1 // SPDX-License-Identifier: GPL-2.0-or-later
2 /*
3 * Squashfs - a compressed read only filesystem for Linux
4 *
5 * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008
6 * Phillip Lougher <phillip@squashfs.org.uk>
7 *
8 * file.c
9 */
10
11 /*
12 * This file contains code for handling regular files. A regular file
13 * consists of a sequence of contiguous compressed blocks, and/or a
14 * compressed fragment block (tail-end packed block). The compressed size
15 * of each datablock is stored in a block list contained within the
16 * file inode (itself stored in one or more compressed metadata blocks).
17 *
18 * To speed up access to datablocks when reading 'large' files (256 Mbytes or
19 * larger), the code implements an index cache that caches the mapping from
20 * block index to datablock location on disk.
21 *
22 * The index cache allows Squashfs to handle large files (up to 1.75 TiB) while
23 * retaining a simple and space-efficient block list on disk. The cache
24 * is split into slots, caching up to eight 224 GiB files (128 KiB blocks).
25 * Larger files use multiple slots, with 1.75 TiB files using all 8 slots.
26 * The index cache is designed to be memory efficient, and by default uses
27 * 16 KiB.
28 */
29
30 #include <linux/fs.h>
31 #include <linux/vfs.h>
32 #include <linux/kernel.h>
33 #include <linux/slab.h>
34 #include <linux/string.h>
35 #include <linux/pagemap.h>
36 #include <linux/mutex.h>
37
38 #include "squashfs_fs.h"
39 #include "squashfs_fs_sb.h"
40 #include "squashfs_fs_i.h"
41 #include "squashfs.h"
42 #include "page_actor.h"
43
44 /*
45 * Locate cache slot in range [offset, index] for specified inode. If
46 * there's more than one return the slot closest to index.
47 */
48 static struct meta_index *locate_meta_index(struct inode *inode, int offset,
49 int index)
50 {
51 struct meta_index *meta = NULL;
52 struct squashfs_sb_info *msblk = inode->i_sb->s_fs_info;
53 int i;
54
55 mutex_lock(&msblk->meta_index_mutex);
56
57 TRACE("locate_meta_index: index %d, offset %d\n", index, offset);
58
59 if (msblk->meta_index == NULL)
60 goto not_allocated;
61
62 for (i = 0; i < SQUASHFS_META_SLOTS; i++) {
63 if (msblk->meta_index[i].inode_number == inode->i_ino &&
64 msblk->meta_index[i].offset >= offset &&
65 msblk->meta_index[i].offset <= index &&
66 msblk->meta_index[i].locked == 0) {
67 TRACE("locate_meta_index: entry %d, offset %d\n", i,
68 msblk->meta_index[i].offset);
69 meta = &msblk->meta_index[i];
70 offset = meta->offset;
71 }
72 }
73
74 if (meta)
75 meta->locked = 1;
76
77 not_allocated:
78 mutex_unlock(&msblk->meta_index_mutex);
79
80 return meta;
81 }
82
83
84 /*
85 * Find and initialise an empty cache slot for index offset.
86 */
87 static struct meta_index *empty_meta_index(struct inode *inode, int offset,
88 int skip)
89 {
90 struct squashfs_sb_info *msblk = inode->i_sb->s_fs_info;
91 struct meta_index *meta = NULL;
92 int i;
93
94 mutex_lock(&msblk->meta_index_mutex);
95
96 TRACE("empty_meta_index: offset %d, skip %d\n", offset, skip);
97
98 if (msblk->meta_index == NULL) {
99 /*
100 * First time cache index has been used, allocate and
101 * initialise. The cache index could be allocated at
102 * mount time but doing it here means it is allocated only
103 * if a 'large' file is read.
104 */
105 msblk->meta_index = kcalloc(SQUASHFS_META_SLOTS,
106 sizeof(*(msblk->meta_index)), GFP_KERNEL);
107 if (msblk->meta_index == NULL) {
108 ERROR("Failed to allocate meta_index\n");
109 goto failed;
110 }
111 for (i = 0; i < SQUASHFS_META_SLOTS; i++) {
112 msblk->meta_index[i].inode_number = 0;
113 msblk->meta_index[i].locked = 0;
114 }
115 msblk->next_meta_index = 0;
116 }
117
118 for (i = SQUASHFS_META_SLOTS; i &&
119 msblk->meta_index[msblk->next_meta_index].locked; i--)
120 msblk->next_meta_index = (msblk->next_meta_index + 1) %
121 SQUASHFS_META_SLOTS;
122
123 if (i == 0) {
124 TRACE("empty_meta_index: failed!\n");
125 goto failed;
126 }
127
128 TRACE("empty_meta_index: returned meta entry %d, %p\n",
129 msblk->next_meta_index,
130 &msblk->meta_index[msblk->next_meta_index]);
131
132 meta = &msblk->meta_index[msblk->next_meta_index];
133 msblk->next_meta_index = (msblk->next_meta_index + 1) %
134 SQUASHFS_META_SLOTS;
135
136 meta->inode_number = inode->i_ino;
137 meta->offset = offset;
138 meta->skip = skip;
139 meta->entries = 0;
140 meta->locked = 1;
141
142 failed:
143 mutex_unlock(&msblk->meta_index_mutex);
144 return meta;
145 }
146
147
148 static void release_meta_index(struct inode *inode, struct meta_index *meta)
149 {
150 struct squashfs_sb_info *msblk = inode->i_sb->s_fs_info;
151 mutex_lock(&msblk->meta_index_mutex);
152 meta->locked = 0;
153 mutex_unlock(&msblk->meta_index_mutex);
154 }
155
156
157 /*
158 * Read the next n blocks from the block list, starting from
159 * metadata block <start_block, offset>.
160 */
161 static long long read_indexes(struct super_block *sb, int n,
162 u64 *start_block, int *offset)
163 {
164 int err, i;
165 long long block = 0;
166 __le32 *blist = kmalloc(PAGE_SIZE, GFP_KERNEL);
167
168 if (blist == NULL) {
169 ERROR("read_indexes: Failed to allocate block_list\n");
170 return -ENOMEM;
171 }
172
173 while (n) {
174 int blocks = min_t(int, n, PAGE_SIZE >> 2);
175
176 err = squashfs_read_metadata(sb, blist, start_block,
177 offset, blocks << 2);
178 if (err < 0) {
179 ERROR("read_indexes: reading block [%llx:%x]\n",
180 *start_block, *offset);
181 goto failure;
182 }
183
184 for (i = 0; i < blocks; i++) {
185 int size = squashfs_block_size(blist[i]);
186 if (size < 0) {
187 err = size;
188 goto failure;
189 }
190 block += SQUASHFS_COMPRESSED_SIZE_BLOCK(size);
191 }
192 n -= blocks;
193 }
194
195 kfree(blist);
196 return block;
197
198 failure:
199 kfree(blist);
200 return err;
201 }
202
203
204 /*
205 * Each cache index slot has SQUASHFS_META_ENTRIES, each of which
206 * can cache one index -> datablock/blocklist-block mapping. We wish
207 * to distribute these over the length of the file, entry[0] maps index x,
208 * entry[1] maps index x + skip, entry[2] maps index x + 2 * skip, and so on.
209 * The larger the file, the greater the skip factor. The skip factor is
210 * limited to the size of the metadata cache (SQUASHFS_CACHED_BLKS) to ensure
211 * the number of metadata blocks that need to be read fits into the cache.
212 * If the skip factor is limited in this way then the file will use multiple
213 * slots.
214 */
215 static inline int calculate_skip(u64 blocks)
216 {
217 u64 skip = blocks / ((SQUASHFS_META_ENTRIES + 1)
218 * SQUASHFS_META_INDEXES);
219 return min((u64) SQUASHFS_CACHED_BLKS - 1, skip + 1);
220 }
221
222
223 /*
224 * Search and grow the index cache for the specified inode, returning the
225 * on-disk locations of the datablock and block list metadata block
226 * <index_block, index_offset> for index (scaled to nearest cache index).
227 */
228 static int fill_meta_index(struct inode *inode, int index,
229 u64 *index_block, int *index_offset, u64 *data_block)
230 {
231 struct squashfs_sb_info *msblk = inode->i_sb->s_fs_info;
232 int skip = calculate_skip(i_size_read(inode) >> msblk->block_log);
233 int offset = 0;
234 struct meta_index *meta;
235 struct meta_entry *meta_entry;
236 u64 cur_index_block = squashfs_i(inode)->block_list_start;
237 int cur_offset = squashfs_i(inode)->offset;
238 u64 cur_data_block = squashfs_i(inode)->start;
239 int err, i;
240
241 /*
242 * Scale index to cache index (cache slot entry)
243 */
244 index /= SQUASHFS_META_INDEXES * skip;
245
246 while (offset < index) {
247 meta = locate_meta_index(inode, offset + 1, index);
248
249 if (meta == NULL) {
250 meta = empty_meta_index(inode, offset + 1, skip);
251 if (meta == NULL)
252 goto all_done;
253 } else {
254 offset = index < meta->offset + meta->entries ? index :
255 meta->offset + meta->entries - 1;
256 meta_entry = &meta->meta_entry[offset - meta->offset];
257 cur_index_block = meta_entry->index_block +
258 msblk->inode_table;
259 cur_offset = meta_entry->offset;
260 cur_data_block = meta_entry->data_block;
261 TRACE("get_meta_index: offset %d, meta->offset %d, "
262 "meta->entries %d\n", offset, meta->offset,
263 meta->entries);
264 TRACE("get_meta_index: index_block 0x%llx, offset 0x%x"
265 " data_block 0x%llx\n", cur_index_block,
266 cur_offset, cur_data_block);
267 }
268
269 /*
270 * If necessary grow cache slot by reading block list. Cache
271 * slot is extended up to index or to the end of the slot, in
272 * which case further slots will be used.
273 */
274 for (i = meta->offset + meta->entries; i <= index &&
275 i < meta->offset + SQUASHFS_META_ENTRIES; i++) {
276 int blocks = skip * SQUASHFS_META_INDEXES;
277 long long res = read_indexes(inode->i_sb, blocks,
278 &cur_index_block, &cur_offset);
279
280 if (res < 0) {
281 if (meta->entries == 0)
282 /*
283 * Don't leave an empty slot on read
284 * error allocated to this inode...
285 */
286 meta->inode_number = 0;
287 err = res;
288 goto failed;
289 }
290
291 cur_data_block += res;
292 meta_entry = &meta->meta_entry[i - meta->offset];
293 meta_entry->index_block = cur_index_block -
294 msblk->inode_table;
295 meta_entry->offset = cur_offset;
296 meta_entry->data_block = cur_data_block;
297 meta->entries++;
298 offset++;
299 }
300
301 TRACE("get_meta_index: meta->offset %d, meta->entries %d\n",
302 meta->offset, meta->entries);
303
304 release_meta_index(inode, meta);
305 }
306
307 all_done:
308 *index_block = cur_index_block;
309 *index_offset = cur_offset;
310 *data_block = cur_data_block;
311
312 /*
313 * Scale cache index (cache slot entry) to index
314 */
315 return offset * SQUASHFS_META_INDEXES * skip;
316
317 failed:
318 release_meta_index(inode, meta);
319 return err;
320 }
321
322
323 /*
324 * Get the on-disk location and compressed size of the datablock
325 * specified by index. Fill_meta_index() does most of the work.
326 */
327 static int read_blocklist(struct inode *inode, int index, u64 *block)
328 {
329 u64 start;
330 long long blks;
331 int offset;
332 __le32 size;
333 int res = fill_meta_index(inode, index, &start, &offset, block);
334
335 TRACE("read_blocklist: res %d, index %d, start 0x%llx, offset"
336 " 0x%x, block 0x%llx\n", res, index, start, offset,
337 *block);
338
339 if (res < 0)
340 return res;
341
342 /*
343 * res contains the index of the mapping returned by fill_meta_index(),
344 * this will likely be less than the desired index (because the
345 * meta_index cache works at a higher granularity). Read any
346 * extra block indexes needed.
347 */
348 if (res < index) {
349 blks = read_indexes(inode->i_sb, index - res, &start, &offset);
350 if (blks < 0)
351 return (int) blks;
352 *block += blks;
353 }
354
355 /*
356 * Read length of block specified by index.
357 */
358 res = squashfs_read_metadata(inode->i_sb, &size, &start, &offset,
359 sizeof(size));
360 if (res < 0)
361 return res;
362 return squashfs_block_size(size);
363 }
364
365 void squashfs_fill_page(struct page *page, struct squashfs_cache_entry *buffer, int offset, int avail)
366 {
367 int copied;
368 void *pageaddr;
369
370 pageaddr = kmap_atomic(page);
371 copied = squashfs_copy_data(pageaddr, buffer, offset, avail);
372 memset(pageaddr + copied, 0, PAGE_SIZE - copied);
373 kunmap_atomic(pageaddr);
374
375 flush_dcache_page(page);
376 if (copied == avail)
377 SetPageUptodate(page);
378 }
379
380 /* Copy data into page cache */
381 void squashfs_copy_cache(struct page *page, struct squashfs_cache_entry *buffer,
382 int bytes, int offset)
383 {
384 struct inode *inode = page->mapping->host;
385 struct squashfs_sb_info *msblk = inode->i_sb->s_fs_info;
386 int i, mask = (1 << (msblk->block_log - PAGE_SHIFT)) - 1;
387 int start_index = page->index & ~mask, end_index = start_index | mask;
388
389 /*
390 * Loop copying datablock into pages. As the datablock likely covers
391 * many PAGE_SIZE pages (default block size is 128 KiB) explicitly
392 * grab the pages from the page cache, except for the page that we've
393 * been called to fill.
394 */
395 for (i = start_index; i <= end_index && bytes > 0; i++,
396 bytes -= PAGE_SIZE, offset += PAGE_SIZE) {
397 struct page *push_page;
398 int avail = buffer ? min_t(int, bytes, PAGE_SIZE) : 0;
399
400 TRACE("bytes %d, i %d, available_bytes %d\n", bytes, i, avail);
401
402 push_page = (i == page->index) ? page :
403 grab_cache_page_nowait(page->mapping, i);
404
405 if (!push_page)
406 continue;
407
408 if (PageUptodate(push_page))
409 goto skip_page;
410
411 squashfs_fill_page(push_page, buffer, offset, avail);
412 skip_page:
413 unlock_page(push_page);
414 if (i != page->index)
415 put_page(push_page);
416 }
417 }
418
419 /* Read datablock stored packed inside a fragment (tail-end packed block) */
420 static int squashfs_readpage_fragment(struct page *page, int expected)
421 {
422 struct inode *inode = page->mapping->host;
423 struct squashfs_cache_entry *buffer = squashfs_get_fragment(inode->i_sb,
424 squashfs_i(inode)->fragment_block,
425 squashfs_i(inode)->fragment_size);
426 int res = buffer->error;
427
428 if (res)
429 ERROR("Unable to read page, block %llx, size %x\n",
430 squashfs_i(inode)->fragment_block,
431 squashfs_i(inode)->fragment_size);
432 else
433 squashfs_copy_cache(page, buffer, expected,
434 squashfs_i(inode)->fragment_offset);
435
436 squashfs_cache_put(buffer);
437 return res;
438 }
439
440 static int squashfs_readpage_sparse(struct page *page, int expected)
441 {
442 squashfs_copy_cache(page, NULL, expected, 0);
443 return 0;
444 }
445
446 static int squashfs_read_folio(struct file *file, struct folio *folio)
447 {
448 struct page *page = &folio->page;
449 struct inode *inode = page->mapping->host;
450 struct squashfs_sb_info *msblk = inode->i_sb->s_fs_info;
451 int index = page->index >> (msblk->block_log - PAGE_SHIFT);
452 int file_end = i_size_read(inode) >> msblk->block_log;
453 int expected = index == file_end ?
454 (i_size_read(inode) & (msblk->block_size - 1)) :
455 msblk->block_size;
456 int res = 0;
457 void *pageaddr;
458
459 TRACE("Entered squashfs_readpage, page index %lx, start block %llx\n",
460 page->index, squashfs_i(inode)->start);
461
462 if (page->index >= ((i_size_read(inode) + PAGE_SIZE - 1) >>
463 PAGE_SHIFT))
464 goto out;
465
466 if (index < file_end || squashfs_i(inode)->fragment_block ==
467 SQUASHFS_INVALID_BLK) {
468 u64 block = 0;
469
470 res = read_blocklist(inode, index, &block);
471 if (res < 0)
472 goto out;
473
474 if (res == 0)
475 res = squashfs_readpage_sparse(page, expected);
476 else
477 res = squashfs_readpage_block(page, block, res, expected);
478 } else
479 res = squashfs_readpage_fragment(page, expected);
480
481 if (!res)
482 return 0;
483
484 out:
485 pageaddr = kmap_atomic(page);
486 memset(pageaddr, 0, PAGE_SIZE);
487 kunmap_atomic(pageaddr);
488 flush_dcache_page(page);
489 if (res == 0)
490 SetPageUptodate(page);
491 unlock_page(page);
492
493 return res;
494 }
495
496 static int squashfs_readahead_fragment(struct page **page,
497 unsigned int pages, unsigned int expected)
498 {
499 struct inode *inode = page[0]->mapping->host;
500 struct squashfs_cache_entry *buffer = squashfs_get_fragment(inode->i_sb,
501 squashfs_i(inode)->fragment_block,
502 squashfs_i(inode)->fragment_size);
503 struct squashfs_sb_info *msblk = inode->i_sb->s_fs_info;
504 unsigned int n, mask = (1 << (msblk->block_log - PAGE_SHIFT)) - 1;
505 int error = buffer->error;
506
507 if (error)
508 goto out;
509
510 expected += squashfs_i(inode)->fragment_offset;
511
512 for (n = 0; n < pages; n++) {
513 unsigned int base = (page[n]->index & mask) << PAGE_SHIFT;
514 unsigned int offset = base + squashfs_i(inode)->fragment_offset;
515
516 if (expected > offset) {
517 unsigned int avail = min_t(unsigned int, expected -
518 offset, PAGE_SIZE);
519
520 squashfs_fill_page(page[n], buffer, offset, avail);
521 }
522
523 unlock_page(page[n]);
524 put_page(page[n]);
525 }
526
527 out:
528 squashfs_cache_put(buffer);
529 return error;
530 }
531
532 static void squashfs_readahead(struct readahead_control *ractl)
533 {
534 struct inode *inode = ractl->mapping->host;
535 struct squashfs_sb_info *msblk = inode->i_sb->s_fs_info;
536 size_t mask = (1UL << msblk->block_log) - 1;
537 unsigned short shift = msblk->block_log - PAGE_SHIFT;
538 loff_t start = readahead_pos(ractl) & ~mask;
539 size_t len = readahead_length(ractl) + readahead_pos(ractl) - start;
540 struct squashfs_page_actor *actor;
541 unsigned int nr_pages = 0;
542 struct page **pages;
543 int i;
544 loff_t file_end = i_size_read(inode) >> msblk->block_log;
545 unsigned int max_pages = 1UL << shift;
546
547 readahead_expand(ractl, start, (len | mask) + 1);
548
549 pages = kmalloc_array(max_pages, sizeof(void *), GFP_KERNEL);
550 if (!pages)
551 return;
552
553 for (;;) {
554 pgoff_t index;
555 int res, bsize;
556 u64 block = 0;
557 unsigned int expected;
558 struct page *last_page;
559
560 expected = start >> msblk->block_log == file_end ?
561 (i_size_read(inode) & (msblk->block_size - 1)) :
562 msblk->block_size;
563
564 max_pages = (expected + PAGE_SIZE - 1) >> PAGE_SHIFT;
565
566 nr_pages = __readahead_batch(ractl, pages, max_pages);
567 if (!nr_pages)
568 break;
569
570 if (readahead_pos(ractl) >= i_size_read(inode))
571 goto skip_pages;
572
573 index = pages[0]->index >> shift;
574
575 if ((pages[nr_pages - 1]->index >> shift) != index)
576 goto skip_pages;
577
578 if (index == file_end && squashfs_i(inode)->fragment_block !=
579 SQUASHFS_INVALID_BLK) {
580 res = squashfs_readahead_fragment(pages, nr_pages,
581 expected);
582 if (res)
583 goto skip_pages;
584 continue;
585 }
586
587 bsize = read_blocklist(inode, index, &block);
588 if (bsize == 0)
589 goto skip_pages;
590
591 actor = squashfs_page_actor_init_special(msblk, pages, nr_pages,
592 expected);
593 if (!actor)
594 goto skip_pages;
595
596 res = squashfs_read_data(inode->i_sb, block, bsize, NULL, actor);
597
598 last_page = squashfs_page_actor_free(actor);
599
600 if (res == expected) {
601 int bytes;
602
603 /* Last page (if present) may have trailing bytes not filled */
604 bytes = res % PAGE_SIZE;
605 if (index == file_end && bytes && last_page)
606 memzero_page(last_page, bytes,
607 PAGE_SIZE - bytes);
608
609 for (i = 0; i < nr_pages; i++) {
610 flush_dcache_page(pages[i]);
611 SetPageUptodate(pages[i]);
612 }
613 }
614
615 for (i = 0; i < nr_pages; i++) {
616 unlock_page(pages[i]);
617 put_page(pages[i]);
618 }
619 }
620
621 kfree(pages);
622 return;
623
624 skip_pages:
625 for (i = 0; i < nr_pages; i++) {
626 unlock_page(pages[i]);
627 put_page(pages[i]);
628 }
629 kfree(pages);
630 }
631
632 const struct address_space_operations squashfs_aops = {
633 .read_folio = squashfs_read_folio,
634 .readahead = squashfs_readahead
635 };
636
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.