Blame view
mm/filemap.c
75.3 KB
1da177e4c Linux-2.6.12-rc2 |
1 2 3 4 5 6 7 8 9 10 11 |
/* * linux/mm/filemap.c * * Copyright (C) 1994-1999 Linus Torvalds */ /* * This file handles the generic file mmap semantics used by * most "normal" filesystems (but you don't /have/ to use this: * the NFS filesystem used to do this differently, for example) */ |
b95f1b31b mm: Map most file... |
12 |
#include <linux/export.h> |
1da177e4c Linux-2.6.12-rc2 |
13 |
#include <linux/compiler.h> |
f9fe48bec dax: support dirt... |
14 |
#include <linux/dax.h> |
1da177e4c Linux-2.6.12-rc2 |
15 |
#include <linux/fs.h> |
c22ce143d [PATCH] x86: cach... |
16 |
#include <linux/uaccess.h> |
c59ede7b7 [PATCH] move capa... |
17 |
#include <linux/capability.h> |
1da177e4c Linux-2.6.12-rc2 |
18 |
#include <linux/kernel_stat.h> |
5a0e3ad6a include cleanup: ... |
19 |
#include <linux/gfp.h> |
1da177e4c Linux-2.6.12-rc2 |
20 21 22 23 24 25 26 27 |
#include <linux/mm.h> #include <linux/swap.h> #include <linux/mman.h> #include <linux/pagemap.h> #include <linux/file.h> #include <linux/uio.h> #include <linux/hash.h> #include <linux/writeback.h> |
53253383f Include <linux/ba... |
28 |
#include <linux/backing-dev.h> |
1da177e4c Linux-2.6.12-rc2 |
29 30 31 |
#include <linux/pagevec.h> #include <linux/blkdev.h> #include <linux/security.h> |
44110fe38 [PATCH] cpuset me... |
32 |
#include <linux/cpuset.h> |
2f718ffc1 mm: buffered writ... |
33 |
#include <linux/hardirq.h> /* for BUG_ON(!in_atomic()) only */ |
00501b531 mm: memcontrol: r... |
34 |
#include <linux/hugetlb.h> |
8a9f3ccd2 Memory controller... |
35 |
#include <linux/memcontrol.h> |
c515e1fd3 mm/fs: add hooks ... |
36 |
#include <linux/cleancache.h> |
f1820361f mm: implement ->m... |
37 |
#include <linux/rmap.h> |
0f8053a50 [PATCH] mm: make ... |
38 |
#include "internal.h" |
fe0bfaaff mm: trace filemap... |
39 40 |
#define CREATE_TRACE_POINTS #include <trace/events/filemap.h> |
1da177e4c Linux-2.6.12-rc2 |
41 |
/* |
1da177e4c Linux-2.6.12-rc2 |
42 43 |
* FIXME: remove all knowledge of the buffer layer from the core VM */ |
148f948ba vfs: Introduce ne... |
44 |
#include <linux/buffer_head.h> /* for try_to_free_buffers */ |
1da177e4c Linux-2.6.12-rc2 |
45 |
|
1da177e4c Linux-2.6.12-rc2 |
46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 |
#include <asm/mman.h> /* * Shared mappings implemented 30.11.1994. It's not fully working yet, * though. * * Shared mappings now work. 15.8.1995 Bruno. * * finished 'unifying' the page and buffer cache and SMP-threaded the * page-cache, 21.05.1999, Ingo Molnar <mingo@redhat.com> * * SMP-threaded pagemap-LRU 1999, Andrea Arcangeli <andrea@suse.de> */ /* * Lock ordering: * |
c8c06efa8 mm: convert i_mma... |
63 |
* ->i_mmap_rwsem (truncate_pagecache) |
1da177e4c Linux-2.6.12-rc2 |
64 |
* ->private_lock (__free_pte->__set_page_dirty_buffers) |
5d337b919 [PATCH] swap: swa... |
65 66 |
* ->swap_lock (exclusive_swap_page, others) * ->mapping->tree_lock |
1da177e4c Linux-2.6.12-rc2 |
67 |
* |
1b1dcc1b5 [PATCH] mutex sub... |
68 |
* ->i_mutex |
c8c06efa8 mm: convert i_mma... |
69 |
* ->i_mmap_rwsem (truncate->unmap_mapping_range) |
1da177e4c Linux-2.6.12-rc2 |
70 71 |
* * ->mmap_sem |
c8c06efa8 mm: convert i_mma... |
72 |
* ->i_mmap_rwsem |
b8072f099 [PATCH] mm: updat... |
73 |
* ->page_table_lock or pte_lock (various, mainly in memory.c) |
1da177e4c Linux-2.6.12-rc2 |
74 75 76 77 78 |
* ->mapping->tree_lock (arch-dependent flush_dcache_mmap_lock) * * ->mmap_sem * ->lock_page (access_process_vm) * |
ccad23656 kill generic_file... |
79 |
* ->i_mutex (generic_perform_write) |
82591e6ea [PATCH] mm: more ... |
80 |
* ->mmap_sem (fault_in_pages_readable->do_page_fault) |
1da177e4c Linux-2.6.12-rc2 |
81 |
* |
f758eeabe writeback: split ... |
82 |
* bdi->wb.list_lock |
a66979aba fs: move i_wb_lis... |
83 |
* sb_lock (fs/fs-writeback.c) |
1da177e4c Linux-2.6.12-rc2 |
84 85 |
* ->mapping->tree_lock (__sync_single_inode) * |
c8c06efa8 mm: convert i_mma... |
86 |
* ->i_mmap_rwsem |
1da177e4c Linux-2.6.12-rc2 |
87 88 89 |
* ->anon_vma.lock (vma_adjust) * * ->anon_vma.lock |
b8072f099 [PATCH] mm: updat... |
90 |
* ->page_table_lock or pte_lock (anon_vma_prepare and various) |
1da177e4c Linux-2.6.12-rc2 |
91 |
* |
b8072f099 [PATCH] mm: updat... |
92 |
* ->page_table_lock or pte_lock |
5d337b919 [PATCH] swap: swa... |
93 |
* ->swap_lock (try_to_unmap_one) |
1da177e4c Linux-2.6.12-rc2 |
94 95 96 |
* ->private_lock (try_to_unmap_one) * ->tree_lock (try_to_unmap_one) * ->zone.lru_lock (follow_page->mark_page_accessed) |
053837fce [PATCH] mm: migra... |
97 |
* ->zone.lru_lock (check_pte_range->isolate_lru_page) |
1da177e4c Linux-2.6.12-rc2 |
98 99 |
* ->private_lock (page_remove_rmap->set_page_dirty) * ->tree_lock (page_remove_rmap->set_page_dirty) |
f758eeabe writeback: split ... |
100 |
* bdi.wb->list_lock (page_remove_rmap->set_page_dirty) |
250df6ed2 fs: protect inode... |
101 |
* ->inode->i_lock (page_remove_rmap->set_page_dirty) |
81f8c3a46 mm: memcontrol: g... |
102 |
* ->memcg->move_lock (page_remove_rmap->lock_page_memcg) |
f758eeabe writeback: split ... |
103 |
* bdi.wb->list_lock (zap_pte_range->set_page_dirty) |
250df6ed2 fs: protect inode... |
104 |
* ->inode->i_lock (zap_pte_range->set_page_dirty) |
1da177e4c Linux-2.6.12-rc2 |
105 106 |
* ->private_lock (zap_pte_range->__set_page_dirty_buffers) * |
c8c06efa8 mm: convert i_mma... |
107 |
* ->i_mmap_rwsem |
9a3c531df mm: update stale ... |
108 |
* ->tasklist_lock (memory_failure, collect_procs_ao) |
1da177e4c Linux-2.6.12-rc2 |
109 |
*/ |
91b0abe36 mm + fs: store sh... |
110 111 112 |
static void page_cache_tree_delete(struct address_space *mapping, struct page *page, void *shadow) { |
449dd6984 mm: keep page cac... |
113 |
struct radix_tree_node *node; |
91b0abe36 mm + fs: store sh... |
114 |
|
449dd6984 mm: keep page cac... |
115 |
VM_BUG_ON(!PageLocked(page)); |
d604c3245 radix-tree: intro... |
116 117 |
node = radix_tree_replace_clear_tags(&mapping->page_tree, page->index, shadow); |
449dd6984 mm: keep page cac... |
118 119 |
if (shadow) { |
f9fe48bec dax: support dirt... |
120 |
mapping->nrexceptional++; |
91b0abe36 mm + fs: store sh... |
121 |
/* |
f9fe48bec dax: support dirt... |
122 |
* Make sure the nrexceptional update is committed before |
91b0abe36 mm + fs: store sh... |
123 124 125 126 127 |
* the nrpages update so that final truncate racing * with reclaim does not see both counters 0 at the * same time and miss a shadow entry. */ smp_wmb(); |
449dd6984 mm: keep page cac... |
128 |
} |
91b0abe36 mm + fs: store sh... |
129 |
mapping->nrpages--; |
449dd6984 mm: keep page cac... |
130 |
|
d604c3245 radix-tree: intro... |
131 |
if (!node) |
449dd6984 mm: keep page cac... |
132 |
return; |
449dd6984 mm: keep page cac... |
133 |
|
449dd6984 mm: keep page cac... |
134 135 136 137 138 139 140 141 |
workingset_node_pages_dec(node); if (shadow) workingset_node_shadows_inc(node); else if (__radix_tree_delete_node(&mapping->page_tree, node)) return; /* |
ac401cc78 dax: New fault lo... |
142 143 144 |
* Track node that only contains shadow entries. DAX mappings contain * no shadow entries and may contain other exceptional entries so skip * those. |
449dd6984 mm: keep page cac... |
145 146 147 148 149 |
* * Avoid acquiring the list_lru lock if already tracked. The * list_empty() test is safe as node->private_list is * protected by mapping->tree_lock. */ |
ac401cc78 dax: New fault lo... |
150 |
if (!dax_mapping(mapping) && !workingset_node_pages(node) && |
449dd6984 mm: keep page cac... |
151 152 153 154 |
list_empty(&node->private_list)) { node->private_data = mapping; list_lru_add(&workingset_shadow_nodes, &node->private_list); } |
91b0abe36 mm + fs: store sh... |
155 |
} |
1da177e4c Linux-2.6.12-rc2 |
156 |
/* |
e64a782fe mm: change __remo... |
157 |
* Delete a page from the page cache and free it. Caller has to make |
1da177e4c Linux-2.6.12-rc2 |
158 |
* sure the page is locked and that nobody else uses it - or that usage |
fdf1cdb91 mm: remove unnece... |
159 |
* is safe. The caller must hold the mapping's tree_lock. |
1da177e4c Linux-2.6.12-rc2 |
160 |
*/ |
62cccb8c8 mm: simplify lock... |
161 |
void __delete_from_page_cache(struct page *page, void *shadow) |
1da177e4c Linux-2.6.12-rc2 |
162 163 |
{ struct address_space *mapping = page->mapping; |
fe0bfaaff mm: trace filemap... |
164 |
trace_mm_filemap_delete_from_page_cache(page); |
c515e1fd3 mm/fs: add hooks ... |
165 166 167 168 169 170 171 172 |
/* * if we're uptodate, flush out into the cleancache, otherwise * invalidate any existing cleancache entries. We can't leave * stale data around in the cleancache once our page is gone */ if (PageUptodate(page) && PageMappedToDisk(page)) cleancache_put_page(page); else |
3167760f8 mm: cleancache: s... |
173 |
cleancache_invalidate_page(mapping, page); |
c515e1fd3 mm/fs: add hooks ... |
174 |
|
06b241f32 mm: __delete_from... |
175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 |
VM_BUG_ON_PAGE(page_mapped(page), page); if (!IS_ENABLED(CONFIG_DEBUG_VM) && unlikely(page_mapped(page))) { int mapcount; pr_alert("BUG: Bad page cache in process %s pfn:%05lx ", current->comm, page_to_pfn(page)); dump_page(page, "still mapped when deleted"); dump_stack(); add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); mapcount = page_mapcount(page); if (mapping_exiting(mapping) && page_count(page) >= mapcount + 2) { /* * All vmas have already been torn down, so it's * a good bet that actually the page is unmapped, * and we'd prefer not to leak it: if we're wrong, * some other bad page check should catch it later. */ page_mapcount_reset(page); |
6d061f9f6 mm/page_ref: use ... |
196 |
page_ref_sub(page, mapcount); |
06b241f32 mm: __delete_from... |
197 198 |
} } |
91b0abe36 mm + fs: store sh... |
199 |
page_cache_tree_delete(mapping, page, shadow); |
1da177e4c Linux-2.6.12-rc2 |
200 |
page->mapping = NULL; |
b85e0effd mm: consistent tr... |
201 |
/* Leave page->index set: truncation lookup relies upon it */ |
91b0abe36 mm + fs: store sh... |
202 |
|
4165b9b46 hugetlb: do not a... |
203 204 205 |
/* hugetlb pages do not participate in page cache accounting. */ if (!PageHuge(page)) __dec_zone_page_state(page, NR_FILE_PAGES); |
4b02108ac mm: oom analysis:... |
206 207 |
if (PageSwapBacked(page)) __dec_zone_page_state(page, NR_SHMEM); |
3a6927906 Do dirty page acc... |
208 209 |
/* |
b9ea25152 page_writeback: c... |
210 211 |
* At this point page must be either written or cleaned by truncate. * Dirty page here signals a bug and loss of unwritten data. |
3a6927906 Do dirty page acc... |
212 |
* |
b9ea25152 page_writeback: c... |
213 214 215 |
* This fixes dirty accounting after removing the page entirely but * leaves PageDirty set: it has no effect for truncated page and * anyway will be cleared before returning page into buddy allocator. |
3a6927906 Do dirty page acc... |
216 |
*/ |
b9ea25152 page_writeback: c... |
217 |
if (WARN_ON_ONCE(PageDirty(page))) |
62cccb8c8 mm: simplify lock... |
218 |
account_page_cleaned(page, mapping, inode_to_wb(mapping->host)); |
1da177e4c Linux-2.6.12-rc2 |
219 |
} |
702cfbf93 mm: goodbye remov... |
220 221 222 223 224 225 226 227 228 |
/** * delete_from_page_cache - delete page from page cache * @page: the page which the kernel is trying to remove from page cache * * This must be called only on pages that have been verified to be in the page * cache and locked. It will never put the page into the free list, the caller * has a reference on the page. */ void delete_from_page_cache(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
229 230 |
{ struct address_space *mapping = page->mapping; |
c4843a759 memcg: add per cg... |
231 |
unsigned long flags; |
6072d13c4 Call the filesyst... |
232 |
void (*freepage)(struct page *); |
1da177e4c Linux-2.6.12-rc2 |
233 |
|
cd7619d6b [PATCH] Extermina... |
234 |
BUG_ON(!PageLocked(page)); |
1da177e4c Linux-2.6.12-rc2 |
235 |
|
6072d13c4 Call the filesyst... |
236 |
freepage = mapping->a_ops->freepage; |
c4843a759 memcg: add per cg... |
237 |
|
c4843a759 memcg: add per cg... |
238 |
spin_lock_irqsave(&mapping->tree_lock, flags); |
62cccb8c8 mm: simplify lock... |
239 |
__delete_from_page_cache(page, NULL); |
c4843a759 memcg: add per cg... |
240 |
spin_unlock_irqrestore(&mapping->tree_lock, flags); |
6072d13c4 Call the filesyst... |
241 242 243 |
if (freepage) freepage(page); |
09cbfeaf1 mm, fs: get rid o... |
244 |
put_page(page); |
97cecb5a2 mm: introduce del... |
245 246 |
} EXPORT_SYMBOL(delete_from_page_cache); |
865ffef37 fs: fix fsync() e... |
247 248 249 250 |
static int filemap_check_errors(struct address_space *mapping) { int ret = 0; /* Check for outstanding write errors */ |
7fcbbaf18 mm/filemap.c: avo... |
251 252 |
if (test_bit(AS_ENOSPC, &mapping->flags) && test_and_clear_bit(AS_ENOSPC, &mapping->flags)) |
865ffef37 fs: fix fsync() e... |
253 |
ret = -ENOSPC; |
7fcbbaf18 mm/filemap.c: avo... |
254 255 |
if (test_bit(AS_EIO, &mapping->flags) && test_and_clear_bit(AS_EIO, &mapping->flags)) |
865ffef37 fs: fix fsync() e... |
256 257 258 |
ret = -EIO; return ret; } |
1da177e4c Linux-2.6.12-rc2 |
259 |
/** |
485bb99b4 [PATCH] kernel-do... |
260 |
* __filemap_fdatawrite_range - start writeback on mapping dirty pages in range |
67be2dd1b [PATCH] DocBook: ... |
261 262 |
* @mapping: address space structure to write * @start: offset in bytes where the range starts |
469eb4d03 [PATCH] filemap_f... |
263 |
* @end: offset in bytes where the range ends (inclusive) |
67be2dd1b [PATCH] DocBook: ... |
264 |
* @sync_mode: enable synchronous operation |
1da177e4c Linux-2.6.12-rc2 |
265 |
* |
485bb99b4 [PATCH] kernel-do... |
266 267 268 |
* Start writeback against all of a mapping's dirty pages that lie * within the byte offsets <start, end> inclusive. * |
1da177e4c Linux-2.6.12-rc2 |
269 |
* If sync_mode is WB_SYNC_ALL then this is a "data integrity" operation, as |
485bb99b4 [PATCH] kernel-do... |
270 |
* opposed to a regular memory cleansing writeback. The difference between |
1da177e4c Linux-2.6.12-rc2 |
271 272 273 |
* these two operations is that if a dirty page/buffer is encountered, it must * be waited upon, and not just skipped over. */ |
ebcf28e1c [PATCH] fadvise()... |
274 275 |
int __filemap_fdatawrite_range(struct address_space *mapping, loff_t start, loff_t end, int sync_mode) |
1da177e4c Linux-2.6.12-rc2 |
276 277 278 279 |
{ int ret; struct writeback_control wbc = { .sync_mode = sync_mode, |
05fe478dd mm: write_cache_p... |
280 |
.nr_to_write = LONG_MAX, |
111ebb6e6 [PATCH] writeback... |
281 282 |
.range_start = start, .range_end = end, |
1da177e4c Linux-2.6.12-rc2 |
283 284 285 286 |
}; if (!mapping_cap_writeback_dirty(mapping)) return 0; |
b16b1deb5 writeback: make w... |
287 |
wbc_attach_fdatawrite_inode(&wbc, mapping->host); |
1da177e4c Linux-2.6.12-rc2 |
288 |
ret = do_writepages(mapping, &wbc); |
b16b1deb5 writeback: make w... |
289 |
wbc_detach_inode(&wbc); |
1da177e4c Linux-2.6.12-rc2 |
290 291 292 293 294 295 |
return ret; } static inline int __filemap_fdatawrite(struct address_space *mapping, int sync_mode) { |
111ebb6e6 [PATCH] writeback... |
296 |
return __filemap_fdatawrite_range(mapping, 0, LLONG_MAX, sync_mode); |
1da177e4c Linux-2.6.12-rc2 |
297 298 299 300 301 302 303 |
} int filemap_fdatawrite(struct address_space *mapping) { return __filemap_fdatawrite(mapping, WB_SYNC_ALL); } EXPORT_SYMBOL(filemap_fdatawrite); |
f4c0a0fdf vfs: export filem... |
304 |
int filemap_fdatawrite_range(struct address_space *mapping, loff_t start, |
ebcf28e1c [PATCH] fadvise()... |
305 |
loff_t end) |
1da177e4c Linux-2.6.12-rc2 |
306 307 308 |
{ return __filemap_fdatawrite_range(mapping, start, end, WB_SYNC_ALL); } |
f4c0a0fdf vfs: export filem... |
309 |
EXPORT_SYMBOL(filemap_fdatawrite_range); |
1da177e4c Linux-2.6.12-rc2 |
310 |
|
485bb99b4 [PATCH] kernel-do... |
311 312 313 314 |
/** * filemap_flush - mostly a non-blocking flush * @mapping: target address_space * |
1da177e4c Linux-2.6.12-rc2 |
315 316 317 318 319 320 321 322 |
* This is a mostly non-blocking flush. Not suitable for data-integrity * purposes - I/O may not be started against all dirty pages. */ int filemap_flush(struct address_space *mapping) { return __filemap_fdatawrite(mapping, WB_SYNC_NONE); } EXPORT_SYMBOL(filemap_flush); |
aa750fd71 mm/filemap.c: mak... |
323 324 |
static int __filemap_fdatawait_range(struct address_space *mapping, loff_t start_byte, loff_t end_byte) |
1da177e4c Linux-2.6.12-rc2 |
325 |
{ |
09cbfeaf1 mm, fs: get rid o... |
326 327 |
pgoff_t index = start_byte >> PAGE_SHIFT; pgoff_t end = end_byte >> PAGE_SHIFT; |
1da177e4c Linux-2.6.12-rc2 |
328 329 |
struct pagevec pvec; int nr_pages; |
aa750fd71 mm/filemap.c: mak... |
330 |
int ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
331 |
|
94004ed72 kill wait_on_page... |
332 |
if (end_byte < start_byte) |
865ffef37 fs: fix fsync() e... |
333 |
goto out; |
1da177e4c Linux-2.6.12-rc2 |
334 335 |
pagevec_init(&pvec, 0); |
1da177e4c Linux-2.6.12-rc2 |
336 337 338 339 340 341 342 343 344 345 346 347 348 349 |
while ((index <= end) && (nr_pages = pagevec_lookup_tag(&pvec, mapping, &index, PAGECACHE_TAG_WRITEBACK, min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1)) != 0) { unsigned i; for (i = 0; i < nr_pages; i++) { struct page *page = pvec.pages[i]; /* until radix tree lookup accepts end_index */ if (page->index > end) continue; wait_on_page_writeback(page); |
212260aa0 mm: clear PageErr... |
350 |
if (TestClearPageError(page)) |
1da177e4c Linux-2.6.12-rc2 |
351 352 353 354 355 |
ret = -EIO; } pagevec_release(&pvec); cond_resched(); } |
865ffef37 fs: fix fsync() e... |
356 |
out: |
aa750fd71 mm/filemap.c: mak... |
357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 |
return ret; } /** * filemap_fdatawait_range - wait for writeback to complete * @mapping: address space structure to wait for * @start_byte: offset in bytes where the range starts * @end_byte: offset in bytes where the range ends (inclusive) * * Walk the list of under-writeback pages of the given address space * in the given range and wait for all of them. Check error status of * the address space and return it. * * Since the error status of the address space is cleared by this function, * callers are responsible for checking the return value and handling and/or * reporting the error. */ int filemap_fdatawait_range(struct address_space *mapping, loff_t start_byte, loff_t end_byte) { int ret, ret2; ret = __filemap_fdatawait_range(mapping, start_byte, end_byte); |
865ffef37 fs: fix fsync() e... |
380 381 382 |
ret2 = filemap_check_errors(mapping); if (!ret) ret = ret2; |
1da177e4c Linux-2.6.12-rc2 |
383 384 385 |
return ret; } |
d3bccb6f4 vfs: Introduce fi... |
386 387 388 |
EXPORT_SYMBOL(filemap_fdatawait_range); /** |
aa750fd71 mm/filemap.c: mak... |
389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 |
* filemap_fdatawait_keep_errors - wait for writeback without clearing errors * @mapping: address space structure to wait for * * Walk the list of under-writeback pages of the given address space * and wait for all of them. Unlike filemap_fdatawait(), this function * does not clear error status of the address space. * * Use this function if callers don't handle errors themselves. Expected * call sites are system-wide / filesystem-wide data flushers: e.g. sync(2), * fsfreeze(8) */ void filemap_fdatawait_keep_errors(struct address_space *mapping) { loff_t i_size = i_size_read(mapping->host); if (i_size == 0) return; __filemap_fdatawait_range(mapping, 0, i_size - 1); } /** |
485bb99b4 [PATCH] kernel-do... |
411 |
* filemap_fdatawait - wait for all under-writeback pages to complete |
1da177e4c Linux-2.6.12-rc2 |
412 |
* @mapping: address space structure to wait for |
485bb99b4 [PATCH] kernel-do... |
413 414 |
* * Walk the list of under-writeback pages of the given address space |
aa750fd71 mm/filemap.c: mak... |
415 416 417 418 419 420 |
* and wait for all of them. Check error status of the address space * and return it. * * Since the error status of the address space is cleared by this function, * callers are responsible for checking the return value and handling and/or * reporting the error. |
1da177e4c Linux-2.6.12-rc2 |
421 422 423 424 425 426 427 |
*/ int filemap_fdatawait(struct address_space *mapping) { loff_t i_size = i_size_read(mapping->host); if (i_size == 0) return 0; |
94004ed72 kill wait_on_page... |
428 |
return filemap_fdatawait_range(mapping, 0, i_size - 1); |
1da177e4c Linux-2.6.12-rc2 |
429 430 431 432 433 |
} EXPORT_SYMBOL(filemap_fdatawait); int filemap_write_and_wait(struct address_space *mapping) { |
28fd12982 [PATCH] Fix and a... |
434 |
int err = 0; |
1da177e4c Linux-2.6.12-rc2 |
435 |
|
7f6d5b529 dax: move writeba... |
436 437 |
if ((!dax_mapping(mapping) && mapping->nrpages) || (dax_mapping(mapping) && mapping->nrexceptional)) { |
28fd12982 [PATCH] Fix and a... |
438 439 440 441 442 443 444 445 446 447 448 449 |
err = filemap_fdatawrite(mapping); /* * Even if the above returned error, the pages may be * written partially (e.g. -ENOSPC), so we wait for it. * But the -EIO is special case, it may indicate the worst * thing (e.g. bug) happened, so we avoid waiting for it. */ if (err != -EIO) { int err2 = filemap_fdatawait(mapping); if (!err) err = err2; } |
865ffef37 fs: fix fsync() e... |
450 451 |
} else { err = filemap_check_errors(mapping); |
1da177e4c Linux-2.6.12-rc2 |
452 |
} |
28fd12982 [PATCH] Fix and a... |
453 |
return err; |
1da177e4c Linux-2.6.12-rc2 |
454 |
} |
28fd12982 [PATCH] Fix and a... |
455 |
EXPORT_SYMBOL(filemap_write_and_wait); |
1da177e4c Linux-2.6.12-rc2 |
456 |
|
485bb99b4 [PATCH] kernel-do... |
457 458 459 460 461 462 |
/** * filemap_write_and_wait_range - write out & wait on a file range * @mapping: the address_space for the pages * @lstart: offset in bytes where the range starts * @lend: offset in bytes where the range ends (inclusive) * |
469eb4d03 [PATCH] filemap_f... |
463 464 465 466 467 |
* Write out and wait upon file offsets lstart->lend, inclusive. * * Note that `lend' is inclusive (describes the last byte to be written) so * that this function can be used to write to the very end-of-file (end = -1). */ |
1da177e4c Linux-2.6.12-rc2 |
468 469 470 |
int filemap_write_and_wait_range(struct address_space *mapping, loff_t lstart, loff_t lend) { |
28fd12982 [PATCH] Fix and a... |
471 |
int err = 0; |
1da177e4c Linux-2.6.12-rc2 |
472 |
|
7f6d5b529 dax: move writeba... |
473 474 |
if ((!dax_mapping(mapping) && mapping->nrpages) || (dax_mapping(mapping) && mapping->nrexceptional)) { |
28fd12982 [PATCH] Fix and a... |
475 476 477 478 |
err = __filemap_fdatawrite_range(mapping, lstart, lend, WB_SYNC_ALL); /* See comment of filemap_write_and_wait() */ if (err != -EIO) { |
94004ed72 kill wait_on_page... |
479 480 |
int err2 = filemap_fdatawait_range(mapping, lstart, lend); |
28fd12982 [PATCH] Fix and a... |
481 482 483 |
if (!err) err = err2; } |
865ffef37 fs: fix fsync() e... |
484 485 |
} else { err = filemap_check_errors(mapping); |
1da177e4c Linux-2.6.12-rc2 |
486 |
} |
28fd12982 [PATCH] Fix and a... |
487 |
return err; |
1da177e4c Linux-2.6.12-rc2 |
488 |
} |
f69955855 Export filemap_wr... |
489 |
EXPORT_SYMBOL(filemap_write_and_wait_range); |
1da177e4c Linux-2.6.12-rc2 |
490 |
|
485bb99b4 [PATCH] kernel-do... |
491 |
/** |
ef6a3c631 mm: add replace_p... |
492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 |
* replace_page_cache_page - replace a pagecache page with a new one * @old: page to be replaced * @new: page to replace with * @gfp_mask: allocation mode * * This function replaces a page in the pagecache with a new one. On * success it acquires the pagecache reference for the new page and * drops it for the old page. Both the old and new pages must be * locked. This function does not add the new page to the LRU, the * caller must do that. * * The remove + add is atomic. The only way this function can fail is * memory allocation failure. */ int replace_page_cache_page(struct page *old, struct page *new, gfp_t gfp_mask) { int error; |
ef6a3c631 mm: add replace_p... |
509 |
|
309381fea mm: dump page whe... |
510 511 512 |
VM_BUG_ON_PAGE(!PageLocked(old), old); VM_BUG_ON_PAGE(!PageLocked(new), new); VM_BUG_ON_PAGE(new->mapping, new); |
ef6a3c631 mm: add replace_p... |
513 |
|
ef6a3c631 mm: add replace_p... |
514 515 516 517 |
error = radix_tree_preload(gfp_mask & ~__GFP_HIGHMEM); if (!error) { struct address_space *mapping = old->mapping; void (*freepage)(struct page *); |
c4843a759 memcg: add per cg... |
518 |
unsigned long flags; |
ef6a3c631 mm: add replace_p... |
519 520 521 |
pgoff_t offset = old->index; freepage = mapping->a_ops->freepage; |
09cbfeaf1 mm, fs: get rid o... |
522 |
get_page(new); |
ef6a3c631 mm: add replace_p... |
523 524 |
new->mapping = mapping; new->index = offset; |
c4843a759 memcg: add per cg... |
525 |
spin_lock_irqsave(&mapping->tree_lock, flags); |
62cccb8c8 mm: simplify lock... |
526 |
__delete_from_page_cache(old, NULL); |
ef6a3c631 mm: add replace_p... |
527 528 529 |
error = radix_tree_insert(&mapping->page_tree, offset, new); BUG_ON(error); mapping->nrpages++; |
4165b9b46 hugetlb: do not a... |
530 531 532 533 534 535 |
/* * hugetlb pages do not participate in page cache accounting. */ if (!PageHuge(new)) __inc_zone_page_state(new, NR_FILE_PAGES); |
ef6a3c631 mm: add replace_p... |
536 537 |
if (PageSwapBacked(new)) __inc_zone_page_state(new, NR_SHMEM); |
c4843a759 memcg: add per cg... |
538 |
spin_unlock_irqrestore(&mapping->tree_lock, flags); |
6a93ca8fd mm: migrate: do n... |
539 |
mem_cgroup_migrate(old, new); |
ef6a3c631 mm: add replace_p... |
540 541 542 |
radix_tree_preload_end(); if (freepage) freepage(old); |
09cbfeaf1 mm, fs: get rid o... |
543 |
put_page(old); |
ef6a3c631 mm: add replace_p... |
544 545 546 547 548 |
} return error; } EXPORT_SYMBOL_GPL(replace_page_cache_page); |
0cd6144aa mm + fs: prepare ... |
549 |
static int page_cache_tree_insert(struct address_space *mapping, |
a528910e1 mm: thrash detect... |
550 |
struct page *page, void **shadowp) |
0cd6144aa mm + fs: prepare ... |
551 |
{ |
449dd6984 mm: keep page cac... |
552 |
struct radix_tree_node *node; |
0cd6144aa mm + fs: prepare ... |
553 554 |
void **slot; int error; |
e61452365 radix_tree: add s... |
555 |
error = __radix_tree_create(&mapping->page_tree, page->index, 0, |
449dd6984 mm: keep page cac... |
556 557 558 559 |
&node, &slot); if (error) return error; if (*slot) { |
0cd6144aa mm + fs: prepare ... |
560 561 562 563 564 |
void *p; p = radix_tree_deref_slot_protected(slot, &mapping->tree_lock); if (!radix_tree_exceptional_entry(p)) return -EEXIST; |
f9fe48bec dax: support dirt... |
565 |
|
f9fe48bec dax: support dirt... |
566 |
mapping->nrexceptional--; |
4f622938a dax: Allow DAX co... |
567 568 569 570 571 572 573 574 575 576 577 578 579 |
if (!dax_mapping(mapping)) { if (shadowp) *shadowp = p; if (node) workingset_node_shadows_dec(node); } else { /* DAX can replace empty locked entry with a hole */ WARN_ON_ONCE(p != (void *)(RADIX_TREE_EXCEPTIONAL_ENTRY | RADIX_DAX_ENTRY_LOCK)); /* DAX accounts exceptional entries as normal pages */ if (node) workingset_node_pages_dec(node); |
ac401cc78 dax: New fault lo... |
580 581 582 |
/* Wakeup waiters for exceptional entry lock */ dax_wake_mapping_entry_waiter(mapping, page->index, false); |
4f622938a dax: Allow DAX co... |
583 |
} |
0cd6144aa mm + fs: prepare ... |
584 |
} |
449dd6984 mm: keep page cac... |
585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 |
radix_tree_replace_slot(slot, page); mapping->nrpages++; if (node) { workingset_node_pages_inc(node); /* * Don't track node that contains actual pages. * * Avoid acquiring the list_lru lock if already * untracked. The list_empty() test is safe as * node->private_list is protected by * mapping->tree_lock. */ if (!list_empty(&node->private_list)) list_lru_del(&workingset_shadow_nodes, &node->private_list); } return 0; |
0cd6144aa mm + fs: prepare ... |
602 |
} |
a528910e1 mm: thrash detect... |
603 604 605 606 |
static int __add_to_page_cache_locked(struct page *page, struct address_space *mapping, pgoff_t offset, gfp_t gfp_mask, void **shadowp) |
1da177e4c Linux-2.6.12-rc2 |
607 |
{ |
00501b531 mm: memcontrol: r... |
608 609 |
int huge = PageHuge(page); struct mem_cgroup *memcg; |
e286781d5 mm: speculative p... |
610 |
int error; |
309381fea mm: dump page whe... |
611 612 |
VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(PageSwapBacked(page), page); |
e286781d5 mm: speculative p... |
613 |
|
00501b531 mm: memcontrol: r... |
614 615 |
if (!huge) { error = mem_cgroup_try_charge(page, current->mm, |
f627c2f53 memcg: adjust to ... |
616 |
gfp_mask, &memcg, false); |
00501b531 mm: memcontrol: r... |
617 618 619 |
if (error) return error; } |
1da177e4c Linux-2.6.12-rc2 |
620 |
|
5e4c0d974 lib/radix-tree.c:... |
621 |
error = radix_tree_maybe_preload(gfp_mask & ~__GFP_HIGHMEM); |
66a0c8ee3 mm: cleanup add_t... |
622 |
if (error) { |
00501b531 mm: memcontrol: r... |
623 |
if (!huge) |
f627c2f53 memcg: adjust to ... |
624 |
mem_cgroup_cancel_charge(page, memcg, false); |
66a0c8ee3 mm: cleanup add_t... |
625 626 |
return error; } |
09cbfeaf1 mm, fs: get rid o... |
627 |
get_page(page); |
66a0c8ee3 mm: cleanup add_t... |
628 629 630 631 |
page->mapping = mapping; page->index = offset; spin_lock_irq(&mapping->tree_lock); |
a528910e1 mm: thrash detect... |
632 |
error = page_cache_tree_insert(mapping, page, shadowp); |
66a0c8ee3 mm: cleanup add_t... |
633 634 635 |
radix_tree_preload_end(); if (unlikely(error)) goto err_insert; |
4165b9b46 hugetlb: do not a... |
636 637 638 639 |
/* hugetlb pages do not participate in page cache accounting. */ if (!huge) __inc_zone_page_state(page, NR_FILE_PAGES); |
66a0c8ee3 mm: cleanup add_t... |
640 |
spin_unlock_irq(&mapping->tree_lock); |
00501b531 mm: memcontrol: r... |
641 |
if (!huge) |
f627c2f53 memcg: adjust to ... |
642 |
mem_cgroup_commit_charge(page, memcg, false, false); |
66a0c8ee3 mm: cleanup add_t... |
643 644 645 646 647 648 |
trace_mm_filemap_add_to_page_cache(page); return 0; err_insert: page->mapping = NULL; /* Leave page->index set: truncation relies upon it */ spin_unlock_irq(&mapping->tree_lock); |
00501b531 mm: memcontrol: r... |
649 |
if (!huge) |
f627c2f53 memcg: adjust to ... |
650 |
mem_cgroup_cancel_charge(page, memcg, false); |
09cbfeaf1 mm, fs: get rid o... |
651 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
652 653 |
return error; } |
a528910e1 mm: thrash detect... |
654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 |
/** * add_to_page_cache_locked - add a locked page to the pagecache * @page: page to add * @mapping: the page's address_space * @offset: page index * @gfp_mask: page allocation mode * * This function is used to add a page to the pagecache. It must be locked. * This function does not add the page to the LRU. The caller must do that. */ int add_to_page_cache_locked(struct page *page, struct address_space *mapping, pgoff_t offset, gfp_t gfp_mask) { return __add_to_page_cache_locked(page, mapping, offset, gfp_mask, NULL); } |
e286781d5 mm: speculative p... |
671 |
EXPORT_SYMBOL(add_to_page_cache_locked); |
1da177e4c Linux-2.6.12-rc2 |
672 673 |
int add_to_page_cache_lru(struct page *page, struct address_space *mapping, |
6daa0e286 [PATCH] gfp_t: mm... |
674 |
pgoff_t offset, gfp_t gfp_mask) |
1da177e4c Linux-2.6.12-rc2 |
675 |
{ |
a528910e1 mm: thrash detect... |
676 |
void *shadow = NULL; |
4f98a2fee vmscan: split LRU... |
677 |
int ret; |
48c935ad8 page-flags: defin... |
678 |
__SetPageLocked(page); |
a528910e1 mm: thrash detect... |
679 680 681 |
ret = __add_to_page_cache_locked(page, mapping, offset, gfp_mask, &shadow); if (unlikely(ret)) |
48c935ad8 page-flags: defin... |
682 |
__ClearPageLocked(page); |
a528910e1 mm: thrash detect... |
683 684 685 686 687 |
else { /* * The page might have been evicted from cache only * recently, in which case it should be activated like * any other repeatedly accessed page. |
f0281a00f mm: workingset: o... |
688 689 690 |
* The exception is pages getting rewritten; evicting other * data from the working set, only to cache data that will * get overwritten with something else, is a waste of memory. |
a528910e1 mm: thrash detect... |
691 |
*/ |
f0281a00f mm: workingset: o... |
692 693 |
if (!(gfp_mask & __GFP_WRITE) && shadow && workingset_refault(shadow)) { |
a528910e1 mm: thrash detect... |
694 695 696 697 698 699 |
SetPageActive(page); workingset_activation(page); } else ClearPageActive(page); lru_cache_add(page); } |
1da177e4c Linux-2.6.12-rc2 |
700 701 |
return ret; } |
18bc0bbd1 Staging: pohmelfs... |
702 |
EXPORT_SYMBOL_GPL(add_to_page_cache_lru); |
1da177e4c Linux-2.6.12-rc2 |
703 |
|
44110fe38 [PATCH] cpuset me... |
704 |
#ifdef CONFIG_NUMA |
2ae88149a [PATCH] mm: clean... |
705 |
struct page *__page_cache_alloc(gfp_t gfp) |
44110fe38 [PATCH] cpuset me... |
706 |
{ |
c0ff7453b cpuset,mm: fix no... |
707 708 |
int n; struct page *page; |
44110fe38 [PATCH] cpuset me... |
709 |
if (cpuset_do_page_mem_spread()) { |
cc9a6c877 cpuset: mm: reduc... |
710 711 |
unsigned int cpuset_mems_cookie; do { |
d26914d11 mm: optimize put_... |
712 |
cpuset_mems_cookie = read_mems_allowed_begin(); |
cc9a6c877 cpuset: mm: reduc... |
713 |
n = cpuset_mem_spread_node(); |
96db800f5 mm: rename alloc_... |
714 |
page = __alloc_pages_node(n, gfp, 0); |
d26914d11 mm: optimize put_... |
715 |
} while (!page && read_mems_allowed_retry(cpuset_mems_cookie)); |
cc9a6c877 cpuset: mm: reduc... |
716 |
|
c0ff7453b cpuset,mm: fix no... |
717 |
return page; |
44110fe38 [PATCH] cpuset me... |
718 |
} |
2ae88149a [PATCH] mm: clean... |
719 |
return alloc_pages(gfp, 0); |
44110fe38 [PATCH] cpuset me... |
720 |
} |
2ae88149a [PATCH] mm: clean... |
721 |
EXPORT_SYMBOL(__page_cache_alloc); |
44110fe38 [PATCH] cpuset me... |
722 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
723 724 725 726 727 728 729 730 731 732 |
/* * In order to wait for pages to become available there must be * waitqueues associated with pages. By using a hash table of * waitqueues where the bucket discipline is to maintain all * waiters on the same queue and wake all when any of the pages * become available, and for the woken contexts to check to be * sure the appropriate page became available, this saves space * at a cost of "thundering herd" phenomena during rare hash * collisions. */ |
a4796e37c MM: export page_w... |
733 |
wait_queue_head_t *page_waitqueue(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
734 735 736 737 738 |
{ const struct zone *zone = page_zone(page); return &zone->wait_table[hash_ptr(page, zone->wait_table_bits)]; } |
a4796e37c MM: export page_w... |
739 |
EXPORT_SYMBOL(page_waitqueue); |
1da177e4c Linux-2.6.12-rc2 |
740 |
|
920c7a5d0 mm: remove fastca... |
741 |
void wait_on_page_bit(struct page *page, int bit_nr) |
1da177e4c Linux-2.6.12-rc2 |
742 743 744 745 |
{ DEFINE_WAIT_BIT(wait, &page->flags, bit_nr); if (test_bit(bit_nr, &page->flags)) |
743162013 sched: Remove pro... |
746 |
__wait_on_bit(page_waitqueue(page), &wait, bit_wait_io, |
1da177e4c Linux-2.6.12-rc2 |
747 748 749 |
TASK_UNINTERRUPTIBLE); } EXPORT_SYMBOL(wait_on_page_bit); |
f62e00cc3 mm: introduce wai... |
750 751 752 753 754 755 756 757 |
int wait_on_page_bit_killable(struct page *page, int bit_nr) { DEFINE_WAIT_BIT(wait, &page->flags, bit_nr); if (!test_bit(bit_nr, &page->flags)) return 0; return __wait_on_bit(page_waitqueue(page), &wait, |
743162013 sched: Remove pro... |
758 |
bit_wait_io, TASK_KILLABLE); |
f62e00cc3 mm: introduce wai... |
759 |
} |
cbbce8220 SCHED: add some "... |
760 761 762 763 764 765 766 767 768 769 770 771 |
int wait_on_page_bit_killable_timeout(struct page *page, int bit_nr, unsigned long timeout) { DEFINE_WAIT_BIT(wait, &page->flags, bit_nr); wait.key.timeout = jiffies + timeout; if (!test_bit(bit_nr, &page->flags)) return 0; return __wait_on_bit(page_waitqueue(page), &wait, bit_wait_io_timeout, TASK_KILLABLE); } EXPORT_SYMBOL_GPL(wait_on_page_bit_killable_timeout); |
1da177e4c Linux-2.6.12-rc2 |
772 |
/** |
385e1ca5f CacheFiles: Permi... |
773 |
* add_page_wait_queue - Add an arbitrary waiter to a page's wait queue |
697f619fc filemap: fix kern... |
774 775 |
* @page: Page defining the wait queue of interest * @waiter: Waiter to add to the queue |
385e1ca5f CacheFiles: Permi... |
776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 |
* * Add an arbitrary @waiter to the wait queue for the nominated @page. */ void add_page_wait_queue(struct page *page, wait_queue_t *waiter) { wait_queue_head_t *q = page_waitqueue(page); unsigned long flags; spin_lock_irqsave(&q->lock, flags); __add_wait_queue(q, waiter); spin_unlock_irqrestore(&q->lock, flags); } EXPORT_SYMBOL_GPL(add_page_wait_queue); /** |
485bb99b4 [PATCH] kernel-do... |
791 |
* unlock_page - unlock a locked page |
1da177e4c Linux-2.6.12-rc2 |
792 793 794 795 |
* @page: the page * * Unlocks the page and wakes up sleepers in ___wait_on_page_locked(). * Also wakes sleepers in wait_on_page_writeback() because the wakeup |
da3dae54e Documentation: Do... |
796 |
* mechanism between PageLocked pages and PageWriteback pages is shared. |
1da177e4c Linux-2.6.12-rc2 |
797 798 |
* But that's OK - sleepers in wait_on_page_writeback() just go back to sleep. * |
8413ac9d8 mm: page lock use... |
799 800 |
* The mb is necessary to enforce ordering between the clear_bit and the read * of the waitqueue (to avoid SMP races with a parallel wait_on_page_locked()). |
1da177e4c Linux-2.6.12-rc2 |
801 |
*/ |
920c7a5d0 mm: remove fastca... |
802 |
void unlock_page(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
803 |
{ |
48c935ad8 page-flags: defin... |
804 |
page = compound_head(page); |
309381fea mm: dump page whe... |
805 |
VM_BUG_ON_PAGE(!PageLocked(page), page); |
8413ac9d8 mm: page lock use... |
806 |
clear_bit_unlock(PG_locked, &page->flags); |
4e857c58e arch: Mass conver... |
807 |
smp_mb__after_atomic(); |
1da177e4c Linux-2.6.12-rc2 |
808 809 810 |
wake_up_page(page, PG_locked); } EXPORT_SYMBOL(unlock_page); |
485bb99b4 [PATCH] kernel-do... |
811 812 813 |
/** * end_page_writeback - end writeback against a page * @page: the page |
1da177e4c Linux-2.6.12-rc2 |
814 815 816 |
*/ void end_page_writeback(struct page *page) { |
888cf2db4 mm: avoid unneces... |
817 818 819 820 821 822 823 824 825 |
/* * TestClearPageReclaim could be used here but it is an atomic * operation and overkill in this particular case. Failing to * shuffle a page marked for immediate reclaim is too mild to * justify taking an atomic operation penalty at the end of * ever page writeback. */ if (PageReclaim(page)) { ClearPageReclaim(page); |
ac6aadb24 mm: rotate_reclai... |
826 |
rotate_reclaimable_page(page); |
888cf2db4 mm: avoid unneces... |
827 |
} |
ac6aadb24 mm: rotate_reclai... |
828 829 830 |
if (!test_clear_page_writeback(page)) BUG(); |
4e857c58e arch: Mass conver... |
831 |
smp_mb__after_atomic(); |
1da177e4c Linux-2.6.12-rc2 |
832 833 834 |
wake_up_page(page, PG_writeback); } EXPORT_SYMBOL(end_page_writeback); |
57d998456 fs/mpage.c: facto... |
835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 |
/* * After completing I/O on a page, call this routine to update the page * flags appropriately */ void page_endio(struct page *page, int rw, int err) { if (rw == READ) { if (!err) { SetPageUptodate(page); } else { ClearPageUptodate(page); SetPageError(page); } unlock_page(page); } else { /* rw == WRITE */ if (err) { SetPageError(page); if (page->mapping) mapping_set_error(page->mapping, err); } end_page_writeback(page); } } EXPORT_SYMBOL_GPL(page_endio); |
485bb99b4 [PATCH] kernel-do... |
859 860 861 |
/** * __lock_page - get a lock on the page, assuming we need to sleep to get it * @page: the page to lock |
1da177e4c Linux-2.6.12-rc2 |
862 |
*/ |
920c7a5d0 mm: remove fastca... |
863 |
void __lock_page(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
864 |
{ |
48c935ad8 page-flags: defin... |
865 866 |
struct page *page_head = compound_head(page); DEFINE_WAIT_BIT(wait, &page_head->flags, PG_locked); |
1da177e4c Linux-2.6.12-rc2 |
867 |
|
48c935ad8 page-flags: defin... |
868 |
__wait_on_bit_lock(page_waitqueue(page_head), &wait, bit_wait_io, |
1da177e4c Linux-2.6.12-rc2 |
869 870 871 |
TASK_UNINTERRUPTIBLE); } EXPORT_SYMBOL(__lock_page); |
b5606c2d4 remove final fast... |
872 |
int __lock_page_killable(struct page *page) |
2687a3569 Add lock_page_kil... |
873 |
{ |
48c935ad8 page-flags: defin... |
874 875 |
struct page *page_head = compound_head(page); DEFINE_WAIT_BIT(wait, &page_head->flags, PG_locked); |
2687a3569 Add lock_page_kil... |
876 |
|
48c935ad8 page-flags: defin... |
877 |
return __wait_on_bit_lock(page_waitqueue(page_head), &wait, |
743162013 sched: Remove pro... |
878 |
bit_wait_io, TASK_KILLABLE); |
2687a3569 Add lock_page_kil... |
879 |
} |
18bc0bbd1 Staging: pohmelfs... |
880 |
EXPORT_SYMBOL_GPL(__lock_page_killable); |
2687a3569 Add lock_page_kil... |
881 |
|
9a95f3cf7 mm: describe mmap... |
882 883 884 885 886 887 888 889 890 891 892 |
/* * Return values: * 1 - page is locked; mmap_sem is still held. * 0 - page is not locked. * mmap_sem has been released (up_read()), unless flags had both * FAULT_FLAG_ALLOW_RETRY and FAULT_FLAG_RETRY_NOWAIT set, in * which case mmap_sem is still held. * * If neither ALLOW_RETRY nor KILLABLE are set, will always return 1 * with the page locked and the mmap_sem unperturbed. */ |
d065bd810 mm: retry page fa... |
893 894 895 |
int __lock_page_or_retry(struct page *page, struct mm_struct *mm, unsigned int flags) { |
37b23e052 x86,mm: make page... |
896 897 898 899 900 901 902 903 904 905 906 907 |
if (flags & FAULT_FLAG_ALLOW_RETRY) { /* * CAUTION! In this case, mmap_sem is not released * even though return 0. */ if (flags & FAULT_FLAG_RETRY_NOWAIT) return 0; up_read(&mm->mmap_sem); if (flags & FAULT_FLAG_KILLABLE) wait_on_page_locked_killable(page); else |
318b275fb mm: allow GUP to ... |
908 |
wait_on_page_locked(page); |
d065bd810 mm: retry page fa... |
909 |
return 0; |
37b23e052 x86,mm: make page... |
910 911 912 913 914 915 916 917 918 919 920 921 |
} else { if (flags & FAULT_FLAG_KILLABLE) { int ret; ret = __lock_page_killable(page); if (ret) { up_read(&mm->mmap_sem); return 0; } } else __lock_page(page); return 1; |
d065bd810 mm: retry page fa... |
922 923 |
} } |
485bb99b4 [PATCH] kernel-do... |
924 |
/** |
e7b563bb2 mm: filemap: move... |
925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 |
* page_cache_next_hole - find the next hole (not-present entry) * @mapping: mapping * @index: index * @max_scan: maximum range to search * * Search the set [index, min(index+max_scan-1, MAX_INDEX)] for the * lowest indexed hole. * * Returns: the index of the hole if found, otherwise returns an index * outside of the set specified (in which case 'return - index >= * max_scan' will be true). In rare cases of index wrap-around, 0 will * be returned. * * page_cache_next_hole may be called under rcu_read_lock. However, * like radix_tree_gang_lookup, this will not atomically search a * snapshot of the tree at a single point in time. For example, if a * hole is created at index 5, then subsequently a hole is created at * index 10, page_cache_next_hole covering both indexes may return 10 * if called under rcu_read_lock. */ pgoff_t page_cache_next_hole(struct address_space *mapping, pgoff_t index, unsigned long max_scan) { unsigned long i; for (i = 0; i < max_scan; i++) { |
0cd6144aa mm + fs: prepare ... |
951 952 953 954 |
struct page *page; page = radix_tree_lookup(&mapping->page_tree, index); if (!page || radix_tree_exceptional_entry(page)) |
e7b563bb2 mm: filemap: move... |
955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 |
break; index++; if (index == 0) break; } return index; } EXPORT_SYMBOL(page_cache_next_hole); /** * page_cache_prev_hole - find the prev hole (not-present entry) * @mapping: mapping * @index: index * @max_scan: maximum range to search * * Search backwards in the range [max(index-max_scan+1, 0), index] for * the first hole. * * Returns: the index of the hole if found, otherwise returns an index * outside of the set specified (in which case 'index - return >= * max_scan' will be true). In rare cases of wrap-around, ULONG_MAX * will be returned. * * page_cache_prev_hole may be called under rcu_read_lock. However, * like radix_tree_gang_lookup, this will not atomically search a * snapshot of the tree at a single point in time. For example, if a * hole is created at index 10, then subsequently a hole is created at * index 5, page_cache_prev_hole covering both indexes may return 5 if * called under rcu_read_lock. */ pgoff_t page_cache_prev_hole(struct address_space *mapping, pgoff_t index, unsigned long max_scan) { unsigned long i; for (i = 0; i < max_scan; i++) { |
0cd6144aa mm + fs: prepare ... |
992 993 994 995 |
struct page *page; page = radix_tree_lookup(&mapping->page_tree, index); if (!page || radix_tree_exceptional_entry(page)) |
e7b563bb2 mm: filemap: move... |
996 997 998 999 1000 1001 1002 1003 1004 1005 1006 |
break; index--; if (index == ULONG_MAX) break; } return index; } EXPORT_SYMBOL(page_cache_prev_hole); /** |
0cd6144aa mm + fs: prepare ... |
1007 |
* find_get_entry - find and get a page cache entry |
485bb99b4 [PATCH] kernel-do... |
1008 |
* @mapping: the address_space to search |
0cd6144aa mm + fs: prepare ... |
1009 1010 1011 1012 |
* @offset: the page cache index * * Looks up the page cache slot at @mapping & @offset. If there is a * page cache page, it is returned with an increased refcount. |
485bb99b4 [PATCH] kernel-do... |
1013 |
* |
139b6a6fb mm: filemap: upda... |
1014 1015 |
* If the slot holds a shadow entry of a previously evicted page, or a * swap entry from shmem/tmpfs, it is returned. |
0cd6144aa mm + fs: prepare ... |
1016 1017 |
* * Otherwise, %NULL is returned. |
1da177e4c Linux-2.6.12-rc2 |
1018 |
*/ |
0cd6144aa mm + fs: prepare ... |
1019 |
struct page *find_get_entry(struct address_space *mapping, pgoff_t offset) |
1da177e4c Linux-2.6.12-rc2 |
1020 |
{ |
a60637c85 mm: lockless page... |
1021 |
void **pagep; |
1da177e4c Linux-2.6.12-rc2 |
1022 |
struct page *page; |
a60637c85 mm: lockless page... |
1023 1024 1025 1026 1027 1028 |
rcu_read_lock(); repeat: page = NULL; pagep = radix_tree_lookup_slot(&mapping->page_tree, offset); if (pagep) { page = radix_tree_deref_slot(pagep); |
27d20fddc radix-tree: fix R... |
1029 1030 |
if (unlikely(!page)) goto out; |
a2c16d6cb mm: let swap use ... |
1031 |
if (radix_tree_exception(page)) { |
8079b1c85 mm: clarify the r... |
1032 1033 1034 |
if (radix_tree_deref_retry(page)) goto repeat; /* |
139b6a6fb mm: filemap: upda... |
1035 1036 1037 |
* A shadow entry of a recently evicted page, * or a swap entry from shmem/tmpfs. Return * it without attempting to raise page count. |
8079b1c85 mm: clarify the r... |
1038 1039 |
*/ goto out; |
a2c16d6cb mm: let swap use ... |
1040 |
} |
a60637c85 mm: lockless page... |
1041 1042 1043 1044 1045 1046 1047 1048 1049 |
if (!page_cache_get_speculative(page)) goto repeat; /* * Has the page moved? * This is part of the lockless pagecache protocol. See * include/linux/pagemap.h for details. */ if (unlikely(page != *pagep)) { |
09cbfeaf1 mm, fs: get rid o... |
1050 |
put_page(page); |
a60637c85 mm: lockless page... |
1051 1052 1053 |
goto repeat; } } |
27d20fddc radix-tree: fix R... |
1054 |
out: |
a60637c85 mm: lockless page... |
1055 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
1056 1057 |
return page; } |
0cd6144aa mm + fs: prepare ... |
1058 |
EXPORT_SYMBOL(find_get_entry); |
1da177e4c Linux-2.6.12-rc2 |
1059 |
|
485bb99b4 [PATCH] kernel-do... |
1060 |
/** |
0cd6144aa mm + fs: prepare ... |
1061 1062 1063 1064 1065 1066 1067 1068 |
* find_lock_entry - locate, pin and lock a page cache entry * @mapping: the address_space to search * @offset: the page cache index * * Looks up the page cache slot at @mapping & @offset. If there is a * page cache page, it is returned locked and with an increased * refcount. * |
139b6a6fb mm: filemap: upda... |
1069 1070 |
* If the slot holds a shadow entry of a previously evicted page, or a * swap entry from shmem/tmpfs, it is returned. |
0cd6144aa mm + fs: prepare ... |
1071 1072 1073 1074 1075 1076 |
* * Otherwise, %NULL is returned. * * find_lock_entry() may sleep. */ struct page *find_lock_entry(struct address_space *mapping, pgoff_t offset) |
1da177e4c Linux-2.6.12-rc2 |
1077 1078 |
{ struct page *page; |
1da177e4c Linux-2.6.12-rc2 |
1079 |
repeat: |
0cd6144aa mm + fs: prepare ... |
1080 |
page = find_get_entry(mapping, offset); |
a2c16d6cb mm: let swap use ... |
1081 |
if (page && !radix_tree_exception(page)) { |
a60637c85 mm: lockless page... |
1082 1083 1084 1085 |
lock_page(page); /* Has the page been truncated? */ if (unlikely(page->mapping != mapping)) { unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
1086 |
put_page(page); |
a60637c85 mm: lockless page... |
1087 |
goto repeat; |
1da177e4c Linux-2.6.12-rc2 |
1088 |
} |
309381fea mm: dump page whe... |
1089 |
VM_BUG_ON_PAGE(page->index != offset, page); |
1da177e4c Linux-2.6.12-rc2 |
1090 |
} |
1da177e4c Linux-2.6.12-rc2 |
1091 1092 |
return page; } |
0cd6144aa mm + fs: prepare ... |
1093 1094 1095 |
EXPORT_SYMBOL(find_lock_entry); /** |
2457aec63 mm: non-atomicall... |
1096 |
* pagecache_get_page - find and get a page reference |
0cd6144aa mm + fs: prepare ... |
1097 1098 |
* @mapping: the address_space to search * @offset: the page index |
2457aec63 mm: non-atomicall... |
1099 |
* @fgp_flags: PCG flags |
45f87de57 mm: get rid of ra... |
1100 |
* @gfp_mask: gfp mask to use for the page cache data page allocation |
0cd6144aa mm + fs: prepare ... |
1101 |
* |
2457aec63 mm: non-atomicall... |
1102 |
* Looks up the page cache slot at @mapping & @offset. |
1da177e4c Linux-2.6.12-rc2 |
1103 |
* |
75325189c mm: fix filemap.c... |
1104 |
* PCG flags modify how the page is returned. |
0cd6144aa mm + fs: prepare ... |
1105 |
* |
2457aec63 mm: non-atomicall... |
1106 1107 1108 |
* FGP_ACCESSED: the page will be marked accessed * FGP_LOCK: Page is return locked * FGP_CREAT: If page is not present then a new page is allocated using |
45f87de57 mm: get rid of ra... |
1109 1110 1111 |
* @gfp_mask and added to the page cache and the VM's LRU * list. The page is returned locked and with an increased * refcount. Otherwise, %NULL is returned. |
1da177e4c Linux-2.6.12-rc2 |
1112 |
* |
2457aec63 mm: non-atomicall... |
1113 1114 |
* If FGP_LOCK or FGP_CREAT are specified then the function may sleep even * if the GFP flags specified for FGP_CREAT are atomic. |
1da177e4c Linux-2.6.12-rc2 |
1115 |
* |
2457aec63 mm: non-atomicall... |
1116 |
* If there is a page cache page, it is returned with an increased refcount. |
1da177e4c Linux-2.6.12-rc2 |
1117 |
*/ |
2457aec63 mm: non-atomicall... |
1118 |
struct page *pagecache_get_page(struct address_space *mapping, pgoff_t offset, |
45f87de57 mm: get rid of ra... |
1119 |
int fgp_flags, gfp_t gfp_mask) |
1da177e4c Linux-2.6.12-rc2 |
1120 |
{ |
eb2be1893 mm: buffered writ... |
1121 |
struct page *page; |
2457aec63 mm: non-atomicall... |
1122 |
|
1da177e4c Linux-2.6.12-rc2 |
1123 |
repeat: |
2457aec63 mm: non-atomicall... |
1124 1125 1126 1127 1128 1129 1130 1131 1132 |
page = find_get_entry(mapping, offset); if (radix_tree_exceptional_entry(page)) page = NULL; if (!page) goto no_page; if (fgp_flags & FGP_LOCK) { if (fgp_flags & FGP_NOWAIT) { if (!trylock_page(page)) { |
09cbfeaf1 mm, fs: get rid o... |
1133 |
put_page(page); |
2457aec63 mm: non-atomicall... |
1134 1135 1136 1137 1138 1139 1140 1141 1142 |
return NULL; } } else { lock_page(page); } /* Has the page been truncated? */ if (unlikely(page->mapping != mapping)) { unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
1143 |
put_page(page); |
2457aec63 mm: non-atomicall... |
1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 |
goto repeat; } VM_BUG_ON_PAGE(page->index != offset, page); } if (page && (fgp_flags & FGP_ACCESSED)) mark_page_accessed(page); no_page: if (!page && (fgp_flags & FGP_CREAT)) { int err; if ((fgp_flags & FGP_WRITE) && mapping_cap_account_dirty(mapping)) |
45f87de57 mm: get rid of ra... |
1156 1157 1158 |
gfp_mask |= __GFP_WRITE; if (fgp_flags & FGP_NOFS) gfp_mask &= ~__GFP_FS; |
2457aec63 mm: non-atomicall... |
1159 |
|
45f87de57 mm: get rid of ra... |
1160 |
page = __page_cache_alloc(gfp_mask); |
eb2be1893 mm: buffered writ... |
1161 1162 |
if (!page) return NULL; |
2457aec63 mm: non-atomicall... |
1163 1164 1165 |
if (WARN_ON_ONCE(!(fgp_flags & FGP_LOCK))) fgp_flags |= FGP_LOCK; |
eb39d618f mm: replace init_... |
1166 |
/* Init accessed so avoid atomic mark_page_accessed later */ |
2457aec63 mm: non-atomicall... |
1167 |
if (fgp_flags & FGP_ACCESSED) |
eb39d618f mm: replace init_... |
1168 |
__SetPageReferenced(page); |
2457aec63 mm: non-atomicall... |
1169 |
|
45f87de57 mm: get rid of ra... |
1170 1171 |
err = add_to_page_cache_lru(page, mapping, offset, gfp_mask & GFP_RECLAIM_MASK); |
eb2be1893 mm: buffered writ... |
1172 |
if (unlikely(err)) { |
09cbfeaf1 mm, fs: get rid o... |
1173 |
put_page(page); |
eb2be1893 mm: buffered writ... |
1174 1175 1176 |
page = NULL; if (err == -EEXIST) goto repeat; |
1da177e4c Linux-2.6.12-rc2 |
1177 |
} |
1da177e4c Linux-2.6.12-rc2 |
1178 |
} |
2457aec63 mm: non-atomicall... |
1179 |
|
1da177e4c Linux-2.6.12-rc2 |
1180 1181 |
return page; } |
2457aec63 mm: non-atomicall... |
1182 |
EXPORT_SYMBOL(pagecache_get_page); |
1da177e4c Linux-2.6.12-rc2 |
1183 1184 |
/** |
0cd6144aa mm + fs: prepare ... |
1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 |
* find_get_entries - gang pagecache lookup * @mapping: The address_space to search * @start: The starting page cache index * @nr_entries: The maximum number of entries * @entries: Where the resulting entries are placed * @indices: The cache indices corresponding to the entries in @entries * * find_get_entries() will search for and return a group of up to * @nr_entries entries in the mapping. The entries are placed at * @entries. find_get_entries() takes a reference against any actual * pages it returns. * * The search returns a group of mapping-contiguous page cache entries * with ascending indexes. There may be holes in the indices due to * not-present pages. * |
139b6a6fb mm: filemap: upda... |
1201 1202 |
* Any shadow entries of evicted pages, or swap entries from * shmem/tmpfs, are included in the returned array. |
0cd6144aa mm + fs: prepare ... |
1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 |
* * find_get_entries() returns the number of pages and shadow entries * which were found. */ unsigned find_get_entries(struct address_space *mapping, pgoff_t start, unsigned int nr_entries, struct page **entries, pgoff_t *indices) { void **slot; unsigned int ret = 0; struct radix_tree_iter iter; if (!nr_entries) return 0; rcu_read_lock(); |
0cd6144aa mm + fs: prepare ... |
1219 1220 1221 1222 1223 1224 1225 |
radix_tree_for_each_slot(slot, &mapping->page_tree, &iter, start) { struct page *page; repeat: page = radix_tree_deref_slot(slot); if (unlikely(!page)) continue; if (radix_tree_exception(page)) { |
2cf938aae mm: use radix_tre... |
1226 1227 1228 1229 |
if (radix_tree_deref_retry(page)) { slot = radix_tree_iter_retry(&iter); continue; } |
0cd6144aa mm + fs: prepare ... |
1230 |
/* |
f9fe48bec dax: support dirt... |
1231 1232 1233 |
* A shadow entry of a recently evicted page, a swap * entry from shmem/tmpfs or a DAX entry. Return it * without attempting to raise page count. |
0cd6144aa mm + fs: prepare ... |
1234 1235 1236 1237 1238 1239 1240 1241 |
*/ goto export; } if (!page_cache_get_speculative(page)) goto repeat; /* Has the page moved? */ if (unlikely(page != *slot)) { |
09cbfeaf1 mm, fs: get rid o... |
1242 |
put_page(page); |
0cd6144aa mm + fs: prepare ... |
1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 |
goto repeat; } export: indices[ret] = iter.index; entries[ret] = page; if (++ret == nr_entries) break; } rcu_read_unlock(); return ret; } /** |
1da177e4c Linux-2.6.12-rc2 |
1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 |
* find_get_pages - gang pagecache lookup * @mapping: The address_space to search * @start: The starting page index * @nr_pages: The maximum number of pages * @pages: Where the resulting pages are placed * * find_get_pages() will search for and return a group of up to * @nr_pages pages in the mapping. The pages are placed at @pages. * find_get_pages() takes a reference against the returned pages. * * The search returns a group of mapping-contiguous pages with ascending * indexes. There may be holes in the indices due to not-present pages. * * find_get_pages() returns the number of pages which were found. */ unsigned find_get_pages(struct address_space *mapping, pgoff_t start, unsigned int nr_pages, struct page **pages) { |
0fc9d1040 radix-tree: use i... |
1274 1275 1276 1277 1278 1279 |
struct radix_tree_iter iter; void **slot; unsigned ret = 0; if (unlikely(!nr_pages)) return 0; |
a60637c85 mm: lockless page... |
1280 1281 |
rcu_read_lock(); |
0fc9d1040 radix-tree: use i... |
1282 |
radix_tree_for_each_slot(slot, &mapping->page_tree, &iter, start) { |
a60637c85 mm: lockless page... |
1283 1284 |
struct page *page; repeat: |
0fc9d1040 radix-tree: use i... |
1285 |
page = radix_tree_deref_slot(slot); |
a60637c85 mm: lockless page... |
1286 1287 |
if (unlikely(!page)) continue; |
9d8aa4ea8 mm: remove worryi... |
1288 |
|
a2c16d6cb mm: let swap use ... |
1289 |
if (radix_tree_exception(page)) { |
8079b1c85 mm: clarify the r... |
1290 |
if (radix_tree_deref_retry(page)) { |
2cf938aae mm: use radix_tre... |
1291 1292 |
slot = radix_tree_iter_retry(&iter); continue; |
8079b1c85 mm: clarify the r... |
1293 |
} |
a2c16d6cb mm: let swap use ... |
1294 |
/* |
139b6a6fb mm: filemap: upda... |
1295 1296 1297 |
* A shadow entry of a recently evicted page, * or a swap entry from shmem/tmpfs. Skip * over it. |
a2c16d6cb mm: let swap use ... |
1298 |
*/ |
8079b1c85 mm: clarify the r... |
1299 |
continue; |
27d20fddc radix-tree: fix R... |
1300 |
} |
a60637c85 mm: lockless page... |
1301 1302 1303 1304 1305 |
if (!page_cache_get_speculative(page)) goto repeat; /* Has the page moved? */ |
0fc9d1040 radix-tree: use i... |
1306 |
if (unlikely(page != *slot)) { |
09cbfeaf1 mm, fs: get rid o... |
1307 |
put_page(page); |
a60637c85 mm: lockless page... |
1308 1309 |
goto repeat; } |
1da177e4c Linux-2.6.12-rc2 |
1310 |
|
a60637c85 mm: lockless page... |
1311 |
pages[ret] = page; |
0fc9d1040 radix-tree: use i... |
1312 1313 |
if (++ret == nr_pages) break; |
a60637c85 mm: lockless page... |
1314 |
} |
5b280c0cc mm: don't return ... |
1315 |
|
a60637c85 mm: lockless page... |
1316 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
1317 1318 |
return ret; } |
ebf43500e [PATCH] Add find_... |
1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 |
/** * find_get_pages_contig - gang contiguous pagecache lookup * @mapping: The address_space to search * @index: The starting page index * @nr_pages: The maximum number of pages * @pages: Where the resulting pages are placed * * find_get_pages_contig() works exactly like find_get_pages(), except * that the returned number of pages are guaranteed to be contiguous. * * find_get_pages_contig() returns the number of pages which were found. */ unsigned find_get_pages_contig(struct address_space *mapping, pgoff_t index, unsigned int nr_pages, struct page **pages) { |
0fc9d1040 radix-tree: use i... |
1334 1335 1336 1337 1338 1339 |
struct radix_tree_iter iter; void **slot; unsigned int ret = 0; if (unlikely(!nr_pages)) return 0; |
a60637c85 mm: lockless page... |
1340 1341 |
rcu_read_lock(); |
0fc9d1040 radix-tree: use i... |
1342 |
radix_tree_for_each_contig(slot, &mapping->page_tree, &iter, index) { |
a60637c85 mm: lockless page... |
1343 1344 |
struct page *page; repeat: |
0fc9d1040 radix-tree: use i... |
1345 1346 |
page = radix_tree_deref_slot(slot); /* The hole, there no reason to continue */ |
a60637c85 mm: lockless page... |
1347 |
if (unlikely(!page)) |
0fc9d1040 radix-tree: use i... |
1348 |
break; |
9d8aa4ea8 mm: remove worryi... |
1349 |
|
a2c16d6cb mm: let swap use ... |
1350 |
if (radix_tree_exception(page)) { |
8079b1c85 mm: clarify the r... |
1351 |
if (radix_tree_deref_retry(page)) { |
2cf938aae mm: use radix_tre... |
1352 1353 |
slot = radix_tree_iter_retry(&iter); continue; |
8079b1c85 mm: clarify the r... |
1354 |
} |
a2c16d6cb mm: let swap use ... |
1355 |
/* |
139b6a6fb mm: filemap: upda... |
1356 1357 1358 |
* A shadow entry of a recently evicted page, * or a swap entry from shmem/tmpfs. Stop * looking for contiguous pages. |
a2c16d6cb mm: let swap use ... |
1359 |
*/ |
8079b1c85 mm: clarify the r... |
1360 |
break; |
a2c16d6cb mm: let swap use ... |
1361 |
} |
ebf43500e [PATCH] Add find_... |
1362 |
|
a60637c85 mm: lockless page... |
1363 1364 1365 1366 |
if (!page_cache_get_speculative(page)) goto repeat; /* Has the page moved? */ |
0fc9d1040 radix-tree: use i... |
1367 |
if (unlikely(page != *slot)) { |
09cbfeaf1 mm, fs: get rid o... |
1368 |
put_page(page); |
a60637c85 mm: lockless page... |
1369 1370 |
goto repeat; } |
9cbb4cb21 mm: find_get_page... |
1371 1372 1373 1374 1375 |
/* * must check mapping and index after taking the ref. * otherwise we can get both false positives and false * negatives, which is just confusing to the caller. */ |
0fc9d1040 radix-tree: use i... |
1376 |
if (page->mapping == NULL || page->index != iter.index) { |
09cbfeaf1 mm, fs: get rid o... |
1377 |
put_page(page); |
9cbb4cb21 mm: find_get_page... |
1378 1379 |
break; } |
a60637c85 mm: lockless page... |
1380 |
pages[ret] = page; |
0fc9d1040 radix-tree: use i... |
1381 1382 |
if (++ret == nr_pages) break; |
ebf43500e [PATCH] Add find_... |
1383 |
} |
a60637c85 mm: lockless page... |
1384 1385 |
rcu_read_unlock(); return ret; |
ebf43500e [PATCH] Add find_... |
1386 |
} |
ef71c15c4 AFS: export a cou... |
1387 |
EXPORT_SYMBOL(find_get_pages_contig); |
ebf43500e [PATCH] Add find_... |
1388 |
|
485bb99b4 [PATCH] kernel-do... |
1389 1390 1391 1392 1393 1394 1395 1396 |
/** * find_get_pages_tag - find and return pages that match @tag * @mapping: the address_space to search * @index: the starting page index * @tag: the tag index * @nr_pages: the maximum number of pages * @pages: where the resulting pages are placed * |
1da177e4c Linux-2.6.12-rc2 |
1397 |
* Like find_get_pages, except we only return pages which are tagged with |
485bb99b4 [PATCH] kernel-do... |
1398 |
* @tag. We update @index to index the next page for the traversal. |
1da177e4c Linux-2.6.12-rc2 |
1399 1400 1401 1402 |
*/ unsigned find_get_pages_tag(struct address_space *mapping, pgoff_t *index, int tag, unsigned int nr_pages, struct page **pages) { |
0fc9d1040 radix-tree: use i... |
1403 1404 1405 1406 1407 1408 |
struct radix_tree_iter iter; void **slot; unsigned ret = 0; if (unlikely(!nr_pages)) return 0; |
a60637c85 mm: lockless page... |
1409 1410 |
rcu_read_lock(); |
0fc9d1040 radix-tree: use i... |
1411 1412 |
radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter, *index, tag) { |
a60637c85 mm: lockless page... |
1413 1414 |
struct page *page; repeat: |
0fc9d1040 radix-tree: use i... |
1415 |
page = radix_tree_deref_slot(slot); |
a60637c85 mm: lockless page... |
1416 1417 |
if (unlikely(!page)) continue; |
9d8aa4ea8 mm: remove worryi... |
1418 |
|
a2c16d6cb mm: let swap use ... |
1419 |
if (radix_tree_exception(page)) { |
8079b1c85 mm: clarify the r... |
1420 |
if (radix_tree_deref_retry(page)) { |
2cf938aae mm: use radix_tre... |
1421 1422 |
slot = radix_tree_iter_retry(&iter); continue; |
8079b1c85 mm: clarify the r... |
1423 |
} |
a2c16d6cb mm: let swap use ... |
1424 |
/* |
139b6a6fb mm: filemap: upda... |
1425 1426 1427 1428 1429 1430 1431 1432 1433 |
* A shadow entry of a recently evicted page. * * Those entries should never be tagged, but * this tree walk is lockless and the tags are * looked up in bulk, one radix tree node at a * time, so there is a sizable window for page * reclaim to evict a page we saw tagged. * * Skip over it. |
a2c16d6cb mm: let swap use ... |
1434 |
*/ |
139b6a6fb mm: filemap: upda... |
1435 |
continue; |
a2c16d6cb mm: let swap use ... |
1436 |
} |
a60637c85 mm: lockless page... |
1437 1438 1439 1440 1441 |
if (!page_cache_get_speculative(page)) goto repeat; /* Has the page moved? */ |
0fc9d1040 radix-tree: use i... |
1442 |
if (unlikely(page != *slot)) { |
09cbfeaf1 mm, fs: get rid o... |
1443 |
put_page(page); |
a60637c85 mm: lockless page... |
1444 1445 1446 1447 |
goto repeat; } pages[ret] = page; |
0fc9d1040 radix-tree: use i... |
1448 1449 |
if (++ret == nr_pages) break; |
a60637c85 mm: lockless page... |
1450 |
} |
5b280c0cc mm: don't return ... |
1451 |
|
a60637c85 mm: lockless page... |
1452 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
1453 |
|
1da177e4c Linux-2.6.12-rc2 |
1454 1455 |
if (ret) *index = pages[ret - 1]->index + 1; |
a60637c85 mm: lockless page... |
1456 |
|
1da177e4c Linux-2.6.12-rc2 |
1457 1458 |
return ret; } |
ef71c15c4 AFS: export a cou... |
1459 |
EXPORT_SYMBOL(find_get_pages_tag); |
1da177e4c Linux-2.6.12-rc2 |
1460 |
|
7e7f77498 mm: add find_get_... |
1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 |
/** * find_get_entries_tag - find and return entries that match @tag * @mapping: the address_space to search * @start: the starting page cache index * @tag: the tag index * @nr_entries: the maximum number of entries * @entries: where the resulting entries are placed * @indices: the cache indices corresponding to the entries in @entries * * Like find_get_entries, except we only return entries which are tagged with * @tag. */ unsigned find_get_entries_tag(struct address_space *mapping, pgoff_t start, int tag, unsigned int nr_entries, struct page **entries, pgoff_t *indices) { void **slot; unsigned int ret = 0; struct radix_tree_iter iter; if (!nr_entries) return 0; rcu_read_lock(); |
7e7f77498 mm: add find_get_... |
1485 1486 1487 1488 1489 1490 1491 1492 1493 |
radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter, start, tag) { struct page *page; repeat: page = radix_tree_deref_slot(slot); if (unlikely(!page)) continue; if (radix_tree_exception(page)) { if (radix_tree_deref_retry(page)) { |
2cf938aae mm: use radix_tre... |
1494 1495 |
slot = radix_tree_iter_retry(&iter); continue; |
7e7f77498 mm: add find_get_... |
1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 |
} /* * A shadow entry of a recently evicted page, a swap * entry from shmem/tmpfs or a DAX entry. Return it * without attempting to raise page count. */ goto export; } if (!page_cache_get_speculative(page)) goto repeat; /* Has the page moved? */ if (unlikely(page != *slot)) { |
09cbfeaf1 mm, fs: get rid o... |
1510 |
put_page(page); |
7e7f77498 mm: add find_get_... |
1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 |
goto repeat; } export: indices[ret] = iter.index; entries[ret] = page; if (++ret == nr_entries) break; } rcu_read_unlock(); return ret; } EXPORT_SYMBOL(find_get_entries_tag); |
76d42bd96 [PATCH] readahead... |
1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 |
/* * CD/DVDs are error prone. When a medium error occurs, the driver may fail * a _large_ part of the i/o request. Imagine the worst scenario: * * ---R__________________________________________B__________ * ^ reading here ^ bad block(assume 4k) * * read(R) => miss => readahead(R...B) => media error => frustrating retries * => failing the whole request => read(R) => read(R+1) => * readahead(R+1...B+1) => bang => read(R+2) => read(R+3) => * readahead(R+3...B+2) => bang => read(R+3) => read(R+4) => * readahead(R+4...B+3) => bang => read(R+4) => read(R+5) => ...... * * It is going insane. Fix it by quickly scaling down the readahead size. */ static void shrink_readahead_size_eio(struct file *filp, struct file_ra_state *ra) { |
76d42bd96 [PATCH] readahead... |
1541 |
ra->ra_pages /= 4; |
76d42bd96 [PATCH] readahead... |
1542 |
} |
485bb99b4 [PATCH] kernel-do... |
1543 |
/** |
36e789144 kill do_generic_m... |
1544 |
* do_generic_file_read - generic file read routine |
485bb99b4 [PATCH] kernel-do... |
1545 1546 |
* @filp: the file to read * @ppos: current file position |
6e58e79db introduce copy_pa... |
1547 1548 |
* @iter: data destination * @written: already copied |
485bb99b4 [PATCH] kernel-do... |
1549 |
* |
1da177e4c Linux-2.6.12-rc2 |
1550 |
* This is a generic file read routine, and uses the |
485bb99b4 [PATCH] kernel-do... |
1551 |
* mapping->a_ops->readpage() function for the actual low-level stuff. |
1da177e4c Linux-2.6.12-rc2 |
1552 1553 1554 |
* * This is really ugly. But the goto's actually try to clarify some * of the logic when it comes to error handling etc. |
1da177e4c Linux-2.6.12-rc2 |
1555 |
*/ |
6e58e79db introduce copy_pa... |
1556 1557 |
static ssize_t do_generic_file_read(struct file *filp, loff_t *ppos, struct iov_iter *iter, ssize_t written) |
1da177e4c Linux-2.6.12-rc2 |
1558 |
{ |
36e789144 kill do_generic_m... |
1559 |
struct address_space *mapping = filp->f_mapping; |
1da177e4c Linux-2.6.12-rc2 |
1560 |
struct inode *inode = mapping->host; |
36e789144 kill do_generic_m... |
1561 |
struct file_ra_state *ra = &filp->f_ra; |
57f6b96c0 filemap: convert ... |
1562 1563 1564 1565 |
pgoff_t index; pgoff_t last_index; pgoff_t prev_index; unsigned long offset; /* offset into pagecache page */ |
ec0f16372 readahead: improv... |
1566 |
unsigned int prev_offset; |
6e58e79db introduce copy_pa... |
1567 |
int error = 0; |
1da177e4c Linux-2.6.12-rc2 |
1568 |
|
09cbfeaf1 mm, fs: get rid o... |
1569 1570 1571 1572 1573 |
index = *ppos >> PAGE_SHIFT; prev_index = ra->prev_pos >> PAGE_SHIFT; prev_offset = ra->prev_pos & (PAGE_SIZE-1); last_index = (*ppos + iter->count + PAGE_SIZE-1) >> PAGE_SHIFT; offset = *ppos & ~PAGE_MASK; |
1da177e4c Linux-2.6.12-rc2 |
1574 |
|
1da177e4c Linux-2.6.12-rc2 |
1575 1576 |
for (;;) { struct page *page; |
57f6b96c0 filemap: convert ... |
1577 |
pgoff_t end_index; |
a32ea1e1f Fix read/truncate... |
1578 |
loff_t isize; |
1da177e4c Linux-2.6.12-rc2 |
1579 |
unsigned long nr, ret; |
1da177e4c Linux-2.6.12-rc2 |
1580 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
1581 1582 |
find_page: page = find_get_page(mapping, index); |
3ea89ee86 readahead: conver... |
1583 |
if (!page) { |
cf914a7d6 readahead: split ... |
1584 |
page_cache_sync_readahead(mapping, |
7ff81078d readahead: remove... |
1585 |
ra, filp, |
3ea89ee86 readahead: conver... |
1586 1587 1588 1589 1590 1591 |
index, last_index - index); page = find_get_page(mapping, index); if (unlikely(page == NULL)) goto no_cached_page; } if (PageReadahead(page)) { |
cf914a7d6 readahead: split ... |
1592 |
page_cache_async_readahead(mapping, |
7ff81078d readahead: remove... |
1593 |
ra, filp, page, |
3ea89ee86 readahead: conver... |
1594 |
index, last_index - index); |
1da177e4c Linux-2.6.12-rc2 |
1595 |
} |
8ab22b9ab vfs: pagecache us... |
1596 |
if (!PageUptodate(page)) { |
ebded0278 mm: filemap: avoi... |
1597 1598 1599 1600 1601 1602 1603 1604 |
/* * See comment in do_read_cache_page on why * wait_on_page_locked is used to avoid unnecessarily * serialisations and why it's safe. */ wait_on_page_locked_killable(page); if (PageUptodate(page)) goto page_ok; |
09cbfeaf1 mm, fs: get rid o... |
1605 |
if (inode->i_blkbits == PAGE_SHIFT || |
8ab22b9ab vfs: pagecache us... |
1606 1607 |
!mapping->a_ops->is_partially_uptodate) goto page_not_up_to_date; |
529ae9aaa mm: rename page t... |
1608 |
if (!trylock_page(page)) |
8ab22b9ab vfs: pagecache us... |
1609 |
goto page_not_up_to_date; |
8d056cb96 mm/vfs: revalidat... |
1610 1611 1612 |
/* Did it get truncated before we got the lock? */ if (!page->mapping) goto page_not_up_to_date_locked; |
8ab22b9ab vfs: pagecache us... |
1613 |
if (!mapping->a_ops->is_partially_uptodate(page, |
6e58e79db introduce copy_pa... |
1614 |
offset, iter->count)) |
8ab22b9ab vfs: pagecache us... |
1615 1616 1617 |
goto page_not_up_to_date_locked; unlock_page(page); } |
1da177e4c Linux-2.6.12-rc2 |
1618 |
page_ok: |
a32ea1e1f Fix read/truncate... |
1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 |
/* * i_size must be checked after we know the page is Uptodate. * * Checking i_size after the check allows us to calculate * the correct value for "nr", which means the zero-filled * part of the page is not copied back to userspace (unless * another truncate extends the file - this is desired though). */ isize = i_size_read(inode); |
09cbfeaf1 mm, fs: get rid o... |
1629 |
end_index = (isize - 1) >> PAGE_SHIFT; |
a32ea1e1f Fix read/truncate... |
1630 |
if (unlikely(!isize || index > end_index)) { |
09cbfeaf1 mm, fs: get rid o... |
1631 |
put_page(page); |
a32ea1e1f Fix read/truncate... |
1632 1633 1634 1635 |
goto out; } /* nr is the maximum number of bytes to copy from this page */ |
09cbfeaf1 mm, fs: get rid o... |
1636 |
nr = PAGE_SIZE; |
a32ea1e1f Fix read/truncate... |
1637 |
if (index == end_index) { |
09cbfeaf1 mm, fs: get rid o... |
1638 |
nr = ((isize - 1) & ~PAGE_MASK) + 1; |
a32ea1e1f Fix read/truncate... |
1639 |
if (nr <= offset) { |
09cbfeaf1 mm, fs: get rid o... |
1640 |
put_page(page); |
a32ea1e1f Fix read/truncate... |
1641 1642 1643 1644 |
goto out; } } nr = nr - offset; |
1da177e4c Linux-2.6.12-rc2 |
1645 1646 1647 1648 1649 1650 1651 1652 1653 |
/* If users can be writing to this page using arbitrary * virtual addresses, take care about potential aliasing * before reading the page on the kernel side. */ if (mapping_writably_mapped(mapping)) flush_dcache_page(page); /* |
ec0f16372 readahead: improv... |
1654 1655 |
* When a sequential read accesses a page several times, * only mark it as accessed the first time. |
1da177e4c Linux-2.6.12-rc2 |
1656 |
*/ |
ec0f16372 readahead: improv... |
1657 |
if (prev_index != index || offset != prev_offset) |
1da177e4c Linux-2.6.12-rc2 |
1658 1659 1660 1661 1662 1663 |
mark_page_accessed(page); prev_index = index; /* * Ok, we have the page, and it's up-to-date, so * now we can copy it to user space... |
1da177e4c Linux-2.6.12-rc2 |
1664 |
*/ |
6e58e79db introduce copy_pa... |
1665 1666 |
ret = copy_page_to_iter(page, offset, nr, iter); |
1da177e4c Linux-2.6.12-rc2 |
1667 |
offset += ret; |
09cbfeaf1 mm, fs: get rid o... |
1668 1669 |
index += offset >> PAGE_SHIFT; offset &= ~PAGE_MASK; |
6ce745ed3 readahead: code c... |
1670 |
prev_offset = offset; |
1da177e4c Linux-2.6.12-rc2 |
1671 |
|
09cbfeaf1 mm, fs: get rid o... |
1672 |
put_page(page); |
6e58e79db introduce copy_pa... |
1673 1674 1675 1676 1677 1678 1679 1680 |
written += ret; if (!iov_iter_count(iter)) goto out; if (ret < nr) { error = -EFAULT; goto out; } continue; |
1da177e4c Linux-2.6.12-rc2 |
1681 1682 1683 |
page_not_up_to_date: /* Get exclusive access to the page ... */ |
854623235 do_generic_file_r... |
1684 1685 1686 |
error = lock_page_killable(page); if (unlikely(error)) goto readpage_error; |
1da177e4c Linux-2.6.12-rc2 |
1687 |
|
8ab22b9ab vfs: pagecache us... |
1688 |
page_not_up_to_date_locked: |
da6052f7b [PATCH] update so... |
1689 |
/* Did it get truncated before we got the lock? */ |
1da177e4c Linux-2.6.12-rc2 |
1690 1691 |
if (!page->mapping) { unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
1692 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 |
continue; } /* Did somebody else fill it already? */ if (PageUptodate(page)) { unlock_page(page); goto page_ok; } readpage: |
91803b499 do_generic_file_r... |
1703 1704 1705 1706 1707 1708 |
/* * A previous I/O error may have been due to temporary * failures, eg. multipath errors. * PG_error will be set again if readpage fails. */ ClearPageError(page); |
1da177e4c Linux-2.6.12-rc2 |
1709 1710 |
/* Start the actual read. The read will unlock the page. */ error = mapping->a_ops->readpage(filp, page); |
994fc28c7 [PATCH] add AOP_T... |
1711 1712 |
if (unlikely(error)) { if (error == AOP_TRUNCATED_PAGE) { |
09cbfeaf1 mm, fs: get rid o... |
1713 |
put_page(page); |
6e58e79db introduce copy_pa... |
1714 |
error = 0; |
994fc28c7 [PATCH] add AOP_T... |
1715 1716 |
goto find_page; } |
1da177e4c Linux-2.6.12-rc2 |
1717 |
goto readpage_error; |
994fc28c7 [PATCH] add AOP_T... |
1718 |
} |
1da177e4c Linux-2.6.12-rc2 |
1719 1720 |
if (!PageUptodate(page)) { |
854623235 do_generic_file_r... |
1721 1722 1723 |
error = lock_page_killable(page); if (unlikely(error)) goto readpage_error; |
1da177e4c Linux-2.6.12-rc2 |
1724 1725 1726 |
if (!PageUptodate(page)) { if (page->mapping == NULL) { /* |
2ecdc82ef kill unused inval... |
1727 |
* invalidate_mapping_pages got it |
1da177e4c Linux-2.6.12-rc2 |
1728 1729 |
*/ unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
1730 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1731 1732 1733 |
goto find_page; } unlock_page(page); |
7ff81078d readahead: remove... |
1734 |
shrink_readahead_size_eio(filp, ra); |
854623235 do_generic_file_r... |
1735 1736 |
error = -EIO; goto readpage_error; |
1da177e4c Linux-2.6.12-rc2 |
1737 1738 1739 |
} unlock_page(page); } |
1da177e4c Linux-2.6.12-rc2 |
1740 1741 1742 1743 |
goto page_ok; readpage_error: /* UHHUH! A synchronous read error occurred. Report it */ |
09cbfeaf1 mm, fs: get rid o... |
1744 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1745 1746 1747 1748 1749 1750 1751 |
goto out; no_cached_page: /* * Ok, it wasn't cached, so we need to create a new * page.. */ |
eb2be1893 mm: buffered writ... |
1752 1753 |
page = page_cache_alloc_cold(mapping); if (!page) { |
6e58e79db introduce copy_pa... |
1754 |
error = -ENOMEM; |
eb2be1893 mm: buffered writ... |
1755 |
goto out; |
1da177e4c Linux-2.6.12-rc2 |
1756 |
} |
6afdb859b mm: do not ignore... |
1757 |
error = add_to_page_cache_lru(page, mapping, index, |
c62d25556 mm, fs: introduce... |
1758 |
mapping_gfp_constraint(mapping, GFP_KERNEL)); |
1da177e4c Linux-2.6.12-rc2 |
1759 |
if (error) { |
09cbfeaf1 mm, fs: get rid o... |
1760 |
put_page(page); |
6e58e79db introduce copy_pa... |
1761 1762 |
if (error == -EEXIST) { error = 0; |
1da177e4c Linux-2.6.12-rc2 |
1763 |
goto find_page; |
6e58e79db introduce copy_pa... |
1764 |
} |
1da177e4c Linux-2.6.12-rc2 |
1765 1766 |
goto out; } |
1da177e4c Linux-2.6.12-rc2 |
1767 1768 1769 1770 |
goto readpage; } out: |
7ff81078d readahead: remove... |
1771 |
ra->prev_pos = prev_index; |
09cbfeaf1 mm, fs: get rid o... |
1772 |
ra->prev_pos <<= PAGE_SHIFT; |
7ff81078d readahead: remove... |
1773 |
ra->prev_pos |= prev_offset; |
1da177e4c Linux-2.6.12-rc2 |
1774 |
|
09cbfeaf1 mm, fs: get rid o... |
1775 |
*ppos = ((loff_t)index << PAGE_SHIFT) + offset; |
0c6aa2639 mm: do_generic_fi... |
1776 |
file_accessed(filp); |
6e58e79db introduce copy_pa... |
1777 |
return written ? written : error; |
1da177e4c Linux-2.6.12-rc2 |
1778 |
} |
485bb99b4 [PATCH] kernel-do... |
1779 |
/** |
6abd23227 bury generic_file... |
1780 |
* generic_file_read_iter - generic filesystem read routine |
485bb99b4 [PATCH] kernel-do... |
1781 |
* @iocb: kernel I/O control block |
6abd23227 bury generic_file... |
1782 |
* @iter: destination for the data read |
485bb99b4 [PATCH] kernel-do... |
1783 |
* |
6abd23227 bury generic_file... |
1784 |
* This is the "read_iter()" routine for all filesystems |
1da177e4c Linux-2.6.12-rc2 |
1785 1786 1787 |
* that can use the page cache directly. */ ssize_t |
ed978a811 new helper: gener... |
1788 |
generic_file_read_iter(struct kiocb *iocb, struct iov_iter *iter) |
1da177e4c Linux-2.6.12-rc2 |
1789 |
{ |
ed978a811 new helper: gener... |
1790 |
struct file *file = iocb->ki_filp; |
cb66a7a1f kill generic_segm... |
1791 |
ssize_t retval = 0; |
e7080a439 mm/filemap: gener... |
1792 1793 1794 1795 |
size_t count = iov_iter_count(iter); if (!count) goto out; /* skip atime */ |
1da177e4c Linux-2.6.12-rc2 |
1796 |
|
2ba48ce51 mirror O_APPEND a... |
1797 |
if (iocb->ki_flags & IOCB_DIRECT) { |
ed978a811 new helper: gener... |
1798 1799 |
struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; |
543ade1fc [PATCH] Streamlin... |
1800 |
loff_t size; |
1da177e4c Linux-2.6.12-rc2 |
1801 |
|
1da177e4c Linux-2.6.12-rc2 |
1802 |
size = i_size_read(inode); |
c64fb5c74 filemap: remove p... |
1803 1804 |
retval = filemap_write_and_wait_range(mapping, iocb->ki_pos, iocb->ki_pos + count - 1); |
9fe55eea7 Fix race when che... |
1805 |
if (!retval) { |
ed978a811 new helper: gener... |
1806 |
struct iov_iter data = *iter; |
c8b8e32d7 direct-io: elimin... |
1807 |
retval = mapping->a_ops->direct_IO(iocb, &data); |
9fe55eea7 Fix race when che... |
1808 |
} |
d8d3d94b8 pass iov_iter to ... |
1809 |
|
9fe55eea7 Fix race when che... |
1810 |
if (retval > 0) { |
c64fb5c74 filemap: remove p... |
1811 |
iocb->ki_pos += retval; |
ed978a811 new helper: gener... |
1812 |
iov_iter_advance(iter, retval); |
9fe55eea7 Fix race when che... |
1813 |
} |
66f998f61 fs: allow short d... |
1814 |
|
9fe55eea7 Fix race when che... |
1815 1816 1817 1818 1819 1820 |
/* * Btrfs can have a short DIO read if we encounter * compressed extents, so if there was an error, or if * we've already read everything we wanted to, or if * there was a short read because we hit EOF, go ahead * and return. Otherwise fallthrough to buffered io for |
fbbbad4bc vfs,ext2: introdu... |
1821 1822 |
* the rest of the read. Buffered reads will not work for * DAX files, so don't bother trying. |
9fe55eea7 Fix race when che... |
1823 |
*/ |
c64fb5c74 filemap: remove p... |
1824 |
if (retval < 0 || !iov_iter_count(iter) || iocb->ki_pos >= size || |
fbbbad4bc vfs,ext2: introdu... |
1825 |
IS_DAX(inode)) { |
ed978a811 new helper: gener... |
1826 |
file_accessed(file); |
9fe55eea7 Fix race when che... |
1827 |
goto out; |
0e0bcae3b [GFS2] Fix direct... |
1828 |
} |
1da177e4c Linux-2.6.12-rc2 |
1829 |
} |
c64fb5c74 filemap: remove p... |
1830 |
retval = do_generic_file_read(file, &iocb->ki_pos, iter, retval); |
1da177e4c Linux-2.6.12-rc2 |
1831 1832 1833 |
out: return retval; } |
ed978a811 new helper: gener... |
1834 |
EXPORT_SYMBOL(generic_file_read_iter); |
1da177e4c Linux-2.6.12-rc2 |
1835 |
|
1da177e4c Linux-2.6.12-rc2 |
1836 |
#ifdef CONFIG_MMU |
485bb99b4 [PATCH] kernel-do... |
1837 1838 1839 1840 |
/** * page_cache_read - adds requested page to the page cache if not already there * @file: file to read * @offset: page index |
62eb320ab mm: fix filemap.c... |
1841 |
* @gfp_mask: memory allocation flags |
485bb99b4 [PATCH] kernel-do... |
1842 |
* |
1da177e4c Linux-2.6.12-rc2 |
1843 1844 1845 |
* This adds the requested page to the page cache if it isn't already there, * and schedules an I/O to read in its contents from disk. */ |
c20cd45eb mm: allow GFP_{FS... |
1846 |
static int page_cache_read(struct file *file, pgoff_t offset, gfp_t gfp_mask) |
1da177e4c Linux-2.6.12-rc2 |
1847 1848 |
{ struct address_space *mapping = file->f_mapping; |
99dadfdde mm/filemap.c: rem... |
1849 |
struct page *page; |
994fc28c7 [PATCH] add AOP_T... |
1850 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
1851 |
|
994fc28c7 [PATCH] add AOP_T... |
1852 |
do { |
c20cd45eb mm: allow GFP_{FS... |
1853 |
page = __page_cache_alloc(gfp_mask|__GFP_COLD); |
994fc28c7 [PATCH] add AOP_T... |
1854 1855 |
if (!page) return -ENOMEM; |
c20cd45eb mm: allow GFP_{FS... |
1856 |
ret = add_to_page_cache_lru(page, mapping, offset, gfp_mask & GFP_KERNEL); |
994fc28c7 [PATCH] add AOP_T... |
1857 1858 1859 1860 |
if (ret == 0) ret = mapping->a_ops->readpage(file, page); else if (ret == -EEXIST) ret = 0; /* losing race to add is OK */ |
1da177e4c Linux-2.6.12-rc2 |
1861 |
|
09cbfeaf1 mm, fs: get rid o... |
1862 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1863 |
|
994fc28c7 [PATCH] add AOP_T... |
1864 |
} while (ret == AOP_TRUNCATED_PAGE); |
99dadfdde mm/filemap.c: rem... |
1865 |
|
994fc28c7 [PATCH] add AOP_T... |
1866 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
1867 1868 1869 |
} #define MMAP_LOTSAMISS (100) |
ef00e08e2 readahead: clean ... |
1870 1871 1872 1873 1874 1875 1876 1877 1878 |
/* * Synchronous readahead happens when we don't even find * a page in the page cache at all. */ static void do_sync_mmap_readahead(struct vm_area_struct *vma, struct file_ra_state *ra, struct file *file, pgoff_t offset) { |
ef00e08e2 readahead: clean ... |
1879 1880 1881 |
struct address_space *mapping = file->f_mapping; /* If we don't want any read-ahead, don't bother */ |
64363aad5 mm: remove unused... |
1882 |
if (vma->vm_flags & VM_RAND_READ) |
ef00e08e2 readahead: clean ... |
1883 |
return; |
275b12bf5 readahead: return... |
1884 1885 |
if (!ra->ra_pages) return; |
ef00e08e2 readahead: clean ... |
1886 |
|
64363aad5 mm: remove unused... |
1887 |
if (vma->vm_flags & VM_SEQ_READ) { |
7ffc59b4d readahead: enforc... |
1888 1889 |
page_cache_sync_readahead(mapping, ra, file, offset, ra->ra_pages); |
ef00e08e2 readahead: clean ... |
1890 1891 |
return; } |
207d04baa readahead: reduce... |
1892 1893 |
/* Avoid banging the cache line if not needed */ if (ra->mmap_miss < MMAP_LOTSAMISS * 10) |
ef00e08e2 readahead: clean ... |
1894 1895 1896 1897 1898 1899 1900 1901 |
ra->mmap_miss++; /* * Do we miss much more than hit in this file? If so, * stop bothering with read-ahead. It will only hurt. */ if (ra->mmap_miss > MMAP_LOTSAMISS) return; |
d30a11004 readahead: record... |
1902 1903 1904 |
/* * mmap read-around */ |
600e19afc mm: use only per-... |
1905 1906 1907 |
ra->start = max_t(long, 0, offset - ra->ra_pages / 2); ra->size = ra->ra_pages; ra->async_size = ra->ra_pages / 4; |
275b12bf5 readahead: return... |
1908 |
ra_submit(ra, mapping, file); |
ef00e08e2 readahead: clean ... |
1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 |
} /* * Asynchronous readahead happens when we find the page and PG_readahead, * so we want to possibly extend the readahead further.. */ static void do_async_mmap_readahead(struct vm_area_struct *vma, struct file_ra_state *ra, struct file *file, struct page *page, pgoff_t offset) { struct address_space *mapping = file->f_mapping; /* If we don't want any read-ahead, don't bother */ |
64363aad5 mm: remove unused... |
1924 |
if (vma->vm_flags & VM_RAND_READ) |
ef00e08e2 readahead: clean ... |
1925 1926 1927 1928 |
return; if (ra->mmap_miss > 0) ra->mmap_miss--; if (PageReadahead(page)) |
2fad6f5de readahead: enforc... |
1929 1930 |
page_cache_async_readahead(mapping, ra, file, page, offset, ra->ra_pages); |
ef00e08e2 readahead: clean ... |
1931 |
} |
485bb99b4 [PATCH] kernel-do... |
1932 |
/** |
54cb8821d mm: merge populat... |
1933 |
* filemap_fault - read in file data for page fault handling |
d0217ac04 mm: fault feedbac... |
1934 1935 |
* @vma: vma in which the fault was taken * @vmf: struct vm_fault containing details of the fault |
485bb99b4 [PATCH] kernel-do... |
1936 |
* |
54cb8821d mm: merge populat... |
1937 |
* filemap_fault() is invoked via the vma operations vector for a |
1da177e4c Linux-2.6.12-rc2 |
1938 1939 1940 1941 1942 |
* mapped memory region to read in file data during a page fault. * * The goto's are kind of ugly, but this streamlines the normal case of having * it in the page cache, and handles the special cases reasonably without * having a lot of duplicated code. |
9a95f3cf7 mm: describe mmap... |
1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 |
* * vma->vm_mm->mmap_sem must be held on entry. * * If our return value has VM_FAULT_RETRY set, it's because * lock_page_or_retry() returned 0. * The mmap_sem has usually been released in this case. * See __lock_page_or_retry() for the exception. * * If our return value does not have VM_FAULT_RETRY set, the mmap_sem * has not been released. * * We never return with VM_FAULT_RETRY and a bit from VM_FAULT_ERROR set. |
1da177e4c Linux-2.6.12-rc2 |
1955 |
*/ |
d0217ac04 mm: fault feedbac... |
1956 |
int filemap_fault(struct vm_area_struct *vma, struct vm_fault *vmf) |
1da177e4c Linux-2.6.12-rc2 |
1957 1958 |
{ int error; |
54cb8821d mm: merge populat... |
1959 |
struct file *file = vma->vm_file; |
1da177e4c Linux-2.6.12-rc2 |
1960 1961 1962 |
struct address_space *mapping = file->f_mapping; struct file_ra_state *ra = &file->f_ra; struct inode *inode = mapping->host; |
ef00e08e2 readahead: clean ... |
1963 |
pgoff_t offset = vmf->pgoff; |
1da177e4c Linux-2.6.12-rc2 |
1964 |
struct page *page; |
99e3e53f4 mm: cleanup size ... |
1965 |
loff_t size; |
83c54070e mm: fault feedbac... |
1966 |
int ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
1967 |
|
09cbfeaf1 mm, fs: get rid o... |
1968 1969 |
size = round_up(i_size_read(inode), PAGE_SIZE); if (offset >= size >> PAGE_SHIFT) |
5307cc1aa Remove broken ptr... |
1970 |
return VM_FAULT_SIGBUS; |
1da177e4c Linux-2.6.12-rc2 |
1971 |
|
1da177e4c Linux-2.6.12-rc2 |
1972 |
/* |
494264208 mm: memcg: handle... |
1973 |
* Do we have something in the page cache already? |
1da177e4c Linux-2.6.12-rc2 |
1974 |
*/ |
ef00e08e2 readahead: clean ... |
1975 |
page = find_get_page(mapping, offset); |
45cac65b0 readahead: fault ... |
1976 |
if (likely(page) && !(vmf->flags & FAULT_FLAG_TRIED)) { |
1da177e4c Linux-2.6.12-rc2 |
1977 |
/* |
ef00e08e2 readahead: clean ... |
1978 1979 |
* We found the page, so try async readahead before * waiting for the lock. |
1da177e4c Linux-2.6.12-rc2 |
1980 |
*/ |
ef00e08e2 readahead: clean ... |
1981 |
do_async_mmap_readahead(vma, ra, file, page, offset); |
45cac65b0 readahead: fault ... |
1982 |
} else if (!page) { |
ef00e08e2 readahead: clean ... |
1983 1984 1985 |
/* No page in the page cache at all */ do_sync_mmap_readahead(vma, ra, file, offset); count_vm_event(PGMAJFAULT); |
456f998ec memcg: add the pa... |
1986 |
mem_cgroup_count_vm_event(vma->vm_mm, PGMAJFAULT); |
ef00e08e2 readahead: clean ... |
1987 1988 |
ret = VM_FAULT_MAJOR; retry_find: |
b522c94da mm: filemap_fault... |
1989 |
page = find_get_page(mapping, offset); |
1da177e4c Linux-2.6.12-rc2 |
1990 1991 1992 |
if (!page) goto no_cached_page; } |
d88c0922f Release page refe... |
1993 |
if (!lock_page_or_retry(page, vma->vm_mm, vmf->flags)) { |
09cbfeaf1 mm, fs: get rid o... |
1994 |
put_page(page); |
d065bd810 mm: retry page fa... |
1995 |
return ret | VM_FAULT_RETRY; |
d88c0922f Release page refe... |
1996 |
} |
b522c94da mm: filemap_fault... |
1997 1998 1999 2000 2001 2002 2003 |
/* Did it get truncated? */ if (unlikely(page->mapping != mapping)) { unlock_page(page); put_page(page); goto retry_find; } |
309381fea mm: dump page whe... |
2004 |
VM_BUG_ON_PAGE(page->index != offset, page); |
b522c94da mm: filemap_fault... |
2005 |
|
1da177e4c Linux-2.6.12-rc2 |
2006 |
/* |
d00806b18 mm: fix fault vs ... |
2007 2008 |
* We have a locked page in the page cache, now we need to check * that it's up-to-date. If not, it is going to be due to an error. |
1da177e4c Linux-2.6.12-rc2 |
2009 |
*/ |
d00806b18 mm: fix fault vs ... |
2010 |
if (unlikely(!PageUptodate(page))) |
1da177e4c Linux-2.6.12-rc2 |
2011 |
goto page_not_uptodate; |
ef00e08e2 readahead: clean ... |
2012 2013 2014 2015 |
/* * Found the page and have a reference on it. * We must recheck i_size under page lock. */ |
09cbfeaf1 mm, fs: get rid o... |
2016 2017 |
size = round_up(i_size_read(inode), PAGE_SIZE); if (unlikely(offset >= size >> PAGE_SHIFT)) { |
d00806b18 mm: fix fault vs ... |
2018 |
unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
2019 |
put_page(page); |
5307cc1aa Remove broken ptr... |
2020 |
return VM_FAULT_SIGBUS; |
d00806b18 mm: fix fault vs ... |
2021 |
} |
d0217ac04 mm: fault feedbac... |
2022 |
vmf->page = page; |
83c54070e mm: fault feedbac... |
2023 |
return ret | VM_FAULT_LOCKED; |
1da177e4c Linux-2.6.12-rc2 |
2024 |
|
1da177e4c Linux-2.6.12-rc2 |
2025 2026 2027 2028 2029 |
no_cached_page: /* * We're only likely to ever get here if MADV_RANDOM is in * effect. */ |
c20cd45eb mm: allow GFP_{FS... |
2030 |
error = page_cache_read(file, offset, vmf->gfp_mask); |
1da177e4c Linux-2.6.12-rc2 |
2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 |
/* * The page we want has now been added to the page cache. * In the unlikely event that someone removed it in the * meantime, we'll just come back here and read it again. */ if (error >= 0) goto retry_find; /* * An error return from page_cache_read can result if the * system is low on memory, or a problem occurs while trying * to schedule I/O. */ if (error == -ENOMEM) |
d0217ac04 mm: fault feedbac... |
2046 2047 |
return VM_FAULT_OOM; return VM_FAULT_SIGBUS; |
1da177e4c Linux-2.6.12-rc2 |
2048 2049 |
page_not_uptodate: |
1da177e4c Linux-2.6.12-rc2 |
2050 2051 2052 2053 2054 2055 |
/* * Umm, take care of errors if the page isn't up-to-date. * Try to re-read it _once_. We do this synchronously, * because there really aren't any performance issues here * and we need to check for errors. */ |
1da177e4c Linux-2.6.12-rc2 |
2056 |
ClearPageError(page); |
994fc28c7 [PATCH] add AOP_T... |
2057 |
error = mapping->a_ops->readpage(file, page); |
3ef0f720e mm: fix infinite ... |
2058 2059 2060 2061 2062 |
if (!error) { wait_on_page_locked(page); if (!PageUptodate(page)) error = -EIO; } |
09cbfeaf1 mm, fs: get rid o... |
2063 |
put_page(page); |
d00806b18 mm: fix fault vs ... |
2064 2065 |
if (!error || error == AOP_TRUNCATED_PAGE) |
994fc28c7 [PATCH] add AOP_T... |
2066 |
goto retry_find; |
1da177e4c Linux-2.6.12-rc2 |
2067 |
|
d00806b18 mm: fix fault vs ... |
2068 |
/* Things didn't work out. Return zero to tell the mm layer so. */ |
76d42bd96 [PATCH] readahead... |
2069 |
shrink_readahead_size_eio(file, ra); |
d0217ac04 mm: fault feedbac... |
2070 |
return VM_FAULT_SIGBUS; |
54cb8821d mm: merge populat... |
2071 2072 |
} EXPORT_SYMBOL(filemap_fault); |
f1820361f mm: implement ->m... |
2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 |
void filemap_map_pages(struct vm_area_struct *vma, struct vm_fault *vmf) { struct radix_tree_iter iter; void **slot; struct file *file = vma->vm_file; struct address_space *mapping = file->f_mapping; loff_t size; struct page *page; unsigned long address = (unsigned long) vmf->virtual_address; unsigned long addr; pte_t *pte; rcu_read_lock(); radix_tree_for_each_slot(slot, &mapping->page_tree, &iter, vmf->pgoff) { if (iter.index > vmf->max_pgoff) break; repeat: page = radix_tree_deref_slot(slot); if (unlikely(!page)) goto next; if (radix_tree_exception(page)) { |
2cf938aae mm: use radix_tre... |
2094 2095 2096 2097 2098 |
if (radix_tree_deref_retry(page)) { slot = radix_tree_iter_retry(&iter); continue; } goto next; |
f1820361f mm: implement ->m... |
2099 2100 2101 2102 2103 2104 2105 |
} if (!page_cache_get_speculative(page)) goto repeat; /* Has the page moved? */ if (unlikely(page != *slot)) { |
09cbfeaf1 mm, fs: get rid o... |
2106 |
put_page(page); |
f1820361f mm: implement ->m... |
2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 |
goto repeat; } if (!PageUptodate(page) || PageReadahead(page) || PageHWPoison(page)) goto skip; if (!trylock_page(page)) goto skip; if (page->mapping != mapping || !PageUptodate(page)) goto unlock; |
09cbfeaf1 mm, fs: get rid o... |
2119 2120 |
size = round_up(i_size_read(mapping->host), PAGE_SIZE); if (page->index >= size >> PAGE_SHIFT) |
f1820361f mm: implement ->m... |
2121 2122 2123 2124 2125 2126 2127 2128 2129 |
goto unlock; pte = vmf->pte + page->index - vmf->pgoff; if (!pte_none(*pte)) goto unlock; if (file->f_ra.mmap_miss > 0) file->f_ra.mmap_miss--; addr = address + (page->index - vmf->pgoff) * PAGE_SIZE; |
5c0a85fad mm: make faultaro... |
2130 |
do_set_pte(vma, addr, page, pte, false, false, true); |
f1820361f mm: implement ->m... |
2131 2132 2133 2134 2135 |
unlock_page(page); goto next; unlock: unlock_page(page); skip: |
09cbfeaf1 mm, fs: get rid o... |
2136 |
put_page(page); |
f1820361f mm: implement ->m... |
2137 2138 2139 2140 2141 2142 2143 |
next: if (iter.index == vmf->max_pgoff) break; } rcu_read_unlock(); } EXPORT_SYMBOL(filemap_map_pages); |
4fcf1c620 mm: Make default ... |
2144 2145 2146 |
int filemap_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { struct page *page = vmf->page; |
496ad9aa8 new helper: file_... |
2147 |
struct inode *inode = file_inode(vma->vm_file); |
4fcf1c620 mm: Make default ... |
2148 |
int ret = VM_FAULT_LOCKED; |
14da92001 fs: Protect write... |
2149 |
sb_start_pagefault(inode->i_sb); |
4fcf1c620 mm: Make default ... |
2150 2151 2152 2153 2154 2155 2156 |
file_update_time(vma->vm_file); lock_page(page); if (page->mapping != inode->i_mapping) { unlock_page(page); ret = VM_FAULT_NOPAGE; goto out; } |
14da92001 fs: Protect write... |
2157 2158 2159 2160 2161 2162 |
/* * We mark the page dirty already here so that when freeze is in * progress, we are guaranteed that writeback during freezing will * see the dirty page and writeprotect it again. */ set_page_dirty(page); |
1d1d1a767 mm: only enforce ... |
2163 |
wait_for_stable_page(page); |
4fcf1c620 mm: Make default ... |
2164 |
out: |
14da92001 fs: Protect write... |
2165 |
sb_end_pagefault(inode->i_sb); |
4fcf1c620 mm: Make default ... |
2166 2167 2168 |
return ret; } EXPORT_SYMBOL(filemap_page_mkwrite); |
f0f37e2f7 const: mark struc... |
2169 |
const struct vm_operations_struct generic_file_vm_ops = { |
54cb8821d mm: merge populat... |
2170 |
.fault = filemap_fault, |
f1820361f mm: implement ->m... |
2171 |
.map_pages = filemap_map_pages, |
4fcf1c620 mm: Make default ... |
2172 |
.page_mkwrite = filemap_page_mkwrite, |
1da177e4c Linux-2.6.12-rc2 |
2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 |
}; /* This is used for a general mmap of a disk file */ int generic_file_mmap(struct file * file, struct vm_area_struct * vma) { struct address_space *mapping = file->f_mapping; if (!mapping->a_ops->readpage) return -ENOEXEC; file_accessed(file); vma->vm_ops = &generic_file_vm_ops; return 0; } |
1da177e4c Linux-2.6.12-rc2 |
2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 |
/* * This is for filesystems which do not implement ->writepage. */ int generic_file_readonly_mmap(struct file *file, struct vm_area_struct *vma) { if ((vma->vm_flags & VM_SHARED) && (vma->vm_flags & VM_MAYWRITE)) return -EINVAL; return generic_file_mmap(file, vma); } #else int generic_file_mmap(struct file * file, struct vm_area_struct * vma) { return -ENOSYS; } int generic_file_readonly_mmap(struct file * file, struct vm_area_struct * vma) { return -ENOSYS; } #endif /* CONFIG_MMU */ EXPORT_SYMBOL(generic_file_mmap); EXPORT_SYMBOL(generic_file_readonly_mmap); |
67f9fd91f mm: remove read_c... |
2210 2211 2212 2213 2214 |
static struct page *wait_on_page_read(struct page *page) { if (!IS_ERR(page)) { wait_on_page_locked(page); if (!PageUptodate(page)) { |
09cbfeaf1 mm, fs: get rid o... |
2215 |
put_page(page); |
67f9fd91f mm: remove read_c... |
2216 2217 2218 2219 2220 |
page = ERR_PTR(-EIO); } } return page; } |
32b635298 mm: filemap: remo... |
2221 |
static struct page *do_read_cache_page(struct address_space *mapping, |
57f6b96c0 filemap: convert ... |
2222 |
pgoff_t index, |
5e5358e7c mm: cleanup descr... |
2223 |
int (*filler)(void *, struct page *), |
0531b2aac mm: add new 'read... |
2224 2225 |
void *data, gfp_t gfp) |
1da177e4c Linux-2.6.12-rc2 |
2226 |
{ |
eb2be1893 mm: buffered writ... |
2227 |
struct page *page; |
1da177e4c Linux-2.6.12-rc2 |
2228 2229 2230 2231 |
int err; repeat: page = find_get_page(mapping, index); if (!page) { |
0531b2aac mm: add new 'read... |
2232 |
page = __page_cache_alloc(gfp | __GFP_COLD); |
eb2be1893 mm: buffered writ... |
2233 2234 |
if (!page) return ERR_PTR(-ENOMEM); |
e6f67b8c0 vfs: __read_cache... |
2235 |
err = add_to_page_cache_lru(page, mapping, index, gfp); |
eb2be1893 mm: buffered writ... |
2236 |
if (unlikely(err)) { |
09cbfeaf1 mm, fs: get rid o... |
2237 |
put_page(page); |
eb2be1893 mm: buffered writ... |
2238 2239 |
if (err == -EEXIST) goto repeat; |
1da177e4c Linux-2.6.12-rc2 |
2240 |
/* Presumably ENOMEM for radix tree node */ |
1da177e4c Linux-2.6.12-rc2 |
2241 2242 |
return ERR_PTR(err); } |
32b635298 mm: filemap: remo... |
2243 2244 |
filler: |
1da177e4c Linux-2.6.12-rc2 |
2245 2246 |
err = filler(data, page); if (err < 0) { |
09cbfeaf1 mm, fs: get rid o... |
2247 |
put_page(page); |
32b635298 mm: filemap: remo... |
2248 |
return ERR_PTR(err); |
1da177e4c Linux-2.6.12-rc2 |
2249 |
} |
1da177e4c Linux-2.6.12-rc2 |
2250 |
|
32b635298 mm: filemap: remo... |
2251 2252 2253 2254 2255 |
page = wait_on_page_read(page); if (IS_ERR(page)) return page; goto out; } |
1da177e4c Linux-2.6.12-rc2 |
2256 2257 |
if (PageUptodate(page)) goto out; |
ebded0278 mm: filemap: avoi... |
2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 |
/* * Page is not up to date and may be locked due one of the following * case a: Page is being filled and the page lock is held * case b: Read/write error clearing the page uptodate status * case c: Truncation in progress (page locked) * case d: Reclaim in progress * * Case a, the page will be up to date when the page is unlocked. * There is no need to serialise on the page lock here as the page * is pinned so the lock gives no additional protection. Even if the * the page is truncated, the data is still valid if PageUptodate as * it's a race vs truncate race. * Case b, the page will not be up to date * Case c, the page may be truncated but in itself, the data may still * be valid after IO completes as it's a read vs truncate race. The * operation must restart if the page is not uptodate on unlock but * otherwise serialising on page lock to stabilise the mapping gives * no additional guarantees to the caller as the page lock is * released before return. * Case d, similar to truncation. If reclaim holds the page lock, it * will be a race with remove_mapping that determines if the mapping * is valid on unlock but otherwise the data is valid and there is * no need to serialise with page lock. * * As the page lock gives no additional guarantee, we optimistically * wait on the page to be unlocked and check if it's up to date and * use the page if it is. Otherwise, the page lock is required to * distinguish between the different cases. The motivation is that we * avoid spurious serialisations and wakeups when multiple processes * wait on the same page for IO to complete. */ wait_on_page_locked(page); if (PageUptodate(page)) goto out; /* Distinguish between all the cases under the safety of the lock */ |
1da177e4c Linux-2.6.12-rc2 |
2294 |
lock_page(page); |
ebded0278 mm: filemap: avoi... |
2295 2296 |
/* Case c or d, restart the operation */ |
1da177e4c Linux-2.6.12-rc2 |
2297 2298 |
if (!page->mapping) { unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
2299 |
put_page(page); |
32b635298 mm: filemap: remo... |
2300 |
goto repeat; |
1da177e4c Linux-2.6.12-rc2 |
2301 |
} |
ebded0278 mm: filemap: avoi... |
2302 2303 |
/* Someone else locked and filled the page in a very small window */ |
1da177e4c Linux-2.6.12-rc2 |
2304 2305 2306 2307 |
if (PageUptodate(page)) { unlock_page(page); goto out; } |
32b635298 mm: filemap: remo... |
2308 |
goto filler; |
c855ff371 Fix a bad error c... |
2309 |
out: |
6fe6900e1 mm: make read_cac... |
2310 2311 2312 |
mark_page_accessed(page); return page; } |
0531b2aac mm: add new 'read... |
2313 2314 |
/** |
67f9fd91f mm: remove read_c... |
2315 |
* read_cache_page - read into page cache, fill it if needed |
0531b2aac mm: add new 'read... |
2316 2317 2318 |
* @mapping: the page's address_space * @index: the page index * @filler: function to perform the read |
5e5358e7c mm: cleanup descr... |
2319 |
* @data: first arg to filler(data, page) function, often left as NULL |
0531b2aac mm: add new 'read... |
2320 |
* |
0531b2aac mm: add new 'read... |
2321 |
* Read into the page cache. If a page already exists, and PageUptodate() is |
67f9fd91f mm: remove read_c... |
2322 |
* not set, try to fill the page and wait for it to become unlocked. |
0531b2aac mm: add new 'read... |
2323 2324 2325 |
* * If the page does not get brought uptodate, return -EIO. */ |
67f9fd91f mm: remove read_c... |
2326 |
struct page *read_cache_page(struct address_space *mapping, |
0531b2aac mm: add new 'read... |
2327 |
pgoff_t index, |
5e5358e7c mm: cleanup descr... |
2328 |
int (*filler)(void *, struct page *), |
0531b2aac mm: add new 'read... |
2329 2330 2331 2332 |
void *data) { return do_read_cache_page(mapping, index, filler, data, mapping_gfp_mask(mapping)); } |
67f9fd91f mm: remove read_c... |
2333 |
EXPORT_SYMBOL(read_cache_page); |
0531b2aac mm: add new 'read... |
2334 2335 2336 2337 2338 2339 2340 2341 |
/** * read_cache_page_gfp - read into page cache, using specified page allocation flags. * @mapping: the page's address_space * @index: the page index * @gfp: the page allocator flags to use if allocating * * This is the same as "read_mapping_page(mapping, index, NULL)", but with |
e6f67b8c0 vfs: __read_cache... |
2342 |
* any new page allocations done using the specified allocation flags. |
0531b2aac mm: add new 'read... |
2343 2344 2345 2346 2347 2348 2349 2350 |
* * If the page does not get brought uptodate, return -EIO. */ struct page *read_cache_page_gfp(struct address_space *mapping, pgoff_t index, gfp_t gfp) { filler_t *filler = (filler_t *)mapping->a_ops->readpage; |
67f9fd91f mm: remove read_c... |
2351 |
return do_read_cache_page(mapping, index, filler, NULL, gfp); |
0531b2aac mm: add new 'read... |
2352 2353 |
} EXPORT_SYMBOL(read_cache_page_gfp); |
2f718ffc1 mm: buffered writ... |
2354 |
/* |
1da177e4c Linux-2.6.12-rc2 |
2355 2356 |
* Performs necessary checks before doing a write * |
485bb99b4 [PATCH] kernel-do... |
2357 |
* Can adjust writing position or amount of bytes to write. |
1da177e4c Linux-2.6.12-rc2 |
2358 2359 2360 |
* Returns appropriate error code that caller should return or * zero in case that write should be allowed. */ |
3309dd04c switch generic_wr... |
2361 |
inline ssize_t generic_write_checks(struct kiocb *iocb, struct iov_iter *from) |
1da177e4c Linux-2.6.12-rc2 |
2362 |
{ |
3309dd04c switch generic_wr... |
2363 |
struct file *file = iocb->ki_filp; |
1da177e4c Linux-2.6.12-rc2 |
2364 |
struct inode *inode = file->f_mapping->host; |
59e99e5b9 mm: use rlimit he... |
2365 |
unsigned long limit = rlimit(RLIMIT_FSIZE); |
3309dd04c switch generic_wr... |
2366 |
loff_t pos; |
1da177e4c Linux-2.6.12-rc2 |
2367 |
|
3309dd04c switch generic_wr... |
2368 2369 |
if (!iov_iter_count(from)) return 0; |
1da177e4c Linux-2.6.12-rc2 |
2370 |
|
0fa6b005a generic_write_che... |
2371 |
/* FIXME: this is for backwards compatibility with 2.4 */ |
2ba48ce51 mirror O_APPEND a... |
2372 |
if (iocb->ki_flags & IOCB_APPEND) |
3309dd04c switch generic_wr... |
2373 |
iocb->ki_pos = i_size_read(inode); |
1da177e4c Linux-2.6.12-rc2 |
2374 |
|
3309dd04c switch generic_wr... |
2375 |
pos = iocb->ki_pos; |
1da177e4c Linux-2.6.12-rc2 |
2376 |
|
0fa6b005a generic_write_che... |
2377 |
if (limit != RLIM_INFINITY) { |
3309dd04c switch generic_wr... |
2378 |
if (iocb->ki_pos >= limit) { |
0fa6b005a generic_write_che... |
2379 2380 |
send_sig(SIGXFSZ, current, 0); return -EFBIG; |
1da177e4c Linux-2.6.12-rc2 |
2381 |
} |
3309dd04c switch generic_wr... |
2382 |
iov_iter_truncate(from, limit - (unsigned long)pos); |
1da177e4c Linux-2.6.12-rc2 |
2383 2384 2385 2386 2387 |
} /* * LFS rule */ |
3309dd04c switch generic_wr... |
2388 |
if (unlikely(pos + iov_iter_count(from) > MAX_NON_LFS && |
1da177e4c Linux-2.6.12-rc2 |
2389 |
!(file->f_flags & O_LARGEFILE))) { |
3309dd04c switch generic_wr... |
2390 |
if (pos >= MAX_NON_LFS) |
1da177e4c Linux-2.6.12-rc2 |
2391 |
return -EFBIG; |
3309dd04c switch generic_wr... |
2392 |
iov_iter_truncate(from, MAX_NON_LFS - (unsigned long)pos); |
1da177e4c Linux-2.6.12-rc2 |
2393 2394 2395 2396 2397 2398 2399 2400 2401 |
} /* * Are we about to exceed the fs block limit ? * * If we have written data it becomes a short write. If we have * exceeded without writing data we send a signal and return EFBIG. * Linus frestrict idea will clean these up nicely.. */ |
3309dd04c switch generic_wr... |
2402 2403 |
if (unlikely(pos >= inode->i_sb->s_maxbytes)) return -EFBIG; |
1da177e4c Linux-2.6.12-rc2 |
2404 |
|
3309dd04c switch generic_wr... |
2405 2406 |
iov_iter_truncate(from, inode->i_sb->s_maxbytes - pos); return iov_iter_count(from); |
1da177e4c Linux-2.6.12-rc2 |
2407 2408 |
} EXPORT_SYMBOL(generic_write_checks); |
afddba49d fs: introduce wri... |
2409 2410 2411 2412 2413 |
int pagecache_write_begin(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned flags, struct page **pagep, void **fsdata) { const struct address_space_operations *aops = mapping->a_ops; |
4e02ed4b4 fs: remove prepar... |
2414 |
return aops->write_begin(file, mapping, pos, len, flags, |
afddba49d fs: introduce wri... |
2415 |
pagep, fsdata); |
afddba49d fs: introduce wri... |
2416 2417 2418 2419 2420 2421 2422 2423 |
} EXPORT_SYMBOL(pagecache_write_begin); int pagecache_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, struct page *page, void *fsdata) { const struct address_space_operations *aops = mapping->a_ops; |
afddba49d fs: introduce wri... |
2424 |
|
4e02ed4b4 fs: remove prepar... |
2425 |
return aops->write_end(file, mapping, pos, len, copied, page, fsdata); |
afddba49d fs: introduce wri... |
2426 2427 |
} EXPORT_SYMBOL(pagecache_write_end); |
1da177e4c Linux-2.6.12-rc2 |
2428 |
ssize_t |
1af5bb491 filemap: remove t... |
2429 |
generic_file_direct_write(struct kiocb *iocb, struct iov_iter *from) |
1da177e4c Linux-2.6.12-rc2 |
2430 2431 2432 2433 |
{ struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; |
1af5bb491 filemap: remove t... |
2434 |
loff_t pos = iocb->ki_pos; |
1da177e4c Linux-2.6.12-rc2 |
2435 |
ssize_t written; |
a969e903a kill generic_file... |
2436 2437 |
size_t write_len; pgoff_t end; |
26978b8b4 give ->direct_IO(... |
2438 |
struct iov_iter data; |
1da177e4c Linux-2.6.12-rc2 |
2439 |
|
0c949334a iov_iter_truncate() |
2440 |
write_len = iov_iter_count(from); |
09cbfeaf1 mm, fs: get rid o... |
2441 |
end = (pos + write_len - 1) >> PAGE_SHIFT; |
a969e903a kill generic_file... |
2442 |
|
48b47c561 mm: direct IO sta... |
2443 |
written = filemap_write_and_wait_range(mapping, pos, pos + write_len - 1); |
a969e903a kill generic_file... |
2444 2445 2446 2447 2448 2449 2450 |
if (written) goto out; /* * After a write we want buffered reads to be sure to go to disk to get * the new data. We invalidate clean cached page from the region we're * about to write. We do this *before* the write so that we can return |
6ccfa806a VFS: fix dio writ... |
2451 |
* without clobbering -EIOCBQUEUED from ->direct_IO(). |
a969e903a kill generic_file... |
2452 2453 2454 |
*/ if (mapping->nrpages) { written = invalidate_inode_pages2_range(mapping, |
09cbfeaf1 mm, fs: get rid o... |
2455 |
pos >> PAGE_SHIFT, end); |
6ccfa806a VFS: fix dio writ... |
2456 2457 2458 2459 2460 2461 2462 |
/* * If a page can not be invalidated, return 0 to fall back * to buffered write. */ if (written) { if (written == -EBUSY) return 0; |
a969e903a kill generic_file... |
2463 |
goto out; |
6ccfa806a VFS: fix dio writ... |
2464 |
} |
a969e903a kill generic_file... |
2465 |
} |
26978b8b4 give ->direct_IO(... |
2466 |
data = *from; |
c8b8e32d7 direct-io: elimin... |
2467 |
written = mapping->a_ops->direct_IO(iocb, &data); |
a969e903a kill generic_file... |
2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 |
/* * Finally, try again to invalidate clean pages which might have been * cached by non-direct readahead, or faulted in by get_user_pages() * if the source of the write was an mmap'ed region of the file * we're writing. Either one is a pretty crazy thing to do, * so we don't support it 100%. If this invalidation * fails, tough, the write still worked... */ if (mapping->nrpages) { invalidate_inode_pages2_range(mapping, |
09cbfeaf1 mm, fs: get rid o... |
2479 |
pos >> PAGE_SHIFT, end); |
a969e903a kill generic_file... |
2480 |
} |
1da177e4c Linux-2.6.12-rc2 |
2481 |
if (written > 0) { |
0116651c8 mm: remove tempor... |
2482 |
pos += written; |
f8579f867 generic_file_dire... |
2483 |
iov_iter_advance(from, written); |
0116651c8 mm: remove tempor... |
2484 2485 |
if (pos > i_size_read(inode) && !S_ISBLK(inode->i_mode)) { i_size_write(inode, pos); |
1da177e4c Linux-2.6.12-rc2 |
2486 2487 |
mark_inode_dirty(inode); } |
5cb6c6c7e generic_file_dire... |
2488 |
iocb->ki_pos = pos; |
1da177e4c Linux-2.6.12-rc2 |
2489 |
} |
a969e903a kill generic_file... |
2490 |
out: |
1da177e4c Linux-2.6.12-rc2 |
2491 2492 2493 |
return written; } EXPORT_SYMBOL(generic_file_direct_write); |
eb2be1893 mm: buffered writ... |
2494 2495 2496 2497 |
/* * Find or create a page at the given pagecache position. Return the locked * page. This function is specifically for buffered writes. */ |
54566b2c1 fs: symlink write... |
2498 2499 |
struct page *grab_cache_page_write_begin(struct address_space *mapping, pgoff_t index, unsigned flags) |
eb2be1893 mm: buffered writ... |
2500 |
{ |
eb2be1893 mm: buffered writ... |
2501 |
struct page *page; |
bbddabe2e mm: filemap: only... |
2502 |
int fgp_flags = FGP_LOCK|FGP_WRITE|FGP_CREAT; |
0faa70cb0 mm: filemap: pass... |
2503 |
|
54566b2c1 fs: symlink write... |
2504 |
if (flags & AOP_FLAG_NOFS) |
2457aec63 mm: non-atomicall... |
2505 2506 2507 |
fgp_flags |= FGP_NOFS; page = pagecache_get_page(mapping, index, fgp_flags, |
45f87de57 mm: get rid of ra... |
2508 |
mapping_gfp_mask(mapping)); |
c585a2678 mm: remove likely... |
2509 |
if (page) |
2457aec63 mm: non-atomicall... |
2510 |
wait_for_stable_page(page); |
eb2be1893 mm: buffered writ... |
2511 |
|
eb2be1893 mm: buffered writ... |
2512 2513 |
return page; } |
54566b2c1 fs: symlink write... |
2514 |
EXPORT_SYMBOL(grab_cache_page_write_begin); |
eb2be1893 mm: buffered writ... |
2515 |
|
3b93f911d export generic_pe... |
2516 |
ssize_t generic_perform_write(struct file *file, |
afddba49d fs: introduce wri... |
2517 2518 2519 2520 2521 2522 |
struct iov_iter *i, loff_t pos) { struct address_space *mapping = file->f_mapping; const struct address_space_operations *a_ops = mapping->a_ops; long status = 0; ssize_t written = 0; |
674b892ed mm: restore KERNE... |
2523 2524 2525 2526 2527 |
unsigned int flags = 0; /* * Copies from kernel address space cannot fail (NFSD is a big user). */ |
777eda2c5 new helper: iter_... |
2528 |
if (!iter_is_iovec(i)) |
674b892ed mm: restore KERNE... |
2529 |
flags |= AOP_FLAG_UNINTERRUPTIBLE; |
afddba49d fs: introduce wri... |
2530 2531 2532 |
do { struct page *page; |
afddba49d fs: introduce wri... |
2533 2534 2535 2536 |
unsigned long offset; /* Offset into pagecache page */ unsigned long bytes; /* Bytes to write to page */ size_t copied; /* Bytes copied from user */ void *fsdata; |
09cbfeaf1 mm, fs: get rid o... |
2537 2538 |
offset = (pos & (PAGE_SIZE - 1)); bytes = min_t(unsigned long, PAGE_SIZE - offset, |
afddba49d fs: introduce wri... |
2539 2540 2541 |
iov_iter_count(i)); again: |
00a3d660c Revert "fs: do no... |
2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 |
/* * Bring in the user page that we will copy from _first_. * Otherwise there's a nasty deadlock on copying from the * same page as we're writing to, without it being marked * up-to-date. * * Not only is this an optimisation, but it is also required * to check that the address is actually valid, when atomic * usercopies are used, below. */ if (unlikely(iov_iter_fault_in_readable(i, bytes))) { status = -EFAULT; break; } |
296291cdd mm: make sendfile... |
2556 2557 2558 2559 |
if (fatal_signal_pending(current)) { status = -EINTR; break; } |
674b892ed mm: restore KERNE... |
2560 |
status = a_ops->write_begin(file, mapping, pos, bytes, flags, |
afddba49d fs: introduce wri... |
2561 |
&page, &fsdata); |
2457aec63 mm: non-atomicall... |
2562 |
if (unlikely(status < 0)) |
afddba49d fs: introduce wri... |
2563 |
break; |
931e80e4b mm: flush dcache ... |
2564 2565 |
if (mapping_writably_mapped(mapping)) flush_dcache_page(page); |
00a3d660c Revert "fs: do no... |
2566 |
|
afddba49d fs: introduce wri... |
2567 |
copied = iov_iter_copy_from_user_atomic(page, i, offset, bytes); |
afddba49d fs: introduce wri... |
2568 2569 2570 2571 2572 2573 2574 2575 2576 |
flush_dcache_page(page); status = a_ops->write_end(file, mapping, pos, bytes, copied, page, fsdata); if (unlikely(status < 0)) break; copied = status; cond_resched(); |
124d3b704 fix writev regres... |
2577 |
iov_iter_advance(i, copied); |
afddba49d fs: introduce wri... |
2578 2579 2580 2581 2582 2583 2584 2585 2586 |
if (unlikely(copied == 0)) { /* * If we were unable to copy any data at all, we must * fall back to a single segment length write. * * If we didn't fallback here, we could livelock * because not all segments in the iov can be copied at * once without a pagefault. */ |
09cbfeaf1 mm, fs: get rid o... |
2587 |
bytes = min_t(unsigned long, PAGE_SIZE - offset, |
afddba49d fs: introduce wri... |
2588 2589 2590 |
iov_iter_single_seg_count(i)); goto again; } |
afddba49d fs: introduce wri... |
2591 2592 2593 2594 |
pos += copied; written += copied; balance_dirty_pages_ratelimited(mapping); |
afddba49d fs: introduce wri... |
2595 2596 2597 2598 |
} while (iov_iter_count(i)); return written ? written : status; } |
3b93f911d export generic_pe... |
2599 |
EXPORT_SYMBOL(generic_perform_write); |
1da177e4c Linux-2.6.12-rc2 |
2600 |
|
e4dd9de3c vfs: Export __gen... |
2601 |
/** |
8174202b3 write_iter varian... |
2602 |
* __generic_file_write_iter - write data to a file |
e4dd9de3c vfs: Export __gen... |
2603 |
* @iocb: IO state structure (file, offset, etc.) |
8174202b3 write_iter varian... |
2604 |
* @from: iov_iter with data to write |
e4dd9de3c vfs: Export __gen... |
2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 |
* * This function does all the work needed for actually writing data to a * file. It does all basic checks, removes SUID from the file, updates * modification times and calls proper subroutines depending on whether we * do direct IO or a standard buffered write. * * It expects i_mutex to be grabbed unless we work on a block device or similar * object which does not need locking at all. * * This function does *not* take care of syncing data in case of O_SYNC write. * A caller has to handle it. This is mainly due to the fact that we want to * avoid syncing under i_mutex. */ |
8174202b3 write_iter varian... |
2618 |
ssize_t __generic_file_write_iter(struct kiocb *iocb, struct iov_iter *from) |
1da177e4c Linux-2.6.12-rc2 |
2619 2620 |
{ struct file *file = iocb->ki_filp; |
fb5527e68 [PATCH] direct-io... |
2621 |
struct address_space * mapping = file->f_mapping; |
1da177e4c Linux-2.6.12-rc2 |
2622 |
struct inode *inode = mapping->host; |
3b93f911d export generic_pe... |
2623 |
ssize_t written = 0; |
1da177e4c Linux-2.6.12-rc2 |
2624 |
ssize_t err; |
3b93f911d export generic_pe... |
2625 |
ssize_t status; |
1da177e4c Linux-2.6.12-rc2 |
2626 |
|
1da177e4c Linux-2.6.12-rc2 |
2627 |
/* We can write back this queue in page reclaim */ |
de1414a65 fs: export inode_... |
2628 |
current->backing_dev_info = inode_to_bdi(inode); |
5fa8e0a1c fs: Rename file_r... |
2629 |
err = file_remove_privs(file); |
1da177e4c Linux-2.6.12-rc2 |
2630 2631 |
if (err) goto out; |
c3b2da314 fs: introduce ino... |
2632 2633 2634 |
err = file_update_time(file); if (err) goto out; |
1da177e4c Linux-2.6.12-rc2 |
2635 |
|
2ba48ce51 mirror O_APPEND a... |
2636 |
if (iocb->ki_flags & IOCB_DIRECT) { |
0b8def9d6 __generic_file_wr... |
2637 |
loff_t pos, endbyte; |
fb5527e68 [PATCH] direct-io... |
2638 |
|
1af5bb491 filemap: remove t... |
2639 |
written = generic_file_direct_write(iocb, from); |
1da177e4c Linux-2.6.12-rc2 |
2640 |
/* |
fbbbad4bc vfs,ext2: introdu... |
2641 2642 2643 2644 2645 |
* If the write stopped short of completing, fall back to * buffered writes. Some filesystems do this for writes to * holes, for example. For DAX files, a buffered write will * not succeed (even if it did, DAX does not handle dirty * page-cache pages correctly). |
1da177e4c Linux-2.6.12-rc2 |
2646 |
*/ |
0b8def9d6 __generic_file_wr... |
2647 |
if (written < 0 || !iov_iter_count(from) || IS_DAX(inode)) |
fbbbad4bc vfs,ext2: introdu... |
2648 |
goto out; |
0b8def9d6 __generic_file_wr... |
2649 |
status = generic_perform_write(file, from, pos = iocb->ki_pos); |
fb5527e68 [PATCH] direct-io... |
2650 |
/* |
3b93f911d export generic_pe... |
2651 |
* If generic_perform_write() returned a synchronous error |
fb5527e68 [PATCH] direct-io... |
2652 2653 2654 2655 2656 |
* then we want to return the number of bytes which were * direct-written, or the error code if that was zero. Note * that this differs from normal direct-io semantics, which * will return -EFOO even if some bytes were written. */ |
60bb45297 __generic_file_wr... |
2657 |
if (unlikely(status < 0)) { |
3b93f911d export generic_pe... |
2658 |
err = status; |
fb5527e68 [PATCH] direct-io... |
2659 2660 |
goto out; } |
fb5527e68 [PATCH] direct-io... |
2661 2662 2663 2664 2665 |
/* * We need to ensure that the page cache pages are written to * disk and invalidated to preserve the expected O_DIRECT * semantics. */ |
3b93f911d export generic_pe... |
2666 |
endbyte = pos + status - 1; |
0b8def9d6 __generic_file_wr... |
2667 |
err = filemap_write_and_wait_range(mapping, pos, endbyte); |
fb5527e68 [PATCH] direct-io... |
2668 |
if (err == 0) { |
0b8def9d6 __generic_file_wr... |
2669 |
iocb->ki_pos = endbyte + 1; |
3b93f911d export generic_pe... |
2670 |
written += status; |
fb5527e68 [PATCH] direct-io... |
2671 |
invalidate_mapping_pages(mapping, |
09cbfeaf1 mm, fs: get rid o... |
2672 2673 |
pos >> PAGE_SHIFT, endbyte >> PAGE_SHIFT); |
fb5527e68 [PATCH] direct-io... |
2674 2675 2676 2677 2678 2679 2680 |
} else { /* * We don't know how much we wrote, so just return * the number of bytes which were direct-written */ } } else { |
0b8def9d6 __generic_file_wr... |
2681 2682 2683 |
written = generic_perform_write(file, from, iocb->ki_pos); if (likely(written > 0)) iocb->ki_pos += written; |
fb5527e68 [PATCH] direct-io... |
2684 |
} |
1da177e4c Linux-2.6.12-rc2 |
2685 2686 2687 2688 |
out: current->backing_dev_info = NULL; return written ? written : err; } |
8174202b3 write_iter varian... |
2689 |
EXPORT_SYMBOL(__generic_file_write_iter); |
e4dd9de3c vfs: Export __gen... |
2690 |
|
e4dd9de3c vfs: Export __gen... |
2691 |
/** |
8174202b3 write_iter varian... |
2692 |
* generic_file_write_iter - write data to a file |
e4dd9de3c vfs: Export __gen... |
2693 |
* @iocb: IO state structure |
8174202b3 write_iter varian... |
2694 |
* @from: iov_iter with data to write |
e4dd9de3c vfs: Export __gen... |
2695 |
* |
8174202b3 write_iter varian... |
2696 |
* This is a wrapper around __generic_file_write_iter() to be used by most |
e4dd9de3c vfs: Export __gen... |
2697 2698 2699 |
* filesystems. It takes care of syncing the file in case of O_SYNC file * and acquires i_mutex as needed. */ |
8174202b3 write_iter varian... |
2700 |
ssize_t generic_file_write_iter(struct kiocb *iocb, struct iov_iter *from) |
1da177e4c Linux-2.6.12-rc2 |
2701 2702 |
{ struct file *file = iocb->ki_filp; |
148f948ba vfs: Introduce ne... |
2703 |
struct inode *inode = file->f_mapping->host; |
1da177e4c Linux-2.6.12-rc2 |
2704 |
ssize_t ret; |
1da177e4c Linux-2.6.12-rc2 |
2705 |
|
5955102c9 wrappers for ->i_... |
2706 |
inode_lock(inode); |
3309dd04c switch generic_wr... |
2707 2708 |
ret = generic_write_checks(iocb, from); if (ret > 0) |
5f380c7fa lift generic_writ... |
2709 |
ret = __generic_file_write_iter(iocb, from); |
5955102c9 wrappers for ->i_... |
2710 |
inode_unlock(inode); |
1da177e4c Linux-2.6.12-rc2 |
2711 |
|
e25922176 fs: simplify the ... |
2712 2713 |
if (ret > 0) ret = generic_write_sync(iocb, ret); |
1da177e4c Linux-2.6.12-rc2 |
2714 2715 |
return ret; } |
8174202b3 write_iter varian... |
2716 |
EXPORT_SYMBOL(generic_file_write_iter); |
1da177e4c Linux-2.6.12-rc2 |
2717 |
|
cf9a2ae8d [PATCH] BLOCK: Mo... |
2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 |
/** * try_to_release_page() - release old fs-specific metadata on a page * * @page: the page which the kernel is trying to free * @gfp_mask: memory allocation flags (and I/O mode) * * The address_space is to try to release any data against the page * (presumably at page->private). If the release was successful, return `1'. * Otherwise return zero. * |
266cf658e FS-Cache: Recruit... |
2728 2729 2730 |
* This may also be called if PG_fscache is set on a page, indicating that the * page is known to the local caching routines. * |
cf9a2ae8d [PATCH] BLOCK: Mo... |
2731 |
* The @gfp_mask argument specifies whether I/O may be performed to release |
71baba4b9 mm, page_alloc: r... |
2732 |
* this page (__GFP_IO), and whether the call may block (__GFP_RECLAIM & __GFP_FS). |
cf9a2ae8d [PATCH] BLOCK: Mo... |
2733 |
* |
cf9a2ae8d [PATCH] BLOCK: Mo... |
2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 |
*/ int try_to_release_page(struct page *page, gfp_t gfp_mask) { struct address_space * const mapping = page->mapping; BUG_ON(!PageLocked(page)); if (PageWriteback(page)) return 0; if (mapping && mapping->a_ops->releasepage) return mapping->a_ops->releasepage(page, gfp_mask); return try_to_free_buffers(page); } EXPORT_SYMBOL(try_to_release_page); |