Blame view
fs/bio.c
40.9 KB
1da177e4c
|
1 |
/* |
0fe234795
|
2 |
* Copyright (C) 2001 Jens Axboe <axboe@kernel.dk> |
1da177e4c
|
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 |
* * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public Licens * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111- * */ #include <linux/mm.h> #include <linux/swap.h> #include <linux/bio.h> #include <linux/blkdev.h> |
852c788f8
|
22 |
#include <linux/iocontext.h> |
1da177e4c
|
23 24 25 |
#include <linux/slab.h> #include <linux/init.h> #include <linux/kernel.h> |
630d9c472
|
26 |
#include <linux/export.h> |
1da177e4c
|
27 28 |
#include <linux/mempool.h> #include <linux/workqueue.h> |
852c788f8
|
29 |
#include <linux/cgroup.h> |
f1970baf6
|
30 |
#include <scsi/sg.h> /* for struct sg_iovec */ |
1da177e4c
|
31 |
|
55782138e
|
32 |
#include <trace/events/block.h> |
0bfc24559
|
33 |
|
392ddc329
|
34 35 36 37 38 |
/* * Test patch to inline a certain number of bi_io_vec's inside the bio * itself, to shrink a bio data allocation from two mempool calls to one */ #define BIO_INLINE_VECS 4 |
6feef531f
|
39 |
static mempool_t *bio_split_pool __read_mostly; |
1da177e4c
|
40 |
|
1da177e4c
|
41 42 43 44 45 |
/* * if you change this list, also change bvec_alloc or things will * break badly! cannot be bigger than what you can fit into an * unsigned short */ |
1da177e4c
|
46 |
#define BV(x) { .nr_vecs = x, .name = "biovec-"__stringify(x) } |
df6771402
|
47 |
static struct biovec_slab bvec_slabs[BIOVEC_NR_POOLS] __read_mostly = { |
1da177e4c
|
48 49 50 51 52 |
BV(1), BV(4), BV(16), BV(64), BV(128), BV(BIO_MAX_PAGES), }; #undef BV /* |
1da177e4c
|
53 54 55 |
* fs_bio_set is the bio_set containing bio and iovec memory pools used by * IO code that does not need private memory pools. */ |
51d654e1d
|
56 |
struct bio_set *fs_bio_set; |
1da177e4c
|
57 |
|
bb799ca02
|
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 |
/* * Our slab pool management */ struct bio_slab { struct kmem_cache *slab; unsigned int slab_ref; unsigned int slab_size; char name[8]; }; static DEFINE_MUTEX(bio_slab_lock); static struct bio_slab *bio_slabs; static unsigned int bio_slab_nr, bio_slab_max; static struct kmem_cache *bio_find_or_create_slab(unsigned int extra_size) { unsigned int sz = sizeof(struct bio) + extra_size; struct kmem_cache *slab = NULL; struct bio_slab *bslab; unsigned int i, entry = -1; mutex_lock(&bio_slab_lock); i = 0; while (i < bio_slab_nr) { |
f06f135d8
|
82 |
bslab = &bio_slabs[i]; |
bb799ca02
|
83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 |
if (!bslab->slab && entry == -1) entry = i; else if (bslab->slab_size == sz) { slab = bslab->slab; bslab->slab_ref++; break; } i++; } if (slab) goto out_unlock; if (bio_slab_nr == bio_slab_max && entry == -1) { bio_slab_max <<= 1; bio_slabs = krealloc(bio_slabs, bio_slab_max * sizeof(struct bio_slab), GFP_KERNEL); if (!bio_slabs) goto out_unlock; } if (entry == -1) entry = bio_slab_nr++; bslab = &bio_slabs[entry]; snprintf(bslab->name, sizeof(bslab->name), "bio-%d", entry); slab = kmem_cache_create(bslab->name, sz, 0, SLAB_HWCACHE_ALIGN, NULL); if (!slab) goto out_unlock; |
80cdc6dae
|
114 115 |
printk(KERN_INFO "bio: create slab <%s> at %d ", bslab->name, entry); |
bb799ca02
|
116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 |
bslab->slab = slab; bslab->slab_ref = 1; bslab->slab_size = sz; out_unlock: mutex_unlock(&bio_slab_lock); return slab; } static void bio_put_slab(struct bio_set *bs) { struct bio_slab *bslab = NULL; unsigned int i; mutex_lock(&bio_slab_lock); for (i = 0; i < bio_slab_nr; i++) { if (bs->bio_slab == bio_slabs[i].slab) { bslab = &bio_slabs[i]; break; } } if (WARN(!bslab, KERN_ERR "bio: unable to find slab! ")) goto out; WARN_ON(!bslab->slab_ref); if (--bslab->slab_ref) goto out; kmem_cache_destroy(bslab->slab); bslab->slab = NULL; out: mutex_unlock(&bio_slab_lock); } |
7ba1ba12e
|
153 154 155 156 |
unsigned int bvec_nr_vecs(unsigned short idx) { return bvec_slabs[idx].nr_vecs; } |
bb799ca02
|
157 158 159 160 161 162 163 164 165 166 167 168 |
void bvec_free_bs(struct bio_set *bs, struct bio_vec *bv, unsigned int idx) { BIO_BUG_ON(idx >= BIOVEC_NR_POOLS); if (idx == BIOVEC_MAX_IDX) mempool_free(bv, bs->bvec_pool); else { struct biovec_slab *bvs = bvec_slabs + idx; kmem_cache_free(bvs->slab, bv); } } |
7ff9345ff
|
169 170 |
struct bio_vec *bvec_alloc_bs(gfp_t gfp_mask, int nr, unsigned long *idx, struct bio_set *bs) |
1da177e4c
|
171 172 |
{ struct bio_vec *bvl; |
1da177e4c
|
173 174 |
/* |
7ff9345ff
|
175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 |
* see comment near bvec_array define! */ switch (nr) { case 1: *idx = 0; break; case 2 ... 4: *idx = 1; break; case 5 ... 16: *idx = 2; break; case 17 ... 64: *idx = 3; break; case 65 ... 128: *idx = 4; break; case 129 ... BIO_MAX_PAGES: *idx = 5; break; default: return NULL; } /* * idx now points to the pool we want to allocate from. only the * 1-vec entry pool is mempool backed. */ if (*idx == BIOVEC_MAX_IDX) { fallback: bvl = mempool_alloc(bs->bvec_pool, gfp_mask); } else { struct biovec_slab *bvs = bvec_slabs + *idx; gfp_t __gfp_mask = gfp_mask & ~(__GFP_WAIT | __GFP_IO); |
0a0d96b03
|
210 |
/* |
7ff9345ff
|
211 212 213 |
* Make this allocation restricted and don't dump info on * allocation failures, since we'll fallback to the mempool * in case of failure. |
0a0d96b03
|
214 |
*/ |
7ff9345ff
|
215 |
__gfp_mask |= __GFP_NOMEMALLOC | __GFP_NORETRY | __GFP_NOWARN; |
1da177e4c
|
216 |
|
0a0d96b03
|
217 |
/* |
7ff9345ff
|
218 219 |
* Try a slab allocation. If this fails and __GFP_WAIT * is set, retry with the 1-entry mempool |
0a0d96b03
|
220 |
*/ |
7ff9345ff
|
221 222 223 224 225 226 |
bvl = kmem_cache_alloc(bvs->slab, __gfp_mask); if (unlikely(!bvl && (gfp_mask & __GFP_WAIT))) { *idx = BIOVEC_MAX_IDX; goto fallback; } } |
1da177e4c
|
227 228 |
return bvl; } |
7ff9345ff
|
229 |
void bio_free(struct bio *bio, struct bio_set *bs) |
1da177e4c
|
230 |
{ |
bb799ca02
|
231 |
void *p; |
1da177e4c
|
232 |
|
392ddc329
|
233 |
if (bio_has_allocated_vec(bio)) |
bb799ca02
|
234 |
bvec_free_bs(bs, bio->bi_io_vec, BIO_POOL_IDX(bio)); |
1da177e4c
|
235 |
|
7ba1ba12e
|
236 |
if (bio_integrity(bio)) |
7878cba9f
|
237 |
bio_integrity_free(bio, bs); |
7ba1ba12e
|
238 |
|
bb799ca02
|
239 240 241 242 243 244 245 246 |
/* * If we have front padding, adjust the bio pointer before freeing */ p = bio; if (bs->front_pad) p -= bs->front_pad; mempool_free(p, bs->bio_pool); |
3676347a5
|
247 |
} |
a112a71d4
|
248 |
EXPORT_SYMBOL(bio_free); |
3676347a5
|
249 |
|
858119e15
|
250 |
void bio_init(struct bio *bio) |
1da177e4c
|
251 |
{ |
2b94de552
|
252 |
memset(bio, 0, sizeof(*bio)); |
1da177e4c
|
253 |
bio->bi_flags = 1 << BIO_UPTODATE; |
1da177e4c
|
254 |
atomic_set(&bio->bi_cnt, 1); |
1da177e4c
|
255 |
} |
a112a71d4
|
256 |
EXPORT_SYMBOL(bio_init); |
1da177e4c
|
257 258 259 260 261 |
/** * bio_alloc_bioset - allocate a bio for I/O * @gfp_mask: the GFP_ mask given to the slab allocator * @nr_iovecs: number of iovecs to pre-allocate |
db18efac0
|
262 |
* @bs: the bio_set to allocate from. |
1da177e4c
|
263 264 |
* * Description: |
db18efac0
|
265 |
* bio_alloc_bioset will try its own mempool to satisfy the allocation. |
1da177e4c
|
266 |
* If %__GFP_WAIT is set then we will block on the internal pool waiting |
db18efac0
|
267 |
* for a &struct bio to become free. |
1da177e4c
|
268 |
* |
af901ca18
|
269 |
* Note that the caller must set ->bi_destructor on successful return |
bb799ca02
|
270 271 |
* of a bio, to do the appropriate freeing of the bio once the reference * count drops to zero. |
1da177e4c
|
272 |
**/ |
dd0fc66fb
|
273 |
struct bio *bio_alloc_bioset(gfp_t gfp_mask, int nr_iovecs, struct bio_set *bs) |
1da177e4c
|
274 |
{ |
451a9ebf6
|
275 |
unsigned long idx = BIO_POOL_NONE; |
34053979f
|
276 |
struct bio_vec *bvl = NULL; |
451a9ebf6
|
277 278 279 280 281 282 283 |
struct bio *bio; void *p; p = mempool_alloc(bs->bio_pool, gfp_mask); if (unlikely(!p)) return NULL; bio = p + bs->front_pad; |
1da177e4c
|
284 |
|
34053979f
|
285 286 287 288 289 290 291 292 293 294 295 296 297 298 |
bio_init(bio); if (unlikely(!nr_iovecs)) goto out_set; if (nr_iovecs <= BIO_INLINE_VECS) { bvl = bio->bi_inline_vecs; nr_iovecs = BIO_INLINE_VECS; } else { bvl = bvec_alloc_bs(gfp_mask, nr_iovecs, &idx, bs); if (unlikely(!bvl)) goto err_free; nr_iovecs = bvec_nr_vecs(idx); |
1da177e4c
|
299 |
} |
451a9ebf6
|
300 |
out_set: |
34053979f
|
301 302 |
bio->bi_flags |= idx << BIO_POOL_OFFSET; bio->bi_max_vecs = nr_iovecs; |
34053979f
|
303 |
bio->bi_io_vec = bvl; |
1da177e4c
|
304 |
return bio; |
34053979f
|
305 306 |
err_free: |
451a9ebf6
|
307 |
mempool_free(p, bs->bio_pool); |
34053979f
|
308 |
return NULL; |
1da177e4c
|
309 |
} |
a112a71d4
|
310 |
EXPORT_SYMBOL(bio_alloc_bioset); |
1da177e4c
|
311 |
|
451a9ebf6
|
312 313 314 315 316 317 318 319 320 321 |
static void bio_fs_destructor(struct bio *bio) { bio_free(bio, fs_bio_set); } /** * bio_alloc - allocate a new bio, memory pool backed * @gfp_mask: allocation mask to use * @nr_iovecs: number of iovecs * |
5f04eeb8a
|
322 323 324 325 326 327 328 329 330 331 |
* bio_alloc will allocate a bio and associated bio_vec array that can hold * at least @nr_iovecs entries. Allocations will be done from the * fs_bio_set. Also see @bio_alloc_bioset and @bio_kmalloc. * * If %__GFP_WAIT is set, then bio_alloc will always be able to allocate * a bio. This is due to the mempool guarantees. To make this work, callers * must never allocate more than 1 bio at a time from this pool. Callers * that need to allocate more than 1 bio must always submit the previously * allocated bio for IO before attempting to allocate a new one. Failure to * do so can cause livelocks under memory pressure. |
451a9ebf6
|
332 333 334 335 |
* * RETURNS: * Pointer to new bio on success, NULL on failure. */ |
121f09941
|
336 |
struct bio *bio_alloc(gfp_t gfp_mask, unsigned int nr_iovecs) |
451a9ebf6
|
337 338 339 340 341 342 343 344 |
{ struct bio *bio = bio_alloc_bioset(gfp_mask, nr_iovecs, fs_bio_set); if (bio) bio->bi_destructor = bio_fs_destructor; return bio; } |
a112a71d4
|
345 |
EXPORT_SYMBOL(bio_alloc); |
451a9ebf6
|
346 347 348 349 |
static void bio_kmalloc_destructor(struct bio *bio) { if (bio_integrity(bio)) |
7878cba9f
|
350 |
bio_integrity_free(bio, fs_bio_set); |
451a9ebf6
|
351 352 |
kfree(bio); } |
86c824b94
|
353 |
/** |
5f04eeb8a
|
354 |
* bio_kmalloc - allocate a bio for I/O using kmalloc() |
86c824b94
|
355 356 357 358 |
* @gfp_mask: the GFP_ mask given to the slab allocator * @nr_iovecs: number of iovecs to pre-allocate * * Description: |
5f04eeb8a
|
359 360 |
* Allocate a new bio with @nr_iovecs bvecs. If @gfp_mask contains * %__GFP_WAIT, the allocation is guaranteed to succeed. |
86c824b94
|
361 362 |
* **/ |
121f09941
|
363 |
struct bio *bio_kmalloc(gfp_t gfp_mask, unsigned int nr_iovecs) |
0a0d96b03
|
364 |
{ |
451a9ebf6
|
365 |
struct bio *bio; |
0a0d96b03
|
366 |
|
f3f63c1c2
|
367 368 |
if (nr_iovecs > UIO_MAXIOV) return NULL; |
451a9ebf6
|
369 370 371 372 373 374 375 376 377 378 |
bio = kmalloc(sizeof(struct bio) + nr_iovecs * sizeof(struct bio_vec), gfp_mask); if (unlikely(!bio)) return NULL; bio_init(bio); bio->bi_flags |= BIO_POOL_NONE << BIO_POOL_OFFSET; bio->bi_max_vecs = nr_iovecs; bio->bi_io_vec = bio->bi_inline_vecs; bio->bi_destructor = bio_kmalloc_destructor; |
0a0d96b03
|
379 380 381 |
return bio; } |
a112a71d4
|
382 |
EXPORT_SYMBOL(bio_kmalloc); |
0a0d96b03
|
383 |
|
1da177e4c
|
384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 |
void zero_fill_bio(struct bio *bio) { unsigned long flags; struct bio_vec *bv; int i; bio_for_each_segment(bv, bio, i) { char *data = bvec_kmap_irq(bv, &flags); memset(data, 0, bv->bv_len); flush_dcache_page(bv->bv_page); bvec_kunmap_irq(data, &flags); } } EXPORT_SYMBOL(zero_fill_bio); /** * bio_put - release a reference to a bio * @bio: bio to release reference to * * Description: * Put a reference to a &struct bio, either one you have gotten with |
ad0bf1107
|
405 |
* bio_alloc, bio_get or bio_clone. The last put of a bio will free it. |
1da177e4c
|
406 407 408 409 410 411 412 413 414 |
**/ void bio_put(struct bio *bio) { BIO_BUG_ON(!atomic_read(&bio->bi_cnt)); /* * last put frees it */ if (atomic_dec_and_test(&bio->bi_cnt)) { |
852c788f8
|
415 |
bio_disassociate_task(bio); |
1da177e4c
|
416 417 418 419 |
bio->bi_next = NULL; bio->bi_destructor(bio); } } |
a112a71d4
|
420 |
EXPORT_SYMBOL(bio_put); |
1da177e4c
|
421 |
|
165125e1e
|
422 |
inline int bio_phys_segments(struct request_queue *q, struct bio *bio) |
1da177e4c
|
423 424 425 426 427 428 |
{ if (unlikely(!bio_flagged(bio, BIO_SEG_VALID))) blk_recount_segments(q, bio); return bio->bi_phys_segments; } |
a112a71d4
|
429 |
EXPORT_SYMBOL(bio_phys_segments); |
1da177e4c
|
430 |
|
1da177e4c
|
431 432 433 434 435 436 437 438 439 |
/** * __bio_clone - clone a bio * @bio: destination bio * @bio_src: bio to clone * * Clone a &bio. Caller will own the returned bio, but not * the actual data it points to. Reference count of returned * bio will be one. */ |
858119e15
|
440 |
void __bio_clone(struct bio *bio, struct bio *bio_src) |
1da177e4c
|
441 |
{ |
e525e153c
|
442 443 |
memcpy(bio->bi_io_vec, bio_src->bi_io_vec, bio_src->bi_max_vecs * sizeof(struct bio_vec)); |
1da177e4c
|
444 |
|
5d84070ee
|
445 446 447 448 |
/* * most users will be overriding ->bi_bdev with a new target, * so we don't set nor calculate new physical/hw segment counts here */ |
1da177e4c
|
449 450 451 452 |
bio->bi_sector = bio_src->bi_sector; bio->bi_bdev = bio_src->bi_bdev; bio->bi_flags |= 1 << BIO_CLONED; bio->bi_rw = bio_src->bi_rw; |
1da177e4c
|
453 454 |
bio->bi_vcnt = bio_src->bi_vcnt; bio->bi_size = bio_src->bi_size; |
a5453be48
|
455 |
bio->bi_idx = bio_src->bi_idx; |
1da177e4c
|
456 |
} |
a112a71d4
|
457 |
EXPORT_SYMBOL(__bio_clone); |
1da177e4c
|
458 459 460 461 462 463 464 465 |
/** * bio_clone - clone a bio * @bio: bio to clone * @gfp_mask: allocation priority * * Like __bio_clone, only also allocates the returned bio */ |
dd0fc66fb
|
466 |
struct bio *bio_clone(struct bio *bio, gfp_t gfp_mask) |
1da177e4c
|
467 468 |
{ struct bio *b = bio_alloc_bioset(gfp_mask, bio->bi_max_vecs, fs_bio_set); |
7ba1ba12e
|
469 470 471 472 473 474 475 476 |
if (!b) return NULL; b->bi_destructor = bio_fs_destructor; __bio_clone(b, bio); if (bio_integrity(bio)) { int ret; |
7878cba9f
|
477 |
ret = bio_integrity_clone(b, bio, gfp_mask, fs_bio_set); |
7ba1ba12e
|
478 |
|
059ea3318
|
479 480 |
if (ret < 0) { bio_put(b); |
7ba1ba12e
|
481 |
return NULL; |
059ea3318
|
482 |
} |
3676347a5
|
483 |
} |
1da177e4c
|
484 485 486 |
return b; } |
a112a71d4
|
487 |
EXPORT_SYMBOL(bio_clone); |
1da177e4c
|
488 489 490 491 492 493 494 495 496 497 498 499 |
/** * bio_get_nr_vecs - return approx number of vecs * @bdev: I/O target * * Return the approximate number of pages we can send to this target. * There's no guarantee that you will be able to fit this number of pages * into a bio, it does not account for dynamic restrictions that vary * on offset. */ int bio_get_nr_vecs(struct block_device *bdev) { |
165125e1e
|
500 |
struct request_queue *q = bdev_get_queue(bdev); |
f908ee946
|
501 502 503 |
int nr_pages; nr_pages = min_t(unsigned, |
5abebfdd0
|
504 505 |
queue_max_segments(q), queue_max_sectors(q) / (PAGE_SIZE >> 9) + 1); |
f908ee946
|
506 507 |
return min_t(unsigned, nr_pages, BIO_MAX_PAGES); |
1da177e4c
|
508 |
} |
a112a71d4
|
509 |
EXPORT_SYMBOL(bio_get_nr_vecs); |
1da177e4c
|
510 |
|
165125e1e
|
511 |
static int __bio_add_page(struct request_queue *q, struct bio *bio, struct page |
defd94b75
|
512 513 |
*page, unsigned int len, unsigned int offset, unsigned short max_sectors) |
1da177e4c
|
514 515 516 517 518 519 520 521 522 |
{ int retried_segments = 0; struct bio_vec *bvec; /* * cloned bio must not modify vec list */ if (unlikely(bio_flagged(bio, BIO_CLONED))) return 0; |
80cfd548e
|
523 |
if (((bio->bi_size + len) >> 9) > max_sectors) |
1da177e4c
|
524 |
return 0; |
80cfd548e
|
525 526 527 528 529 530 531 532 533 534 |
/* * For filesystems with a blocksize smaller than the pagesize * we will often be called with the same page as last time and * a consecutive offset. Optimize this special case. */ if (bio->bi_vcnt > 0) { struct bio_vec *prev = &bio->bi_io_vec[bio->bi_vcnt - 1]; if (page == prev->bv_page && offset == prev->bv_offset + prev->bv_len) { |
1d6165851
|
535 |
unsigned int prev_bv_len = prev->bv_len; |
80cfd548e
|
536 |
prev->bv_len += len; |
cc371e66e
|
537 538 539 |
if (q->merge_bvec_fn) { struct bvec_merge_data bvm = { |
1d6165851
|
540 541 542 543 |
/* prev_bvec is already charged in bi_size, discharge it in order to simulate merging updated prev_bvec as new bvec. */ |
cc371e66e
|
544 545 |
.bi_bdev = bio->bi_bdev, .bi_sector = bio->bi_sector, |
1d6165851
|
546 |
.bi_size = bio->bi_size - prev_bv_len, |
cc371e66e
|
547 548 |
.bi_rw = bio->bi_rw, }; |
8bf8c376a
|
549 |
if (q->merge_bvec_fn(q, &bvm, prev) < prev->bv_len) { |
cc371e66e
|
550 551 552 |
prev->bv_len -= len; return 0; } |
80cfd548e
|
553 554 555 556 557 558 559 |
} goto done; } } if (bio->bi_vcnt >= bio->bi_max_vecs) |
1da177e4c
|
560 561 562 563 564 565 |
return 0; /* * we might lose a segment or two here, but rather that than * make this too complex. */ |
8a78362c4
|
566 |
while (bio->bi_phys_segments >= queue_max_segments(q)) { |
1da177e4c
|
567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 |
if (retried_segments) return 0; retried_segments = 1; blk_recount_segments(q, bio); } /* * setup the new entry, we might clear it again later if we * cannot add the page */ bvec = &bio->bi_io_vec[bio->bi_vcnt]; bvec->bv_page = page; bvec->bv_len = len; bvec->bv_offset = offset; /* * if queue has other restrictions (eg varying max sector size * depending on offset), it can specify a merge_bvec_fn in the * queue to get further control */ if (q->merge_bvec_fn) { |
cc371e66e
|
590 591 592 593 594 595 |
struct bvec_merge_data bvm = { .bi_bdev = bio->bi_bdev, .bi_sector = bio->bi_sector, .bi_size = bio->bi_size, .bi_rw = bio->bi_rw, }; |
1da177e4c
|
596 597 598 599 |
/* * merge_bvec_fn() returns number of bytes it can accept * at this offset */ |
8bf8c376a
|
600 |
if (q->merge_bvec_fn(q, &bvm, bvec) < bvec->bv_len) { |
1da177e4c
|
601 602 603 604 605 606 607 608 |
bvec->bv_page = NULL; bvec->bv_len = 0; bvec->bv_offset = 0; return 0; } } /* If we may be able to merge these biovecs, force a recount */ |
b8b3e16cf
|
609 |
if (bio->bi_vcnt && (BIOVEC_PHYS_MERGEABLE(bvec-1, bvec))) |
1da177e4c
|
610 611 612 613 |
bio->bi_flags &= ~(1 << BIO_SEG_VALID); bio->bi_vcnt++; bio->bi_phys_segments++; |
80cfd548e
|
614 |
done: |
1da177e4c
|
615 616 617 618 619 |
bio->bi_size += len; return len; } /** |
6e68af666
|
620 |
* bio_add_pc_page - attempt to add page to bio |
fddfdeafa
|
621 |
* @q: the target queue |
6e68af666
|
622 623 624 625 626 627 |
* @bio: destination bio * @page: page to add * @len: vec entry length * @offset: vec entry offset * * Attempt to add a page to the bio_vec maplist. This can fail for a |
c64280845
|
628 629 630 631 632 |
* number of reasons, such as the bio being full or target block device * limitations. The target block device must allow bio's up to PAGE_SIZE, * so it is always possible to add a single page to an empty bio. * * This should only be used by REQ_PC bios. |
6e68af666
|
633 |
*/ |
165125e1e
|
634 |
int bio_add_pc_page(struct request_queue *q, struct bio *bio, struct page *page, |
6e68af666
|
635 636 |
unsigned int len, unsigned int offset) { |
ae03bf639
|
637 638 |
return __bio_add_page(q, bio, page, len, offset, queue_max_hw_sectors(q)); |
6e68af666
|
639 |
} |
a112a71d4
|
640 |
EXPORT_SYMBOL(bio_add_pc_page); |
6e68af666
|
641 642 |
/** |
1da177e4c
|
643 644 645 646 647 648 649 |
* bio_add_page - attempt to add page to bio * @bio: destination bio * @page: page to add * @len: vec entry length * @offset: vec entry offset * * Attempt to add a page to the bio_vec maplist. This can fail for a |
c64280845
|
650 651 652 |
* number of reasons, such as the bio being full or target block device * limitations. The target block device must allow bio's up to PAGE_SIZE, * so it is always possible to add a single page to an empty bio. |
1da177e4c
|
653 654 655 656 |
*/ int bio_add_page(struct bio *bio, struct page *page, unsigned int len, unsigned int offset) { |
defd94b75
|
657 |
struct request_queue *q = bdev_get_queue(bio->bi_bdev); |
ae03bf639
|
658 |
return __bio_add_page(q, bio, page, len, offset, queue_max_sectors(q)); |
1da177e4c
|
659 |
} |
a112a71d4
|
660 |
EXPORT_SYMBOL(bio_add_page); |
1da177e4c
|
661 662 663 |
struct bio_map_data { struct bio_vec *iovecs; |
c5dec1c30
|
664 |
struct sg_iovec *sgvecs; |
152e283fd
|
665 666 |
int nr_sgvecs; int is_our_pages; |
1da177e4c
|
667 |
}; |
c5dec1c30
|
668 |
static void bio_set_map_data(struct bio_map_data *bmd, struct bio *bio, |
152e283fd
|
669 670 |
struct sg_iovec *iov, int iov_count, int is_our_pages) |
1da177e4c
|
671 672 |
{ memcpy(bmd->iovecs, bio->bi_io_vec, sizeof(struct bio_vec) * bio->bi_vcnt); |
c5dec1c30
|
673 674 |
memcpy(bmd->sgvecs, iov, sizeof(struct sg_iovec) * iov_count); bmd->nr_sgvecs = iov_count; |
152e283fd
|
675 |
bmd->is_our_pages = is_our_pages; |
1da177e4c
|
676 677 678 679 680 681 |
bio->bi_private = bmd; } static void bio_free_map_data(struct bio_map_data *bmd) { kfree(bmd->iovecs); |
c5dec1c30
|
682 |
kfree(bmd->sgvecs); |
1da177e4c
|
683 684 |
kfree(bmd); } |
121f09941
|
685 686 |
static struct bio_map_data *bio_alloc_map_data(int nr_segs, unsigned int iov_count, |
76029ff37
|
687 |
gfp_t gfp_mask) |
1da177e4c
|
688 |
{ |
f3f63c1c2
|
689 690 691 692 |
struct bio_map_data *bmd; if (iov_count > UIO_MAXIOV) return NULL; |
1da177e4c
|
693 |
|
f3f63c1c2
|
694 |
bmd = kmalloc(sizeof(*bmd), gfp_mask); |
1da177e4c
|
695 696 |
if (!bmd) return NULL; |
76029ff37
|
697 |
bmd->iovecs = kmalloc(sizeof(struct bio_vec) * nr_segs, gfp_mask); |
c5dec1c30
|
698 699 700 701 |
if (!bmd->iovecs) { kfree(bmd); return NULL; } |
76029ff37
|
702 |
bmd->sgvecs = kmalloc(sizeof(struct sg_iovec) * iov_count, gfp_mask); |
c5dec1c30
|
703 |
if (bmd->sgvecs) |
1da177e4c
|
704 |
return bmd; |
c5dec1c30
|
705 |
kfree(bmd->iovecs); |
1da177e4c
|
706 707 708 |
kfree(bmd); return NULL; } |
aefcc28a3
|
709 |
static int __bio_copy_iov(struct bio *bio, struct bio_vec *iovecs, |
ecb554a84
|
710 711 |
struct sg_iovec *iov, int iov_count, int to_user, int from_user, int do_free_page) |
c5dec1c30
|
712 713 714 715 716 |
{ int ret = 0, i; struct bio_vec *bvec; int iov_idx = 0; unsigned int iov_off = 0; |
c5dec1c30
|
717 718 719 |
__bio_for_each_segment(bvec, bio, i, 0) { char *bv_addr = page_address(bvec->bv_page); |
aefcc28a3
|
720 |
unsigned int bv_len = iovecs[i].bv_len; |
c5dec1c30
|
721 722 723 |
while (bv_len && iov_idx < iov_count) { unsigned int bytes; |
0e0c62123
|
724 |
char __user *iov_addr; |
c5dec1c30
|
725 726 727 728 729 730 |
bytes = min_t(unsigned int, iov[iov_idx].iov_len - iov_off, bv_len); iov_addr = iov[iov_idx].iov_base + iov_off; if (!ret) { |
ecb554a84
|
731 |
if (to_user) |
c5dec1c30
|
732 733 |
ret = copy_to_user(iov_addr, bv_addr, bytes); |
ecb554a84
|
734 735 736 |
if (from_user) ret = copy_from_user(bv_addr, iov_addr, bytes); |
c5dec1c30
|
737 738 739 740 741 742 743 744 745 746 747 748 749 750 |
if (ret) ret = -EFAULT; } bv_len -= bytes; bv_addr += bytes; iov_addr += bytes; iov_off += bytes; if (iov[iov_idx].iov_len == iov_off) { iov_idx++; iov_off = 0; } } |
152e283fd
|
751 |
if (do_free_page) |
c5dec1c30
|
752 753 754 755 756 |
__free_page(bvec->bv_page); } return ret; } |
1da177e4c
|
757 758 759 760 761 762 763 764 765 766 |
/** * bio_uncopy_user - finish previously mapped bio * @bio: bio being terminated * * Free pages allocated from bio_copy_user() and write back data * to user space in case of a read. */ int bio_uncopy_user(struct bio *bio) { struct bio_map_data *bmd = bio->bi_private; |
818827669
|
767 |
int ret = 0; |
1da177e4c
|
768 |
|
818827669
|
769 770 |
if (!bio_flagged(bio, BIO_NULL_MAPPED)) ret = __bio_copy_iov(bio, bmd->iovecs, bmd->sgvecs, |
ecb554a84
|
771 772 |
bmd->nr_sgvecs, bio_data_dir(bio) == READ, 0, bmd->is_our_pages); |
1da177e4c
|
773 774 775 776 |
bio_free_map_data(bmd); bio_put(bio); return ret; } |
a112a71d4
|
777 |
EXPORT_SYMBOL(bio_uncopy_user); |
1da177e4c
|
778 779 |
/** |
c5dec1c30
|
780 |
* bio_copy_user_iov - copy user data to bio |
1da177e4c
|
781 |
* @q: destination block queue |
152e283fd
|
782 |
* @map_data: pointer to the rq_map_data holding pages (if necessary) |
c5dec1c30
|
783 784 |
* @iov: the iovec. * @iov_count: number of elements in the iovec |
1da177e4c
|
785 |
* @write_to_vm: bool indicating writing to pages or not |
a3bce90ed
|
786 |
* @gfp_mask: memory allocation flags |
1da177e4c
|
787 788 789 790 791 |
* * Prepares and returns a bio for indirect user io, bouncing data * to/from kernel pages as necessary. Must be paired with * call bio_uncopy_user() on io completion. */ |
152e283fd
|
792 793 794 795 |
struct bio *bio_copy_user_iov(struct request_queue *q, struct rq_map_data *map_data, struct sg_iovec *iov, int iov_count, int write_to_vm, gfp_t gfp_mask) |
1da177e4c
|
796 |
{ |
1da177e4c
|
797 798 799 800 801 |
struct bio_map_data *bmd; struct bio_vec *bvec; struct page *page; struct bio *bio; int i, ret; |
c5dec1c30
|
802 803 |
int nr_pages = 0; unsigned int len = 0; |
56c451f4b
|
804 |
unsigned int offset = map_data ? map_data->offset & ~PAGE_MASK : 0; |
1da177e4c
|
805 |
|
c5dec1c30
|
806 807 808 809 810 811 812 813 |
for (i = 0; i < iov_count; i++) { unsigned long uaddr; unsigned long end; unsigned long start; uaddr = (unsigned long)iov[i].iov_base; end = (uaddr + iov[i].iov_len + PAGE_SIZE - 1) >> PAGE_SHIFT; start = uaddr >> PAGE_SHIFT; |
cb4644cac
|
814 815 816 817 818 |
/* * Overflow, abort */ if (end < start) return ERR_PTR(-EINVAL); |
c5dec1c30
|
819 820 821 |
nr_pages += end - start; len += iov[i].iov_len; } |
69838727b
|
822 823 |
if (offset) nr_pages++; |
a3bce90ed
|
824 |
bmd = bio_alloc_map_data(nr_pages, iov_count, gfp_mask); |
1da177e4c
|
825 826 |
if (!bmd) return ERR_PTR(-ENOMEM); |
1da177e4c
|
827 |
ret = -ENOMEM; |
a9e9dc24b
|
828 |
bio = bio_kmalloc(gfp_mask, nr_pages); |
1da177e4c
|
829 830 |
if (!bio) goto out_bmd; |
7b6d91dae
|
831 832 |
if (!write_to_vm) bio->bi_rw |= REQ_WRITE; |
1da177e4c
|
833 834 |
ret = 0; |
56c451f4b
|
835 836 |
if (map_data) { |
e623ddb4e
|
837 |
nr_pages = 1 << map_data->page_order; |
56c451f4b
|
838 839 |
i = map_data->offset / PAGE_SIZE; } |
1da177e4c
|
840 |
while (len) { |
e623ddb4e
|
841 |
unsigned int bytes = PAGE_SIZE; |
1da177e4c
|
842 |
|
56c451f4b
|
843 |
bytes -= offset; |
1da177e4c
|
844 845 |
if (bytes > len) bytes = len; |
152e283fd
|
846 |
if (map_data) { |
e623ddb4e
|
847 |
if (i == map_data->nr_entries * nr_pages) { |
152e283fd
|
848 849 850 |
ret = -ENOMEM; break; } |
e623ddb4e
|
851 852 853 854 855 856 |
page = map_data->pages[i / nr_pages]; page += (i % nr_pages); i++; } else { |
152e283fd
|
857 |
page = alloc_page(q->bounce_gfp | gfp_mask); |
e623ddb4e
|
858 859 860 861 |
if (!page) { ret = -ENOMEM; break; } |
1da177e4c
|
862 |
} |
56c451f4b
|
863 |
if (bio_add_pc_page(q, bio, page, bytes, offset) < bytes) |
1da177e4c
|
864 |
break; |
1da177e4c
|
865 866 |
len -= bytes; |
56c451f4b
|
867 |
offset = 0; |
1da177e4c
|
868 869 870 871 872 873 874 875 |
} if (ret) goto cleanup; /* * success */ |
ecb554a84
|
876 877 878 |
if ((!write_to_vm && (!map_data || !map_data->null_mapped)) || (map_data && map_data->from_user)) { ret = __bio_copy_iov(bio, bio->bi_io_vec, iov, iov_count, 0, 1, 0); |
c5dec1c30
|
879 880 |
if (ret) goto cleanup; |
1da177e4c
|
881 |
} |
152e283fd
|
882 |
bio_set_map_data(bmd, bio, iov, iov_count, map_data ? 0 : 1); |
1da177e4c
|
883 884 |
return bio; cleanup: |
152e283fd
|
885 886 887 |
if (!map_data) bio_for_each_segment(bvec, bio, i) __free_page(bvec->bv_page); |
1da177e4c
|
888 889 890 891 892 893 |
bio_put(bio); out_bmd: bio_free_map_data(bmd); return ERR_PTR(ret); } |
c5dec1c30
|
894 895 896 |
/** * bio_copy_user - copy user data to bio * @q: destination block queue |
152e283fd
|
897 |
* @map_data: pointer to the rq_map_data holding pages (if necessary) |
c5dec1c30
|
898 899 900 |
* @uaddr: start of user address * @len: length in bytes * @write_to_vm: bool indicating writing to pages or not |
a3bce90ed
|
901 |
* @gfp_mask: memory allocation flags |
c5dec1c30
|
902 903 904 905 906 |
* * Prepares and returns a bio for indirect user io, bouncing data * to/from kernel pages as necessary. Must be paired with * call bio_uncopy_user() on io completion. */ |
152e283fd
|
907 908 909 |
struct bio *bio_copy_user(struct request_queue *q, struct rq_map_data *map_data, unsigned long uaddr, unsigned int len, int write_to_vm, gfp_t gfp_mask) |
c5dec1c30
|
910 911 912 913 914 |
{ struct sg_iovec iov; iov.iov_base = (void __user *)uaddr; iov.iov_len = len; |
152e283fd
|
915 |
return bio_copy_user_iov(q, map_data, &iov, 1, write_to_vm, gfp_mask); |
c5dec1c30
|
916 |
} |
a112a71d4
|
917 |
EXPORT_SYMBOL(bio_copy_user); |
c5dec1c30
|
918 |
|
165125e1e
|
919 |
static struct bio *__bio_map_user_iov(struct request_queue *q, |
f1970baf6
|
920 921 |
struct block_device *bdev, struct sg_iovec *iov, int iov_count, |
a3bce90ed
|
922 |
int write_to_vm, gfp_t gfp_mask) |
1da177e4c
|
923 |
{ |
f1970baf6
|
924 925 |
int i, j; int nr_pages = 0; |
1da177e4c
|
926 927 |
struct page **pages; struct bio *bio; |
f1970baf6
|
928 929 |
int cur_page = 0; int ret, offset; |
1da177e4c
|
930 |
|
f1970baf6
|
931 932 933 934 935 |
for (i = 0; i < iov_count; i++) { unsigned long uaddr = (unsigned long)iov[i].iov_base; unsigned long len = iov[i].iov_len; unsigned long end = (uaddr + len + PAGE_SIZE - 1) >> PAGE_SHIFT; unsigned long start = uaddr >> PAGE_SHIFT; |
cb4644cac
|
936 937 938 939 940 |
/* * Overflow, abort */ if (end < start) return ERR_PTR(-EINVAL); |
f1970baf6
|
941 942 |
nr_pages += end - start; /* |
ad2d72257
|
943 |
* buffer must be aligned to at least hardsector size for now |
f1970baf6
|
944 |
*/ |
ad2d72257
|
945 |
if (uaddr & queue_dma_alignment(q)) |
f1970baf6
|
946 947 948 949 |
return ERR_PTR(-EINVAL); } if (!nr_pages) |
1da177e4c
|
950 |
return ERR_PTR(-EINVAL); |
a9e9dc24b
|
951 |
bio = bio_kmalloc(gfp_mask, nr_pages); |
1da177e4c
|
952 953 954 955 |
if (!bio) return ERR_PTR(-ENOMEM); ret = -ENOMEM; |
a3bce90ed
|
956 |
pages = kcalloc(nr_pages, sizeof(struct page *), gfp_mask); |
1da177e4c
|
957 958 |
if (!pages) goto out; |
f1970baf6
|
959 960 961 962 963 964 965 |
for (i = 0; i < iov_count; i++) { unsigned long uaddr = (unsigned long)iov[i].iov_base; unsigned long len = iov[i].iov_len; unsigned long end = (uaddr + len + PAGE_SIZE - 1) >> PAGE_SHIFT; unsigned long start = uaddr >> PAGE_SHIFT; const int local_nr_pages = end - start; const int page_limit = cur_page + local_nr_pages; |
cb4644cac
|
966 |
|
f5dd33c49
|
967 968 |
ret = get_user_pages_fast(uaddr, local_nr_pages, write_to_vm, &pages[cur_page]); |
991721572
|
969 970 |
if (ret < local_nr_pages) { ret = -EFAULT; |
f1970baf6
|
971 |
goto out_unmap; |
991721572
|
972 |
} |
f1970baf6
|
973 974 975 976 977 978 979 980 981 982 983 984 985 986 |
offset = uaddr & ~PAGE_MASK; for (j = cur_page; j < page_limit; j++) { unsigned int bytes = PAGE_SIZE - offset; if (len <= 0) break; if (bytes > len) bytes = len; /* * sorry... */ |
defd94b75
|
987 988 |
if (bio_add_pc_page(q, bio, pages[j], bytes, offset) < bytes) |
f1970baf6
|
989 990 991 992 993 |
break; len -= bytes; offset = 0; } |
1da177e4c
|
994 |
|
f1970baf6
|
995 |
cur_page = j; |
1da177e4c
|
996 |
/* |
f1970baf6
|
997 |
* release the pages we didn't map into the bio, if any |
1da177e4c
|
998 |
*/ |
f1970baf6
|
999 1000 |
while (j < page_limit) page_cache_release(pages[j++]); |
1da177e4c
|
1001 |
} |
1da177e4c
|
1002 1003 1004 1005 1006 1007 |
kfree(pages); /* * set data direction, and check if mapped pages need bouncing */ if (!write_to_vm) |
7b6d91dae
|
1008 |
bio->bi_rw |= REQ_WRITE; |
1da177e4c
|
1009 |
|
f1970baf6
|
1010 |
bio->bi_bdev = bdev; |
1da177e4c
|
1011 1012 |
bio->bi_flags |= (1 << BIO_USER_MAPPED); return bio; |
f1970baf6
|
1013 1014 1015 1016 1017 1018 1019 1020 |
out_unmap: for (i = 0; i < nr_pages; i++) { if(!pages[i]) break; page_cache_release(pages[i]); } out: |
1da177e4c
|
1021 1022 1023 1024 1025 1026 1027 |
kfree(pages); bio_put(bio); return ERR_PTR(ret); } /** * bio_map_user - map user address into bio |
165125e1e
|
1028 |
* @q: the struct request_queue for the bio |
1da177e4c
|
1029 1030 1031 1032 |
* @bdev: destination block device * @uaddr: start of user address * @len: length in bytes * @write_to_vm: bool indicating writing to pages or not |
a3bce90ed
|
1033 |
* @gfp_mask: memory allocation flags |
1da177e4c
|
1034 1035 1036 1037 |
* * Map the user space address into a bio suitable for io to a block * device. Returns an error pointer in case of error. */ |
165125e1e
|
1038 |
struct bio *bio_map_user(struct request_queue *q, struct block_device *bdev, |
a3bce90ed
|
1039 1040 |
unsigned long uaddr, unsigned int len, int write_to_vm, gfp_t gfp_mask) |
1da177e4c
|
1041 |
{ |
f1970baf6
|
1042 |
struct sg_iovec iov; |
3f70353ea
|
1043 |
iov.iov_base = (void __user *)uaddr; |
f1970baf6
|
1044 |
iov.iov_len = len; |
a3bce90ed
|
1045 |
return bio_map_user_iov(q, bdev, &iov, 1, write_to_vm, gfp_mask); |
f1970baf6
|
1046 |
} |
a112a71d4
|
1047 |
EXPORT_SYMBOL(bio_map_user); |
f1970baf6
|
1048 1049 1050 |
/** * bio_map_user_iov - map user sg_iovec table into bio |
165125e1e
|
1051 |
* @q: the struct request_queue for the bio |
f1970baf6
|
1052 1053 1054 1055 |
* @bdev: destination block device * @iov: the iovec. * @iov_count: number of elements in the iovec * @write_to_vm: bool indicating writing to pages or not |
a3bce90ed
|
1056 |
* @gfp_mask: memory allocation flags |
f1970baf6
|
1057 1058 1059 1060 |
* * Map the user space address into a bio suitable for io to a block * device. Returns an error pointer in case of error. */ |
165125e1e
|
1061 |
struct bio *bio_map_user_iov(struct request_queue *q, struct block_device *bdev, |
f1970baf6
|
1062 |
struct sg_iovec *iov, int iov_count, |
a3bce90ed
|
1063 |
int write_to_vm, gfp_t gfp_mask) |
f1970baf6
|
1064 |
{ |
1da177e4c
|
1065 |
struct bio *bio; |
a3bce90ed
|
1066 1067 |
bio = __bio_map_user_iov(q, bdev, iov, iov_count, write_to_vm, gfp_mask); |
1da177e4c
|
1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 |
if (IS_ERR(bio)) return bio; /* * subtle -- if __bio_map_user() ended up bouncing a bio, * it would normally disappear when its bi_end_io is run. * however, we need it for the unmap, so grab an extra * reference to it */ bio_get(bio); |
0e75f9063
|
1078 |
return bio; |
1da177e4c
|
1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 |
} static void __bio_unmap_user(struct bio *bio) { struct bio_vec *bvec; int i; /* * make sure we dirty pages we wrote to */ __bio_for_each_segment(bvec, bio, i, 0) { if (bio_data_dir(bio) == READ) set_page_dirty_lock(bvec->bv_page); page_cache_release(bvec->bv_page); } bio_put(bio); } /** * bio_unmap_user - unmap a bio * @bio: the bio being unmapped * * Unmap a bio previously mapped by bio_map_user(). Must be called with * a process context. * * bio_unmap_user() may sleep. */ void bio_unmap_user(struct bio *bio) { __bio_unmap_user(bio); bio_put(bio); } |
a112a71d4
|
1113 |
EXPORT_SYMBOL(bio_unmap_user); |
1da177e4c
|
1114 |
|
6712ecf8f
|
1115 |
static void bio_map_kern_endio(struct bio *bio, int err) |
b823825e8
|
1116 |
{ |
b823825e8
|
1117 |
bio_put(bio); |
b823825e8
|
1118 |
} |
165125e1e
|
1119 |
static struct bio *__bio_map_kern(struct request_queue *q, void *data, |
27496a8c6
|
1120 |
unsigned int len, gfp_t gfp_mask) |
df46b9a44
|
1121 1122 1123 1124 1125 1126 1127 |
{ unsigned long kaddr = (unsigned long)data; unsigned long end = (kaddr + len + PAGE_SIZE - 1) >> PAGE_SHIFT; unsigned long start = kaddr >> PAGE_SHIFT; const int nr_pages = end - start; int offset, i; struct bio *bio; |
a9e9dc24b
|
1128 |
bio = bio_kmalloc(gfp_mask, nr_pages); |
df46b9a44
|
1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 |
if (!bio) return ERR_PTR(-ENOMEM); offset = offset_in_page(kaddr); for (i = 0; i < nr_pages; i++) { unsigned int bytes = PAGE_SIZE - offset; if (len <= 0) break; if (bytes > len) bytes = len; |
defd94b75
|
1141 1142 |
if (bio_add_pc_page(q, bio, virt_to_page(data), bytes, offset) < bytes) |
df46b9a44
|
1143 1144 1145 1146 1147 1148 |
break; data += bytes; len -= bytes; offset = 0; } |
b823825e8
|
1149 |
bio->bi_end_io = bio_map_kern_endio; |
df46b9a44
|
1150 1151 1152 1153 1154 |
return bio; } /** * bio_map_kern - map kernel address into bio |
165125e1e
|
1155 |
* @q: the struct request_queue for the bio |
df46b9a44
|
1156 1157 1158 1159 1160 1161 1162 |
* @data: pointer to buffer to map * @len: length in bytes * @gfp_mask: allocation flags for bio allocation * * Map the kernel address into a bio suitable for io to a block * device. Returns an error pointer in case of error. */ |
165125e1e
|
1163 |
struct bio *bio_map_kern(struct request_queue *q, void *data, unsigned int len, |
27496a8c6
|
1164 |
gfp_t gfp_mask) |
df46b9a44
|
1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 |
{ struct bio *bio; bio = __bio_map_kern(q, data, len, gfp_mask); if (IS_ERR(bio)) return bio; if (bio->bi_size == len) return bio; /* * Don't support partial mappings. */ bio_put(bio); return ERR_PTR(-EINVAL); } |
a112a71d4
|
1181 |
EXPORT_SYMBOL(bio_map_kern); |
df46b9a44
|
1182 |
|
68154e90c
|
1183 1184 1185 1186 |
static void bio_copy_kern_endio(struct bio *bio, int err) { struct bio_vec *bvec; const int read = bio_data_dir(bio) == READ; |
76029ff37
|
1187 |
struct bio_map_data *bmd = bio->bi_private; |
68154e90c
|
1188 |
int i; |
76029ff37
|
1189 |
char *p = bmd->sgvecs[0].iov_base; |
68154e90c
|
1190 1191 1192 |
__bio_for_each_segment(bvec, bio, i, 0) { char *addr = page_address(bvec->bv_page); |
76029ff37
|
1193 |
int len = bmd->iovecs[i].bv_len; |
68154e90c
|
1194 |
|
4fc981ef9
|
1195 |
if (read) |
76029ff37
|
1196 |
memcpy(p, addr, len); |
68154e90c
|
1197 1198 |
__free_page(bvec->bv_page); |
76029ff37
|
1199 |
p += len; |
68154e90c
|
1200 |
} |
76029ff37
|
1201 |
bio_free_map_data(bmd); |
68154e90c
|
1202 1203 1204 1205 1206 1207 1208 1209 1210 |
bio_put(bio); } /** * bio_copy_kern - copy kernel address into bio * @q: the struct request_queue for the bio * @data: pointer to buffer to copy * @len: length in bytes * @gfp_mask: allocation flags for bio and page allocation |
ffee0259c
|
1211 |
* @reading: data direction is READ |
68154e90c
|
1212 1213 1214 1215 1216 1217 1218 |
* * copy the kernel address into a bio suitable for io to a block * device. Returns an error pointer in case of error. */ struct bio *bio_copy_kern(struct request_queue *q, void *data, unsigned int len, gfp_t gfp_mask, int reading) { |
68154e90c
|
1219 1220 |
struct bio *bio; struct bio_vec *bvec; |
4d8ab62e0
|
1221 |
int i; |
68154e90c
|
1222 |
|
4d8ab62e0
|
1223 1224 1225 |
bio = bio_copy_user(q, NULL, (unsigned long)data, len, 1, gfp_mask); if (IS_ERR(bio)) return bio; |
68154e90c
|
1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 |
if (!reading) { void *p = data; bio_for_each_segment(bvec, bio, i) { char *addr = page_address(bvec->bv_page); memcpy(addr, p, bvec->bv_len); p += bvec->bv_len; } } |
68154e90c
|
1237 |
bio->bi_end_io = bio_copy_kern_endio; |
76029ff37
|
1238 |
|
68154e90c
|
1239 |
return bio; |
68154e90c
|
1240 |
} |
a112a71d4
|
1241 |
EXPORT_SYMBOL(bio_copy_kern); |
68154e90c
|
1242 |
|
1da177e4c
|
1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 |
/* * bio_set_pages_dirty() and bio_check_pages_dirty() are support functions * for performing direct-IO in BIOs. * * The problem is that we cannot run set_page_dirty() from interrupt context * because the required locks are not interrupt-safe. So what we can do is to * mark the pages dirty _before_ performing IO. And in interrupt context, * check that the pages are still dirty. If so, fine. If not, redirty them * in process context. * * We special-case compound pages here: normally this means reads into hugetlb * pages. The logic in here doesn't really work right for compound pages * because the VM does not uniformly chase down the head page in all cases. * But dirtiness of compound pages is pretty meaningless anyway: the VM doesn't * handle them at all. So we skip compound pages here at an early stage. * * Note that this code is very hard to test under normal circumstances because * direct-io pins the pages with get_user_pages(). This makes * is_page_cache_freeable return false, and the VM will not clean the pages. |
0d5c3eba2
|
1262 |
* But other code (eg, flusher threads) could clean the pages if they are mapped |
1da177e4c
|
1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 |
* pagecache. * * Simply disabling the call to bio_set_pages_dirty() is a good way to test the * deferred bio dirtying paths. */ /* * bio_set_pages_dirty() will mark all the bio's pages as dirty. */ void bio_set_pages_dirty(struct bio *bio) { struct bio_vec *bvec = bio->bi_io_vec; int i; for (i = 0; i < bio->bi_vcnt; i++) { struct page *page = bvec[i].bv_page; if (page && !PageCompound(page)) set_page_dirty_lock(page); } } |
86b6c7a7f
|
1284 |
static void bio_release_pages(struct bio *bio) |
1da177e4c
|
1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 |
{ struct bio_vec *bvec = bio->bi_io_vec; int i; for (i = 0; i < bio->bi_vcnt; i++) { struct page *page = bvec[i].bv_page; if (page) put_page(page); } } /* * bio_check_pages_dirty() will check that all the BIO's pages are still dirty. * If they are, then fine. If, however, some pages are clean then they must * have been written out during the direct-IO read. So we take another ref on * the BIO and the offending pages and re-dirty the pages in process context. * * It is expected that bio_check_pages_dirty() will wholly own the BIO from * here on. It will run one page_cache_release() against each page and will * run one bio_put() against the BIO. */ |
65f27f384
|
1307 |
static void bio_dirty_fn(struct work_struct *work); |
1da177e4c
|
1308 |
|
65f27f384
|
1309 |
static DECLARE_WORK(bio_dirty_work, bio_dirty_fn); |
1da177e4c
|
1310 1311 1312 1313 1314 1315 |
static DEFINE_SPINLOCK(bio_dirty_lock); static struct bio *bio_dirty_list; /* * This runs in process context */ |
65f27f384
|
1316 |
static void bio_dirty_fn(struct work_struct *work) |
1da177e4c
|
1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 |
{ unsigned long flags; struct bio *bio; spin_lock_irqsave(&bio_dirty_lock, flags); bio = bio_dirty_list; bio_dirty_list = NULL; spin_unlock_irqrestore(&bio_dirty_lock, flags); while (bio) { struct bio *next = bio->bi_private; bio_set_pages_dirty(bio); bio_release_pages(bio); bio_put(bio); bio = next; } } void bio_check_pages_dirty(struct bio *bio) { struct bio_vec *bvec = bio->bi_io_vec; int nr_clean_pages = 0; int i; for (i = 0; i < bio->bi_vcnt; i++) { struct page *page = bvec[i].bv_page; if (PageDirty(page) || PageCompound(page)) { page_cache_release(page); bvec[i].bv_page = NULL; } else { nr_clean_pages++; } } if (nr_clean_pages) { unsigned long flags; spin_lock_irqsave(&bio_dirty_lock, flags); bio->bi_private = bio_dirty_list; bio_dirty_list = bio; spin_unlock_irqrestore(&bio_dirty_lock, flags); schedule_work(&bio_dirty_work); } else { bio_put(bio); } } |
2d4dc890b
|
1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 |
#if ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE void bio_flush_dcache_pages(struct bio *bi) { int i; struct bio_vec *bvec; bio_for_each_segment(bvec, bi, i) flush_dcache_page(bvec->bv_page); } EXPORT_SYMBOL(bio_flush_dcache_pages); #endif |
1da177e4c
|
1376 1377 1378 |
/** * bio_endio - end I/O on a bio * @bio: bio |
1da177e4c
|
1379 1380 1381 |
* @error: error, if any * * Description: |
6712ecf8f
|
1382 |
* bio_endio() will end I/O on the whole bio. bio_endio() is the |
5bb23a688
|
1383 1384 1385 |
* preferred way to end I/O on a bio, it takes care of clearing * BIO_UPTODATE on error. @error is 0 on success, and and one of the * established -Exxxx (-EIO, for instance) error values in case |
25985edce
|
1386 |
* something went wrong. No one should call bi_end_io() directly on a |
5bb23a688
|
1387 1388 |
* bio unless they own it and thus know that it has an end_io * function. |
1da177e4c
|
1389 |
**/ |
6712ecf8f
|
1390 |
void bio_endio(struct bio *bio, int error) |
1da177e4c
|
1391 1392 1393 |
{ if (error) clear_bit(BIO_UPTODATE, &bio->bi_flags); |
9cc54d40b
|
1394 1395 |
else if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) error = -EIO; |
1da177e4c
|
1396 |
|
5bb23a688
|
1397 |
if (bio->bi_end_io) |
6712ecf8f
|
1398 |
bio->bi_end_io(bio, error); |
1da177e4c
|
1399 |
} |
a112a71d4
|
1400 |
EXPORT_SYMBOL(bio_endio); |
1da177e4c
|
1401 1402 1403 1404 1405 |
void bio_pair_release(struct bio_pair *bp) { if (atomic_dec_and_test(&bp->cnt)) { struct bio *master = bp->bio1.bi_private; |
6712ecf8f
|
1406 |
bio_endio(master, bp->error); |
1da177e4c
|
1407 1408 1409 |
mempool_free(bp, bp->bio2.bi_private); } } |
a112a71d4
|
1410 |
EXPORT_SYMBOL(bio_pair_release); |
1da177e4c
|
1411 |
|
6712ecf8f
|
1412 |
static void bio_pair_end_1(struct bio *bi, int err) |
1da177e4c
|
1413 1414 1415 1416 1417 |
{ struct bio_pair *bp = container_of(bi, struct bio_pair, bio1); if (err) bp->error = err; |
1da177e4c
|
1418 |
bio_pair_release(bp); |
1da177e4c
|
1419 |
} |
6712ecf8f
|
1420 |
static void bio_pair_end_2(struct bio *bi, int err) |
1da177e4c
|
1421 1422 1423 1424 1425 |
{ struct bio_pair *bp = container_of(bi, struct bio_pair, bio2); if (err) bp->error = err; |
1da177e4c
|
1426 |
bio_pair_release(bp); |
1da177e4c
|
1427 1428 1429 |
} /* |
c7eee1b83
|
1430 |
* split a bio - only worry about a bio with a single page in its iovec |
1da177e4c
|
1431 |
*/ |
6feef531f
|
1432 |
struct bio_pair *bio_split(struct bio *bi, int first_sectors) |
1da177e4c
|
1433 |
{ |
6feef531f
|
1434 |
struct bio_pair *bp = mempool_alloc(bio_split_pool, GFP_NOIO); |
1da177e4c
|
1435 1436 1437 |
if (!bp) return bp; |
5f3ea37c7
|
1438 |
trace_block_split(bdev_get_queue(bi->bi_bdev), bi, |
2056a782f
|
1439 |
bi->bi_sector + first_sectors); |
1da177e4c
|
1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 |
BUG_ON(bi->bi_vcnt != 1); BUG_ON(bi->bi_idx != 0); atomic_set(&bp->cnt, 3); bp->error = 0; bp->bio1 = *bi; bp->bio2 = *bi; bp->bio2.bi_sector += first_sectors; bp->bio2.bi_size -= first_sectors << 9; bp->bio1.bi_size = first_sectors << 9; bp->bv1 = bi->bi_io_vec[0]; bp->bv2 = bi->bi_io_vec[0]; bp->bv2.bv_offset += first_sectors << 9; bp->bv2.bv_len -= first_sectors << 9; bp->bv1.bv_len = first_sectors << 9; bp->bio1.bi_io_vec = &bp->bv1; bp->bio2.bi_io_vec = &bp->bv2; |
a2eb0c101
|
1458 1459 |
bp->bio1.bi_max_vecs = 1; bp->bio2.bi_max_vecs = 1; |
1da177e4c
|
1460 1461 1462 1463 |
bp->bio1.bi_end_io = bio_pair_end_1; bp->bio2.bi_end_io = bio_pair_end_2; bp->bio1.bi_private = bi; |
6feef531f
|
1464 |
bp->bio2.bi_private = bio_split_pool; |
1da177e4c
|
1465 |
|
7ba1ba12e
|
1466 1467 |
if (bio_integrity(bi)) bio_integrity_split(bi, bp, first_sectors); |
1da177e4c
|
1468 1469 |
return bp; } |
a112a71d4
|
1470 |
EXPORT_SYMBOL(bio_split); |
1da177e4c
|
1471 |
|
ad3316bf4
|
1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 |
/** * bio_sector_offset - Find hardware sector offset in bio * @bio: bio to inspect * @index: bio_vec index * @offset: offset in bv_page * * Return the number of hardware sectors between beginning of bio * and an end point indicated by a bio_vec index and an offset * within that vector's page. */ sector_t bio_sector_offset(struct bio *bio, unsigned short index, unsigned int offset) { |
e1defc4ff
|
1485 |
unsigned int sector_sz; |
ad3316bf4
|
1486 1487 1488 |
struct bio_vec *bv; sector_t sectors; int i; |
e1defc4ff
|
1489 |
sector_sz = queue_logical_block_size(bio->bi_bdev->bd_disk->queue); |
ad3316bf4
|
1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 |
sectors = 0; if (index >= bio->bi_idx) index = bio->bi_vcnt - 1; __bio_for_each_segment(bv, bio, i, 0) { if (i == index) { if (offset > bv->bv_offset) sectors += (offset - bv->bv_offset) / sector_sz; break; } sectors += bv->bv_len / sector_sz; } return sectors; } EXPORT_SYMBOL(bio_sector_offset); |
1da177e4c
|
1508 1509 1510 1511 1512 |
/* * create memory pools for biovec's in a bio_set. * use the global biovec slabs created for general use. */ |
5972511b7
|
1513 |
static int biovec_create_pools(struct bio_set *bs, int pool_entries) |
1da177e4c
|
1514 |
{ |
7ff9345ff
|
1515 |
struct biovec_slab *bp = bvec_slabs + BIOVEC_MAX_IDX; |
1da177e4c
|
1516 |
|
7ff9345ff
|
1517 1518 1519 |
bs->bvec_pool = mempool_create_slab_pool(pool_entries, bp->slab); if (!bs->bvec_pool) return -ENOMEM; |
1da177e4c
|
1520 |
|
1da177e4c
|
1521 1522 1523 1524 1525 |
return 0; } static void biovec_free_pools(struct bio_set *bs) { |
7ff9345ff
|
1526 |
mempool_destroy(bs->bvec_pool); |
1da177e4c
|
1527 1528 1529 1530 1531 1532 |
} void bioset_free(struct bio_set *bs) { if (bs->bio_pool) mempool_destroy(bs->bio_pool); |
7878cba9f
|
1533 |
bioset_integrity_free(bs); |
1da177e4c
|
1534 |
biovec_free_pools(bs); |
bb799ca02
|
1535 |
bio_put_slab(bs); |
1da177e4c
|
1536 1537 1538 |
kfree(bs); } |
a112a71d4
|
1539 |
EXPORT_SYMBOL(bioset_free); |
1da177e4c
|
1540 |
|
bb799ca02
|
1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 |
/** * bioset_create - Create a bio_set * @pool_size: Number of bio and bio_vecs to cache in the mempool * @front_pad: Number of bytes to allocate in front of the returned bio * * Description: * Set up a bio_set to be used with @bio_alloc_bioset. Allows the caller * to ask for a number of bytes to be allocated in front of the bio. * Front pad allocation is useful for embedding the bio inside * another structure, to avoid allocating extra data to go with the bio. * Note that the bio must be embedded at the END of that structure always, * or things will break badly. */ struct bio_set *bioset_create(unsigned int pool_size, unsigned int front_pad) |
1da177e4c
|
1555 |
{ |
392ddc329
|
1556 |
unsigned int back_pad = BIO_INLINE_VECS * sizeof(struct bio_vec); |
1b4344986
|
1557 |
struct bio_set *bs; |
1da177e4c
|
1558 |
|
1b4344986
|
1559 |
bs = kzalloc(sizeof(*bs), GFP_KERNEL); |
1da177e4c
|
1560 1561 |
if (!bs) return NULL; |
bb799ca02
|
1562 |
bs->front_pad = front_pad; |
1b4344986
|
1563 |
|
392ddc329
|
1564 |
bs->bio_slab = bio_find_or_create_slab(front_pad + back_pad); |
bb799ca02
|
1565 1566 1567 1568 1569 1570 |
if (!bs->bio_slab) { kfree(bs); return NULL; } bs->bio_pool = mempool_create_slab_pool(pool_size, bs->bio_slab); |
1da177e4c
|
1571 1572 |
if (!bs->bio_pool) goto bad; |
bb799ca02
|
1573 |
if (!biovec_create_pools(bs, pool_size)) |
1da177e4c
|
1574 1575 1576 1577 1578 1579 |
return bs; bad: bioset_free(bs); return NULL; } |
a112a71d4
|
1580 |
EXPORT_SYMBOL(bioset_create); |
1da177e4c
|
1581 |
|
852c788f8
|
1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 |
#ifdef CONFIG_BLK_CGROUP /** * bio_associate_current - associate a bio with %current * @bio: target bio * * Associate @bio with %current if it hasn't been associated yet. Block * layer will treat @bio as if it were issued by %current no matter which * task actually issues it. * * This function takes an extra reference of @task's io_context and blkcg * which will be put when @bio is released. The caller must own @bio, * ensure %current->io_context exists, and is responsible for synchronizing * calls to this function. */ int bio_associate_current(struct bio *bio) { struct io_context *ioc; struct cgroup_subsys_state *css; if (bio->bi_ioc) return -EBUSY; ioc = current->io_context; if (!ioc) return -ENOENT; /* acquire active ref on @ioc and associate */ get_io_context_active(ioc); bio->bi_ioc = ioc; /* associate blkcg if exists */ rcu_read_lock(); css = task_subsys_state(current, blkio_subsys_id); if (css && css_tryget(css)) bio->bi_css = css; rcu_read_unlock(); return 0; } /** * bio_disassociate_task - undo bio_associate_current() * @bio: target bio */ void bio_disassociate_task(struct bio *bio) { if (bio->bi_ioc) { put_io_context(bio->bi_ioc); bio->bi_ioc = NULL; } if (bio->bi_css) { css_put(bio->bi_css); bio->bi_css = NULL; } } #endif /* CONFIG_BLK_CGROUP */ |
1da177e4c
|
1639 1640 1641 1642 1643 1644 1645 |
static void __init biovec_init_slabs(void) { int i; for (i = 0; i < BIOVEC_NR_POOLS; i++) { int size; struct biovec_slab *bvs = bvec_slabs + i; |
a7fcd37cd
|
1646 1647 1648 1649 |
if (bvs->nr_vecs <= BIO_INLINE_VECS) { bvs->slab = NULL; continue; } |
a7fcd37cd
|
1650 |
|
1da177e4c
|
1651 1652 |
size = bvs->nr_vecs * sizeof(struct bio_vec); bvs->slab = kmem_cache_create(bvs->name, size, 0, |
20c2df83d
|
1653 |
SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL); |
1da177e4c
|
1654 1655 1656 1657 1658 |
} } static int __init init_bio(void) { |
bb799ca02
|
1659 1660 1661 1662 1663 1664 |
bio_slab_max = 2; bio_slab_nr = 0; bio_slabs = kzalloc(bio_slab_max * sizeof(struct bio_slab), GFP_KERNEL); if (!bio_slabs) panic("bio: can't allocate bios "); |
1da177e4c
|
1665 |
|
7878cba9f
|
1666 |
bio_integrity_init(); |
1da177e4c
|
1667 |
biovec_init_slabs(); |
bb799ca02
|
1668 |
fs_bio_set = bioset_create(BIO_POOL_SIZE, 0); |
1da177e4c
|
1669 1670 1671 |
if (!fs_bio_set) panic("bio: can't allocate bios "); |
a91a2785b
|
1672 1673 1674 |
if (bioset_integrity_create(fs_bio_set, BIO_POOL_SIZE)) panic("bio: can't create integrity pool "); |
0eaae62ab
|
1675 1676 |
bio_split_pool = mempool_create_kmalloc_pool(BIO_SPLIT_ENTRIES, sizeof(struct bio_pair)); |
1da177e4c
|
1677 1678 1679 1680 1681 1682 |
if (!bio_split_pool) panic("bio: can't create split pool "); return 0; } |
1da177e4c
|
1683 |
subsys_initcall(init_bio); |