Blame view
fs/splice.c
46.3 KB
5274f052e
|
1 2 3 4 5 6 7 8 9 10 11 |
/* * "splice": joining two ropes together by interweaving their strands. * * This is the "extended pipe" functionality, where a pipe is used as * an arbitrary in-memory buffer. Think of a pipe as a small kernel * buffer that you can use to transfer data from one end to the other. * * The traditional unix read/write is extended with a "splice()" operation * that transfers data buffers to or from a pipe buffer. * * Named by Larry McVoy, original implementation from Linus, extended by |
c2058e061
|
12 13 |
* Jens to support splicing to files, network, direct splicing, etc and * fixing lots of bugs. |
5274f052e
|
14 |
* |
0fe234795
|
15 |
* Copyright (C) 2005-2006 Jens Axboe <axboe@kernel.dk> |
c2058e061
|
16 17 |
* Copyright (C) 2005-2006 Linus Torvalds <torvalds@osdl.org> * Copyright (C) 2006 Ingo Molnar <mingo@elte.hu> |
5274f052e
|
18 19 20 21 22 |
* */ #include <linux/fs.h> #include <linux/file.h> #include <linux/pagemap.h> |
d6b29d7ce
|
23 |
#include <linux/splice.h> |
08e552c69
|
24 |
#include <linux/memcontrol.h> |
5274f052e
|
25 |
#include <linux/mm_inline.h> |
5abc97aa2
|
26 |
#include <linux/swap.h> |
4f6f0bd2f
|
27 28 |
#include <linux/writeback.h> #include <linux/buffer_head.h> |
a0f067802
|
29 |
#include <linux/module.h> |
4f6f0bd2f
|
30 |
#include <linux/syscalls.h> |
912d35f86
|
31 |
#include <linux/uio.h> |
29ce20586
|
32 |
#include <linux/security.h> |
5a0e3ad6a
|
33 |
#include <linux/gfp.h> |
5274f052e
|
34 |
|
83f9135bd
|
35 36 37 38 39 40 |
/* * Attempt to steal a page from a pipe buffer. This should perhaps go into * a vm helper function, it's already simplified quite a bit by the * addition of remove_mapping(). If success is returned, the caller may * attempt to reuse this page for another destination. */ |
76ad4d111
|
41 |
static int page_cache_pipe_buf_steal(struct pipe_inode_info *pipe, |
5abc97aa2
|
42 43 44 |
struct pipe_buffer *buf) { struct page *page = buf->page; |
9e94cd4fd
|
45 |
struct address_space *mapping; |
5abc97aa2
|
46 |
|
9e0267c26
|
47 |
lock_page(page); |
9e94cd4fd
|
48 49 50 |
mapping = page_mapping(page); if (mapping) { WARN_ON(!PageUptodate(page)); |
5abc97aa2
|
51 |
|
9e94cd4fd
|
52 53 54 55 56 57 58 59 60 |
/* * At least for ext2 with nobh option, we need to wait on * writeback completing on this page, since we'll remove it * from the pagecache. Otherwise truncate wont wait on the * page, allowing the disk blocks to be reused by someone else * before we actually wrote our data to them. fs corruption * ensues. */ wait_on_page_writeback(page); |
ad8d6f0a7
|
61 |
|
266cf658e
|
62 63 |
if (page_has_private(page) && !try_to_release_page(page, GFP_KERNEL)) |
ca39d651d
|
64 |
goto out_unlock; |
4f6f0bd2f
|
65 |
|
9e94cd4fd
|
66 67 68 69 70 71 72 73 |
/* * If we succeeded in removing the mapping, set LRU flag * and return good. */ if (remove_mapping(mapping, page)) { buf->flags |= PIPE_BUF_FLAG_LRU; return 0; } |
9e0267c26
|
74 |
} |
5abc97aa2
|
75 |
|
9e94cd4fd
|
76 77 78 79 |
/* * Raced with truncate or failed to remove page from current * address space, unlock and return failure. */ |
ca39d651d
|
80 |
out_unlock: |
9e94cd4fd
|
81 82 |
unlock_page(page); return 1; |
5abc97aa2
|
83 |
} |
76ad4d111
|
84 |
static void page_cache_pipe_buf_release(struct pipe_inode_info *pipe, |
5274f052e
|
85 86 87 |
struct pipe_buffer *buf) { page_cache_release(buf->page); |
1432873af
|
88 |
buf->flags &= ~PIPE_BUF_FLAG_LRU; |
5274f052e
|
89 |
} |
0845718da
|
90 91 92 93 |
/* * Check whether the contents of buf is OK to access. Since the content * is a page cache page, IO may be in flight. */ |
cac36bb06
|
94 95 |
static int page_cache_pipe_buf_confirm(struct pipe_inode_info *pipe, struct pipe_buffer *buf) |
5274f052e
|
96 97 |
{ struct page *page = buf->page; |
49d0b21be
|
98 |
int err; |
5274f052e
|
99 100 |
if (!PageUptodate(page)) { |
49d0b21be
|
101 102 103 104 |
lock_page(page); /* * Page got truncated/unhashed. This will cause a 0-byte |
73d62d83e
|
105 |
* splice, if this is the first page. |
49d0b21be
|
106 107 108 109 110 |
*/ if (!page->mapping) { err = -ENODATA; goto error; } |
5274f052e
|
111 |
|
49d0b21be
|
112 |
/* |
73d62d83e
|
113 |
* Uh oh, read-error from disk. |
49d0b21be
|
114 115 116 117 118 119 120 |
*/ if (!PageUptodate(page)) { err = -EIO; goto error; } /* |
f84d75199
|
121 |
* Page is ok afterall, we are done. |
49d0b21be
|
122 |
*/ |
5274f052e
|
123 |
unlock_page(page); |
5274f052e
|
124 |
} |
f84d75199
|
125 |
return 0; |
49d0b21be
|
126 127 |
error: unlock_page(page); |
f84d75199
|
128 |
return err; |
70524490e
|
129 |
} |
708e3508c
|
130 |
const struct pipe_buf_operations page_cache_pipe_buf_ops = { |
5274f052e
|
131 |
.can_merge = 0, |
f84d75199
|
132 133 |
.map = generic_pipe_buf_map, .unmap = generic_pipe_buf_unmap, |
cac36bb06
|
134 |
.confirm = page_cache_pipe_buf_confirm, |
5274f052e
|
135 |
.release = page_cache_pipe_buf_release, |
5abc97aa2
|
136 |
.steal = page_cache_pipe_buf_steal, |
f84d75199
|
137 |
.get = generic_pipe_buf_get, |
5274f052e
|
138 |
}; |
912d35f86
|
139 140 141 |
static int user_page_pipe_buf_steal(struct pipe_inode_info *pipe, struct pipe_buffer *buf) { |
7afa6fd03
|
142 143 |
if (!(buf->flags & PIPE_BUF_FLAG_GIFT)) return 1; |
1432873af
|
144 |
buf->flags |= PIPE_BUF_FLAG_LRU; |
330ab7161
|
145 |
return generic_pipe_buf_steal(pipe, buf); |
912d35f86
|
146 |
} |
d4c3cca94
|
147 |
static const struct pipe_buf_operations user_page_pipe_buf_ops = { |
912d35f86
|
148 |
.can_merge = 0, |
f84d75199
|
149 150 |
.map = generic_pipe_buf_map, .unmap = generic_pipe_buf_unmap, |
cac36bb06
|
151 |
.confirm = generic_pipe_buf_confirm, |
912d35f86
|
152 153 |
.release = page_cache_pipe_buf_release, .steal = user_page_pipe_buf_steal, |
f84d75199
|
154 |
.get = generic_pipe_buf_get, |
912d35f86
|
155 |
}; |
825cdcb1a
|
156 157 158 159 160 161 162 |
static void wakeup_pipe_readers(struct pipe_inode_info *pipe) { smp_mb(); if (waitqueue_active(&pipe->wait)) wake_up_interruptible(&pipe->wait); kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); } |
932cc6d4f
|
163 164 165 166 167 168 |
/** * splice_to_pipe - fill passed data into a pipe * @pipe: pipe to fill * @spd: data to fill * * Description: |
79685b8de
|
169 |
* @spd contains a map of pages and len/offset tuples, along with |
932cc6d4f
|
170 171 172 |
* the struct pipe_buf_operations associated with these pages. This * function will link that data to the pipe. * |
83f9135bd
|
173 |
*/ |
d6b29d7ce
|
174 175 |
ssize_t splice_to_pipe(struct pipe_inode_info *pipe, struct splice_pipe_desc *spd) |
5274f052e
|
176 |
{ |
00de00bda
|
177 |
unsigned int spd_pages = spd->nr_pages; |
912d35f86
|
178 |
int ret, do_wakeup, page_nr; |
5274f052e
|
179 180 181 |
ret = 0; do_wakeup = 0; |
912d35f86
|
182 |
page_nr = 0; |
5274f052e
|
183 |
|
61e0d47c3
|
184 |
pipe_lock(pipe); |
5274f052e
|
185 |
|
5274f052e
|
186 |
for (;;) { |
3a326a2ce
|
187 |
if (!pipe->readers) { |
5274f052e
|
188 189 190 191 192 |
send_sig(SIGPIPE, current, 0); if (!ret) ret = -EPIPE; break; } |
35f3d14db
|
193 194 |
if (pipe->nrbufs < pipe->buffers) { int newbuf = (pipe->curbuf + pipe->nrbufs) & (pipe->buffers - 1); |
3a326a2ce
|
195 |
struct pipe_buffer *buf = pipe->bufs + newbuf; |
5274f052e
|
196 |
|
912d35f86
|
197 198 199 |
buf->page = spd->pages[page_nr]; buf->offset = spd->partial[page_nr].offset; buf->len = spd->partial[page_nr].len; |
497f9625c
|
200 |
buf->private = spd->partial[page_nr].private; |
912d35f86
|
201 |
buf->ops = spd->ops; |
7afa6fd03
|
202 203 |
if (spd->flags & SPLICE_F_GIFT) buf->flags |= PIPE_BUF_FLAG_GIFT; |
6f767b042
|
204 |
pipe->nrbufs++; |
912d35f86
|
205 206 |
page_nr++; ret += buf->len; |
6f767b042
|
207 208 |
if (pipe->inode) do_wakeup = 1; |
5274f052e
|
209 |
|
912d35f86
|
210 |
if (!--spd->nr_pages) |
5274f052e
|
211 |
break; |
35f3d14db
|
212 |
if (pipe->nrbufs < pipe->buffers) |
5274f052e
|
213 214 215 216 |
continue; break; } |
912d35f86
|
217 |
if (spd->flags & SPLICE_F_NONBLOCK) { |
29e350944
|
218 219 220 221 |
if (!ret) ret = -EAGAIN; break; } |
5274f052e
|
222 223 224 225 226 227 228 |
if (signal_pending(current)) { if (!ret) ret = -ERESTARTSYS; break; } if (do_wakeup) { |
c0bd1f650
|
229 |
smp_mb(); |
3a326a2ce
|
230 231 232 |
if (waitqueue_active(&pipe->wait)) wake_up_interruptible_sync(&pipe->wait); kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); |
5274f052e
|
233 234 |
do_wakeup = 0; } |
3a326a2ce
|
235 236 237 |
pipe->waiting_writers++; pipe_wait(pipe); pipe->waiting_writers--; |
5274f052e
|
238 |
} |
61e0d47c3
|
239 |
pipe_unlock(pipe); |
5274f052e
|
240 |
|
825cdcb1a
|
241 242 |
if (do_wakeup) wakeup_pipe_readers(pipe); |
5274f052e
|
243 |
|
00de00bda
|
244 |
while (page_nr < spd_pages) |
bbdfc2f70
|
245 |
spd->spd_release(spd, page_nr++); |
5274f052e
|
246 247 248 |
return ret; } |
708e3508c
|
249 |
void spd_release_page(struct splice_pipe_desc *spd, unsigned int i) |
bbdfc2f70
|
250 251 252 |
{ page_cache_release(spd->pages[i]); } |
35f3d14db
|
253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 |
/* * Check if we need to grow the arrays holding pages and partial page * descriptions. */ int splice_grow_spd(struct pipe_inode_info *pipe, struct splice_pipe_desc *spd) { if (pipe->buffers <= PIPE_DEF_BUFFERS) return 0; spd->pages = kmalloc(pipe->buffers * sizeof(struct page *), GFP_KERNEL); spd->partial = kmalloc(pipe->buffers * sizeof(struct partial_page), GFP_KERNEL); if (spd->pages && spd->partial) return 0; kfree(spd->pages); kfree(spd->partial); return -ENOMEM; } void splice_shrink_spd(struct pipe_inode_info *pipe, struct splice_pipe_desc *spd) { if (pipe->buffers <= PIPE_DEF_BUFFERS) return; kfree(spd->pages); kfree(spd->partial); } |
3a326a2ce
|
282 |
static int |
cbb7e577e
|
283 284 285 |
__generic_file_splice_read(struct file *in, loff_t *ppos, struct pipe_inode_info *pipe, size_t len, unsigned int flags) |
5274f052e
|
286 287 |
{ struct address_space *mapping = in->f_mapping; |
d8983910a
|
288 |
unsigned int loff, nr_pages, req_pages; |
35f3d14db
|
289 290 |
struct page *pages[PIPE_DEF_BUFFERS]; struct partial_page partial[PIPE_DEF_BUFFERS]; |
5274f052e
|
291 |
struct page *page; |
91ad66ef4
|
292 293 |
pgoff_t index, end_index; loff_t isize; |
eb20796bf
|
294 |
int error, page_nr; |
912d35f86
|
295 296 297 298 299 |
struct splice_pipe_desc spd = { .pages = pages, .partial = partial, .flags = flags, .ops = &page_cache_pipe_buf_ops, |
bbdfc2f70
|
300 |
.spd_release = spd_release_page, |
912d35f86
|
301 |
}; |
5274f052e
|
302 |
|
35f3d14db
|
303 304 |
if (splice_grow_spd(pipe, &spd)) return -ENOMEM; |
cbb7e577e
|
305 |
index = *ppos >> PAGE_CACHE_SHIFT; |
912d35f86
|
306 |
loff = *ppos & ~PAGE_CACHE_MASK; |
d8983910a
|
307 |
req_pages = (len + loff + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; |
35f3d14db
|
308 |
nr_pages = min(req_pages, pipe->buffers); |
5274f052e
|
309 310 |
/* |
eb20796bf
|
311 312 |
* Lookup the (hopefully) full range of pages we need. */ |
35f3d14db
|
313 |
spd.nr_pages = find_get_pages_contig(mapping, index, nr_pages, spd.pages); |
431a4820b
|
314 |
index += spd.nr_pages; |
82aa5d618
|
315 |
|
eb20796bf
|
316 317 |
/* * If find_get_pages_contig() returned fewer pages than we needed, |
431a4820b
|
318 |
* readahead/allocate the rest and fill in the holes. |
eb20796bf
|
319 |
*/ |
431a4820b
|
320 |
if (spd.nr_pages < nr_pages) |
cf914a7d6
|
321 322 |
page_cache_sync_readahead(mapping, &in->f_ra, in, index, req_pages - spd.nr_pages); |
431a4820b
|
323 |
|
932cc6d4f
|
324 |
error = 0; |
eb20796bf
|
325 |
while (spd.nr_pages < nr_pages) { |
82aa5d618
|
326 |
/* |
eb20796bf
|
327 328 |
* Page could be there, find_get_pages_contig() breaks on * the first hole. |
5274f052e
|
329 |
*/ |
7480a9043
|
330 331 |
page = find_get_page(mapping, index); if (!page) { |
e27dedd84
|
332 |
/* |
eb20796bf
|
333 |
* page didn't exist, allocate one. |
7480a9043
|
334 335 336 337 338 339 |
*/ page = page_cache_alloc_cold(mapping); if (!page) break; error = add_to_page_cache_lru(page, mapping, index, |
0ae0b5d05
|
340 |
GFP_KERNEL); |
7480a9043
|
341 342 |
if (unlikely(error)) { page_cache_release(page); |
a0548871e
|
343 344 |
if (error == -EEXIST) continue; |
7480a9043
|
345 346 |
break; } |
eb20796bf
|
347 348 349 350 351 |
/* * add_to_page_cache() locks the page, unlock it * to avoid convoluting the logic below even more. */ unlock_page(page); |
7480a9043
|
352 |
} |
35f3d14db
|
353 |
spd.pages[spd.nr_pages++] = page; |
eb20796bf
|
354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 |
index++; } /* * Now loop over the map and see if we need to start IO on any * pages, fill in the partial map, etc. */ index = *ppos >> PAGE_CACHE_SHIFT; nr_pages = spd.nr_pages; spd.nr_pages = 0; for (page_nr = 0; page_nr < nr_pages; page_nr++) { unsigned int this_len; if (!len) break; /* * this_len is the max we'll use from this page */ this_len = min_t(unsigned long, len, PAGE_CACHE_SIZE - loff); |
35f3d14db
|
374 |
page = spd.pages[page_nr]; |
eb20796bf
|
375 |
|
a08a166fe
|
376 |
if (PageReadahead(page)) |
cf914a7d6
|
377 |
page_cache_async_readahead(mapping, &in->f_ra, in, |
d8983910a
|
378 |
page, index, req_pages - page_nr); |
a08a166fe
|
379 |
|
7480a9043
|
380 381 382 383 |
/* * If the page isn't uptodate, we may need to start io on it */ if (!PageUptodate(page)) { |
6965031d3
|
384 |
lock_page(page); |
7480a9043
|
385 386 |
/* |
32502b841
|
387 388 389 390 |
* Page was truncated, or invalidated by the * filesystem. Redo the find/create, but this time the * page is kept locked, so there's no chance of another * race with truncate/invalidate. |
7480a9043
|
391 392 393 |
*/ if (!page->mapping) { unlock_page(page); |
32502b841
|
394 395 396 397 398 399 400 |
page = find_or_create_page(mapping, index, mapping_gfp_mask(mapping)); if (!page) { error = -ENOMEM; break; } |
35f3d14db
|
401 402 |
page_cache_release(spd.pages[page_nr]); spd.pages[page_nr] = page; |
7480a9043
|
403 404 405 406 407 408 409 410 |
} /* * page was already under io and is now done, great */ if (PageUptodate(page)) { unlock_page(page); goto fill_it; } |
5274f052e
|
411 |
|
7480a9043
|
412 413 414 415 |
/* * need to read in the page */ error = mapping->a_ops->readpage(in, page); |
5274f052e
|
416 |
if (unlikely(error)) { |
eb20796bf
|
417 418 419 420 421 422 |
/* * We really should re-lookup the page here, * but it complicates things a lot. Instead * lets just do what we already stored, and * we'll get it the next time we are called. */ |
7480a9043
|
423 |
if (error == AOP_TRUNCATED_PAGE) |
eb20796bf
|
424 |
error = 0; |
5274f052e
|
425 426 |
break; } |
620a324b7
|
427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 |
} fill_it: /* * i_size must be checked after PageUptodate. */ isize = i_size_read(mapping->host); end_index = (isize - 1) >> PAGE_CACHE_SHIFT; if (unlikely(!isize || index > end_index)) break; /* * if this is the last page, see if we need to shrink * the length and stop */ if (end_index == index) { unsigned int plen; |
91ad66ef4
|
443 444 |
/* |
620a324b7
|
445 |
* max good bytes in this page |
91ad66ef4
|
446 |
*/ |
620a324b7
|
447 448 |
plen = ((isize - 1) & ~PAGE_CACHE_MASK) + 1; if (plen <= loff) |
91ad66ef4
|
449 |
break; |
91ad66ef4
|
450 451 |
/* |
620a324b7
|
452 |
* force quit after adding this page |
91ad66ef4
|
453 |
*/ |
620a324b7
|
454 455 |
this_len = min(this_len, plen - loff); len = this_len; |
5274f052e
|
456 |
} |
620a324b7
|
457 |
|
35f3d14db
|
458 459 |
spd.partial[page_nr].offset = loff; spd.partial[page_nr].len = this_len; |
82aa5d618
|
460 |
len -= this_len; |
91ad66ef4
|
461 |
loff = 0; |
eb20796bf
|
462 463 |
spd.nr_pages++; index++; |
5274f052e
|
464 |
} |
eb20796bf
|
465 |
/* |
475ecade6
|
466 |
* Release any pages at the end, if we quit early. 'page_nr' is how far |
eb20796bf
|
467 468 469 |
* we got, 'nr_pages' is how many pages are in the map. */ while (page_nr < nr_pages) |
35f3d14db
|
470 |
page_cache_release(spd.pages[page_nr++]); |
f4e6b498d
|
471 |
in->f_ra.prev_pos = (loff_t)index << PAGE_CACHE_SHIFT; |
eb20796bf
|
472 |
|
912d35f86
|
473 |
if (spd.nr_pages) |
35f3d14db
|
474 |
error = splice_to_pipe(pipe, &spd); |
5274f052e
|
475 |
|
35f3d14db
|
476 |
splice_shrink_spd(pipe, &spd); |
7480a9043
|
477 |
return error; |
5274f052e
|
478 |
} |
83f9135bd
|
479 480 481 |
/** * generic_file_splice_read - splice data from file to a pipe * @in: file to splice from |
932cc6d4f
|
482 |
* @ppos: position in @in |
83f9135bd
|
483 484 485 486 |
* @pipe: pipe to splice to * @len: number of bytes to splice * @flags: splice modifier flags * |
932cc6d4f
|
487 488 489 490 491 |
* Description: * Will read pages from given file and fill them into a pipe. Can be * used as long as the address_space operations for the source implements * a readpage() hook. * |
83f9135bd
|
492 |
*/ |
cbb7e577e
|
493 494 495 |
ssize_t generic_file_splice_read(struct file *in, loff_t *ppos, struct pipe_inode_info *pipe, size_t len, unsigned int flags) |
5274f052e
|
496 |
{ |
d366d3988
|
497 |
loff_t isize, left; |
8191ecd1d
|
498 |
int ret; |
d366d3988
|
499 500 501 502 503 504 505 506 |
isize = i_size_read(in->f_mapping->host); if (unlikely(*ppos >= isize)) return 0; left = isize - *ppos; if (unlikely(left < len)) len = left; |
5274f052e
|
507 |
|
8191ecd1d
|
508 |
ret = __generic_file_splice_read(in, ppos, pipe, len, flags); |
723590ed5
|
509 |
if (ret > 0) { |
cbb7e577e
|
510 |
*ppos += ret; |
723590ed5
|
511 512 |
file_accessed(in); } |
5274f052e
|
513 514 515 |
return ret; } |
059a8f373
|
516 |
EXPORT_SYMBOL(generic_file_splice_read); |
6818173bd
|
517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 |
static const struct pipe_buf_operations default_pipe_buf_ops = { .can_merge = 0, .map = generic_pipe_buf_map, .unmap = generic_pipe_buf_unmap, .confirm = generic_pipe_buf_confirm, .release = generic_pipe_buf_release, .steal = generic_pipe_buf_steal, .get = generic_pipe_buf_get, }; static ssize_t kernel_readv(struct file *file, const struct iovec *vec, unsigned long vlen, loff_t offset) { mm_segment_t old_fs; loff_t pos = offset; ssize_t res; old_fs = get_fs(); set_fs(get_ds()); /* The cast to a user pointer is valid due to the set_fs() */ res = vfs_readv(file, (const struct iovec __user *)vec, vlen, &pos); set_fs(old_fs); return res; } |
b2858d7d1
|
542 543 |
static ssize_t kernel_write(struct file *file, const char *buf, size_t count, loff_t pos) |
0b0a47f5c
|
544 545 546 547 548 549 550 |
{ mm_segment_t old_fs; ssize_t res; old_fs = get_fs(); set_fs(get_ds()); /* The cast to a user pointer is valid due to the set_fs() */ |
b2858d7d1
|
551 |
res = vfs_write(file, (const char __user *)buf, count, &pos); |
0b0a47f5c
|
552 553 554 555 |
set_fs(old_fs); return res; } |
6818173bd
|
556 557 558 559 560 561 562 |
ssize_t default_file_splice_read(struct file *in, loff_t *ppos, struct pipe_inode_info *pipe, size_t len, unsigned int flags) { unsigned int nr_pages; unsigned int nr_freed; size_t offset; |
35f3d14db
|
563 564 565 |
struct page *pages[PIPE_DEF_BUFFERS]; struct partial_page partial[PIPE_DEF_BUFFERS]; struct iovec *vec, __vec[PIPE_DEF_BUFFERS]; |
6818173bd
|
566 567 568 569 570 571 572 573 574 575 576 |
ssize_t res; size_t this_len; int error; int i; struct splice_pipe_desc spd = { .pages = pages, .partial = partial, .flags = flags, .ops = &default_pipe_buf_ops, .spd_release = spd_release_page, }; |
35f3d14db
|
577 578 579 580 581 582 583 584 585 586 |
if (splice_grow_spd(pipe, &spd)) return -ENOMEM; res = -ENOMEM; vec = __vec; if (pipe->buffers > PIPE_DEF_BUFFERS) { vec = kmalloc(pipe->buffers * sizeof(struct iovec), GFP_KERNEL); if (!vec) goto shrink_ret; } |
6818173bd
|
587 588 |
offset = *ppos & ~PAGE_CACHE_MASK; nr_pages = (len + offset + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; |
35f3d14db
|
589 |
for (i = 0; i < nr_pages && i < pipe->buffers && len; i++) { |
6818173bd
|
590 |
struct page *page; |
4f2312285
|
591 |
page = alloc_page(GFP_USER); |
6818173bd
|
592 593 594 595 596 |
error = -ENOMEM; if (!page) goto err; this_len = min_t(size_t, len, PAGE_CACHE_SIZE - offset); |
4f2312285
|
597 |
vec[i].iov_base = (void __user *) page_address(page); |
6818173bd
|
598 |
vec[i].iov_len = this_len; |
35f3d14db
|
599 |
spd.pages[i] = page; |
6818173bd
|
600 601 602 603 604 605 |
spd.nr_pages++; len -= this_len; offset = 0; } res = kernel_readv(in, vec, spd.nr_pages, *ppos); |
77f6bf57b
|
606 607 |
if (res < 0) { error = res; |
6818173bd
|
608 |
goto err; |
77f6bf57b
|
609 |
} |
6818173bd
|
610 611 612 613 614 615 616 |
error = 0; if (!res) goto err; nr_freed = 0; for (i = 0; i < spd.nr_pages; i++) { |
6818173bd
|
617 |
this_len = min_t(size_t, vec[i].iov_len, res); |
35f3d14db
|
618 619 |
spd.partial[i].offset = 0; spd.partial[i].len = this_len; |
6818173bd
|
620 |
if (!this_len) { |
35f3d14db
|
621 622 |
__free_page(spd.pages[i]); spd.pages[i] = NULL; |
6818173bd
|
623 624 625 626 627 628 629 630 631 |
nr_freed++; } res -= this_len; } spd.nr_pages -= nr_freed; res = splice_to_pipe(pipe, &spd); if (res > 0) *ppos += res; |
35f3d14db
|
632 633 634 635 |
shrink_ret: if (vec != __vec) kfree(vec); splice_shrink_spd(pipe, &spd); |
6818173bd
|
636 637 638 |
return res; err: |
4f2312285
|
639 |
for (i = 0; i < spd.nr_pages; i++) |
35f3d14db
|
640 |
__free_page(spd.pages[i]); |
4f2312285
|
641 |
|
35f3d14db
|
642 643 |
res = error; goto shrink_ret; |
6818173bd
|
644 645 |
} EXPORT_SYMBOL(default_file_splice_read); |
5274f052e
|
646 |
/* |
4f6f0bd2f
|
647 |
* Send 'sd->len' bytes to socket from 'sd->file' at position 'sd->pos' |
016b661e2
|
648 |
* using sendpage(). Return the number of bytes sent. |
5274f052e
|
649 |
*/ |
76ad4d111
|
650 |
static int pipe_to_sendpage(struct pipe_inode_info *pipe, |
5274f052e
|
651 652 |
struct pipe_buffer *buf, struct splice_desc *sd) { |
6a14b90bb
|
653 |
struct file *file = sd->u.file; |
5274f052e
|
654 |
loff_t pos = sd->pos; |
a8adbe378
|
655 |
int more; |
5274f052e
|
656 |
|
a8adbe378
|
657 658 659 660 661 662 |
if (!likely(file->f_op && file->f_op->sendpage)) return -EINVAL; more = (sd->flags & SPLICE_F_MORE) || sd->len < sd->total_len; return file->f_op->sendpage(file, buf->page, buf->offset, sd->len, &pos, more); |
5274f052e
|
663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 |
} /* * This is a little more tricky than the file -> pipe splicing. There are * basically three cases: * * - Destination page already exists in the address space and there * are users of it. For that case we have no other option that * copying the data. Tough luck. * - Destination page already exists in the address space, but there * are no users of it. Make sure it's uptodate, then drop it. Fall * through to last case. * - Destination page does not exist, we can add the pipe page to * the page cache and avoid the copy. * |
83f9135bd
|
678 679 680 681 682 683 |
* If asked to move pages to the output file (SPLICE_F_MOVE is set in * sd->flags), we attempt to migrate pages from the pipe to the output * file address space page cache. This is possible if no one else has * the pipe page referenced outside of the pipe and page cache. If * SPLICE_F_MOVE isn't set, or we cannot move the page, we simply create * a new page in the output file page cache and fill/dirty that. |
5274f052e
|
684 |
*/ |
328eaaba4
|
685 686 |
int pipe_to_file(struct pipe_inode_info *pipe, struct pipe_buffer *buf, struct splice_desc *sd) |
5274f052e
|
687 |
{ |
6a14b90bb
|
688 |
struct file *file = sd->u.file; |
5274f052e
|
689 |
struct address_space *mapping = file->f_mapping; |
016b661e2
|
690 |
unsigned int offset, this_len; |
5274f052e
|
691 |
struct page *page; |
afddba49d
|
692 |
void *fsdata; |
3e7ee3e7b
|
693 |
int ret; |
5274f052e
|
694 |
|
5274f052e
|
695 |
offset = sd->pos & ~PAGE_CACHE_MASK; |
016b661e2
|
696 697 698 |
this_len = sd->len; if (this_len + offset > PAGE_CACHE_SIZE) this_len = PAGE_CACHE_SIZE - offset; |
afddba49d
|
699 700 701 702 |
ret = pagecache_write_begin(file, mapping, sd->pos, this_len, AOP_FLAG_UNINTERRUPTIBLE, &page, &fsdata); if (unlikely(ret)) goto out; |
5274f052e
|
703 |
|
0568b409c
|
704 |
if (buf->page != page) { |
f84d75199
|
705 706 707 |
/* * Careful, ->map() uses KM_USER0! */ |
76ad4d111
|
708 |
char *src = buf->ops->map(pipe, buf, 1); |
f84d75199
|
709 |
char *dst = kmap_atomic(page, KM_USER1); |
5abc97aa2
|
710 |
|
016b661e2
|
711 |
memcpy(dst + offset, src + buf->offset, this_len); |
5abc97aa2
|
712 |
flush_dcache_page(page); |
f84d75199
|
713 |
kunmap_atomic(dst, KM_USER1); |
76ad4d111
|
714 |
buf->ops->unmap(pipe, buf, src); |
5abc97aa2
|
715 |
} |
afddba49d
|
716 717 |
ret = pagecache_write_end(file, mapping, sd->pos, this_len, this_len, page, fsdata); |
5274f052e
|
718 |
out: |
5274f052e
|
719 720 |
return ret; } |
328eaaba4
|
721 |
EXPORT_SYMBOL(pipe_to_file); |
5274f052e
|
722 |
|
b3c2d2ddd
|
723 724 725 726 727 728 729 |
static void wakeup_pipe_writers(struct pipe_inode_info *pipe) { smp_mb(); if (waitqueue_active(&pipe->wait)) wake_up_interruptible(&pipe->wait); kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); } |
932cc6d4f
|
730 |
/** |
b3c2d2ddd
|
731 |
* splice_from_pipe_feed - feed available data from a pipe to a file |
932cc6d4f
|
732 733 734 735 736 |
* @pipe: pipe to splice from * @sd: information to @actor * @actor: handler that splices the data * * Description: |
b3c2d2ddd
|
737 738 739 740 741 742 743 |
* This function loops over the pipe and calls @actor to do the * actual moving of a single struct pipe_buffer to the desired * destination. It returns when there's no more buffers left in * the pipe or if the requested number of bytes (@sd->total_len) * have been copied. It returns a positive number (one) if the * pipe needs to be filled with more data, zero if the required * number of bytes have been copied and -errno on error. |
932cc6d4f
|
744 |
* |
b3c2d2ddd
|
745 746 747 748 |
* This, together with splice_from_pipe_{begin,end,next}, may be * used to implement the functionality of __splice_from_pipe() when * locking is required around copying the pipe buffers to the * destination. |
83f9135bd
|
749 |
*/ |
b3c2d2ddd
|
750 751 |
int splice_from_pipe_feed(struct pipe_inode_info *pipe, struct splice_desc *sd, splice_actor *actor) |
5274f052e
|
752 |
{ |
b3c2d2ddd
|
753 |
int ret; |
5274f052e
|
754 |
|
b3c2d2ddd
|
755 756 757 |
while (pipe->nrbufs) { struct pipe_buffer *buf = pipe->bufs + pipe->curbuf; const struct pipe_buf_operations *ops = buf->ops; |
5274f052e
|
758 |
|
b3c2d2ddd
|
759 760 761 |
sd->len = buf->len; if (sd->len > sd->total_len) sd->len = sd->total_len; |
5274f052e
|
762 |
|
a8adbe378
|
763 764 |
ret = buf->ops->confirm(pipe, buf); if (unlikely(ret)) { |
b3c2d2ddd
|
765 766 767 768 |
if (ret == -ENODATA) ret = 0; return ret; } |
a8adbe378
|
769 770 771 772 |
ret = actor(pipe, buf, sd); if (ret <= 0) return ret; |
b3c2d2ddd
|
773 774 775 776 777 778 779 780 781 782 783 |
buf->offset += ret; buf->len -= ret; sd->num_spliced += ret; sd->len -= ret; sd->pos += ret; sd->total_len -= ret; if (!buf->len) { buf->ops = NULL; ops->release(pipe, buf); |
35f3d14db
|
784 |
pipe->curbuf = (pipe->curbuf + 1) & (pipe->buffers - 1); |
b3c2d2ddd
|
785 786 787 788 |
pipe->nrbufs--; if (pipe->inode) sd->need_wakeup = true; } |
5274f052e
|
789 |
|
b3c2d2ddd
|
790 791 792 |
if (!sd->total_len) return 0; } |
5274f052e
|
793 |
|
b3c2d2ddd
|
794 795 796 |
return 1; } EXPORT_SYMBOL(splice_from_pipe_feed); |
5274f052e
|
797 |
|
b3c2d2ddd
|
798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 |
/** * splice_from_pipe_next - wait for some data to splice from * @pipe: pipe to splice from * @sd: information about the splice operation * * Description: * This function will wait for some data and return a positive * value (one) if pipe buffers are available. It will return zero * or -errno if no more data needs to be spliced. */ int splice_from_pipe_next(struct pipe_inode_info *pipe, struct splice_desc *sd) { while (!pipe->nrbufs) { if (!pipe->writers) return 0; |
016b661e2
|
813 |
|
b3c2d2ddd
|
814 815 |
if (!pipe->waiting_writers && sd->num_spliced) return 0; |
73d62d83e
|
816 |
|
b3c2d2ddd
|
817 818 |
if (sd->flags & SPLICE_F_NONBLOCK) return -EAGAIN; |
5274f052e
|
819 |
|
b3c2d2ddd
|
820 821 |
if (signal_pending(current)) return -ERESTARTSYS; |
5274f052e
|
822 |
|
b3c2d2ddd
|
823 824 825 |
if (sd->need_wakeup) { wakeup_pipe_writers(pipe); sd->need_wakeup = false; |
5274f052e
|
826 |
} |
b3c2d2ddd
|
827 828 |
pipe_wait(pipe); } |
29e350944
|
829 |
|
b3c2d2ddd
|
830 831 832 |
return 1; } EXPORT_SYMBOL(splice_from_pipe_next); |
5274f052e
|
833 |
|
b3c2d2ddd
|
834 835 |
/** * splice_from_pipe_begin - start splicing from pipe |
b80901bbf
|
836 |
* @sd: information about the splice operation |
b3c2d2ddd
|
837 838 839 840 841 842 843 844 845 846 847 848 |
* * Description: * This function should be called before a loop containing * splice_from_pipe_next() and splice_from_pipe_feed() to * initialize the necessary fields of @sd. */ void splice_from_pipe_begin(struct splice_desc *sd) { sd->num_spliced = 0; sd->need_wakeup = false; } EXPORT_SYMBOL(splice_from_pipe_begin); |
5274f052e
|
849 |
|
b3c2d2ddd
|
850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 |
/** * splice_from_pipe_end - finish splicing from pipe * @pipe: pipe to splice from * @sd: information about the splice operation * * Description: * This function will wake up pipe writers if necessary. It should * be called after a loop containing splice_from_pipe_next() and * splice_from_pipe_feed(). */ void splice_from_pipe_end(struct pipe_inode_info *pipe, struct splice_desc *sd) { if (sd->need_wakeup) wakeup_pipe_writers(pipe); } EXPORT_SYMBOL(splice_from_pipe_end); |
5274f052e
|
866 |
|
b3c2d2ddd
|
867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 |
/** * __splice_from_pipe - splice data from a pipe to given actor * @pipe: pipe to splice from * @sd: information to @actor * @actor: handler that splices the data * * Description: * This function does little more than loop over the pipe and call * @actor to do the actual moving of a single struct pipe_buffer to * the desired destination. See pipe_to_file, pipe_to_sendpage, or * pipe_to_user. * */ ssize_t __splice_from_pipe(struct pipe_inode_info *pipe, struct splice_desc *sd, splice_actor *actor) { int ret; |
5274f052e
|
884 |
|
b3c2d2ddd
|
885 886 887 888 889 890 891 892 893 |
splice_from_pipe_begin(sd); do { ret = splice_from_pipe_next(pipe, sd); if (ret > 0) ret = splice_from_pipe_feed(pipe, sd, actor); } while (ret > 0); splice_from_pipe_end(pipe, sd); return sd->num_spliced ? sd->num_spliced : ret; |
5274f052e
|
894 |
} |
40bee44ea
|
895 |
EXPORT_SYMBOL(__splice_from_pipe); |
5274f052e
|
896 |
|
932cc6d4f
|
897 898 899 900 901 902 903 904 905 906 |
/** * splice_from_pipe - splice data from a pipe to a file * @pipe: pipe to splice from * @out: file to splice to * @ppos: position in @out * @len: how many bytes to splice * @flags: splice modifier flags * @actor: handler that splices the data * * Description: |
2933970b9
|
907 |
* See __splice_from_pipe. This function locks the pipe inode, |
932cc6d4f
|
908 909 910 |
* otherwise it's identical to __splice_from_pipe(). * */ |
6da618098
|
911 912 913 914 915 |
ssize_t splice_from_pipe(struct pipe_inode_info *pipe, struct file *out, loff_t *ppos, size_t len, unsigned int flags, splice_actor *actor) { ssize_t ret; |
c66ab6fa7
|
916 917 918 919 |
struct splice_desc sd = { .total_len = len, .flags = flags, .pos = *ppos, |
6a14b90bb
|
920 |
.u.file = out, |
c66ab6fa7
|
921 |
}; |
6da618098
|
922 |
|
61e0d47c3
|
923 |
pipe_lock(pipe); |
c66ab6fa7
|
924 |
ret = __splice_from_pipe(pipe, &sd, actor); |
61e0d47c3
|
925 |
pipe_unlock(pipe); |
6da618098
|
926 927 928 929 930 |
return ret; } /** |
83f9135bd
|
931 |
* generic_file_splice_write - splice data from a pipe to a file |
3a326a2ce
|
932 |
* @pipe: pipe info |
83f9135bd
|
933 |
* @out: file to write to |
932cc6d4f
|
934 |
* @ppos: position in @out |
83f9135bd
|
935 936 937 |
* @len: number of bytes to splice * @flags: splice modifier flags * |
932cc6d4f
|
938 939 940 |
* Description: * Will either move or copy pages (determined by @flags options) from * the given pipe inode to the given file. |
83f9135bd
|
941 942 |
* */ |
3a326a2ce
|
943 944 |
ssize_t generic_file_splice_write(struct pipe_inode_info *pipe, struct file *out, |
cbb7e577e
|
945 |
loff_t *ppos, size_t len, unsigned int flags) |
5274f052e
|
946 |
{ |
4f6f0bd2f
|
947 |
struct address_space *mapping = out->f_mapping; |
8c34e2d63
|
948 |
struct inode *inode = mapping->host; |
7f3d4ee10
|
949 950 951 952 953 954 |
struct splice_desc sd = { .total_len = len, .flags = flags, .pos = *ppos, .u.file = out, }; |
3a326a2ce
|
955 |
ssize_t ret; |
61e0d47c3
|
956 |
pipe_lock(pipe); |
eb443e5a2
|
957 958 959 960 961 962 963 964 965 |
splice_from_pipe_begin(&sd); do { ret = splice_from_pipe_next(pipe, &sd); if (ret <= 0) break; mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD); ret = file_remove_suid(out); |
723590ed5
|
966 967 |
if (!ret) { file_update_time(out); |
eb443e5a2
|
968 |
ret = splice_from_pipe_feed(pipe, &sd, pipe_to_file); |
723590ed5
|
969 |
} |
eb443e5a2
|
970 971 972 |
mutex_unlock(&inode->i_mutex); } while (ret > 0); splice_from_pipe_end(pipe, &sd); |
61e0d47c3
|
973 |
pipe_unlock(pipe); |
eb443e5a2
|
974 975 976 |
if (sd.num_spliced) ret = sd.num_spliced; |
a4514ebd8
|
977 |
if (ret > 0) { |
17ee4f49a
|
978 |
unsigned long nr_pages; |
148f948ba
|
979 |
int err; |
17ee4f49a
|
980 |
|
17ee4f49a
|
981 |
nr_pages = (ret + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; |
a4514ebd8
|
982 |
|
148f948ba
|
983 984 985 986 987 |
err = generic_write_sync(out, *ppos, ret); if (err) ret = err; else *ppos += ret; |
17ee4f49a
|
988 |
balance_dirty_pages_ratelimited_nr(mapping, nr_pages); |
4f6f0bd2f
|
989 990 991 |
} return ret; |
5274f052e
|
992 |
} |
059a8f373
|
993 |
EXPORT_SYMBOL(generic_file_splice_write); |
b2858d7d1
|
994 995 |
static int write_pipe_buf(struct pipe_inode_info *pipe, struct pipe_buffer *buf, struct splice_desc *sd) |
0b0a47f5c
|
996 |
{ |
b2858d7d1
|
997 998 |
int ret; void *data; |
b2858d7d1
|
999 1000 1001 1002 1003 |
data = buf->ops->map(pipe, buf, 0); ret = kernel_write(sd->u.file, data + buf->offset, sd->len, sd->pos); buf->ops->unmap(pipe, buf, data); return ret; |
0b0a47f5c
|
1004 1005 1006 1007 1008 1009 |
} static ssize_t default_file_splice_write(struct pipe_inode_info *pipe, struct file *out, loff_t *ppos, size_t len, unsigned int flags) { |
b2858d7d1
|
1010 |
ssize_t ret; |
0b0a47f5c
|
1011 |
|
b2858d7d1
|
1012 1013 1014 |
ret = splice_from_pipe(pipe, out, ppos, len, flags, write_pipe_buf); if (ret > 0) *ppos += ret; |
0b0a47f5c
|
1015 |
|
b2858d7d1
|
1016 |
return ret; |
0b0a47f5c
|
1017 |
} |
83f9135bd
|
1018 1019 |
/** * generic_splice_sendpage - splice data from a pipe to a socket |
932cc6d4f
|
1020 |
* @pipe: pipe to splice from |
83f9135bd
|
1021 |
* @out: socket to write to |
932cc6d4f
|
1022 |
* @ppos: position in @out |
83f9135bd
|
1023 1024 1025 |
* @len: number of bytes to splice * @flags: splice modifier flags * |
932cc6d4f
|
1026 1027 1028 |
* Description: * Will send @len bytes from the pipe to a network socket. No data copying * is involved. |
83f9135bd
|
1029 1030 |
* */ |
3a326a2ce
|
1031 |
ssize_t generic_splice_sendpage(struct pipe_inode_info *pipe, struct file *out, |
cbb7e577e
|
1032 |
loff_t *ppos, size_t len, unsigned int flags) |
5274f052e
|
1033 |
{ |
00522fb41
|
1034 |
return splice_from_pipe(pipe, out, ppos, len, flags, pipe_to_sendpage); |
5274f052e
|
1035 |
} |
059a8f373
|
1036 |
EXPORT_SYMBOL(generic_splice_sendpage); |
a0f067802
|
1037 |
|
83f9135bd
|
1038 1039 1040 |
/* * Attempt to initiate a splice from pipe to file. */ |
3a326a2ce
|
1041 |
static long do_splice_from(struct pipe_inode_info *pipe, struct file *out, |
cbb7e577e
|
1042 |
loff_t *ppos, size_t len, unsigned int flags) |
5274f052e
|
1043 |
{ |
0b0a47f5c
|
1044 1045 |
ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, loff_t *, size_t, unsigned int); |
5274f052e
|
1046 |
int ret; |
49570e9b2
|
1047 |
if (unlikely(!(out->f_mode & FMODE_WRITE))) |
5274f052e
|
1048 |
return -EBADF; |
efc968d45
|
1049 1050 |
if (unlikely(out->f_flags & O_APPEND)) return -EINVAL; |
cbb7e577e
|
1051 |
ret = rw_verify_area(WRITE, out, ppos, len); |
5274f052e
|
1052 1053 |
if (unlikely(ret < 0)) return ret; |
cc56f7de7
|
1054 1055 1056 |
if (out->f_op && out->f_op->splice_write) splice_write = out->f_op->splice_write; else |
0b0a47f5c
|
1057 1058 1059 |
splice_write = default_file_splice_write; return splice_write(pipe, out, ppos, len, flags); |
5274f052e
|
1060 |
} |
83f9135bd
|
1061 1062 1063 |
/* * Attempt to initiate a splice from a file to a pipe. */ |
cbb7e577e
|
1064 1065 1066 |
static long do_splice_to(struct file *in, loff_t *ppos, struct pipe_inode_info *pipe, size_t len, unsigned int flags) |
5274f052e
|
1067 |
{ |
6818173bd
|
1068 1069 |
ssize_t (*splice_read)(struct file *, loff_t *, struct pipe_inode_info *, size_t, unsigned int); |
5274f052e
|
1070 |
int ret; |
49570e9b2
|
1071 |
if (unlikely(!(in->f_mode & FMODE_READ))) |
5274f052e
|
1072 |
return -EBADF; |
cbb7e577e
|
1073 |
ret = rw_verify_area(READ, in, ppos, len); |
5274f052e
|
1074 1075 |
if (unlikely(ret < 0)) return ret; |
cc56f7de7
|
1076 1077 1078 |
if (in->f_op && in->f_op->splice_read) splice_read = in->f_op->splice_read; else |
6818173bd
|
1079 1080 1081 |
splice_read = default_file_splice_read; return splice_read(in, ppos, pipe, len, flags); |
5274f052e
|
1082 |
} |
932cc6d4f
|
1083 1084 1085 1086 1087 1088 1089 1090 1091 |
/** * splice_direct_to_actor - splices data directly between two non-pipes * @in: file to splice from * @sd: actor information on where to splice to * @actor: handles the data splicing * * Description: * This is a special case helper to splice directly between two * points, without requiring an explicit pipe. Internally an allocated |
79685b8de
|
1092 |
* pipe is cached in the process, and reused during the lifetime of |
932cc6d4f
|
1093 1094 |
* that process. * |
c66ab6fa7
|
1095 1096 1097 |
*/ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd, splice_direct_actor *actor) |
b92ce5589
|
1098 1099 1100 1101 |
{ struct pipe_inode_info *pipe; long ret, bytes; umode_t i_mode; |
c66ab6fa7
|
1102 1103 |
size_t len; int i, flags; |
b92ce5589
|
1104 1105 1106 1107 1108 1109 |
/* * We require the input being a regular file, as we don't want to * randomly drop data for eg socket -> socket splicing. Use the * piped splicing for that! */ |
0f7fc9e4d
|
1110 |
i_mode = in->f_path.dentry->d_inode->i_mode; |
b92ce5589
|
1111 1112 1113 1114 1115 1116 1117 1118 |
if (unlikely(!S_ISREG(i_mode) && !S_ISBLK(i_mode))) return -EINVAL; /* * neither in nor out is a pipe, setup an internal pipe attached to * 'out' and transfer the wanted data from 'in' to 'out' through that */ pipe = current->splice_pipe; |
49570e9b2
|
1119 |
if (unlikely(!pipe)) { |
b92ce5589
|
1120 1121 1122 1123 1124 1125 |
pipe = alloc_pipe_info(NULL); if (!pipe) return -ENOMEM; /* * We don't have an immediate reader, but we'll read the stuff |
00522fb41
|
1126 |
* out of the pipe right after the splice_to_pipe(). So set |
b92ce5589
|
1127 1128 1129 1130 1131 1132 1133 1134 |
* PIPE_READERS appropriately. */ pipe->readers = 1; current->splice_pipe = pipe; } /* |
73d62d83e
|
1135 |
* Do the splice. |
b92ce5589
|
1136 1137 1138 |
*/ ret = 0; bytes = 0; |
c66ab6fa7
|
1139 1140 1141 1142 1143 1144 1145 |
len = sd->total_len; flags = sd->flags; /* * Don't block on output, we have to drain the direct pipe. */ sd->flags &= ~SPLICE_F_NONBLOCK; |
b92ce5589
|
1146 1147 |
while (len) { |
51a92c0f6
|
1148 |
size_t read_len; |
a82c53a0e
|
1149 |
loff_t pos = sd->pos, prev_pos = pos; |
b92ce5589
|
1150 |
|
bcd4f3acb
|
1151 |
ret = do_splice_to(in, &pos, pipe, len, flags); |
51a92c0f6
|
1152 |
if (unlikely(ret <= 0)) |
b92ce5589
|
1153 1154 1155 |
goto out_release; read_len = ret; |
c66ab6fa7
|
1156 |
sd->total_len = read_len; |
b92ce5589
|
1157 1158 1159 1160 1161 1162 |
/* * NOTE: nonblocking mode only applies to the input. We * must not do the output in nonblocking mode as then we * could get stuck data in the internal pipe: */ |
c66ab6fa7
|
1163 |
ret = actor(pipe, sd); |
a82c53a0e
|
1164 1165 |
if (unlikely(ret <= 0)) { sd->pos = prev_pos; |
b92ce5589
|
1166 |
goto out_release; |
a82c53a0e
|
1167 |
} |
b92ce5589
|
1168 1169 1170 |
bytes += ret; len -= ret; |
bcd4f3acb
|
1171 |
sd->pos = pos; |
b92ce5589
|
1172 |
|
a82c53a0e
|
1173 1174 |
if (ret < read_len) { sd->pos = prev_pos + ret; |
51a92c0f6
|
1175 |
goto out_release; |
a82c53a0e
|
1176 |
} |
b92ce5589
|
1177 |
} |
9e97198db
|
1178 |
done: |
b92ce5589
|
1179 |
pipe->nrbufs = pipe->curbuf = 0; |
808487085
|
1180 |
file_accessed(in); |
b92ce5589
|
1181 1182 1183 1184 1185 1186 1187 |
return bytes; out_release: /* * If we did an incomplete transfer we must release * the pipe buffers in question: */ |
35f3d14db
|
1188 |
for (i = 0; i < pipe->buffers; i++) { |
b92ce5589
|
1189 1190 1191 1192 1193 1194 1195 |
struct pipe_buffer *buf = pipe->bufs + i; if (buf->ops) { buf->ops->release(pipe, buf); buf->ops = NULL; } } |
b92ce5589
|
1196 |
|
9e97198db
|
1197 1198 |
if (!bytes) bytes = ret; |
c66ab6fa7
|
1199 |
|
9e97198db
|
1200 |
goto done; |
c66ab6fa7
|
1201 1202 1203 1204 1205 1206 |
} EXPORT_SYMBOL(splice_direct_to_actor); static int direct_splice_actor(struct pipe_inode_info *pipe, struct splice_desc *sd) { |
6a14b90bb
|
1207 |
struct file *file = sd->u.file; |
c66ab6fa7
|
1208 |
|
2cb4b05e7
|
1209 1210 |
return do_splice_from(pipe, file, &file->f_pos, sd->total_len, sd->flags); |
c66ab6fa7
|
1211 |
} |
932cc6d4f
|
1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 |
/** * do_splice_direct - splices data directly between two files * @in: file to splice from * @ppos: input file offset * @out: file to splice to * @len: number of bytes to splice * @flags: splice modifier flags * * Description: * For use by do_sendfile(). splice can easily emulate sendfile, but * doing it in the application would incur an extra system call * (splice in + splice out, as compared to just sendfile()). So this helper * can splice directly through a process-private pipe. * */ |
c66ab6fa7
|
1227 1228 1229 1230 1231 1232 1233 1234 |
long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, size_t len, unsigned int flags) { struct splice_desc sd = { .len = len, .total_len = len, .flags = flags, .pos = *ppos, |
6a14b90bb
|
1235 |
.u.file = out, |
c66ab6fa7
|
1236 |
}; |
51a92c0f6
|
1237 |
long ret; |
c66ab6fa7
|
1238 1239 |
ret = splice_direct_to_actor(in, &sd, direct_splice_actor); |
51a92c0f6
|
1240 |
if (ret > 0) |
a82c53a0e
|
1241 |
*ppos = sd.pos; |
51a92c0f6
|
1242 |
|
c66ab6fa7
|
1243 |
return ret; |
b92ce5589
|
1244 |
} |
7c77f0b3f
|
1245 1246 1247 |
static int splice_pipe_to_pipe(struct pipe_inode_info *ipipe, struct pipe_inode_info *opipe, size_t len, unsigned int flags); |
ddac0d39c
|
1248 1249 |
/* |
83f9135bd
|
1250 1251 |
* Determine where to splice to/from. */ |
529565dcb
|
1252 1253 1254 |
static long do_splice(struct file *in, loff_t __user *off_in, struct file *out, loff_t __user *off_out, size_t len, unsigned int flags) |
5274f052e
|
1255 |
{ |
7c77f0b3f
|
1256 1257 |
struct pipe_inode_info *ipipe; struct pipe_inode_info *opipe; |
cbb7e577e
|
1258 |
loff_t offset, *off; |
a4514ebd8
|
1259 |
long ret; |
5274f052e
|
1260 |
|
71993e62a
|
1261 1262 |
ipipe = get_pipe_info(in); opipe = get_pipe_info(out); |
7c77f0b3f
|
1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 |
if (ipipe && opipe) { if (off_in || off_out) return -ESPIPE; if (!(in->f_mode & FMODE_READ)) return -EBADF; if (!(out->f_mode & FMODE_WRITE)) return -EBADF; /* Splicing to self would be fun, but... */ if (ipipe == opipe) return -EINVAL; return splice_pipe_to_pipe(ipipe, opipe, len, flags); } if (ipipe) { |
529565dcb
|
1282 1283 |
if (off_in) return -ESPIPE; |
b92ce5589
|
1284 |
if (off_out) { |
19c9a49b4
|
1285 |
if (!(out->f_mode & FMODE_PWRITE)) |
b92ce5589
|
1286 |
return -EINVAL; |
cbb7e577e
|
1287 |
if (copy_from_user(&offset, off_out, sizeof(loff_t))) |
b92ce5589
|
1288 |
return -EFAULT; |
cbb7e577e
|
1289 1290 1291 |
off = &offset; } else off = &out->f_pos; |
529565dcb
|
1292 |
|
7c77f0b3f
|
1293 |
ret = do_splice_from(ipipe, out, off, len, flags); |
a4514ebd8
|
1294 1295 1296 1297 1298 |
if (off_out && copy_to_user(off_out, off, sizeof(loff_t))) ret = -EFAULT; return ret; |
529565dcb
|
1299 |
} |
5274f052e
|
1300 |
|
7c77f0b3f
|
1301 |
if (opipe) { |
529565dcb
|
1302 1303 |
if (off_out) return -ESPIPE; |
b92ce5589
|
1304 |
if (off_in) { |
19c9a49b4
|
1305 |
if (!(in->f_mode & FMODE_PREAD)) |
b92ce5589
|
1306 |
return -EINVAL; |
cbb7e577e
|
1307 |
if (copy_from_user(&offset, off_in, sizeof(loff_t))) |
b92ce5589
|
1308 |
return -EFAULT; |
cbb7e577e
|
1309 1310 1311 |
off = &offset; } else off = &in->f_pos; |
529565dcb
|
1312 |
|
7c77f0b3f
|
1313 |
ret = do_splice_to(in, off, opipe, len, flags); |
a4514ebd8
|
1314 1315 1316 1317 1318 |
if (off_in && copy_to_user(off_in, off, sizeof(loff_t))) ret = -EFAULT; return ret; |
529565dcb
|
1319 |
} |
5274f052e
|
1320 1321 1322 |
return -EINVAL; } |
912d35f86
|
1323 1324 1325 1326 1327 1328 1329 1330 1331 |
/* * Map an iov into an array of pages and offset/length tupples. With the * partial_page structure, we can map several non-contiguous ranges into * our ones pages[] map instead of splitting that operation into pieces. * Could easily be exported as a generic helper for other users, in which * case one would probably want to add a 'max_nr_pages' parameter as well. */ static int get_iovec_page_array(const struct iovec __user *iov, unsigned int nr_vecs, struct page **pages, |
35f3d14db
|
1332 1333 |
struct partial_page *partial, int aligned, unsigned int pipe_buffers) |
912d35f86
|
1334 1335 |
{ int buffers = 0, error = 0; |
912d35f86
|
1336 1337 |
while (nr_vecs) { unsigned long off, npages; |
757239576
|
1338 |
struct iovec entry; |
912d35f86
|
1339 1340 1341 |
void __user *base; size_t len; int i; |
757239576
|
1342 |
error = -EFAULT; |
bc40d73c9
|
1343 |
if (copy_from_user(&entry, iov, sizeof(entry))) |
912d35f86
|
1344 |
break; |
757239576
|
1345 1346 |
base = entry.iov_base; len = entry.iov_len; |
912d35f86
|
1347 1348 1349 |
/* * Sanity check this iovec. 0 read succeeds. */ |
757239576
|
1350 |
error = 0; |
912d35f86
|
1351 1352 1353 |
if (unlikely(!len)) break; error = -EFAULT; |
712a30e63
|
1354 |
if (!access_ok(VERIFY_READ, base, len)) |
912d35f86
|
1355 1356 1357 1358 1359 1360 1361 |
break; /* * Get this base offset and number of pages, then map * in the user pages. */ off = (unsigned long) base & ~PAGE_MASK; |
7afa6fd03
|
1362 1363 1364 1365 1366 1367 1368 1369 |
/* * If asked for alignment, the offset must be zero and the * length a multiple of the PAGE_SIZE. */ error = -EINVAL; if (aligned && (off || len & ~PAGE_MASK)) break; |
912d35f86
|
1370 |
npages = (off + len + PAGE_SIZE - 1) >> PAGE_SHIFT; |
35f3d14db
|
1371 1372 |
if (npages > pipe_buffers - buffers) npages = pipe_buffers - buffers; |
912d35f86
|
1373 |
|
bc40d73c9
|
1374 1375 |
error = get_user_pages_fast((unsigned long)base, npages, 0, &pages[buffers]); |
912d35f86
|
1376 1377 1378 1379 1380 1381 1382 1383 |
if (unlikely(error <= 0)) break; /* * Fill this contiguous range into the partial page map. */ for (i = 0; i < error; i++) { |
7591489a8
|
1384 |
const int plen = min_t(size_t, len, PAGE_SIZE - off); |
912d35f86
|
1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 |
partial[buffers].offset = off; partial[buffers].len = plen; off = 0; len -= plen; buffers++; } /* * We didn't complete this iov, stop here since it probably * means we have to move some of this into a pipe to * be able to continue. */ if (len) break; /* * Don't continue if we mapped fewer pages than we asked for, * or if we mapped the max number of pages that we have * room for. */ |
35f3d14db
|
1407 |
if (error < npages || buffers == pipe_buffers) |
912d35f86
|
1408 1409 1410 1411 1412 |
break; nr_vecs--; iov++; } |
912d35f86
|
1413 1414 1415 1416 1417 |
if (buffers) return buffers; return error; } |
6a14b90bb
|
1418 1419 1420 1421 1422 |
static int pipe_to_user(struct pipe_inode_info *pipe, struct pipe_buffer *buf, struct splice_desc *sd) { char *src; int ret; |
6a14b90bb
|
1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 |
/* * See if we can use the atomic maps, by prefaulting in the * pages and doing an atomic copy */ if (!fault_in_pages_writeable(sd->u.userptr, sd->len)) { src = buf->ops->map(pipe, buf, 1); ret = __copy_to_user_inatomic(sd->u.userptr, src + buf->offset, sd->len); buf->ops->unmap(pipe, buf, src); if (!ret) { ret = sd->len; goto out; } } /* * No dice, use slow non-atomic map and copy */ src = buf->ops->map(pipe, buf, 0); ret = sd->len; if (copy_to_user(sd->u.userptr, src + buf->offset, sd->len)) ret = -EFAULT; |
6866bef40
|
1446 |
buf->ops->unmap(pipe, buf, src); |
6a14b90bb
|
1447 1448 1449 |
out: if (ret > 0) sd->u.userptr += ret; |
6a14b90bb
|
1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 |
return ret; } /* * For lack of a better implementation, implement vmsplice() to userspace * as a simple copy of the pipes pages to the user iov. */ static long vmsplice_to_user(struct file *file, const struct iovec __user *iov, unsigned long nr_segs, unsigned int flags) { struct pipe_inode_info *pipe; struct splice_desc sd; ssize_t size; int error; long ret; |
71993e62a
|
1465 |
pipe = get_pipe_info(file); |
6a14b90bb
|
1466 1467 |
if (!pipe) return -EBADF; |
61e0d47c3
|
1468 |
pipe_lock(pipe); |
6a14b90bb
|
1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 |
error = ret = 0; while (nr_segs) { void __user *base; size_t len; /* * Get user address base and length for this iovec. */ error = get_user(base, &iov->iov_base); if (unlikely(error)) break; error = get_user(len, &iov->iov_len); if (unlikely(error)) break; /* * Sanity check this iovec. 0 read succeeds. */ if (unlikely(!len)) break; if (unlikely(!base)) { error = -EFAULT; break; } |
8811930dc
|
1494 1495 1496 1497 |
if (unlikely(!access_ok(VERIFY_WRITE, base, len))) { error = -EFAULT; break; } |
6a14b90bb
|
1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 |
sd.len = 0; sd.total_len = len; sd.flags = flags; sd.u.userptr = base; sd.pos = 0; size = __splice_from_pipe(pipe, &sd, pipe_to_user); if (size < 0) { if (!ret) ret = size; break; } ret += size; if (size < len) break; nr_segs--; iov++; } |
61e0d47c3
|
1520 |
pipe_unlock(pipe); |
6a14b90bb
|
1521 1522 1523 1524 1525 1526 |
if (!ret) ret = error; return ret; } |
912d35f86
|
1527 1528 1529 1530 |
/* * vmsplice splices a user address range into a pipe. It can be thought of * as splice-from-memory, where the regular splice is splice-from-file (or * to file). In both cases the output is a pipe, naturally. |
912d35f86
|
1531 |
*/ |
6a14b90bb
|
1532 1533 |
static long vmsplice_to_pipe(struct file *file, const struct iovec __user *iov, unsigned long nr_segs, unsigned int flags) |
912d35f86
|
1534 |
{ |
ddac0d39c
|
1535 |
struct pipe_inode_info *pipe; |
35f3d14db
|
1536 1537 |
struct page *pages[PIPE_DEF_BUFFERS]; struct partial_page partial[PIPE_DEF_BUFFERS]; |
912d35f86
|
1538 1539 1540 1541 1542 |
struct splice_pipe_desc spd = { .pages = pages, .partial = partial, .flags = flags, .ops = &user_page_pipe_buf_ops, |
bbdfc2f70
|
1543 |
.spd_release = spd_release_page, |
912d35f86
|
1544 |
}; |
35f3d14db
|
1545 |
long ret; |
912d35f86
|
1546 |
|
71993e62a
|
1547 |
pipe = get_pipe_info(file); |
ddac0d39c
|
1548 |
if (!pipe) |
912d35f86
|
1549 |
return -EBADF; |
912d35f86
|
1550 |
|
35f3d14db
|
1551 1552 1553 1554 1555 1556 |
if (splice_grow_spd(pipe, &spd)) return -ENOMEM; spd.nr_pages = get_iovec_page_array(iov, nr_segs, spd.pages, spd.partial, flags & SPLICE_F_GIFT, pipe->buffers); |
912d35f86
|
1557 |
if (spd.nr_pages <= 0) |
35f3d14db
|
1558 1559 1560 |
ret = spd.nr_pages; else ret = splice_to_pipe(pipe, &spd); |
912d35f86
|
1561 |
|
35f3d14db
|
1562 1563 |
splice_shrink_spd(pipe, &spd); return ret; |
912d35f86
|
1564 |
} |
6a14b90bb
|
1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 |
/* * Note that vmsplice only really supports true splicing _from_ user memory * to a pipe, not the other way around. Splicing from user memory is a simple * operation that can be supported without any funky alignment restrictions * or nasty vm tricks. We simply map in the user memory and fill them into * a pipe. The reverse isn't quite as easy, though. There are two possible * solutions for that: * * - memcpy() the data internally, at which point we might as well just * do a regular read() on the buffer anyway. * - Lots of nasty vm tricks, that are neither fast nor flexible (it * has restriction limitations on both ends of the pipe). * * Currently we punt and implement it as a normal copy, see pipe_to_user(). * */ |
836f92adf
|
1581 1582 |
SYSCALL_DEFINE4(vmsplice, int, fd, const struct iovec __user *, iov, unsigned long, nr_segs, unsigned int, flags) |
912d35f86
|
1583 1584 1585 1586 |
{ struct file *file; long error; int fput; |
6a14b90bb
|
1587 1588 1589 1590 |
if (unlikely(nr_segs > UIO_MAXIOV)) return -EINVAL; else if (unlikely(!nr_segs)) return 0; |
912d35f86
|
1591 1592 1593 1594 |
error = -EBADF; file = fget_light(fd, &fput); if (file) { if (file->f_mode & FMODE_WRITE) |
6a14b90bb
|
1595 1596 1597 |
error = vmsplice_to_pipe(file, iov, nr_segs, flags); else if (file->f_mode & FMODE_READ) error = vmsplice_to_user(file, iov, nr_segs, flags); |
912d35f86
|
1598 1599 1600 1601 1602 1603 |
fput_light(file, fput); } return error; } |
836f92adf
|
1604 1605 1606 |
SYSCALL_DEFINE6(splice, int, fd_in, loff_t __user *, off_in, int, fd_out, loff_t __user *, off_out, size_t, len, unsigned int, flags) |
5274f052e
|
1607 1608 1609 1610 1611 1612 1613 1614 1615 |
{ long error; struct file *in, *out; int fput_in, fput_out; if (unlikely(!len)) return 0; error = -EBADF; |
529565dcb
|
1616 |
in = fget_light(fd_in, &fput_in); |
5274f052e
|
1617 1618 |
if (in) { if (in->f_mode & FMODE_READ) { |
529565dcb
|
1619 |
out = fget_light(fd_out, &fput_out); |
5274f052e
|
1620 1621 |
if (out) { if (out->f_mode & FMODE_WRITE) |
529565dcb
|
1622 1623 1624 |
error = do_splice(in, off_in, out, off_out, len, flags); |
5274f052e
|
1625 1626 1627 1628 1629 1630 1631 1632 1633 |
fput_light(out, fput_out); } } fput_light(in, fput_in); } return error; } |
70524490e
|
1634 1635 |
/* |
aadd06e5c
|
1636 1637 1638 |
* Make sure there's data to read. Wait for input if we can, otherwise * return an appropriate error. */ |
7c77f0b3f
|
1639 |
static int ipipe_prep(struct pipe_inode_info *pipe, unsigned int flags) |
aadd06e5c
|
1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 |
{ int ret; /* * Check ->nrbufs without the inode lock first. This function * is speculative anyways, so missing one is ok. */ if (pipe->nrbufs) return 0; ret = 0; |
61e0d47c3
|
1651 |
pipe_lock(pipe); |
aadd06e5c
|
1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 |
while (!pipe->nrbufs) { if (signal_pending(current)) { ret = -ERESTARTSYS; break; } if (!pipe->writers) break; if (!pipe->waiting_writers) { if (flags & SPLICE_F_NONBLOCK) { ret = -EAGAIN; break; } } pipe_wait(pipe); } |
61e0d47c3
|
1668 |
pipe_unlock(pipe); |
aadd06e5c
|
1669 1670 1671 1672 1673 1674 1675 |
return ret; } /* * Make sure there's writeable room. Wait for room if we can, otherwise * return an appropriate error. */ |
7c77f0b3f
|
1676 |
static int opipe_prep(struct pipe_inode_info *pipe, unsigned int flags) |
aadd06e5c
|
1677 1678 1679 1680 1681 1682 1683 |
{ int ret; /* * Check ->nrbufs without the inode lock first. This function * is speculative anyways, so missing one is ok. */ |
35f3d14db
|
1684 |
if (pipe->nrbufs < pipe->buffers) |
aadd06e5c
|
1685 1686 1687 |
return 0; ret = 0; |
61e0d47c3
|
1688 |
pipe_lock(pipe); |
aadd06e5c
|
1689 |
|
35f3d14db
|
1690 |
while (pipe->nrbufs >= pipe->buffers) { |
aadd06e5c
|
1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 |
if (!pipe->readers) { send_sig(SIGPIPE, current, 0); ret = -EPIPE; break; } if (flags & SPLICE_F_NONBLOCK) { ret = -EAGAIN; break; } if (signal_pending(current)) { ret = -ERESTARTSYS; break; } pipe->waiting_writers++; pipe_wait(pipe); pipe->waiting_writers--; } |
61e0d47c3
|
1708 |
pipe_unlock(pipe); |
aadd06e5c
|
1709 1710 1711 1712 |
return ret; } /* |
7c77f0b3f
|
1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 |
* Splice contents of ipipe to opipe. */ static int splice_pipe_to_pipe(struct pipe_inode_info *ipipe, struct pipe_inode_info *opipe, size_t len, unsigned int flags) { struct pipe_buffer *ibuf, *obuf; int ret = 0, nbuf; bool input_wakeup = false; retry: ret = ipipe_prep(ipipe, flags); if (ret) return ret; ret = opipe_prep(opipe, flags); if (ret) return ret; /* * Potential ABBA deadlock, work around it by ordering lock * grabbing by pipe info address. Otherwise two different processes * could deadlock (one doing tee from A -> B, the other from B -> A). */ pipe_double_lock(ipipe, opipe); do { if (!opipe->readers) { send_sig(SIGPIPE, current, 0); if (!ret) ret = -EPIPE; break; } if (!ipipe->nrbufs && !ipipe->writers) break; /* * Cannot make any progress, because either the input * pipe is empty or the output pipe is full. */ |
35f3d14db
|
1755 |
if (!ipipe->nrbufs || opipe->nrbufs >= opipe->buffers) { |
7c77f0b3f
|
1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 |
/* Already processed some buffers, break */ if (ret) break; if (flags & SPLICE_F_NONBLOCK) { ret = -EAGAIN; break; } /* * We raced with another reader/writer and haven't * managed to process any buffers. A zero return * value means EOF, so retry instead. */ pipe_unlock(ipipe); pipe_unlock(opipe); goto retry; } ibuf = ipipe->bufs + ipipe->curbuf; |
35f3d14db
|
1776 |
nbuf = (opipe->curbuf + opipe->nrbufs) & (opipe->buffers - 1); |
7c77f0b3f
|
1777 1778 1779 1780 1781 1782 1783 1784 1785 |
obuf = opipe->bufs + nbuf; if (len >= ibuf->len) { /* * Simply move the whole buffer from ipipe to opipe */ *obuf = *ibuf; ibuf->ops = NULL; opipe->nrbufs++; |
35f3d14db
|
1786 |
ipipe->curbuf = (ipipe->curbuf + 1) & (ipipe->buffers - 1); |
7c77f0b3f
|
1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 |
ipipe->nrbufs--; input_wakeup = true; } else { /* * Get a reference to this pipe buffer, * so we can copy the contents over. */ ibuf->ops->get(ipipe, ibuf); *obuf = *ibuf; /* * Don't inherit the gift flag, we need to * prevent multiple steals of this page. */ obuf->flags &= ~PIPE_BUF_FLAG_GIFT; obuf->len = len; opipe->nrbufs++; ibuf->offset += obuf->len; ibuf->len -= obuf->len; } ret += obuf->len; len -= obuf->len; } while (len); pipe_unlock(ipipe); pipe_unlock(opipe); /* * If we put data in the output pipe, wakeup any potential readers. */ |
825cdcb1a
|
1818 1819 |
if (ret > 0) wakeup_pipe_readers(opipe); |
7c77f0b3f
|
1820 1821 1822 1823 1824 1825 1826 |
if (input_wakeup) wakeup_pipe_writers(ipipe); return ret; } /* |
70524490e
|
1827 1828 1829 1830 1831 1832 1833 |
* Link contents of ipipe to opipe. */ static int link_pipe(struct pipe_inode_info *ipipe, struct pipe_inode_info *opipe, size_t len, unsigned int flags) { struct pipe_buffer *ibuf, *obuf; |
aadd06e5c
|
1834 |
int ret = 0, i = 0, nbuf; |
70524490e
|
1835 1836 1837 |
/* * Potential ABBA deadlock, work around it by ordering lock |
61e0d47c3
|
1838 |
* grabbing by pipe info address. Otherwise two different processes |
70524490e
|
1839 1840 |
* could deadlock (one doing tee from A -> B, the other from B -> A). */ |
61e0d47c3
|
1841 |
pipe_double_lock(ipipe, opipe); |
70524490e
|
1842 |
|
aadd06e5c
|
1843 |
do { |
70524490e
|
1844 1845 1846 1847 1848 1849 |
if (!opipe->readers) { send_sig(SIGPIPE, current, 0); if (!ret) ret = -EPIPE; break; } |
70524490e
|
1850 |
|
aadd06e5c
|
1851 1852 1853 1854 |
/* * If we have iterated all input buffers or ran out of * output room, break. */ |
35f3d14db
|
1855 |
if (i >= ipipe->nrbufs || opipe->nrbufs >= opipe->buffers) |
aadd06e5c
|
1856 |
break; |
70524490e
|
1857 |
|
35f3d14db
|
1858 1859 |
ibuf = ipipe->bufs + ((ipipe->curbuf + i) & (ipipe->buffers-1)); nbuf = (opipe->curbuf + opipe->nrbufs) & (opipe->buffers - 1); |
70524490e
|
1860 1861 |
/* |
aadd06e5c
|
1862 1863 |
* Get a reference to this pipe buffer, * so we can copy the contents over. |
70524490e
|
1864 |
*/ |
aadd06e5c
|
1865 1866 1867 1868 |
ibuf->ops->get(ipipe, ibuf); obuf = opipe->bufs + nbuf; *obuf = *ibuf; |
2a27250e6
|
1869 |
/* |
aadd06e5c
|
1870 1871 |
* Don't inherit the gift flag, we need to * prevent multiple steals of this page. |
2a27250e6
|
1872 |
*/ |
aadd06e5c
|
1873 |
obuf->flags &= ~PIPE_BUF_FLAG_GIFT; |
70524490e
|
1874 |
|
aadd06e5c
|
1875 1876 |
if (obuf->len > len) obuf->len = len; |
70524490e
|
1877 |
|
aadd06e5c
|
1878 1879 1880 1881 1882 |
opipe->nrbufs++; ret += obuf->len; len -= obuf->len; i++; } while (len); |
70524490e
|
1883 |
|
02cf01aea
|
1884 1885 1886 1887 1888 1889 |
/* * return EAGAIN if we have the potential of some data in the * future, otherwise just return 0 */ if (!ret && ipipe->waiting_writers && (flags & SPLICE_F_NONBLOCK)) ret = -EAGAIN; |
61e0d47c3
|
1890 1891 |
pipe_unlock(ipipe); pipe_unlock(opipe); |
70524490e
|
1892 |
|
aadd06e5c
|
1893 1894 1895 |
/* * If we put data in the output pipe, wakeup any potential readers. */ |
825cdcb1a
|
1896 1897 |
if (ret > 0) wakeup_pipe_readers(opipe); |
70524490e
|
1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 |
return ret; } /* * This is a tee(1) implementation that works on pipes. It doesn't copy * any data, it simply references the 'in' pages on the 'out' pipe. * The 'flags' used are the SPLICE_F_* variants, currently the only * applicable one is SPLICE_F_NONBLOCK. */ static long do_tee(struct file *in, struct file *out, size_t len, unsigned int flags) { |
71993e62a
|
1911 1912 |
struct pipe_inode_info *ipipe = get_pipe_info(in); struct pipe_inode_info *opipe = get_pipe_info(out); |
aadd06e5c
|
1913 |
int ret = -EINVAL; |
70524490e
|
1914 1915 |
/* |
aadd06e5c
|
1916 1917 |
* Duplicate the contents of ipipe to opipe without actually * copying the data. |
70524490e
|
1918 |
*/ |
aadd06e5c
|
1919 1920 1921 1922 1923 |
if (ipipe && opipe && ipipe != opipe) { /* * Keep going, unless we encounter an error. The ipipe/opipe * ordering doesn't really matter. */ |
7c77f0b3f
|
1924 |
ret = ipipe_prep(ipipe, flags); |
aadd06e5c
|
1925 |
if (!ret) { |
7c77f0b3f
|
1926 |
ret = opipe_prep(opipe, flags); |
02cf01aea
|
1927 |
if (!ret) |
aadd06e5c
|
1928 |
ret = link_pipe(ipipe, opipe, len, flags); |
aadd06e5c
|
1929 1930 |
} } |
70524490e
|
1931 |
|
aadd06e5c
|
1932 |
return ret; |
70524490e
|
1933 |
} |
836f92adf
|
1934 |
SYSCALL_DEFINE4(tee, int, fdin, int, fdout, size_t, len, unsigned int, flags) |
70524490e
|
1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 |
{ struct file *in; int error, fput_in; if (unlikely(!len)) return 0; error = -EBADF; in = fget_light(fdin, &fput_in); if (in) { if (in->f_mode & FMODE_READ) { int fput_out; struct file *out = fget_light(fdout, &fput_out); if (out) { if (out->f_mode & FMODE_WRITE) error = do_tee(in, out, len, flags); fput_light(out, fput_out); } } fput_light(in, fput_in); } return error; } |