vfs_bio.c revision 1.242
1 /* $NetBSD: vfs_bio.c,v 1.242 2012/12/30 09:19:24 hannken Exp $ */ 2 3 /*- 4 * Copyright (c) 2007, 2008, 2009 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Andrew Doran, and by Wasabi Systems, Inc. 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 * 19 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 20 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 21 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 22 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 23 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 24 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 25 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 26 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 27 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 28 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 29 * POSSIBILITY OF SUCH DAMAGE. 30 */ 31 32 /*- 33 * Copyright (c) 1982, 1986, 1989, 1993 34 * The Regents of the University of California. All rights reserved. 35 * (c) UNIX System Laboratories, Inc. 36 * All or some portions of this file are derived from material licensed 37 * to the University of California by American Telephone and Telegraph 38 * Co. or Unix System Laboratories, Inc. and are reproduced herein with 39 * the permission of UNIX System Laboratories, Inc. 40 * 41 * Redistribution and use in source and binary forms, with or without 42 * modification, are permitted provided that the following conditions 43 * are met: 44 * 1. Redistributions of source code must retain the above copyright 45 * notice, this list of conditions and the following disclaimer. 46 * 2. Redistributions in binary form must reproduce the above copyright 47 * notice, this list of conditions and the following disclaimer in the 48 * documentation and/or other materials provided with the distribution. 49 * 3. Neither the name of the University nor the names of its contributors 50 * may be used to endorse or promote products derived from this software 51 * without specific prior written permission. 52 * 53 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 54 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 55 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 56 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 57 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 58 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 59 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 60 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 61 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 62 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 63 * SUCH DAMAGE. 64 * 65 * @(#)vfs_bio.c 8.6 (Berkeley) 1/11/94 66 */ 67 68 /*- 69 * Copyright (c) 1994 Christopher G. Demetriou 70 * 71 * Redistribution and use in source and binary forms, with or without 72 * modification, are permitted provided that the following conditions 73 * are met: 74 * 1. Redistributions of source code must retain the above copyright 75 * notice, this list of conditions and the following disclaimer. 76 * 2. Redistributions in binary form must reproduce the above copyright 77 * notice, this list of conditions and the following disclaimer in the 78 * documentation and/or other materials provided with the distribution. 79 * 3. All advertising materials mentioning features or use of this software 80 * must display the following acknowledgement: 81 * This product includes software developed by the University of 82 * California, Berkeley and its contributors. 83 * 4. Neither the name of the University nor the names of its contributors 84 * may be used to endorse or promote products derived from this software 85 * without specific prior written permission. 86 * 87 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 88 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 89 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 90 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 91 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 92 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 93 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 94 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 95 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 96 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 97 * SUCH DAMAGE. 98 * 99 * @(#)vfs_bio.c 8.6 (Berkeley) 1/11/94 100 */ 101 102 /* 103 * The buffer cache subsystem. 104 * 105 * Some references: 106 * Bach: The Design of the UNIX Operating System (Prentice Hall, 1986) 107 * Leffler, et al.: The Design and Implementation of the 4.3BSD 108 * UNIX Operating System (Addison Welley, 1989) 109 * 110 * Locking 111 * 112 * There are three locks: 113 * - bufcache_lock: protects global buffer cache state. 114 * - BC_BUSY: a long term per-buffer lock. 115 * - buf_t::b_objlock: lock on completion (biowait vs biodone). 116 * 117 * For buffers associated with vnodes (a most common case) b_objlock points 118 * to the vnode_t::v_interlock. Otherwise, it points to generic buffer_lock. 119 * 120 * Lock order: 121 * bufcache_lock -> 122 * buf_t::b_objlock 123 */ 124 125 #include <sys/cdefs.h> 126 __KERNEL_RCSID(0, "$NetBSD: vfs_bio.c,v 1.242 2012/12/30 09:19:24 hannken Exp $"); 127 128 #include "opt_bufcache.h" 129 130 #include <sys/param.h> 131 #include <sys/systm.h> 132 #include <sys/kernel.h> 133 #include <sys/proc.h> 134 #include <sys/buf.h> 135 #include <sys/vnode.h> 136 #include <sys/mount.h> 137 #include <sys/resourcevar.h> 138 #include <sys/sysctl.h> 139 #include <sys/conf.h> 140 #include <sys/kauth.h> 141 #include <sys/fstrans.h> 142 #include <sys/intr.h> 143 #include <sys/cpu.h> 144 #include <sys/wapbl.h> 145 #include <sys/bitops.h> 146 147 #include <uvm/uvm.h> /* extern struct uvm uvm */ 148 149 #include <miscfs/specfs/specdev.h> 150 151 #ifndef BUFPAGES 152 # define BUFPAGES 0 153 #endif 154 155 #ifdef BUFCACHE 156 # if (BUFCACHE < 5) || (BUFCACHE > 95) 157 # error BUFCACHE is not between 5 and 95 158 # endif 159 #else 160 # define BUFCACHE 15 161 #endif 162 163 u_int nbuf; /* desired number of buffer headers */ 164 u_int bufpages = BUFPAGES; /* optional hardwired count */ 165 u_int bufcache = BUFCACHE; /* max % of RAM to use for buffer cache */ 166 167 /* Function prototypes */ 168 struct bqueue; 169 170 static void buf_setwm(void); 171 static int buf_trim(void); 172 static void *bufpool_page_alloc(struct pool *, int); 173 static void bufpool_page_free(struct pool *, void *); 174 static buf_t *bio_doread(struct vnode *, daddr_t, int, 175 kauth_cred_t, int); 176 static buf_t *getnewbuf(int, int, int); 177 static int buf_lotsfree(void); 178 static int buf_canrelease(void); 179 static u_long buf_mempoolidx(u_long); 180 static u_long buf_roundsize(u_long); 181 static void *buf_alloc(size_t); 182 static void buf_mrelease(void *, size_t); 183 static void binsheadfree(buf_t *, struct bqueue *); 184 static void binstailfree(buf_t *, struct bqueue *); 185 #ifdef DEBUG 186 static int checkfreelist(buf_t *, struct bqueue *, int); 187 #endif 188 static void biointr(void *); 189 static void biodone2(buf_t *); 190 static void bref(buf_t *); 191 static void brele(buf_t *); 192 static void sysctl_kern_buf_setup(void); 193 static void sysctl_vm_buf_setup(void); 194 195 /* 196 * Definitions for the buffer hash lists. 197 */ 198 #define BUFHASH(dvp, lbn) \ 199 (&bufhashtbl[(((long)(dvp) >> 8) + (int)(lbn)) & bufhash]) 200 LIST_HEAD(bufhashhdr, buf) *bufhashtbl, invalhash; 201 u_long bufhash; 202 struct bqueue bufqueues[BQUEUES]; 203 204 static kcondvar_t needbuffer_cv; 205 206 /* 207 * Buffer queue lock. 208 */ 209 kmutex_t bufcache_lock; 210 kmutex_t buffer_lock; 211 212 /* Software ISR for completed transfers. */ 213 static void *biodone_sih; 214 215 /* Buffer pool for I/O buffers. */ 216 static pool_cache_t buf_cache; 217 static pool_cache_t bufio_cache; 218 219 #define MEMPOOL_INDEX_OFFSET (ilog2(DEV_BSIZE)) /* smallest pool is 512 bytes */ 220 #define NMEMPOOLS (ilog2(MAXBSIZE) - MEMPOOL_INDEX_OFFSET + 1) 221 __CTASSERT((1 << (NMEMPOOLS + MEMPOOL_INDEX_OFFSET - 1)) == MAXBSIZE); 222 223 /* Buffer memory pools */ 224 static struct pool bmempools[NMEMPOOLS]; 225 226 static struct vm_map *buf_map; 227 228 /* 229 * Buffer memory pool allocator. 230 */ 231 static void * 232 bufpool_page_alloc(struct pool *pp, int flags) 233 { 234 235 return (void *)uvm_km_alloc(buf_map, 236 MAXBSIZE, MAXBSIZE, 237 ((flags & PR_WAITOK) ? 0 : UVM_KMF_NOWAIT|UVM_KMF_TRYLOCK) 238 | UVM_KMF_WIRED); 239 } 240 241 static void 242 bufpool_page_free(struct pool *pp, void *v) 243 { 244 245 uvm_km_free(buf_map, (vaddr_t)v, MAXBSIZE, UVM_KMF_WIRED); 246 } 247 248 static struct pool_allocator bufmempool_allocator = { 249 .pa_alloc = bufpool_page_alloc, 250 .pa_free = bufpool_page_free, 251 .pa_pagesz = MAXBSIZE, 252 }; 253 254 /* Buffer memory management variables */ 255 u_long bufmem_valimit; 256 u_long bufmem_hiwater; 257 u_long bufmem_lowater; 258 u_long bufmem; 259 260 /* 261 * MD code can call this to set a hard limit on the amount 262 * of virtual memory used by the buffer cache. 263 */ 264 int 265 buf_setvalimit(vsize_t sz) 266 { 267 268 /* We need to accommodate at least NMEMPOOLS of MAXBSIZE each */ 269 if (sz < NMEMPOOLS * MAXBSIZE) 270 return EINVAL; 271 272 bufmem_valimit = sz; 273 return 0; 274 } 275 276 static void 277 buf_setwm(void) 278 { 279 280 bufmem_hiwater = buf_memcalc(); 281 /* lowater is approx. 2% of memory (with bufcache = 15) */ 282 #define BUFMEM_WMSHIFT 3 283 #define BUFMEM_HIWMMIN (64 * 1024 << BUFMEM_WMSHIFT) 284 if (bufmem_hiwater < BUFMEM_HIWMMIN) 285 /* Ensure a reasonable minimum value */ 286 bufmem_hiwater = BUFMEM_HIWMMIN; 287 bufmem_lowater = bufmem_hiwater >> BUFMEM_WMSHIFT; 288 } 289 290 #ifdef DEBUG 291 int debug_verify_freelist = 0; 292 static int 293 checkfreelist(buf_t *bp, struct bqueue *dp, int ison) 294 { 295 buf_t *b; 296 297 if (!debug_verify_freelist) 298 return 1; 299 300 TAILQ_FOREACH(b, &dp->bq_queue, b_freelist) { 301 if (b == bp) 302 return ison ? 1 : 0; 303 } 304 305 return ison ? 0 : 1; 306 } 307 #endif 308 309 /* 310 * Insq/Remq for the buffer hash lists. 311 * Call with buffer queue locked. 312 */ 313 static void 314 binsheadfree(buf_t *bp, struct bqueue *dp) 315 { 316 317 KASSERT(mutex_owned(&bufcache_lock)); 318 KASSERT(bp->b_freelistindex == -1); 319 TAILQ_INSERT_HEAD(&dp->bq_queue, bp, b_freelist); 320 dp->bq_bytes += bp->b_bufsize; 321 bp->b_freelistindex = dp - bufqueues; 322 } 323 324 static void 325 binstailfree(buf_t *bp, struct bqueue *dp) 326 { 327 328 KASSERT(mutex_owned(&bufcache_lock)); 329 KASSERT(bp->b_freelistindex == -1); 330 TAILQ_INSERT_TAIL(&dp->bq_queue, bp, b_freelist); 331 dp->bq_bytes += bp->b_bufsize; 332 bp->b_freelistindex = dp - bufqueues; 333 } 334 335 void 336 bremfree(buf_t *bp) 337 { 338 struct bqueue *dp; 339 int bqidx = bp->b_freelistindex; 340 341 KASSERT(mutex_owned(&bufcache_lock)); 342 343 KASSERT(bqidx != -1); 344 dp = &bufqueues[bqidx]; 345 KDASSERT(checkfreelist(bp, dp, 1)); 346 KASSERT(dp->bq_bytes >= bp->b_bufsize); 347 TAILQ_REMOVE(&dp->bq_queue, bp, b_freelist); 348 dp->bq_bytes -= bp->b_bufsize; 349 350 /* For the sysctl helper. */ 351 if (bp == dp->bq_marker) 352 dp->bq_marker = NULL; 353 354 #if defined(DIAGNOSTIC) 355 bp->b_freelistindex = -1; 356 #endif /* defined(DIAGNOSTIC) */ 357 } 358 359 /* 360 * Add a reference to an buffer structure that came from buf_cache. 361 */ 362 static inline void 363 bref(buf_t *bp) 364 { 365 366 KASSERT(mutex_owned(&bufcache_lock)); 367 KASSERT(bp->b_refcnt > 0); 368 369 bp->b_refcnt++; 370 } 371 372 /* 373 * Free an unused buffer structure that came from buf_cache. 374 */ 375 static inline void 376 brele(buf_t *bp) 377 { 378 379 KASSERT(mutex_owned(&bufcache_lock)); 380 KASSERT(bp->b_refcnt > 0); 381 382 if (bp->b_refcnt-- == 1) { 383 buf_destroy(bp); 384 #ifdef DEBUG 385 memset((char *)bp, 0, sizeof(*bp)); 386 #endif 387 pool_cache_put(buf_cache, bp); 388 } 389 } 390 391 /* 392 * note that for some ports this is used by pmap bootstrap code to 393 * determine kva size. 394 */ 395 u_long 396 buf_memcalc(void) 397 { 398 u_long n; 399 400 /* 401 * Determine the upper bound of memory to use for buffers. 402 * 403 * - If bufpages is specified, use that as the number 404 * pages. 405 * 406 * - Otherwise, use bufcache as the percentage of 407 * physical memory. 408 */ 409 if (bufpages != 0) { 410 n = bufpages; 411 } else { 412 if (bufcache < 5) { 413 printf("forcing bufcache %d -> 5", bufcache); 414 bufcache = 5; 415 } 416 if (bufcache > 95) { 417 printf("forcing bufcache %d -> 95", bufcache); 418 bufcache = 95; 419 } 420 n = calc_cache_size(buf_map, bufcache, 421 (buf_map != kernel_map) ? 100 : BUFCACHE_VA_MAXPCT) 422 / PAGE_SIZE; 423 } 424 425 n <<= PAGE_SHIFT; 426 if (bufmem_valimit != 0 && n > bufmem_valimit) 427 n = bufmem_valimit; 428 429 return (n); 430 } 431 432 /* 433 * Initialize buffers and hash links for buffers. 434 */ 435 void 436 bufinit(void) 437 { 438 struct bqueue *dp; 439 int use_std; 440 u_int i; 441 442 mutex_init(&bufcache_lock, MUTEX_DEFAULT, IPL_NONE); 443 mutex_init(&buffer_lock, MUTEX_DEFAULT, IPL_NONE); 444 cv_init(&needbuffer_cv, "needbuf"); 445 446 if (bufmem_valimit != 0) { 447 vaddr_t minaddr = 0, maxaddr; 448 buf_map = uvm_km_suballoc(kernel_map, &minaddr, &maxaddr, 449 bufmem_valimit, 0, false, 0); 450 if (buf_map == NULL) 451 panic("bufinit: cannot allocate submap"); 452 } else 453 buf_map = kernel_map; 454 455 /* 456 * Initialize buffer cache memory parameters. 457 */ 458 bufmem = 0; 459 buf_setwm(); 460 461 /* On "small" machines use small pool page sizes where possible */ 462 use_std = (physmem < atop(16*1024*1024)); 463 464 /* 465 * Also use them on systems that can map the pool pages using 466 * a direct-mapped segment. 467 */ 468 #ifdef PMAP_MAP_POOLPAGE 469 use_std = 1; 470 #endif 471 472 buf_cache = pool_cache_init(sizeof(buf_t), 0, 0, 0, 473 "bufpl", NULL, IPL_SOFTBIO, NULL, NULL, NULL); 474 bufio_cache = pool_cache_init(sizeof(buf_t), 0, 0, 0, 475 "biopl", NULL, IPL_BIO, NULL, NULL, NULL); 476 477 for (i = 0; i < NMEMPOOLS; i++) { 478 struct pool_allocator *pa; 479 struct pool *pp = &bmempools[i]; 480 u_int size = 1 << (i + MEMPOOL_INDEX_OFFSET); 481 char *name = kmem_alloc(8, KM_SLEEP); /* XXX: never freed */ 482 if (__predict_false(size >= 1048576)) 483 (void)snprintf(name, 8, "buf%um", size / 1048576); 484 else if (__predict_true(size >= 1024)) 485 (void)snprintf(name, 8, "buf%uk", size / 1024); 486 else 487 (void)snprintf(name, 8, "buf%ub", size); 488 pa = (size <= PAGE_SIZE && use_std) 489 ? &pool_allocator_nointr 490 : &bufmempool_allocator; 491 pool_init(pp, size, 0, 0, 0, name, pa, IPL_NONE); 492 pool_setlowat(pp, 1); 493 pool_sethiwat(pp, 1); 494 } 495 496 /* Initialize the buffer queues */ 497 for (dp = bufqueues; dp < &bufqueues[BQUEUES]; dp++) { 498 TAILQ_INIT(&dp->bq_queue); 499 dp->bq_bytes = 0; 500 } 501 502 /* 503 * Estimate hash table size based on the amount of memory we 504 * intend to use for the buffer cache. The average buffer 505 * size is dependent on our clients (i.e. filesystems). 506 * 507 * For now, use an empirical 3K per buffer. 508 */ 509 nbuf = (bufmem_hiwater / 1024) / 3; 510 bufhashtbl = hashinit(nbuf, HASH_LIST, true, &bufhash); 511 512 sysctl_kern_buf_setup(); 513 sysctl_vm_buf_setup(); 514 } 515 516 void 517 bufinit2(void) 518 { 519 520 biodone_sih = softint_establish(SOFTINT_BIO | SOFTINT_MPSAFE, biointr, 521 NULL); 522 if (biodone_sih == NULL) 523 panic("bufinit2: can't establish soft interrupt"); 524 } 525 526 static int 527 buf_lotsfree(void) 528 { 529 int try, thresh; 530 531 /* Always allocate if less than the low water mark. */ 532 if (bufmem < bufmem_lowater) 533 return 1; 534 535 /* Never allocate if greater than the high water mark. */ 536 if (bufmem > bufmem_hiwater) 537 return 0; 538 539 /* If there's anything on the AGE list, it should be eaten. */ 540 if (TAILQ_FIRST(&bufqueues[BQ_AGE].bq_queue) != NULL) 541 return 0; 542 543 /* 544 * The probabily of getting a new allocation is inversely 545 * proportional to the current size of the cache, using 546 * a granularity of 16 steps. 547 */ 548 try = random() & 0x0000000fL; 549 550 /* Don't use "16 * bufmem" here to avoid a 32-bit overflow. */ 551 thresh = (bufmem - bufmem_lowater) / 552 ((bufmem_hiwater - bufmem_lowater) / 16); 553 554 if (try >= thresh) 555 return 1; 556 557 /* Otherwise don't allocate. */ 558 return 0; 559 } 560 561 /* 562 * Return estimate of bytes we think need to be 563 * released to help resolve low memory conditions. 564 * 565 * => called with bufcache_lock held. 566 */ 567 static int 568 buf_canrelease(void) 569 { 570 int pagedemand, ninvalid = 0; 571 572 KASSERT(mutex_owned(&bufcache_lock)); 573 574 if (bufmem < bufmem_lowater) 575 return 0; 576 577 if (bufmem > bufmem_hiwater) 578 return bufmem - bufmem_hiwater; 579 580 ninvalid += bufqueues[BQ_AGE].bq_bytes; 581 582 pagedemand = uvmexp.freetarg - uvmexp.free; 583 if (pagedemand < 0) 584 return ninvalid; 585 return MAX(ninvalid, MIN(2 * MAXBSIZE, 586 MIN((bufmem - bufmem_lowater) / 16, pagedemand * PAGE_SIZE))); 587 } 588 589 /* 590 * Buffer memory allocation helper functions 591 */ 592 static u_long 593 buf_mempoolidx(u_long size) 594 { 595 u_int n = 0; 596 597 size -= 1; 598 size >>= MEMPOOL_INDEX_OFFSET; 599 while (size) { 600 size >>= 1; 601 n += 1; 602 } 603 if (n >= NMEMPOOLS) 604 panic("buf mem pool index %d", n); 605 return n; 606 } 607 608 static u_long 609 buf_roundsize(u_long size) 610 { 611 /* Round up to nearest power of 2 */ 612 return (1 << (buf_mempoolidx(size) + MEMPOOL_INDEX_OFFSET)); 613 } 614 615 static void * 616 buf_alloc(size_t size) 617 { 618 u_int n = buf_mempoolidx(size); 619 void *addr; 620 621 while (1) { 622 addr = pool_get(&bmempools[n], PR_NOWAIT); 623 if (addr != NULL) 624 break; 625 626 /* No memory, see if we can free some. If so, try again */ 627 mutex_enter(&bufcache_lock); 628 if (buf_drain(1) > 0) { 629 mutex_exit(&bufcache_lock); 630 continue; 631 } 632 633 if (curlwp == uvm.pagedaemon_lwp) { 634 mutex_exit(&bufcache_lock); 635 return NULL; 636 } 637 638 /* Wait for buffers to arrive on the LRU queue */ 639 cv_timedwait(&needbuffer_cv, &bufcache_lock, hz / 4); 640 mutex_exit(&bufcache_lock); 641 } 642 643 return addr; 644 } 645 646 static void 647 buf_mrelease(void *addr, size_t size) 648 { 649 650 pool_put(&bmempools[buf_mempoolidx(size)], addr); 651 } 652 653 /* 654 * bread()/breadn() helper. 655 */ 656 static buf_t * 657 bio_doread(struct vnode *vp, daddr_t blkno, int size, kauth_cred_t cred, 658 int async) 659 { 660 buf_t *bp; 661 struct mount *mp; 662 663 bp = getblk(vp, blkno, size, 0, 0); 664 665 /* 666 * getblk() may return NULL if we are the pagedaemon. 667 */ 668 if (bp == NULL) { 669 KASSERT(curlwp == uvm.pagedaemon_lwp); 670 return NULL; 671 } 672 673 /* 674 * If buffer does not have data valid, start a read. 675 * Note that if buffer is BC_INVAL, getblk() won't return it. 676 * Therefore, it's valid if its I/O has completed or been delayed. 677 */ 678 if (!ISSET(bp->b_oflags, (BO_DONE | BO_DELWRI))) { 679 /* Start I/O for the buffer. */ 680 SET(bp->b_flags, B_READ | async); 681 if (async) 682 BIO_SETPRIO(bp, BPRIO_TIMELIMITED); 683 else 684 BIO_SETPRIO(bp, BPRIO_TIMECRITICAL); 685 VOP_STRATEGY(vp, bp); 686 687 /* Pay for the read. */ 688 curlwp->l_ru.ru_inblock++; 689 } else if (async) 690 brelse(bp, 0); 691 692 if (vp->v_type == VBLK) 693 mp = vp->v_specmountpoint; 694 else 695 mp = vp->v_mount; 696 697 /* 698 * Collect statistics on synchronous and asynchronous reads. 699 * Reads from block devices are charged to their associated 700 * filesystem (if any). 701 */ 702 if (mp != NULL) { 703 if (async == 0) 704 mp->mnt_stat.f_syncreads++; 705 else 706 mp->mnt_stat.f_asyncreads++; 707 } 708 709 return (bp); 710 } 711 712 /* 713 * Read a disk block. 714 * This algorithm described in Bach (p.54). 715 */ 716 int 717 bread(struct vnode *vp, daddr_t blkno, int size, kauth_cred_t cred, 718 int flags, buf_t **bpp) 719 { 720 buf_t *bp; 721 int error; 722 723 /* Get buffer for block. */ 724 bp = *bpp = bio_doread(vp, blkno, size, cred, 0); 725 if (bp == NULL) 726 return ENOMEM; 727 728 /* Wait for the read to complete, and return result. */ 729 error = biowait(bp); 730 if (error == 0 && (flags & B_MODIFY) != 0) 731 error = fscow_run(bp, true); 732 if (error) { 733 brelse(bp, 0); 734 *bpp = NULL; 735 } 736 737 return error; 738 } 739 740 /* 741 * Read-ahead multiple disk blocks. The first is sync, the rest async. 742 * Trivial modification to the breada algorithm presented in Bach (p.55). 743 */ 744 int 745 breadn(struct vnode *vp, daddr_t blkno, int size, daddr_t *rablks, 746 int *rasizes, int nrablks, kauth_cred_t cred, int flags, buf_t **bpp) 747 { 748 buf_t *bp; 749 int error, i; 750 751 bp = *bpp = bio_doread(vp, blkno, size, cred, 0); 752 if (bp == NULL) 753 return ENOMEM; 754 755 /* 756 * For each of the read-ahead blocks, start a read, if necessary. 757 */ 758 mutex_enter(&bufcache_lock); 759 for (i = 0; i < nrablks; i++) { 760 /* If it's in the cache, just go on to next one. */ 761 if (incore(vp, rablks[i])) 762 continue; 763 764 /* Get a buffer for the read-ahead block */ 765 mutex_exit(&bufcache_lock); 766 (void) bio_doread(vp, rablks[i], rasizes[i], cred, B_ASYNC); 767 mutex_enter(&bufcache_lock); 768 } 769 mutex_exit(&bufcache_lock); 770 771 /* Otherwise, we had to start a read for it; wait until it's valid. */ 772 error = biowait(bp); 773 if (error == 0 && (flags & B_MODIFY) != 0) 774 error = fscow_run(bp, true); 775 if (error) { 776 brelse(bp, 0); 777 *bpp = NULL; 778 } 779 780 return error; 781 } 782 783 /* 784 * Block write. Described in Bach (p.56) 785 */ 786 int 787 bwrite(buf_t *bp) 788 { 789 int rv, sync, wasdelayed; 790 struct vnode *vp; 791 struct mount *mp; 792 793 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 794 KASSERT(!cv_has_waiters(&bp->b_done)); 795 796 vp = bp->b_vp; 797 if (vp != NULL) { 798 KASSERT(bp->b_objlock == vp->v_interlock); 799 if (vp->v_type == VBLK) 800 mp = vp->v_specmountpoint; 801 else 802 mp = vp->v_mount; 803 } else { 804 mp = NULL; 805 } 806 807 if (mp && mp->mnt_wapbl) { 808 if (bp->b_iodone != mp->mnt_wapbl_op->wo_wapbl_biodone) { 809 bdwrite(bp); 810 return 0; 811 } 812 } 813 814 /* 815 * Remember buffer type, to switch on it later. If the write was 816 * synchronous, but the file system was mounted with MNT_ASYNC, 817 * convert it to a delayed write. 818 * XXX note that this relies on delayed tape writes being converted 819 * to async, not sync writes (which is safe, but ugly). 820 */ 821 sync = !ISSET(bp->b_flags, B_ASYNC); 822 if (sync && mp != NULL && ISSET(mp->mnt_flag, MNT_ASYNC)) { 823 bdwrite(bp); 824 return (0); 825 } 826 827 /* 828 * Collect statistics on synchronous and asynchronous writes. 829 * Writes to block devices are charged to their associated 830 * filesystem (if any). 831 */ 832 if (mp != NULL) { 833 if (sync) 834 mp->mnt_stat.f_syncwrites++; 835 else 836 mp->mnt_stat.f_asyncwrites++; 837 } 838 839 /* 840 * Pay for the I/O operation and make sure the buf is on the correct 841 * vnode queue. 842 */ 843 bp->b_error = 0; 844 wasdelayed = ISSET(bp->b_oflags, BO_DELWRI); 845 CLR(bp->b_flags, B_READ); 846 if (wasdelayed) { 847 mutex_enter(&bufcache_lock); 848 mutex_enter(bp->b_objlock); 849 CLR(bp->b_oflags, BO_DONE | BO_DELWRI); 850 reassignbuf(bp, bp->b_vp); 851 mutex_exit(&bufcache_lock); 852 } else { 853 curlwp->l_ru.ru_oublock++; 854 mutex_enter(bp->b_objlock); 855 CLR(bp->b_oflags, BO_DONE | BO_DELWRI); 856 } 857 if (vp != NULL) 858 vp->v_numoutput++; 859 mutex_exit(bp->b_objlock); 860 861 /* Initiate disk write. */ 862 if (sync) 863 BIO_SETPRIO(bp, BPRIO_TIMECRITICAL); 864 else 865 BIO_SETPRIO(bp, BPRIO_TIMELIMITED); 866 867 VOP_STRATEGY(vp, bp); 868 869 if (sync) { 870 /* If I/O was synchronous, wait for it to complete. */ 871 rv = biowait(bp); 872 873 /* Release the buffer. */ 874 brelse(bp, 0); 875 876 return (rv); 877 } else { 878 return (0); 879 } 880 } 881 882 int 883 vn_bwrite(void *v) 884 { 885 struct vop_bwrite_args *ap = v; 886 887 return (bwrite(ap->a_bp)); 888 } 889 890 /* 891 * Delayed write. 892 * 893 * The buffer is marked dirty, but is not queued for I/O. 894 * This routine should be used when the buffer is expected 895 * to be modified again soon, typically a small write that 896 * partially fills a buffer. 897 * 898 * NB: magnetic tapes cannot be delayed; they must be 899 * written in the order that the writes are requested. 900 * 901 * Described in Leffler, et al. (pp. 208-213). 902 */ 903 void 904 bdwrite(buf_t *bp) 905 { 906 907 KASSERT(bp->b_vp == NULL || bp->b_vp->v_tag != VT_UFS || 908 bp->b_vp->v_type == VBLK || ISSET(bp->b_flags, B_COWDONE)); 909 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 910 KASSERT(!cv_has_waiters(&bp->b_done)); 911 912 /* If this is a tape block, write the block now. */ 913 if (bdev_type(bp->b_dev) == D_TAPE) { 914 bawrite(bp); 915 return; 916 } 917 918 if (wapbl_vphaswapbl(bp->b_vp)) { 919 struct mount *mp = wapbl_vptomp(bp->b_vp); 920 921 if (bp->b_iodone != mp->mnt_wapbl_op->wo_wapbl_biodone) { 922 WAPBL_ADD_BUF(mp, bp); 923 } 924 } 925 926 /* 927 * If the block hasn't been seen before: 928 * (1) Mark it as having been seen, 929 * (2) Charge for the write, 930 * (3) Make sure it's on its vnode's correct block list. 931 */ 932 KASSERT(bp->b_vp == NULL || bp->b_objlock == bp->b_vp->v_interlock); 933 934 if (!ISSET(bp->b_oflags, BO_DELWRI)) { 935 mutex_enter(&bufcache_lock); 936 mutex_enter(bp->b_objlock); 937 SET(bp->b_oflags, BO_DELWRI); 938 curlwp->l_ru.ru_oublock++; 939 reassignbuf(bp, bp->b_vp); 940 mutex_exit(&bufcache_lock); 941 } else { 942 mutex_enter(bp->b_objlock); 943 } 944 /* Otherwise, the "write" is done, so mark and release the buffer. */ 945 CLR(bp->b_oflags, BO_DONE); 946 mutex_exit(bp->b_objlock); 947 948 brelse(bp, 0); 949 } 950 951 /* 952 * Asynchronous block write; just an asynchronous bwrite(). 953 */ 954 void 955 bawrite(buf_t *bp) 956 { 957 958 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 959 KASSERT(bp->b_vp != NULL); 960 961 SET(bp->b_flags, B_ASYNC); 962 VOP_BWRITE(bp->b_vp, bp); 963 } 964 965 /* 966 * Release a buffer on to the free lists. 967 * Described in Bach (p. 46). 968 */ 969 void 970 brelsel(buf_t *bp, int set) 971 { 972 struct bqueue *bufq; 973 struct vnode *vp; 974 975 KASSERT(bp != NULL); 976 KASSERT(mutex_owned(&bufcache_lock)); 977 KASSERT(!cv_has_waiters(&bp->b_done)); 978 KASSERT(bp->b_refcnt > 0); 979 980 SET(bp->b_cflags, set); 981 982 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 983 KASSERT(bp->b_iodone == NULL); 984 985 /* Wake up any processes waiting for any buffer to become free. */ 986 cv_signal(&needbuffer_cv); 987 988 /* Wake up any proceeses waiting for _this_ buffer to become */ 989 if (ISSET(bp->b_cflags, BC_WANTED)) 990 CLR(bp->b_cflags, BC_WANTED|BC_AGE); 991 992 /* If it's clean clear the copy-on-write flag. */ 993 if (ISSET(bp->b_flags, B_COWDONE)) { 994 mutex_enter(bp->b_objlock); 995 if (!ISSET(bp->b_oflags, BO_DELWRI)) 996 CLR(bp->b_flags, B_COWDONE); 997 mutex_exit(bp->b_objlock); 998 } 999 1000 /* 1001 * Determine which queue the buffer should be on, then put it there. 1002 */ 1003 1004 /* If it's locked, don't report an error; try again later. */ 1005 if (ISSET(bp->b_flags, B_LOCKED)) 1006 bp->b_error = 0; 1007 1008 /* If it's not cacheable, or an error, mark it invalid. */ 1009 if (ISSET(bp->b_cflags, BC_NOCACHE) || bp->b_error != 0) 1010 SET(bp->b_cflags, BC_INVAL); 1011 1012 if (ISSET(bp->b_cflags, BC_VFLUSH)) { 1013 /* 1014 * This is a delayed write buffer that was just flushed to 1015 * disk. It is still on the LRU queue. If it's become 1016 * invalid, then we need to move it to a different queue; 1017 * otherwise leave it in its current position. 1018 */ 1019 CLR(bp->b_cflags, BC_VFLUSH); 1020 if (!ISSET(bp->b_cflags, BC_INVAL|BC_AGE) && 1021 !ISSET(bp->b_flags, B_LOCKED) && bp->b_error == 0) { 1022 KDASSERT(checkfreelist(bp, &bufqueues[BQ_LRU], 1)); 1023 goto already_queued; 1024 } else { 1025 bremfree(bp); 1026 } 1027 } 1028 1029 KDASSERT(checkfreelist(bp, &bufqueues[BQ_AGE], 0)); 1030 KDASSERT(checkfreelist(bp, &bufqueues[BQ_LRU], 0)); 1031 KDASSERT(checkfreelist(bp, &bufqueues[BQ_LOCKED], 0)); 1032 1033 if ((bp->b_bufsize <= 0) || ISSET(bp->b_cflags, BC_INVAL)) { 1034 /* 1035 * If it's invalid or empty, dissociate it from its vnode 1036 * and put on the head of the appropriate queue. 1037 */ 1038 if (ISSET(bp->b_flags, B_LOCKED)) { 1039 if (wapbl_vphaswapbl(vp = bp->b_vp)) { 1040 struct mount *mp = wapbl_vptomp(vp); 1041 1042 KASSERT(bp->b_iodone 1043 != mp->mnt_wapbl_op->wo_wapbl_biodone); 1044 WAPBL_REMOVE_BUF(mp, bp); 1045 } 1046 } 1047 1048 mutex_enter(bp->b_objlock); 1049 CLR(bp->b_oflags, BO_DONE|BO_DELWRI); 1050 if ((vp = bp->b_vp) != NULL) { 1051 KASSERT(bp->b_objlock == vp->v_interlock); 1052 reassignbuf(bp, bp->b_vp); 1053 brelvp(bp); 1054 mutex_exit(vp->v_interlock); 1055 } else { 1056 KASSERT(bp->b_objlock == &buffer_lock); 1057 mutex_exit(bp->b_objlock); 1058 } 1059 1060 if (bp->b_bufsize <= 0) 1061 /* no data */ 1062 goto already_queued; 1063 else 1064 /* invalid data */ 1065 bufq = &bufqueues[BQ_AGE]; 1066 binsheadfree(bp, bufq); 1067 } else { 1068 /* 1069 * It has valid data. Put it on the end of the appropriate 1070 * queue, so that it'll stick around for as long as possible. 1071 * If buf is AGE, but has dependencies, must put it on last 1072 * bufqueue to be scanned, ie LRU. This protects against the 1073 * livelock where BQ_AGE only has buffers with dependencies, 1074 * and we thus never get to the dependent buffers in BQ_LRU. 1075 */ 1076 if (ISSET(bp->b_flags, B_LOCKED)) { 1077 /* locked in core */ 1078 bufq = &bufqueues[BQ_LOCKED]; 1079 } else if (!ISSET(bp->b_cflags, BC_AGE)) { 1080 /* valid data */ 1081 bufq = &bufqueues[BQ_LRU]; 1082 } else { 1083 /* stale but valid data */ 1084 bufq = &bufqueues[BQ_AGE]; 1085 } 1086 binstailfree(bp, bufq); 1087 } 1088 already_queued: 1089 /* Unlock the buffer. */ 1090 CLR(bp->b_cflags, BC_AGE|BC_BUSY|BC_NOCACHE); 1091 CLR(bp->b_flags, B_ASYNC); 1092 cv_broadcast(&bp->b_busy); 1093 1094 if (bp->b_bufsize <= 0) 1095 brele(bp); 1096 } 1097 1098 void 1099 brelse(buf_t *bp, int set) 1100 { 1101 1102 mutex_enter(&bufcache_lock); 1103 brelsel(bp, set); 1104 mutex_exit(&bufcache_lock); 1105 } 1106 1107 /* 1108 * Determine if a block is in the cache. 1109 * Just look on what would be its hash chain. If it's there, return 1110 * a pointer to it, unless it's marked invalid. If it's marked invalid, 1111 * we normally don't return the buffer, unless the caller explicitly 1112 * wants us to. 1113 */ 1114 buf_t * 1115 incore(struct vnode *vp, daddr_t blkno) 1116 { 1117 buf_t *bp; 1118 1119 KASSERT(mutex_owned(&bufcache_lock)); 1120 1121 /* Search hash chain */ 1122 LIST_FOREACH(bp, BUFHASH(vp, blkno), b_hash) { 1123 if (bp->b_lblkno == blkno && bp->b_vp == vp && 1124 !ISSET(bp->b_cflags, BC_INVAL)) { 1125 KASSERT(bp->b_objlock == vp->v_interlock); 1126 return (bp); 1127 } 1128 } 1129 1130 return (NULL); 1131 } 1132 1133 /* 1134 * Get a block of requested size that is associated with 1135 * a given vnode and block offset. If it is found in the 1136 * block cache, mark it as having been found, make it busy 1137 * and return it. Otherwise, return an empty block of the 1138 * correct size. It is up to the caller to insure that the 1139 * cached blocks be of the correct size. 1140 */ 1141 buf_t * 1142 getblk(struct vnode *vp, daddr_t blkno, int size, int slpflag, int slptimeo) 1143 { 1144 int err, preserve; 1145 buf_t *bp; 1146 1147 mutex_enter(&bufcache_lock); 1148 loop: 1149 bp = incore(vp, blkno); 1150 if (bp != NULL) { 1151 err = bbusy(bp, ((slpflag & PCATCH) != 0), slptimeo, NULL); 1152 if (err != 0) { 1153 if (err == EPASSTHROUGH) 1154 goto loop; 1155 mutex_exit(&bufcache_lock); 1156 return (NULL); 1157 } 1158 KASSERT(!cv_has_waiters(&bp->b_done)); 1159 #ifdef DIAGNOSTIC 1160 if (ISSET(bp->b_oflags, BO_DONE|BO_DELWRI) && 1161 bp->b_bcount < size && vp->v_type != VBLK) 1162 panic("getblk: block size invariant failed"); 1163 #endif 1164 bremfree(bp); 1165 preserve = 1; 1166 } else { 1167 if ((bp = getnewbuf(slpflag, slptimeo, 0)) == NULL) 1168 goto loop; 1169 1170 if (incore(vp, blkno) != NULL) { 1171 /* The block has come into memory in the meantime. */ 1172 brelsel(bp, 0); 1173 goto loop; 1174 } 1175 1176 LIST_INSERT_HEAD(BUFHASH(vp, blkno), bp, b_hash); 1177 bp->b_blkno = bp->b_lblkno = bp->b_rawblkno = blkno; 1178 mutex_enter(vp->v_interlock); 1179 bgetvp(vp, bp); 1180 mutex_exit(vp->v_interlock); 1181 preserve = 0; 1182 } 1183 mutex_exit(&bufcache_lock); 1184 1185 /* 1186 * LFS can't track total size of B_LOCKED buffer (locked_queue_bytes) 1187 * if we re-size buffers here. 1188 */ 1189 if (ISSET(bp->b_flags, B_LOCKED)) { 1190 KASSERT(bp->b_bufsize >= size); 1191 } else { 1192 if (allocbuf(bp, size, preserve)) { 1193 mutex_enter(&bufcache_lock); 1194 LIST_REMOVE(bp, b_hash); 1195 mutex_exit(&bufcache_lock); 1196 brelse(bp, BC_INVAL); 1197 return NULL; 1198 } 1199 } 1200 BIO_SETPRIO(bp, BPRIO_DEFAULT); 1201 return (bp); 1202 } 1203 1204 /* 1205 * Get an empty, disassociated buffer of given size. 1206 */ 1207 buf_t * 1208 geteblk(int size) 1209 { 1210 buf_t *bp; 1211 int error; 1212 1213 mutex_enter(&bufcache_lock); 1214 while ((bp = getnewbuf(0, 0, 0)) == NULL) 1215 ; 1216 1217 SET(bp->b_cflags, BC_INVAL); 1218 LIST_INSERT_HEAD(&invalhash, bp, b_hash); 1219 mutex_exit(&bufcache_lock); 1220 BIO_SETPRIO(bp, BPRIO_DEFAULT); 1221 error = allocbuf(bp, size, 0); 1222 KASSERT(error == 0); 1223 return (bp); 1224 } 1225 1226 /* 1227 * Expand or contract the actual memory allocated to a buffer. 1228 * 1229 * If the buffer shrinks, data is lost, so it's up to the 1230 * caller to have written it out *first*; this routine will not 1231 * start a write. If the buffer grows, it's the callers 1232 * responsibility to fill out the buffer's additional contents. 1233 */ 1234 int 1235 allocbuf(buf_t *bp, int size, int preserve) 1236 { 1237 void *addr; 1238 vsize_t oldsize, desired_size; 1239 int oldcount; 1240 int delta; 1241 1242 desired_size = buf_roundsize(size); 1243 if (desired_size > MAXBSIZE) 1244 printf("allocbuf: buffer larger than MAXBSIZE requested"); 1245 1246 oldcount = bp->b_bcount; 1247 1248 bp->b_bcount = size; 1249 1250 oldsize = bp->b_bufsize; 1251 if (oldsize == desired_size) { 1252 /* 1253 * Do not short cut the WAPBL resize, as the buffer length 1254 * could still have changed and this would corrupt the 1255 * tracking of the transaction length. 1256 */ 1257 goto out; 1258 } 1259 1260 /* 1261 * If we want a buffer of a different size, re-allocate the 1262 * buffer's memory; copy old content only if needed. 1263 */ 1264 addr = buf_alloc(desired_size); 1265 if (addr == NULL) 1266 return ENOMEM; 1267 if (preserve) 1268 memcpy(addr, bp->b_data, MIN(oldsize,desired_size)); 1269 if (bp->b_data != NULL) 1270 buf_mrelease(bp->b_data, oldsize); 1271 bp->b_data = addr; 1272 bp->b_bufsize = desired_size; 1273 1274 /* 1275 * Update overall buffer memory counter (protected by bufcache_lock) 1276 */ 1277 delta = (long)desired_size - (long)oldsize; 1278 1279 mutex_enter(&bufcache_lock); 1280 if ((bufmem += delta) > bufmem_hiwater) { 1281 /* 1282 * Need to trim overall memory usage. 1283 */ 1284 while (buf_canrelease()) { 1285 if (curcpu()->ci_schedstate.spc_flags & 1286 SPCF_SHOULDYIELD) { 1287 mutex_exit(&bufcache_lock); 1288 preempt(); 1289 mutex_enter(&bufcache_lock); 1290 } 1291 if (buf_trim() == 0) 1292 break; 1293 } 1294 } 1295 mutex_exit(&bufcache_lock); 1296 1297 out: 1298 if (wapbl_vphaswapbl(bp->b_vp)) 1299 WAPBL_RESIZE_BUF(wapbl_vptomp(bp->b_vp), bp, oldsize, oldcount); 1300 1301 return 0; 1302 } 1303 1304 /* 1305 * Find a buffer which is available for use. 1306 * Select something from a free list. 1307 * Preference is to AGE list, then LRU list. 1308 * 1309 * Called with the buffer queues locked. 1310 * Return buffer locked. 1311 */ 1312 buf_t * 1313 getnewbuf(int slpflag, int slptimeo, int from_bufq) 1314 { 1315 buf_t *bp; 1316 struct vnode *vp; 1317 1318 start: 1319 KASSERT(mutex_owned(&bufcache_lock)); 1320 1321 /* 1322 * Get a new buffer from the pool. 1323 */ 1324 if (!from_bufq && buf_lotsfree()) { 1325 mutex_exit(&bufcache_lock); 1326 bp = pool_cache_get(buf_cache, PR_NOWAIT); 1327 if (bp != NULL) { 1328 memset((char *)bp, 0, sizeof(*bp)); 1329 buf_init(bp); 1330 SET(bp->b_cflags, BC_BUSY); /* mark buffer busy */ 1331 mutex_enter(&bufcache_lock); 1332 #if defined(DIAGNOSTIC) 1333 bp->b_freelistindex = -1; 1334 #endif /* defined(DIAGNOSTIC) */ 1335 return (bp); 1336 } 1337 mutex_enter(&bufcache_lock); 1338 } 1339 1340 KASSERT(mutex_owned(&bufcache_lock)); 1341 if ((bp = TAILQ_FIRST(&bufqueues[BQ_AGE].bq_queue)) != NULL || 1342 (bp = TAILQ_FIRST(&bufqueues[BQ_LRU].bq_queue)) != NULL) { 1343 KASSERT(!ISSET(bp->b_cflags, BC_BUSY) || ISSET(bp->b_cflags, BC_VFLUSH)); 1344 bremfree(bp); 1345 1346 /* Buffer is no longer on free lists. */ 1347 SET(bp->b_cflags, BC_BUSY); 1348 } else { 1349 /* 1350 * XXX: !from_bufq should be removed. 1351 */ 1352 if (!from_bufq || curlwp != uvm.pagedaemon_lwp) { 1353 /* wait for a free buffer of any kind */ 1354 if ((slpflag & PCATCH) != 0) 1355 (void)cv_timedwait_sig(&needbuffer_cv, 1356 &bufcache_lock, slptimeo); 1357 else 1358 (void)cv_timedwait(&needbuffer_cv, 1359 &bufcache_lock, slptimeo); 1360 } 1361 return (NULL); 1362 } 1363 1364 #ifdef DIAGNOSTIC 1365 if (bp->b_bufsize <= 0) 1366 panic("buffer %p: on queue but empty", bp); 1367 #endif 1368 1369 if (ISSET(bp->b_cflags, BC_VFLUSH)) { 1370 /* 1371 * This is a delayed write buffer being flushed to disk. Make 1372 * sure it gets aged out of the queue when it's finished, and 1373 * leave it off the LRU queue. 1374 */ 1375 CLR(bp->b_cflags, BC_VFLUSH); 1376 SET(bp->b_cflags, BC_AGE); 1377 goto start; 1378 } 1379 1380 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 1381 KASSERT(bp->b_refcnt > 0); 1382 KASSERT(!cv_has_waiters(&bp->b_done)); 1383 1384 /* 1385 * If buffer was a delayed write, start it and return NULL 1386 * (since we might sleep while starting the write). 1387 */ 1388 if (ISSET(bp->b_oflags, BO_DELWRI)) { 1389 /* 1390 * This buffer has gone through the LRU, so make sure it gets 1391 * reused ASAP. 1392 */ 1393 SET(bp->b_cflags, BC_AGE); 1394 mutex_exit(&bufcache_lock); 1395 bawrite(bp); 1396 mutex_enter(&bufcache_lock); 1397 return (NULL); 1398 } 1399 1400 vp = bp->b_vp; 1401 1402 /* clear out various other fields */ 1403 bp->b_cflags = BC_BUSY; 1404 bp->b_oflags = 0; 1405 bp->b_flags = 0; 1406 bp->b_dev = NODEV; 1407 bp->b_blkno = 0; 1408 bp->b_lblkno = 0; 1409 bp->b_rawblkno = 0; 1410 bp->b_iodone = 0; 1411 bp->b_error = 0; 1412 bp->b_resid = 0; 1413 bp->b_bcount = 0; 1414 1415 LIST_REMOVE(bp, b_hash); 1416 1417 /* Disassociate us from our vnode, if we had one... */ 1418 if (vp != NULL) { 1419 mutex_enter(vp->v_interlock); 1420 brelvp(bp); 1421 mutex_exit(vp->v_interlock); 1422 } 1423 1424 return (bp); 1425 } 1426 1427 /* 1428 * Attempt to free an aged buffer off the queues. 1429 * Called with queue lock held. 1430 * Returns the amount of buffer memory freed. 1431 */ 1432 static int 1433 buf_trim(void) 1434 { 1435 buf_t *bp; 1436 long size = 0; 1437 1438 KASSERT(mutex_owned(&bufcache_lock)); 1439 1440 /* Instruct getnewbuf() to get buffers off the queues */ 1441 if ((bp = getnewbuf(PCATCH, 1, 1)) == NULL) 1442 return 0; 1443 1444 KASSERT((bp->b_cflags & BC_WANTED) == 0); 1445 size = bp->b_bufsize; 1446 bufmem -= size; 1447 if (size > 0) { 1448 buf_mrelease(bp->b_data, size); 1449 bp->b_bcount = bp->b_bufsize = 0; 1450 } 1451 /* brelse() will return the buffer to the global buffer pool */ 1452 brelsel(bp, 0); 1453 return size; 1454 } 1455 1456 int 1457 buf_drain(int n) 1458 { 1459 int size = 0, sz; 1460 1461 KASSERT(mutex_owned(&bufcache_lock)); 1462 1463 while (size < n && bufmem > bufmem_lowater) { 1464 sz = buf_trim(); 1465 if (sz <= 0) 1466 break; 1467 size += sz; 1468 } 1469 1470 return size; 1471 } 1472 1473 /* 1474 * Wait for operations on the buffer to complete. 1475 * When they do, extract and return the I/O's error value. 1476 */ 1477 int 1478 biowait(buf_t *bp) 1479 { 1480 1481 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 1482 KASSERT(bp->b_refcnt > 0); 1483 1484 mutex_enter(bp->b_objlock); 1485 while (!ISSET(bp->b_oflags, BO_DONE | BO_DELWRI)) 1486 cv_wait(&bp->b_done, bp->b_objlock); 1487 mutex_exit(bp->b_objlock); 1488 1489 return bp->b_error; 1490 } 1491 1492 /* 1493 * Mark I/O complete on a buffer. 1494 * 1495 * If a callback has been requested, e.g. the pageout 1496 * daemon, do so. Otherwise, awaken waiting processes. 1497 * 1498 * [ Leffler, et al., says on p.247: 1499 * "This routine wakes up the blocked process, frees the buffer 1500 * for an asynchronous write, or, for a request by the pagedaemon 1501 * process, invokes a procedure specified in the buffer structure" ] 1502 * 1503 * In real life, the pagedaemon (or other system processes) wants 1504 * to do async stuff to, and doesn't want the buffer brelse()'d. 1505 * (for swap pager, that puts swap buffers on the free lists (!!!), 1506 * for the vn device, that puts allocated buffers on the free lists!) 1507 */ 1508 void 1509 biodone(buf_t *bp) 1510 { 1511 int s; 1512 1513 KASSERT(!ISSET(bp->b_oflags, BO_DONE)); 1514 1515 if (cpu_intr_p()) { 1516 /* From interrupt mode: defer to a soft interrupt. */ 1517 s = splvm(); 1518 TAILQ_INSERT_TAIL(&curcpu()->ci_data.cpu_biodone, bp, b_actq); 1519 softint_schedule(biodone_sih); 1520 splx(s); 1521 } else { 1522 /* Process now - the buffer may be freed soon. */ 1523 biodone2(bp); 1524 } 1525 } 1526 1527 static void 1528 biodone2(buf_t *bp) 1529 { 1530 void (*callout)(buf_t *); 1531 1532 mutex_enter(bp->b_objlock); 1533 /* Note that the transfer is done. */ 1534 if (ISSET(bp->b_oflags, BO_DONE)) 1535 panic("biodone2 already"); 1536 CLR(bp->b_flags, B_COWDONE); 1537 SET(bp->b_oflags, BO_DONE); 1538 BIO_SETPRIO(bp, BPRIO_DEFAULT); 1539 1540 /* Wake up waiting writers. */ 1541 if (!ISSET(bp->b_flags, B_READ)) 1542 vwakeup(bp); 1543 1544 if ((callout = bp->b_iodone) != NULL) { 1545 /* Note callout done, then call out. */ 1546 KASSERT(!cv_has_waiters(&bp->b_done)); 1547 KERNEL_LOCK(1, NULL); /* XXXSMP */ 1548 bp->b_iodone = NULL; 1549 mutex_exit(bp->b_objlock); 1550 (*callout)(bp); 1551 KERNEL_UNLOCK_ONE(NULL); /* XXXSMP */ 1552 } else if (ISSET(bp->b_flags, B_ASYNC)) { 1553 /* If async, release. */ 1554 KASSERT(!cv_has_waiters(&bp->b_done)); 1555 mutex_exit(bp->b_objlock); 1556 brelse(bp, 0); 1557 } else { 1558 /* Otherwise just wake up waiters in biowait(). */ 1559 cv_broadcast(&bp->b_done); 1560 mutex_exit(bp->b_objlock); 1561 } 1562 } 1563 1564 static void 1565 biointr(void *cookie) 1566 { 1567 struct cpu_info *ci; 1568 buf_t *bp; 1569 int s; 1570 1571 ci = curcpu(); 1572 1573 while (!TAILQ_EMPTY(&ci->ci_data.cpu_biodone)) { 1574 KASSERT(curcpu() == ci); 1575 1576 s = splvm(); 1577 bp = TAILQ_FIRST(&ci->ci_data.cpu_biodone); 1578 TAILQ_REMOVE(&ci->ci_data.cpu_biodone, bp, b_actq); 1579 splx(s); 1580 1581 biodone2(bp); 1582 } 1583 } 1584 1585 /* 1586 * Wait for all buffers to complete I/O 1587 * Return the number of "stuck" buffers. 1588 */ 1589 int 1590 buf_syncwait(void) 1591 { 1592 buf_t *bp; 1593 int iter, nbusy, nbusy_prev = 0, dcount, ihash; 1594 1595 dcount = 10000; 1596 for (iter = 0; iter < 20;) { 1597 mutex_enter(&bufcache_lock); 1598 nbusy = 0; 1599 for (ihash = 0; ihash < bufhash+1; ihash++) { 1600 LIST_FOREACH(bp, &bufhashtbl[ihash], b_hash) { 1601 if ((bp->b_cflags & (BC_BUSY|BC_INVAL)) == BC_BUSY) 1602 nbusy += ((bp->b_flags & B_READ) == 0); 1603 } 1604 } 1605 mutex_exit(&bufcache_lock); 1606 1607 if (nbusy == 0) 1608 break; 1609 if (nbusy_prev == 0) 1610 nbusy_prev = nbusy; 1611 printf("%d ", nbusy); 1612 kpause("bflush", false, MAX(1, hz / 25 * iter), NULL); 1613 if (nbusy >= nbusy_prev) /* we didn't flush anything */ 1614 iter++; 1615 else 1616 nbusy_prev = nbusy; 1617 } 1618 1619 if (nbusy) { 1620 #if defined(DEBUG) || defined(DEBUG_HALT_BUSY) 1621 printf("giving up\nPrinting vnodes for busy buffers\n"); 1622 for (ihash = 0; ihash < bufhash+1; ihash++) { 1623 LIST_FOREACH(bp, &bufhashtbl[ihash], b_hash) { 1624 if ((bp->b_cflags & (BC_BUSY|BC_INVAL)) == BC_BUSY && 1625 (bp->b_flags & B_READ) == 0) 1626 vprint(NULL, bp->b_vp); 1627 } 1628 } 1629 #endif 1630 } 1631 1632 return nbusy; 1633 } 1634 1635 static void 1636 sysctl_fillbuf(buf_t *i, struct buf_sysctl *o) 1637 { 1638 1639 o->b_flags = i->b_flags | i->b_cflags | i->b_oflags; 1640 o->b_error = i->b_error; 1641 o->b_prio = i->b_prio; 1642 o->b_dev = i->b_dev; 1643 o->b_bufsize = i->b_bufsize; 1644 o->b_bcount = i->b_bcount; 1645 o->b_resid = i->b_resid; 1646 o->b_addr = PTRTOUINT64(i->b_data); 1647 o->b_blkno = i->b_blkno; 1648 o->b_rawblkno = i->b_rawblkno; 1649 o->b_iodone = PTRTOUINT64(i->b_iodone); 1650 o->b_proc = PTRTOUINT64(i->b_proc); 1651 o->b_vp = PTRTOUINT64(i->b_vp); 1652 o->b_saveaddr = PTRTOUINT64(i->b_saveaddr); 1653 o->b_lblkno = i->b_lblkno; 1654 } 1655 1656 #define KERN_BUFSLOP 20 1657 static int 1658 sysctl_dobuf(SYSCTLFN_ARGS) 1659 { 1660 buf_t *bp; 1661 struct buf_sysctl bs; 1662 struct bqueue *bq; 1663 char *dp; 1664 u_int i, op, arg; 1665 size_t len, needed, elem_size, out_size; 1666 int error, elem_count, retries; 1667 1668 if (namelen == 1 && name[0] == CTL_QUERY) 1669 return (sysctl_query(SYSCTLFN_CALL(rnode))); 1670 1671 if (namelen != 4) 1672 return (EINVAL); 1673 1674 retries = 100; 1675 retry: 1676 dp = oldp; 1677 len = (oldp != NULL) ? *oldlenp : 0; 1678 op = name[0]; 1679 arg = name[1]; 1680 elem_size = name[2]; 1681 elem_count = name[3]; 1682 out_size = MIN(sizeof(bs), elem_size); 1683 1684 /* 1685 * at the moment, these are just "placeholders" to make the 1686 * API for retrieving kern.buf data more extensible in the 1687 * future. 1688 * 1689 * XXX kern.buf currently has "netbsd32" issues. hopefully 1690 * these will be resolved at a later point. 1691 */ 1692 if (op != KERN_BUF_ALL || arg != KERN_BUF_ALL || 1693 elem_size < 1 || elem_count < 0) 1694 return (EINVAL); 1695 1696 error = 0; 1697 needed = 0; 1698 sysctl_unlock(); 1699 mutex_enter(&bufcache_lock); 1700 for (i = 0; i < BQUEUES; i++) { 1701 bq = &bufqueues[i]; 1702 TAILQ_FOREACH(bp, &bq->bq_queue, b_freelist) { 1703 bq->bq_marker = bp; 1704 if (len >= elem_size && elem_count > 0) { 1705 sysctl_fillbuf(bp, &bs); 1706 mutex_exit(&bufcache_lock); 1707 error = copyout(&bs, dp, out_size); 1708 mutex_enter(&bufcache_lock); 1709 if (error) 1710 break; 1711 if (bq->bq_marker != bp) { 1712 /* 1713 * This sysctl node is only for 1714 * statistics. Retry; if the 1715 * queue keeps changing, then 1716 * bail out. 1717 */ 1718 if (retries-- == 0) { 1719 error = EAGAIN; 1720 break; 1721 } 1722 mutex_exit(&bufcache_lock); 1723 sysctl_relock(); 1724 goto retry; 1725 } 1726 dp += elem_size; 1727 len -= elem_size; 1728 } 1729 needed += elem_size; 1730 if (elem_count > 0 && elem_count != INT_MAX) 1731 elem_count--; 1732 } 1733 if (error != 0) 1734 break; 1735 } 1736 mutex_exit(&bufcache_lock); 1737 sysctl_relock(); 1738 1739 *oldlenp = needed; 1740 if (oldp == NULL) 1741 *oldlenp += KERN_BUFSLOP * sizeof(buf_t); 1742 1743 return (error); 1744 } 1745 1746 static int 1747 sysctl_bufvm_update(SYSCTLFN_ARGS) 1748 { 1749 int error, rv; 1750 struct sysctlnode node; 1751 unsigned int temp_bufcache; 1752 unsigned long temp_water; 1753 1754 /* Take a copy of the supplied node and its data */ 1755 node = *rnode; 1756 if (node.sysctl_data == &bufcache) { 1757 node.sysctl_data = &temp_bufcache; 1758 temp_bufcache = *(unsigned int *)rnode->sysctl_data; 1759 } else { 1760 node.sysctl_data = &temp_water; 1761 temp_water = *(unsigned long *)rnode->sysctl_data; 1762 } 1763 1764 /* Update the copy */ 1765 error = sysctl_lookup(SYSCTLFN_CALL(&node)); 1766 if (error || newp == NULL) 1767 return (error); 1768 1769 if (rnode->sysctl_data == &bufcache) { 1770 if (temp_bufcache > 100) 1771 return (EINVAL); 1772 bufcache = temp_bufcache; 1773 buf_setwm(); 1774 } else if (rnode->sysctl_data == &bufmem_lowater) { 1775 if (bufmem_hiwater - temp_water < 16) 1776 return (EINVAL); 1777 bufmem_lowater = temp_water; 1778 } else if (rnode->sysctl_data == &bufmem_hiwater) { 1779 if (temp_water - bufmem_lowater < 16) 1780 return (EINVAL); 1781 bufmem_hiwater = temp_water; 1782 } else 1783 return (EINVAL); 1784 1785 /* Drain until below new high water mark */ 1786 sysctl_unlock(); 1787 mutex_enter(&bufcache_lock); 1788 while (bufmem > bufmem_hiwater) { 1789 rv = buf_drain((bufmem - bufmem_hiwater) / (2 * 1024)); 1790 if (rv <= 0) 1791 break; 1792 } 1793 mutex_exit(&bufcache_lock); 1794 sysctl_relock(); 1795 1796 return 0; 1797 } 1798 1799 static struct sysctllog *vfsbio_sysctllog; 1800 1801 static void 1802 sysctl_kern_buf_setup(void) 1803 { 1804 1805 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1806 CTLFLAG_PERMANENT, 1807 CTLTYPE_NODE, "kern", NULL, 1808 NULL, 0, NULL, 0, 1809 CTL_KERN, CTL_EOL); 1810 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1811 CTLFLAG_PERMANENT, 1812 CTLTYPE_NODE, "buf", 1813 SYSCTL_DESCR("Kernel buffer cache information"), 1814 sysctl_dobuf, 0, NULL, 0, 1815 CTL_KERN, KERN_BUF, CTL_EOL); 1816 } 1817 1818 static void 1819 sysctl_vm_buf_setup(void) 1820 { 1821 1822 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1823 CTLFLAG_PERMANENT, 1824 CTLTYPE_NODE, "vm", NULL, 1825 NULL, 0, NULL, 0, 1826 CTL_VM, CTL_EOL); 1827 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1828 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 1829 CTLTYPE_INT, "bufcache", 1830 SYSCTL_DESCR("Percentage of physical memory to use for " 1831 "buffer cache"), 1832 sysctl_bufvm_update, 0, &bufcache, 0, 1833 CTL_VM, CTL_CREATE, CTL_EOL); 1834 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1835 CTLFLAG_PERMANENT|CTLFLAG_READONLY, 1836 CTLTYPE_LONG, "bufmem", 1837 SYSCTL_DESCR("Amount of kernel memory used by buffer " 1838 "cache"), 1839 NULL, 0, &bufmem, 0, 1840 CTL_VM, CTL_CREATE, CTL_EOL); 1841 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1842 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 1843 CTLTYPE_LONG, "bufmem_lowater", 1844 SYSCTL_DESCR("Minimum amount of kernel memory to " 1845 "reserve for buffer cache"), 1846 sysctl_bufvm_update, 0, &bufmem_lowater, 0, 1847 CTL_VM, CTL_CREATE, CTL_EOL); 1848 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1849 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 1850 CTLTYPE_LONG, "bufmem_hiwater", 1851 SYSCTL_DESCR("Maximum amount of kernel memory to use " 1852 "for buffer cache"), 1853 sysctl_bufvm_update, 0, &bufmem_hiwater, 0, 1854 CTL_VM, CTL_CREATE, CTL_EOL); 1855 } 1856 1857 #ifdef DEBUG 1858 /* 1859 * Print out statistics on the current allocation of the buffer pool. 1860 * Can be enabled to print out on every ``sync'' by setting "syncprt" 1861 * in vfs_syscalls.c using sysctl. 1862 */ 1863 void 1864 vfs_bufstats(void) 1865 { 1866 int i, j, count; 1867 buf_t *bp; 1868 struct bqueue *dp; 1869 int counts[(MAXBSIZE / PAGE_SIZE) + 1]; 1870 static const char *bname[BQUEUES] = { "LOCKED", "LRU", "AGE" }; 1871 1872 for (dp = bufqueues, i = 0; dp < &bufqueues[BQUEUES]; dp++, i++) { 1873 count = 0; 1874 for (j = 0; j <= MAXBSIZE/PAGE_SIZE; j++) 1875 counts[j] = 0; 1876 TAILQ_FOREACH(bp, &dp->bq_queue, b_freelist) { 1877 counts[bp->b_bufsize/PAGE_SIZE]++; 1878 count++; 1879 } 1880 printf("%s: total-%d", bname[i], count); 1881 for (j = 0; j <= MAXBSIZE/PAGE_SIZE; j++) 1882 if (counts[j] != 0) 1883 printf(", %d-%d", j * PAGE_SIZE, counts[j]); 1884 printf("\n"); 1885 } 1886 } 1887 #endif /* DEBUG */ 1888 1889 /* ------------------------------ */ 1890 1891 buf_t * 1892 getiobuf(struct vnode *vp, bool waitok) 1893 { 1894 buf_t *bp; 1895 1896 bp = pool_cache_get(bufio_cache, (waitok ? PR_WAITOK : PR_NOWAIT)); 1897 if (bp == NULL) 1898 return bp; 1899 1900 buf_init(bp); 1901 1902 if ((bp->b_vp = vp) == NULL) 1903 bp->b_objlock = &buffer_lock; 1904 else 1905 bp->b_objlock = vp->v_interlock; 1906 1907 return bp; 1908 } 1909 1910 void 1911 putiobuf(buf_t *bp) 1912 { 1913 1914 buf_destroy(bp); 1915 pool_cache_put(bufio_cache, bp); 1916 } 1917 1918 /* 1919 * nestiobuf_iodone: b_iodone callback for nested buffers. 1920 */ 1921 1922 void 1923 nestiobuf_iodone(buf_t *bp) 1924 { 1925 buf_t *mbp = bp->b_private; 1926 int error; 1927 int donebytes; 1928 1929 KASSERT(bp->b_bcount <= bp->b_bufsize); 1930 KASSERT(mbp != bp); 1931 1932 error = bp->b_error; 1933 if (bp->b_error == 0 && 1934 (bp->b_bcount < bp->b_bufsize || bp->b_resid > 0)) { 1935 /* 1936 * Not all got transfered, raise an error. We have no way to 1937 * propagate these conditions to mbp. 1938 */ 1939 error = EIO; 1940 } 1941 1942 donebytes = bp->b_bufsize; 1943 1944 putiobuf(bp); 1945 nestiobuf_done(mbp, donebytes, error); 1946 } 1947 1948 /* 1949 * nestiobuf_setup: setup a "nested" buffer. 1950 * 1951 * => 'mbp' is a "master" buffer which is being divided into sub pieces. 1952 * => 'bp' should be a buffer allocated by getiobuf. 1953 * => 'offset' is a byte offset in the master buffer. 1954 * => 'size' is a size in bytes of this nested buffer. 1955 */ 1956 1957 void 1958 nestiobuf_setup(buf_t *mbp, buf_t *bp, int offset, size_t size) 1959 { 1960 const int b_read = mbp->b_flags & B_READ; 1961 struct vnode *vp = mbp->b_vp; 1962 1963 KASSERT(mbp->b_bcount >= offset + size); 1964 bp->b_vp = vp; 1965 bp->b_dev = mbp->b_dev; 1966 bp->b_objlock = mbp->b_objlock; 1967 bp->b_cflags = BC_BUSY; 1968 bp->b_flags = B_ASYNC | b_read; 1969 bp->b_iodone = nestiobuf_iodone; 1970 bp->b_data = (char *)mbp->b_data + offset; 1971 bp->b_resid = bp->b_bcount = size; 1972 bp->b_bufsize = bp->b_bcount; 1973 bp->b_private = mbp; 1974 BIO_COPYPRIO(bp, mbp); 1975 if (!b_read && vp != NULL) { 1976 mutex_enter(vp->v_interlock); 1977 vp->v_numoutput++; 1978 mutex_exit(vp->v_interlock); 1979 } 1980 } 1981 1982 /* 1983 * nestiobuf_done: propagate completion to the master buffer. 1984 * 1985 * => 'donebytes' specifies how many bytes in the 'mbp' is completed. 1986 * => 'error' is an errno(2) that 'donebytes' has been completed with. 1987 */ 1988 1989 void 1990 nestiobuf_done(buf_t *mbp, int donebytes, int error) 1991 { 1992 1993 if (donebytes == 0) { 1994 return; 1995 } 1996 mutex_enter(mbp->b_objlock); 1997 KASSERT(mbp->b_resid >= donebytes); 1998 mbp->b_resid -= donebytes; 1999 if (error) 2000 mbp->b_error = error; 2001 if (mbp->b_resid == 0) { 2002 if (mbp->b_error) 2003 mbp->b_resid = mbp->b_bcount; 2004 mutex_exit(mbp->b_objlock); 2005 biodone(mbp); 2006 } else 2007 mutex_exit(mbp->b_objlock); 2008 } 2009 2010 void 2011 buf_init(buf_t *bp) 2012 { 2013 2014 cv_init(&bp->b_busy, "biolock"); 2015 cv_init(&bp->b_done, "biowait"); 2016 bp->b_dev = NODEV; 2017 bp->b_error = 0; 2018 bp->b_flags = 0; 2019 bp->b_cflags = 0; 2020 bp->b_oflags = 0; 2021 bp->b_objlock = &buffer_lock; 2022 bp->b_iodone = NULL; 2023 bp->b_refcnt = 1; 2024 bp->b_dev = NODEV; 2025 bp->b_vnbufs.le_next = NOLIST; 2026 BIO_SETPRIO(bp, BPRIO_DEFAULT); 2027 } 2028 2029 void 2030 buf_destroy(buf_t *bp) 2031 { 2032 2033 cv_destroy(&bp->b_done); 2034 cv_destroy(&bp->b_busy); 2035 } 2036 2037 int 2038 bbusy(buf_t *bp, bool intr, int timo, kmutex_t *interlock) 2039 { 2040 int error; 2041 2042 KASSERT(mutex_owned(&bufcache_lock)); 2043 2044 if ((bp->b_cflags & BC_BUSY) != 0) { 2045 if (curlwp == uvm.pagedaemon_lwp) 2046 return EDEADLK; 2047 bp->b_cflags |= BC_WANTED; 2048 bref(bp); 2049 if (interlock != NULL) 2050 mutex_exit(interlock); 2051 if (intr) { 2052 error = cv_timedwait_sig(&bp->b_busy, &bufcache_lock, 2053 timo); 2054 } else { 2055 error = cv_timedwait(&bp->b_busy, &bufcache_lock, 2056 timo); 2057 } 2058 brele(bp); 2059 if (interlock != NULL) 2060 mutex_enter(interlock); 2061 if (error != 0) 2062 return error; 2063 return EPASSTHROUGH; 2064 } 2065 bp->b_cflags |= BC_BUSY; 2066 2067 return 0; 2068 } 2069
Indexes created Thu Sep 25 16:09:42 GMT 2025