h> 53 #include <sys/proc.h> 54 #include <sys/errno.h> 55 #include <sys/buf.h> 56 #include <sys/malloc.h> 57 #include <sys/ioctl.h> 58 #include <sys/disklabel.h> 59 #include <sys/device.h> 60 #include <sys/disk.h> 61 #include <sys/stat.h> 62 #include <sys/mount.h> 63 #include <sys/vnode.h> 64 #include <sys/file.h> 65 #include <sys/uio.h> 66 #include <sys/conf.h> 67 #include <sys/kthread.h> 68 #include <sys/fstrans.h> 69 70 #include <miscfs/specfs/specdev.h> 71 72 #include <dev/fssvar.h> 73 74 #include <machine/stdarg.h> 75 76 #ifdef DEBUG 77 #define FSS_STATISTICS 78 #endif 79 80 #ifdef FSS_STATISTICS 81 struct fss_stat { 82 u_int64_t cow_calls; 83 u_int64_t cow_copied; 84 u_int64_t cow_cache_full; 85 u_int64_t indir_read; 86 u_int64_t indir_write; 87 }; 88 89 static struct fss_stat fss_stat[NFSS]; 90 91 #define FSS_STAT_INC(sc, field) \ 92 do { \ 93 fss_stat[sc->sc_unit].field++; \ 94 } while (0) 95 #define FSS_STAT_SET(sc, field, value) \ 96 do { \ 97 fss_stat[sc->sc_unit].field = value; \ 98 } while (0) 99 #define FSS_STAT_ADD(sc, field, value) \ 100 do { \ 101 fss_stat[sc->sc_unit].field += value; \ 102 } while (0) 103 #define FSS_STAT_VAL(sc, field) fss_stat[sc->sc_unit].field 104 #define FSS_STAT_CLEAR(sc) \ 105 do { \ 106 memset(&fss_stat[sc->sc_unit], 0, \ 107 sizeof(struct fss_stat)); \ 108 } while (0) 109 #else /* FSS_STATISTICS */ 110 #define FSS_STAT_INC(sc, field) 111 #define FSS_STAT_SET(sc, field, value) 112 #define FSS_STAT_ADD(sc, field, value) 113 #define FSS_STAT_CLEAR(sc) 114 #endif /* FSS_STATISTICS */ 115 116 static struct fss_softc fss_softc[NFSS]; 117 118 void fssattach(int); 119 120 dev_type_open(fss_open); 121 dev_type_close(fss_close); 122 dev_type_read(fss_read); 123 dev_type_write(fss_write); 124 dev_type_ioctl(fss_ioctl); 125 dev_type_strategy(fss_strategy); 126 dev_type_dump(fss_dump); 127 dev_type_size(fss_size); 128 129 static int fss_copy_on_write(void *, struct buf *); 130 static inline void fss_error(struct fss_softc *, const char *, ...); 131 static int fss_create_files(struct fss_softc *, struct fss_set *, 132 off_t *, struct lwp *); 133 static int fss_create_snapshot(struct fss_softc *, struct fss_set *, 134 struct lwp *); 135 static int fss_delete_snapshot(struct fss_softc *, struct lwp *); 136 static int fss_softc_alloc(struct fss_softc *); 137 static void fss_softc_free(struct fss_softc *); 138 static void fss_cluster_iodone(struct buf *); 139 static void fss_read_cluster(struct fss_softc *, u_int32_t); 140 static void fss_bs_thread(void *); 141 static int fss_bs_io(struct fss_softc *, fss_io_type, 142 u_int32_t, off_t, int, void *); 143 static u_int32_t *fss_bs_indir(struct fss_softc *, u_int32_t); 144 145 const struct bdevsw fss_bdevsw = { 146 fss_open, fss_close, fss_strategy, fss_ioctl, 147 fss_dump, fss_size, D_DISK 148 }; 149 150 const struct cdevsw fss_cdevsw = { 151 fss_open, fss_close, fss_read, fss_write, fss_ioctl, 152 nostop, notty, nopoll, nommap, nokqfilter, D_DISK 153 }; 154 155 void 156 fssattach(int num) 157 { 158 int i; 159 struct fss_softc *sc; 160 161 for (i = 0; i < NFSS; i++) { 162 sc = &fss_softc[i]; 163 sc->sc_unit = i; 164 sc->sc_bdev = NODEV; 165 simple_lock_init(&sc->sc_slock); 166 mutex_init(&sc->sc_lock, MUTEX_DRIVER, IPL_NONE); 167 bufq_alloc(&sc->sc_bufq, "fcfs", 0); 168 } 169 } 170 171 int 172 fss_open(dev_t dev, int flags, int mode, struct lwp *l) 173 { 174 int s, mflag; 175 struct fss_softc *sc; 176 177 mflag = (mode == S_IFCHR ? FSS_CDEV_OPEN : FSS_BDEV_OPEN); 178 179 if ((sc = FSS_DEV_TO_SOFTC(dev)) == NULL) 180 return ENODEV; 181 182 FSS_LOCK(sc, s); 183 184 sc->sc_flags |= mflag; 185 186 FSS_UNLOCK(sc, s); 187 188 return 0; 189 } 190 191 int 192 fss_close(dev_t dev, int flags, int mode, struct lwp *l) 193 { 194 int s, mflag, error; 195 struct fss_softc *sc; 196 197 mflag = (mode == S_IFCHR ? FSS_CDEV_OPEN : FSS_BDEV_OPEN); 198 199 if ((sc = FSS_DEV_TO_SOFTC(dev)) == NULL) 200 return ENODEV; 201 202 FSS_LOCK(sc, s); 203 204 if ((sc->sc_flags & (FSS_CDEV_OPEN|FSS_BDEV_OPEN)) == mflag) { 205 if ((sc->sc_uflags & FSS_UNCONFIG_ON_CLOSE) != 0 && 206 (sc->sc_flags & FSS_ACTIVE) != 0) { 207 FSS_UNLOCK(sc, s); 208 error = fss_ioctl(dev, FSSIOCCLR, NULL, FWRITE, l); 209 if (error) 210 return error; 211 FSS_LOCK(sc, s); 212 } 213 sc->sc_uflags &= ~FSS_UNCONFIG_ON_CLOSE; 214 } 215 216 sc->sc_flags &= ~mflag; 217 218 FSS_UNLOCK(sc, s); 219 220 return 0; 221 } 222 223 void 224 fss_strategy(struct buf *bp) 225 { 226 int s; 227 struct fss_softc *sc; 228 229 sc = FSS_DEV_TO_SOFTC(bp->b_dev); 230 231 FSS_LOCK(sc, s); 232 233 if ((bp->b_flags & B_READ) != B_READ || 234 sc == NULL || !FSS_ISVALID(sc)) { 235 236 FSS_UNLOCK(sc, s); 237 238 bp->b_error = (sc == NULL ? ENODEV : EROFS); 239 bp->b_resid = bp->b_bcount; 240 biodone(bp); 241 return; 242 } 243 244 bp->b_rawblkno = bp->b_blkno; 245 BUFQ_PUT(sc->sc_bufq, bp); 246 wakeup(&sc->sc_bs_lwp); 247 248 FSS_UNLOCK(sc, s); 249 } 250 251 int 252 fss_read(dev_t dev, struct uio *uio, int flags) 253 { 254 return physio(fss_strategy, NULL, dev, B_READ, minphys, uio); 255 } 256 257 int 258 fss_write(dev_t dev, struct uio *uio, int flags) 259 { 260 return physio(fss_strategy, NULL, dev, B_WRITE, minphys, uio); 261 } 262 263 int 264 fss_ioctl(dev_t dev, u_long cmd, void *data, int flag, struct lwp *l) 265 { 266 int error; 267 struct fss_softc *sc; 268 struct fss_set *fss = (struct fss_set *)data; 269 struct fss_get *fsg = (struct fss_get *)data; 270 271 if ((sc = FSS_DEV_TO_SOFTC(dev)) == NULL) 272 return ENODEV; 273 274 switch (cmd) { 275 case FSSIOCSET: 276 mutex_enter(&sc->sc_lock); 277 if ((flag & FWRITE) == 0) 278 error = EPERM; 279 else if ((sc->sc_flags & FSS_ACTIVE) != 0) 280 error = EBUSY; 281 else 282 error = fss_create_snapshot(sc, fss, l); 283 mutex_exit(&sc->sc_lock); 284 break; 285 286 case FSSIOCCLR: 287 mutex_enter(&sc->sc_lock); 288 if ((flag & FWRITE) == 0) 289 error = EPERM; 290 else if ((sc->sc_flags & FSS_ACTIVE) == 0) 291 error = ENXIO; 292 else 293 error = fss_delete_snapshot(sc, l); 294 mutex_exit(&sc->sc_lock); 295 break; 296 297 case FSSIOCGET: 298 mutex_enter(&sc->sc_lock); 299 switch (sc->sc_flags & (FSS_PERSISTENT | FSS_ACTIVE)) { 300 case FSS_ACTIVE: 301 memcpy(fsg->fsg_mount, sc->sc_mntname, MNAMELEN); 302 fsg->fsg_csize = FSS_CLSIZE(sc); 303 fsg->fsg_time = sc->sc_time; 304 fsg->fsg_mount_size = sc->sc_clcount; 305 fsg->fsg_bs_size = sc->sc_clnext; 306 error = 0; 307 break; 308 case FSS_PERSISTENT | FSS_ACTIVE: 309 memcpy(fsg->fsg_mount, sc->sc_mntname, MNAMELEN); 310 fsg->fsg_csize = 0; 311 fsg->fsg_time = sc->sc_time; 312 fsg->fsg_mount_size = 0; 313 fsg->fsg_bs_size = 0; 314 error = 0; 315 break; 316 default: 317 error = ENXIO; 318 break; 319 } 320 mutex_exit(&sc->sc_lock); 321 break; 322 323 case FSSIOFSET: 324 sc->sc_uflags = *(int *)data; 325 error = 0; 326 break; 327 328 case FSSIOFGET: 329 *(int *)data = sc->sc_uflags; 330 error = 0; 331 break; 332 333 default: 334 error = EINVAL; 335 break; 336 } 337 338 return error; 339 } 340 341 int 342 fss_size(dev_t dev) 343 { 344 return -1; 345 } 346 347 int 348 fss_dump(dev_t dev, daddr_t blkno, void *va, 349 size_t size) 350 { 351 return EROFS; 352 } 353 354 /* 355 * An error occurred reading or writing the snapshot or backing store. 356 * If it is the first error log to console. 357 * The caller holds the simplelock. 358 */ 359 static inline void 360 fss_error(struct fss_softc *sc, const char *fmt, ...) 361 { 362 va_list ap; 363 364 if ((sc->sc_flags & (FSS_ACTIVE|FSS_ERROR)) == FSS_ACTIVE) { 365 va_start(ap, fmt); 366 printf("fss%d: snapshot invalid: ", sc->sc_unit); 367 vprintf(fmt, ap); 368 printf("\n"); 369 va_end(ap); 370 } 371 if ((sc->sc_flags & FSS_ACTIVE) == FSS_ACTIVE) 372 sc->sc_flags |= FSS_ERROR; 373 } 374 375 /* 376 * Allocate the variable sized parts of the softc and 377 * fork the kernel thread. 378 * 379 * The fields sc_clcount, sc_clshift, sc_cache_size and sc_indir_size 380 * must be initialized. 381 */ 382 static int 383 fss_softc_alloc(struct fss_softc *sc) 384 { 385 int i, len, error; 386 387 len = (sc->sc_clcount+NBBY-1)/NBBY; 388 sc->sc_copied = malloc(len, M_TEMP, M_ZERO|M_WAITOK|M_CANFAIL); 389 if (sc->sc_copied == NULL) 390 return(ENOMEM); 391 392 len = sc->sc_cache_size*sizeof(struct fss_cache); 393 sc->sc_cache = malloc(len, M_TEMP, M_ZERO|M_WAITOK|M_CANFAIL); 394 if (sc->sc_cache == NULL) 395 return(ENOMEM); 396 397 len = FSS_CLSIZE(sc); 398 for (i = 0; i < sc->sc_cache_size; i++) { 399 sc->sc_cache[i].fc_type = FSS_CACHE_FREE; 400 sc->sc_cache[i].fc_softc = sc; 401 sc->sc_cache[i].fc_xfercount = 0; 402 sc->sc_cache[i].fc_data = malloc(len, M_TEMP, 403 M_WAITOK|M_CANFAIL); 404 if (sc->sc_cache[i].fc_data == NULL) 405 return(ENOMEM); 406 } 407 408 len = (sc->sc_indir_size+NBBY-1)/NBBY; 409 sc->sc_indir_valid = malloc(len, M_TEMP, M_ZERO|M_WAITOK|M_CANFAIL); 410 if (sc->sc_indir_valid == NULL) 411 return(ENOMEM); 412 413 len = FSS_CLSIZE(sc); 414 sc->sc_indir_data = malloc(len, M_TEMP, M_ZERO|M_WAITOK|M_CANFAIL); 415 if (sc->sc_indir_data == NULL) 416 return(ENOMEM); 417 418 if ((error = kthread_create(PRI_BIO, 0, NULL, fss_bs_thread, sc, 419 &sc->sc_bs_lwp, "fssbs%d", sc->sc_unit)) != 0) 420 return error; 421 422 sc->sc_flags |= FSS_BS_THREAD; 423 return 0; 424 } 425 426 /* 427 * Free the variable sized parts of the softc. 428 */ 429 static void 430 fss_softc_free(struct fss_softc *sc) 431 { 432 int s, i; 433 434 if ((sc->sc_flags & FSS_BS_THREAD) != 0) { 435 FSS_LOCK(sc, s); 436 sc->sc_flags &= ~FSS_BS_THREAD; 437 wakeup(&sc->sc_bs_lwp); 438 while (sc->sc_bs_lwp != NULL) 439 ltsleep(&sc->sc_bs_lwp, PRIBIO, "fssthread", 0, 440 &sc->sc_slock); 441 FSS_UNLOCK(sc, s); 442 } 443 444 if (sc->sc_copied != NULL) 445 free(sc->sc_copied, M_TEMP); 446 sc->sc_copied = NULL; 447 448 if (sc->sc_cache != NULL) { 449 for (i = 0; i < sc->sc_cache_size; i++) 450 if (sc->sc_cache[i].fc_data != NULL) 451 free(sc->sc_cache[i].fc_data, M_TEMP); 452 free(sc->sc_cache, M_TEMP); 453 } 454 sc->sc_cache = NULL; 455 456 if (sc->sc_indir_valid != NULL) 457 free(sc->sc_indir_valid, M_TEMP); 458 sc->sc_indir_valid = NULL; 459 460 if (sc->sc_indir_data != NULL) 461 free(sc->sc_indir_data, M_TEMP); 462 sc->sc_indir_data = NULL; 463 } 464 465 /* 466 * Check if an unmount is ok. If forced, set this snapshot into ERROR state. 467 */ 468 int 469 fss_umount_hook(struct mount *mp, int forced) 470 { 471 int i, s; 472 473 for (i = 0; i < NFSS; i++) { 474 FSS_LOCK(&fss_softc[i], s); 475 if ((fss_softc[i].sc_flags & FSS_ACTIVE) != 0 && 476 fss_softc[i].sc_mount == mp) { 477 if (forced) 478 fss_error(&fss_softc[i], "forced unmount"); 479 else { 480 FSS_UNLOCK(&fss_softc[i], s); 481 return EBUSY; 482 } 483 } 484 FSS_UNLOCK(&fss_softc[i], s); 485 } 486 487 return 0; 488 } 489 490 /* 491 * A buffer is written to the snapshotted block device. Copy to 492 * backing store if needed. 493 */ 494 static int 495 fss_copy_on_write(void *v, struct buf *bp) 496 { 497 int s; 498 u_int32_t cl, ch, c; 499 struct fss_softc *sc = v; 500 501 FSS_LOCK(sc, s); 502 if (!FSS_ISVALID(sc)) { 503 FSS_UNLOCK(sc, s); 504 return 0; 505 } 506 507 FSS_UNLOCK(sc, s); 508 509 FSS_STAT_INC(sc, cow_calls); 510 511 cl = FSS_BTOCL(sc, dbtob(bp->b_blkno)); 512 ch = FSS_BTOCL(sc, dbtob(bp->b_blkno)+bp->b_bcount-1); 513 514 for (c = cl; c <= ch; c++) 515 fss_read_cluster(sc, c); 516 517 return 0; 518 } 519 520 /* 521 * Lookup and open needed files. 522 * 523 * For file system internal snapshot initializes sc_mntname, sc_mount, 524 * sc_bs_vp and sc_time. 525 * 526 * Otherwise returns dev and size of the underlying block device. 527 * Initializes sc_mntname, sc_mount, sc_bdev, sc_bs_vp and sc_mount 528 */ 529 static int 530 fss_create_files(struct fss_softc *sc, struct fss_set *fss, 531 off_t *bsize, struct lwp *l) 532 { 533 int error, bits, fsbsize; 534 struct timespec ts; 535 struct partinfo dpart; 536 struct vattr va; 537 struct nameidata nd; 538 539 /* 540 * Get the mounted file system. 541 */ 542 543 NDINIT(&nd, LOOKUP, FOLLOW, UIO_USERSPACE, fss->fss_mount, l); 544 if ((error = namei(&nd)) != 0) 545 return error; 546 547 if ((nd.ni_vp->v_vflag & VV_ROOT) != VV_ROOT) { 548 vrele(nd.ni_vp); 549 return EINVAL; 550 } 551 552 sc->sc_mount = nd.ni_vp->v_mount; 553 memcpy(sc->sc_mntname, sc->sc_mount->mnt_stat.f_mntonname, MNAMELEN); 554 555 vrele(nd.ni_vp); 556 557 /* 558 * Check for file system internal snapshot. 559 */ 560 561 NDINIT(&nd, LOOKUP, FOLLOW, UIO_USERSPACE, fss->fss_bstore, l); 562 if ((error = namei(&nd)) != 0) 563 return error; 564 565 if (nd.ni_vp->v_type == VREG && nd.ni_vp->v_mount == sc->sc_mount) { 566 vrele(nd.ni_vp); 567 sc->sc_flags |= FSS_PERSISTENT; 568 569 NDINIT(&nd, LOOKUP, FOLLOW, UIO_USERSPACE, fss->fss_bstore, l); 570 if ((error = vn_open(&nd, FREAD, 0)) != 0) 571 return error; 572 sc->sc_bs_vp = nd.ni_vp; 573 574 fsbsize = sc->sc_bs_vp->v_mount->mnt_stat.f_iosize; 575 bits = sizeof(sc->sc_bs_bshift)*NBBY; 576 for (sc->sc_bs_bshift = 1; sc->sc_bs_bshift < bits; 577 sc->sc_bs_bshift++) 578 if (FSS_FSBSIZE(sc) == fsbsize) 579 break; 580 if (sc->sc_bs_bshift >= bits) { 581 VOP_UNLOCK(sc->sc_bs_vp, 0); 582 return EINVAL; 583 } 584 585 sc->sc_bs_bmask = FSS_FSBSIZE(sc)-1; 586 sc->sc_clshift = 0; 587 588 error = VFS_SNAPSHOT(sc->sc_mount, sc->sc_bs_vp, &ts); 589 TIMESPEC_TO_TIMEVAL(&sc->sc_time, &ts); 590 591 VOP_UNLOCK(sc->sc_bs_vp, 0); 592 593 return error; 594 } 595 vrele(nd.ni_vp); 596 597 /* 598 * Get the block device it is mounted on. 599 */ 600 601 NDINIT(&nd, LOOKUP, FOLLOW, UIO_SYSSPACE, 602 sc->sc_mount->mnt_stat.f_mntfromname, l); 603 if ((error = namei(&nd)) != 0) 604 return error; 605 606 if (nd.ni_vp->v_type != VBLK) { 607 vrele(nd.ni_vp); 608 return EINVAL; 609 } 610 611 error = VOP_IOCTL(nd.ni_vp, DIOCGPART, &dpart, FREAD, 612 l->l_cred, l); 613 if (error) { 614 vrele(nd.ni_vp); 615 return error; 616 } 617 618 sc->sc_bdev = nd.ni_vp->v_rdev; 619 *bsize = (off_t)dpart.disklab->d_secsize*dpart.part->p_size; 620 vrele(nd.ni_vp); 621 622 /* 623 * Get the backing store 624 */ 625 626 NDINIT(&nd, LOOKUP, FOLLOW, UIO_USERSPACE, fss->fss_bstore, l); 627 if ((error = vn_open(&nd, FREAD|FWRITE, 0)) != 0) 628 return error; 629 VOP_UNLOCK(nd.ni_vp, 0); 630 631 sc->sc_bs_vp = nd.ni_vp; 632 633 if (nd.ni_vp->v_type != VREG && nd.ni_vp->v_type != VCHR) 634 return EINVAL; 635 636 if (sc->sc_bs_vp->v_type == VREG) { 637 error = VOP_GETATTR(sc->sc_bs_vp, &va, l->l_cred, l); 638 if (error != 0) 639 return error; 640 sc->sc_bs_size = va.va_size; 641 fsbsize = sc->sc_bs_vp->v_mount->mnt_stat.f_iosize; 642 if (fsbsize & (fsbsize-1)) /* No power of two */ 643 return EINVAL; 644 for (sc->sc_bs_bshift = 1; sc->sc_bs_bshift < 32; 645 sc->sc_bs_bshift++) 646 if (FSS_FSBSIZE(sc) == fsbsize) 647 break; 648 if (sc->sc_bs_bshift >= 32) 649 return EINVAL; 650 sc->sc_bs_bmask = FSS_FSBSIZE(sc)-1; 651 } else { 652 sc->sc_bs_bshift = DEV_BSHIFT; 653 sc->sc_bs_bmask = FSS_FSBSIZE(sc)-1; 654 } 655 656 /* 657 * As all IO to from/to the backing store goes through 658 * VOP_STRATEGY() clean the buffer cache to prevent 659 * cache incoherencies. 660 */ 661 if ((error = vinvalbuf(sc->sc_bs_vp, V_SAVE, l->l_cred, l, 0, 0)) != 0) 662 return error; 663 664 return 0; 665 } 666 667 /* 668 * Create a snapshot. 669 */ 670 static int 671 fss_create_snapshot(struct fss_softc *sc, struct fss_set *fss, struct lwp *l) 672 { 673 int len, error; 674 u_int32_t csize; 675 off_t bsize; 676 677 bsize = 0; /* XXX gcc */ 678 679 /* 680 * Open needed files. 681 */ 682 if ((error = fss_create_files(sc, fss, &bsize, l)) != 0) 683 goto bad; 684 685 if (sc->sc_flags & FSS_PERSISTENT) { 686 fss_softc_alloc(sc); 687 sc->sc_flags |= FSS_ACTIVE; 688 return 0; 689 } 690 691 /* 692 * Set cluster size. Must be a power of two and 693 * a multiple of backing store block size. 694 */ 695 if (fss->fss_csize <= 0) 696 csize = MAXPHYS; 697 else 698 csize = fss->fss_csize; 699 if (bsize/csize > FSS_CLUSTER_MAX) 700 csize = bsize/FSS_CLUSTER_MAX+1; 701 702 for (sc->sc_clshift = sc->sc_bs_bshift; sc->sc_clshift < 32; 703 sc->sc_clshift++) 704 if (FSS_CLSIZE(sc) >= csize) 705 break; 706 if (sc->sc_clshift >= 32) { 707 error = EINVAL; 708 goto bad; 709 } 710 sc->sc_clmask = FSS_CLSIZE(sc)-1; 711 712 /* 713 * Set number of cache slots. 714 */ 715 if (FSS_CLSIZE(sc) <= 8192) 716 sc->sc_cache_size = 32; 717 else if (FSS_CLSIZE(sc) <= 65536) 718 sc->sc_cache_size = 8; 719 else 720 sc->sc_cache_size = 4; 721 722 /* 723 * Set number of clusters and size of last cluster. 724 */ 725 sc->sc_clcount = FSS_BTOCL(sc, bsize-1)+1; 726 sc->sc_clresid = FSS_CLOFF(sc, bsize-1)+1; 727 728 /* 729 * Set size of indirect table. 730 */ 731 len = sc->sc_clcount*sizeof(u_int32_t); 732 sc->sc_indir_size = FSS_BTOCL(sc, len)+1; 733 sc->sc_clnext = sc->sc_indir_size; 734 sc->sc_indir_cur = 0; 735 736 if ((error = fss_softc_alloc(sc)) != 0) 737 goto bad; 738 739 /* 740 * Activate the snapshot. 741 */ 742 743 if ((error = vfs_suspend(sc->sc_mount, 0)) != 0) 744 goto bad; 745 746 microtime(&sc->sc_time); 747 748 if (error == 0) 749 error = fscow_establish(sc->sc_mount, 750 fss_copy_on_write, sc); 751 if (error == 0) 752 sc->sc_flags |= FSS_ACTIVE; 753 754 vfs_resume(sc->sc_mount); 755 756 if (error != 0) 757 goto bad; 758 759 #ifdef DEBUG 760 printf("fss%d: %s snapshot active\n", sc->sc_unit, sc->sc_mntname); 761 printf("fss%d: %u clusters of %u, %u cache slots, %u indir clusters\n", 762 sc->sc_unit, sc->sc_clcount, FSS_CLSIZE(sc), 763 sc->sc_cache_size, sc->sc_indir_size); 764 #endif 765 766 return 0; 767 768 bad: 769 fss_softc_free(sc); 770 if (sc->sc_bs_vp != NULL) { 771 if (sc->sc_flags & FSS_PERSISTENT) 772 vn_close(sc->sc_bs_vp, FREAD, l->l_cred, l); 773 else 774 vn_close(sc->sc_bs_vp, FREAD|FWRITE, l->l_cred, l); 775 } 776 sc->sc_bs_vp = NULL; 777 778 return error; 779 } 780 781 /* 782 * Delete a snapshot. 783 */ 784 static int 785 fss_delete_snapshot(struct fss_softc *sc, struct lwp *l) 786 { 787 int s; 788 789 if ((sc->sc_flags & FSS_PERSISTENT) == 0) 790 fscow_disestablish(sc->sc_mount, fss_copy_on_write, sc); 791 792 FSS_LOCK(sc, s); 793 sc->sc_flags &= ~(FSS_ACTIVE|FSS_ERROR); 794 sc->sc_mount = NULL; 795 sc->sc_bdev = NODEV; 796 FSS_UNLOCK(sc, s); 797 798 fss_softc_free(sc); 799 if (sc->sc_flags & FSS_PERSISTENT) 800 vn_close(sc->sc_bs_vp, FREAD, l->l_cred, l); 801 else 802 vn_close(sc->sc_bs_vp, FREAD|FWRITE, l->l_cred, l); 803 sc->sc_bs_vp = NULL; 804 sc->sc_flags &= ~FSS_PERSISTENT; 805 806 FSS_STAT_CLEAR(sc); 807 808 return 0; 809 } 810 811 /* 812 * A read from the snapshotted block device has completed. 813 */ 814 static void 815 fss_cluster_iodone(struct buf *bp) 816 { 817 int s; 818 struct fss_cache *scp = bp->b_private; 819 820 KASSERT(bp->b_vp == NULL); 821 822 FSS_LOCK(scp->fc_softc, s); 823 824 if (bp->b_error != 0) 825 fss_error(scp->fc_softc, "fs read error %d", bp->b_error); 826 827 if (--scp->fc_xfercount == 0) 828 wakeup(&scp->fc_data); 829 830 FSS_UNLOCK(scp->fc_softc, s); 831 832 putiobuf(bp); 833 } 834 835 /* 836 * Read a cluster from the snapshotted block device to the cache. 837 */ 838 static void 839 fss_read_cluster(struct fss_softc *sc, u_int32_t cl) 840 { 841 int s, todo, len; 842 char *addr; 843 daddr_t dblk; 844 struct buf *bp; 845 struct fss_cache *scp, *scl; 846 847 /* 848 * Get a free cache slot. 849 */ 850 scl = sc->sc_cache+sc->sc_cache_size; 851 852 FSS_LOCK(sc, s); 853 854 restart: 855 if (isset(sc->sc_copied, cl) || !FSS_ISVALID(sc)) { 856 FSS_UNLOCK(sc, s); 857 return; 858 } 859 860 for (scp = sc->sc_cache; scp < scl; scp++) 861 if (scp->fc_type != FSS_CACHE_FREE && 862 scp->fc_cluster == cl) { 863 ltsleep(&scp->fc_type, PRIBIO, "cowwait2", 0, 864 &sc->sc_slock); 865 goto restart; 866 } 867 868 for (scp = sc->sc_cache; scp < scl; scp++) 869 if (scp->fc_type == FSS_CACHE_FREE) { 870 scp->fc_type = FSS_CACHE_BUSY; 871 scp->fc_cluster = cl; 872 break; 873 } 874 if (scp >= scl) { 875 FSS_STAT_INC(sc, cow_cache_full); 876 ltsleep(&sc->sc_cache, PRIBIO, "cowwait3", 0, &sc->sc_slock); 877 goto restart; 878 } 879 880 FSS_UNLOCK(sc, s); 881 882 /* 883 * Start the read. 884 */ 885 FSS_STAT_INC(sc, cow_copied); 886 887 dblk = btodb(FSS_CLTOB(sc, cl)); 888 addr = scp->fc_data; 889 if (cl == sc->sc_clcount-1) { 890 todo = sc->sc_clresid; 891 memset((char *)addr + todo, 0, FSS_CLSIZE(sc) - todo); 892 } else 893 todo = FSS_CLSIZE(sc); 894 while (todo > 0) { 895 len = todo; 896 if (len > MAXPHYS) 897 len = MAXPHYS; 898 899 bp = getiobuf(); 900 bp->b_flags = B_READ|B_CALL; 901 bp->b_bcount = len; 902 bp->b_bufsize = bp->b_bcount; 903 bp->b_error = 0; 904 bp->b_data = addr; 905 bp->b_blkno = dblk; 906 bp->b_proc = NULL; 907 bp->b_dev = sc->sc_bdev; 908 bp->b_vp = NULLVP; 909 bp->b_private = scp; 910 bp->b_iodone = fss_cluster_iodone; 911 912 bdev_strategy(bp); 913 914 FSS_LOCK(sc, s); 915 scp->fc_xfercount++; 916 FSS_UNLOCK(sc, s); 917 918 dblk += btodb(len); 919 addr += len; 920 todo -= len; 921 } 922 923 /* 924 * Wait for all read requests to complete. 925 */ 926 FSS_LOCK(sc, s); 927 while (scp->fc_xfercount > 0) 928 ltsleep(&scp->fc_data, PRIBIO, "cowwait", 0, &sc->sc_slock); 929 930 scp->fc_type = FSS_CACHE_VALID; 931 setbit(sc->sc_copied, scp->fc_cluster); 932 FSS_UNLOCK(sc, s); 933 934 wakeup(&sc->sc_bs_lwp); 935 } 936 937 /* 938 * Read/write clusters from/to backing store. 939 * For persistent snapshots must be called with cl == 0. off is the 940 * offset into the snapshot. 941 */ 942 static int 943 fss_bs_io(struct fss_softc *sc, fss_io_type rw, 944 u_int32_t cl, off_t off, int len, void *data) 945 { 946 int error; 947 948 off += FSS_CLTOB(sc, cl); 949 950 vn_lock(sc->sc_bs_vp, LK_EXCLUSIVE|LK_RETRY); 951 952 error = vn_rdwr((rw == FSS_READ ? UIO_READ : UIO_WRITE), sc->sc_bs_vp, 953 data, len, off, UIO_SYSSPACE, IO_UNIT|IO_NODELOCKED, 954 sc->sc_bs_lwp->l_cred, NULL, NULL); 955 if (error == 0) { 956 simple_lock(&sc->sc_bs_vp->v_interlock); 957 error = VOP_PUTPAGES(sc->sc_bs_vp, trunc_page(off), 958 round_page(off+len), PGO_CLEANIT|PGO_SYNCIO|PGO_FREE); 959 } 960 961 VOP_UNLOCK(sc->sc_bs_vp, 0); 962 963 return error; 964 } 965 966 /* 967 * Get a pointer to the indirect slot for this cluster. 968 */ 969 static u_int32_t * 970 fss_bs_indir(struct fss_softc *sc, u_int32_t cl) 971 { 972 u_int32_t icl; 973 int ioff; 974 975 icl = cl/(FSS_CLSIZE(sc)/sizeof(u_int32_t)); 976 ioff = cl%(FSS_CLSIZE(sc)/sizeof(u_int32_t)); 977 978 if (sc->sc_indir_cur == icl) 979 return &sc->sc_indir_data[ioff]; 980 981 if (sc->sc_indir_dirty) { 982 FSS_STAT_INC(sc, indir_write); 983 if (fss_bs_io(sc, FSS_WRITE, sc->sc_indir_cur, 0, 984 FSS_CLSIZE(sc), (void *)sc->sc_indir_data) != 0) 985 return NULL; 986 setbit(sc->sc_indir_valid, sc->sc_indir_cur); 987 } 988 989 sc->sc_indir_dirty = 0; 990 sc->sc_indir_cur = icl; 991 992 if (isset(sc->sc_indir_valid, sc->sc_indir_cur)) { 993 FSS_STAT_INC(sc, indir_read); 994 if (fss_bs_io(sc, FSS_READ, sc->sc_indir_cur, 0, 995 FSS_CLSIZE(sc), (void *)sc->sc_indir_data) != 0) 996 return NULL; 997 } else 998 memset(sc->sc_indir_data, 0, FSS_CLSIZE(sc)); 999 1000 return &sc->sc_indir_data[ioff]; 1001 } 1002 1003 /* 1004 * The kernel thread (one for every active snapshot). 1005 * 1006 * After wakeup it cleans the cache and runs the I/O requests. 1007 */ 1008 static void 1009 fss_bs_thread(void *arg) 1010 { 1011 int error, len, nfreed, nio, s; 1012 long off; 1013 char *addr; 1014 u_int32_t c, cl, ch, *indirp; 1015 struct buf *bp, *nbp; 1016 struct fss_softc *sc; 1017 struct fss_cache *scp, *scl; 1018 1019 sc = arg; 1020 1021 scl = sc->sc_cache+sc->sc_cache_size; 1022 1023 nbp = getiobuf(); 1024 1025 nfreed = nio = 1; /* Dont sleep the first time */ 1026 1027 FSS_LOCK(sc, s); 1028 1029 for (;;) { 1030 if (nfreed == 0 && nio == 0) 1031 ltsleep(&sc->sc_bs_lwp, PVM-1, "fssbs", 0, 1032 &sc->sc_slock); 1033 1034 if ((sc->sc_flags & FSS_BS_THREAD) == 0) { 1035 sc->sc_bs_lwp = NULL; 1036 wakeup(&sc->sc_bs_lwp); 1037 1038 FSS_UNLOCK(sc, s); 1039 1040 putiobuf(nbp); 1041 #ifdef FSS_STATISTICS 1042 if ((sc->sc_flags & FSS_PERSISTENT) == 0) { 1043 printf("fss%d: cow called %" PRId64 " times," 1044 " copied %" PRId64 " clusters," 1045 " cache full %" PRId64 " times\n", 1046 sc->sc_unit, 1047 FSS_STAT_VAL(sc, cow_calls), 1048 FSS_STAT_VAL(sc, cow_copied), 1049 FSS_STAT_VAL(sc, cow_cache_full)); 1050 printf("fss%d: %" PRId64 " indir reads," 1051 " %" PRId64 " indir writes\n", 1052 sc->sc_unit, 1053 FSS_STAT_VAL(sc, indir_read), 1054 FSS_STAT_VAL(sc, indir_write)); 1055 } 1056 #endif /* FSS_STATISTICS */ 1057 kthread_exit(0); 1058 } 1059 1060 /* 1061 * Process I/O requests (persistent) 1062 */ 1063 1064 if (sc->sc_flags & FSS_PERSISTENT) { 1065 nfreed = nio = 0; 1066 1067 if ((bp = BUFQ_GET(sc->sc_bufq)) == NULL) 1068 continue; 1069 1070 nio++; 1071 1072 if (FSS_ISVALID(sc)) { 1073 FSS_UNLOCK(sc, s); 1074 1075 error = fss_bs_io(sc, FSS_READ, 0, 1076 dbtob(bp->b_blkno), bp->b_bcount, 1077 bp->b_data); 1078 1079 FSS_LOCK(sc, s); 1080 } else 1081 error = ENXIO; 1082 1083 if (error) { 1084 bp->b_error = error; 1085 bp->b_resid = bp->b_bcount; 1086 } else 1087 bp->b_resid = 0; 1088 1089 biodone(bp); 1090 1091 continue; 1092 } 1093 1094 /* 1095 * Clean the cache 1096 */ 1097 nfreed = 0; 1098 for (scp = sc->sc_cache; scp < scl; scp++) { 1099 if (scp->fc_type != FSS_CACHE_VALID) 1100 continue; 1101 1102 FSS_UNLOCK(sc, s); 1103 1104 indirp = fss_bs_indir(sc, scp->fc_cluster); 1105 if (indirp != NULL) { 1106 error = fss_bs_io(sc, FSS_WRITE, sc->sc_clnext, 1107 0, FSS_CLSIZE(sc), scp->fc_data); 1108 } else 1109 error = EIO; 1110 1111 FSS_LOCK(sc, s); 1112 1113 if (error == 0) { 1114 *indirp = sc->sc_clnext++; 1115 sc->sc_indir_dirty = 1; 1116 } else 1117 fss_error(sc, "write bs error %d", error); 1118 1119 scp->fc_type = FSS_CACHE_FREE; 1120 nfreed++; 1121 wakeup(&scp->fc_type); 1122 } 1123 1124 if (nfreed) 1125 wakeup(&sc->sc_cache); 1126 1127 /* 1128 * Process I/O requests 1129 */ 1130 nio = 0; 1131 1132 if ((bp = BUFQ_GET(sc->sc_bufq)) == NULL) 1133 continue; 1134 1135 nio++; 1136 1137 if (!FSS_ISVALID(sc)) { 1138 bp->b_error = ENXIO; 1139 bp->b_resid = bp->b_bcount; 1140 biodone(bp); 1141 continue; 1142 } 1143 1144 /* 1145 * First read from the snapshotted block device. 1146 * XXX Split to only read those parts that have not 1147 * been saved to backing store? 1148 */ 1149 1150 FSS_UNLOCK(sc, s); 1151 1152 BUF_INIT(nbp); 1153 nbp->b_flags = B_READ; 1154 nbp->b_bcount = bp->b_bcount; 1155 nbp->b_bufsize = bp->b_bcount; 1156 nbp->b_error = 0; 1157 nbp->b_data = bp->b_data; 1158 nbp->b_blkno = bp->b_blkno; 1159 nbp->b_proc = bp->b_proc; 1160 nbp->b_dev = sc->sc_bdev; 1161 nbp->b_vp = NULLVP; 1162 1163 bdev_strategy(nbp); 1164 1165 if (biowait(nbp) != 0) { 1166 bp->b_resid = bp->b_bcount; 1167 bp->b_error = nbp->b_error; 1168 biodone(bp); 1169 FSS_LOCK(sc, s); 1170 continue; 1171 } 1172 1173 cl = FSS_BTOCL(sc, dbtob(bp->b_blkno)); 1174 off = FSS_CLOFF(sc, dbtob(bp->b_blkno)); 1175 ch = FSS_BTOCL(sc, dbtob(bp->b_blkno)+bp->b_bcount-1); 1176 bp->b_resid = bp->b_bcount; 1177 addr = bp->b_data; 1178 1179 FSS_LOCK(sc, s); 1180 1181 /* 1182 * Replace those parts that have been saved to backing store. 1183 */ 1184 1185 for (c = cl; c <= ch; 1186 c++, off = 0, bp->b_resid -= len, addr += len) { 1187 len = FSS_CLSIZE(sc)-off; 1188 if (len > bp->b_resid) 1189 len = bp->b_resid; 1190 1191 if (isclr(sc->sc_copied, c)) 1192 continue; 1193 1194 FSS_UNLOCK(sc, s); 1195 1196 indirp = fss_bs_indir(sc, c); 1197 1198 FSS_LOCK(sc, s); 1199 1200 if (indirp == NULL || *indirp == 0) { 1201 /* 1202 * Not on backing store. Either in cache 1203 * or hole in the snapshotted block device. 1204 */ 1205 for (scp = sc->sc_cache; scp < scl; scp++) 1206 if (scp->fc_type == FSS_CACHE_VALID && 1207 scp->fc_cluster == c) 1208 break; 1209 if (scp < scl) 1210 memcpy(addr, (char *)scp->fc_data+off, len); 1211 else 1212 memset(addr, 0, len); 1213 continue; 1214 } 1215 /* 1216 * Read from backing store. 1217 */ 1218 1219 FSS_UNLOCK(sc, s); 1220 1221 if ((error = fss_bs_io(sc, FSS_READ, *indirp, 1222 off, len, addr)) != 0) { 1223 bp->b_resid = bp->b_bcount; 1224 bp->b_error = error; 1225 FSS_LOCK(sc, s); 1226 break; 1227 } 1228 1229 FSS_LOCK(sc, s); 1230 1231 } 1232 1233 biodone(bp); 1234 } 1235 } 1236