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