1 /* $NetBSD: vfs_bio.c,v 1.248 2013/10/25 20:36:08 martin 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 1.248 2013/10/25 20:36:08 martin 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 vsize_t mapsz = 0; 400 401 /* 402 * Determine the upper bound of memory to use for buffers. 403 * 404 * - If bufpages is specified, use that as the number 405 * pages. 406 * 407 * - Otherwise, use bufcache as the percentage of 408 * physical memory. 409 */ 410 if (bufpages != 0) { 411 n = bufpages; 412 } else { 413 if (bufcache < 5) { 414 printf("forcing bufcache %d -> 5", bufcache); 415 bufcache = 5; 416 } 417 if (bufcache > 95) { 418 printf("forcing bufcache %d -> 95", bufcache); 419 bufcache = 95; 420 } 421 if (buf_map != NULL) 422 mapsz = vm_map_max(buf_map) - vm_map_min(buf_map); 423 n = calc_cache_size(mapsz, bufcache, 424 (buf_map != kernel_map) ? 100 : BUFCACHE_VA_MAXPCT) 425 / PAGE_SIZE; 426 } 427 428 n <<= PAGE_SHIFT; 429 if (bufmem_valimit != 0 && n > bufmem_valimit) 430 n = bufmem_valimit; 431 432 return (n); 433 } 434 435 /* 436 * Initialize buffers and hash links for buffers. 437 */ 438 void 439 bufinit(void) 440 { 441 struct bqueue *dp; 442 int use_std; 443 u_int i; 444 445 mutex_init(&bufcache_lock, MUTEX_DEFAULT, IPL_NONE); 446 mutex_init(&buffer_lock, MUTEX_DEFAULT, IPL_NONE); 447 cv_init(&needbuffer_cv, "needbuf"); 448 449 if (bufmem_valimit != 0) { 450 vaddr_t minaddr = 0, maxaddr; 451 buf_map = uvm_km_suballoc(kernel_map, &minaddr, &maxaddr, 452 bufmem_valimit, 0, false, 0); 453 if (buf_map == NULL) 454 panic("bufinit: cannot allocate submap"); 455 } else 456 buf_map = kernel_map; 457 458 /* 459 * Initialize buffer cache memory parameters. 460 */ 461 bufmem = 0; 462 buf_setwm(); 463 464 /* On "small" machines use small pool page sizes where possible */ 465 use_std = (physmem < atop(16*1024*1024)); 466 467 /* 468 * Also use them on systems that can map the pool pages using 469 * a direct-mapped segment. 470 */ 471 #ifdef PMAP_MAP_POOLPAGE 472 use_std = 1; 473 #endif 474 475 buf_cache = pool_cache_init(sizeof(buf_t), 0, 0, 0, 476 "bufpl", NULL, IPL_SOFTBIO, NULL, NULL, NULL); 477 bufio_cache = pool_cache_init(sizeof(buf_t), 0, 0, 0, 478 "biopl", NULL, IPL_BIO, NULL, NULL, NULL); 479 480 for (i = 0; i < NMEMPOOLS; i++) { 481 struct pool_allocator *pa; 482 struct pool *pp = &bmempools[i]; 483 u_int size = 1 << (i + MEMPOOL_INDEX_OFFSET); 484 char *name = kmem_alloc(8, KM_SLEEP); /* XXX: never freed */ 485 if (__predict_false(size >= 1048576)) 486 (void)snprintf(name, 8, "buf%um", size / 1048576); 487 else if (__predict_true(size >= 1024)) 488 (void)snprintf(name, 8, "buf%uk", size / 1024); 489 else 490 (void)snprintf(name, 8, "buf%ub", size); 491 pa = (size <= PAGE_SIZE && use_std) 492 ? &pool_allocator_nointr 493 : &bufmempool_allocator; 494 pool_init(pp, size, 0, 0, 0, name, pa, IPL_NONE); 495 pool_setlowat(pp, 1); 496 pool_sethiwat(pp, 1); 497 } 498 499 /* Initialize the buffer queues */ 500 for (dp = bufqueues; dp < &bufqueues[BQUEUES]; dp++) { 501 TAILQ_INIT(&dp->bq_queue); 502 dp->bq_bytes = 0; 503 } 504 505 /* 506 * Estimate hash table size based on the amount of memory we 507 * intend to use for the buffer cache. The average buffer 508 * size is dependent on our clients (i.e. filesystems). 509 * 510 * For now, use an empirical 3K per buffer. 511 */ 512 nbuf = (bufmem_hiwater / 1024) / 3; 513 bufhashtbl = hashinit(nbuf, HASH_LIST, true, &bufhash); 514 515 sysctl_kern_buf_setup(); 516 sysctl_vm_buf_setup(); 517 } 518 519 void 520 bufinit2(void) 521 { 522 523 biodone_sih = softint_establish(SOFTINT_BIO | SOFTINT_MPSAFE, biointr, 524 NULL); 525 if (biodone_sih == NULL) 526 panic("bufinit2: can't establish soft interrupt"); 527 } 528 529 static int 530 buf_lotsfree(void) 531 { 532 int try, thresh; 533 534 /* Always allocate if less than the low water mark. */ 535 if (bufmem < bufmem_lowater) 536 return 1; 537 538 /* Never allocate if greater than the high water mark. */ 539 if (bufmem > bufmem_hiwater) 540 return 0; 541 542 /* If there's anything on the AGE list, it should be eaten. */ 543 if (TAILQ_FIRST(&bufqueues[BQ_AGE].bq_queue) != NULL) 544 return 0; 545 546 /* 547 * The probabily of getting a new allocation is inversely 548 * proportional to the current size of the cache, using 549 * a granularity of 16 steps. 550 */ 551 try = random() & 0x0000000fL; 552 553 /* Don't use "16 * bufmem" here to avoid a 32-bit overflow. */ 554 thresh = (bufmem - bufmem_lowater) / 555 ((bufmem_hiwater - bufmem_lowater) / 16); 556 557 if (try >= thresh) 558 return 1; 559 560 /* Otherwise don't allocate. */ 561 return 0; 562 } 563 564 /* 565 * Return estimate of bytes we think need to be 566 * released to help resolve low memory conditions. 567 * 568 * => called with bufcache_lock held. 569 */ 570 static int 571 buf_canrelease(void) 572 { 573 int pagedemand, ninvalid = 0; 574 575 KASSERT(mutex_owned(&bufcache_lock)); 576 577 if (bufmem < bufmem_lowater) 578 return 0; 579 580 if (bufmem > bufmem_hiwater) 581 return bufmem - bufmem_hiwater; 582 583 ninvalid += bufqueues[BQ_AGE].bq_bytes; 584 585 pagedemand = uvmexp.freetarg - uvmexp.free; 586 if (pagedemand < 0) 587 return ninvalid; 588 return MAX(ninvalid, MIN(2 * MAXBSIZE, 589 MIN((bufmem - bufmem_lowater) / 16, pagedemand * PAGE_SIZE))); 590 } 591 592 /* 593 * Buffer memory allocation helper functions 594 */ 595 static u_long 596 buf_mempoolidx(u_long size) 597 { 598 u_int n = 0; 599 600 size -= 1; 601 size >>= MEMPOOL_INDEX_OFFSET; 602 while (size) { 603 size >>= 1; 604 n += 1; 605 } 606 if (n >= NMEMPOOLS) 607 panic("buf mem pool index %d", n); 608 return n; 609 } 610 611 static u_long 612 buf_roundsize(u_long size) 613 { 614 /* Round up to nearest power of 2 */ 615 return (1 << (buf_mempoolidx(size) + MEMPOOL_INDEX_OFFSET)); 616 } 617 618 static void * 619 buf_alloc(size_t size) 620 { 621 u_int n = buf_mempoolidx(size); 622 void *addr; 623 624 while (1) { 625 addr = pool_get(&bmempools[n], PR_NOWAIT); 626 if (addr != NULL) 627 break; 628 629 /* No memory, see if we can free some. If so, try again */ 630 mutex_enter(&bufcache_lock); 631 if (buf_drain(1) > 0) { 632 mutex_exit(&bufcache_lock); 633 continue; 634 } 635 636 if (curlwp == uvm.pagedaemon_lwp) { 637 mutex_exit(&bufcache_lock); 638 return NULL; 639 } 640 641 /* Wait for buffers to arrive on the LRU queue */ 642 cv_timedwait(&needbuffer_cv, &bufcache_lock, hz / 4); 643 mutex_exit(&bufcache_lock); 644 } 645 646 return addr; 647 } 648 649 static void 650 buf_mrelease(void *addr, size_t size) 651 { 652 653 pool_put(&bmempools[buf_mempoolidx(size)], addr); 654 } 655 656 /* 657 * bread()/breadn() helper. 658 */ 659 static buf_t * 660 bio_doread(struct vnode *vp, daddr_t blkno, int size, kauth_cred_t cred, 661 int async) 662 { 663 buf_t *bp; 664 struct mount *mp; 665 666 bp = getblk(vp, blkno, size, 0, 0); 667 668 /* 669 * getblk() may return NULL if we are the pagedaemon. 670 */ 671 if (bp == NULL) { 672 KASSERT(curlwp == uvm.pagedaemon_lwp); 673 return NULL; 674 } 675 676 /* 677 * If buffer does not have data valid, start a read. 678 * Note that if buffer is BC_INVAL, getblk() won't return it. 679 * Therefore, it's valid if its I/O has completed or been delayed. 680 */ 681 if (!ISSET(bp->b_oflags, (BO_DONE | BO_DELWRI))) { 682 /* Start I/O for the buffer. */ 683 SET(bp->b_flags, B_READ | async); 684 if (async) 685 BIO_SETPRIO(bp, BPRIO_TIMELIMITED); 686 else 687 BIO_SETPRIO(bp, BPRIO_TIMECRITICAL); 688 VOP_STRATEGY(vp, bp); 689 690 /* Pay for the read. */ 691 curlwp->l_ru.ru_inblock++; 692 } else if (async) 693 brelse(bp, 0); 694 695 if (vp->v_type == VBLK) 696 mp = spec_node_getmountedfs(vp); 697 else 698 mp = vp->v_mount; 699 700 /* 701 * Collect statistics on synchronous and asynchronous reads. 702 * Reads from block devices are charged to their associated 703 * filesystem (if any). 704 */ 705 if (mp != NULL) { 706 if (async == 0) 707 mp->mnt_stat.f_syncreads++; 708 else 709 mp->mnt_stat.f_asyncreads++; 710 } 711 712 return (bp); 713 } 714 715 /* 716 * Read a disk block. 717 * This algorithm described in Bach (p.54). 718 */ 719 int 720 bread(struct vnode *vp, daddr_t blkno, int size, kauth_cred_t cred, 721 int flags, buf_t **bpp) 722 { 723 buf_t *bp; 724 int error; 725 726 /* Get buffer for block. */ 727 bp = *bpp = bio_doread(vp, blkno, size, cred, 0); 728 if (bp == NULL) 729 return ENOMEM; 730 731 /* Wait for the read to complete, and return result. */ 732 error = biowait(bp); 733 if (error == 0 && (flags & B_MODIFY) != 0) 734 error = fscow_run(bp, true); 735 if (error) { 736 brelse(bp, 0); 737 *bpp = NULL; 738 } 739 740 return error; 741 } 742 743 /* 744 * Read-ahead multiple disk blocks. The first is sync, the rest async. 745 * Trivial modification to the breada algorithm presented in Bach (p.55). 746 */ 747 int 748 breadn(struct vnode *vp, daddr_t blkno, int size, daddr_t *rablks, 749 int *rasizes, int nrablks, kauth_cred_t cred, int flags, buf_t **bpp) 750 { 751 buf_t *bp; 752 int error, i; 753 754 bp = *bpp = bio_doread(vp, blkno, size, cred, 0); 755 if (bp == NULL) 756 return ENOMEM; 757 758 /* 759 * For each of the read-ahead blocks, start a read, if necessary. 760 */ 761 mutex_enter(&bufcache_lock); 762 for (i = 0; i < nrablks; i++) { 763 /* If it's in the cache, just go on to next one. */ 764 if (incore(vp, rablks[i])) 765 continue; 766 767 /* Get a buffer for the read-ahead block */ 768 mutex_exit(&bufcache_lock); 769 (void) bio_doread(vp, rablks[i], rasizes[i], cred, B_ASYNC); 770 mutex_enter(&bufcache_lock); 771 } 772 mutex_exit(&bufcache_lock); 773 774 /* Otherwise, we had to start a read for it; wait until it's valid. */ 775 error = biowait(bp); 776 if (error == 0 && (flags & B_MODIFY) != 0) 777 error = fscow_run(bp, true); 778 if (error) { 779 brelse(bp, 0); 780 *bpp = NULL; 781 } 782 783 return error; 784 } 785 786 /* 787 * Block write. Described in Bach (p.56) 788 */ 789 int 790 bwrite(buf_t *bp) 791 { 792 int rv, sync, wasdelayed; 793 struct vnode *vp; 794 struct mount *mp; 795 796 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 797 KASSERT(!cv_has_waiters(&bp->b_done)); 798 799 vp = bp->b_vp; 800 if (vp != NULL) { 801 KASSERT(bp->b_objlock == vp->v_interlock); 802 if (vp->v_type == VBLK) 803 mp = spec_node_getmountedfs(vp); 804 else 805 mp = vp->v_mount; 806 } else { 807 mp = NULL; 808 } 809 810 if (mp && mp->mnt_wapbl) { 811 if (bp->b_iodone != mp->mnt_wapbl_op->wo_wapbl_biodone) { 812 bdwrite(bp); 813 return 0; 814 } 815 } 816 817 /* 818 * Remember buffer type, to switch on it later. If the write was 819 * synchronous, but the file system was mounted with MNT_ASYNC, 820 * convert it to a delayed write. 821 * XXX note that this relies on delayed tape writes being converted 822 * to async, not sync writes (which is safe, but ugly). 823 */ 824 sync = !ISSET(bp->b_flags, B_ASYNC); 825 if (sync && mp != NULL && ISSET(mp->mnt_flag, MNT_ASYNC)) { 826 bdwrite(bp); 827 return (0); 828 } 829 830 /* 831 * Collect statistics on synchronous and asynchronous writes. 832 * Writes to block devices are charged to their associated 833 * filesystem (if any). 834 */ 835 if (mp != NULL) { 836 if (sync) 837 mp->mnt_stat.f_syncwrites++; 838 else 839 mp->mnt_stat.f_asyncwrites++; 840 } 841 842 /* 843 * Pay for the I/O operation and make sure the buf is on the correct 844 * vnode queue. 845 */ 846 bp->b_error = 0; 847 wasdelayed = ISSET(bp->b_oflags, BO_DELWRI); 848 CLR(bp->b_flags, B_READ); 849 if (wasdelayed) { 850 mutex_enter(&bufcache_lock); 851 mutex_enter(bp->b_objlock); 852 CLR(bp->b_oflags, BO_DONE | BO_DELWRI); 853 reassignbuf(bp, bp->b_vp); 854 mutex_exit(&bufcache_lock); 855 } else { 856 curlwp->l_ru.ru_oublock++; 857 mutex_enter(bp->b_objlock); 858 CLR(bp->b_oflags, BO_DONE | BO_DELWRI); 859 } 860 if (vp != NULL) 861 vp->v_numoutput++; 862 mutex_exit(bp->b_objlock); 863 864 /* Initiate disk write. */ 865 if (sync) 866 BIO_SETPRIO(bp, BPRIO_TIMECRITICAL); 867 else 868 BIO_SETPRIO(bp, BPRIO_TIMELIMITED); 869 870 VOP_STRATEGY(vp, bp); 871 872 if (sync) { 873 /* If I/O was synchronous, wait for it to complete. */ 874 rv = biowait(bp); 875 876 /* Release the buffer. */ 877 brelse(bp, 0); 878 879 return (rv); 880 } else { 881 return (0); 882 } 883 } 884 885 int 886 vn_bwrite(void *v) 887 { 888 struct vop_bwrite_args *ap = v; 889 890 return (bwrite(ap->a_bp)); 891 } 892 893 /* 894 * Delayed write. 895 * 896 * The buffer is marked dirty, but is not queued for I/O. 897 * This routine should be used when the buffer is expected 898 * to be modified again soon, typically a small write that 899 * partially fills a buffer. 900 * 901 * NB: magnetic tapes cannot be delayed; they must be 902 * written in the order that the writes are requested. 903 * 904 * Described in Leffler, et al. (pp. 208-213). 905 */ 906 void 907 bdwrite(buf_t *bp) 908 { 909 910 KASSERT(bp->b_vp == NULL || bp->b_vp->v_tag != VT_UFS || 911 bp->b_vp->v_type == VBLK || ISSET(bp->b_flags, B_COWDONE)); 912 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 913 KASSERT(!cv_has_waiters(&bp->b_done)); 914 915 /* If this is a tape block, write the block now. */ 916 if (bdev_type(bp->b_dev) == D_TAPE) { 917 bawrite(bp); 918 return; 919 } 920 921 if (wapbl_vphaswapbl(bp->b_vp)) { 922 struct mount *mp = wapbl_vptomp(bp->b_vp); 923 924 if (bp->b_iodone != mp->mnt_wapbl_op->wo_wapbl_biodone) { 925 WAPBL_ADD_BUF(mp, bp); 926 } 927 } 928 929 /* 930 * If the block hasn't been seen before: 931 * (1) Mark it as having been seen, 932 * (2) Charge for the write, 933 * (3) Make sure it's on its vnode's correct block list. 934 */ 935 KASSERT(bp->b_vp == NULL || bp->b_objlock == bp->b_vp->v_interlock); 936 937 if (!ISSET(bp->b_oflags, BO_DELWRI)) { 938 mutex_enter(&bufcache_lock); 939 mutex_enter(bp->b_objlock); 940 SET(bp->b_oflags, BO_DELWRI); 941 curlwp->l_ru.ru_oublock++; 942 reassignbuf(bp, bp->b_vp); 943 mutex_exit(&bufcache_lock); 944 } else { 945 mutex_enter(bp->b_objlock); 946 } 947 /* Otherwise, the "write" is done, so mark and release the buffer. */ 948 CLR(bp->b_oflags, BO_DONE); 949 mutex_exit(bp->b_objlock); 950 951 brelse(bp, 0); 952 } 953 954 /* 955 * Asynchronous block write; just an asynchronous bwrite(). 956 */ 957 void 958 bawrite(buf_t *bp) 959 { 960 961 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 962 KASSERT(bp->b_vp != NULL); 963 964 SET(bp->b_flags, B_ASYNC); 965 VOP_BWRITE(bp->b_vp, bp); 966 } 967 968 /* 969 * Release a buffer on to the free lists. 970 * Described in Bach (p. 46). 971 */ 972 void 973 brelsel(buf_t *bp, int set) 974 { 975 struct bqueue *bufq; 976 struct vnode *vp; 977 978 KASSERT(bp != NULL); 979 KASSERT(mutex_owned(&bufcache_lock)); 980 KASSERT(!cv_has_waiters(&bp->b_done)); 981 KASSERT(bp->b_refcnt > 0); 982 983 SET(bp->b_cflags, set); 984 985 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 986 KASSERT(bp->b_iodone == NULL); 987 988 /* Wake up any processes waiting for any buffer to become free. */ 989 cv_signal(&needbuffer_cv); 990 991 /* Wake up any proceeses waiting for _this_ buffer to become */ 992 if (ISSET(bp->b_cflags, BC_WANTED)) 993 CLR(bp->b_cflags, BC_WANTED|BC_AGE); 994 995 /* If it's clean clear the copy-on-write flag. */ 996 if (ISSET(bp->b_flags, B_COWDONE)) { 997 mutex_enter(bp->b_objlock); 998 if (!ISSET(bp->b_oflags, BO_DELWRI)) 999 CLR(bp->b_flags, B_COWDONE); 1000 mutex_exit(bp->b_objlock); 1001 } 1002 1003 /* 1004 * Determine which queue the buffer should be on, then put it there. 1005 */ 1006 1007 /* If it's locked, don't report an error; try again later. */ 1008 if (ISSET(bp->b_flags, B_LOCKED)) 1009 bp->b_error = 0; 1010 1011 /* If it's not cacheable, or an error, mark it invalid. */ 1012 if (ISSET(bp->b_cflags, BC_NOCACHE) || bp->b_error != 0) 1013 SET(bp->b_cflags, BC_INVAL); 1014 1015 if (ISSET(bp->b_cflags, BC_VFLUSH)) { 1016 /* 1017 * This is a delayed write buffer that was just flushed to 1018 * disk. It is still on the LRU queue. If it's become 1019 * invalid, then we need to move it to a different queue; 1020 * otherwise leave it in its current position. 1021 */ 1022 CLR(bp->b_cflags, BC_VFLUSH); 1023 if (!ISSET(bp->b_cflags, BC_INVAL|BC_AGE) && 1024 !ISSET(bp->b_flags, B_LOCKED) && bp->b_error == 0) { 1025 KDASSERT(checkfreelist(bp, &bufqueues[BQ_LRU], 1)); 1026 goto already_queued; 1027 } else { 1028 bremfree(bp); 1029 } 1030 } 1031 1032 KDASSERT(checkfreelist(bp, &bufqueues[BQ_AGE], 0)); 1033 KDASSERT(checkfreelist(bp, &bufqueues[BQ_LRU], 0)); 1034 KDASSERT(checkfreelist(bp, &bufqueues[BQ_LOCKED], 0)); 1035 1036 if ((bp->b_bufsize <= 0) || ISSET(bp->b_cflags, BC_INVAL)) { 1037 /* 1038 * If it's invalid or empty, dissociate it from its vnode 1039 * and put on the head of the appropriate queue. 1040 */ 1041 if (ISSET(bp->b_flags, B_LOCKED)) { 1042 if (wapbl_vphaswapbl(vp = bp->b_vp)) { 1043 struct mount *mp = wapbl_vptomp(vp); 1044 1045 KASSERT(bp->b_iodone 1046 != mp->mnt_wapbl_op->wo_wapbl_biodone); 1047 WAPBL_REMOVE_BUF(mp, bp); 1048 } 1049 } 1050 1051 mutex_enter(bp->b_objlock); 1052 CLR(bp->b_oflags, BO_DONE|BO_DELWRI); 1053 if ((vp = bp->b_vp) != NULL) { 1054 KASSERT(bp->b_objlock == vp->v_interlock); 1055 reassignbuf(bp, bp->b_vp); 1056 brelvp(bp); 1057 mutex_exit(vp->v_interlock); 1058 } else { 1059 KASSERT(bp->b_objlock == &buffer_lock); 1060 mutex_exit(bp->b_objlock); 1061 } 1062 1063 if (bp->b_bufsize <= 0) 1064 /* no data */ 1065 goto already_queued; 1066 else 1067 /* invalid data */ 1068 bufq = &bufqueues[BQ_AGE]; 1069 binsheadfree(bp, bufq); 1070 } else { 1071 /* 1072 * It has valid data. Put it on the end of the appropriate 1073 * queue, so that it'll stick around for as long as possible. 1074 * If buf is AGE, but has dependencies, must put it on last 1075 * bufqueue to be scanned, ie LRU. This protects against the 1076 * livelock where BQ_AGE only has buffers with dependencies, 1077 * and we thus never get to the dependent buffers in BQ_LRU. 1078 */ 1079 if (ISSET(bp->b_flags, B_LOCKED)) { 1080 /* locked in core */ 1081 bufq = &bufqueues[BQ_LOCKED]; 1082 } else if (!ISSET(bp->b_cflags, BC_AGE)) { 1083 /* valid data */ 1084 bufq = &bufqueues[BQ_LRU]; 1085 } else { 1086 /* stale but valid data */ 1087 bufq = &bufqueues[BQ_AGE]; 1088 } 1089 binstailfree(bp, bufq); 1090 } 1091 already_queued: 1092 /* Unlock the buffer. */ 1093 CLR(bp->b_cflags, BC_AGE|BC_BUSY|BC_NOCACHE); 1094 CLR(bp->b_flags, B_ASYNC); 1095 cv_broadcast(&bp->b_busy); 1096 1097 if (bp->b_bufsize <= 0) 1098 brele(bp); 1099 } 1100 1101 void 1102 brelse(buf_t *bp, int set) 1103 { 1104 1105 mutex_enter(&bufcache_lock); 1106 brelsel(bp, set); 1107 mutex_exit(&bufcache_lock); 1108 } 1109 1110 /* 1111 * Determine if a block is in the cache. 1112 * Just look on what would be its hash chain. If it's there, return 1113 * a pointer to it, unless it's marked invalid. If it's marked invalid, 1114 * we normally don't return the buffer, unless the caller explicitly 1115 * wants us to. 1116 */ 1117 buf_t * 1118 incore(struct vnode *vp, daddr_t blkno) 1119 { 1120 buf_t *bp; 1121 1122 KASSERT(mutex_owned(&bufcache_lock)); 1123 1124 /* Search hash chain */ 1125 LIST_FOREACH(bp, BUFHASH(vp, blkno), b_hash) { 1126 if (bp->b_lblkno == blkno && bp->b_vp == vp && 1127 !ISSET(bp->b_cflags, BC_INVAL)) { 1128 KASSERT(bp->b_objlock == vp->v_interlock); 1129 return (bp); 1130 } 1131 } 1132 1133 return (NULL); 1134 } 1135 1136 /* 1137 * Get a block of requested size that is associated with 1138 * a given vnode and block offset. If it is found in the 1139 * block cache, mark it as having been found, make it busy 1140 * and return it. Otherwise, return an empty block of the 1141 * correct size. It is up to the caller to insure that the 1142 * cached blocks be of the correct size. 1143 */ 1144 buf_t * 1145 getblk(struct vnode *vp, daddr_t blkno, int size, int slpflag, int slptimeo) 1146 { 1147 int err, preserve; 1148 buf_t *bp; 1149 1150 mutex_enter(&bufcache_lock); 1151 loop: 1152 bp = incore(vp, blkno); 1153 if (bp != NULL) { 1154 err = bbusy(bp, ((slpflag & PCATCH) != 0), slptimeo, NULL); 1155 if (err != 0) { 1156 if (err == EPASSTHROUGH) 1157 goto loop; 1158 mutex_exit(&bufcache_lock); 1159 return (NULL); 1160 } 1161 KASSERT(!cv_has_waiters(&bp->b_done)); 1162 #ifdef DIAGNOSTIC 1163 if (ISSET(bp->b_oflags, BO_DONE|BO_DELWRI) && 1164 bp->b_bcount < size && vp->v_type != VBLK) 1165 panic("getblk: block size invariant failed"); 1166 #endif 1167 bremfree(bp); 1168 preserve = 1; 1169 } else { 1170 if ((bp = getnewbuf(slpflag, slptimeo, 0)) == NULL) 1171 goto loop; 1172 1173 if (incore(vp, blkno) != NULL) { 1174 /* The block has come into memory in the meantime. */ 1175 brelsel(bp, 0); 1176 goto loop; 1177 } 1178 1179 LIST_INSERT_HEAD(BUFHASH(vp, blkno), bp, b_hash); 1180 bp->b_blkno = bp->b_lblkno = bp->b_rawblkno = blkno; 1181 mutex_enter(vp->v_interlock); 1182 bgetvp(vp, bp); 1183 mutex_exit(vp->v_interlock); 1184 preserve = 0; 1185 } 1186 mutex_exit(&bufcache_lock); 1187 1188 /* 1189 * LFS can't track total size of B_LOCKED buffer (locked_queue_bytes) 1190 * if we re-size buffers here. 1191 */ 1192 if (ISSET(bp->b_flags, B_LOCKED)) { 1193 KASSERT(bp->b_bufsize >= size); 1194 } else { 1195 if (allocbuf(bp, size, preserve)) { 1196 mutex_enter(&bufcache_lock); 1197 LIST_REMOVE(bp, b_hash); 1198 mutex_exit(&bufcache_lock); 1199 brelse(bp, BC_INVAL); 1200 return NULL; 1201 } 1202 } 1203 BIO_SETPRIO(bp, BPRIO_DEFAULT); 1204 return (bp); 1205 } 1206 1207 /* 1208 * Get an empty, disassociated buffer of given size. 1209 */ 1210 buf_t * 1211 geteblk(int size) 1212 { 1213 buf_t *bp; 1214 int error __diagused; 1215 1216 mutex_enter(&bufcache_lock); 1217 while ((bp = getnewbuf(0, 0, 0)) == NULL) 1218 ; 1219 1220 SET(bp->b_cflags, BC_INVAL); 1221 LIST_INSERT_HEAD(&invalhash, bp, b_hash); 1222 mutex_exit(&bufcache_lock); 1223 BIO_SETPRIO(bp, BPRIO_DEFAULT); 1224 error = allocbuf(bp, size, 0); 1225 KASSERT(error == 0); 1226 return (bp); 1227 } 1228 1229 /* 1230 * Expand or contract the actual memory allocated to a buffer. 1231 * 1232 * If the buffer shrinks, data is lost, so it's up to the 1233 * caller to have written it out *first*; this routine will not 1234 * start a write. If the buffer grows, it's the callers 1235 * responsibility to fill out the buffer's additional contents. 1236 */ 1237 int 1238 allocbuf(buf_t *bp, int size, int preserve) 1239 { 1240 void *addr; 1241 vsize_t oldsize, desired_size; 1242 int oldcount; 1243 int delta; 1244 1245 desired_size = buf_roundsize(size); 1246 if (desired_size > MAXBSIZE) 1247 printf("allocbuf: buffer larger than MAXBSIZE requested"); 1248 1249 oldcount = bp->b_bcount; 1250 1251 bp->b_bcount = size; 1252 1253 oldsize = bp->b_bufsize; 1254 if (oldsize == desired_size) { 1255 /* 1256 * Do not short cut the WAPBL resize, as the buffer length 1257 * could still have changed and this would corrupt the 1258 * tracking of the transaction length. 1259 */ 1260 goto out; 1261 } 1262 1263 /* 1264 * If we want a buffer of a different size, re-allocate the 1265 * buffer's memory; copy old content only if needed. 1266 */ 1267 addr = buf_alloc(desired_size); 1268 if (addr == NULL) 1269 return ENOMEM; 1270 if (preserve) 1271 memcpy(addr, bp->b_data, MIN(oldsize,desired_size)); 1272 if (bp->b_data != NULL) 1273 buf_mrelease(bp->b_data, oldsize); 1274 bp->b_data = addr; 1275 bp->b_bufsize = desired_size; 1276 1277 /* 1278 * Update overall buffer memory counter (protected by bufcache_lock) 1279 */ 1280 delta = (long)desired_size - (long)oldsize; 1281 1282 mutex_enter(&bufcache_lock); 1283 if ((bufmem += delta) > bufmem_hiwater) { 1284 /* 1285 * Need to trim overall memory usage. 1286 */ 1287 while (buf_canrelease()) { 1288 if (curcpu()->ci_schedstate.spc_flags & 1289 SPCF_SHOULDYIELD) { 1290 mutex_exit(&bufcache_lock); 1291 preempt(); 1292 mutex_enter(&bufcache_lock); 1293 } 1294 if (buf_trim() == 0) 1295 break; 1296 } 1297 } 1298 mutex_exit(&bufcache_lock); 1299 1300 out: 1301 if (wapbl_vphaswapbl(bp->b_vp)) 1302 WAPBL_RESIZE_BUF(wapbl_vptomp(bp->b_vp), bp, oldsize, oldcount); 1303 1304 return 0; 1305 } 1306 1307 /* 1308 * Find a buffer which is available for use. 1309 * Select something from a free list. 1310 * Preference is to AGE list, then LRU list. 1311 * 1312 * Called with the buffer queues locked. 1313 * Return buffer locked. 1314 */ 1315 buf_t * 1316 getnewbuf(int slpflag, int slptimeo, int from_bufq) 1317 { 1318 buf_t *bp; 1319 struct vnode *vp; 1320 1321 start: 1322 KASSERT(mutex_owned(&bufcache_lock)); 1323 1324 /* 1325 * Get a new buffer from the pool. 1326 */ 1327 if (!from_bufq && buf_lotsfree()) { 1328 mutex_exit(&bufcache_lock); 1329 bp = pool_cache_get(buf_cache, PR_NOWAIT); 1330 if (bp != NULL) { 1331 memset((char *)bp, 0, sizeof(*bp)); 1332 buf_init(bp); 1333 SET(bp->b_cflags, BC_BUSY); /* mark buffer busy */ 1334 mutex_enter(&bufcache_lock); 1335 #if defined(DIAGNOSTIC) 1336 bp->b_freelistindex = -1; 1337 #endif /* defined(DIAGNOSTIC) */ 1338 return (bp); 1339 } 1340 mutex_enter(&bufcache_lock); 1341 } 1342 1343 KASSERT(mutex_owned(&bufcache_lock)); 1344 if ((bp = TAILQ_FIRST(&bufqueues[BQ_AGE].bq_queue)) != NULL || 1345 (bp = TAILQ_FIRST(&bufqueues[BQ_LRU].bq_queue)) != NULL) { 1346 KASSERT(!ISSET(bp->b_cflags, BC_BUSY) || ISSET(bp->b_cflags, BC_VFLUSH)); 1347 bremfree(bp); 1348 1349 /* Buffer is no longer on free lists. */ 1350 SET(bp->b_cflags, BC_BUSY); 1351 } else { 1352 /* 1353 * XXX: !from_bufq should be removed. 1354 */ 1355 if (!from_bufq || curlwp != uvm.pagedaemon_lwp) { 1356 /* wait for a free buffer of any kind */ 1357 if ((slpflag & PCATCH) != 0) 1358 (void)cv_timedwait_sig(&needbuffer_cv, 1359 &bufcache_lock, slptimeo); 1360 else 1361 (void)cv_timedwait(&needbuffer_cv, 1362 &bufcache_lock, slptimeo); 1363 } 1364 return (NULL); 1365 } 1366 1367 #ifdef DIAGNOSTIC 1368 if (bp->b_bufsize <= 0) 1369 panic("buffer %p: on queue but empty", bp); 1370 #endif 1371 1372 if (ISSET(bp->b_cflags, BC_VFLUSH)) { 1373 /* 1374 * This is a delayed write buffer being flushed to disk. Make 1375 * sure it gets aged out of the queue when it's finished, and 1376 * leave it off the LRU queue. 1377 */ 1378 CLR(bp->b_cflags, BC_VFLUSH); 1379 SET(bp->b_cflags, BC_AGE); 1380 goto start; 1381 } 1382 1383 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 1384 KASSERT(bp->b_refcnt > 0); 1385 KASSERT(!cv_has_waiters(&bp->b_done)); 1386 1387 /* 1388 * If buffer was a delayed write, start it and return NULL 1389 * (since we might sleep while starting the write). 1390 */ 1391 if (ISSET(bp->b_oflags, BO_DELWRI)) { 1392 /* 1393 * This buffer has gone through the LRU, so make sure it gets 1394 * reused ASAP. 1395 */ 1396 SET(bp->b_cflags, BC_AGE); 1397 mutex_exit(&bufcache_lock); 1398 bawrite(bp); 1399 mutex_enter(&bufcache_lock); 1400 return (NULL); 1401 } 1402 1403 vp = bp->b_vp; 1404 1405 /* clear out various other fields */ 1406 bp->b_cflags = BC_BUSY; 1407 bp->b_oflags = 0; 1408 bp->b_flags = 0; 1409 bp->b_dev = NODEV; 1410 bp->b_blkno = 0; 1411 bp->b_lblkno = 0; 1412 bp->b_rawblkno = 0; 1413 bp->b_iodone = 0; 1414 bp->b_error = 0; 1415 bp->b_resid = 0; 1416 bp->b_bcount = 0; 1417 1418 LIST_REMOVE(bp, b_hash); 1419 1420 /* Disassociate us from our vnode, if we had one... */ 1421 if (vp != NULL) { 1422 mutex_enter(vp->v_interlock); 1423 brelvp(bp); 1424 mutex_exit(vp->v_interlock); 1425 } 1426 1427 return (bp); 1428 } 1429 1430 /* 1431 * Attempt to free an aged buffer off the queues. 1432 * Called with queue lock held. 1433 * Returns the amount of buffer memory freed. 1434 */ 1435 static int 1436 buf_trim(void) 1437 { 1438 buf_t *bp; 1439 long size; 1440 1441 KASSERT(mutex_owned(&bufcache_lock)); 1442 1443 /* Instruct getnewbuf() to get buffers off the queues */ 1444 if ((bp = getnewbuf(PCATCH, 1, 1)) == NULL) 1445 return 0; 1446 1447 KASSERT((bp->b_cflags & BC_WANTED) == 0); 1448 size = bp->b_bufsize; 1449 bufmem -= size; 1450 if (size > 0) { 1451 buf_mrelease(bp->b_data, size); 1452 bp->b_bcount = bp->b_bufsize = 0; 1453 } 1454 /* brelse() will return the buffer to the global buffer pool */ 1455 brelsel(bp, 0); 1456 return size; 1457 } 1458 1459 int 1460 buf_drain(int n) 1461 { 1462 int size = 0, sz; 1463 1464 KASSERT(mutex_owned(&bufcache_lock)); 1465 1466 while (size < n && bufmem > bufmem_lowater) { 1467 sz = buf_trim(); 1468 if (sz <= 0) 1469 break; 1470 size += sz; 1471 } 1472 1473 return size; 1474 } 1475 1476 /* 1477 * Wait for operations on the buffer to complete. 1478 * When they do, extract and return the I/O's error value. 1479 */ 1480 int 1481 biowait(buf_t *bp) 1482 { 1483 1484 KASSERT(ISSET(bp->b_cflags, BC_BUSY)); 1485 KASSERT(bp->b_refcnt > 0); 1486 1487 mutex_enter(bp->b_objlock); 1488 while (!ISSET(bp->b_oflags, BO_DONE | BO_DELWRI)) 1489 cv_wait(&bp->b_done, bp->b_objlock); 1490 mutex_exit(bp->b_objlock); 1491 1492 return bp->b_error; 1493 } 1494 1495 /* 1496 * Mark I/O complete on a buffer. 1497 * 1498 * If a callback has been requested, e.g. the pageout 1499 * daemon, do so. Otherwise, awaken waiting processes. 1500 * 1501 * [ Leffler, et al., says on p.247: 1502 * "This routine wakes up the blocked process, frees the buffer 1503 * for an asynchronous write, or, for a request by the pagedaemon 1504 * process, invokes a procedure specified in the buffer structure" ] 1505 * 1506 * In real life, the pagedaemon (or other system processes) wants 1507 * to do async stuff to, and doesn't want the buffer brelse()'d. 1508 * (for swap pager, that puts swap buffers on the free lists (!!!), 1509 * for the vn device, that puts allocated buffers on the free lists!) 1510 */ 1511 void 1512 biodone(buf_t *bp) 1513 { 1514 int s; 1515 1516 KASSERT(!ISSET(bp->b_oflags, BO_DONE)); 1517 1518 if (cpu_intr_p()) { 1519 /* From interrupt mode: defer to a soft interrupt. */ 1520 s = splvm(); 1521 TAILQ_INSERT_TAIL(&curcpu()->ci_data.cpu_biodone, bp, b_actq); 1522 softint_schedule(biodone_sih); 1523 splx(s); 1524 } else { 1525 /* Process now - the buffer may be freed soon. */ 1526 biodone2(bp); 1527 } 1528 } 1529 1530 static void 1531 biodone2(buf_t *bp) 1532 { 1533 void (*callout)(buf_t *); 1534 1535 mutex_enter(bp->b_objlock); 1536 /* Note that the transfer is done. */ 1537 if (ISSET(bp->b_oflags, BO_DONE)) 1538 panic("biodone2 already"); 1539 CLR(bp->b_flags, B_COWDONE); 1540 SET(bp->b_oflags, BO_DONE); 1541 BIO_SETPRIO(bp, BPRIO_DEFAULT); 1542 1543 /* Wake up waiting writers. */ 1544 if (!ISSET(bp->b_flags, B_READ)) 1545 vwakeup(bp); 1546 1547 if ((callout = bp->b_iodone) != NULL) { 1548 /* Note callout done, then call out. */ 1549 KASSERT(!cv_has_waiters(&bp->b_done)); 1550 KERNEL_LOCK(1, NULL); /* XXXSMP */ 1551 bp->b_iodone = NULL; 1552 mutex_exit(bp->b_objlock); 1553 (*callout)(bp); 1554 KERNEL_UNLOCK_ONE(NULL); /* XXXSMP */ 1555 } else if (ISSET(bp->b_flags, B_ASYNC)) { 1556 /* If async, release. */ 1557 KASSERT(!cv_has_waiters(&bp->b_done)); 1558 mutex_exit(bp->b_objlock); 1559 brelse(bp, 0); 1560 } else { 1561 /* Otherwise just wake up waiters in biowait(). */ 1562 cv_broadcast(&bp->b_done); 1563 mutex_exit(bp->b_objlock); 1564 } 1565 } 1566 1567 static void 1568 biointr(void *cookie) 1569 { 1570 struct cpu_info *ci; 1571 buf_t *bp; 1572 int s; 1573 1574 ci = curcpu(); 1575 1576 while (!TAILQ_EMPTY(&ci->ci_data.cpu_biodone)) { 1577 KASSERT(curcpu() == ci); 1578 1579 s = splvm(); 1580 bp = TAILQ_FIRST(&ci->ci_data.cpu_biodone); 1581 TAILQ_REMOVE(&ci->ci_data.cpu_biodone, bp, b_actq); 1582 splx(s); 1583 1584 biodone2(bp); 1585 } 1586 } 1587 1588 /* 1589 * Wait for all buffers to complete I/O 1590 * Return the number of "stuck" buffers. 1591 */ 1592 int 1593 buf_syncwait(void) 1594 { 1595 buf_t *bp; 1596 int iter, nbusy, nbusy_prev = 0, ihash; 1597 1598 for (iter = 0; iter < 20;) { 1599 mutex_enter(&bufcache_lock); 1600 nbusy = 0; 1601 for (ihash = 0; ihash < bufhash+1; ihash++) { 1602 LIST_FOREACH(bp, &bufhashtbl[ihash], b_hash) { 1603 if ((bp->b_cflags & (BC_BUSY|BC_INVAL)) == BC_BUSY) 1604 nbusy += ((bp->b_flags & B_READ) == 0); 1605 } 1606 } 1607 mutex_exit(&bufcache_lock); 1608 1609 if (nbusy == 0) 1610 break; 1611 if (nbusy_prev == 0) 1612 nbusy_prev = nbusy; 1613 printf("%d ", nbusy); 1614 kpause("bflush", false, MAX(1, hz / 25 * iter), NULL); 1615 if (nbusy >= nbusy_prev) /* we didn't flush anything */ 1616 iter++; 1617 else 1618 nbusy_prev = nbusy; 1619 } 1620 1621 if (nbusy) { 1622 #if defined(DEBUG) || defined(DEBUG_HALT_BUSY) 1623 printf("giving up\nPrinting vnodes for busy buffers\n"); 1624 for (ihash = 0; ihash < bufhash+1; ihash++) { 1625 LIST_FOREACH(bp, &bufhashtbl[ihash], b_hash) { 1626 if ((bp->b_cflags & (BC_BUSY|BC_INVAL)) == BC_BUSY && 1627 (bp->b_flags & B_READ) == 0) 1628 vprint(NULL, bp->b_vp); 1629 } 1630 } 1631 #endif 1632 } 1633 1634 return nbusy; 1635 } 1636 1637 static void 1638 sysctl_fillbuf(buf_t *i, struct buf_sysctl *o) 1639 { 1640 1641 o->b_flags = i->b_flags | i->b_cflags | i->b_oflags; 1642 o->b_error = i->b_error; 1643 o->b_prio = i->b_prio; 1644 o->b_dev = i->b_dev; 1645 o->b_bufsize = i->b_bufsize; 1646 o->b_bcount = i->b_bcount; 1647 o->b_resid = i->b_resid; 1648 o->b_addr = PTRTOUINT64(i->b_data); 1649 o->b_blkno = i->b_blkno; 1650 o->b_rawblkno = i->b_rawblkno; 1651 o->b_iodone = PTRTOUINT64(i->b_iodone); 1652 o->b_proc = PTRTOUINT64(i->b_proc); 1653 o->b_vp = PTRTOUINT64(i->b_vp); 1654 o->b_saveaddr = PTRTOUINT64(i->b_saveaddr); 1655 o->b_lblkno = i->b_lblkno; 1656 } 1657 1658 #define KERN_BUFSLOP 20 1659 static int 1660 sysctl_dobuf(SYSCTLFN_ARGS) 1661 { 1662 buf_t *bp; 1663 struct buf_sysctl bs; 1664 struct bqueue *bq; 1665 char *dp; 1666 u_int i, op, arg; 1667 size_t len, needed, elem_size, out_size; 1668 int error, elem_count, retries; 1669 1670 if (namelen == 1 && name[0] == CTL_QUERY) 1671 return (sysctl_query(SYSCTLFN_CALL(rnode))); 1672 1673 if (namelen != 4) 1674 return (EINVAL); 1675 1676 retries = 100; 1677 retry: 1678 dp = oldp; 1679 len = (oldp != NULL) ? *oldlenp : 0; 1680 op = name[0]; 1681 arg = name[1]; 1682 elem_size = name[2]; 1683 elem_count = name[3]; 1684 out_size = MIN(sizeof(bs), elem_size); 1685 1686 /* 1687 * at the moment, these are just "placeholders" to make the 1688 * API for retrieving kern.buf data more extensible in the 1689 * future. 1690 * 1691 * XXX kern.buf currently has "netbsd32" issues. hopefully 1692 * these will be resolved at a later point. 1693 */ 1694 if (op != KERN_BUF_ALL || arg != KERN_BUF_ALL || 1695 elem_size < 1 || elem_count < 0) 1696 return (EINVAL); 1697 1698 error = 0; 1699 needed = 0; 1700 sysctl_unlock(); 1701 mutex_enter(&bufcache_lock); 1702 for (i = 0; i < BQUEUES; i++) { 1703 bq = &bufqueues[i]; 1704 TAILQ_FOREACH(bp, &bq->bq_queue, b_freelist) { 1705 bq->bq_marker = bp; 1706 if (len >= elem_size && elem_count > 0) { 1707 sysctl_fillbuf(bp, &bs); 1708 mutex_exit(&bufcache_lock); 1709 error = copyout(&bs, dp, out_size); 1710 mutex_enter(&bufcache_lock); 1711 if (error) 1712 break; 1713 if (bq->bq_marker != bp) { 1714 /* 1715 * This sysctl node is only for 1716 * statistics. Retry; if the 1717 * queue keeps changing, then 1718 * bail out. 1719 */ 1720 if (retries-- == 0) { 1721 error = EAGAIN; 1722 break; 1723 } 1724 mutex_exit(&bufcache_lock); 1725 sysctl_relock(); 1726 goto retry; 1727 } 1728 dp += elem_size; 1729 len -= elem_size; 1730 } 1731 needed += elem_size; 1732 if (elem_count > 0 && elem_count != INT_MAX) 1733 elem_count--; 1734 } 1735 if (error != 0) 1736 break; 1737 } 1738 mutex_exit(&bufcache_lock); 1739 sysctl_relock(); 1740 1741 *oldlenp = needed; 1742 if (oldp == NULL) 1743 *oldlenp += KERN_BUFSLOP * sizeof(buf_t); 1744 1745 return (error); 1746 } 1747 1748 static int 1749 sysctl_bufvm_update(SYSCTLFN_ARGS) 1750 { 1751 int error, rv; 1752 struct sysctlnode node; 1753 unsigned int temp_bufcache; 1754 unsigned long temp_water; 1755 1756 /* Take a copy of the supplied node and its data */ 1757 node = *rnode; 1758 if (node.sysctl_data == &bufcache) { 1759 node.sysctl_data = &temp_bufcache; 1760 temp_bufcache = *(unsigned int *)rnode->sysctl_data; 1761 } else { 1762 node.sysctl_data = &temp_water; 1763 temp_water = *(unsigned long *)rnode->sysctl_data; 1764 } 1765 1766 /* Update the copy */ 1767 error = sysctl_lookup(SYSCTLFN_CALL(&node)); 1768 if (error || newp == NULL) 1769 return (error); 1770 1771 if (rnode->sysctl_data == &bufcache) { 1772 if (temp_bufcache > 100) 1773 return (EINVAL); 1774 bufcache = temp_bufcache; 1775 buf_setwm(); 1776 } else if (rnode->sysctl_data == &bufmem_lowater) { 1777 if (bufmem_hiwater - temp_water < 16) 1778 return (EINVAL); 1779 bufmem_lowater = temp_water; 1780 } else if (rnode->sysctl_data == &bufmem_hiwater) { 1781 if (temp_water - bufmem_lowater < 16) 1782 return (EINVAL); 1783 bufmem_hiwater = temp_water; 1784 } else 1785 return (EINVAL); 1786 1787 /* Drain until below new high water mark */ 1788 sysctl_unlock(); 1789 mutex_enter(&bufcache_lock); 1790 while (bufmem > bufmem_hiwater) { 1791 rv = buf_drain((bufmem - bufmem_hiwater) / (2 * 1024)); 1792 if (rv <= 0) 1793 break; 1794 } 1795 mutex_exit(&bufcache_lock); 1796 sysctl_relock(); 1797 1798 return 0; 1799 } 1800 1801 static struct sysctllog *vfsbio_sysctllog; 1802 1803 static void 1804 sysctl_kern_buf_setup(void) 1805 { 1806 1807 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1808 CTLFLAG_PERMANENT, 1809 CTLTYPE_NODE, "kern", NULL, 1810 NULL, 0, NULL, 0, 1811 CTL_KERN, CTL_EOL); 1812 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1813 CTLFLAG_PERMANENT, 1814 CTLTYPE_NODE, "buf", 1815 SYSCTL_DESCR("Kernel buffer cache information"), 1816 sysctl_dobuf, 0, NULL, 0, 1817 CTL_KERN, KERN_BUF, CTL_EOL); 1818 } 1819 1820 static void 1821 sysctl_vm_buf_setup(void) 1822 { 1823 1824 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1825 CTLFLAG_PERMANENT, 1826 CTLTYPE_NODE, "vm", NULL, 1827 NULL, 0, NULL, 0, 1828 CTL_VM, CTL_EOL); 1829 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1830 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 1831 CTLTYPE_INT, "bufcache", 1832 SYSCTL_DESCR("Percentage of physical memory to use for " 1833 "buffer cache"), 1834 sysctl_bufvm_update, 0, &bufcache, 0, 1835 CTL_VM, CTL_CREATE, CTL_EOL); 1836 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1837 CTLFLAG_PERMANENT|CTLFLAG_READONLY, 1838 CTLTYPE_LONG, "bufmem", 1839 SYSCTL_DESCR("Amount of kernel memory used by buffer " 1840 "cache"), 1841 NULL, 0, &bufmem, 0, 1842 CTL_VM, CTL_CREATE, CTL_EOL); 1843 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1844 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 1845 CTLTYPE_LONG, "bufmem_lowater", 1846 SYSCTL_DESCR("Minimum amount of kernel memory to " 1847 "reserve for buffer cache"), 1848 sysctl_bufvm_update, 0, &bufmem_lowater, 0, 1849 CTL_VM, CTL_CREATE, CTL_EOL); 1850 sysctl_createv(&vfsbio_sysctllog, 0, NULL, NULL, 1851 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 1852 CTLTYPE_LONG, "bufmem_hiwater", 1853 SYSCTL_DESCR("Maximum amount of kernel memory to use " 1854 "for buffer cache"), 1855 sysctl_bufvm_update, 0, &bufmem_hiwater, 0, 1856 CTL_VM, CTL_CREATE, CTL_EOL); 1857 } 1858 1859 #ifdef DEBUG 1860 /* 1861 * Print out statistics on the current allocation of the buffer pool. 1862 * Can be enabled to print out on every ``sync'' by setting "syncprt" 1863 * in vfs_syscalls.c using sysctl. 1864 */ 1865 void 1866 vfs_bufstats(void) 1867 { 1868 int i, j, count; 1869 buf_t *bp; 1870 struct bqueue *dp; 1871 int counts[(MAXBSIZE / PAGE_SIZE) + 1]; 1872 static const char *bname[BQUEUES] = { "LOCKED", "LRU", "AGE" }; 1873 1874 for (dp = bufqueues, i = 0; dp < &bufqueues[BQUEUES]; dp++, i++) { 1875 count = 0; 1876 for (j = 0; j <= MAXBSIZE/PAGE_SIZE; j++) 1877 counts[j] = 0; 1878 TAILQ_FOREACH(bp, &dp->bq_queue, b_freelist) { 1879 counts[bp->b_bufsize/PAGE_SIZE]++; 1880 count++; 1881 } 1882 printf("%s: total-%d", bname[i], count); 1883 for (j = 0; j <= MAXBSIZE/PAGE_SIZE; j++) 1884 if (counts[j] != 0) 1885 printf(", %d-%d", j * PAGE_SIZE, counts[j]); 1886 printf("\n"); 1887 } 1888 } 1889 #endif /* DEBUG */ 1890 1891 /* ------------------------------ */ 1892 1893 buf_t * 1894 getiobuf(struct vnode *vp, bool waitok) 1895 { 1896 buf_t *bp; 1897 1898 bp = pool_cache_get(bufio_cache, (waitok ? PR_WAITOK : PR_NOWAIT)); 1899 if (bp == NULL) 1900 return bp; 1901 1902 buf_init(bp); 1903 1904 if ((bp->b_vp = vp) == NULL) 1905 bp->b_objlock = &buffer_lock; 1906 else 1907 bp->b_objlock = vp->v_interlock; 1908 1909 return bp; 1910 } 1911 1912 void 1913 putiobuf(buf_t *bp) 1914 { 1915 1916 buf_destroy(bp); 1917 pool_cache_put(bufio_cache, bp); 1918 } 1919 1920 /* 1921 * nestiobuf_iodone: b_iodone callback for nested buffers. 1922 */ 1923 1924 void 1925 nestiobuf_iodone(buf_t *bp) 1926 { 1927 buf_t *mbp = bp->b_private; 1928 int error; 1929 int donebytes; 1930 1931 KASSERT(bp->b_bcount <= bp->b_bufsize); 1932 KASSERT(mbp != bp); 1933 1934 error = bp->b_error; 1935 if (bp->b_error == 0 && 1936 (bp->b_bcount < bp->b_bufsize || bp->b_resid > 0)) { 1937 /* 1938 * Not all got transfered, raise an error. We have no way to 1939 * propagate these conditions to mbp. 1940 */ 1941 error = EIO; 1942 } 1943 1944 donebytes = bp->b_bufsize; 1945 1946 putiobuf(bp); 1947 nestiobuf_done(mbp, donebytes, error); 1948 } 1949 1950 /* 1951 * nestiobuf_setup: setup a "nested" buffer. 1952 * 1953 * => 'mbp' is a "master" buffer which is being divided into sub pieces. 1954 * => 'bp' should be a buffer allocated by getiobuf. 1955 * => 'offset' is a byte offset in the master buffer. 1956 * => 'size' is a size in bytes of this nested buffer. 1957 */ 1958 1959 void 1960 nestiobuf_setup(buf_t *mbp, buf_t *bp, int offset, size_t size) 1961 { 1962 const int b_read = mbp->b_flags & B_READ; 1963 struct vnode *vp = mbp->b_vp; 1964 1965 KASSERT(mbp->b_bcount >= offset + size); 1966 bp->b_vp = vp; 1967 bp->b_dev = mbp->b_dev; 1968 bp->b_objlock = mbp->b_objlock; 1969 bp->b_cflags = BC_BUSY; 1970 bp->b_flags = B_ASYNC | b_read; 1971 bp->b_iodone = nestiobuf_iodone; 1972 bp->b_data = (char *)mbp->b_data + offset; 1973 bp->b_resid = bp->b_bcount = size; 1974 bp->b_bufsize = bp->b_bcount; 1975 bp->b_private = mbp; 1976 BIO_COPYPRIO(bp, mbp); 1977 if (!b_read && vp != NULL) { 1978 mutex_enter(vp->v_interlock); 1979 vp->v_numoutput++; 1980 mutex_exit(vp->v_interlock); 1981 } 1982 } 1983 1984 /* 1985 * nestiobuf_done: propagate completion to the master buffer. 1986 * 1987 * => 'donebytes' specifies how many bytes in the 'mbp' is completed. 1988 * => 'error' is an errno(2) that 'donebytes' has been completed with. 1989 */ 1990 1991 void 1992 nestiobuf_done(buf_t *mbp, int donebytes, int error) 1993 { 1994 1995 if (donebytes == 0) { 1996 return; 1997 } 1998 mutex_enter(mbp->b_objlock); 1999 KASSERT(mbp->b_resid >= donebytes); 2000 mbp->b_resid -= donebytes; 2001 if (error) 2002 mbp->b_error = error; 2003 if (mbp->b_resid == 0) { 2004 if (mbp->b_error) 2005 mbp->b_resid = mbp->b_bcount; 2006 mutex_exit(mbp->b_objlock); 2007 biodone(mbp); 2008 } else 2009 mutex_exit(mbp->b_objlock); 2010 } 2011 2012 void 2013 buf_init(buf_t *bp) 2014 { 2015 2016 cv_init(&bp->b_busy, "biolock"); 2017 cv_init(&bp->b_done, "biowait"); 2018 bp->b_dev = NODEV; 2019 bp->b_error = 0; 2020 bp->b_flags = 0; 2021 bp->b_cflags = 0; 2022 bp->b_oflags = 0; 2023 bp->b_objlock = &buffer_lock; 2024 bp->b_iodone = NULL; 2025 bp->b_refcnt = 1; 2026 bp->b_dev = NODEV; 2027 bp->b_vnbufs.le_next = NOLIST; 2028 BIO_SETPRIO(bp, BPRIO_DEFAULT); 2029 } 2030 2031 void 2032 buf_destroy(buf_t *bp) 2033 { 2034 2035 cv_destroy(&bp->b_done); 2036 cv_destroy(&bp->b_busy); 2037 } 2038 2039 int 2040 bbusy(buf_t *bp, bool intr, int timo, kmutex_t *interlock) 2041 { 2042 int error; 2043 2044 KASSERT(mutex_owned(&bufcache_lock)); 2045 2046 if ((bp->b_cflags & BC_BUSY) != 0) { 2047 if (curlwp == uvm.pagedaemon_lwp) 2048 return EDEADLK; 2049 bp->b_cflags |= BC_WANTED; 2050 bref(bp); 2051 if (interlock != NULL) 2052 mutex_exit(interlock); 2053 if (intr) { 2054 error = cv_timedwait_sig(&bp->b_busy, &bufcache_lock, 2055 timo); 2056 } else { 2057 error = cv_timedwait(&bp->b_busy, &bufcache_lock, 2058 timo); 2059 } 2060 brele(bp); 2061 if (interlock != NULL) 2062 mutex_enter(interlock); 2063 if (error != 0) 2064 return error; 2065 return EPASSTHROUGH; 2066 } 2067 bp->b_cflags |= BC_BUSY; 2068 2069 return 0; 2070 } 2071