forked from swarren/WSL2-Linux-Kernel
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathfile.c
1592 lines (1449 loc) · 53.5 KB
/
file.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
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
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/*
* Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README
*/
#include <linux/time.h>
#include <linux/reiserfs_fs.h>
#include <linux/reiserfs_acl.h>
#include <linux/reiserfs_xattr.h>
#include <linux/smp_lock.h>
#include <asm/uaccess.h>
#include <linux/pagemap.h>
#include <linux/swap.h>
#include <linux/writeback.h>
#include <linux/blkdev.h>
#include <linux/buffer_head.h>
#include <linux/quotaops.h>
/*
** We pack the tails of files on file close, not at the time they are written.
** This implies an unnecessary copy of the tail and an unnecessary indirect item
** insertion/balancing, for files that are written in one write.
** It avoids unnecessary tail packings (balances) for files that are written in
** multiple writes and are small enough to have tails.
**
** file_release is called by the VFS layer when the file is closed. If
** this is the last open file descriptor, and the file
** small enough to have a tail, and the tail is currently in an
** unformatted node, the tail is converted back into a direct item.
**
** We use reiserfs_truncate_file to pack the tail, since it already has
** all the conditions coded.
*/
static int reiserfs_file_release(struct inode *inode, struct file *filp)
{
struct reiserfs_transaction_handle th;
int err;
int jbegin_failure = 0;
BUG_ON(!S_ISREG(inode->i_mode));
/* fast out for when nothing needs to be done */
if ((atomic_read(&inode->i_count) > 1 ||
!(REISERFS_I(inode)->i_flags & i_pack_on_close_mask) ||
!tail_has_to_be_packed(inode)) &&
REISERFS_I(inode)->i_prealloc_count <= 0) {
return 0;
}
mutex_lock(&inode->i_mutex);
reiserfs_write_lock(inode->i_sb);
/* freeing preallocation only involves relogging blocks that
* are already in the current transaction. preallocation gets
* freed at the end of each transaction, so it is impossible for
* us to log any additional blocks (including quota blocks)
*/
err = journal_begin(&th, inode->i_sb, 1);
if (err) {
/* uh oh, we can't allow the inode to go away while there
* is still preallocation blocks pending. Try to join the
* aborted transaction
*/
jbegin_failure = err;
err = journal_join_abort(&th, inode->i_sb, 1);
if (err) {
/* hmpf, our choices here aren't good. We can pin the inode
* which will disallow unmount from every happening, we can
* do nothing, which will corrupt random memory on unmount,
* or we can forcibly remove the file from the preallocation
* list, which will leak blocks on disk. Lets pin the inode
* and let the admin know what is going on.
*/
igrab(inode);
reiserfs_warning(inode->i_sb,
"pinning inode %lu because the "
"preallocation can't be freed",
inode->i_ino);
goto out;
}
}
reiserfs_update_inode_transaction(inode);
#ifdef REISERFS_PREALLOCATE
reiserfs_discard_prealloc(&th, inode);
#endif
err = journal_end(&th, inode->i_sb, 1);
/* copy back the error code from journal_begin */
if (!err)
err = jbegin_failure;
if (!err && atomic_read(&inode->i_count) <= 1 &&
(REISERFS_I(inode)->i_flags & i_pack_on_close_mask) &&
tail_has_to_be_packed(inode)) {
/* if regular file is released by last holder and it has been
appended (we append by unformatted node only) or its direct
item(s) had to be converted, then it may have to be
indirect2direct converted */
err = reiserfs_truncate_file(inode, 0);
}
out:
mutex_unlock(&inode->i_mutex);
reiserfs_write_unlock(inode->i_sb);
return err;
}
static void reiserfs_vfs_truncate_file(struct inode *inode)
{
reiserfs_truncate_file(inode, 1);
}
/* Sync a reiserfs file. */
/*
* FIXME: sync_mapping_buffers() never has anything to sync. Can
* be removed...
*/
static int reiserfs_sync_file(struct file *p_s_filp,
struct dentry *p_s_dentry, int datasync)
{
struct inode *p_s_inode = p_s_dentry->d_inode;
int n_err;
int barrier_done;
BUG_ON(!S_ISREG(p_s_inode->i_mode));
n_err = sync_mapping_buffers(p_s_inode->i_mapping);
reiserfs_write_lock(p_s_inode->i_sb);
barrier_done = reiserfs_commit_for_inode(p_s_inode);
reiserfs_write_unlock(p_s_inode->i_sb);
if (barrier_done != 1 && reiserfs_barrier_flush(p_s_inode->i_sb))
blkdev_issue_flush(p_s_inode->i_sb->s_bdev, NULL);
if (barrier_done < 0)
return barrier_done;
return (n_err < 0) ? -EIO : 0;
}
/* I really do not want to play with memory shortage right now, so
to simplify the code, we are not going to write more than this much pages at
a time. This still should considerably improve performance compared to 4k
at a time case. This is 32 pages of 4k size. */
#define REISERFS_WRITE_PAGES_AT_A_TIME (128 * 1024) / PAGE_CACHE_SIZE
/* Allocates blocks for a file to fulfil write request.
Maps all unmapped but prepared pages from the list.
Updates metadata with newly allocated blocknumbers as needed */
static int reiserfs_allocate_blocks_for_region(struct reiserfs_transaction_handle *th, struct inode *inode, /* Inode we work with */
loff_t pos, /* Writing position */
int num_pages, /* number of pages write going
to touch */
int write_bytes, /* amount of bytes to write */
struct page **prepared_pages, /* array of
prepared pages
*/
int blocks_to_allocate /* Amount of blocks we
need to allocate to
fit the data into file
*/
)
{
struct cpu_key key; // cpu key of item that we are going to deal with
struct item_head *ih; // pointer to item head that we are going to deal with
struct buffer_head *bh; // Buffer head that contains items that we are going to deal with
__le32 *item; // pointer to item we are going to deal with
INITIALIZE_PATH(path); // path to item, that we are going to deal with.
b_blocknr_t *allocated_blocks; // Pointer to a place where allocated blocknumbers would be stored.
reiserfs_blocknr_hint_t hint; // hint structure for block allocator.
size_t res; // return value of various functions that we call.
int curr_block; // current block used to keep track of unmapped blocks.
int i; // loop counter
int itempos; // position in item
unsigned int from = (pos & (PAGE_CACHE_SIZE - 1)); // writing position in
// first page
unsigned int to = ((pos + write_bytes - 1) & (PAGE_CACHE_SIZE - 1)) + 1; /* last modified byte offset in last page */
__u64 hole_size; // amount of blocks for a file hole, if it needed to be created.
int modifying_this_item = 0; // Flag for items traversal code to keep track
// of the fact that we already prepared
// current block for journal
int will_prealloc = 0;
RFALSE(!blocks_to_allocate,
"green-9004: tried to allocate zero blocks?");
/* only preallocate if this is a small write */
if (REISERFS_I(inode)->i_prealloc_count ||
(!(write_bytes & (inode->i_sb->s_blocksize - 1)) &&
blocks_to_allocate <
REISERFS_SB(inode->i_sb)->s_alloc_options.preallocsize))
will_prealloc =
REISERFS_SB(inode->i_sb)->s_alloc_options.preallocsize;
allocated_blocks = kmalloc((blocks_to_allocate + will_prealloc) *
sizeof(b_blocknr_t), GFP_NOFS);
if (!allocated_blocks)
return -ENOMEM;
/* First we compose a key to point at the writing position, we want to do
that outside of any locking region. */
make_cpu_key(&key, inode, pos + 1, TYPE_ANY, 3 /*key length */ );
/* If we came here, it means we absolutely need to open a transaction,
since we need to allocate some blocks */
reiserfs_write_lock(inode->i_sb); // Journaling stuff and we need that.
res = journal_begin(th, inode->i_sb, JOURNAL_PER_BALANCE_CNT * 3 + 1 + 2 * REISERFS_QUOTA_TRANS_BLOCKS(inode->i_sb)); // Wish I know if this number enough
if (res)
goto error_exit;
reiserfs_update_inode_transaction(inode);
/* Look for the in-tree position of our write, need path for block allocator */
res = search_for_position_by_key(inode->i_sb, &key, &path);
if (res == IO_ERROR) {
res = -EIO;
goto error_exit;
}
/* Allocate blocks */
/* First fill in "hint" structure for block allocator */
hint.th = th; // transaction handle.
hint.path = &path; // Path, so that block allocator can determine packing locality or whatever it needs to determine.
hint.inode = inode; // Inode is needed by block allocator too.
hint.search_start = 0; // We have no hint on where to search free blocks for block allocator.
hint.key = key.on_disk_key; // on disk key of file.
hint.block = inode->i_blocks >> (inode->i_sb->s_blocksize_bits - 9); // Number of disk blocks this file occupies already.
hint.formatted_node = 0; // We are allocating blocks for unformatted node.
hint.preallocate = will_prealloc;
/* Call block allocator to allocate blocks */
res =
reiserfs_allocate_blocknrs(&hint, allocated_blocks,
blocks_to_allocate, blocks_to_allocate);
if (res != CARRY_ON) {
if (res == NO_DISK_SPACE) {
/* We flush the transaction in case of no space. This way some
blocks might become free */
SB_JOURNAL(inode->i_sb)->j_must_wait = 1;
res = restart_transaction(th, inode, &path);
if (res)
goto error_exit;
/* We might have scheduled, so search again */
res =
search_for_position_by_key(inode->i_sb, &key,
&path);
if (res == IO_ERROR) {
res = -EIO;
goto error_exit;
}
/* update changed info for hint structure. */
res =
reiserfs_allocate_blocknrs(&hint, allocated_blocks,
blocks_to_allocate,
blocks_to_allocate);
if (res != CARRY_ON) {
res = res == QUOTA_EXCEEDED ? -EDQUOT : -ENOSPC;
pathrelse(&path);
goto error_exit;
}
} else {
res = res == QUOTA_EXCEEDED ? -EDQUOT : -ENOSPC;
pathrelse(&path);
goto error_exit;
}
}
#ifdef __BIG_ENDIAN
// Too bad, I have not found any way to convert a given region from
// cpu format to little endian format
{
int i;
for (i = 0; i < blocks_to_allocate; i++)
allocated_blocks[i] = cpu_to_le32(allocated_blocks[i]);
}
#endif
/* Blocks allocating well might have scheduled and tree might have changed,
let's search the tree again */
/* find where in the tree our write should go */
res = search_for_position_by_key(inode->i_sb, &key, &path);
if (res == IO_ERROR) {
res = -EIO;
goto error_exit_free_blocks;
}
bh = get_last_bh(&path); // Get a bufferhead for last element in path.
ih = get_ih(&path); // Get a pointer to last item head in path.
item = get_item(&path); // Get a pointer to last item in path
/* Let's see what we have found */
if (res != POSITION_FOUND) { /* position not found, this means that we
might need to append file with holes
first */
// Since we are writing past the file's end, we need to find out if
// there is a hole that needs to be inserted before our writing
// position, and how many blocks it is going to cover (we need to
// populate pointers to file blocks representing the hole with zeros)
{
int item_offset = 1;
/*
* if ih is stat data, its offset is 0 and we don't want to
* add 1 to pos in the hole_size calculation
*/
if (is_statdata_le_ih(ih))
item_offset = 0;
hole_size = (pos + item_offset -
(le_key_k_offset
(get_inode_item_key_version(inode),
&(ih->ih_key)) + op_bytes_number(ih,
inode->
i_sb->
s_blocksize)))
>> inode->i_sb->s_blocksize_bits;
}
if (hole_size > 0) {
int to_paste = min_t(__u64, hole_size, MAX_ITEM_LEN(inode->i_sb->s_blocksize) / UNFM_P_SIZE); // How much data to insert first time.
/* area filled with zeroes, to supply as list of zero blocknumbers
We allocate it outside of loop just in case loop would spin for
several iterations. */
char *zeros = kmalloc(to_paste * UNFM_P_SIZE, GFP_ATOMIC); // We cannot insert more than MAX_ITEM_LEN bytes anyway.
if (!zeros) {
res = -ENOMEM;
goto error_exit_free_blocks;
}
memset(zeros, 0, to_paste * UNFM_P_SIZE);
do {
to_paste =
min_t(__u64, hole_size,
MAX_ITEM_LEN(inode->i_sb->
s_blocksize) /
UNFM_P_SIZE);
if (is_indirect_le_ih(ih)) {
/* Ok, there is existing indirect item already. Need to append it */
/* Calculate position past inserted item */
make_cpu_key(&key, inode,
le_key_k_offset
(get_inode_item_key_version
(inode),
&(ih->ih_key)) +
op_bytes_number(ih,
inode->
i_sb->
s_blocksize),
TYPE_INDIRECT, 3);
res =
reiserfs_paste_into_item(th, &path,
&key,
inode,
(char *)
zeros,
UNFM_P_SIZE
*
to_paste);
if (res) {
kfree(zeros);
goto error_exit_free_blocks;
}
} else if (is_statdata_le_ih(ih)) {
/* No existing item, create it */
/* item head for new item */
struct item_head ins_ih;
/* create a key for our new item */
make_cpu_key(&key, inode, 1,
TYPE_INDIRECT, 3);
/* Create new item head for our new item */
make_le_item_head(&ins_ih, &key,
key.version, 1,
TYPE_INDIRECT,
to_paste *
UNFM_P_SIZE,
0 /* free space */ );
/* Find where such item should live in the tree */
res =
search_item(inode->i_sb, &key,
&path);
if (res != ITEM_NOT_FOUND) {
/* item should not exist, otherwise we have error */
if (res != -ENOSPC) {
reiserfs_warning(inode->
i_sb,
"green-9008: search_by_key (%K) returned %d",
&key,
res);
}
res = -EIO;
kfree(zeros);
goto error_exit_free_blocks;
}
res =
reiserfs_insert_item(th, &path,
&key, &ins_ih,
inode,
(char *)zeros);
} else {
reiserfs_panic(inode->i_sb,
"green-9011: Unexpected key type %K\n",
&key);
}
if (res) {
kfree(zeros);
goto error_exit_free_blocks;
}
/* Now we want to check if transaction is too full, and if it is
we restart it. This will also free the path. */
if (journal_transaction_should_end
(th, th->t_blocks_allocated)) {
res =
restart_transaction(th, inode,
&path);
if (res) {
pathrelse(&path);
kfree(zeros);
goto error_exit;
}
}
/* Well, need to recalculate path and stuff */
set_cpu_key_k_offset(&key,
cpu_key_k_offset(&key) +
(to_paste << inode->
i_blkbits));
res =
search_for_position_by_key(inode->i_sb,
&key, &path);
if (res == IO_ERROR) {
res = -EIO;
kfree(zeros);
goto error_exit_free_blocks;
}
bh = get_last_bh(&path);
ih = get_ih(&path);
item = get_item(&path);
hole_size -= to_paste;
} while (hole_size);
kfree(zeros);
}
}
// Go through existing indirect items first
// replace all zeroes with blocknumbers from list
// Note that if no corresponding item was found, by previous search,
// it means there are no existing in-tree representation for file area
// we are going to overwrite, so there is nothing to scan through for holes.
for (curr_block = 0, itempos = path.pos_in_item;
curr_block < blocks_to_allocate && res == POSITION_FOUND;) {
retry:
if (itempos >= ih_item_len(ih) / UNFM_P_SIZE) {
/* We run out of data in this indirect item, let's look for another
one. */
/* First if we are already modifying current item, log it */
if (modifying_this_item) {
journal_mark_dirty(th, inode->i_sb, bh);
modifying_this_item = 0;
}
/* Then set the key to look for a new indirect item (offset of old
item is added to old item length */
set_cpu_key_k_offset(&key,
le_key_k_offset
(get_inode_item_key_version(inode),
&(ih->ih_key)) +
op_bytes_number(ih,
inode->i_sb->
s_blocksize));
/* Search ofor position of new key in the tree. */
res =
search_for_position_by_key(inode->i_sb, &key,
&path);
if (res == IO_ERROR) {
res = -EIO;
goto error_exit_free_blocks;
}
bh = get_last_bh(&path);
ih = get_ih(&path);
item = get_item(&path);
itempos = path.pos_in_item;
continue; // loop to check all kinds of conditions and so on.
}
/* Ok, we have correct position in item now, so let's see if it is
representing file hole (blocknumber is zero) and fill it if needed */
if (!item[itempos]) {
/* Ok, a hole. Now we need to check if we already prepared this
block to be journaled */
while (!modifying_this_item) { // loop until succeed
/* Well, this item is not journaled yet, so we must prepare
it for journal first, before we can change it */
struct item_head tmp_ih; // We copy item head of found item,
// here to detect if fs changed under
// us while we were preparing for
// journal.
int fs_gen; // We store fs generation here to find if someone
// changes fs under our feet
copy_item_head(&tmp_ih, ih); // Remember itemhead
fs_gen = get_generation(inode->i_sb); // remember fs generation
reiserfs_prepare_for_journal(inode->i_sb, bh, 1); // Prepare a buffer within which indirect item is stored for changing.
if (fs_changed(fs_gen, inode->i_sb)
&& item_moved(&tmp_ih, &path)) {
// Sigh, fs was changed under us, we need to look for new
// location of item we are working with
/* unmark prepaerd area as journaled and search for it's
new position */
reiserfs_restore_prepared_buffer(inode->
i_sb,
bh);
res =
search_for_position_by_key(inode->
i_sb,
&key,
&path);
if (res == IO_ERROR) {
res = -EIO;
goto error_exit_free_blocks;
}
bh = get_last_bh(&path);
ih = get_ih(&path);
item = get_item(&path);
itempos = path.pos_in_item;
goto retry;
}
modifying_this_item = 1;
}
item[itempos] = allocated_blocks[curr_block]; // Assign new block
curr_block++;
}
itempos++;
}
if (modifying_this_item) { // We need to log last-accessed block, if it
// was modified, but not logged yet.
journal_mark_dirty(th, inode->i_sb, bh);
}
if (curr_block < blocks_to_allocate) {
// Oh, well need to append to indirect item, or to create indirect item
// if there weren't any
if (is_indirect_le_ih(ih)) {
// Existing indirect item - append. First calculate key for append
// position. We do not need to recalculate path as it should
// already point to correct place.
make_cpu_key(&key, inode,
le_key_k_offset(get_inode_item_key_version
(inode),
&(ih->ih_key)) +
op_bytes_number(ih,
inode->i_sb->s_blocksize),
TYPE_INDIRECT, 3);
res =
reiserfs_paste_into_item(th, &path, &key, inode,
(char *)(allocated_blocks +
curr_block),
UNFM_P_SIZE *
(blocks_to_allocate -
curr_block));
if (res) {
goto error_exit_free_blocks;
}
} else if (is_statdata_le_ih(ih)) {
// Last found item was statdata. That means we need to create indirect item.
struct item_head ins_ih; /* itemhead for new item */
/* create a key for our new item */
make_cpu_key(&key, inode, 1, TYPE_INDIRECT, 3); // Position one,
// because that's
// where first
// indirect item
// begins
/* Create new item head for our new item */
make_le_item_head(&ins_ih, &key, key.version, 1,
TYPE_INDIRECT,
(blocks_to_allocate -
curr_block) * UNFM_P_SIZE,
0 /* free space */ );
/* Find where such item should live in the tree */
res = search_item(inode->i_sb, &key, &path);
if (res != ITEM_NOT_FOUND) {
/* Well, if we have found such item already, or some error
occured, we need to warn user and return error */
if (res != -ENOSPC) {
reiserfs_warning(inode->i_sb,
"green-9009: search_by_key (%K) "
"returned %d", &key,
res);
}
res = -EIO;
goto error_exit_free_blocks;
}
/* Insert item into the tree with the data as its body */
res =
reiserfs_insert_item(th, &path, &key, &ins_ih,
inode,
(char *)(allocated_blocks +
curr_block));
} else {
reiserfs_panic(inode->i_sb,
"green-9010: unexpected item type for key %K\n",
&key);
}
}
// the caller is responsible for closing the transaction
// unless we return an error, they are also responsible for logging
// the inode.
//
pathrelse(&path);
/*
* cleanup prellocation from previous writes
* if this is a partial block write
*/
if (write_bytes & (inode->i_sb->s_blocksize - 1))
reiserfs_discard_prealloc(th, inode);
reiserfs_write_unlock(inode->i_sb);
// go through all the pages/buffers and map the buffers to newly allocated
// blocks (so that system knows where to write these pages later).
curr_block = 0;
for (i = 0; i < num_pages; i++) {
struct page *page = prepared_pages[i]; //current page
struct buffer_head *head = page_buffers(page); // first buffer for a page
int block_start, block_end; // in-page offsets for buffers.
if (!page_buffers(page))
reiserfs_panic(inode->i_sb,
"green-9005: No buffers for prepared page???");
/* For each buffer in page */
for (bh = head, block_start = 0; bh != head || !block_start;
block_start = block_end, bh = bh->b_this_page) {
if (!bh)
reiserfs_panic(inode->i_sb,
"green-9006: Allocated but absent buffer for a page?");
block_end = block_start + inode->i_sb->s_blocksize;
if (i == 0 && block_end <= from)
/* if this buffer is before requested data to map, skip it */
continue;
if (i == num_pages - 1 && block_start >= to)
/* If this buffer is after requested data to map, abort
processing of current page */
break;
if (!buffer_mapped(bh)) { // Ok, unmapped buffer, need to map it
map_bh(bh, inode->i_sb,
le32_to_cpu(allocated_blocks
[curr_block]));
curr_block++;
set_buffer_new(bh);
}
}
}
RFALSE(curr_block > blocks_to_allocate,
"green-9007: Used too many blocks? weird");
kfree(allocated_blocks);
return 0;
// Need to deal with transaction here.
error_exit_free_blocks:
pathrelse(&path);
// free blocks
for (i = 0; i < blocks_to_allocate; i++)
reiserfs_free_block(th, inode, le32_to_cpu(allocated_blocks[i]),
1);
error_exit:
if (th->t_trans_id) {
int err;
// update any changes we made to blk count
mark_inode_dirty(inode);
err =
journal_end(th, inode->i_sb,
JOURNAL_PER_BALANCE_CNT * 3 + 1 +
2 * REISERFS_QUOTA_TRANS_BLOCKS(inode->i_sb));
if (err)
res = err;
}
reiserfs_write_unlock(inode->i_sb);
kfree(allocated_blocks);
return res;
}
/* Unlock pages prepared by reiserfs_prepare_file_region_for_write */
static void reiserfs_unprepare_pages(struct page **prepared_pages, /* list of locked pages */
size_t num_pages /* amount of pages */ )
{
int i; // loop counter
for (i = 0; i < num_pages; i++) {
struct page *page = prepared_pages[i];
try_to_free_buffers(page);
unlock_page(page);
page_cache_release(page);
}
}
/* This function will copy data from userspace to specified pages within
supplied byte range */
static int reiserfs_copy_from_user_to_file_region(loff_t pos, /* In-file position */
int num_pages, /* Number of pages affected */
int write_bytes, /* Amount of bytes to write */
struct page **prepared_pages, /* pointer to
array to
prepared pages
*/
const char __user * buf /* Pointer to user-supplied
data */
)
{
long page_fault = 0; // status of copy_from_user.
int i; // loop counter.
int offset; // offset in page
for (i = 0, offset = (pos & (PAGE_CACHE_SIZE - 1)); i < num_pages;
i++, offset = 0) {
size_t count = min_t(size_t, PAGE_CACHE_SIZE - offset, write_bytes); // How much of bytes to write to this page
struct page *page = prepared_pages[i]; // Current page we process.
fault_in_pages_readable(buf, count);
/* Copy data from userspace to the current page */
kmap(page);
page_fault = __copy_from_user(page_address(page) + offset, buf, count); // Copy the data.
/* Flush processor's dcache for this page */
flush_dcache_page(page);
kunmap(page);
buf += count;
write_bytes -= count;
if (page_fault)
break; // Was there a fault? abort.
}
return page_fault ? -EFAULT : 0;
}
/* taken fs/buffer.c:__block_commit_write */
int reiserfs_commit_page(struct inode *inode, struct page *page,
unsigned from, unsigned to)
{
unsigned block_start, block_end;
int partial = 0;
unsigned blocksize;
struct buffer_head *bh, *head;
unsigned long i_size_index = inode->i_size >> PAGE_CACHE_SHIFT;
int new;
int logit = reiserfs_file_data_log(inode);
struct super_block *s = inode->i_sb;
int bh_per_page = PAGE_CACHE_SIZE / s->s_blocksize;
struct reiserfs_transaction_handle th;
int ret = 0;
th.t_trans_id = 0;
blocksize = 1 << inode->i_blkbits;
if (logit) {
reiserfs_write_lock(s);
ret = journal_begin(&th, s, bh_per_page + 1);
if (ret)
goto drop_write_lock;
reiserfs_update_inode_transaction(inode);
}
for (bh = head = page_buffers(page), block_start = 0;
bh != head || !block_start;
block_start = block_end, bh = bh->b_this_page) {
new = buffer_new(bh);
clear_buffer_new(bh);
block_end = block_start + blocksize;
if (block_end <= from || block_start >= to) {
if (!buffer_uptodate(bh))
partial = 1;
} else {
set_buffer_uptodate(bh);
if (logit) {
reiserfs_prepare_for_journal(s, bh, 1);
journal_mark_dirty(&th, s, bh);
} else if (!buffer_dirty(bh)) {
mark_buffer_dirty(bh);
/* do data=ordered on any page past the end
* of file and any buffer marked BH_New.
*/
if (reiserfs_data_ordered(inode->i_sb) &&
(new || page->index >= i_size_index)) {
reiserfs_add_ordered_list(inode, bh);
}
}
}
}
if (logit) {
ret = journal_end(&th, s, bh_per_page + 1);
drop_write_lock:
reiserfs_write_unlock(s);
}
/*
* If this is a partial write which happened to make all buffers
* uptodate then we can optimize away a bogus readpage() for
* the next read(). Here we 'discover' whether the page went
* uptodate as a result of this (potentially partial) write.
*/
if (!partial)
SetPageUptodate(page);
return ret;
}
/* Submit pages for write. This was separated from actual file copying
because we might want to allocate block numbers in-between.
This function assumes that caller will adjust file size to correct value. */
static int reiserfs_submit_file_region_for_write(struct reiserfs_transaction_handle *th, struct inode *inode, loff_t pos, /* Writing position offset */
size_t num_pages, /* Number of pages to write */
size_t write_bytes, /* number of bytes to write */
struct page **prepared_pages /* list of pages */
)
{
int status; // return status of block_commit_write.
int retval = 0; // Return value we are going to return.
int i; // loop counter
int offset; // Writing offset in page.
int orig_write_bytes = write_bytes;
int sd_update = 0;
for (i = 0, offset = (pos & (PAGE_CACHE_SIZE - 1)); i < num_pages;
i++, offset = 0) {
int count = min_t(int, PAGE_CACHE_SIZE - offset, write_bytes); // How much of bytes to write to this page
struct page *page = prepared_pages[i]; // Current page we process.
status =
reiserfs_commit_page(inode, page, offset, offset + count);
if (status)
retval = status; // To not overcomplicate matters We are going to
// submit all the pages even if there was error.
// we only remember error status to report it on
// exit.
write_bytes -= count;
}
/* now that we've gotten all the ordered buffers marked dirty,
* we can safely update i_size and close any running transaction
*/
if (pos + orig_write_bytes > inode->i_size) {
inode->i_size = pos + orig_write_bytes; // Set new size
/* If the file have grown so much that tail packing is no
* longer possible, reset "need to pack" flag */
if ((have_large_tails(inode->i_sb) &&
inode->i_size > i_block_size(inode) * 4) ||
(have_small_tails(inode->i_sb) &&
inode->i_size > i_block_size(inode)))
REISERFS_I(inode)->i_flags &= ~i_pack_on_close_mask;
else if ((have_large_tails(inode->i_sb) &&
inode->i_size < i_block_size(inode) * 4) ||
(have_small_tails(inode->i_sb) &&
inode->i_size < i_block_size(inode)))
REISERFS_I(inode)->i_flags |= i_pack_on_close_mask;
if (th->t_trans_id) {
reiserfs_write_lock(inode->i_sb);
// this sets the proper flags for O_SYNC to trigger a commit
mark_inode_dirty(inode);
reiserfs_write_unlock(inode->i_sb);
} else {
reiserfs_write_lock(inode->i_sb);
reiserfs_update_inode_transaction(inode);
mark_inode_dirty(inode);
reiserfs_write_unlock(inode->i_sb);
}
sd_update = 1;
}
if (th->t_trans_id) {
reiserfs_write_lock(inode->i_sb);
if (!sd_update)
mark_inode_dirty(inode);
status = journal_end(th, th->t_super, th->t_blocks_allocated);
if (status)
retval = status;
reiserfs_write_unlock(inode->i_sb);
}
th->t_trans_id = 0;
/*
* we have to unlock the pages after updating i_size, otherwise
* we race with writepage
*/
for (i = 0; i < num_pages; i++) {
struct page *page = prepared_pages[i];
unlock_page(page);
mark_page_accessed(page);
page_cache_release(page);
}
return retval;
}
/* Look if passed writing region is going to touch file's tail
(if it is present). And if it is, convert the tail to unformatted node */
static int reiserfs_check_for_tail_and_convert(struct inode *inode, /* inode to deal with */
loff_t pos, /* Writing position */
int write_bytes /* amount of bytes to write */
)
{
INITIALIZE_PATH(path); // needed for search_for_position
struct cpu_key key; // Key that would represent last touched writing byte.
struct item_head *ih; // item header of found block;
int res; // Return value of various functions we call.
int cont_expand_offset; // We will put offset for generic_cont_expand here
// This can be int just because tails are created
// only for small files.
/* this embodies a dependency on a particular tail policy */
if (inode->i_size >= inode->i_sb->s_blocksize * 4) {
/* such a big files do not have tails, so we won't bother ourselves
to look for tails, simply return */
return 0;
}
reiserfs_write_lock(inode->i_sb);
/* find the item containing the last byte to be written, or if
* writing past the end of the file then the last item of the
* file (and then we check its type). */
make_cpu_key(&key, inode, pos + write_bytes + 1, TYPE_ANY,
3 /*key length */ );
res = search_for_position_by_key(inode->i_sb, &key, &path);
if (res == IO_ERROR) {
reiserfs_write_unlock(inode->i_sb);
return -EIO;
}
ih = get_ih(&path);
res = 0;
if (is_direct_le_ih(ih)) {
/* Ok, closest item is file tail (tails are stored in "direct"
* items), so we need to unpack it. */
/* To not overcomplicate matters, we just call generic_cont_expand
which will in turn call other stuff and finally will boil down to
reiserfs_get_block() that would do necessary conversion. */
cont_expand_offset =
le_key_k_offset(get_inode_item_key_version(inode),
&(ih->ih_key));
pathrelse(&path);
res = generic_cont_expand(inode, cont_expand_offset);
} else
pathrelse(&path);
reiserfs_write_unlock(inode->i_sb);
return res;
}
/* This function locks pages starting from @pos for @inode.
@num_pages pages are locked and stored in
@prepared_pages array. Also buffers are allocated for these pages.
First and last page of the region is read if it is overwritten only
partially. If last page did not exist before write (file hole or file
append), it is zeroed, then.
Returns number of unallocated blocks that should be allocated to cover
new file data.*/
static int reiserfs_prepare_file_region_for_write(struct inode *inode
/* Inode of the file */ ,
loff_t pos, /* position in the file */
size_t num_pages, /* number of pages to
prepare */
size_t write_bytes, /* Amount of bytes to be
overwritten from
@pos */
struct page **prepared_pages /* pointer to array
where to store
prepared pages */
)
{
int res = 0; // Return values of different functions we call.
unsigned long index = pos >> PAGE_CACHE_SHIFT; // Offset in file in pages.
int from = (pos & (PAGE_CACHE_SIZE - 1)); // Writing offset in first page
int to = ((pos + write_bytes - 1) & (PAGE_CACHE_SIZE - 1)) + 1;
/* offset of last modified byte in last
page */
struct address_space *mapping = inode->i_mapping; // Pages are mapped here.
int i; // Simple counter
int blocks = 0; /* Return value (blocks that should be allocated) */
struct buffer_head *bh, *head; // Current bufferhead and first bufferhead
// of a page.
unsigned block_start, block_end; // Starting and ending offsets of current
// buffer in the page.
struct buffer_head *wait[2], **wait_bh = wait; // Buffers for page, if
// Page appeared to be not up
// to date. Note how we have
// at most 2 buffers, this is
// because we at most may
// partially overwrite two
// buffers for one page. One at // the beginning of write area
// and one at the end.
// Everything inthe middle gets // overwritten totally.
struct cpu_key key; // cpu key of item that we are going to deal with
struct item_head *ih = NULL; // pointer to item head that we are going to deal with
struct buffer_head *itembuf = NULL; // Buffer head that contains items that we are going to deal with
INITIALIZE_PATH(path); // path to item, that we are going to deal with.
__le32 *item = NULL; // pointer to item we are going to deal with
int item_pos = -1; /* Position in indirect item */
if (num_pages < 1) {
reiserfs_warning(inode->i_sb,