Blame view
fs/splice.c
38.9 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> |
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 |
|
ca39d651d
|
61 62 |
if (PagePrivate(page) && !try_to_release_page(page, GFP_KERNEL)) goto out_unlock; |
4f6f0bd2f
|
63 |
|
9e94cd4fd
|
64 65 66 67 68 69 70 71 |
/* * 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
|
72 |
} |
5abc97aa2
|
73 |
|
9e94cd4fd
|
74 75 76 77 |
/* * Raced with truncate or failed to remove page from current * address space, unlock and return failure. */ |
ca39d651d
|
78 |
out_unlock: |
9e94cd4fd
|
79 80 |
unlock_page(page); return 1; |
5abc97aa2
|
81 |
} |
76ad4d111
|
82 |
static void page_cache_pipe_buf_release(struct pipe_inode_info *pipe, |
5274f052e
|
83 84 85 |
struct pipe_buffer *buf) { page_cache_release(buf->page); |
1432873af
|
86 |
buf->flags &= ~PIPE_BUF_FLAG_LRU; |
5274f052e
|
87 |
} |
0845718da
|
88 89 90 91 |
/* * Check whether the contents of buf is OK to access. Since the content * is a page cache page, IO may be in flight. */ |
cac36bb06
|
92 93 |
static int page_cache_pipe_buf_confirm(struct pipe_inode_info *pipe, struct pipe_buffer *buf) |
5274f052e
|
94 95 |
{ struct page *page = buf->page; |
49d0b21be
|
96 |
int err; |
5274f052e
|
97 98 |
if (!PageUptodate(page)) { |
49d0b21be
|
99 100 101 102 |
lock_page(page); /* * Page got truncated/unhashed. This will cause a 0-byte |
73d62d83e
|
103 |
* splice, if this is the first page. |
49d0b21be
|
104 105 106 107 108 |
*/ if (!page->mapping) { err = -ENODATA; goto error; } |
5274f052e
|
109 |
|
49d0b21be
|
110 |
/* |
73d62d83e
|
111 |
* Uh oh, read-error from disk. |
49d0b21be
|
112 113 114 115 116 117 118 |
*/ if (!PageUptodate(page)) { err = -EIO; goto error; } /* |
f84d75199
|
119 |
* Page is ok afterall, we are done. |
49d0b21be
|
120 |
*/ |
5274f052e
|
121 |
unlock_page(page); |
5274f052e
|
122 |
} |
f84d75199
|
123 |
return 0; |
49d0b21be
|
124 125 |
error: unlock_page(page); |
f84d75199
|
126 |
return err; |
70524490e
|
127 |
} |
d4c3cca94
|
128 |
static const struct pipe_buf_operations page_cache_pipe_buf_ops = { |
5274f052e
|
129 |
.can_merge = 0, |
f84d75199
|
130 131 |
.map = generic_pipe_buf_map, .unmap = generic_pipe_buf_unmap, |
cac36bb06
|
132 |
.confirm = page_cache_pipe_buf_confirm, |
5274f052e
|
133 |
.release = page_cache_pipe_buf_release, |
5abc97aa2
|
134 |
.steal = page_cache_pipe_buf_steal, |
f84d75199
|
135 |
.get = generic_pipe_buf_get, |
5274f052e
|
136 |
}; |
912d35f86
|
137 138 139 |
static int user_page_pipe_buf_steal(struct pipe_inode_info *pipe, struct pipe_buffer *buf) { |
7afa6fd03
|
140 141 |
if (!(buf->flags & PIPE_BUF_FLAG_GIFT)) return 1; |
1432873af
|
142 |
buf->flags |= PIPE_BUF_FLAG_LRU; |
330ab7161
|
143 |
return generic_pipe_buf_steal(pipe, buf); |
912d35f86
|
144 |
} |
d4c3cca94
|
145 |
static const struct pipe_buf_operations user_page_pipe_buf_ops = { |
912d35f86
|
146 |
.can_merge = 0, |
f84d75199
|
147 148 |
.map = generic_pipe_buf_map, .unmap = generic_pipe_buf_unmap, |
cac36bb06
|
149 |
.confirm = generic_pipe_buf_confirm, |
912d35f86
|
150 151 |
.release = page_cache_pipe_buf_release, .steal = user_page_pipe_buf_steal, |
f84d75199
|
152 |
.get = generic_pipe_buf_get, |
912d35f86
|
153 |
}; |
932cc6d4f
|
154 155 156 157 158 159 |
/** * splice_to_pipe - fill passed data into a pipe * @pipe: pipe to fill * @spd: data to fill * * Description: |
79685b8de
|
160 |
* @spd contains a map of pages and len/offset tuples, along with |
932cc6d4f
|
161 162 163 |
* the struct pipe_buf_operations associated with these pages. This * function will link that data to the pipe. * |
83f9135bd
|
164 |
*/ |
d6b29d7ce
|
165 166 |
ssize_t splice_to_pipe(struct pipe_inode_info *pipe, struct splice_pipe_desc *spd) |
5274f052e
|
167 |
{ |
00de00bda
|
168 |
unsigned int spd_pages = spd->nr_pages; |
912d35f86
|
169 |
int ret, do_wakeup, page_nr; |
5274f052e
|
170 171 172 |
ret = 0; do_wakeup = 0; |
912d35f86
|
173 |
page_nr = 0; |
5274f052e
|
174 |
|
3a326a2ce
|
175 176 |
if (pipe->inode) mutex_lock(&pipe->inode->i_mutex); |
5274f052e
|
177 |
|
5274f052e
|
178 |
for (;;) { |
3a326a2ce
|
179 |
if (!pipe->readers) { |
5274f052e
|
180 181 182 183 184 |
send_sig(SIGPIPE, current, 0); if (!ret) ret = -EPIPE; break; } |
6f767b042
|
185 186 |
if (pipe->nrbufs < PIPE_BUFFERS) { int newbuf = (pipe->curbuf + pipe->nrbufs) & (PIPE_BUFFERS - 1); |
3a326a2ce
|
187 |
struct pipe_buffer *buf = pipe->bufs + newbuf; |
5274f052e
|
188 |
|
912d35f86
|
189 190 191 |
buf->page = spd->pages[page_nr]; buf->offset = spd->partial[page_nr].offset; buf->len = spd->partial[page_nr].len; |
497f9625c
|
192 |
buf->private = spd->partial[page_nr].private; |
912d35f86
|
193 |
buf->ops = spd->ops; |
7afa6fd03
|
194 195 |
if (spd->flags & SPLICE_F_GIFT) buf->flags |= PIPE_BUF_FLAG_GIFT; |
6f767b042
|
196 |
pipe->nrbufs++; |
912d35f86
|
197 198 |
page_nr++; ret += buf->len; |
6f767b042
|
199 200 |
if (pipe->inode) do_wakeup = 1; |
5274f052e
|
201 |
|
912d35f86
|
202 |
if (!--spd->nr_pages) |
5274f052e
|
203 |
break; |
6f767b042
|
204 |
if (pipe->nrbufs < PIPE_BUFFERS) |
5274f052e
|
205 206 207 208 |
continue; break; } |
912d35f86
|
209 |
if (spd->flags & SPLICE_F_NONBLOCK) { |
29e350944
|
210 211 212 213 |
if (!ret) ret = -EAGAIN; break; } |
5274f052e
|
214 215 216 217 218 219 220 |
if (signal_pending(current)) { if (!ret) ret = -ERESTARTSYS; break; } if (do_wakeup) { |
c0bd1f650
|
221 |
smp_mb(); |
3a326a2ce
|
222 223 224 |
if (waitqueue_active(&pipe->wait)) wake_up_interruptible_sync(&pipe->wait); kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); |
5274f052e
|
225 226 |
do_wakeup = 0; } |
3a326a2ce
|
227 228 229 |
pipe->waiting_writers++; pipe_wait(pipe); pipe->waiting_writers--; |
5274f052e
|
230 |
} |
02676e5ae
|
231 |
if (pipe->inode) { |
3a326a2ce
|
232 |
mutex_unlock(&pipe->inode->i_mutex); |
5274f052e
|
233 |
|
02676e5ae
|
234 235 236 237 238 239 |
if (do_wakeup) { smp_mb(); if (waitqueue_active(&pipe->wait)) wake_up_interruptible(&pipe->wait); kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); } |
5274f052e
|
240 |
} |
00de00bda
|
241 |
while (page_nr < spd_pages) |
bbdfc2f70
|
242 |
spd->spd_release(spd, page_nr++); |
5274f052e
|
243 244 245 |
return ret; } |
bbdfc2f70
|
246 247 248 249 |
static void spd_release_page(struct splice_pipe_desc *spd, unsigned int i) { page_cache_release(spd->pages[i]); } |
3a326a2ce
|
250 |
static int |
cbb7e577e
|
251 252 253 |
__generic_file_splice_read(struct file *in, loff_t *ppos, struct pipe_inode_info *pipe, size_t len, unsigned int flags) |
5274f052e
|
254 255 |
{ struct address_space *mapping = in->f_mapping; |
d8983910a
|
256 |
unsigned int loff, nr_pages, req_pages; |
16c523dda
|
257 |
struct page *pages[PIPE_BUFFERS]; |
912d35f86
|
258 |
struct partial_page partial[PIPE_BUFFERS]; |
5274f052e
|
259 |
struct page *page; |
91ad66ef4
|
260 261 |
pgoff_t index, end_index; loff_t isize; |
eb20796bf
|
262 |
int error, page_nr; |
912d35f86
|
263 264 265 266 267 |
struct splice_pipe_desc spd = { .pages = pages, .partial = partial, .flags = flags, .ops = &page_cache_pipe_buf_ops, |
bbdfc2f70
|
268 |
.spd_release = spd_release_page, |
912d35f86
|
269 |
}; |
5274f052e
|
270 |
|
cbb7e577e
|
271 |
index = *ppos >> PAGE_CACHE_SHIFT; |
912d35f86
|
272 |
loff = *ppos & ~PAGE_CACHE_MASK; |
d8983910a
|
273 274 |
req_pages = (len + loff + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; nr_pages = min(req_pages, (unsigned)PIPE_BUFFERS); |
5274f052e
|
275 276 |
/* |
eb20796bf
|
277 278 279 |
* Lookup the (hopefully) full range of pages we need. */ spd.nr_pages = find_get_pages_contig(mapping, index, nr_pages, pages); |
431a4820b
|
280 |
index += spd.nr_pages; |
82aa5d618
|
281 |
|
eb20796bf
|
282 283 |
/* * If find_get_pages_contig() returned fewer pages than we needed, |
431a4820b
|
284 |
* readahead/allocate the rest and fill in the holes. |
eb20796bf
|
285 |
*/ |
431a4820b
|
286 |
if (spd.nr_pages < nr_pages) |
cf914a7d6
|
287 288 |
page_cache_sync_readahead(mapping, &in->f_ra, in, index, req_pages - spd.nr_pages); |
431a4820b
|
289 |
|
932cc6d4f
|
290 |
error = 0; |
eb20796bf
|
291 |
while (spd.nr_pages < nr_pages) { |
82aa5d618
|
292 |
/* |
eb20796bf
|
293 294 |
* Page could be there, find_get_pages_contig() breaks on * the first hole. |
5274f052e
|
295 |
*/ |
7480a9043
|
296 297 |
page = find_get_page(mapping, index); if (!page) { |
e27dedd84
|
298 |
/* |
eb20796bf
|
299 |
* page didn't exist, allocate one. |
7480a9043
|
300 301 302 303 304 305 |
*/ page = page_cache_alloc_cold(mapping); if (!page) break; error = add_to_page_cache_lru(page, mapping, index, |
4cd135046
|
306 |
mapping_gfp_mask(mapping)); |
7480a9043
|
307 308 |
if (unlikely(error)) { page_cache_release(page); |
a0548871e
|
309 310 |
if (error == -EEXIST) continue; |
7480a9043
|
311 312 |
break; } |
eb20796bf
|
313 314 315 316 317 |
/* * add_to_page_cache() locks the page, unlock it * to avoid convoluting the logic below even more. */ unlock_page(page); |
7480a9043
|
318 |
} |
eb20796bf
|
319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 |
pages[spd.nr_pages++] = page; 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); page = pages[page_nr]; |
a08a166fe
|
341 |
if (PageReadahead(page)) |
cf914a7d6
|
342 |
page_cache_async_readahead(mapping, &in->f_ra, in, |
d8983910a
|
343 |
page, index, req_pages - page_nr); |
a08a166fe
|
344 |
|
7480a9043
|
345 346 347 348 |
/* * If the page isn't uptodate, we may need to start io on it */ if (!PageUptodate(page)) { |
c4f895cbe
|
349 350 351 352 |
/* * If in nonblock mode then dont block on waiting * for an in-flight io page */ |
9ae9d68cb
|
353 |
if (flags & SPLICE_F_NONBLOCK) { |
529ae9aaa
|
354 |
if (!trylock_page(page)) { |
8191ecd1d
|
355 |
error = -EAGAIN; |
9ae9d68cb
|
356 |
break; |
8191ecd1d
|
357 |
} |
9ae9d68cb
|
358 359 |
} else lock_page(page); |
7480a9043
|
360 361 |
/* |
32502b841
|
362 363 364 365 |
* 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
|
366 367 368 |
*/ if (!page->mapping) { unlock_page(page); |
32502b841
|
369 370 371 372 373 374 375 376 377 |
page = find_or_create_page(mapping, index, mapping_gfp_mask(mapping)); if (!page) { error = -ENOMEM; break; } page_cache_release(pages[page_nr]); pages[page_nr] = page; |
7480a9043
|
378 379 380 381 382 383 384 385 |
} /* * page was already under io and is now done, great */ if (PageUptodate(page)) { unlock_page(page); goto fill_it; } |
5274f052e
|
386 |
|
7480a9043
|
387 388 389 390 |
/* * need to read in the page */ error = mapping->a_ops->readpage(in, page); |
5274f052e
|
391 |
if (unlikely(error)) { |
eb20796bf
|
392 393 394 395 396 397 |
/* * 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
|
398 |
if (error == AOP_TRUNCATED_PAGE) |
eb20796bf
|
399 |
error = 0; |
5274f052e
|
400 401 |
break; } |
620a324b7
|
402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 |
} 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
|
418 419 |
/* |
620a324b7
|
420 |
* max good bytes in this page |
91ad66ef4
|
421 |
*/ |
620a324b7
|
422 423 |
plen = ((isize - 1) & ~PAGE_CACHE_MASK) + 1; if (plen <= loff) |
91ad66ef4
|
424 |
break; |
91ad66ef4
|
425 426 |
/* |
620a324b7
|
427 |
* force quit after adding this page |
91ad66ef4
|
428 |
*/ |
620a324b7
|
429 430 |
this_len = min(this_len, plen - loff); len = this_len; |
5274f052e
|
431 |
} |
620a324b7
|
432 |
|
eb20796bf
|
433 434 |
partial[page_nr].offset = loff; partial[page_nr].len = this_len; |
82aa5d618
|
435 |
len -= this_len; |
91ad66ef4
|
436 |
loff = 0; |
eb20796bf
|
437 438 |
spd.nr_pages++; index++; |
5274f052e
|
439 |
} |
eb20796bf
|
440 |
/* |
475ecade6
|
441 |
* Release any pages at the end, if we quit early. 'page_nr' is how far |
eb20796bf
|
442 443 444 445 |
* we got, 'nr_pages' is how many pages are in the map. */ while (page_nr < nr_pages) page_cache_release(pages[page_nr++]); |
f4e6b498d
|
446 |
in->f_ra.prev_pos = (loff_t)index << PAGE_CACHE_SHIFT; |
eb20796bf
|
447 |
|
912d35f86
|
448 |
if (spd.nr_pages) |
00522fb41
|
449 |
return splice_to_pipe(pipe, &spd); |
5274f052e
|
450 |
|
7480a9043
|
451 |
return error; |
5274f052e
|
452 |
} |
83f9135bd
|
453 454 455 |
/** * generic_file_splice_read - splice data from file to a pipe * @in: file to splice from |
932cc6d4f
|
456 |
* @ppos: position in @in |
83f9135bd
|
457 458 459 460 |
* @pipe: pipe to splice to * @len: number of bytes to splice * @flags: splice modifier flags * |
932cc6d4f
|
461 462 463 464 465 |
* 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
|
466 |
*/ |
cbb7e577e
|
467 468 469 |
ssize_t generic_file_splice_read(struct file *in, loff_t *ppos, struct pipe_inode_info *pipe, size_t len, unsigned int flags) |
5274f052e
|
470 |
{ |
d366d3988
|
471 |
loff_t isize, left; |
8191ecd1d
|
472 |
int ret; |
d366d3988
|
473 474 475 476 477 478 479 480 |
isize = i_size_read(in->f_mapping->host); if (unlikely(*ppos >= isize)) return 0; left = isize - *ppos; if (unlikely(left < len)) len = left; |
5274f052e
|
481 |
|
8191ecd1d
|
482 483 |
ret = __generic_file_splice_read(in, ppos, pipe, len, flags); if (ret > 0) |
cbb7e577e
|
484 |
*ppos += ret; |
5274f052e
|
485 486 487 |
return ret; } |
059a8f373
|
488 |
EXPORT_SYMBOL(generic_file_splice_read); |
5274f052e
|
489 |
/* |
4f6f0bd2f
|
490 |
* Send 'sd->len' bytes to socket from 'sd->file' at position 'sd->pos' |
016b661e2
|
491 |
* using sendpage(). Return the number of bytes sent. |
5274f052e
|
492 |
*/ |
76ad4d111
|
493 |
static int pipe_to_sendpage(struct pipe_inode_info *pipe, |
5274f052e
|
494 495 |
struct pipe_buffer *buf, struct splice_desc *sd) { |
6a14b90bb
|
496 |
struct file *file = sd->u.file; |
5274f052e
|
497 |
loff_t pos = sd->pos; |
f84d75199
|
498 |
int ret, more; |
5274f052e
|
499 |
|
cac36bb06
|
500 |
ret = buf->ops->confirm(pipe, buf); |
f84d75199
|
501 502 |
if (!ret) { more = (sd->flags & SPLICE_F_MORE) || sd->len < sd->total_len; |
5274f052e
|
503 |
|
f84d75199
|
504 505 506 |
ret = file->f_op->sendpage(file, buf->page, buf->offset, sd->len, &pos, more); } |
5274f052e
|
507 |
|
016b661e2
|
508 |
return ret; |
5274f052e
|
509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 |
} /* * 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
|
524 525 526 527 528 529 |
* 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
|
530 |
*/ |
76ad4d111
|
531 |
static int pipe_to_file(struct pipe_inode_info *pipe, struct pipe_buffer *buf, |
5274f052e
|
532 533 |
struct splice_desc *sd) { |
6a14b90bb
|
534 |
struct file *file = sd->u.file; |
5274f052e
|
535 |
struct address_space *mapping = file->f_mapping; |
016b661e2
|
536 |
unsigned int offset, this_len; |
5274f052e
|
537 |
struct page *page; |
afddba49d
|
538 |
void *fsdata; |
3e7ee3e7b
|
539 |
int ret; |
5274f052e
|
540 541 |
/* |
49d0b21be
|
542 |
* make sure the data in this buffer is uptodate |
5274f052e
|
543 |
*/ |
cac36bb06
|
544 |
ret = buf->ops->confirm(pipe, buf); |
f84d75199
|
545 546 |
if (unlikely(ret)) return ret; |
5274f052e
|
547 |
|
5274f052e
|
548 |
offset = sd->pos & ~PAGE_CACHE_MASK; |
016b661e2
|
549 550 551 |
this_len = sd->len; if (this_len + offset > PAGE_CACHE_SIZE) this_len = PAGE_CACHE_SIZE - offset; |
afddba49d
|
552 553 554 555 |
ret = pagecache_write_begin(file, mapping, sd->pos, this_len, AOP_FLAG_UNINTERRUPTIBLE, &page, &fsdata); if (unlikely(ret)) goto out; |
5274f052e
|
556 |
|
0568b409c
|
557 |
if (buf->page != page) { |
f84d75199
|
558 559 560 |
/* * Careful, ->map() uses KM_USER0! */ |
76ad4d111
|
561 |
char *src = buf->ops->map(pipe, buf, 1); |
f84d75199
|
562 |
char *dst = kmap_atomic(page, KM_USER1); |
5abc97aa2
|
563 |
|
016b661e2
|
564 |
memcpy(dst + offset, src + buf->offset, this_len); |
5abc97aa2
|
565 |
flush_dcache_page(page); |
f84d75199
|
566 |
kunmap_atomic(dst, KM_USER1); |
76ad4d111
|
567 |
buf->ops->unmap(pipe, buf, src); |
5abc97aa2
|
568 |
} |
afddba49d
|
569 570 |
ret = pagecache_write_end(file, mapping, sd->pos, this_len, this_len, page, fsdata); |
5274f052e
|
571 |
out: |
5274f052e
|
572 573 |
return ret; } |
932cc6d4f
|
574 575 576 577 578 579 580 581 582 583 584 585 |
/** * __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. * |
83f9135bd
|
586 |
*/ |
c66ab6fa7
|
587 588 |
ssize_t __splice_from_pipe(struct pipe_inode_info *pipe, struct splice_desc *sd, splice_actor *actor) |
5274f052e
|
589 |
{ |
5274f052e
|
590 |
int ret, do_wakeup, err; |
5274f052e
|
591 592 593 |
ret = 0; do_wakeup = 0; |
5274f052e
|
594 |
for (;;) { |
6f767b042
|
595 596 |
if (pipe->nrbufs) { struct pipe_buffer *buf = pipe->bufs + pipe->curbuf; |
d4c3cca94
|
597 |
const struct pipe_buf_operations *ops = buf->ops; |
5274f052e
|
598 |
|
c66ab6fa7
|
599 600 601 |
sd->len = buf->len; if (sd->len > sd->total_len) sd->len = sd->total_len; |
5274f052e
|
602 |
|
c66ab6fa7
|
603 |
err = actor(pipe, buf, sd); |
016b661e2
|
604 |
if (err <= 0) { |
5274f052e
|
605 606 607 608 609 |
if (!ret && err != -ENODATA) ret = err; break; } |
016b661e2
|
610 611 612 |
ret += err; buf->offset += err; buf->len -= err; |
c66ab6fa7
|
613 614 615 616 |
sd->len -= err; sd->pos += err; sd->total_len -= err; if (sd->len) |
016b661e2
|
617 |
continue; |
73d62d83e
|
618 |
|
5274f052e
|
619 620 |
if (!buf->len) { buf->ops = NULL; |
3a326a2ce
|
621 |
ops->release(pipe, buf); |
6f767b042
|
622 623 624 625 |
pipe->curbuf = (pipe->curbuf + 1) & (PIPE_BUFFERS - 1); pipe->nrbufs--; if (pipe->inode) do_wakeup = 1; |
5274f052e
|
626 |
} |
c66ab6fa7
|
627 |
if (!sd->total_len) |
5274f052e
|
628 629 |
break; } |
6f767b042
|
630 |
if (pipe->nrbufs) |
5274f052e
|
631 |
continue; |
3a326a2ce
|
632 |
if (!pipe->writers) |
5274f052e
|
633 |
break; |
3a326a2ce
|
634 |
if (!pipe->waiting_writers) { |
5274f052e
|
635 636 637 |
if (ret) break; } |
c66ab6fa7
|
638 |
if (sd->flags & SPLICE_F_NONBLOCK) { |
29e350944
|
639 640 641 642 |
if (!ret) ret = -EAGAIN; break; } |
5274f052e
|
643 644 645 646 647 648 649 |
if (signal_pending(current)) { if (!ret) ret = -ERESTARTSYS; break; } if (do_wakeup) { |
c0bd1f650
|
650 |
smp_mb(); |
3a326a2ce
|
651 652 653 |
if (waitqueue_active(&pipe->wait)) wake_up_interruptible_sync(&pipe->wait); kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); |
5274f052e
|
654 655 |
do_wakeup = 0; } |
3a326a2ce
|
656 |
pipe_wait(pipe); |
5274f052e
|
657 |
} |
5274f052e
|
658 |
if (do_wakeup) { |
c0bd1f650
|
659 |
smp_mb(); |
3a326a2ce
|
660 661 662 |
if (waitqueue_active(&pipe->wait)) wake_up_interruptible(&pipe->wait); kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); |
5274f052e
|
663 |
} |
5274f052e
|
664 |
return ret; |
5274f052e
|
665 |
} |
40bee44ea
|
666 |
EXPORT_SYMBOL(__splice_from_pipe); |
5274f052e
|
667 |
|
932cc6d4f
|
668 669 670 671 672 673 674 675 676 677 678 679 680 681 |
/** * 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: * See __splice_from_pipe. This function locks the input and output inodes, * otherwise it's identical to __splice_from_pipe(). * */ |
6da618098
|
682 683 684 685 686 687 |
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; struct inode *inode = out->f_mapping->host; |
c66ab6fa7
|
688 689 690 691 |
struct splice_desc sd = { .total_len = len, .flags = flags, .pos = *ppos, |
6a14b90bb
|
692 |
.u.file = out, |
c66ab6fa7
|
693 |
}; |
6da618098
|
694 695 |
/* |
4e02ed4b4
|
696 697 |
* The actor worker might be calling ->write_begin and * ->write_end. Most of the time, these expect i_mutex to |
6da618098
|
698 699 700 701 |
* be held. Since this may result in an ABBA deadlock with * pipe->inode, we have to order lock acquiry here. */ inode_double_lock(inode, pipe->inode); |
c66ab6fa7
|
702 |
ret = __splice_from_pipe(pipe, &sd, actor); |
6da618098
|
703 704 705 706 707 708 709 710 711 |
inode_double_unlock(inode, pipe->inode); return ret; } /** * generic_file_splice_write_nolock - generic_file_splice_write without mutexes * @pipe: pipe info * @out: file to write to |
932cc6d4f
|
712 |
* @ppos: position in @out |
6da618098
|
713 714 715 |
* @len: number of bytes to splice * @flags: splice modifier flags * |
932cc6d4f
|
716 717 718 719 |
* Description: * Will either move or copy pages (determined by @flags options) from * the given pipe inode to the given file. The caller is responsible * for acquiring i_mutex on both inodes. |
6da618098
|
720 721 722 723 724 725 726 727 |
* */ ssize_t generic_file_splice_write_nolock(struct pipe_inode_info *pipe, struct file *out, loff_t *ppos, size_t len, unsigned int flags) { struct address_space *mapping = out->f_mapping; struct inode *inode = mapping->host; |
c66ab6fa7
|
728 729 730 731 |
struct splice_desc sd = { .total_len = len, .flags = flags, .pos = *ppos, |
6a14b90bb
|
732 |
.u.file = out, |
c66ab6fa7
|
733 |
}; |
6da618098
|
734 735 |
ssize_t ret; int err; |
2f1936b87
|
736 |
err = file_remove_suid(out); |
8c34e2d63
|
737 738 |
if (unlikely(err)) return err; |
c66ab6fa7
|
739 |
ret = __splice_from_pipe(pipe, &sd, pipe_to_file); |
6da618098
|
740 |
if (ret > 0) { |
17ee4f49a
|
741 |
unsigned long nr_pages; |
6da618098
|
742 |
*ppos += ret; |
17ee4f49a
|
743 |
nr_pages = (ret + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; |
6da618098
|
744 745 746 747 748 749 750 751 752 753 754 755 |
/* * If file or inode is SYNC and we actually wrote some data, * sync it. */ if (unlikely((out->f_flags & O_SYNC) || IS_SYNC(inode))) { err = generic_osync_inode(inode, mapping, OSYNC_METADATA|OSYNC_DATA); if (err) ret = err; } |
17ee4f49a
|
756 |
balance_dirty_pages_ratelimited_nr(mapping, nr_pages); |
6da618098
|
757 758 759 760 761 762 |
} return ret; } EXPORT_SYMBOL(generic_file_splice_write_nolock); |
83f9135bd
|
763 764 |
/** * generic_file_splice_write - splice data from a pipe to a file |
3a326a2ce
|
765 |
* @pipe: pipe info |
83f9135bd
|
766 |
* @out: file to write to |
932cc6d4f
|
767 |
* @ppos: position in @out |
83f9135bd
|
768 769 770 |
* @len: number of bytes to splice * @flags: splice modifier flags * |
932cc6d4f
|
771 772 773 |
* Description: * Will either move or copy pages (determined by @flags options) from * the given pipe inode to the given file. |
83f9135bd
|
774 775 |
* */ |
3a326a2ce
|
776 777 |
ssize_t generic_file_splice_write(struct pipe_inode_info *pipe, struct file *out, |
cbb7e577e
|
778 |
loff_t *ppos, size_t len, unsigned int flags) |
5274f052e
|
779 |
{ |
4f6f0bd2f
|
780 |
struct address_space *mapping = out->f_mapping; |
8c34e2d63
|
781 |
struct inode *inode = mapping->host; |
7f3d4ee10
|
782 783 784 785 786 787 |
struct splice_desc sd = { .total_len = len, .flags = flags, .pos = *ppos, .u.file = out, }; |
3a326a2ce
|
788 |
ssize_t ret; |
7f3d4ee10
|
789 |
inode_double_lock(inode, pipe->inode); |
2f1936b87
|
790 |
ret = file_remove_suid(out); |
7f3d4ee10
|
791 792 793 |
if (likely(!ret)) ret = __splice_from_pipe(pipe, &sd, pipe_to_file); inode_double_unlock(inode, pipe->inode); |
a4514ebd8
|
794 |
if (ret > 0) { |
17ee4f49a
|
795 |
unsigned long nr_pages; |
a4514ebd8
|
796 |
*ppos += ret; |
17ee4f49a
|
797 |
nr_pages = (ret + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; |
a4514ebd8
|
798 799 800 801 802 803 |
/* * If file or inode is SYNC and we actually wrote some data, * sync it. */ if (unlikely((out->f_flags & O_SYNC) || IS_SYNC(inode))) { |
7f3d4ee10
|
804 |
int err; |
a4514ebd8
|
805 806 807 808 |
mutex_lock(&inode->i_mutex); err = generic_osync_inode(inode, mapping, OSYNC_METADATA|OSYNC_DATA); mutex_unlock(&inode->i_mutex); |
4f6f0bd2f
|
809 |
|
a4514ebd8
|
810 811 812 |
if (err) ret = err; } |
17ee4f49a
|
813 |
balance_dirty_pages_ratelimited_nr(mapping, nr_pages); |
4f6f0bd2f
|
814 815 816 |
} return ret; |
5274f052e
|
817 |
} |
059a8f373
|
818 |
EXPORT_SYMBOL(generic_file_splice_write); |
83f9135bd
|
819 820 |
/** * generic_splice_sendpage - splice data from a pipe to a socket |
932cc6d4f
|
821 |
* @pipe: pipe to splice from |
83f9135bd
|
822 |
* @out: socket to write to |
932cc6d4f
|
823 |
* @ppos: position in @out |
83f9135bd
|
824 825 826 |
* @len: number of bytes to splice * @flags: splice modifier flags * |
932cc6d4f
|
827 828 829 |
* Description: * Will send @len bytes from the pipe to a network socket. No data copying * is involved. |
83f9135bd
|
830 831 |
* */ |
3a326a2ce
|
832 |
ssize_t generic_splice_sendpage(struct pipe_inode_info *pipe, struct file *out, |
cbb7e577e
|
833 |
loff_t *ppos, size_t len, unsigned int flags) |
5274f052e
|
834 |
{ |
00522fb41
|
835 |
return splice_from_pipe(pipe, out, ppos, len, flags, pipe_to_sendpage); |
5274f052e
|
836 |
} |
059a8f373
|
837 |
EXPORT_SYMBOL(generic_splice_sendpage); |
a0f067802
|
838 |
|
83f9135bd
|
839 840 841 |
/* * Attempt to initiate a splice from pipe to file. */ |
3a326a2ce
|
842 |
static long do_splice_from(struct pipe_inode_info *pipe, struct file *out, |
cbb7e577e
|
843 |
loff_t *ppos, size_t len, unsigned int flags) |
5274f052e
|
844 |
{ |
5274f052e
|
845 |
int ret; |
49570e9b2
|
846 |
if (unlikely(!out->f_op || !out->f_op->splice_write)) |
5274f052e
|
847 |
return -EINVAL; |
49570e9b2
|
848 |
if (unlikely(!(out->f_mode & FMODE_WRITE))) |
5274f052e
|
849 |
return -EBADF; |
efc968d45
|
850 851 |
if (unlikely(out->f_flags & O_APPEND)) return -EINVAL; |
cbb7e577e
|
852 |
ret = rw_verify_area(WRITE, out, ppos, len); |
5274f052e
|
853 854 |
if (unlikely(ret < 0)) return ret; |
cbb7e577e
|
855 |
return out->f_op->splice_write(pipe, out, ppos, len, flags); |
5274f052e
|
856 |
} |
83f9135bd
|
857 858 859 |
/* * Attempt to initiate a splice from a file to a pipe. */ |
cbb7e577e
|
860 861 862 |
static long do_splice_to(struct file *in, loff_t *ppos, struct pipe_inode_info *pipe, size_t len, unsigned int flags) |
5274f052e
|
863 |
{ |
5274f052e
|
864 |
int ret; |
49570e9b2
|
865 |
if (unlikely(!in->f_op || !in->f_op->splice_read)) |
5274f052e
|
866 |
return -EINVAL; |
49570e9b2
|
867 |
if (unlikely(!(in->f_mode & FMODE_READ))) |
5274f052e
|
868 |
return -EBADF; |
cbb7e577e
|
869 |
ret = rw_verify_area(READ, in, ppos, len); |
5274f052e
|
870 871 |
if (unlikely(ret < 0)) return ret; |
cbb7e577e
|
872 |
return in->f_op->splice_read(in, ppos, pipe, len, flags); |
5274f052e
|
873 |
} |
932cc6d4f
|
874 875 876 877 878 879 880 881 882 |
/** * 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
|
883 |
* pipe is cached in the process, and reused during the lifetime of |
932cc6d4f
|
884 885 |
* that process. * |
c66ab6fa7
|
886 887 888 |
*/ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd, splice_direct_actor *actor) |
b92ce5589
|
889 890 891 892 |
{ struct pipe_inode_info *pipe; long ret, bytes; umode_t i_mode; |
c66ab6fa7
|
893 894 |
size_t len; int i, flags; |
b92ce5589
|
895 896 897 898 899 900 |
/* * 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
|
901 |
i_mode = in->f_path.dentry->d_inode->i_mode; |
b92ce5589
|
902 903 904 905 906 907 908 909 |
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
|
910 |
if (unlikely(!pipe)) { |
b92ce5589
|
911 912 913 914 915 916 |
pipe = alloc_pipe_info(NULL); if (!pipe) return -ENOMEM; /* * We don't have an immediate reader, but we'll read the stuff |
00522fb41
|
917 |
* out of the pipe right after the splice_to_pipe(). So set |
b92ce5589
|
918 919 920 921 922 923 924 925 |
* PIPE_READERS appropriately. */ pipe->readers = 1; current->splice_pipe = pipe; } /* |
73d62d83e
|
926 |
* Do the splice. |
b92ce5589
|
927 928 929 |
*/ ret = 0; bytes = 0; |
c66ab6fa7
|
930 931 932 933 934 935 936 |
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
|
937 938 |
while (len) { |
51a92c0f6
|
939 |
size_t read_len; |
a82c53a0e
|
940 |
loff_t pos = sd->pos, prev_pos = pos; |
b92ce5589
|
941 |
|
bcd4f3acb
|
942 |
ret = do_splice_to(in, &pos, pipe, len, flags); |
51a92c0f6
|
943 |
if (unlikely(ret <= 0)) |
b92ce5589
|
944 945 946 |
goto out_release; read_len = ret; |
c66ab6fa7
|
947 |
sd->total_len = read_len; |
b92ce5589
|
948 949 950 951 952 953 |
/* * 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
|
954 |
ret = actor(pipe, sd); |
a82c53a0e
|
955 956 |
if (unlikely(ret <= 0)) { sd->pos = prev_pos; |
b92ce5589
|
957 |
goto out_release; |
a82c53a0e
|
958 |
} |
b92ce5589
|
959 960 961 |
bytes += ret; len -= ret; |
bcd4f3acb
|
962 |
sd->pos = pos; |
b92ce5589
|
963 |
|
a82c53a0e
|
964 965 |
if (ret < read_len) { sd->pos = prev_pos + ret; |
51a92c0f6
|
966 |
goto out_release; |
a82c53a0e
|
967 |
} |
b92ce5589
|
968 |
} |
9e97198db
|
969 |
done: |
b92ce5589
|
970 |
pipe->nrbufs = pipe->curbuf = 0; |
808487085
|
971 |
file_accessed(in); |
b92ce5589
|
972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 |
return bytes; out_release: /* * If we did an incomplete transfer we must release * the pipe buffers in question: */ for (i = 0; i < PIPE_BUFFERS; i++) { struct pipe_buffer *buf = pipe->bufs + i; if (buf->ops) { buf->ops->release(pipe, buf); buf->ops = NULL; } } |
b92ce5589
|
987 |
|
9e97198db
|
988 989 |
if (!bytes) bytes = ret; |
c66ab6fa7
|
990 |
|
9e97198db
|
991 |
goto done; |
c66ab6fa7
|
992 993 994 995 996 997 |
} EXPORT_SYMBOL(splice_direct_to_actor); static int direct_splice_actor(struct pipe_inode_info *pipe, struct splice_desc *sd) { |
6a14b90bb
|
998 |
struct file *file = sd->u.file; |
c66ab6fa7
|
999 1000 1001 |
return do_splice_from(pipe, file, &sd->pos, sd->total_len, sd->flags); } |
932cc6d4f
|
1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 |
/** * 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
|
1017 1018 1019 1020 1021 1022 1023 1024 |
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
|
1025 |
.u.file = out, |
c66ab6fa7
|
1026 |
}; |
51a92c0f6
|
1027 |
long ret; |
c66ab6fa7
|
1028 1029 |
ret = splice_direct_to_actor(in, &sd, direct_splice_actor); |
51a92c0f6
|
1030 |
if (ret > 0) |
a82c53a0e
|
1031 |
*ppos = sd.pos; |
51a92c0f6
|
1032 |
|
c66ab6fa7
|
1033 |
return ret; |
b92ce5589
|
1034 |
} |
83f9135bd
|
1035 |
/* |
ddac0d39c
|
1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 |
* After the inode slimming patch, i_pipe/i_bdev/i_cdev share the same * location, so checking ->i_pipe is not enough to verify that this is a * pipe. */ static inline struct pipe_inode_info *pipe_info(struct inode *inode) { if (S_ISFIFO(inode->i_mode)) return inode->i_pipe; return NULL; } /* |
83f9135bd
|
1049 1050 |
* Determine where to splice to/from. */ |
529565dcb
|
1051 1052 1053 |
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
|
1054 |
{ |
3a326a2ce
|
1055 |
struct pipe_inode_info *pipe; |
cbb7e577e
|
1056 |
loff_t offset, *off; |
a4514ebd8
|
1057 |
long ret; |
5274f052e
|
1058 |
|
0f7fc9e4d
|
1059 |
pipe = pipe_info(in->f_path.dentry->d_inode); |
529565dcb
|
1060 1061 1062 |
if (pipe) { if (off_in) return -ESPIPE; |
b92ce5589
|
1063 1064 1065 |
if (off_out) { if (out->f_op->llseek == no_llseek) return -EINVAL; |
cbb7e577e
|
1066 |
if (copy_from_user(&offset, off_out, sizeof(loff_t))) |
b92ce5589
|
1067 |
return -EFAULT; |
cbb7e577e
|
1068 1069 1070 |
off = &offset; } else off = &out->f_pos; |
529565dcb
|
1071 |
|
a4514ebd8
|
1072 1073 1074 1075 1076 1077 |
ret = do_splice_from(pipe, out, off, len, flags); if (off_out && copy_to_user(off_out, off, sizeof(loff_t))) ret = -EFAULT; return ret; |
529565dcb
|
1078 |
} |
5274f052e
|
1079 |
|
0f7fc9e4d
|
1080 |
pipe = pipe_info(out->f_path.dentry->d_inode); |
529565dcb
|
1081 1082 1083 |
if (pipe) { if (off_out) return -ESPIPE; |
b92ce5589
|
1084 1085 1086 |
if (off_in) { if (in->f_op->llseek == no_llseek) return -EINVAL; |
cbb7e577e
|
1087 |
if (copy_from_user(&offset, off_in, sizeof(loff_t))) |
b92ce5589
|
1088 |
return -EFAULT; |
cbb7e577e
|
1089 1090 1091 |
off = &offset; } else off = &in->f_pos; |
529565dcb
|
1092 |
|
a4514ebd8
|
1093 1094 1095 1096 1097 1098 |
ret = do_splice_to(in, off, pipe, len, flags); if (off_in && copy_to_user(off_in, off, sizeof(loff_t))) ret = -EFAULT; return ret; |
529565dcb
|
1099 |
} |
5274f052e
|
1100 1101 1102 |
return -EINVAL; } |
912d35f86
|
1103 1104 1105 1106 1107 1108 1109 1110 1111 |
/* * 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, |
7afa6fd03
|
1112 |
struct partial_page *partial, int aligned) |
912d35f86
|
1113 1114 |
{ int buffers = 0, error = 0; |
912d35f86
|
1115 1116 |
while (nr_vecs) { unsigned long off, npages; |
757239576
|
1117 |
struct iovec entry; |
912d35f86
|
1118 1119 1120 |
void __user *base; size_t len; int i; |
757239576
|
1121 |
error = -EFAULT; |
bc40d73c9
|
1122 |
if (copy_from_user(&entry, iov, sizeof(entry))) |
912d35f86
|
1123 |
break; |
757239576
|
1124 1125 |
base = entry.iov_base; len = entry.iov_len; |
912d35f86
|
1126 1127 1128 |
/* * Sanity check this iovec. 0 read succeeds. */ |
757239576
|
1129 |
error = 0; |
912d35f86
|
1130 1131 1132 |
if (unlikely(!len)) break; error = -EFAULT; |
712a30e63
|
1133 |
if (!access_ok(VERIFY_READ, base, len)) |
912d35f86
|
1134 1135 1136 1137 1138 1139 1140 |
break; /* * Get this base offset and number of pages, then map * in the user pages. */ off = (unsigned long) base & ~PAGE_MASK; |
7afa6fd03
|
1141 1142 1143 1144 1145 1146 1147 1148 |
/* * 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
|
1149 1150 1151 |
npages = (off + len + PAGE_SIZE - 1) >> PAGE_SHIFT; if (npages > PIPE_BUFFERS - buffers) npages = PIPE_BUFFERS - buffers; |
bc40d73c9
|
1152 1153 |
error = get_user_pages_fast((unsigned long)base, npages, 0, &pages[buffers]); |
912d35f86
|
1154 1155 1156 1157 1158 1159 1160 1161 |
if (unlikely(error <= 0)) break; /* * Fill this contiguous range into the partial page map. */ for (i = 0; i < error; i++) { |
7591489a8
|
1162 |
const int plen = min_t(size_t, len, PAGE_SIZE - off); |
912d35f86
|
1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 |
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. */ if (error < npages || buffers == PIPE_BUFFERS) break; nr_vecs--; iov++; } |
912d35f86
|
1191 1192 1193 1194 1195 |
if (buffers) return buffers; return error; } |
6a14b90bb
|
1196 1197 1198 1199 1200 |
static int pipe_to_user(struct pipe_inode_info *pipe, struct pipe_buffer *buf, struct splice_desc *sd) { char *src; int ret; |
cac36bb06
|
1201 |
ret = buf->ops->confirm(pipe, buf); |
6a14b90bb
|
1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 |
if (unlikely(ret)) return ret; /* * 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
|
1228 |
buf->ops->unmap(pipe, buf, src); |
6a14b90bb
|
1229 1230 1231 |
out: if (ret > 0) sd->u.userptr += ret; |
6a14b90bb
|
1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 |
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; pipe = pipe_info(file->f_path.dentry->d_inode); if (!pipe) return -EBADF; if (pipe->inode) mutex_lock(&pipe->inode->i_mutex); 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
|
1279 1280 1281 1282 |
if (unlikely(!access_ok(VERIFY_WRITE, base, len))) { error = -EFAULT; break; } |
6a14b90bb
|
1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 |
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++; } if (pipe->inode) mutex_unlock(&pipe->inode->i_mutex); if (!ret) ret = error; return ret; } |
912d35f86
|
1314 1315 1316 1317 |
/* * 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
|
1318 |
*/ |
6a14b90bb
|
1319 1320 |
static long vmsplice_to_pipe(struct file *file, const struct iovec __user *iov, unsigned long nr_segs, unsigned int flags) |
912d35f86
|
1321 |
{ |
ddac0d39c
|
1322 |
struct pipe_inode_info *pipe; |
912d35f86
|
1323 1324 1325 1326 1327 1328 1329 |
struct page *pages[PIPE_BUFFERS]; struct partial_page partial[PIPE_BUFFERS]; struct splice_pipe_desc spd = { .pages = pages, .partial = partial, .flags = flags, .ops = &user_page_pipe_buf_ops, |
bbdfc2f70
|
1330 |
.spd_release = spd_release_page, |
912d35f86
|
1331 |
}; |
0f7fc9e4d
|
1332 |
pipe = pipe_info(file->f_path.dentry->d_inode); |
ddac0d39c
|
1333 |
if (!pipe) |
912d35f86
|
1334 |
return -EBADF; |
912d35f86
|
1335 |
|
7afa6fd03
|
1336 1337 |
spd.nr_pages = get_iovec_page_array(iov, nr_segs, pages, partial, flags & SPLICE_F_GIFT); |
912d35f86
|
1338 1339 |
if (spd.nr_pages <= 0) return spd.nr_pages; |
00522fb41
|
1340 |
return splice_to_pipe(pipe, &spd); |
912d35f86
|
1341 |
} |
6a14b90bb
|
1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 |
/* * 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
|
1358 1359 |
SYSCALL_DEFINE4(vmsplice, int, fd, const struct iovec __user *, iov, unsigned long, nr_segs, unsigned int, flags) |
912d35f86
|
1360 1361 1362 1363 |
{ struct file *file; long error; int fput; |
6a14b90bb
|
1364 1365 1366 1367 |
if (unlikely(nr_segs > UIO_MAXIOV)) return -EINVAL; else if (unlikely(!nr_segs)) return 0; |
912d35f86
|
1368 1369 1370 1371 |
error = -EBADF; file = fget_light(fd, &fput); if (file) { if (file->f_mode & FMODE_WRITE) |
6a14b90bb
|
1372 1373 1374 |
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
|
1375 1376 1377 1378 1379 1380 |
fput_light(file, fput); } return error; } |
836f92adf
|
1381 1382 1383 |
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
|
1384 1385 1386 1387 1388 1389 1390 1391 1392 |
{ long error; struct file *in, *out; int fput_in, fput_out; if (unlikely(!len)) return 0; error = -EBADF; |
529565dcb
|
1393 |
in = fget_light(fd_in, &fput_in); |
5274f052e
|
1394 1395 |
if (in) { if (in->f_mode & FMODE_READ) { |
529565dcb
|
1396 |
out = fget_light(fd_out, &fput_out); |
5274f052e
|
1397 1398 |
if (out) { if (out->f_mode & FMODE_WRITE) |
529565dcb
|
1399 1400 1401 |
error = do_splice(in, off_in, out, off_out, len, flags); |
5274f052e
|
1402 1403 1404 1405 1406 1407 1408 1409 1410 |
fput_light(out, fput_out); } } fput_light(in, fput_in); } return error; } |
70524490e
|
1411 1412 |
/* |
aadd06e5c
|
1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 |
* Make sure there's data to read. Wait for input if we can, otherwise * return an appropriate error. */ static int link_ipipe_prep(struct pipe_inode_info *pipe, unsigned int flags) { 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; mutex_lock(&pipe->inode->i_mutex); 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); } mutex_unlock(&pipe->inode->i_mutex); return ret; } /* * Make sure there's writeable room. Wait for room if we can, otherwise * return an appropriate error. */ static int link_opipe_prep(struct pipe_inode_info *pipe, unsigned int flags) { int ret; /* * Check ->nrbufs without the inode lock first. This function * is speculative anyways, so missing one is ok. */ if (pipe->nrbufs < PIPE_BUFFERS) return 0; ret = 0; mutex_lock(&pipe->inode->i_mutex); while (pipe->nrbufs >= PIPE_BUFFERS) { 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--; } mutex_unlock(&pipe->inode->i_mutex); return ret; } /* |
70524490e
|
1492 1493 1494 1495 1496 1497 1498 |
* 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
|
1499 |
int ret = 0, i = 0, nbuf; |
70524490e
|
1500 1501 1502 1503 1504 1505 |
/* * Potential ABBA deadlock, work around it by ordering lock * grabbing by inode address. Otherwise two different processes * could deadlock (one doing tee from A -> B, the other from B -> A). */ |
62752ee19
|
1506 |
inode_double_lock(ipipe->inode, opipe->inode); |
70524490e
|
1507 |
|
aadd06e5c
|
1508 |
do { |
70524490e
|
1509 1510 1511 1512 1513 1514 |
if (!opipe->readers) { send_sig(SIGPIPE, current, 0); if (!ret) ret = -EPIPE; break; } |
70524490e
|
1515 |
|
aadd06e5c
|
1516 1517 1518 1519 1520 1521 |
/* * If we have iterated all input buffers or ran out of * output room, break. */ if (i >= ipipe->nrbufs || opipe->nrbufs >= PIPE_BUFFERS) break; |
70524490e
|
1522 |
|
aadd06e5c
|
1523 1524 |
ibuf = ipipe->bufs + ((ipipe->curbuf + i) & (PIPE_BUFFERS - 1)); nbuf = (opipe->curbuf + opipe->nrbufs) & (PIPE_BUFFERS - 1); |
70524490e
|
1525 1526 |
/* |
aadd06e5c
|
1527 1528 |
* Get a reference to this pipe buffer, * so we can copy the contents over. |
70524490e
|
1529 |
*/ |
aadd06e5c
|
1530 1531 1532 1533 |
ibuf->ops->get(ipipe, ibuf); obuf = opipe->bufs + nbuf; *obuf = *ibuf; |
2a27250e6
|
1534 |
/* |
aadd06e5c
|
1535 1536 |
* Don't inherit the gift flag, we need to * prevent multiple steals of this page. |
2a27250e6
|
1537 |
*/ |
aadd06e5c
|
1538 |
obuf->flags &= ~PIPE_BUF_FLAG_GIFT; |
70524490e
|
1539 |
|
aadd06e5c
|
1540 1541 |
if (obuf->len > len) obuf->len = len; |
70524490e
|
1542 |
|
aadd06e5c
|
1543 1544 1545 1546 1547 |
opipe->nrbufs++; ret += obuf->len; len -= obuf->len; i++; } while (len); |
70524490e
|
1548 |
|
02cf01aea
|
1549 1550 1551 1552 1553 1554 |
/* * 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; |
62752ee19
|
1555 |
inode_double_unlock(ipipe->inode, opipe->inode); |
70524490e
|
1556 |
|
aadd06e5c
|
1557 1558 1559 1560 |
/* * If we put data in the output pipe, wakeup any potential readers. */ if (ret > 0) { |
70524490e
|
1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 |
smp_mb(); if (waitqueue_active(&opipe->wait)) wake_up_interruptible(&opipe->wait); kill_fasync(&opipe->fasync_readers, SIGIO, POLL_IN); } 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) { |
0f7fc9e4d
|
1579 1580 |
struct pipe_inode_info *ipipe = pipe_info(in->f_path.dentry->d_inode); struct pipe_inode_info *opipe = pipe_info(out->f_path.dentry->d_inode); |
aadd06e5c
|
1581 |
int ret = -EINVAL; |
70524490e
|
1582 1583 |
/* |
aadd06e5c
|
1584 1585 |
* Duplicate the contents of ipipe to opipe without actually * copying the data. |
70524490e
|
1586 |
*/ |
aadd06e5c
|
1587 1588 1589 1590 1591 1592 1593 1594 |
if (ipipe && opipe && ipipe != opipe) { /* * Keep going, unless we encounter an error. The ipipe/opipe * ordering doesn't really matter. */ ret = link_ipipe_prep(ipipe, flags); if (!ret) { ret = link_opipe_prep(opipe, flags); |
02cf01aea
|
1595 |
if (!ret) |
aadd06e5c
|
1596 |
ret = link_pipe(ipipe, opipe, len, flags); |
aadd06e5c
|
1597 1598 |
} } |
70524490e
|
1599 |
|
aadd06e5c
|
1600 |
return ret; |
70524490e
|
1601 |
} |
836f92adf
|
1602 |
SYSCALL_DEFINE4(tee, int, fdin, int, fdout, size_t, len, unsigned int, flags) |
70524490e
|
1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 |
{ 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; } |