Blame view
mm/filemap.c
77.6 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 |
* ->private_lock (try_to_unmap_one) * ->tree_lock (try_to_unmap_one) |
a52633d8e mm, vmscan: move ... |
96 97 |
* ->zone_lru_lock(zone) (follow_page->mark_page_accessed) * ->zone_lru_lock(zone) (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 |
*/ |
22f2ac51b mm: workingset: f... |
110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 |
static int page_cache_tree_insert(struct address_space *mapping, struct page *page, void **shadowp) { struct radix_tree_node *node; void **slot; int error; error = __radix_tree_create(&mapping->page_tree, page->index, 0, &node, &slot); if (error) return error; if (*slot) { void *p; p = radix_tree_deref_slot_protected(slot, &mapping->tree_lock); if (!radix_tree_exceptional_entry(p)) return -EEXIST; mapping->nrexceptional--; 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); /* Wakeup waiters for exceptional entry lock */ dax_wake_mapping_entry_waiter(mapping, page->index, |
87fa6f37f dax: fix deadlock... |
144 |
true); |
22f2ac51b mm: workingset: f... |
145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 |
} } 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; } |
91b0abe36 mm + fs: store sh... |
165 166 167 |
static void page_cache_tree_delete(struct address_space *mapping, struct page *page, void *shadow) { |
83929372f filemap: prepare ... |
168 |
int i, nr = PageHuge(page) ? 1 : hpage_nr_pages(page); |
91b0abe36 mm + fs: store sh... |
169 |
|
83929372f filemap: prepare ... |
170 171 172 |
VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(PageTail(page), page); VM_BUG_ON_PAGE(nr != 1 && shadow, page); |
449dd6984 mm: keep page cac... |
173 |
|
83929372f filemap: prepare ... |
174 |
for (i = 0; i < nr; i++) { |
d3798ae8c mm: filemap: don'... |
175 176 177 178 179 180 181 |
struct radix_tree_node *node; void **slot; __radix_tree_lookup(&mapping->page_tree, page->index + i, &node, &slot); radix_tree_clear_tags(&mapping->page_tree, node, slot); |
83929372f filemap: prepare ... |
182 183 |
if (!node) { VM_BUG_ON_PAGE(nr != 1, page); |
d3798ae8c mm: filemap: don'... |
184 185 186 187 188 |
/* * We need a node to properly account shadow * entries. Don't plant any without. XXX */ shadow = NULL; |
83929372f filemap: prepare ... |
189 |
} |
449dd6984 mm: keep page cac... |
190 |
|
d3798ae8c mm: filemap: don'... |
191 192 193 194 |
radix_tree_replace_slot(slot, shadow); if (!node) break; |
83929372f filemap: prepare ... |
195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 |
workingset_node_pages_dec(node); if (shadow) workingset_node_shadows_inc(node); else if (__radix_tree_delete_node(&mapping->page_tree, node)) continue; /* * Track node that only contains shadow entries. DAX mappings * contain no shadow entries and may contain other exceptional * entries so skip those. * * 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. */ if (!dax_mapping(mapping) && !workingset_node_pages(node) && list_empty(&node->private_list)) { node->private_data = mapping; list_lru_add(&workingset_shadow_nodes, &node->private_list); } |
449dd6984 mm: keep page cac... |
217 |
} |
d3798ae8c mm: filemap: don'... |
218 219 220 221 222 223 224 225 226 227 228 229 |
if (shadow) { mapping->nrexceptional += nr; /* * Make sure the nrexceptional update is committed before * 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(); } mapping->nrpages -= nr; |
91b0abe36 mm + fs: store sh... |
230 |
} |
1da177e4c Linux-2.6.12-rc2 |
231 |
/* |
e64a782fe mm: change __remo... |
232 |
* Delete a page from the page cache and free it. Caller has to make |
1da177e4c Linux-2.6.12-rc2 |
233 |
* sure the page is locked and that nobody else uses it - or that usage |
fdf1cdb91 mm: remove unnece... |
234 |
* is safe. The caller must hold the mapping's tree_lock. |
1da177e4c Linux-2.6.12-rc2 |
235 |
*/ |
62cccb8c8 mm: simplify lock... |
236 |
void __delete_from_page_cache(struct page *page, void *shadow) |
1da177e4c Linux-2.6.12-rc2 |
237 238 |
{ struct address_space *mapping = page->mapping; |
83929372f filemap: prepare ... |
239 |
int nr = hpage_nr_pages(page); |
1da177e4c Linux-2.6.12-rc2 |
240 |
|
fe0bfaaff mm: trace filemap... |
241 |
trace_mm_filemap_delete_from_page_cache(page); |
c515e1fd3 mm/fs: add hooks ... |
242 243 244 245 246 247 248 249 |
/* * 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... |
250 |
cleancache_invalidate_page(mapping, page); |
c515e1fd3 mm/fs: add hooks ... |
251 |
|
83929372f filemap: prepare ... |
252 |
VM_BUG_ON_PAGE(PageTail(page), page); |
06b241f32 mm: __delete_from... |
253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 |
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 ... |
274 |
page_ref_sub(page, mapcount); |
06b241f32 mm: __delete_from... |
275 276 |
} } |
91b0abe36 mm + fs: store sh... |
277 |
page_cache_tree_delete(mapping, page, shadow); |
1da177e4c Linux-2.6.12-rc2 |
278 |
page->mapping = NULL; |
b85e0effd mm: consistent tr... |
279 |
/* Leave page->index set: truncation lookup relies upon it */ |
91b0abe36 mm + fs: store sh... |
280 |
|
4165b9b46 hugetlb: do not a... |
281 282 |
/* hugetlb pages do not participate in page cache accounting. */ if (!PageHuge(page)) |
11fb99898 mm: move most fil... |
283 |
__mod_node_page_state(page_pgdat(page), NR_FILE_PAGES, -nr); |
800d8c63b shmem: add huge p... |
284 |
if (PageSwapBacked(page)) { |
11fb99898 mm: move most fil... |
285 |
__mod_node_page_state(page_pgdat(page), NR_SHMEM, -nr); |
800d8c63b shmem: add huge p... |
286 |
if (PageTransHuge(page)) |
11fb99898 mm: move most fil... |
287 |
__dec_node_page_state(page, NR_SHMEM_THPS); |
800d8c63b shmem: add huge p... |
288 289 290 |
} else { VM_BUG_ON_PAGE(PageTransHuge(page) && !PageHuge(page), page); } |
3a6927906 Do dirty page acc... |
291 292 |
/* |
b9ea25152 page_writeback: c... |
293 294 |
* 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... |
295 |
* |
b9ea25152 page_writeback: c... |
296 297 298 |
* 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... |
299 |
*/ |
b9ea25152 page_writeback: c... |
300 |
if (WARN_ON_ONCE(PageDirty(page))) |
62cccb8c8 mm: simplify lock... |
301 |
account_page_cleaned(page, mapping, inode_to_wb(mapping->host)); |
1da177e4c Linux-2.6.12-rc2 |
302 |
} |
702cfbf93 mm: goodbye remov... |
303 304 305 306 307 308 309 310 311 |
/** * 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 |
312 |
{ |
83929372f filemap: prepare ... |
313 |
struct address_space *mapping = page_mapping(page); |
c4843a759 memcg: add per cg... |
314 |
unsigned long flags; |
6072d13c4 Call the filesyst... |
315 |
void (*freepage)(struct page *); |
1da177e4c Linux-2.6.12-rc2 |
316 |
|
cd7619d6b [PATCH] Extermina... |
317 |
BUG_ON(!PageLocked(page)); |
1da177e4c Linux-2.6.12-rc2 |
318 |
|
6072d13c4 Call the filesyst... |
319 |
freepage = mapping->a_ops->freepage; |
c4843a759 memcg: add per cg... |
320 |
|
c4843a759 memcg: add per cg... |
321 |
spin_lock_irqsave(&mapping->tree_lock, flags); |
62cccb8c8 mm: simplify lock... |
322 |
__delete_from_page_cache(page, NULL); |
c4843a759 memcg: add per cg... |
323 |
spin_unlock_irqrestore(&mapping->tree_lock, flags); |
6072d13c4 Call the filesyst... |
324 325 326 |
if (freepage) freepage(page); |
83929372f filemap: prepare ... |
327 328 329 330 331 332 333 |
if (PageTransHuge(page) && !PageHuge(page)) { page_ref_sub(page, HPAGE_PMD_NR); VM_BUG_ON_PAGE(page_count(page) <= 0, page); } else { put_page(page); } |
97cecb5a2 mm: introduce del... |
334 335 |
} EXPORT_SYMBOL(delete_from_page_cache); |
d72d9e2a5 mm: export filema... |
336 |
int filemap_check_errors(struct address_space *mapping) |
865ffef37 fs: fix fsync() e... |
337 338 339 |
{ int ret = 0; /* Check for outstanding write errors */ |
7fcbbaf18 mm/filemap.c: avo... |
340 341 |
if (test_bit(AS_ENOSPC, &mapping->flags) && test_and_clear_bit(AS_ENOSPC, &mapping->flags)) |
865ffef37 fs: fix fsync() e... |
342 |
ret = -ENOSPC; |
7fcbbaf18 mm/filemap.c: avo... |
343 344 |
if (test_bit(AS_EIO, &mapping->flags) && test_and_clear_bit(AS_EIO, &mapping->flags)) |
865ffef37 fs: fix fsync() e... |
345 346 347 |
ret = -EIO; return ret; } |
d72d9e2a5 mm: export filema... |
348 |
EXPORT_SYMBOL(filemap_check_errors); |
865ffef37 fs: fix fsync() e... |
349 |
|
1da177e4c Linux-2.6.12-rc2 |
350 |
/** |
485bb99b4 [PATCH] kernel-do... |
351 |
* __filemap_fdatawrite_range - start writeback on mapping dirty pages in range |
67be2dd1b [PATCH] DocBook: ... |
352 353 |
* @mapping: address space structure to write * @start: offset in bytes where the range starts |
469eb4d03 [PATCH] filemap_f... |
354 |
* @end: offset in bytes where the range ends (inclusive) |
67be2dd1b [PATCH] DocBook: ... |
355 |
* @sync_mode: enable synchronous operation |
1da177e4c Linux-2.6.12-rc2 |
356 |
* |
485bb99b4 [PATCH] kernel-do... |
357 358 359 |
* 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 |
360 |
* If sync_mode is WB_SYNC_ALL then this is a "data integrity" operation, as |
485bb99b4 [PATCH] kernel-do... |
361 |
* opposed to a regular memory cleansing writeback. The difference between |
1da177e4c Linux-2.6.12-rc2 |
362 363 364 |
* 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()... |
365 366 |
int __filemap_fdatawrite_range(struct address_space *mapping, loff_t start, loff_t end, int sync_mode) |
1da177e4c Linux-2.6.12-rc2 |
367 368 369 370 |
{ int ret; struct writeback_control wbc = { .sync_mode = sync_mode, |
05fe478dd mm: write_cache_p... |
371 |
.nr_to_write = LONG_MAX, |
111ebb6e6 [PATCH] writeback... |
372 373 |
.range_start = start, .range_end = end, |
1da177e4c Linux-2.6.12-rc2 |
374 375 376 377 |
}; if (!mapping_cap_writeback_dirty(mapping)) return 0; |
b16b1deb5 writeback: make w... |
378 |
wbc_attach_fdatawrite_inode(&wbc, mapping->host); |
1da177e4c Linux-2.6.12-rc2 |
379 |
ret = do_writepages(mapping, &wbc); |
b16b1deb5 writeback: make w... |
380 |
wbc_detach_inode(&wbc); |
1da177e4c Linux-2.6.12-rc2 |
381 382 383 384 385 386 |
return ret; } static inline int __filemap_fdatawrite(struct address_space *mapping, int sync_mode) { |
111ebb6e6 [PATCH] writeback... |
387 |
return __filemap_fdatawrite_range(mapping, 0, LLONG_MAX, sync_mode); |
1da177e4c Linux-2.6.12-rc2 |
388 389 390 391 392 393 394 |
} int filemap_fdatawrite(struct address_space *mapping) { return __filemap_fdatawrite(mapping, WB_SYNC_ALL); } EXPORT_SYMBOL(filemap_fdatawrite); |
f4c0a0fdf vfs: export filem... |
395 |
int filemap_fdatawrite_range(struct address_space *mapping, loff_t start, |
ebcf28e1c [PATCH] fadvise()... |
396 |
loff_t end) |
1da177e4c Linux-2.6.12-rc2 |
397 398 399 |
{ return __filemap_fdatawrite_range(mapping, start, end, WB_SYNC_ALL); } |
f4c0a0fdf vfs: export filem... |
400 |
EXPORT_SYMBOL(filemap_fdatawrite_range); |
1da177e4c Linux-2.6.12-rc2 |
401 |
|
485bb99b4 [PATCH] kernel-do... |
402 403 404 405 |
/** * filemap_flush - mostly a non-blocking flush * @mapping: target address_space * |
1da177e4c Linux-2.6.12-rc2 |
406 407 408 409 410 411 412 413 |
* 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... |
414 415 |
static int __filemap_fdatawait_range(struct address_space *mapping, loff_t start_byte, loff_t end_byte) |
1da177e4c Linux-2.6.12-rc2 |
416 |
{ |
09cbfeaf1 mm, fs: get rid o... |
417 418 |
pgoff_t index = start_byte >> PAGE_SHIFT; pgoff_t end = end_byte >> PAGE_SHIFT; |
1da177e4c Linux-2.6.12-rc2 |
419 420 |
struct pagevec pvec; int nr_pages; |
aa750fd71 mm/filemap.c: mak... |
421 |
int ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
422 |
|
94004ed72 kill wait_on_page... |
423 |
if (end_byte < start_byte) |
865ffef37 fs: fix fsync() e... |
424 |
goto out; |
1da177e4c Linux-2.6.12-rc2 |
425 426 |
pagevec_init(&pvec, 0); |
1da177e4c Linux-2.6.12-rc2 |
427 428 429 430 431 432 433 434 435 436 437 438 439 440 |
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... |
441 |
if (TestClearPageError(page)) |
1da177e4c Linux-2.6.12-rc2 |
442 443 444 445 446 |
ret = -EIO; } pagevec_release(&pvec); cond_resched(); } |
865ffef37 fs: fix fsync() e... |
447 |
out: |
aa750fd71 mm/filemap.c: mak... |
448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 |
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... |
471 472 473 |
ret2 = filemap_check_errors(mapping); if (!ret) ret = ret2; |
1da177e4c Linux-2.6.12-rc2 |
474 475 476 |
return ret; } |
d3bccb6f4 vfs: Introduce fi... |
477 478 479 |
EXPORT_SYMBOL(filemap_fdatawait_range); /** |
aa750fd71 mm/filemap.c: mak... |
480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 |
* 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... |
502 |
* filemap_fdatawait - wait for all under-writeback pages to complete |
1da177e4c Linux-2.6.12-rc2 |
503 |
* @mapping: address space structure to wait for |
485bb99b4 [PATCH] kernel-do... |
504 505 |
* * Walk the list of under-writeback pages of the given address space |
aa750fd71 mm/filemap.c: mak... |
506 507 508 509 510 511 |
* 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 |
512 513 514 515 516 517 518 |
*/ 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... |
519 |
return filemap_fdatawait_range(mapping, 0, i_size - 1); |
1da177e4c Linux-2.6.12-rc2 |
520 521 522 523 524 |
} EXPORT_SYMBOL(filemap_fdatawait); int filemap_write_and_wait(struct address_space *mapping) { |
28fd12982 [PATCH] Fix and a... |
525 |
int err = 0; |
1da177e4c Linux-2.6.12-rc2 |
526 |
|
7f6d5b529 dax: move writeba... |
527 528 |
if ((!dax_mapping(mapping) && mapping->nrpages) || (dax_mapping(mapping) && mapping->nrexceptional)) { |
28fd12982 [PATCH] Fix and a... |
529 530 531 532 533 534 535 536 537 538 539 540 |
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... |
541 542 |
} else { err = filemap_check_errors(mapping); |
1da177e4c Linux-2.6.12-rc2 |
543 |
} |
28fd12982 [PATCH] Fix and a... |
544 |
return err; |
1da177e4c Linux-2.6.12-rc2 |
545 |
} |
28fd12982 [PATCH] Fix and a... |
546 |
EXPORT_SYMBOL(filemap_write_and_wait); |
1da177e4c Linux-2.6.12-rc2 |
547 |
|
485bb99b4 [PATCH] kernel-do... |
548 549 550 551 552 553 |
/** * 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... |
554 555 556 557 558 |
* 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 |
559 560 561 |
int filemap_write_and_wait_range(struct address_space *mapping, loff_t lstart, loff_t lend) { |
28fd12982 [PATCH] Fix and a... |
562 |
int err = 0; |
1da177e4c Linux-2.6.12-rc2 |
563 |
|
7f6d5b529 dax: move writeba... |
564 565 |
if ((!dax_mapping(mapping) && mapping->nrpages) || (dax_mapping(mapping) && mapping->nrexceptional)) { |
28fd12982 [PATCH] Fix and a... |
566 567 568 569 |
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... |
570 571 |
int err2 = filemap_fdatawait_range(mapping, lstart, lend); |
28fd12982 [PATCH] Fix and a... |
572 573 574 |
if (!err) err = err2; } |
865ffef37 fs: fix fsync() e... |
575 576 |
} else { err = filemap_check_errors(mapping); |
1da177e4c Linux-2.6.12-rc2 |
577 |
} |
28fd12982 [PATCH] Fix and a... |
578 |
return err; |
1da177e4c Linux-2.6.12-rc2 |
579 |
} |
f69955855 Export filemap_wr... |
580 |
EXPORT_SYMBOL(filemap_write_and_wait_range); |
1da177e4c Linux-2.6.12-rc2 |
581 |
|
485bb99b4 [PATCH] kernel-do... |
582 |
/** |
ef6a3c631 mm: add replace_p... |
583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 |
* 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... |
600 |
|
309381fea mm: dump page whe... |
601 602 603 |
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... |
604 |
|
ef6a3c631 mm: add replace_p... |
605 606 607 608 |
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... |
609 |
unsigned long flags; |
ef6a3c631 mm: add replace_p... |
610 611 612 |
pgoff_t offset = old->index; freepage = mapping->a_ops->freepage; |
09cbfeaf1 mm, fs: get rid o... |
613 |
get_page(new); |
ef6a3c631 mm: add replace_p... |
614 615 |
new->mapping = mapping; new->index = offset; |
c4843a759 memcg: add per cg... |
616 |
spin_lock_irqsave(&mapping->tree_lock, flags); |
62cccb8c8 mm: simplify lock... |
617 |
__delete_from_page_cache(old, NULL); |
22f2ac51b mm: workingset: f... |
618 |
error = page_cache_tree_insert(mapping, new, NULL); |
ef6a3c631 mm: add replace_p... |
619 |
BUG_ON(error); |
4165b9b46 hugetlb: do not a... |
620 621 622 623 624 |
/* * hugetlb pages do not participate in page cache accounting. */ if (!PageHuge(new)) |
11fb99898 mm: move most fil... |
625 |
__inc_node_page_state(new, NR_FILE_PAGES); |
ef6a3c631 mm: add replace_p... |
626 |
if (PageSwapBacked(new)) |
11fb99898 mm: move most fil... |
627 |
__inc_node_page_state(new, NR_SHMEM); |
c4843a759 memcg: add per cg... |
628 |
spin_unlock_irqrestore(&mapping->tree_lock, flags); |
6a93ca8fd mm: migrate: do n... |
629 |
mem_cgroup_migrate(old, new); |
ef6a3c631 mm: add replace_p... |
630 631 632 |
radix_tree_preload_end(); if (freepage) freepage(old); |
09cbfeaf1 mm, fs: get rid o... |
633 |
put_page(old); |
ef6a3c631 mm: add replace_p... |
634 635 636 637 638 |
} return error; } EXPORT_SYMBOL_GPL(replace_page_cache_page); |
a528910e1 mm: thrash detect... |
639 640 641 642 |
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 |
643 |
{ |
00501b531 mm: memcontrol: r... |
644 645 |
int huge = PageHuge(page); struct mem_cgroup *memcg; |
e286781d5 mm: speculative p... |
646 |
int error; |
309381fea mm: dump page whe... |
647 648 |
VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(PageSwapBacked(page), page); |
e286781d5 mm: speculative p... |
649 |
|
00501b531 mm: memcontrol: r... |
650 651 |
if (!huge) { error = mem_cgroup_try_charge(page, current->mm, |
f627c2f53 memcg: adjust to ... |
652 |
gfp_mask, &memcg, false); |
00501b531 mm: memcontrol: r... |
653 654 655 |
if (error) return error; } |
1da177e4c Linux-2.6.12-rc2 |
656 |
|
5e4c0d974 lib/radix-tree.c:... |
657 |
error = radix_tree_maybe_preload(gfp_mask & ~__GFP_HIGHMEM); |
66a0c8ee3 mm: cleanup add_t... |
658 |
if (error) { |
00501b531 mm: memcontrol: r... |
659 |
if (!huge) |
f627c2f53 memcg: adjust to ... |
660 |
mem_cgroup_cancel_charge(page, memcg, false); |
66a0c8ee3 mm: cleanup add_t... |
661 662 |
return error; } |
09cbfeaf1 mm, fs: get rid o... |
663 |
get_page(page); |
66a0c8ee3 mm: cleanup add_t... |
664 665 666 667 |
page->mapping = mapping; page->index = offset; spin_lock_irq(&mapping->tree_lock); |
a528910e1 mm: thrash detect... |
668 |
error = page_cache_tree_insert(mapping, page, shadowp); |
66a0c8ee3 mm: cleanup add_t... |
669 670 671 |
radix_tree_preload_end(); if (unlikely(error)) goto err_insert; |
4165b9b46 hugetlb: do not a... |
672 673 674 |
/* hugetlb pages do not participate in page cache accounting. */ if (!huge) |
11fb99898 mm: move most fil... |
675 |
__inc_node_page_state(page, NR_FILE_PAGES); |
66a0c8ee3 mm: cleanup add_t... |
676 |
spin_unlock_irq(&mapping->tree_lock); |
00501b531 mm: memcontrol: r... |
677 |
if (!huge) |
f627c2f53 memcg: adjust to ... |
678 |
mem_cgroup_commit_charge(page, memcg, false, false); |
66a0c8ee3 mm: cleanup add_t... |
679 680 681 682 683 684 |
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... |
685 |
if (!huge) |
f627c2f53 memcg: adjust to ... |
686 |
mem_cgroup_cancel_charge(page, memcg, false); |
09cbfeaf1 mm, fs: get rid o... |
687 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
688 689 |
return error; } |
a528910e1 mm: thrash detect... |
690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 |
/** * 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... |
707 |
EXPORT_SYMBOL(add_to_page_cache_locked); |
1da177e4c Linux-2.6.12-rc2 |
708 709 |
int add_to_page_cache_lru(struct page *page, struct address_space *mapping, |
6daa0e286 [PATCH] gfp_t: mm... |
710 |
pgoff_t offset, gfp_t gfp_mask) |
1da177e4c Linux-2.6.12-rc2 |
711 |
{ |
a528910e1 mm: thrash detect... |
712 |
void *shadow = NULL; |
4f98a2fee vmscan: split LRU... |
713 |
int ret; |
48c935ad8 page-flags: defin... |
714 |
__SetPageLocked(page); |
a528910e1 mm: thrash detect... |
715 716 717 |
ret = __add_to_page_cache_locked(page, mapping, offset, gfp_mask, &shadow); if (unlikely(ret)) |
48c935ad8 page-flags: defin... |
718 |
__ClearPageLocked(page); |
a528910e1 mm: thrash detect... |
719 720 721 722 723 |
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... |
724 725 726 |
* 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... |
727 |
*/ |
f0281a00f mm: workingset: o... |
728 729 |
if (!(gfp_mask & __GFP_WRITE) && shadow && workingset_refault(shadow)) { |
a528910e1 mm: thrash detect... |
730 731 732 733 734 735 |
SetPageActive(page); workingset_activation(page); } else ClearPageActive(page); lru_cache_add(page); } |
1da177e4c Linux-2.6.12-rc2 |
736 737 |
return ret; } |
18bc0bbd1 Staging: pohmelfs... |
738 |
EXPORT_SYMBOL_GPL(add_to_page_cache_lru); |
1da177e4c Linux-2.6.12-rc2 |
739 |
|
44110fe38 [PATCH] cpuset me... |
740 |
#ifdef CONFIG_NUMA |
2ae88149a [PATCH] mm: clean... |
741 |
struct page *__page_cache_alloc(gfp_t gfp) |
44110fe38 [PATCH] cpuset me... |
742 |
{ |
c0ff7453b cpuset,mm: fix no... |
743 744 |
int n; struct page *page; |
44110fe38 [PATCH] cpuset me... |
745 |
if (cpuset_do_page_mem_spread()) { |
cc9a6c877 cpuset: mm: reduc... |
746 747 |
unsigned int cpuset_mems_cookie; do { |
d26914d11 mm: optimize put_... |
748 |
cpuset_mems_cookie = read_mems_allowed_begin(); |
cc9a6c877 cpuset: mm: reduc... |
749 |
n = cpuset_mem_spread_node(); |
96db800f5 mm: rename alloc_... |
750 |
page = __alloc_pages_node(n, gfp, 0); |
d26914d11 mm: optimize put_... |
751 |
} while (!page && read_mems_allowed_retry(cpuset_mems_cookie)); |
cc9a6c877 cpuset: mm: reduc... |
752 |
|
c0ff7453b cpuset,mm: fix no... |
753 |
return page; |
44110fe38 [PATCH] cpuset me... |
754 |
} |
2ae88149a [PATCH] mm: clean... |
755 |
return alloc_pages(gfp, 0); |
44110fe38 [PATCH] cpuset me... |
756 |
} |
2ae88149a [PATCH] mm: clean... |
757 |
EXPORT_SYMBOL(__page_cache_alloc); |
44110fe38 [PATCH] cpuset me... |
758 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
759 760 761 762 763 764 765 766 767 768 |
/* * 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... |
769 |
wait_queue_head_t *page_waitqueue(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
770 |
{ |
9dcb8b685 mm: remove per-zo... |
771 |
return bit_waitqueue(page, 0); |
1da177e4c Linux-2.6.12-rc2 |
772 |
} |
a4796e37c MM: export page_w... |
773 |
EXPORT_SYMBOL(page_waitqueue); |
1da177e4c Linux-2.6.12-rc2 |
774 |
|
920c7a5d0 mm: remove fastca... |
775 |
void wait_on_page_bit(struct page *page, int bit_nr) |
1da177e4c Linux-2.6.12-rc2 |
776 777 778 779 |
{ DEFINE_WAIT_BIT(wait, &page->flags, bit_nr); if (test_bit(bit_nr, &page->flags)) |
743162013 sched: Remove pro... |
780 |
__wait_on_bit(page_waitqueue(page), &wait, bit_wait_io, |
1da177e4c Linux-2.6.12-rc2 |
781 782 783 |
TASK_UNINTERRUPTIBLE); } EXPORT_SYMBOL(wait_on_page_bit); |
f62e00cc3 mm: introduce wai... |
784 785 786 787 788 789 790 791 |
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... |
792 |
bit_wait_io, TASK_KILLABLE); |
f62e00cc3 mm: introduce wai... |
793 |
} |
cbbce8220 SCHED: add some "... |
794 795 796 797 798 799 800 801 802 803 804 805 |
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 |
806 |
/** |
385e1ca5f CacheFiles: Permi... |
807 |
* add_page_wait_queue - Add an arbitrary waiter to a page's wait queue |
697f619fc filemap: fix kern... |
808 809 |
* @page: Page defining the wait queue of interest * @waiter: Waiter to add to the queue |
385e1ca5f CacheFiles: Permi... |
810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 |
* * 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... |
825 |
* unlock_page - unlock a locked page |
1da177e4c Linux-2.6.12-rc2 |
826 827 828 829 |
* @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... |
830 |
* mechanism between PageLocked pages and PageWriteback pages is shared. |
1da177e4c Linux-2.6.12-rc2 |
831 832 |
* But that's OK - sleepers in wait_on_page_writeback() just go back to sleep. * |
8413ac9d8 mm: page lock use... |
833 834 |
* 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 |
835 |
*/ |
920c7a5d0 mm: remove fastca... |
836 |
void unlock_page(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
837 |
{ |
48c935ad8 page-flags: defin... |
838 |
page = compound_head(page); |
309381fea mm: dump page whe... |
839 |
VM_BUG_ON_PAGE(!PageLocked(page), page); |
8413ac9d8 mm: page lock use... |
840 |
clear_bit_unlock(PG_locked, &page->flags); |
4e857c58e arch: Mass conver... |
841 |
smp_mb__after_atomic(); |
1da177e4c Linux-2.6.12-rc2 |
842 843 844 |
wake_up_page(page, PG_locked); } EXPORT_SYMBOL(unlock_page); |
485bb99b4 [PATCH] kernel-do... |
845 846 847 |
/** * end_page_writeback - end writeback against a page * @page: the page |
1da177e4c Linux-2.6.12-rc2 |
848 849 850 |
*/ void end_page_writeback(struct page *page) { |
888cf2db4 mm: avoid unneces... |
851 852 853 854 855 856 857 858 859 |
/* * 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... |
860 |
rotate_reclaimable_page(page); |
888cf2db4 mm: avoid unneces... |
861 |
} |
ac6aadb24 mm: rotate_reclai... |
862 863 864 |
if (!test_clear_page_writeback(page)) BUG(); |
4e857c58e arch: Mass conver... |
865 |
smp_mb__after_atomic(); |
1da177e4c Linux-2.6.12-rc2 |
866 867 868 |
wake_up_page(page, PG_writeback); } EXPORT_SYMBOL(end_page_writeback); |
57d998456 fs/mpage.c: facto... |
869 870 871 872 |
/* * After completing I/O on a page, call this routine to update the page * flags appropriately */ |
c11f0c0b5 block/mm: make bd... |
873 |
void page_endio(struct page *page, bool is_write, int err) |
57d998456 fs/mpage.c: facto... |
874 |
{ |
c11f0c0b5 block/mm: make bd... |
875 |
if (!is_write) { |
57d998456 fs/mpage.c: facto... |
876 877 878 879 880 881 882 |
if (!err) { SetPageUptodate(page); } else { ClearPageUptodate(page); SetPageError(page); } unlock_page(page); |
abf545484 mm/block: convert... |
883 |
} else { |
57d998456 fs/mpage.c: facto... |
884 885 886 887 888 889 890 891 892 |
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... |
893 894 895 |
/** * __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 |
896 |
*/ |
920c7a5d0 mm: remove fastca... |
897 |
void __lock_page(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
898 |
{ |
48c935ad8 page-flags: defin... |
899 900 |
struct page *page_head = compound_head(page); DEFINE_WAIT_BIT(wait, &page_head->flags, PG_locked); |
1da177e4c Linux-2.6.12-rc2 |
901 |
|
48c935ad8 page-flags: defin... |
902 |
__wait_on_bit_lock(page_waitqueue(page_head), &wait, bit_wait_io, |
1da177e4c Linux-2.6.12-rc2 |
903 904 905 |
TASK_UNINTERRUPTIBLE); } EXPORT_SYMBOL(__lock_page); |
b5606c2d4 remove final fast... |
906 |
int __lock_page_killable(struct page *page) |
2687a3569 Add lock_page_kil... |
907 |
{ |
48c935ad8 page-flags: defin... |
908 909 |
struct page *page_head = compound_head(page); DEFINE_WAIT_BIT(wait, &page_head->flags, PG_locked); |
2687a3569 Add lock_page_kil... |
910 |
|
48c935ad8 page-flags: defin... |
911 |
return __wait_on_bit_lock(page_waitqueue(page_head), &wait, |
743162013 sched: Remove pro... |
912 |
bit_wait_io, TASK_KILLABLE); |
2687a3569 Add lock_page_kil... |
913 |
} |
18bc0bbd1 Staging: pohmelfs... |
914 |
EXPORT_SYMBOL_GPL(__lock_page_killable); |
2687a3569 Add lock_page_kil... |
915 |
|
9a95f3cf7 mm: describe mmap... |
916 917 918 919 920 921 922 923 924 925 926 |
/* * 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... |
927 928 929 |
int __lock_page_or_retry(struct page *page, struct mm_struct *mm, unsigned int flags) { |
37b23e052 x86,mm: make page... |
930 931 932 933 934 935 936 937 938 939 940 941 |
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 ... |
942 |
wait_on_page_locked(page); |
d065bd810 mm: retry page fa... |
943 |
return 0; |
37b23e052 x86,mm: make page... |
944 945 946 947 948 949 950 951 952 953 954 955 |
} 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... |
956 957 |
} } |
485bb99b4 [PATCH] kernel-do... |
958 |
/** |
e7b563bb2 mm: filemap: move... |
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 |
* 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 ... |
985 986 987 988 |
struct page *page; page = radix_tree_lookup(&mapping->page_tree, index); if (!page || radix_tree_exceptional_entry(page)) |
e7b563bb2 mm: filemap: move... |
989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 |
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 ... |
1026 1027 1028 1029 |
struct page *page; page = radix_tree_lookup(&mapping->page_tree, index); if (!page || radix_tree_exceptional_entry(page)) |
e7b563bb2 mm: filemap: move... |
1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 |
break; index--; if (index == ULONG_MAX) break; } return index; } EXPORT_SYMBOL(page_cache_prev_hole); /** |
0cd6144aa mm + fs: prepare ... |
1041 |
* find_get_entry - find and get a page cache entry |
485bb99b4 [PATCH] kernel-do... |
1042 |
* @mapping: the address_space to search |
0cd6144aa mm + fs: prepare ... |
1043 1044 1045 1046 |
* @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... |
1047 |
* |
139b6a6fb mm: filemap: upda... |
1048 1049 |
* 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 ... |
1050 1051 |
* * Otherwise, %NULL is returned. |
1da177e4c Linux-2.6.12-rc2 |
1052 |
*/ |
0cd6144aa mm + fs: prepare ... |
1053 |
struct page *find_get_entry(struct address_space *mapping, pgoff_t offset) |
1da177e4c Linux-2.6.12-rc2 |
1054 |
{ |
a60637c85 mm: lockless page... |
1055 |
void **pagep; |
83929372f filemap: prepare ... |
1056 |
struct page *head, *page; |
1da177e4c Linux-2.6.12-rc2 |
1057 |
|
a60637c85 mm: lockless page... |
1058 1059 1060 1061 1062 1063 |
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... |
1064 1065 |
if (unlikely(!page)) goto out; |
a2c16d6cb mm: let swap use ... |
1066 |
if (radix_tree_exception(page)) { |
8079b1c85 mm: clarify the r... |
1067 1068 1069 |
if (radix_tree_deref_retry(page)) goto repeat; /* |
139b6a6fb mm: filemap: upda... |
1070 1071 1072 |
* 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... |
1073 1074 |
*/ goto out; |
a2c16d6cb mm: let swap use ... |
1075 |
} |
83929372f filemap: prepare ... |
1076 1077 1078 1079 1080 1081 1082 1083 |
head = compound_head(page); if (!page_cache_get_speculative(head)) goto repeat; /* The page was split under us? */ if (compound_head(page) != head) { put_page(head); |
a60637c85 mm: lockless page... |
1084 |
goto repeat; |
83929372f filemap: prepare ... |
1085 |
} |
a60637c85 mm: lockless page... |
1086 1087 1088 1089 1090 1091 1092 |
/* * Has the page moved? * This is part of the lockless pagecache protocol. See * include/linux/pagemap.h for details. */ if (unlikely(page != *pagep)) { |
83929372f filemap: prepare ... |
1093 |
put_page(head); |
a60637c85 mm: lockless page... |
1094 1095 1096 |
goto repeat; } } |
27d20fddc radix-tree: fix R... |
1097 |
out: |
a60637c85 mm: lockless page... |
1098 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
1099 1100 |
return page; } |
0cd6144aa mm + fs: prepare ... |
1101 |
EXPORT_SYMBOL(find_get_entry); |
1da177e4c Linux-2.6.12-rc2 |
1102 |
|
485bb99b4 [PATCH] kernel-do... |
1103 |
/** |
0cd6144aa mm + fs: prepare ... |
1104 1105 1106 1107 1108 1109 1110 1111 |
* 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... |
1112 1113 |
* 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 ... |
1114 1115 1116 1117 1118 1119 |
* * 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 |
1120 1121 |
{ struct page *page; |
1da177e4c Linux-2.6.12-rc2 |
1122 |
repeat: |
0cd6144aa mm + fs: prepare ... |
1123 |
page = find_get_entry(mapping, offset); |
a2c16d6cb mm: let swap use ... |
1124 |
if (page && !radix_tree_exception(page)) { |
a60637c85 mm: lockless page... |
1125 1126 |
lock_page(page); /* Has the page been truncated? */ |
83929372f filemap: prepare ... |
1127 |
if (unlikely(page_mapping(page) != mapping)) { |
a60637c85 mm: lockless page... |
1128 |
unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
1129 |
put_page(page); |
a60637c85 mm: lockless page... |
1130 |
goto repeat; |
1da177e4c Linux-2.6.12-rc2 |
1131 |
} |
83929372f filemap: prepare ... |
1132 |
VM_BUG_ON_PAGE(page_to_pgoff(page) != offset, page); |
1da177e4c Linux-2.6.12-rc2 |
1133 |
} |
1da177e4c Linux-2.6.12-rc2 |
1134 1135 |
return page; } |
0cd6144aa mm + fs: prepare ... |
1136 1137 1138 |
EXPORT_SYMBOL(find_lock_entry); /** |
2457aec63 mm: non-atomicall... |
1139 |
* pagecache_get_page - find and get a page reference |
0cd6144aa mm + fs: prepare ... |
1140 1141 |
* @mapping: the address_space to search * @offset: the page index |
2457aec63 mm: non-atomicall... |
1142 |
* @fgp_flags: PCG flags |
45f87de57 mm: get rid of ra... |
1143 |
* @gfp_mask: gfp mask to use for the page cache data page allocation |
0cd6144aa mm + fs: prepare ... |
1144 |
* |
2457aec63 mm: non-atomicall... |
1145 |
* Looks up the page cache slot at @mapping & @offset. |
1da177e4c Linux-2.6.12-rc2 |
1146 |
* |
75325189c mm: fix filemap.c... |
1147 |
* PCG flags modify how the page is returned. |
0cd6144aa mm + fs: prepare ... |
1148 |
* |
2457aec63 mm: non-atomicall... |
1149 1150 1151 |
* 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... |
1152 1153 1154 |
* @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 |
1155 |
* |
2457aec63 mm: non-atomicall... |
1156 1157 |
* 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 |
1158 |
* |
2457aec63 mm: non-atomicall... |
1159 |
* If there is a page cache page, it is returned with an increased refcount. |
1da177e4c Linux-2.6.12-rc2 |
1160 |
*/ |
2457aec63 mm: non-atomicall... |
1161 |
struct page *pagecache_get_page(struct address_space *mapping, pgoff_t offset, |
45f87de57 mm: get rid of ra... |
1162 |
int fgp_flags, gfp_t gfp_mask) |
1da177e4c Linux-2.6.12-rc2 |
1163 |
{ |
eb2be1893 mm: buffered writ... |
1164 |
struct page *page; |
2457aec63 mm: non-atomicall... |
1165 |
|
1da177e4c Linux-2.6.12-rc2 |
1166 |
repeat: |
2457aec63 mm: non-atomicall... |
1167 1168 1169 1170 1171 1172 1173 1174 1175 |
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... |
1176 |
put_page(page); |
2457aec63 mm: non-atomicall... |
1177 1178 1179 1180 1181 1182 1183 1184 1185 |
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... |
1186 |
put_page(page); |
2457aec63 mm: non-atomicall... |
1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 |
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... |
1199 1200 1201 |
gfp_mask |= __GFP_WRITE; if (fgp_flags & FGP_NOFS) gfp_mask &= ~__GFP_FS; |
2457aec63 mm: non-atomicall... |
1202 |
|
45f87de57 mm: get rid of ra... |
1203 |
page = __page_cache_alloc(gfp_mask); |
eb2be1893 mm: buffered writ... |
1204 1205 |
if (!page) return NULL; |
2457aec63 mm: non-atomicall... |
1206 1207 1208 |
if (WARN_ON_ONCE(!(fgp_flags & FGP_LOCK))) fgp_flags |= FGP_LOCK; |
eb39d618f mm: replace init_... |
1209 |
/* Init accessed so avoid atomic mark_page_accessed later */ |
2457aec63 mm: non-atomicall... |
1210 |
if (fgp_flags & FGP_ACCESSED) |
eb39d618f mm: replace init_... |
1211 |
__SetPageReferenced(page); |
2457aec63 mm: non-atomicall... |
1212 |
|
45f87de57 mm: get rid of ra... |
1213 1214 |
err = add_to_page_cache_lru(page, mapping, offset, gfp_mask & GFP_RECLAIM_MASK); |
eb2be1893 mm: buffered writ... |
1215 |
if (unlikely(err)) { |
09cbfeaf1 mm, fs: get rid o... |
1216 |
put_page(page); |
eb2be1893 mm: buffered writ... |
1217 1218 1219 |
page = NULL; if (err == -EEXIST) goto repeat; |
1da177e4c Linux-2.6.12-rc2 |
1220 |
} |
1da177e4c Linux-2.6.12-rc2 |
1221 |
} |
2457aec63 mm: non-atomicall... |
1222 |
|
1da177e4c Linux-2.6.12-rc2 |
1223 1224 |
return page; } |
2457aec63 mm: non-atomicall... |
1225 |
EXPORT_SYMBOL(pagecache_get_page); |
1da177e4c Linux-2.6.12-rc2 |
1226 1227 |
/** |
0cd6144aa mm + fs: prepare ... |
1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 |
* 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... |
1244 1245 |
* Any shadow entries of evicted pages, or swap entries from * shmem/tmpfs, are included in the returned array. |
0cd6144aa mm + fs: prepare ... |
1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 |
* * 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 ... |
1262 |
radix_tree_for_each_slot(slot, &mapping->page_tree, &iter, start) { |
83929372f filemap: prepare ... |
1263 |
struct page *head, *page; |
0cd6144aa mm + fs: prepare ... |
1264 1265 1266 1267 1268 |
repeat: page = radix_tree_deref_slot(slot); if (unlikely(!page)) continue; if (radix_tree_exception(page)) { |
2cf938aae mm: use radix_tre... |
1269 1270 1271 1272 |
if (radix_tree_deref_retry(page)) { slot = radix_tree_iter_retry(&iter); continue; } |
0cd6144aa mm + fs: prepare ... |
1273 |
/* |
f9fe48bec dax: support dirt... |
1274 1275 1276 |
* 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 ... |
1277 1278 1279 |
*/ goto export; } |
83929372f filemap: prepare ... |
1280 1281 1282 1283 1284 1285 1286 1287 |
head = compound_head(page); if (!page_cache_get_speculative(head)) goto repeat; /* The page was split under us? */ if (compound_head(page) != head) { put_page(head); |
0cd6144aa mm + fs: prepare ... |
1288 |
goto repeat; |
83929372f filemap: prepare ... |
1289 |
} |
0cd6144aa mm + fs: prepare ... |
1290 1291 1292 |
/* Has the page moved? */ if (unlikely(page != *slot)) { |
83929372f filemap: prepare ... |
1293 |
put_page(head); |
0cd6144aa mm + fs: prepare ... |
1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 |
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 |
1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 |
* 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... |
1325 1326 1327 1328 1329 1330 |
struct radix_tree_iter iter; void **slot; unsigned ret = 0; if (unlikely(!nr_pages)) return 0; |
a60637c85 mm: lockless page... |
1331 1332 |
rcu_read_lock(); |
0fc9d1040 radix-tree: use i... |
1333 |
radix_tree_for_each_slot(slot, &mapping->page_tree, &iter, start) { |
83929372f filemap: prepare ... |
1334 |
struct page *head, *page; |
a60637c85 mm: lockless page... |
1335 |
repeat: |
0fc9d1040 radix-tree: use i... |
1336 |
page = radix_tree_deref_slot(slot); |
a60637c85 mm: lockless page... |
1337 1338 |
if (unlikely(!page)) continue; |
9d8aa4ea8 mm: remove worryi... |
1339 |
|
a2c16d6cb mm: let swap use ... |
1340 |
if (radix_tree_exception(page)) { |
8079b1c85 mm: clarify the r... |
1341 |
if (radix_tree_deref_retry(page)) { |
2cf938aae mm: use radix_tre... |
1342 1343 |
slot = radix_tree_iter_retry(&iter); continue; |
8079b1c85 mm: clarify the r... |
1344 |
} |
a2c16d6cb mm: let swap use ... |
1345 |
/* |
139b6a6fb mm: filemap: upda... |
1346 1347 1348 |
* A shadow entry of a recently evicted page, * or a swap entry from shmem/tmpfs. Skip * over it. |
a2c16d6cb mm: let swap use ... |
1349 |
*/ |
8079b1c85 mm: clarify the r... |
1350 |
continue; |
27d20fddc radix-tree: fix R... |
1351 |
} |
a60637c85 mm: lockless page... |
1352 |
|
83929372f filemap: prepare ... |
1353 1354 1355 1356 1357 1358 1359 |
head = compound_head(page); if (!page_cache_get_speculative(head)) goto repeat; /* The page was split under us? */ if (compound_head(page) != head) { put_page(head); |
a60637c85 mm: lockless page... |
1360 |
goto repeat; |
83929372f filemap: prepare ... |
1361 |
} |
a60637c85 mm: lockless page... |
1362 1363 |
/* Has the page moved? */ |
0fc9d1040 radix-tree: use i... |
1364 |
if (unlikely(page != *slot)) { |
83929372f filemap: prepare ... |
1365 |
put_page(head); |
a60637c85 mm: lockless page... |
1366 1367 |
goto repeat; } |
1da177e4c Linux-2.6.12-rc2 |
1368 |
|
a60637c85 mm: lockless page... |
1369 |
pages[ret] = page; |
0fc9d1040 radix-tree: use i... |
1370 1371 |
if (++ret == nr_pages) break; |
a60637c85 mm: lockless page... |
1372 |
} |
5b280c0cc mm: don't return ... |
1373 |
|
a60637c85 mm: lockless page... |
1374 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
1375 1376 |
return ret; } |
ebf43500e [PATCH] Add find_... |
1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 |
/** * 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... |
1392 1393 1394 1395 1396 1397 |
struct radix_tree_iter iter; void **slot; unsigned int ret = 0; if (unlikely(!nr_pages)) return 0; |
a60637c85 mm: lockless page... |
1398 1399 |
rcu_read_lock(); |
0fc9d1040 radix-tree: use i... |
1400 |
radix_tree_for_each_contig(slot, &mapping->page_tree, &iter, index) { |
83929372f filemap: prepare ... |
1401 |
struct page *head, *page; |
a60637c85 mm: lockless page... |
1402 |
repeat: |
0fc9d1040 radix-tree: use i... |
1403 1404 |
page = radix_tree_deref_slot(slot); /* The hole, there no reason to continue */ |
a60637c85 mm: lockless page... |
1405 |
if (unlikely(!page)) |
0fc9d1040 radix-tree: use i... |
1406 |
break; |
9d8aa4ea8 mm: remove worryi... |
1407 |
|
a2c16d6cb mm: let swap use ... |
1408 |
if (radix_tree_exception(page)) { |
8079b1c85 mm: clarify the r... |
1409 |
if (radix_tree_deref_retry(page)) { |
2cf938aae mm: use radix_tre... |
1410 1411 |
slot = radix_tree_iter_retry(&iter); continue; |
8079b1c85 mm: clarify the r... |
1412 |
} |
a2c16d6cb mm: let swap use ... |
1413 |
/* |
139b6a6fb mm: filemap: upda... |
1414 1415 1416 |
* 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 ... |
1417 |
*/ |
8079b1c85 mm: clarify the r... |
1418 |
break; |
a2c16d6cb mm: let swap use ... |
1419 |
} |
ebf43500e [PATCH] Add find_... |
1420 |
|
83929372f filemap: prepare ... |
1421 1422 1423 1424 1425 1426 1427 |
head = compound_head(page); if (!page_cache_get_speculative(head)) goto repeat; /* The page was split under us? */ if (compound_head(page) != head) { put_page(head); |
a60637c85 mm: lockless page... |
1428 |
goto repeat; |
83929372f filemap: prepare ... |
1429 |
} |
a60637c85 mm: lockless page... |
1430 1431 |
/* Has the page moved? */ |
0fc9d1040 radix-tree: use i... |
1432 |
if (unlikely(page != *slot)) { |
83929372f filemap: prepare ... |
1433 |
put_page(head); |
a60637c85 mm: lockless page... |
1434 1435 |
goto repeat; } |
9cbb4cb21 mm: find_get_page... |
1436 1437 1438 1439 1440 |
/* * 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. */ |
83929372f filemap: prepare ... |
1441 |
if (page->mapping == NULL || page_to_pgoff(page) != iter.index) { |
09cbfeaf1 mm, fs: get rid o... |
1442 |
put_page(page); |
9cbb4cb21 mm: find_get_page... |
1443 1444 |
break; } |
a60637c85 mm: lockless page... |
1445 |
pages[ret] = page; |
0fc9d1040 radix-tree: use i... |
1446 1447 |
if (++ret == nr_pages) break; |
ebf43500e [PATCH] Add find_... |
1448 |
} |
a60637c85 mm: lockless page... |
1449 1450 |
rcu_read_unlock(); return ret; |
ebf43500e [PATCH] Add find_... |
1451 |
} |
ef71c15c4 AFS: export a cou... |
1452 |
EXPORT_SYMBOL(find_get_pages_contig); |
ebf43500e [PATCH] Add find_... |
1453 |
|
485bb99b4 [PATCH] kernel-do... |
1454 1455 1456 1457 1458 1459 1460 1461 |
/** * 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 |
1462 |
* Like find_get_pages, except we only return pages which are tagged with |
485bb99b4 [PATCH] kernel-do... |
1463 |
* @tag. We update @index to index the next page for the traversal. |
1da177e4c Linux-2.6.12-rc2 |
1464 1465 1466 1467 |
*/ 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... |
1468 1469 1470 1471 1472 1473 |
struct radix_tree_iter iter; void **slot; unsigned ret = 0; if (unlikely(!nr_pages)) return 0; |
a60637c85 mm: lockless page... |
1474 1475 |
rcu_read_lock(); |
0fc9d1040 radix-tree: use i... |
1476 1477 |
radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter, *index, tag) { |
83929372f filemap: prepare ... |
1478 |
struct page *head, *page; |
a60637c85 mm: lockless page... |
1479 |
repeat: |
0fc9d1040 radix-tree: use i... |
1480 |
page = radix_tree_deref_slot(slot); |
a60637c85 mm: lockless page... |
1481 1482 |
if (unlikely(!page)) continue; |
9d8aa4ea8 mm: remove worryi... |
1483 |
|
a2c16d6cb mm: let swap use ... |
1484 |
if (radix_tree_exception(page)) { |
8079b1c85 mm: clarify the r... |
1485 |
if (radix_tree_deref_retry(page)) { |
2cf938aae mm: use radix_tre... |
1486 1487 |
slot = radix_tree_iter_retry(&iter); continue; |
8079b1c85 mm: clarify the r... |
1488 |
} |
a2c16d6cb mm: let swap use ... |
1489 |
/* |
139b6a6fb mm: filemap: upda... |
1490 1491 1492 1493 1494 1495 1496 1497 1498 |
* 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 ... |
1499 |
*/ |
139b6a6fb mm: filemap: upda... |
1500 |
continue; |
a2c16d6cb mm: let swap use ... |
1501 |
} |
a60637c85 mm: lockless page... |
1502 |
|
83929372f filemap: prepare ... |
1503 1504 |
head = compound_head(page); if (!page_cache_get_speculative(head)) |
a60637c85 mm: lockless page... |
1505 |
goto repeat; |
83929372f filemap: prepare ... |
1506 1507 1508 1509 1510 |
/* The page was split under us? */ if (compound_head(page) != head) { put_page(head); goto repeat; } |
a60637c85 mm: lockless page... |
1511 |
/* Has the page moved? */ |
0fc9d1040 radix-tree: use i... |
1512 |
if (unlikely(page != *slot)) { |
83929372f filemap: prepare ... |
1513 |
put_page(head); |
a60637c85 mm: lockless page... |
1514 1515 1516 1517 |
goto repeat; } pages[ret] = page; |
0fc9d1040 radix-tree: use i... |
1518 1519 |
if (++ret == nr_pages) break; |
a60637c85 mm: lockless page... |
1520 |
} |
5b280c0cc mm: don't return ... |
1521 |
|
a60637c85 mm: lockless page... |
1522 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
1523 |
|
1da177e4c Linux-2.6.12-rc2 |
1524 1525 |
if (ret) *index = pages[ret - 1]->index + 1; |
a60637c85 mm: lockless page... |
1526 |
|
1da177e4c Linux-2.6.12-rc2 |
1527 1528 |
return ret; } |
ef71c15c4 AFS: export a cou... |
1529 |
EXPORT_SYMBOL(find_get_pages_tag); |
1da177e4c Linux-2.6.12-rc2 |
1530 |
|
7e7f77498 mm: add find_get_... |
1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 |
/** * 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_... |
1555 1556 |
radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter, start, tag) { |
83929372f filemap: prepare ... |
1557 |
struct page *head, *page; |
7e7f77498 mm: add find_get_... |
1558 1559 1560 1561 1562 1563 |
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... |
1564 1565 |
slot = radix_tree_iter_retry(&iter); continue; |
7e7f77498 mm: add find_get_... |
1566 1567 1568 1569 1570 1571 1572 1573 1574 |
} /* * 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; } |
83929372f filemap: prepare ... |
1575 1576 1577 |
head = compound_head(page); if (!page_cache_get_speculative(head)) |
7e7f77498 mm: add find_get_... |
1578 |
goto repeat; |
83929372f filemap: prepare ... |
1579 1580 1581 1582 1583 |
/* The page was split under us? */ if (compound_head(page) != head) { put_page(head); goto repeat; } |
7e7f77498 mm: add find_get_... |
1584 1585 |
/* Has the page moved? */ if (unlikely(page != *slot)) { |
83929372f filemap: prepare ... |
1586 |
put_page(head); |
7e7f77498 mm: add find_get_... |
1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 |
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... |
1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 |
/* * 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... |
1617 |
ra->ra_pages /= 4; |
76d42bd96 [PATCH] readahead... |
1618 |
} |
485bb99b4 [PATCH] kernel-do... |
1619 |
/** |
36e789144 kill do_generic_m... |
1620 |
* do_generic_file_read - generic file read routine |
485bb99b4 [PATCH] kernel-do... |
1621 1622 |
* @filp: the file to read * @ppos: current file position |
6e58e79db introduce copy_pa... |
1623 1624 |
* @iter: data destination * @written: already copied |
485bb99b4 [PATCH] kernel-do... |
1625 |
* |
1da177e4c Linux-2.6.12-rc2 |
1626 |
* This is a generic file read routine, and uses the |
485bb99b4 [PATCH] kernel-do... |
1627 |
* mapping->a_ops->readpage() function for the actual low-level stuff. |
1da177e4c Linux-2.6.12-rc2 |
1628 1629 1630 |
* * 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 |
1631 |
*/ |
6e58e79db introduce copy_pa... |
1632 1633 |
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 |
1634 |
{ |
36e789144 kill do_generic_m... |
1635 |
struct address_space *mapping = filp->f_mapping; |
1da177e4c Linux-2.6.12-rc2 |
1636 |
struct inode *inode = mapping->host; |
36e789144 kill do_generic_m... |
1637 |
struct file_ra_state *ra = &filp->f_ra; |
57f6b96c0 filemap: convert ... |
1638 1639 1640 1641 |
pgoff_t index; pgoff_t last_index; pgoff_t prev_index; unsigned long offset; /* offset into pagecache page */ |
ec0f16372 readahead: improv... |
1642 |
unsigned int prev_offset; |
6e58e79db introduce copy_pa... |
1643 |
int error = 0; |
1da177e4c Linux-2.6.12-rc2 |
1644 |
|
c2a9737f4 vfs,mm: fix a dea... |
1645 |
if (unlikely(*ppos >= inode->i_sb->s_maxbytes)) |
48466c477 vfs,mm: fix retur... |
1646 |
return 0; |
c2a9737f4 vfs,mm: fix a dea... |
1647 |
iov_iter_truncate(iter, inode->i_sb->s_maxbytes); |
09cbfeaf1 mm, fs: get rid o... |
1648 1649 1650 1651 1652 |
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 |
1653 |
|
1da177e4c Linux-2.6.12-rc2 |
1654 1655 |
for (;;) { struct page *page; |
57f6b96c0 filemap: convert ... |
1656 |
pgoff_t end_index; |
a32ea1e1f Fix read/truncate... |
1657 |
loff_t isize; |
1da177e4c Linux-2.6.12-rc2 |
1658 |
unsigned long nr, ret; |
1da177e4c Linux-2.6.12-rc2 |
1659 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
1660 |
find_page: |
b67c7d39b mm, fs: check for... |
1661 1662 1663 1664 |
if (fatal_signal_pending(current)) { error = -EINTR; goto out; } |
1da177e4c Linux-2.6.12-rc2 |
1665 |
page = find_get_page(mapping, index); |
3ea89ee86 readahead: conver... |
1666 |
if (!page) { |
cf914a7d6 readahead: split ... |
1667 |
page_cache_sync_readahead(mapping, |
7ff81078d readahead: remove... |
1668 |
ra, filp, |
3ea89ee86 readahead: conver... |
1669 1670 1671 1672 1673 1674 |
index, last_index - index); page = find_get_page(mapping, index); if (unlikely(page == NULL)) goto no_cached_page; } if (PageReadahead(page)) { |
cf914a7d6 readahead: split ... |
1675 |
page_cache_async_readahead(mapping, |
7ff81078d readahead: remove... |
1676 |
ra, filp, page, |
3ea89ee86 readahead: conver... |
1677 |
index, last_index - index); |
1da177e4c Linux-2.6.12-rc2 |
1678 |
} |
8ab22b9ab vfs: pagecache us... |
1679 |
if (!PageUptodate(page)) { |
ebded0278 mm: filemap: avoi... |
1680 1681 1682 1683 1684 |
/* * See comment in do_read_cache_page on why * wait_on_page_locked is used to avoid unnecessarily * serialisations and why it's safe. */ |
c4b209a42 do_generic_file_r... |
1685 1686 1687 |
error = wait_on_page_locked_killable(page); if (unlikely(error)) goto readpage_error; |
ebded0278 mm: filemap: avoi... |
1688 1689 |
if (PageUptodate(page)) goto page_ok; |
09cbfeaf1 mm, fs: get rid o... |
1690 |
if (inode->i_blkbits == PAGE_SHIFT || |
8ab22b9ab vfs: pagecache us... |
1691 1692 |
!mapping->a_ops->is_partially_uptodate) goto page_not_up_to_date; |
6d6d36bc6 mm/filemap: don't... |
1693 1694 1695 |
/* pipes can't handle partially uptodate pages */ if (unlikely(iter->type & ITER_PIPE)) goto page_not_up_to_date; |
529ae9aaa mm: rename page t... |
1696 |
if (!trylock_page(page)) |
8ab22b9ab vfs: pagecache us... |
1697 |
goto page_not_up_to_date; |
8d056cb96 mm/vfs: revalidat... |
1698 1699 1700 |
/* Did it get truncated before we got the lock? */ if (!page->mapping) goto page_not_up_to_date_locked; |
8ab22b9ab vfs: pagecache us... |
1701 |
if (!mapping->a_ops->is_partially_uptodate(page, |
6e58e79db introduce copy_pa... |
1702 |
offset, iter->count)) |
8ab22b9ab vfs: pagecache us... |
1703 1704 1705 |
goto page_not_up_to_date_locked; unlock_page(page); } |
1da177e4c Linux-2.6.12-rc2 |
1706 |
page_ok: |
a32ea1e1f Fix read/truncate... |
1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 |
/* * 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... |
1717 |
end_index = (isize - 1) >> PAGE_SHIFT; |
a32ea1e1f Fix read/truncate... |
1718 |
if (unlikely(!isize || index > end_index)) { |
09cbfeaf1 mm, fs: get rid o... |
1719 |
put_page(page); |
a32ea1e1f Fix read/truncate... |
1720 1721 1722 1723 |
goto out; } /* nr is the maximum number of bytes to copy from this page */ |
09cbfeaf1 mm, fs: get rid o... |
1724 |
nr = PAGE_SIZE; |
a32ea1e1f Fix read/truncate... |
1725 |
if (index == end_index) { |
09cbfeaf1 mm, fs: get rid o... |
1726 |
nr = ((isize - 1) & ~PAGE_MASK) + 1; |
a32ea1e1f Fix read/truncate... |
1727 |
if (nr <= offset) { |
09cbfeaf1 mm, fs: get rid o... |
1728 |
put_page(page); |
a32ea1e1f Fix read/truncate... |
1729 1730 1731 1732 |
goto out; } } nr = nr - offset; |
1da177e4c Linux-2.6.12-rc2 |
1733 1734 1735 1736 1737 1738 1739 1740 1741 |
/* 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... |
1742 1743 |
* When a sequential read accesses a page several times, * only mark it as accessed the first time. |
1da177e4c Linux-2.6.12-rc2 |
1744 |
*/ |
ec0f16372 readahead: improv... |
1745 |
if (prev_index != index || offset != prev_offset) |
1da177e4c Linux-2.6.12-rc2 |
1746 1747 1748 1749 1750 1751 |
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 |
1752 |
*/ |
6e58e79db introduce copy_pa... |
1753 1754 |
ret = copy_page_to_iter(page, offset, nr, iter); |
1da177e4c Linux-2.6.12-rc2 |
1755 |
offset += ret; |
09cbfeaf1 mm, fs: get rid o... |
1756 1757 |
index += offset >> PAGE_SHIFT; offset &= ~PAGE_MASK; |
6ce745ed3 readahead: code c... |
1758 |
prev_offset = offset; |
1da177e4c Linux-2.6.12-rc2 |
1759 |
|
09cbfeaf1 mm, fs: get rid o... |
1760 |
put_page(page); |
6e58e79db introduce copy_pa... |
1761 1762 1763 1764 1765 1766 1767 1768 |
written += ret; if (!iov_iter_count(iter)) goto out; if (ret < nr) { error = -EFAULT; goto out; } continue; |
1da177e4c Linux-2.6.12-rc2 |
1769 1770 1771 |
page_not_up_to_date: /* Get exclusive access to the page ... */ |
854623235 do_generic_file_r... |
1772 1773 1774 |
error = lock_page_killable(page); if (unlikely(error)) goto readpage_error; |
1da177e4c Linux-2.6.12-rc2 |
1775 |
|
8ab22b9ab vfs: pagecache us... |
1776 |
page_not_up_to_date_locked: |
da6052f7b [PATCH] update so... |
1777 |
/* Did it get truncated before we got the lock? */ |
1da177e4c Linux-2.6.12-rc2 |
1778 1779 |
if (!page->mapping) { unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
1780 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 |
continue; } /* Did somebody else fill it already? */ if (PageUptodate(page)) { unlock_page(page); goto page_ok; } readpage: |
91803b499 do_generic_file_r... |
1791 1792 1793 1794 1795 1796 |
/* * 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 |
1797 1798 |
/* Start the actual read. The read will unlock the page. */ error = mapping->a_ops->readpage(filp, page); |
994fc28c7 [PATCH] add AOP_T... |
1799 1800 |
if (unlikely(error)) { if (error == AOP_TRUNCATED_PAGE) { |
09cbfeaf1 mm, fs: get rid o... |
1801 |
put_page(page); |
6e58e79db introduce copy_pa... |
1802 |
error = 0; |
994fc28c7 [PATCH] add AOP_T... |
1803 1804 |
goto find_page; } |
1da177e4c Linux-2.6.12-rc2 |
1805 |
goto readpage_error; |
994fc28c7 [PATCH] add AOP_T... |
1806 |
} |
1da177e4c Linux-2.6.12-rc2 |
1807 1808 |
if (!PageUptodate(page)) { |
854623235 do_generic_file_r... |
1809 1810 1811 |
error = lock_page_killable(page); if (unlikely(error)) goto readpage_error; |
1da177e4c Linux-2.6.12-rc2 |
1812 1813 1814 |
if (!PageUptodate(page)) { if (page->mapping == NULL) { /* |
2ecdc82ef kill unused inval... |
1815 |
* invalidate_mapping_pages got it |
1da177e4c Linux-2.6.12-rc2 |
1816 1817 |
*/ unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
1818 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1819 1820 1821 |
goto find_page; } unlock_page(page); |
7ff81078d readahead: remove... |
1822 |
shrink_readahead_size_eio(filp, ra); |
854623235 do_generic_file_r... |
1823 1824 |
error = -EIO; goto readpage_error; |
1da177e4c Linux-2.6.12-rc2 |
1825 1826 1827 |
} unlock_page(page); } |
1da177e4c Linux-2.6.12-rc2 |
1828 1829 1830 1831 |
goto page_ok; readpage_error: /* UHHUH! A synchronous read error occurred. Report it */ |
09cbfeaf1 mm, fs: get rid o... |
1832 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1833 1834 1835 1836 1837 1838 1839 |
goto out; no_cached_page: /* * Ok, it wasn't cached, so we need to create a new * page.. */ |
eb2be1893 mm: buffered writ... |
1840 1841 |
page = page_cache_alloc_cold(mapping); if (!page) { |
6e58e79db introduce copy_pa... |
1842 |
error = -ENOMEM; |
eb2be1893 mm: buffered writ... |
1843 |
goto out; |
1da177e4c Linux-2.6.12-rc2 |
1844 |
} |
6afdb859b mm: do not ignore... |
1845 |
error = add_to_page_cache_lru(page, mapping, index, |
c62d25556 mm, fs: introduce... |
1846 |
mapping_gfp_constraint(mapping, GFP_KERNEL)); |
1da177e4c Linux-2.6.12-rc2 |
1847 |
if (error) { |
09cbfeaf1 mm, fs: get rid o... |
1848 |
put_page(page); |
6e58e79db introduce copy_pa... |
1849 1850 |
if (error == -EEXIST) { error = 0; |
1da177e4c Linux-2.6.12-rc2 |
1851 |
goto find_page; |
6e58e79db introduce copy_pa... |
1852 |
} |
1da177e4c Linux-2.6.12-rc2 |
1853 1854 |
goto out; } |
1da177e4c Linux-2.6.12-rc2 |
1855 1856 1857 1858 |
goto readpage; } out: |
7ff81078d readahead: remove... |
1859 |
ra->prev_pos = prev_index; |
09cbfeaf1 mm, fs: get rid o... |
1860 |
ra->prev_pos <<= PAGE_SHIFT; |
7ff81078d readahead: remove... |
1861 |
ra->prev_pos |= prev_offset; |
1da177e4c Linux-2.6.12-rc2 |
1862 |
|
09cbfeaf1 mm, fs: get rid o... |
1863 |
*ppos = ((loff_t)index << PAGE_SHIFT) + offset; |
0c6aa2639 mm: do_generic_fi... |
1864 |
file_accessed(filp); |
6e58e79db introduce copy_pa... |
1865 |
return written ? written : error; |
1da177e4c Linux-2.6.12-rc2 |
1866 |
} |
485bb99b4 [PATCH] kernel-do... |
1867 |
/** |
6abd23227 bury generic_file... |
1868 |
* generic_file_read_iter - generic filesystem read routine |
485bb99b4 [PATCH] kernel-do... |
1869 |
* @iocb: kernel I/O control block |
6abd23227 bury generic_file... |
1870 |
* @iter: destination for the data read |
485bb99b4 [PATCH] kernel-do... |
1871 |
* |
6abd23227 bury generic_file... |
1872 |
* This is the "read_iter()" routine for all filesystems |
1da177e4c Linux-2.6.12-rc2 |
1873 1874 1875 |
* that can use the page cache directly. */ ssize_t |
ed978a811 new helper: gener... |
1876 |
generic_file_read_iter(struct kiocb *iocb, struct iov_iter *iter) |
1da177e4c Linux-2.6.12-rc2 |
1877 |
{ |
ed978a811 new helper: gener... |
1878 |
struct file *file = iocb->ki_filp; |
cb66a7a1f kill generic_segm... |
1879 |
ssize_t retval = 0; |
e7080a439 mm/filemap: gener... |
1880 1881 1882 1883 |
size_t count = iov_iter_count(iter); if (!count) goto out; /* skip atime */ |
1da177e4c Linux-2.6.12-rc2 |
1884 |
|
2ba48ce51 mirror O_APPEND a... |
1885 |
if (iocb->ki_flags & IOCB_DIRECT) { |
ed978a811 new helper: gener... |
1886 1887 |
struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; |
0d5b0cf24 fs: update atime ... |
1888 |
struct iov_iter data = *iter; |
543ade1fc [PATCH] Streamlin... |
1889 |
loff_t size; |
1da177e4c Linux-2.6.12-rc2 |
1890 |
|
1da177e4c Linux-2.6.12-rc2 |
1891 |
size = i_size_read(inode); |
c64fb5c74 filemap: remove p... |
1892 1893 |
retval = filemap_write_and_wait_range(mapping, iocb->ki_pos, iocb->ki_pos + count - 1); |
0d5b0cf24 fs: update atime ... |
1894 1895 |
if (retval < 0) goto out; |
d8d3d94b8 pass iov_iter to ... |
1896 |
|
0d5b0cf24 fs: update atime ... |
1897 1898 1899 |
file_accessed(file); retval = mapping->a_ops->direct_IO(iocb, &data); |
c3a690240 fix ITER_PIPE int... |
1900 |
if (retval >= 0) { |
c64fb5c74 filemap: remove p... |
1901 |
iocb->ki_pos += retval; |
ed978a811 new helper: gener... |
1902 |
iov_iter_advance(iter, retval); |
9fe55eea7 Fix race when che... |
1903 |
} |
66f998f61 fs: allow short d... |
1904 |
|
9fe55eea7 Fix race when che... |
1905 1906 1907 1908 1909 1910 |
/* * 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... |
1911 1912 |
* the rest of the read. Buffered reads will not work for * DAX files, so don't bother trying. |
9fe55eea7 Fix race when che... |
1913 |
*/ |
c64fb5c74 filemap: remove p... |
1914 |
if (retval < 0 || !iov_iter_count(iter) || iocb->ki_pos >= size || |
0d5b0cf24 fs: update atime ... |
1915 |
IS_DAX(inode)) |
9fe55eea7 Fix race when che... |
1916 |
goto out; |
1da177e4c Linux-2.6.12-rc2 |
1917 |
} |
c64fb5c74 filemap: remove p... |
1918 |
retval = do_generic_file_read(file, &iocb->ki_pos, iter, retval); |
1da177e4c Linux-2.6.12-rc2 |
1919 1920 1921 |
out: return retval; } |
ed978a811 new helper: gener... |
1922 |
EXPORT_SYMBOL(generic_file_read_iter); |
1da177e4c Linux-2.6.12-rc2 |
1923 |
|
1da177e4c Linux-2.6.12-rc2 |
1924 |
#ifdef CONFIG_MMU |
485bb99b4 [PATCH] kernel-do... |
1925 1926 1927 1928 |
/** * 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... |
1929 |
* @gfp_mask: memory allocation flags |
485bb99b4 [PATCH] kernel-do... |
1930 |
* |
1da177e4c Linux-2.6.12-rc2 |
1931 1932 1933 |
* 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... |
1934 |
static int page_cache_read(struct file *file, pgoff_t offset, gfp_t gfp_mask) |
1da177e4c Linux-2.6.12-rc2 |
1935 1936 |
{ struct address_space *mapping = file->f_mapping; |
99dadfdde mm/filemap.c: rem... |
1937 |
struct page *page; |
994fc28c7 [PATCH] add AOP_T... |
1938 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
1939 |
|
994fc28c7 [PATCH] add AOP_T... |
1940 |
do { |
c20cd45eb mm: allow GFP_{FS... |
1941 |
page = __page_cache_alloc(gfp_mask|__GFP_COLD); |
994fc28c7 [PATCH] add AOP_T... |
1942 1943 |
if (!page) return -ENOMEM; |
c20cd45eb mm: allow GFP_{FS... |
1944 |
ret = add_to_page_cache_lru(page, mapping, offset, gfp_mask & GFP_KERNEL); |
994fc28c7 [PATCH] add AOP_T... |
1945 1946 1947 1948 |
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 |
1949 |
|
09cbfeaf1 mm, fs: get rid o... |
1950 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1951 |
|
994fc28c7 [PATCH] add AOP_T... |
1952 |
} while (ret == AOP_TRUNCATED_PAGE); |
99dadfdde mm/filemap.c: rem... |
1953 |
|
994fc28c7 [PATCH] add AOP_T... |
1954 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
1955 1956 1957 |
} #define MMAP_LOTSAMISS (100) |
ef00e08e2 readahead: clean ... |
1958 1959 1960 1961 1962 1963 1964 1965 1966 |
/* * 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 ... |
1967 1968 1969 |
struct address_space *mapping = file->f_mapping; /* If we don't want any read-ahead, don't bother */ |
64363aad5 mm: remove unused... |
1970 |
if (vma->vm_flags & VM_RAND_READ) |
ef00e08e2 readahead: clean ... |
1971 |
return; |
275b12bf5 readahead: return... |
1972 1973 |
if (!ra->ra_pages) return; |
ef00e08e2 readahead: clean ... |
1974 |
|
64363aad5 mm: remove unused... |
1975 |
if (vma->vm_flags & VM_SEQ_READ) { |
7ffc59b4d readahead: enforc... |
1976 1977 |
page_cache_sync_readahead(mapping, ra, file, offset, ra->ra_pages); |
ef00e08e2 readahead: clean ... |
1978 1979 |
return; } |
207d04baa readahead: reduce... |
1980 1981 |
/* Avoid banging the cache line if not needed */ if (ra->mmap_miss < MMAP_LOTSAMISS * 10) |
ef00e08e2 readahead: clean ... |
1982 1983 1984 1985 1986 1987 1988 1989 |
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... |
1990 1991 1992 |
/* * mmap read-around */ |
600e19afc mm: use only per-... |
1993 1994 1995 |
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... |
1996 |
ra_submit(ra, mapping, file); |
ef00e08e2 readahead: clean ... |
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 |
} /* * 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... |
2012 |
if (vma->vm_flags & VM_RAND_READ) |
ef00e08e2 readahead: clean ... |
2013 2014 2015 2016 |
return; if (ra->mmap_miss > 0) ra->mmap_miss--; if (PageReadahead(page)) |
2fad6f5de readahead: enforc... |
2017 2018 |
page_cache_async_readahead(mapping, ra, file, page, offset, ra->ra_pages); |
ef00e08e2 readahead: clean ... |
2019 |
} |
485bb99b4 [PATCH] kernel-do... |
2020 |
/** |
54cb8821d mm: merge populat... |
2021 |
* filemap_fault - read in file data for page fault handling |
d0217ac04 mm: fault feedbac... |
2022 2023 |
* @vma: vma in which the fault was taken * @vmf: struct vm_fault containing details of the fault |
485bb99b4 [PATCH] kernel-do... |
2024 |
* |
54cb8821d mm: merge populat... |
2025 |
* filemap_fault() is invoked via the vma operations vector for a |
1da177e4c Linux-2.6.12-rc2 |
2026 2027 2028 2029 2030 |
* 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... |
2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 |
* * 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 |
2043 |
*/ |
d0217ac04 mm: fault feedbac... |
2044 |
int filemap_fault(struct vm_area_struct *vma, struct vm_fault *vmf) |
1da177e4c Linux-2.6.12-rc2 |
2045 2046 |
{ int error; |
54cb8821d mm: merge populat... |
2047 |
struct file *file = vma->vm_file; |
1da177e4c Linux-2.6.12-rc2 |
2048 2049 2050 |
struct address_space *mapping = file->f_mapping; struct file_ra_state *ra = &file->f_ra; struct inode *inode = mapping->host; |
ef00e08e2 readahead: clean ... |
2051 |
pgoff_t offset = vmf->pgoff; |
1da177e4c Linux-2.6.12-rc2 |
2052 |
struct page *page; |
99e3e53f4 mm: cleanup size ... |
2053 |
loff_t size; |
83c54070e mm: fault feedbac... |
2054 |
int ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
2055 |
|
09cbfeaf1 mm, fs: get rid o... |
2056 2057 |
size = round_up(i_size_read(inode), PAGE_SIZE); if (offset >= size >> PAGE_SHIFT) |
5307cc1aa Remove broken ptr... |
2058 |
return VM_FAULT_SIGBUS; |
1da177e4c Linux-2.6.12-rc2 |
2059 |
|
1da177e4c Linux-2.6.12-rc2 |
2060 |
/* |
494264208 mm: memcg: handle... |
2061 |
* Do we have something in the page cache already? |
1da177e4c Linux-2.6.12-rc2 |
2062 |
*/ |
ef00e08e2 readahead: clean ... |
2063 |
page = find_get_page(mapping, offset); |
45cac65b0 readahead: fault ... |
2064 |
if (likely(page) && !(vmf->flags & FAULT_FLAG_TRIED)) { |
1da177e4c Linux-2.6.12-rc2 |
2065 |
/* |
ef00e08e2 readahead: clean ... |
2066 2067 |
* We found the page, so try async readahead before * waiting for the lock. |
1da177e4c Linux-2.6.12-rc2 |
2068 |
*/ |
ef00e08e2 readahead: clean ... |
2069 |
do_async_mmap_readahead(vma, ra, file, page, offset); |
45cac65b0 readahead: fault ... |
2070 |
} else if (!page) { |
ef00e08e2 readahead: clean ... |
2071 2072 2073 |
/* 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... |
2074 |
mem_cgroup_count_vm_event(vma->vm_mm, PGMAJFAULT); |
ef00e08e2 readahead: clean ... |
2075 2076 |
ret = VM_FAULT_MAJOR; retry_find: |
b522c94da mm: filemap_fault... |
2077 |
page = find_get_page(mapping, offset); |
1da177e4c Linux-2.6.12-rc2 |
2078 2079 2080 |
if (!page) goto no_cached_page; } |
d88c0922f Release page refe... |
2081 |
if (!lock_page_or_retry(page, vma->vm_mm, vmf->flags)) { |
09cbfeaf1 mm, fs: get rid o... |
2082 |
put_page(page); |
d065bd810 mm: retry page fa... |
2083 |
return ret | VM_FAULT_RETRY; |
d88c0922f Release page refe... |
2084 |
} |
b522c94da mm: filemap_fault... |
2085 2086 2087 2088 2089 2090 2091 |
/* Did it get truncated? */ if (unlikely(page->mapping != mapping)) { unlock_page(page); put_page(page); goto retry_find; } |
309381fea mm: dump page whe... |
2092 |
VM_BUG_ON_PAGE(page->index != offset, page); |
b522c94da mm: filemap_fault... |
2093 |
|
1da177e4c Linux-2.6.12-rc2 |
2094 |
/* |
d00806b18 mm: fix fault vs ... |
2095 2096 |
* 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 |
2097 |
*/ |
d00806b18 mm: fix fault vs ... |
2098 |
if (unlikely(!PageUptodate(page))) |
1da177e4c Linux-2.6.12-rc2 |
2099 |
goto page_not_uptodate; |
ef00e08e2 readahead: clean ... |
2100 2101 2102 2103 |
/* * Found the page and have a reference on it. * We must recheck i_size under page lock. */ |
09cbfeaf1 mm, fs: get rid o... |
2104 2105 |
size = round_up(i_size_read(inode), PAGE_SIZE); if (unlikely(offset >= size >> PAGE_SHIFT)) { |
d00806b18 mm: fix fault vs ... |
2106 |
unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
2107 |
put_page(page); |
5307cc1aa Remove broken ptr... |
2108 |
return VM_FAULT_SIGBUS; |
d00806b18 mm: fix fault vs ... |
2109 |
} |
d0217ac04 mm: fault feedbac... |
2110 |
vmf->page = page; |
83c54070e mm: fault feedbac... |
2111 |
return ret | VM_FAULT_LOCKED; |
1da177e4c Linux-2.6.12-rc2 |
2112 |
|
1da177e4c Linux-2.6.12-rc2 |
2113 2114 2115 2116 2117 |
no_cached_page: /* * We're only likely to ever get here if MADV_RANDOM is in * effect. */ |
c20cd45eb mm: allow GFP_{FS... |
2118 |
error = page_cache_read(file, offset, vmf->gfp_mask); |
1da177e4c Linux-2.6.12-rc2 |
2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 |
/* * 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... |
2134 2135 |
return VM_FAULT_OOM; return VM_FAULT_SIGBUS; |
1da177e4c Linux-2.6.12-rc2 |
2136 2137 |
page_not_uptodate: |
1da177e4c Linux-2.6.12-rc2 |
2138 2139 2140 2141 2142 2143 |
/* * 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 |
2144 |
ClearPageError(page); |
994fc28c7 [PATCH] add AOP_T... |
2145 |
error = mapping->a_ops->readpage(file, page); |
3ef0f720e mm: fix infinite ... |
2146 2147 2148 2149 2150 |
if (!error) { wait_on_page_locked(page); if (!PageUptodate(page)) error = -EIO; } |
09cbfeaf1 mm, fs: get rid o... |
2151 |
put_page(page); |
d00806b18 mm: fix fault vs ... |
2152 2153 |
if (!error || error == AOP_TRUNCATED_PAGE) |
994fc28c7 [PATCH] add AOP_T... |
2154 |
goto retry_find; |
1da177e4c Linux-2.6.12-rc2 |
2155 |
|
d00806b18 mm: fix fault vs ... |
2156 |
/* Things didn't work out. Return zero to tell the mm layer so. */ |
76d42bd96 [PATCH] readahead... |
2157 |
shrink_readahead_size_eio(file, ra); |
d0217ac04 mm: fault feedbac... |
2158 |
return VM_FAULT_SIGBUS; |
54cb8821d mm: merge populat... |
2159 2160 |
} EXPORT_SYMBOL(filemap_fault); |
bae473a42 mm: introduce fau... |
2161 2162 |
void filemap_map_pages(struct fault_env *fe, pgoff_t start_pgoff, pgoff_t end_pgoff) |
f1820361f mm: implement ->m... |
2163 2164 2165 |
{ struct radix_tree_iter iter; void **slot; |
bae473a42 mm: introduce fau... |
2166 |
struct file *file = fe->vma->vm_file; |
f1820361f mm: implement ->m... |
2167 |
struct address_space *mapping = file->f_mapping; |
bae473a42 mm: introduce fau... |
2168 |
pgoff_t last_pgoff = start_pgoff; |
f1820361f mm: implement ->m... |
2169 |
loff_t size; |
83929372f filemap: prepare ... |
2170 |
struct page *head, *page; |
f1820361f mm: implement ->m... |
2171 2172 |
rcu_read_lock(); |
bae473a42 mm: introduce fau... |
2173 2174 2175 |
radix_tree_for_each_slot(slot, &mapping->page_tree, &iter, start_pgoff) { if (iter.index > end_pgoff) |
f1820361f mm: implement ->m... |
2176 2177 2178 2179 2180 2181 |
break; repeat: page = radix_tree_deref_slot(slot); if (unlikely(!page)) goto next; if (radix_tree_exception(page)) { |
2cf938aae mm: use radix_tre... |
2182 2183 2184 2185 2186 |
if (radix_tree_deref_retry(page)) { slot = radix_tree_iter_retry(&iter); continue; } goto next; |
f1820361f mm: implement ->m... |
2187 |
} |
83929372f filemap: prepare ... |
2188 2189 |
head = compound_head(page); if (!page_cache_get_speculative(head)) |
f1820361f mm: implement ->m... |
2190 |
goto repeat; |
83929372f filemap: prepare ... |
2191 2192 2193 2194 2195 |
/* The page was split under us? */ if (compound_head(page) != head) { put_page(head); goto repeat; } |
f1820361f mm: implement ->m... |
2196 2197 |
/* Has the page moved? */ if (unlikely(page != *slot)) { |
83929372f filemap: prepare ... |
2198 |
put_page(head); |
f1820361f mm: implement ->m... |
2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 |
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... |
2211 2212 |
size = round_up(i_size_read(mapping->host), PAGE_SIZE); if (page->index >= size >> PAGE_SHIFT) |
f1820361f mm: implement ->m... |
2213 |
goto unlock; |
f1820361f mm: implement ->m... |
2214 2215 |
if (file->f_ra.mmap_miss > 0) file->f_ra.mmap_miss--; |
7267ec008 mm: postpone page... |
2216 2217 2218 2219 2220 2221 2222 |
fe->address += (iter.index - last_pgoff) << PAGE_SHIFT; if (fe->pte) fe->pte += iter.index - last_pgoff; last_pgoff = iter.index; if (alloc_set_pte(fe, NULL, page)) goto unlock; |
f1820361f mm: implement ->m... |
2223 2224 2225 2226 2227 |
unlock_page(page); goto next; unlock: unlock_page(page); skip: |
09cbfeaf1 mm, fs: get rid o... |
2228 |
put_page(page); |
f1820361f mm: implement ->m... |
2229 |
next: |
7267ec008 mm: postpone page... |
2230 2231 2232 |
/* Huge page is mapped? No need to proceed. */ if (pmd_trans_huge(*fe->pmd)) break; |
bae473a42 mm: introduce fau... |
2233 |
if (iter.index == end_pgoff) |
f1820361f mm: implement ->m... |
2234 2235 2236 2237 2238 |
break; } rcu_read_unlock(); } EXPORT_SYMBOL(filemap_map_pages); |
4fcf1c620 mm: Make default ... |
2239 2240 2241 |
int filemap_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { struct page *page = vmf->page; |
496ad9aa8 new helper: file_... |
2242 |
struct inode *inode = file_inode(vma->vm_file); |
4fcf1c620 mm: Make default ... |
2243 |
int ret = VM_FAULT_LOCKED; |
14da92001 fs: Protect write... |
2244 |
sb_start_pagefault(inode->i_sb); |
4fcf1c620 mm: Make default ... |
2245 2246 2247 2248 2249 2250 2251 |
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... |
2252 2253 2254 2255 2256 2257 |
/* * 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 ... |
2258 |
wait_for_stable_page(page); |
4fcf1c620 mm: Make default ... |
2259 |
out: |
14da92001 fs: Protect write... |
2260 |
sb_end_pagefault(inode->i_sb); |
4fcf1c620 mm: Make default ... |
2261 2262 2263 |
return ret; } EXPORT_SYMBOL(filemap_page_mkwrite); |
f0f37e2f7 const: mark struc... |
2264 |
const struct vm_operations_struct generic_file_vm_ops = { |
54cb8821d mm: merge populat... |
2265 |
.fault = filemap_fault, |
f1820361f mm: implement ->m... |
2266 |
.map_pages = filemap_map_pages, |
4fcf1c620 mm: Make default ... |
2267 |
.page_mkwrite = filemap_page_mkwrite, |
1da177e4c Linux-2.6.12-rc2 |
2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 |
}; /* 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 |
2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 |
/* * 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... |
2305 2306 2307 2308 2309 |
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... |
2310 |
put_page(page); |
67f9fd91f mm: remove read_c... |
2311 2312 2313 2314 2315 |
page = ERR_PTR(-EIO); } } return page; } |
32b635298 mm: filemap: remo... |
2316 |
static struct page *do_read_cache_page(struct address_space *mapping, |
57f6b96c0 filemap: convert ... |
2317 |
pgoff_t index, |
5e5358e7c mm: cleanup descr... |
2318 |
int (*filler)(void *, struct page *), |
0531b2aac mm: add new 'read... |
2319 2320 |
void *data, gfp_t gfp) |
1da177e4c Linux-2.6.12-rc2 |
2321 |
{ |
eb2be1893 mm: buffered writ... |
2322 |
struct page *page; |
1da177e4c Linux-2.6.12-rc2 |
2323 2324 2325 2326 |
int err; repeat: page = find_get_page(mapping, index); if (!page) { |
0531b2aac mm: add new 'read... |
2327 |
page = __page_cache_alloc(gfp | __GFP_COLD); |
eb2be1893 mm: buffered writ... |
2328 2329 |
if (!page) return ERR_PTR(-ENOMEM); |
e6f67b8c0 vfs: __read_cache... |
2330 |
err = add_to_page_cache_lru(page, mapping, index, gfp); |
eb2be1893 mm: buffered writ... |
2331 |
if (unlikely(err)) { |
09cbfeaf1 mm, fs: get rid o... |
2332 |
put_page(page); |
eb2be1893 mm: buffered writ... |
2333 2334 |
if (err == -EEXIST) goto repeat; |
1da177e4c Linux-2.6.12-rc2 |
2335 |
/* Presumably ENOMEM for radix tree node */ |
1da177e4c Linux-2.6.12-rc2 |
2336 2337 |
return ERR_PTR(err); } |
32b635298 mm: filemap: remo... |
2338 2339 |
filler: |
1da177e4c Linux-2.6.12-rc2 |
2340 2341 |
err = filler(data, page); if (err < 0) { |
09cbfeaf1 mm, fs: get rid o... |
2342 |
put_page(page); |
32b635298 mm: filemap: remo... |
2343 |
return ERR_PTR(err); |
1da177e4c Linux-2.6.12-rc2 |
2344 |
} |
1da177e4c Linux-2.6.12-rc2 |
2345 |
|
32b635298 mm: filemap: remo... |
2346 2347 2348 2349 2350 |
page = wait_on_page_read(page); if (IS_ERR(page)) return page; goto out; } |
1da177e4c Linux-2.6.12-rc2 |
2351 2352 |
if (PageUptodate(page)) goto out; |
ebded0278 mm: filemap: avoi... |
2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 |
/* * 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 |
2389 |
lock_page(page); |
ebded0278 mm: filemap: avoi... |
2390 2391 |
/* Case c or d, restart the operation */ |
1da177e4c Linux-2.6.12-rc2 |
2392 2393 |
if (!page->mapping) { unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
2394 |
put_page(page); |
32b635298 mm: filemap: remo... |
2395 |
goto repeat; |
1da177e4c Linux-2.6.12-rc2 |
2396 |
} |
ebded0278 mm: filemap: avoi... |
2397 2398 |
/* Someone else locked and filled the page in a very small window */ |
1da177e4c Linux-2.6.12-rc2 |
2399 2400 2401 2402 |
if (PageUptodate(page)) { unlock_page(page); goto out; } |
32b635298 mm: filemap: remo... |
2403 |
goto filler; |
c855ff371 Fix a bad error c... |
2404 |
out: |
6fe6900e1 mm: make read_cac... |
2405 2406 2407 |
mark_page_accessed(page); return page; } |
0531b2aac mm: add new 'read... |
2408 2409 |
/** |
67f9fd91f mm: remove read_c... |
2410 |
* read_cache_page - read into page cache, fill it if needed |
0531b2aac mm: add new 'read... |
2411 2412 2413 |
* @mapping: the page's address_space * @index: the page index * @filler: function to perform the read |
5e5358e7c mm: cleanup descr... |
2414 |
* @data: first arg to filler(data, page) function, often left as NULL |
0531b2aac mm: add new 'read... |
2415 |
* |
0531b2aac mm: add new 'read... |
2416 |
* Read into the page cache. If a page already exists, and PageUptodate() is |
67f9fd91f mm: remove read_c... |
2417 |
* not set, try to fill the page and wait for it to become unlocked. |
0531b2aac mm: add new 'read... |
2418 2419 2420 |
* * If the page does not get brought uptodate, return -EIO. */ |
67f9fd91f mm: remove read_c... |
2421 |
struct page *read_cache_page(struct address_space *mapping, |
0531b2aac mm: add new 'read... |
2422 |
pgoff_t index, |
5e5358e7c mm: cleanup descr... |
2423 |
int (*filler)(void *, struct page *), |
0531b2aac mm: add new 'read... |
2424 2425 2426 2427 |
void *data) { return do_read_cache_page(mapping, index, filler, data, mapping_gfp_mask(mapping)); } |
67f9fd91f mm: remove read_c... |
2428 |
EXPORT_SYMBOL(read_cache_page); |
0531b2aac mm: add new 'read... |
2429 2430 2431 2432 2433 2434 2435 2436 |
/** * 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... |
2437 |
* any new page allocations done using the specified allocation flags. |
0531b2aac mm: add new 'read... |
2438 2439 2440 2441 2442 2443 2444 2445 |
* * 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... |
2446 |
return do_read_cache_page(mapping, index, filler, NULL, gfp); |
0531b2aac mm: add new 'read... |
2447 2448 |
} EXPORT_SYMBOL(read_cache_page_gfp); |
2f718ffc1 mm: buffered writ... |
2449 |
/* |
1da177e4c Linux-2.6.12-rc2 |
2450 2451 |
* Performs necessary checks before doing a write * |
485bb99b4 [PATCH] kernel-do... |
2452 |
* Can adjust writing position or amount of bytes to write. |
1da177e4c Linux-2.6.12-rc2 |
2453 2454 2455 |
* Returns appropriate error code that caller should return or * zero in case that write should be allowed. */ |
3309dd04c switch generic_wr... |
2456 |
inline ssize_t generic_write_checks(struct kiocb *iocb, struct iov_iter *from) |
1da177e4c Linux-2.6.12-rc2 |
2457 |
{ |
3309dd04c switch generic_wr... |
2458 |
struct file *file = iocb->ki_filp; |
1da177e4c Linux-2.6.12-rc2 |
2459 |
struct inode *inode = file->f_mapping->host; |
59e99e5b9 mm: use rlimit he... |
2460 |
unsigned long limit = rlimit(RLIMIT_FSIZE); |
3309dd04c switch generic_wr... |
2461 |
loff_t pos; |
1da177e4c Linux-2.6.12-rc2 |
2462 |
|
3309dd04c switch generic_wr... |
2463 2464 |
if (!iov_iter_count(from)) return 0; |
1da177e4c Linux-2.6.12-rc2 |
2465 |
|
0fa6b005a generic_write_che... |
2466 |
/* FIXME: this is for backwards compatibility with 2.4 */ |
2ba48ce51 mirror O_APPEND a... |
2467 |
if (iocb->ki_flags & IOCB_APPEND) |
3309dd04c switch generic_wr... |
2468 |
iocb->ki_pos = i_size_read(inode); |
1da177e4c Linux-2.6.12-rc2 |
2469 |
|
3309dd04c switch generic_wr... |
2470 |
pos = iocb->ki_pos; |
1da177e4c Linux-2.6.12-rc2 |
2471 |
|
0fa6b005a generic_write_che... |
2472 |
if (limit != RLIM_INFINITY) { |
3309dd04c switch generic_wr... |
2473 |
if (iocb->ki_pos >= limit) { |
0fa6b005a generic_write_che... |
2474 2475 |
send_sig(SIGXFSZ, current, 0); return -EFBIG; |
1da177e4c Linux-2.6.12-rc2 |
2476 |
} |
3309dd04c switch generic_wr... |
2477 |
iov_iter_truncate(from, limit - (unsigned long)pos); |
1da177e4c Linux-2.6.12-rc2 |
2478 2479 2480 2481 2482 |
} /* * LFS rule */ |
3309dd04c switch generic_wr... |
2483 |
if (unlikely(pos + iov_iter_count(from) > MAX_NON_LFS && |
1da177e4c Linux-2.6.12-rc2 |
2484 |
!(file->f_flags & O_LARGEFILE))) { |
3309dd04c switch generic_wr... |
2485 |
if (pos >= MAX_NON_LFS) |
1da177e4c Linux-2.6.12-rc2 |
2486 |
return -EFBIG; |
3309dd04c switch generic_wr... |
2487 |
iov_iter_truncate(from, MAX_NON_LFS - (unsigned long)pos); |
1da177e4c Linux-2.6.12-rc2 |
2488 2489 2490 2491 2492 2493 2494 2495 2496 |
} /* * 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... |
2497 2498 |
if (unlikely(pos >= inode->i_sb->s_maxbytes)) return -EFBIG; |
1da177e4c Linux-2.6.12-rc2 |
2499 |
|
3309dd04c switch generic_wr... |
2500 2501 |
iov_iter_truncate(from, inode->i_sb->s_maxbytes - pos); return iov_iter_count(from); |
1da177e4c Linux-2.6.12-rc2 |
2502 2503 |
} EXPORT_SYMBOL(generic_write_checks); |
afddba49d fs: introduce wri... |
2504 2505 2506 2507 2508 |
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... |
2509 |
return aops->write_begin(file, mapping, pos, len, flags, |
afddba49d fs: introduce wri... |
2510 |
pagep, fsdata); |
afddba49d fs: introduce wri... |
2511 2512 2513 2514 2515 2516 2517 2518 |
} 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... |
2519 |
|
4e02ed4b4 fs: remove prepar... |
2520 |
return aops->write_end(file, mapping, pos, len, copied, page, fsdata); |
afddba49d fs: introduce wri... |
2521 2522 |
} EXPORT_SYMBOL(pagecache_write_end); |
1da177e4c Linux-2.6.12-rc2 |
2523 |
ssize_t |
1af5bb491 filemap: remove t... |
2524 |
generic_file_direct_write(struct kiocb *iocb, struct iov_iter *from) |
1da177e4c Linux-2.6.12-rc2 |
2525 2526 2527 2528 |
{ struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; |
1af5bb491 filemap: remove t... |
2529 |
loff_t pos = iocb->ki_pos; |
1da177e4c Linux-2.6.12-rc2 |
2530 |
ssize_t written; |
a969e903a kill generic_file... |
2531 2532 |
size_t write_len; pgoff_t end; |
26978b8b4 give ->direct_IO(... |
2533 |
struct iov_iter data; |
1da177e4c Linux-2.6.12-rc2 |
2534 |
|
0c949334a iov_iter_truncate() |
2535 |
write_len = iov_iter_count(from); |
09cbfeaf1 mm, fs: get rid o... |
2536 |
end = (pos + write_len - 1) >> PAGE_SHIFT; |
a969e903a kill generic_file... |
2537 |
|
48b47c561 mm: direct IO sta... |
2538 |
written = filemap_write_and_wait_range(mapping, pos, pos + write_len - 1); |
a969e903a kill generic_file... |
2539 2540 2541 2542 2543 2544 2545 |
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... |
2546 |
* without clobbering -EIOCBQUEUED from ->direct_IO(). |
a969e903a kill generic_file... |
2547 2548 2549 |
*/ if (mapping->nrpages) { written = invalidate_inode_pages2_range(mapping, |
09cbfeaf1 mm, fs: get rid o... |
2550 |
pos >> PAGE_SHIFT, end); |
6ccfa806a VFS: fix dio writ... |
2551 2552 2553 2554 2555 2556 2557 |
/* * 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... |
2558 |
goto out; |
6ccfa806a VFS: fix dio writ... |
2559 |
} |
a969e903a kill generic_file... |
2560 |
} |
26978b8b4 give ->direct_IO(... |
2561 |
data = *from; |
c8b8e32d7 direct-io: elimin... |
2562 |
written = mapping->a_ops->direct_IO(iocb, &data); |
a969e903a kill generic_file... |
2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 |
/* * 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... |
2574 |
pos >> PAGE_SHIFT, end); |
a969e903a kill generic_file... |
2575 |
} |
1da177e4c Linux-2.6.12-rc2 |
2576 |
if (written > 0) { |
0116651c8 mm: remove tempor... |
2577 |
pos += written; |
f8579f867 generic_file_dire... |
2578 |
iov_iter_advance(from, written); |
0116651c8 mm: remove tempor... |
2579 2580 |
if (pos > i_size_read(inode) && !S_ISBLK(inode->i_mode)) { i_size_write(inode, pos); |
1da177e4c Linux-2.6.12-rc2 |
2581 2582 |
mark_inode_dirty(inode); } |
5cb6c6c7e generic_file_dire... |
2583 |
iocb->ki_pos = pos; |
1da177e4c Linux-2.6.12-rc2 |
2584 |
} |
a969e903a kill generic_file... |
2585 |
out: |
1da177e4c Linux-2.6.12-rc2 |
2586 2587 2588 |
return written; } EXPORT_SYMBOL(generic_file_direct_write); |
eb2be1893 mm: buffered writ... |
2589 2590 2591 2592 |
/* * 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... |
2593 2594 |
struct page *grab_cache_page_write_begin(struct address_space *mapping, pgoff_t index, unsigned flags) |
eb2be1893 mm: buffered writ... |
2595 |
{ |
eb2be1893 mm: buffered writ... |
2596 |
struct page *page; |
bbddabe2e mm: filemap: only... |
2597 |
int fgp_flags = FGP_LOCK|FGP_WRITE|FGP_CREAT; |
0faa70cb0 mm: filemap: pass... |
2598 |
|
54566b2c1 fs: symlink write... |
2599 |
if (flags & AOP_FLAG_NOFS) |
2457aec63 mm: non-atomicall... |
2600 2601 2602 |
fgp_flags |= FGP_NOFS; page = pagecache_get_page(mapping, index, fgp_flags, |
45f87de57 mm: get rid of ra... |
2603 |
mapping_gfp_mask(mapping)); |
c585a2678 mm: remove likely... |
2604 |
if (page) |
2457aec63 mm: non-atomicall... |
2605 |
wait_for_stable_page(page); |
eb2be1893 mm: buffered writ... |
2606 |
|
eb2be1893 mm: buffered writ... |
2607 2608 |
return page; } |
54566b2c1 fs: symlink write... |
2609 |
EXPORT_SYMBOL(grab_cache_page_write_begin); |
eb2be1893 mm: buffered writ... |
2610 |
|
3b93f911d export generic_pe... |
2611 |
ssize_t generic_perform_write(struct file *file, |
afddba49d fs: introduce wri... |
2612 2613 2614 2615 2616 2617 |
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... |
2618 2619 2620 2621 2622 |
unsigned int flags = 0; /* * Copies from kernel address space cannot fail (NFSD is a big user). */ |
777eda2c5 new helper: iter_... |
2623 |
if (!iter_is_iovec(i)) |
674b892ed mm: restore KERNE... |
2624 |
flags |= AOP_FLAG_UNINTERRUPTIBLE; |
afddba49d fs: introduce wri... |
2625 2626 2627 |
do { struct page *page; |
afddba49d fs: introduce wri... |
2628 2629 2630 2631 |
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... |
2632 2633 |
offset = (pos & (PAGE_SIZE - 1)); bytes = min_t(unsigned long, PAGE_SIZE - offset, |
afddba49d fs: introduce wri... |
2634 2635 2636 |
iov_iter_count(i)); again: |
00a3d660c Revert "fs: do no... |
2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 |
/* * 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... |
2651 2652 2653 2654 |
if (fatal_signal_pending(current)) { status = -EINTR; break; } |
674b892ed mm: restore KERNE... |
2655 |
status = a_ops->write_begin(file, mapping, pos, bytes, flags, |
afddba49d fs: introduce wri... |
2656 |
&page, &fsdata); |
2457aec63 mm: non-atomicall... |
2657 |
if (unlikely(status < 0)) |
afddba49d fs: introduce wri... |
2658 |
break; |
931e80e4b mm: flush dcache ... |
2659 2660 |
if (mapping_writably_mapped(mapping)) flush_dcache_page(page); |
00a3d660c Revert "fs: do no... |
2661 |
|
afddba49d fs: introduce wri... |
2662 |
copied = iov_iter_copy_from_user_atomic(page, i, offset, bytes); |
afddba49d fs: introduce wri... |
2663 2664 2665 2666 2667 2668 2669 2670 2671 |
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... |
2672 |
iov_iter_advance(i, copied); |
afddba49d fs: introduce wri... |
2673 2674 2675 2676 2677 2678 2679 2680 2681 |
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... |
2682 |
bytes = min_t(unsigned long, PAGE_SIZE - offset, |
afddba49d fs: introduce wri... |
2683 2684 2685 |
iov_iter_single_seg_count(i)); goto again; } |
afddba49d fs: introduce wri... |
2686 2687 2688 2689 |
pos += copied; written += copied; balance_dirty_pages_ratelimited(mapping); |
afddba49d fs: introduce wri... |
2690 2691 2692 2693 |
} while (iov_iter_count(i)); return written ? written : status; } |
3b93f911d export generic_pe... |
2694 |
EXPORT_SYMBOL(generic_perform_write); |
1da177e4c Linux-2.6.12-rc2 |
2695 |
|
e4dd9de3c vfs: Export __gen... |
2696 |
/** |
8174202b3 write_iter varian... |
2697 |
* __generic_file_write_iter - write data to a file |
e4dd9de3c vfs: Export __gen... |
2698 |
* @iocb: IO state structure (file, offset, etc.) |
8174202b3 write_iter varian... |
2699 |
* @from: iov_iter with data to write |
e4dd9de3c vfs: Export __gen... |
2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 |
* * 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... |
2713 |
ssize_t __generic_file_write_iter(struct kiocb *iocb, struct iov_iter *from) |
1da177e4c Linux-2.6.12-rc2 |
2714 2715 |
{ struct file *file = iocb->ki_filp; |
fb5527e68 [PATCH] direct-io... |
2716 |
struct address_space * mapping = file->f_mapping; |
1da177e4c Linux-2.6.12-rc2 |
2717 |
struct inode *inode = mapping->host; |
3b93f911d export generic_pe... |
2718 |
ssize_t written = 0; |
1da177e4c Linux-2.6.12-rc2 |
2719 |
ssize_t err; |
3b93f911d export generic_pe... |
2720 |
ssize_t status; |
1da177e4c Linux-2.6.12-rc2 |
2721 |
|
1da177e4c Linux-2.6.12-rc2 |
2722 |
/* We can write back this queue in page reclaim */ |
de1414a65 fs: export inode_... |
2723 |
current->backing_dev_info = inode_to_bdi(inode); |
5fa8e0a1c fs: Rename file_r... |
2724 |
err = file_remove_privs(file); |
1da177e4c Linux-2.6.12-rc2 |
2725 2726 |
if (err) goto out; |
c3b2da314 fs: introduce ino... |
2727 2728 2729 |
err = file_update_time(file); if (err) goto out; |
1da177e4c Linux-2.6.12-rc2 |
2730 |
|
2ba48ce51 mirror O_APPEND a... |
2731 |
if (iocb->ki_flags & IOCB_DIRECT) { |
0b8def9d6 __generic_file_wr... |
2732 |
loff_t pos, endbyte; |
fb5527e68 [PATCH] direct-io... |
2733 |
|
1af5bb491 filemap: remove t... |
2734 |
written = generic_file_direct_write(iocb, from); |
1da177e4c Linux-2.6.12-rc2 |
2735 |
/* |
fbbbad4bc vfs,ext2: introdu... |
2736 2737 2738 2739 2740 |
* 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 |
2741 |
*/ |
0b8def9d6 __generic_file_wr... |
2742 |
if (written < 0 || !iov_iter_count(from) || IS_DAX(inode)) |
fbbbad4bc vfs,ext2: introdu... |
2743 |
goto out; |
0b8def9d6 __generic_file_wr... |
2744 |
status = generic_perform_write(file, from, pos = iocb->ki_pos); |
fb5527e68 [PATCH] direct-io... |
2745 |
/* |
3b93f911d export generic_pe... |
2746 |
* If generic_perform_write() returned a synchronous error |
fb5527e68 [PATCH] direct-io... |
2747 2748 2749 2750 2751 |
* 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... |
2752 |
if (unlikely(status < 0)) { |
3b93f911d export generic_pe... |
2753 |
err = status; |
fb5527e68 [PATCH] direct-io... |
2754 2755 |
goto out; } |
fb5527e68 [PATCH] direct-io... |
2756 2757 2758 2759 2760 |
/* * 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... |
2761 |
endbyte = pos + status - 1; |
0b8def9d6 __generic_file_wr... |
2762 |
err = filemap_write_and_wait_range(mapping, pos, endbyte); |
fb5527e68 [PATCH] direct-io... |
2763 |
if (err == 0) { |
0b8def9d6 __generic_file_wr... |
2764 |
iocb->ki_pos = endbyte + 1; |
3b93f911d export generic_pe... |
2765 |
written += status; |
fb5527e68 [PATCH] direct-io... |
2766 |
invalidate_mapping_pages(mapping, |
09cbfeaf1 mm, fs: get rid o... |
2767 2768 |
pos >> PAGE_SHIFT, endbyte >> PAGE_SHIFT); |
fb5527e68 [PATCH] direct-io... |
2769 2770 2771 2772 2773 2774 2775 |
} 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... |
2776 2777 2778 |
written = generic_perform_write(file, from, iocb->ki_pos); if (likely(written > 0)) iocb->ki_pos += written; |
fb5527e68 [PATCH] direct-io... |
2779 |
} |
1da177e4c Linux-2.6.12-rc2 |
2780 2781 2782 2783 |
out: current->backing_dev_info = NULL; return written ? written : err; } |
8174202b3 write_iter varian... |
2784 |
EXPORT_SYMBOL(__generic_file_write_iter); |
e4dd9de3c vfs: Export __gen... |
2785 |
|
e4dd9de3c vfs: Export __gen... |
2786 |
/** |
8174202b3 write_iter varian... |
2787 |
* generic_file_write_iter - write data to a file |
e4dd9de3c vfs: Export __gen... |
2788 |
* @iocb: IO state structure |
8174202b3 write_iter varian... |
2789 |
* @from: iov_iter with data to write |
e4dd9de3c vfs: Export __gen... |
2790 |
* |
8174202b3 write_iter varian... |
2791 |
* This is a wrapper around __generic_file_write_iter() to be used by most |
e4dd9de3c vfs: Export __gen... |
2792 2793 2794 |
* filesystems. It takes care of syncing the file in case of O_SYNC file * and acquires i_mutex as needed. */ |
8174202b3 write_iter varian... |
2795 |
ssize_t generic_file_write_iter(struct kiocb *iocb, struct iov_iter *from) |
1da177e4c Linux-2.6.12-rc2 |
2796 2797 |
{ struct file *file = iocb->ki_filp; |
148f948ba vfs: Introduce ne... |
2798 |
struct inode *inode = file->f_mapping->host; |
1da177e4c Linux-2.6.12-rc2 |
2799 |
ssize_t ret; |
1da177e4c Linux-2.6.12-rc2 |
2800 |
|
5955102c9 wrappers for ->i_... |
2801 |
inode_lock(inode); |
3309dd04c switch generic_wr... |
2802 2803 |
ret = generic_write_checks(iocb, from); if (ret > 0) |
5f380c7fa lift generic_writ... |
2804 |
ret = __generic_file_write_iter(iocb, from); |
5955102c9 wrappers for ->i_... |
2805 |
inode_unlock(inode); |
1da177e4c Linux-2.6.12-rc2 |
2806 |
|
e25922176 fs: simplify the ... |
2807 2808 |
if (ret > 0) ret = generic_write_sync(iocb, ret); |
1da177e4c Linux-2.6.12-rc2 |
2809 2810 |
return ret; } |
8174202b3 write_iter varian... |
2811 |
EXPORT_SYMBOL(generic_file_write_iter); |
1da177e4c Linux-2.6.12-rc2 |
2812 |
|
cf9a2ae8d [PATCH] BLOCK: Mo... |
2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 |
/** * 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... |
2823 2824 2825 |
* 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... |
2826 |
* The @gfp_mask argument specifies whether I/O may be performed to release |
71baba4b9 mm, page_alloc: r... |
2827 |
* this page (__GFP_IO), and whether the call may block (__GFP_RECLAIM & __GFP_FS). |
cf9a2ae8d [PATCH] BLOCK: Mo... |
2828 |
* |
cf9a2ae8d [PATCH] BLOCK: Mo... |
2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 |
*/ 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); |