Linux Audio

Check our new training course

Loading...
v4.6
 
  1/*
  2 * Copyright (C) 2007 Oracle.  All rights reserved.
  3 *
  4 * This program is free software; you can redistribute it and/or
  5 * modify it under the terms of the GNU General Public
  6 * License v2 as published by the Free Software Foundation.
  7 *
  8 * This program is distributed in the hope that it will be useful,
  9 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 10 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 11 * General Public License for more details.
 12 *
 13 * You should have received a copy of the GNU General Public
 14 * License along with this program; if not, write to the
 15 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
 16 * Boston, MA 021110-1307, USA.
 17 */
 18
 19#include <linux/bio.h>
 20#include <linux/slab.h>
 21#include <linux/pagemap.h>
 22#include <linux/highmem.h>
 
 
 
 23#include "ctree.h"
 24#include "disk-io.h"
 25#include "transaction.h"
 26#include "volumes.h"
 27#include "print-tree.h"
 28#include "compression.h"
 
 
 
 29
 30#define __MAX_CSUM_ITEMS(r, size) ((unsigned long)(((BTRFS_LEAF_DATA_SIZE(r) - \
 31				   sizeof(struct btrfs_item) * 2) / \
 32				  size) - 1))
 33
 34#define MAX_CSUM_ITEMS(r, size) (min_t(u32, __MAX_CSUM_ITEMS(r, size), \
 35				       PAGE_SIZE))
 36
 37#define MAX_ORDERED_SUM_BYTES(r) ((PAGE_SIZE - \
 38				   sizeof(struct btrfs_ordered_sum)) / \
 39				   sizeof(u32) * (r)->sectorsize)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 40
 41int btrfs_insert_file_extent(struct btrfs_trans_handle *trans,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 42			     struct btrfs_root *root,
 43			     u64 objectid, u64 pos,
 44			     u64 disk_offset, u64 disk_num_bytes,
 45			     u64 num_bytes, u64 offset, u64 ram_bytes,
 46			     u8 compression, u8 encryption, u16 other_encoding)
 47{
 48	int ret = 0;
 49	struct btrfs_file_extent_item *item;
 50	struct btrfs_key file_key;
 51	struct btrfs_path *path;
 52	struct extent_buffer *leaf;
 53
 54	path = btrfs_alloc_path();
 55	if (!path)
 56		return -ENOMEM;
 57	file_key.objectid = objectid;
 58	file_key.offset = pos;
 59	file_key.type = BTRFS_EXTENT_DATA_KEY;
 60
 61	path->leave_spinning = 1;
 62	ret = btrfs_insert_empty_item(trans, root, path, &file_key,
 63				      sizeof(*item));
 64	if (ret < 0)
 65		goto out;
 66	BUG_ON(ret); /* Can't happen */
 67	leaf = path->nodes[0];
 68	item = btrfs_item_ptr(leaf, path->slots[0],
 69			      struct btrfs_file_extent_item);
 70	btrfs_set_file_extent_disk_bytenr(leaf, item, disk_offset);
 71	btrfs_set_file_extent_disk_num_bytes(leaf, item, disk_num_bytes);
 72	btrfs_set_file_extent_offset(leaf, item, offset);
 73	btrfs_set_file_extent_num_bytes(leaf, item, num_bytes);
 74	btrfs_set_file_extent_ram_bytes(leaf, item, ram_bytes);
 75	btrfs_set_file_extent_generation(leaf, item, trans->transid);
 76	btrfs_set_file_extent_type(leaf, item, BTRFS_FILE_EXTENT_REG);
 77	btrfs_set_file_extent_compression(leaf, item, compression);
 78	btrfs_set_file_extent_encryption(leaf, item, encryption);
 79	btrfs_set_file_extent_other_encoding(leaf, item, other_encoding);
 80
 81	btrfs_mark_buffer_dirty(leaf);
 82out:
 83	btrfs_free_path(path);
 84	return ret;
 85}
 86
 87static struct btrfs_csum_item *
 88btrfs_lookup_csum(struct btrfs_trans_handle *trans,
 89		  struct btrfs_root *root,
 90		  struct btrfs_path *path,
 91		  u64 bytenr, int cow)
 92{
 
 93	int ret;
 94	struct btrfs_key file_key;
 95	struct btrfs_key found_key;
 96	struct btrfs_csum_item *item;
 97	struct extent_buffer *leaf;
 98	u64 csum_offset = 0;
 99	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
100	int csums_in_item;
101
102	file_key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
103	file_key.offset = bytenr;
104	file_key.type = BTRFS_EXTENT_CSUM_KEY;
105	ret = btrfs_search_slot(trans, root, &file_key, path, 0, cow);
106	if (ret < 0)
107		goto fail;
108	leaf = path->nodes[0];
109	if (ret > 0) {
110		ret = 1;
111		if (path->slots[0] == 0)
112			goto fail;
113		path->slots[0]--;
114		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
115		if (found_key.type != BTRFS_EXTENT_CSUM_KEY)
116			goto fail;
117
118		csum_offset = (bytenr - found_key.offset) >>
119				root->fs_info->sb->s_blocksize_bits;
120		csums_in_item = btrfs_item_size_nr(leaf, path->slots[0]);
121		csums_in_item /= csum_size;
122
123		if (csum_offset == csums_in_item) {
124			ret = -EFBIG;
125			goto fail;
126		} else if (csum_offset > csums_in_item) {
127			goto fail;
128		}
129	}
130	item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_csum_item);
131	item = (struct btrfs_csum_item *)((unsigned char *)item +
132					  csum_offset * csum_size);
133	return item;
134fail:
135	if (ret > 0)
136		ret = -ENOENT;
137	return ERR_PTR(ret);
138}
139
140int btrfs_lookup_file_extent(struct btrfs_trans_handle *trans,
141			     struct btrfs_root *root,
142			     struct btrfs_path *path, u64 objectid,
143			     u64 offset, int mod)
144{
145	int ret;
146	struct btrfs_key file_key;
147	int ins_len = mod < 0 ? -1 : 0;
148	int cow = mod != 0;
149
150	file_key.objectid = objectid;
151	file_key.offset = offset;
152	file_key.type = BTRFS_EXTENT_DATA_KEY;
153	ret = btrfs_search_slot(trans, root, &file_key, path, ins_len, cow);
154	return ret;
155}
156
157static void btrfs_io_bio_endio_readpage(struct btrfs_io_bio *bio, int err)
 
 
 
 
 
 
 
 
 
 
 
158{
159	kfree(bio->csum_allocated);
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
160}
161
162static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
163				   struct inode *inode, struct bio *bio,
164				   u64 logical_offset, u32 *dst, int dio)
 
 
 
165{
166	struct bio_vec *bvec = bio->bi_io_vec;
167	struct btrfs_io_bio *btrfs_bio = btrfs_io_bio(bio);
168	struct btrfs_csum_item *item = NULL;
169	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
170	struct btrfs_path *path;
171	u8 *csum;
172	u64 offset = 0;
173	u64 item_start_offset = 0;
174	u64 item_last_offset = 0;
175	u64 disk_bytenr;
176	u64 page_bytes_left;
177	u32 diff;
178	int nblocks;
179	int bio_index = 0;
180	int count;
181	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
182
 
 
 
 
 
 
 
 
 
 
 
 
 
183	path = btrfs_alloc_path();
184	if (!path)
185		return -ENOMEM;
186
187	nblocks = bio->bi_iter.bi_size >> inode->i_sb->s_blocksize_bits;
188	if (!dst) {
189		if (nblocks * csum_size > BTRFS_BIO_INLINE_CSUM_SIZE) {
190			btrfs_bio->csum_allocated = kmalloc_array(nblocks,
191					csum_size, GFP_NOFS);
192			if (!btrfs_bio->csum_allocated) {
193				btrfs_free_path(path);
194				return -ENOMEM;
195			}
196			btrfs_bio->csum = btrfs_bio->csum_allocated;
197			btrfs_bio->end_io = btrfs_io_bio_endio_readpage;
198		} else {
199			btrfs_bio->csum = btrfs_bio->csum_inline;
200		}
201		csum = btrfs_bio->csum;
202	} else {
203		csum = (u8 *)dst;
204	}
205
206	if (bio->bi_iter.bi_size > PAGE_SIZE * 8)
 
 
 
 
207		path->reada = READA_FORWARD;
208
209	WARN_ON(bio->bi_vcnt <= 0);
210
211	/*
212	 * the free space stuff is only read when it hasn't been
213	 * updated in the current transaction.  So, we can safely
214	 * read from the commit root and sidestep a nasty deadlock
215	 * between reading the free space cache and updating the csum tree.
216	 */
217	if (btrfs_is_free_space_inode(inode)) {
218		path->search_commit_root = 1;
219		path->skip_locking = 1;
220	}
221
222	disk_bytenr = (u64)bio->bi_iter.bi_sector << 9;
223	if (dio)
224		offset = logical_offset;
225
226	page_bytes_left = bvec->bv_len;
227	while (bio_index < bio->bi_vcnt) {
228		if (!dio)
229			offset = page_offset(bvec->bv_page) + bvec->bv_offset;
230		count = btrfs_find_ordered_sum(inode, offset, disk_bytenr,
231					       (u32 *)csum, nblocks);
232		if (count)
233			goto found;
234
235		if (!item || disk_bytenr < item_start_offset ||
236		    disk_bytenr >= item_last_offset) {
237			struct btrfs_key found_key;
238			u32 item_size;
239
240			if (item)
241				btrfs_release_path(path);
242			item = btrfs_lookup_csum(NULL, root->fs_info->csum_root,
243						 path, disk_bytenr, 0);
244			if (IS_ERR(item)) {
245				count = 1;
246				memset(csum, 0, csum_size);
247				if (BTRFS_I(inode)->root->root_key.objectid ==
248				    BTRFS_DATA_RELOC_TREE_OBJECTID) {
249					set_extent_bits(io_tree, offset,
250						offset + root->sectorsize - 1,
251						EXTENT_NODATASUM, GFP_NOFS);
252				} else {
253					btrfs_info(BTRFS_I(inode)->root->fs_info,
254						   "no csum found for inode %llu start %llu",
255					       btrfs_ino(inode), offset);
256				}
257				item = NULL;
258				btrfs_release_path(path);
259				goto found;
260			}
261			btrfs_item_key_to_cpu(path->nodes[0], &found_key,
262					      path->slots[0]);
263
264			item_start_offset = found_key.offset;
265			item_size = btrfs_item_size_nr(path->nodes[0],
266						       path->slots[0]);
267			item_last_offset = item_start_offset +
268				(item_size / csum_size) *
269				root->sectorsize;
270			item = btrfs_item_ptr(path->nodes[0], path->slots[0],
271					      struct btrfs_csum_item);
272		}
 
273		/*
274		 * this byte range must be able to fit inside
275		 * a single leaf so it will also fit inside a u32
 
 
 
 
 
 
276		 */
277		diff = disk_bytenr - item_start_offset;
278		diff = diff / root->sectorsize;
279		diff = diff * csum_size;
280		count = min_t(int, nblocks, (item_last_offset - disk_bytenr) >>
281					    inode->i_sb->s_blocksize_bits);
282		read_extent_buffer(path->nodes[0], csum,
283				   ((unsigned long)item) + diff,
284				   csum_size * count);
285found:
286		csum += count * csum_size;
287		nblocks -= count;
288
289		while (count--) {
290			disk_bytenr += root->sectorsize;
291			offset += root->sectorsize;
292			page_bytes_left -= root->sectorsize;
293			if (!page_bytes_left) {
294				bio_index++;
295				/*
296				 * make sure we're still inside the
297				 * bio before we update page_bytes_left
298				 */
299				if (bio_index >= bio->bi_vcnt) {
300					WARN_ON_ONCE(count);
301					goto done;
302				}
303				bvec++;
304				page_bytes_left = bvec->bv_len;
305			}
306
307		}
 
308	}
309
310done:
311	btrfs_free_path(path);
312	return 0;
313}
314
315int btrfs_lookup_bio_sums(struct btrfs_root *root, struct inode *inode,
316			  struct bio *bio, u32 *dst)
317{
318	return __btrfs_lookup_bio_sums(root, inode, bio, 0, dst, 0);
319}
320
321int btrfs_lookup_bio_sums_dio(struct btrfs_root *root, struct inode *inode,
322			      struct bio *bio, u64 offset)
323{
324	return __btrfs_lookup_bio_sums(root, inode, bio, offset, NULL, 1);
325}
326
327int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
328			     struct list_head *list, int search_commit)
 
329{
 
330	struct btrfs_key key;
331	struct btrfs_path *path;
332	struct extent_buffer *leaf;
333	struct btrfs_ordered_sum *sums;
334	struct btrfs_csum_item *item;
335	LIST_HEAD(tmplist);
336	unsigned long offset;
337	int ret;
338	size_t size;
339	u64 csum_end;
340	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
341
342	ASSERT(IS_ALIGNED(start, root->sectorsize) &&
343	       IS_ALIGNED(end + 1, root->sectorsize));
344
345	path = btrfs_alloc_path();
346	if (!path)
347		return -ENOMEM;
348
 
349	if (search_commit) {
350		path->skip_locking = 1;
351		path->reada = READA_FORWARD;
352		path->search_commit_root = 1;
353	}
354
355	key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
356	key.offset = start;
357	key.type = BTRFS_EXTENT_CSUM_KEY;
358
359	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
360	if (ret < 0)
361		goto fail;
362	if (ret > 0 && path->slots[0] > 0) {
363		leaf = path->nodes[0];
364		btrfs_item_key_to_cpu(leaf, &key, path->slots[0] - 1);
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
365		if (key.objectid == BTRFS_EXTENT_CSUM_OBJECTID &&
366		    key.type == BTRFS_EXTENT_CSUM_KEY) {
367			offset = (start - key.offset) >>
368				 root->fs_info->sb->s_blocksize_bits;
369			if (offset * csum_size <
370			    btrfs_item_size_nr(leaf, path->slots[0] - 1))
371				path->slots[0]--;
372		}
373	}
374
375	while (start <= end) {
 
 
376		leaf = path->nodes[0];
377		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
378			ret = btrfs_next_leaf(root, path);
379			if (ret < 0)
380				goto fail;
381			if (ret > 0)
382				break;
383			leaf = path->nodes[0];
384		}
385
386		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
387		if (key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
388		    key.type != BTRFS_EXTENT_CSUM_KEY ||
389		    key.offset > end)
390			break;
391
392		if (key.offset > start)
393			start = key.offset;
394
395		size = btrfs_item_size_nr(leaf, path->slots[0]);
396		csum_end = key.offset + (size / csum_size) * root->sectorsize;
397		if (csum_end <= start) {
398			path->slots[0]++;
399			continue;
400		}
401
402		csum_end = min(csum_end, end + 1);
403		item = btrfs_item_ptr(path->nodes[0], path->slots[0],
404				      struct btrfs_csum_item);
405		while (start < csum_end) {
 
 
 
406			size = min_t(size_t, csum_end - start,
407				     MAX_ORDERED_SUM_BYTES(root));
408			sums = kzalloc(btrfs_ordered_sum_size(root, size),
409				       GFP_NOFS);
410			if (!sums) {
411				ret = -ENOMEM;
412				goto fail;
413			}
414
415			sums->bytenr = start;
416			sums->len = (int)size;
417
418			offset = (start - key.offset) >>
419				root->fs_info->sb->s_blocksize_bits;
420			offset *= csum_size;
421			size >>= root->fs_info->sb->s_blocksize_bits;
422
423			read_extent_buffer(path->nodes[0],
424					   sums->sums,
425					   ((unsigned long)item) + offset,
426					   csum_size * size);
427
428			start += root->sectorsize * size;
429			list_add_tail(&sums->list, &tmplist);
430		}
431		path->slots[0]++;
432	}
433	ret = 0;
434fail:
435	while (ret < 0 && !list_empty(&tmplist)) {
436		sums = list_entry(tmplist.next, struct btrfs_ordered_sum, list);
437		list_del(&sums->list);
438		kfree(sums);
439	}
440	list_splice_tail(&tmplist, list);
441
442	btrfs_free_path(path);
443	return ret;
444}
445
446int btrfs_csum_one_bio(struct btrfs_root *root, struct inode *inode,
447		       struct bio *bio, u64 file_start, int contig)
 
 
 
 
 
 
 
 
 
 
448{
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
449	struct btrfs_ordered_sum *sums;
450	struct btrfs_ordered_extent *ordered;
451	char *data;
452	struct bio_vec *bvec = bio->bi_io_vec;
453	int bio_index = 0;
454	int index;
455	int nr_sectors;
456	int i;
457	unsigned long total_bytes = 0;
458	unsigned long this_sum_bytes = 0;
459	u64 offset;
460
461	WARN_ON(bio->bi_vcnt <= 0);
462	sums = kzalloc(btrfs_ordered_sum_size(root, bio->bi_iter.bi_size),
463		       GFP_NOFS);
464	if (!sums)
465		return -ENOMEM;
466
467	sums->len = bio->bi_iter.bi_size;
468	INIT_LIST_HEAD(&sums->list);
469
470	if (contig)
471		offset = file_start;
472	else
473		offset = page_offset(bvec->bv_page) + bvec->bv_offset;
474
475	ordered = btrfs_lookup_ordered_extent(inode, offset);
476	BUG_ON(!ordered); /* Logic error */
477	sums->bytenr = (u64)bio->bi_iter.bi_sector << 9;
478	index = 0;
479
480	while (bio_index < bio->bi_vcnt) {
481		if (!contig)
482			offset = page_offset(bvec->bv_page) + bvec->bv_offset;
483
484		data = kmap_atomic(bvec->bv_page);
485
486		nr_sectors = BTRFS_BYTES_TO_BLKS(root->fs_info,
487						bvec->bv_len + root->sectorsize
488						- 1);
489
490		for (i = 0; i < nr_sectors; i++) {
491			if (offset >= ordered->file_offset + ordered->len ||
492				offset < ordered->file_offset) {
493				unsigned long bytes_left;
494
495				kunmap_atomic(data);
496				sums->len = this_sum_bytes;
497				this_sum_bytes = 0;
498				btrfs_add_ordered_sum(inode, ordered, sums);
499				btrfs_put_ordered_extent(ordered);
500
501				bytes_left = bio->bi_iter.bi_size - total_bytes;
502
503				sums = kzalloc(btrfs_ordered_sum_size(root, bytes_left),
504					GFP_NOFS);
505				BUG_ON(!sums); /* -ENOMEM */
506				sums->len = bytes_left;
507				ordered = btrfs_lookup_ordered_extent(inode,
508								offset);
509				ASSERT(ordered); /* Logic error */
510				sums->bytenr = ((u64)bio->bi_iter.bi_sector << 9)
511					+ total_bytes;
512				index = 0;
513
514				data = kmap_atomic(bvec->bv_page);
515			}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
516
517			sums->sums[index] = ~(u32)0;
518			sums->sums[index]
519				= btrfs_csum_data(data + bvec->bv_offset
520						+ (i * root->sectorsize),
521						sums->sums[index],
522						root->sectorsize);
523			btrfs_csum_final(sums->sums[index],
524					(char *)(sums->sums + index));
525			index++;
526			offset += root->sectorsize;
527			this_sum_bytes += root->sectorsize;
528			total_bytes += root->sectorsize;
529		}
530
531		kunmap_atomic(data);
532
533		bio_index++;
534		bvec++;
535	}
536	this_sum_bytes = 0;
537	btrfs_add_ordered_sum(inode, ordered, sums);
538	btrfs_put_ordered_extent(ordered);
539	return 0;
540}
541
542/*
543 * helper function for csum removal, this expects the
544 * key to describe the csum pointed to by the path, and it expects
545 * the csum to overlap the range [bytenr, len]
 
546 *
547 * The csum should not be entirely contained in the range and the
548 * range should not be entirely contained in the csum.
549 *
550 * This calls btrfs_truncate_item with the correct args based on the
551 * overlap, and fixes up the key as required.
552 */
553static noinline void truncate_one_csum(struct btrfs_root *root,
554				       struct btrfs_path *path,
555				       struct btrfs_key *key,
556				       u64 bytenr, u64 len)
557{
 
558	struct extent_buffer *leaf;
559	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
560	u64 csum_end;
561	u64 end_byte = bytenr + len;
562	u32 blocksize_bits = root->fs_info->sb->s_blocksize_bits;
563
564	leaf = path->nodes[0];
565	csum_end = btrfs_item_size_nr(leaf, path->slots[0]) / csum_size;
566	csum_end <<= root->fs_info->sb->s_blocksize_bits;
567	csum_end += key->offset;
568
569	if (key->offset < bytenr && csum_end <= end_byte) {
570		/*
571		 *         [ bytenr - len ]
572		 *         [   ]
573		 *   [csum     ]
574		 *   A simple truncate off the end of the item
575		 */
576		u32 new_size = (bytenr - key->offset) >> blocksize_bits;
577		new_size *= csum_size;
578		btrfs_truncate_item(root, path, new_size, 1);
579	} else if (key->offset >= bytenr && csum_end > end_byte &&
580		   end_byte > key->offset) {
581		/*
582		 *         [ bytenr - len ]
583		 *                 [ ]
584		 *                 [csum     ]
585		 * we need to truncate from the beginning of the csum
586		 */
587		u32 new_size = (csum_end - end_byte) >> blocksize_bits;
588		new_size *= csum_size;
589
590		btrfs_truncate_item(root, path, new_size, 0);
591
592		key->offset = end_byte;
593		btrfs_set_item_key_safe(root->fs_info, path, key);
594	} else {
595		BUG();
596	}
597}
598
599/*
600 * deletes the csum items from the csum tree for a given
601 * range of bytes.
602 */
603int btrfs_del_csums(struct btrfs_trans_handle *trans,
604		    struct btrfs_root *root, u64 bytenr, u64 len)
605{
 
606	struct btrfs_path *path;
607	struct btrfs_key key;
608	u64 end_byte = bytenr + len;
609	u64 csum_end;
610	struct extent_buffer *leaf;
611	int ret;
612	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
613	int blocksize_bits = root->fs_info->sb->s_blocksize_bits;
614
615	root = root->fs_info->csum_root;
 
616
617	path = btrfs_alloc_path();
618	if (!path)
619		return -ENOMEM;
620
621	while (1) {
622		key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
623		key.offset = end_byte - 1;
624		key.type = BTRFS_EXTENT_CSUM_KEY;
625
626		path->leave_spinning = 1;
627		ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
628		if (ret > 0) {
 
629			if (path->slots[0] == 0)
630				break;
631			path->slots[0]--;
632		} else if (ret < 0) {
633			break;
634		}
635
636		leaf = path->nodes[0];
637		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
638
639		if (key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
640		    key.type != BTRFS_EXTENT_CSUM_KEY) {
641			break;
642		}
643
644		if (key.offset >= end_byte)
645			break;
646
647		csum_end = btrfs_item_size_nr(leaf, path->slots[0]) / csum_size;
648		csum_end <<= blocksize_bits;
649		csum_end += key.offset;
650
651		/* this csum ends before we start, we're done */
652		if (csum_end <= bytenr)
653			break;
654
655		/* delete the entire item, it is inside our range */
656		if (key.offset >= bytenr && csum_end <= end_byte) {
657			ret = btrfs_del_item(trans, root, path);
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
658			if (ret)
659				goto out;
660			if (key.offset == bytenr)
661				break;
662		} else if (key.offset < bytenr && csum_end > end_byte) {
663			unsigned long offset;
664			unsigned long shift_len;
665			unsigned long item_offset;
666			/*
667			 *        [ bytenr - len ]
668			 *     [csum                ]
669			 *
670			 * Our bytes are in the middle of the csum,
671			 * we need to split this item and insert a new one.
672			 *
673			 * But we can't drop the path because the
674			 * csum could change, get removed, extended etc.
675			 *
676			 * The trick here is the max size of a csum item leaves
677			 * enough room in the tree block for a single
678			 * item header.  So, we split the item in place,
679			 * adding a new header pointing to the existing
680			 * bytes.  Then we loop around again and we have
681			 * a nicely formed csum item that we can neatly
682			 * truncate.
683			 */
684			offset = (bytenr - key.offset) >> blocksize_bits;
685			offset *= csum_size;
686
687			shift_len = (len >> blocksize_bits) * csum_size;
688
689			item_offset = btrfs_item_ptr_offset(leaf,
690							    path->slots[0]);
691
692			memset_extent_buffer(leaf, 0, item_offset + offset,
693					     shift_len);
694			key.offset = bytenr;
695
696			/*
697			 * btrfs_split_item returns -EAGAIN when the
698			 * item changed size or key
699			 */
700			ret = btrfs_split_item(trans, root, path, &key, offset);
701			if (ret && ret != -EAGAIN) {
702				btrfs_abort_transaction(trans, root, ret);
703				goto out;
704			}
 
705
706			key.offset = end_byte - 1;
707		} else {
708			truncate_one_csum(root, path, &key, bytenr, len);
709			if (key.offset < bytenr)
710				break;
711		}
712		btrfs_release_path(path);
713	}
714	ret = 0;
715out:
716	btrfs_free_path(path);
717	return ret;
718}
719
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
720int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans,
721			   struct btrfs_root *root,
722			   struct btrfs_ordered_sum *sums)
723{
 
724	struct btrfs_key file_key;
725	struct btrfs_key found_key;
726	struct btrfs_path *path;
727	struct btrfs_csum_item *item;
728	struct btrfs_csum_item *item_end;
729	struct extent_buffer *leaf = NULL;
730	u64 next_offset;
731	u64 total_bytes = 0;
732	u64 csum_offset;
733	u64 bytenr;
734	u32 nritems;
735	u32 ins_size;
736	int index = 0;
737	int found_next;
738	int ret;
739	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
740
741	path = btrfs_alloc_path();
742	if (!path)
743		return -ENOMEM;
744again:
745	next_offset = (u64)-1;
746	found_next = 0;
747	bytenr = sums->bytenr + total_bytes;
748	file_key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
749	file_key.offset = bytenr;
750	file_key.type = BTRFS_EXTENT_CSUM_KEY;
751
752	item = btrfs_lookup_csum(trans, root, path, bytenr, 1);
753	if (!IS_ERR(item)) {
754		ret = 0;
755		leaf = path->nodes[0];
756		item_end = btrfs_item_ptr(leaf, path->slots[0],
757					  struct btrfs_csum_item);
758		item_end = (struct btrfs_csum_item *)((char *)item_end +
759			   btrfs_item_size_nr(leaf, path->slots[0]));
760		goto found;
761	}
762	ret = PTR_ERR(item);
763	if (ret != -EFBIG && ret != -ENOENT)
764		goto fail_unlock;
765
766	if (ret == -EFBIG) {
767		u32 item_size;
768		/* we found one, but it isn't big enough yet */
769		leaf = path->nodes[0];
770		item_size = btrfs_item_size_nr(leaf, path->slots[0]);
771		if ((item_size / csum_size) >=
772		    MAX_CSUM_ITEMS(root, csum_size)) {
773			/* already at max size, make a new one */
774			goto insert;
775		}
776	} else {
777		int slot = path->slots[0] + 1;
778		/* we didn't find a csum item, insert one */
779		nritems = btrfs_header_nritems(path->nodes[0]);
780		if (!nritems || (path->slots[0] >= nritems - 1)) {
781			ret = btrfs_next_leaf(root, path);
782			if (ret == 1)
783				found_next = 1;
784			if (ret != 0)
785				goto insert;
786			slot = path->slots[0];
787		}
788		btrfs_item_key_to_cpu(path->nodes[0], &found_key, slot);
789		if (found_key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
790		    found_key.type != BTRFS_EXTENT_CSUM_KEY) {
791			found_next = 1;
792			goto insert;
793		}
794		next_offset = found_key.offset;
795		found_next = 1;
796		goto insert;
797	}
798
799	/*
800	 * at this point, we know the tree has an item, but it isn't big
801	 * enough yet to put our csum in.  Grow it
 
 
 
 
 
 
802	 */
 
 
 
 
 
 
 
803	btrfs_release_path(path);
 
804	ret = btrfs_search_slot(trans, root, &file_key, path,
805				csum_size, 1);
 
806	if (ret < 0)
807		goto fail_unlock;
808
809	if (ret > 0) {
810		if (path->slots[0] == 0)
811			goto insert;
812		path->slots[0]--;
813	}
814
815	leaf = path->nodes[0];
816	btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
817	csum_offset = (bytenr - found_key.offset) >>
818			root->fs_info->sb->s_blocksize_bits;
819
820	if (found_key.type != BTRFS_EXTENT_CSUM_KEY ||
821	    found_key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
822	    csum_offset >= MAX_CSUM_ITEMS(root, csum_size)) {
823		goto insert;
824	}
825
826	if (csum_offset == btrfs_item_size_nr(leaf, path->slots[0]) /
 
827	    csum_size) {
828		int extend_nr;
829		u64 tmp;
830		u32 diff;
831		u32 free_space;
832
833		if (btrfs_leaf_free_space(root, leaf) <
834				 sizeof(struct btrfs_item) + csum_size * 2)
835			goto insert;
836
837		free_space = btrfs_leaf_free_space(root, leaf) -
838					 sizeof(struct btrfs_item) - csum_size;
839		tmp = sums->len - total_bytes;
840		tmp >>= root->fs_info->sb->s_blocksize_bits;
841		WARN_ON(tmp < 1);
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
842
843		extend_nr = max_t(int, 1, (int)tmp);
844		diff = (csum_offset + extend_nr) * csum_size;
845		diff = min(diff, MAX_CSUM_ITEMS(root, csum_size) * csum_size);
 
846
847		diff = diff - btrfs_item_size_nr(leaf, path->slots[0]);
848		diff = min(free_space, diff);
849		diff /= csum_size;
850		diff *= csum_size;
851
852		btrfs_extend_item(root, path, diff);
853		ret = 0;
854		goto csum;
855	}
856
857insert:
858	btrfs_release_path(path);
859	csum_offset = 0;
860	if (found_next) {
861		u64 tmp;
862
863		tmp = sums->len - total_bytes;
864		tmp >>= root->fs_info->sb->s_blocksize_bits;
865		tmp = min(tmp, (next_offset - file_key.offset) >>
866					 root->fs_info->sb->s_blocksize_bits);
867
868		tmp = max((u64)1, tmp);
869		tmp = min(tmp, (u64)MAX_CSUM_ITEMS(root, csum_size));
870		ins_size = csum_size * tmp;
871	} else {
872		ins_size = csum_size;
873	}
874	path->leave_spinning = 1;
875	ret = btrfs_insert_empty_item(trans, root, path, &file_key,
876				      ins_size);
877	path->leave_spinning = 0;
878	if (ret < 0)
879		goto fail_unlock;
880	if (WARN_ON(ret != 0))
881		goto fail_unlock;
882	leaf = path->nodes[0];
883csum:
884	item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_csum_item);
885	item_end = (struct btrfs_csum_item *)((unsigned char *)item +
886				      btrfs_item_size_nr(leaf, path->slots[0]));
887	item = (struct btrfs_csum_item *)((unsigned char *)item +
888					  csum_offset * csum_size);
889found:
890	ins_size = (u32)(sums->len - total_bytes) >>
891		   root->fs_info->sb->s_blocksize_bits;
892	ins_size *= csum_size;
893	ins_size = min_t(u32, (unsigned long)item_end - (unsigned long)item,
894			      ins_size);
895	write_extent_buffer(leaf, sums->sums + index, (unsigned long)item,
896			    ins_size);
897
898	ins_size /= csum_size;
899	total_bytes += ins_size * root->sectorsize;
900	index += ins_size;
 
 
901
902	btrfs_mark_buffer_dirty(path->nodes[0]);
903	if (total_bytes < sums->len) {
904		btrfs_release_path(path);
905		cond_resched();
906		goto again;
907	}
908out:
909	btrfs_free_path(path);
910	return ret;
911
912fail_unlock:
913	goto out;
914}
915
916void btrfs_extent_item_to_extent_map(struct inode *inode,
917				     const struct btrfs_path *path,
918				     struct btrfs_file_extent_item *fi,
919				     const bool new_inline,
920				     struct extent_map *em)
921{
922	struct btrfs_root *root = BTRFS_I(inode)->root;
 
923	struct extent_buffer *leaf = path->nodes[0];
924	const int slot = path->slots[0];
925	struct btrfs_key key;
926	u64 extent_start, extent_end;
927	u64 bytenr;
928	u8 type = btrfs_file_extent_type(leaf, fi);
929	int compress_type = btrfs_file_extent_compression(leaf, fi);
930
931	em->bdev = root->fs_info->fs_devices->latest_bdev;
932	btrfs_item_key_to_cpu(leaf, &key, slot);
933	extent_start = key.offset;
934
935	if (type == BTRFS_FILE_EXTENT_REG ||
936	    type == BTRFS_FILE_EXTENT_PREALLOC) {
937		extent_end = extent_start +
938			btrfs_file_extent_num_bytes(leaf, fi);
939	} else if (type == BTRFS_FILE_EXTENT_INLINE) {
940		size_t size;
941		size = btrfs_file_extent_inline_len(leaf, slot, fi);
942		extent_end = ALIGN(extent_start + size, root->sectorsize);
943	}
944
945	em->ram_bytes = btrfs_file_extent_ram_bytes(leaf, fi);
 
946	if (type == BTRFS_FILE_EXTENT_REG ||
947	    type == BTRFS_FILE_EXTENT_PREALLOC) {
948		em->start = extent_start;
949		em->len = extent_end - extent_start;
950		em->orig_start = extent_start -
951			btrfs_file_extent_offset(leaf, fi);
952		em->orig_block_len = btrfs_file_extent_disk_num_bytes(leaf, fi);
953		bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
954		if (bytenr == 0) {
955			em->block_start = EXTENT_MAP_HOLE;
956			return;
957		}
958		if (compress_type != BTRFS_COMPRESS_NONE) {
959			set_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
960			em->compress_type = compress_type;
961			em->block_start = bytenr;
962			em->block_len = em->orig_block_len;
963		} else {
964			bytenr += btrfs_file_extent_offset(leaf, fi);
965			em->block_start = bytenr;
966			em->block_len = em->len;
967			if (type == BTRFS_FILE_EXTENT_PREALLOC)
968				set_bit(EXTENT_FLAG_PREALLOC, &em->flags);
969		}
970	} else if (type == BTRFS_FILE_EXTENT_INLINE) {
971		em->block_start = EXTENT_MAP_INLINE;
972		em->start = extent_start;
973		em->len = extent_end - extent_start;
974		/*
975		 * Initialize orig_start and block_len with the same values
976		 * as in inode.c:btrfs_get_extent().
977		 */
978		em->orig_start = EXTENT_MAP_HOLE;
979		em->block_len = (u64)-1;
980		if (!new_inline && compress_type != BTRFS_COMPRESS_NONE) {
981			set_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
982			em->compress_type = compress_type;
983		}
984	} else {
985		btrfs_err(root->fs_info,
986			  "unknown file extent item type %d, inode %llu, offset %llu, root %llu",
987			  type, btrfs_ino(inode), extent_start,
988			  root->root_key.objectid);
989	}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
990}
v6.9.4
   1// SPDX-License-Identifier: GPL-2.0
   2/*
   3 * Copyright (C) 2007 Oracle.  All rights reserved.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
   4 */
   5
   6#include <linux/bio.h>
   7#include <linux/slab.h>
   8#include <linux/pagemap.h>
   9#include <linux/highmem.h>
  10#include <linux/sched/mm.h>
  11#include <crypto/hash.h>
  12#include "messages.h"
  13#include "ctree.h"
  14#include "disk-io.h"
  15#include "transaction.h"
  16#include "bio.h"
 
  17#include "compression.h"
  18#include "fs.h"
  19#include "accessors.h"
  20#include "file-item.h"
  21
  22#define __MAX_CSUM_ITEMS(r, size) ((unsigned long)(((BTRFS_LEAF_DATA_SIZE(r) - \
  23				   sizeof(struct btrfs_item) * 2) / \
  24				  size) - 1))
  25
  26#define MAX_CSUM_ITEMS(r, size) (min_t(u32, __MAX_CSUM_ITEMS(r, size), \
  27				       PAGE_SIZE))
  28
  29/*
  30 * Set inode's size according to filesystem options.
  31 *
  32 * @inode:      inode we want to update the disk_i_size for
  33 * @new_i_size: i_size we want to set to, 0 if we use i_size
  34 *
  35 * With NO_HOLES set this simply sets the disk_is_size to whatever i_size_read()
  36 * returns as it is perfectly fine with a file that has holes without hole file
  37 * extent items.
  38 *
  39 * However without NO_HOLES we need to only return the area that is contiguous
  40 * from the 0 offset of the file.  Otherwise we could end up adjust i_size up
  41 * to an extent that has a gap in between.
  42 *
  43 * Finally new_i_size should only be set in the case of truncate where we're not
  44 * ready to use i_size_read() as the limiter yet.
  45 */
  46void btrfs_inode_safe_disk_i_size_write(struct btrfs_inode *inode, u64 new_i_size)
  47{
  48	struct btrfs_fs_info *fs_info = inode->root->fs_info;
  49	u64 start, end, i_size;
  50	int ret;
  51
  52	spin_lock(&inode->lock);
  53	i_size = new_i_size ?: i_size_read(&inode->vfs_inode);
  54	if (btrfs_fs_incompat(fs_info, NO_HOLES)) {
  55		inode->disk_i_size = i_size;
  56		goto out_unlock;
  57	}
  58
  59	ret = find_contiguous_extent_bit(inode->file_extent_tree, 0, &start,
  60					 &end, EXTENT_DIRTY);
  61	if (!ret && start == 0)
  62		i_size = min(i_size, end + 1);
  63	else
  64		i_size = 0;
  65	inode->disk_i_size = i_size;
  66out_unlock:
  67	spin_unlock(&inode->lock);
  68}
  69
  70/*
  71 * Mark range within a file as having a new extent inserted.
  72 *
  73 * @inode: inode being modified
  74 * @start: start file offset of the file extent we've inserted
  75 * @len:   logical length of the file extent item
  76 *
  77 * Call when we are inserting a new file extent where there was none before.
  78 * Does not need to call this in the case where we're replacing an existing file
  79 * extent, however if not sure it's fine to call this multiple times.
  80 *
  81 * The start and len must match the file extent item, so thus must be sectorsize
  82 * aligned.
  83 */
  84int btrfs_inode_set_file_extent_range(struct btrfs_inode *inode, u64 start,
  85				      u64 len)
  86{
  87	if (len == 0)
  88		return 0;
  89
  90	ASSERT(IS_ALIGNED(start + len, inode->root->fs_info->sectorsize));
  91
  92	if (btrfs_fs_incompat(inode->root->fs_info, NO_HOLES))
  93		return 0;
  94	return set_extent_bit(inode->file_extent_tree, start, start + len - 1,
  95			      EXTENT_DIRTY, NULL);
  96}
  97
  98/*
  99 * Mark an inode range as not having a backing extent.
 100 *
 101 * @inode: inode being modified
 102 * @start: start file offset of the file extent we've inserted
 103 * @len:   logical length of the file extent item
 104 *
 105 * Called when we drop a file extent, for example when we truncate.  Doesn't
 106 * need to be called for cases where we're replacing a file extent, like when
 107 * we've COWed a file extent.
 108 *
 109 * The start and len must match the file extent item, so thus must be sectorsize
 110 * aligned.
 111 */
 112int btrfs_inode_clear_file_extent_range(struct btrfs_inode *inode, u64 start,
 113					u64 len)
 114{
 115	if (len == 0)
 116		return 0;
 117
 118	ASSERT(IS_ALIGNED(start + len, inode->root->fs_info->sectorsize) ||
 119	       len == (u64)-1);
 120
 121	if (btrfs_fs_incompat(inode->root->fs_info, NO_HOLES))
 122		return 0;
 123	return clear_extent_bit(inode->file_extent_tree, start,
 124				start + len - 1, EXTENT_DIRTY, NULL);
 125}
 126
 127static size_t bytes_to_csum_size(const struct btrfs_fs_info *fs_info, u32 bytes)
 128{
 129	ASSERT(IS_ALIGNED(bytes, fs_info->sectorsize));
 130
 131	return (bytes >> fs_info->sectorsize_bits) * fs_info->csum_size;
 132}
 133
 134static size_t csum_size_to_bytes(const struct btrfs_fs_info *fs_info, u32 csum_size)
 135{
 136	ASSERT(IS_ALIGNED(csum_size, fs_info->csum_size));
 137
 138	return (csum_size / fs_info->csum_size) << fs_info->sectorsize_bits;
 139}
 140
 141static inline u32 max_ordered_sum_bytes(const struct btrfs_fs_info *fs_info)
 142{
 143	u32 max_csum_size = round_down(PAGE_SIZE - sizeof(struct btrfs_ordered_sum),
 144				       fs_info->csum_size);
 145
 146	return csum_size_to_bytes(fs_info, max_csum_size);
 147}
 148
 149/*
 150 * Calculate the total size needed to allocate for an ordered sum structure
 151 * spanning @bytes in the file.
 152 */
 153static int btrfs_ordered_sum_size(struct btrfs_fs_info *fs_info, unsigned long bytes)
 154{
 155	return sizeof(struct btrfs_ordered_sum) + bytes_to_csum_size(fs_info, bytes);
 156}
 157
 158int btrfs_insert_hole_extent(struct btrfs_trans_handle *trans,
 159			     struct btrfs_root *root,
 160			     u64 objectid, u64 pos, u64 num_bytes)
 
 
 
 161{
 162	int ret = 0;
 163	struct btrfs_file_extent_item *item;
 164	struct btrfs_key file_key;
 165	struct btrfs_path *path;
 166	struct extent_buffer *leaf;
 167
 168	path = btrfs_alloc_path();
 169	if (!path)
 170		return -ENOMEM;
 171	file_key.objectid = objectid;
 172	file_key.offset = pos;
 173	file_key.type = BTRFS_EXTENT_DATA_KEY;
 174
 
 175	ret = btrfs_insert_empty_item(trans, root, path, &file_key,
 176				      sizeof(*item));
 177	if (ret < 0)
 178		goto out;
 
 179	leaf = path->nodes[0];
 180	item = btrfs_item_ptr(leaf, path->slots[0],
 181			      struct btrfs_file_extent_item);
 182	btrfs_set_file_extent_disk_bytenr(leaf, item, 0);
 183	btrfs_set_file_extent_disk_num_bytes(leaf, item, 0);
 184	btrfs_set_file_extent_offset(leaf, item, 0);
 185	btrfs_set_file_extent_num_bytes(leaf, item, num_bytes);
 186	btrfs_set_file_extent_ram_bytes(leaf, item, num_bytes);
 187	btrfs_set_file_extent_generation(leaf, item, trans->transid);
 188	btrfs_set_file_extent_type(leaf, item, BTRFS_FILE_EXTENT_REG);
 189	btrfs_set_file_extent_compression(leaf, item, 0);
 190	btrfs_set_file_extent_encryption(leaf, item, 0);
 191	btrfs_set_file_extent_other_encoding(leaf, item, 0);
 192
 193	btrfs_mark_buffer_dirty(trans, leaf);
 194out:
 195	btrfs_free_path(path);
 196	return ret;
 197}
 198
 199static struct btrfs_csum_item *
 200btrfs_lookup_csum(struct btrfs_trans_handle *trans,
 201		  struct btrfs_root *root,
 202		  struct btrfs_path *path,
 203		  u64 bytenr, int cow)
 204{
 205	struct btrfs_fs_info *fs_info = root->fs_info;
 206	int ret;
 207	struct btrfs_key file_key;
 208	struct btrfs_key found_key;
 209	struct btrfs_csum_item *item;
 210	struct extent_buffer *leaf;
 211	u64 csum_offset = 0;
 212	const u32 csum_size = fs_info->csum_size;
 213	int csums_in_item;
 214
 215	file_key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
 216	file_key.offset = bytenr;
 217	file_key.type = BTRFS_EXTENT_CSUM_KEY;
 218	ret = btrfs_search_slot(trans, root, &file_key, path, 0, cow);
 219	if (ret < 0)
 220		goto fail;
 221	leaf = path->nodes[0];
 222	if (ret > 0) {
 223		ret = 1;
 224		if (path->slots[0] == 0)
 225			goto fail;
 226		path->slots[0]--;
 227		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
 228		if (found_key.type != BTRFS_EXTENT_CSUM_KEY)
 229			goto fail;
 230
 231		csum_offset = (bytenr - found_key.offset) >>
 232				fs_info->sectorsize_bits;
 233		csums_in_item = btrfs_item_size(leaf, path->slots[0]);
 234		csums_in_item /= csum_size;
 235
 236		if (csum_offset == csums_in_item) {
 237			ret = -EFBIG;
 238			goto fail;
 239		} else if (csum_offset > csums_in_item) {
 240			goto fail;
 241		}
 242	}
 243	item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_csum_item);
 244	item = (struct btrfs_csum_item *)((unsigned char *)item +
 245					  csum_offset * csum_size);
 246	return item;
 247fail:
 248	if (ret > 0)
 249		ret = -ENOENT;
 250	return ERR_PTR(ret);
 251}
 252
 253int btrfs_lookup_file_extent(struct btrfs_trans_handle *trans,
 254			     struct btrfs_root *root,
 255			     struct btrfs_path *path, u64 objectid,
 256			     u64 offset, int mod)
 257{
 
 258	struct btrfs_key file_key;
 259	int ins_len = mod < 0 ? -1 : 0;
 260	int cow = mod != 0;
 261
 262	file_key.objectid = objectid;
 263	file_key.offset = offset;
 264	file_key.type = BTRFS_EXTENT_DATA_KEY;
 265
 266	return btrfs_search_slot(trans, root, &file_key, path, ins_len, cow);
 267}
 268
 269/*
 270 * Find checksums for logical bytenr range [disk_bytenr, disk_bytenr + len) and
 271 * store the result to @dst.
 272 *
 273 * Return >0 for the number of sectors we found.
 274 * Return 0 for the range [disk_bytenr, disk_bytenr + sectorsize) has no csum
 275 * for it. Caller may want to try next sector until one range is hit.
 276 * Return <0 for fatal error.
 277 */
 278static int search_csum_tree(struct btrfs_fs_info *fs_info,
 279			    struct btrfs_path *path, u64 disk_bytenr,
 280			    u64 len, u8 *dst)
 281{
 282	struct btrfs_root *csum_root;
 283	struct btrfs_csum_item *item = NULL;
 284	struct btrfs_key key;
 285	const u32 sectorsize = fs_info->sectorsize;
 286	const u32 csum_size = fs_info->csum_size;
 287	u32 itemsize;
 288	int ret;
 289	u64 csum_start;
 290	u64 csum_len;
 291
 292	ASSERT(IS_ALIGNED(disk_bytenr, sectorsize) &&
 293	       IS_ALIGNED(len, sectorsize));
 294
 295	/* Check if the current csum item covers disk_bytenr */
 296	if (path->nodes[0]) {
 297		item = btrfs_item_ptr(path->nodes[0], path->slots[0],
 298				      struct btrfs_csum_item);
 299		btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
 300		itemsize = btrfs_item_size(path->nodes[0], path->slots[0]);
 301
 302		csum_start = key.offset;
 303		csum_len = (itemsize / csum_size) * sectorsize;
 304
 305		if (in_range(disk_bytenr, csum_start, csum_len))
 306			goto found;
 307	}
 308
 309	/* Current item doesn't contain the desired range, search again */
 310	btrfs_release_path(path);
 311	csum_root = btrfs_csum_root(fs_info, disk_bytenr);
 312	item = btrfs_lookup_csum(NULL, csum_root, path, disk_bytenr, 0);
 313	if (IS_ERR(item)) {
 314		ret = PTR_ERR(item);
 315		goto out;
 316	}
 317	btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
 318	itemsize = btrfs_item_size(path->nodes[0], path->slots[0]);
 319
 320	csum_start = key.offset;
 321	csum_len = (itemsize / csum_size) * sectorsize;
 322	ASSERT(in_range(disk_bytenr, csum_start, csum_len));
 323
 324found:
 325	ret = (min(csum_start + csum_len, disk_bytenr + len) -
 326		   disk_bytenr) >> fs_info->sectorsize_bits;
 327	read_extent_buffer(path->nodes[0], dst, (unsigned long)item,
 328			ret * csum_size);
 329out:
 330	if (ret == -ENOENT || ret == -EFBIG)
 331		ret = 0;
 332	return ret;
 333}
 334
 335/*
 336 * Lookup the checksum for the read bio in csum tree.
 337 *
 338 * Return: BLK_STS_RESOURCE if allocating memory fails, BLK_STS_OK otherwise.
 339 */
 340blk_status_t btrfs_lookup_bio_sums(struct btrfs_bio *bbio)
 341{
 342	struct btrfs_inode *inode = bbio->inode;
 343	struct btrfs_fs_info *fs_info = inode->root->fs_info;
 344	struct bio *bio = &bbio->bio;
 
 345	struct btrfs_path *path;
 346	const u32 sectorsize = fs_info->sectorsize;
 347	const u32 csum_size = fs_info->csum_size;
 348	u32 orig_len = bio->bi_iter.bi_size;
 349	u64 orig_disk_bytenr = bio->bi_iter.bi_sector << SECTOR_SHIFT;
 350	const unsigned int nblocks = orig_len >> fs_info->sectorsize_bits;
 351	blk_status_t ret = BLK_STS_OK;
 352	u32 bio_offset = 0;
 353
 354	if ((inode->flags & BTRFS_INODE_NODATASUM) ||
 355	    test_bit(BTRFS_FS_STATE_NO_CSUMS, &fs_info->fs_state))
 356		return BLK_STS_OK;
 357
 358	/*
 359	 * This function is only called for read bio.
 360	 *
 361	 * This means two things:
 362	 * - All our csums should only be in csum tree
 363	 *   No ordered extents csums, as ordered extents are only for write
 364	 *   path.
 365	 * - No need to bother any other info from bvec
 366	 *   Since we're looking up csums, the only important info is the
 367	 *   disk_bytenr and the length, which can be extracted from bi_iter
 368	 *   directly.
 369	 */
 370	ASSERT(bio_op(bio) == REQ_OP_READ);
 371	path = btrfs_alloc_path();
 372	if (!path)
 373		return BLK_STS_RESOURCE;
 374
 375	if (nblocks * csum_size > BTRFS_BIO_INLINE_CSUM_SIZE) {
 376		bbio->csum = kmalloc_array(nblocks, csum_size, GFP_NOFS);
 377		if (!bbio->csum) {
 378			btrfs_free_path(path);
 379			return BLK_STS_RESOURCE;
 
 
 
 
 
 
 
 
 380		}
 
 381	} else {
 382		bbio->csum = bbio->csum_inline;
 383	}
 384
 385	/*
 386	 * If requested number of sectors is larger than one leaf can contain,
 387	 * kick the readahead for csum tree.
 388	 */
 389	if (nblocks > fs_info->csums_per_leaf)
 390		path->reada = READA_FORWARD;
 391
 
 
 392	/*
 393	 * the free space stuff is only read when it hasn't been
 394	 * updated in the current transaction.  So, we can safely
 395	 * read from the commit root and sidestep a nasty deadlock
 396	 * between reading the free space cache and updating the csum tree.
 397	 */
 398	if (btrfs_is_free_space_inode(inode)) {
 399		path->search_commit_root = 1;
 400		path->skip_locking = 1;
 401	}
 402
 403	while (bio_offset < orig_len) {
 404		int count;
 405		u64 cur_disk_bytenr = orig_disk_bytenr + bio_offset;
 406		u8 *csum_dst = bbio->csum +
 407			(bio_offset >> fs_info->sectorsize_bits) * csum_size;
 408
 409		count = search_csum_tree(fs_info, path, cur_disk_bytenr,
 410					 orig_len - bio_offset, csum_dst);
 411		if (count < 0) {
 412			ret = errno_to_blk_status(count);
 413			if (bbio->csum != bbio->csum_inline)
 414				kfree(bbio->csum);
 415			bbio->csum = NULL;
 416			break;
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 417		}
 418
 419		/*
 420		 * We didn't find a csum for this range.  We need to make sure
 421		 * we complain loudly about this, because we are not NODATASUM.
 422		 *
 423		 * However for the DATA_RELOC inode we could potentially be
 424		 * relocating data extents for a NODATASUM inode, so the inode
 425		 * itself won't be marked with NODATASUM, but the extent we're
 426		 * copying is in fact NODATASUM.  If we don't find a csum we
 427		 * assume this is the case.
 428		 */
 429		if (count == 0) {
 430			memset(csum_dst, 0, csum_size);
 431			count = 1;
 432
 433			if (inode->root->root_key.objectid ==
 434			    BTRFS_DATA_RELOC_TREE_OBJECTID) {
 435				u64 file_offset = bbio->file_offset + bio_offset;
 436
 437				set_extent_bit(&inode->io_tree, file_offset,
 438					       file_offset + sectorsize - 1,
 439					       EXTENT_NODATASUM, NULL);
 440			} else {
 441				btrfs_warn_rl(fs_info,
 442			"csum hole found for disk bytenr range [%llu, %llu)",
 443				cur_disk_bytenr, cur_disk_bytenr + sectorsize);
 
 
 
 
 
 
 
 
 
 
 
 
 
 444			}
 
 445		}
 446		bio_offset += count * sectorsize;
 447	}
 448
 
 449	btrfs_free_path(path);
 450	return ret;
 
 
 
 
 
 
 
 
 
 
 
 
 451}
 452
 453int btrfs_lookup_csums_list(struct btrfs_root *root, u64 start, u64 end,
 454			    struct list_head *list, int search_commit,
 455			    bool nowait)
 456{
 457	struct btrfs_fs_info *fs_info = root->fs_info;
 458	struct btrfs_key key;
 459	struct btrfs_path *path;
 460	struct extent_buffer *leaf;
 461	struct btrfs_ordered_sum *sums;
 462	struct btrfs_csum_item *item;
 463	LIST_HEAD(tmplist);
 
 464	int ret;
 
 
 
 465
 466	ASSERT(IS_ALIGNED(start, fs_info->sectorsize) &&
 467	       IS_ALIGNED(end + 1, fs_info->sectorsize));
 468
 469	path = btrfs_alloc_path();
 470	if (!path)
 471		return -ENOMEM;
 472
 473	path->nowait = nowait;
 474	if (search_commit) {
 475		path->skip_locking = 1;
 476		path->reada = READA_FORWARD;
 477		path->search_commit_root = 1;
 478	}
 479
 480	key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
 481	key.offset = start;
 482	key.type = BTRFS_EXTENT_CSUM_KEY;
 483
 484	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 485	if (ret < 0)
 486		goto fail;
 487	if (ret > 0 && path->slots[0] > 0) {
 488		leaf = path->nodes[0];
 489		btrfs_item_key_to_cpu(leaf, &key, path->slots[0] - 1);
 490
 491		/*
 492		 * There are two cases we can hit here for the previous csum
 493		 * item:
 494		 *
 495		 *		|<- search range ->|
 496		 *	|<- csum item ->|
 497		 *
 498		 * Or
 499		 *				|<- search range ->|
 500		 *	|<- csum item ->|
 501		 *
 502		 * Check if the previous csum item covers the leading part of
 503		 * the search range.  If so we have to start from previous csum
 504		 * item.
 505		 */
 506		if (key.objectid == BTRFS_EXTENT_CSUM_OBJECTID &&
 507		    key.type == BTRFS_EXTENT_CSUM_KEY) {
 508			if (bytes_to_csum_size(fs_info, start - key.offset) <
 509			    btrfs_item_size(leaf, path->slots[0] - 1))
 
 
 510				path->slots[0]--;
 511		}
 512	}
 513
 514	while (start <= end) {
 515		u64 csum_end;
 516
 517		leaf = path->nodes[0];
 518		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
 519			ret = btrfs_next_leaf(root, path);
 520			if (ret < 0)
 521				goto fail;
 522			if (ret > 0)
 523				break;
 524			leaf = path->nodes[0];
 525		}
 526
 527		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
 528		if (key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
 529		    key.type != BTRFS_EXTENT_CSUM_KEY ||
 530		    key.offset > end)
 531			break;
 532
 533		if (key.offset > start)
 534			start = key.offset;
 535
 536		csum_end = key.offset + csum_size_to_bytes(fs_info,
 537					btrfs_item_size(leaf, path->slots[0]));
 538		if (csum_end <= start) {
 539			path->slots[0]++;
 540			continue;
 541		}
 542
 543		csum_end = min(csum_end, end + 1);
 544		item = btrfs_item_ptr(path->nodes[0], path->slots[0],
 545				      struct btrfs_csum_item);
 546		while (start < csum_end) {
 547			unsigned long offset;
 548			size_t size;
 549
 550			size = min_t(size_t, csum_end - start,
 551				     max_ordered_sum_bytes(fs_info));
 552			sums = kzalloc(btrfs_ordered_sum_size(fs_info, size),
 553				       GFP_NOFS);
 554			if (!sums) {
 555				ret = -ENOMEM;
 556				goto fail;
 557			}
 558
 559			sums->logical = start;
 560			sums->len = size;
 561
 562			offset = bytes_to_csum_size(fs_info, start - key.offset);
 
 
 
 563
 564			read_extent_buffer(path->nodes[0],
 565					   sums->sums,
 566					   ((unsigned long)item) + offset,
 567					   bytes_to_csum_size(fs_info, size));
 568
 569			start += size;
 570			list_add_tail(&sums->list, &tmplist);
 571		}
 572		path->slots[0]++;
 573	}
 574	ret = 0;
 575fail:
 576	while (ret < 0 && !list_empty(&tmplist)) {
 577		sums = list_entry(tmplist.next, struct btrfs_ordered_sum, list);
 578		list_del(&sums->list);
 579		kfree(sums);
 580	}
 581	list_splice_tail(&tmplist, list);
 582
 583	btrfs_free_path(path);
 584	return ret;
 585}
 586
 587/*
 588 * Do the same work as btrfs_lookup_csums_list(), the difference is in how
 589 * we return the result.
 590 *
 591 * This version will set the corresponding bits in @csum_bitmap to represent
 592 * that there is a csum found.
 593 * Each bit represents a sector. Thus caller should ensure @csum_buf passed
 594 * in is large enough to contain all csums.
 595 */
 596int btrfs_lookup_csums_bitmap(struct btrfs_root *root, struct btrfs_path *path,
 597			      u64 start, u64 end, u8 *csum_buf,
 598			      unsigned long *csum_bitmap)
 599{
 600	struct btrfs_fs_info *fs_info = root->fs_info;
 601	struct btrfs_key key;
 602	struct extent_buffer *leaf;
 603	struct btrfs_csum_item *item;
 604	const u64 orig_start = start;
 605	bool free_path = false;
 606	int ret;
 607
 608	ASSERT(IS_ALIGNED(start, fs_info->sectorsize) &&
 609	       IS_ALIGNED(end + 1, fs_info->sectorsize));
 610
 611	if (!path) {
 612		path = btrfs_alloc_path();
 613		if (!path)
 614			return -ENOMEM;
 615		free_path = true;
 616	}
 617
 618	/* Check if we can reuse the previous path. */
 619	if (path->nodes[0]) {
 620		btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
 621
 622		if (key.objectid == BTRFS_EXTENT_CSUM_OBJECTID &&
 623		    key.type == BTRFS_EXTENT_CSUM_KEY &&
 624		    key.offset <= start)
 625			goto search_forward;
 626		btrfs_release_path(path);
 627	}
 628
 629	key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
 630	key.type = BTRFS_EXTENT_CSUM_KEY;
 631	key.offset = start;
 632
 633	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 634	if (ret < 0)
 635		goto fail;
 636	if (ret > 0 && path->slots[0] > 0) {
 637		leaf = path->nodes[0];
 638		btrfs_item_key_to_cpu(leaf, &key, path->slots[0] - 1);
 639
 640		/*
 641		 * There are two cases we can hit here for the previous csum
 642		 * item:
 643		 *
 644		 *		|<- search range ->|
 645		 *	|<- csum item ->|
 646		 *
 647		 * Or
 648		 *				|<- search range ->|
 649		 *	|<- csum item ->|
 650		 *
 651		 * Check if the previous csum item covers the leading part of
 652		 * the search range.  If so we have to start from previous csum
 653		 * item.
 654		 */
 655		if (key.objectid == BTRFS_EXTENT_CSUM_OBJECTID &&
 656		    key.type == BTRFS_EXTENT_CSUM_KEY) {
 657			if (bytes_to_csum_size(fs_info, start - key.offset) <
 658			    btrfs_item_size(leaf, path->slots[0] - 1))
 659				path->slots[0]--;
 660		}
 661	}
 662
 663search_forward:
 664	while (start <= end) {
 665		u64 csum_end;
 666
 667		leaf = path->nodes[0];
 668		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
 669			ret = btrfs_next_leaf(root, path);
 670			if (ret < 0)
 671				goto fail;
 672			if (ret > 0)
 673				break;
 674			leaf = path->nodes[0];
 675		}
 676
 677		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
 678		if (key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
 679		    key.type != BTRFS_EXTENT_CSUM_KEY ||
 680		    key.offset > end)
 681			break;
 682
 683		if (key.offset > start)
 684			start = key.offset;
 685
 686		csum_end = key.offset + csum_size_to_bytes(fs_info,
 687					btrfs_item_size(leaf, path->slots[0]));
 688		if (csum_end <= start) {
 689			path->slots[0]++;
 690			continue;
 691		}
 692
 693		csum_end = min(csum_end, end + 1);
 694		item = btrfs_item_ptr(path->nodes[0], path->slots[0],
 695				      struct btrfs_csum_item);
 696		while (start < csum_end) {
 697			unsigned long offset;
 698			size_t size;
 699			u8 *csum_dest = csum_buf + bytes_to_csum_size(fs_info,
 700						start - orig_start);
 701
 702			size = min_t(size_t, csum_end - start, end + 1 - start);
 703
 704			offset = bytes_to_csum_size(fs_info, start - key.offset);
 705
 706			read_extent_buffer(path->nodes[0], csum_dest,
 707					   ((unsigned long)item) + offset,
 708					   bytes_to_csum_size(fs_info, size));
 709
 710			bitmap_set(csum_bitmap,
 711				(start - orig_start) >> fs_info->sectorsize_bits,
 712				size >> fs_info->sectorsize_bits);
 713
 714			start += size;
 715		}
 716		path->slots[0]++;
 717	}
 718	ret = 0;
 719fail:
 720	if (free_path)
 721		btrfs_free_path(path);
 722	return ret;
 723}
 724
 725/*
 726 * Calculate checksums of the data contained inside a bio.
 727 */
 728blk_status_t btrfs_csum_one_bio(struct btrfs_bio *bbio)
 729{
 730	struct btrfs_ordered_extent *ordered = bbio->ordered;
 731	struct btrfs_inode *inode = bbio->inode;
 732	struct btrfs_fs_info *fs_info = inode->root->fs_info;
 733	SHASH_DESC_ON_STACK(shash, fs_info->csum_shash);
 734	struct bio *bio = &bbio->bio;
 735	struct btrfs_ordered_sum *sums;
 
 736	char *data;
 737	struct bvec_iter iter;
 738	struct bio_vec bvec;
 739	int index;
 740	unsigned int blockcount;
 741	int i;
 742	unsigned nofs_flag;
 743
 744	nofs_flag = memalloc_nofs_save();
 745	sums = kvzalloc(btrfs_ordered_sum_size(fs_info, bio->bi_iter.bi_size),
 746		       GFP_KERNEL);
 747	memalloc_nofs_restore(nofs_flag);
 748
 749	if (!sums)
 750		return BLK_STS_RESOURCE;
 751
 752	sums->len = bio->bi_iter.bi_size;
 753	INIT_LIST_HEAD(&sums->list);
 754
 755	sums->logical = bio->bi_iter.bi_sector << SECTOR_SHIFT;
 
 
 
 
 
 
 
 756	index = 0;
 757
 758	shash->tfm = fs_info->csum_shash;
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 759
 760	bio_for_each_segment(bvec, bio, iter) {
 761		blockcount = BTRFS_BYTES_TO_BLKS(fs_info,
 762						 bvec.bv_len + fs_info->sectorsize
 763						 - 1);
 764
 765		for (i = 0; i < blockcount; i++) {
 766			data = bvec_kmap_local(&bvec);
 767			crypto_shash_digest(shash,
 768					    data + (i * fs_info->sectorsize),
 769					    fs_info->sectorsize,
 770					    sums->sums + index);
 771			kunmap_local(data);
 772			index += fs_info->csum_size;
 773		}
 774
 775	}
 776
 777	bbio->sums = sums;
 778	btrfs_add_ordered_sum(ordered, sums);
 779	return 0;
 780}
 781
 782/*
 783 * Nodatasum I/O on zoned file systems still requires an btrfs_ordered_sum to
 784 * record the updated logical address on Zone Append completion.
 785 * Allocate just the structure with an empty sums array here for that case.
 786 */
 787blk_status_t btrfs_alloc_dummy_sum(struct btrfs_bio *bbio)
 788{
 789	bbio->sums = kmalloc(sizeof(*bbio->sums), GFP_NOFS);
 790	if (!bbio->sums)
 791		return BLK_STS_RESOURCE;
 792	bbio->sums->len = bbio->bio.bi_iter.bi_size;
 793	bbio->sums->logical = bbio->bio.bi_iter.bi_sector << SECTOR_SHIFT;
 794	btrfs_add_ordered_sum(bbio->ordered, bbio->sums);
 
 
 
 
 
 
 
 
 
 795	return 0;
 796}
 797
 798/*
 799 * Remove one checksum overlapping a range.
 800 *
 801 * This expects the key to describe the csum pointed to by the path, and it
 802 * expects the csum to overlap the range [bytenr, len]
 803 *
 804 * The csum should not be entirely contained in the range and the range should
 805 * not be entirely contained in the csum.
 806 *
 807 * This calls btrfs_truncate_item with the correct args based on the overlap,
 808 * and fixes up the key as required.
 809 */
 810static noinline void truncate_one_csum(struct btrfs_trans_handle *trans,
 811				       struct btrfs_path *path,
 812				       struct btrfs_key *key,
 813				       u64 bytenr, u64 len)
 814{
 815	struct btrfs_fs_info *fs_info = trans->fs_info;
 816	struct extent_buffer *leaf;
 817	const u32 csum_size = fs_info->csum_size;
 818	u64 csum_end;
 819	u64 end_byte = bytenr + len;
 820	u32 blocksize_bits = fs_info->sectorsize_bits;
 821
 822	leaf = path->nodes[0];
 823	csum_end = btrfs_item_size(leaf, path->slots[0]) / csum_size;
 824	csum_end <<= blocksize_bits;
 825	csum_end += key->offset;
 826
 827	if (key->offset < bytenr && csum_end <= end_byte) {
 828		/*
 829		 *         [ bytenr - len ]
 830		 *         [   ]
 831		 *   [csum     ]
 832		 *   A simple truncate off the end of the item
 833		 */
 834		u32 new_size = (bytenr - key->offset) >> blocksize_bits;
 835		new_size *= csum_size;
 836		btrfs_truncate_item(trans, path, new_size, 1);
 837	} else if (key->offset >= bytenr && csum_end > end_byte &&
 838		   end_byte > key->offset) {
 839		/*
 840		 *         [ bytenr - len ]
 841		 *                 [ ]
 842		 *                 [csum     ]
 843		 * we need to truncate from the beginning of the csum
 844		 */
 845		u32 new_size = (csum_end - end_byte) >> blocksize_bits;
 846		new_size *= csum_size;
 847
 848		btrfs_truncate_item(trans, path, new_size, 0);
 849
 850		key->offset = end_byte;
 851		btrfs_set_item_key_safe(trans, path, key);
 852	} else {
 853		BUG();
 854	}
 855}
 856
 857/*
 858 * Delete the csum items from the csum tree for a given range of bytes.
 
 859 */
 860int btrfs_del_csums(struct btrfs_trans_handle *trans,
 861		    struct btrfs_root *root, u64 bytenr, u64 len)
 862{
 863	struct btrfs_fs_info *fs_info = trans->fs_info;
 864	struct btrfs_path *path;
 865	struct btrfs_key key;
 866	u64 end_byte = bytenr + len;
 867	u64 csum_end;
 868	struct extent_buffer *leaf;
 869	int ret = 0;
 870	const u32 csum_size = fs_info->csum_size;
 871	u32 blocksize_bits = fs_info->sectorsize_bits;
 872
 873	ASSERT(root->root_key.objectid == BTRFS_CSUM_TREE_OBJECTID ||
 874	       root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID);
 875
 876	path = btrfs_alloc_path();
 877	if (!path)
 878		return -ENOMEM;
 879
 880	while (1) {
 881		key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
 882		key.offset = end_byte - 1;
 883		key.type = BTRFS_EXTENT_CSUM_KEY;
 884
 
 885		ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
 886		if (ret > 0) {
 887			ret = 0;
 888			if (path->slots[0] == 0)
 889				break;
 890			path->slots[0]--;
 891		} else if (ret < 0) {
 892			break;
 893		}
 894
 895		leaf = path->nodes[0];
 896		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
 897
 898		if (key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
 899		    key.type != BTRFS_EXTENT_CSUM_KEY) {
 900			break;
 901		}
 902
 903		if (key.offset >= end_byte)
 904			break;
 905
 906		csum_end = btrfs_item_size(leaf, path->slots[0]) / csum_size;
 907		csum_end <<= blocksize_bits;
 908		csum_end += key.offset;
 909
 910		/* this csum ends before we start, we're done */
 911		if (csum_end <= bytenr)
 912			break;
 913
 914		/* delete the entire item, it is inside our range */
 915		if (key.offset >= bytenr && csum_end <= end_byte) {
 916			int del_nr = 1;
 917
 918			/*
 919			 * Check how many csum items preceding this one in this
 920			 * leaf correspond to our range and then delete them all
 921			 * at once.
 922			 */
 923			if (key.offset > bytenr && path->slots[0] > 0) {
 924				int slot = path->slots[0] - 1;
 925
 926				while (slot >= 0) {
 927					struct btrfs_key pk;
 928
 929					btrfs_item_key_to_cpu(leaf, &pk, slot);
 930					if (pk.offset < bytenr ||
 931					    pk.type != BTRFS_EXTENT_CSUM_KEY ||
 932					    pk.objectid !=
 933					    BTRFS_EXTENT_CSUM_OBJECTID)
 934						break;
 935					path->slots[0] = slot;
 936					del_nr++;
 937					key.offset = pk.offset;
 938					slot--;
 939				}
 940			}
 941			ret = btrfs_del_items(trans, root, path,
 942					      path->slots[0], del_nr);
 943			if (ret)
 944				break;
 945			if (key.offset == bytenr)
 946				break;
 947		} else if (key.offset < bytenr && csum_end > end_byte) {
 948			unsigned long offset;
 949			unsigned long shift_len;
 950			unsigned long item_offset;
 951			/*
 952			 *        [ bytenr - len ]
 953			 *     [csum                ]
 954			 *
 955			 * Our bytes are in the middle of the csum,
 956			 * we need to split this item and insert a new one.
 957			 *
 958			 * But we can't drop the path because the
 959			 * csum could change, get removed, extended etc.
 960			 *
 961			 * The trick here is the max size of a csum item leaves
 962			 * enough room in the tree block for a single
 963			 * item header.  So, we split the item in place,
 964			 * adding a new header pointing to the existing
 965			 * bytes.  Then we loop around again and we have
 966			 * a nicely formed csum item that we can neatly
 967			 * truncate.
 968			 */
 969			offset = (bytenr - key.offset) >> blocksize_bits;
 970			offset *= csum_size;
 971
 972			shift_len = (len >> blocksize_bits) * csum_size;
 973
 974			item_offset = btrfs_item_ptr_offset(leaf,
 975							    path->slots[0]);
 976
 977			memzero_extent_buffer(leaf, item_offset + offset,
 978					     shift_len);
 979			key.offset = bytenr;
 980
 981			/*
 982			 * btrfs_split_item returns -EAGAIN when the
 983			 * item changed size or key
 984			 */
 985			ret = btrfs_split_item(trans, root, path, &key, offset);
 986			if (ret && ret != -EAGAIN) {
 987				btrfs_abort_transaction(trans, ret);
 988				break;
 989			}
 990			ret = 0;
 991
 992			key.offset = end_byte - 1;
 993		} else {
 994			truncate_one_csum(trans, path, &key, bytenr, len);
 995			if (key.offset < bytenr)
 996				break;
 997		}
 998		btrfs_release_path(path);
 999	}
 
 
1000	btrfs_free_path(path);
1001	return ret;
1002}
1003
1004static int find_next_csum_offset(struct btrfs_root *root,
1005				 struct btrfs_path *path,
1006				 u64 *next_offset)
1007{
1008	const u32 nritems = btrfs_header_nritems(path->nodes[0]);
1009	struct btrfs_key found_key;
1010	int slot = path->slots[0] + 1;
1011	int ret;
1012
1013	if (nritems == 0 || slot >= nritems) {
1014		ret = btrfs_next_leaf(root, path);
1015		if (ret < 0) {
1016			return ret;
1017		} else if (ret > 0) {
1018			*next_offset = (u64)-1;
1019			return 0;
1020		}
1021		slot = path->slots[0];
1022	}
1023
1024	btrfs_item_key_to_cpu(path->nodes[0], &found_key, slot);
1025
1026	if (found_key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
1027	    found_key.type != BTRFS_EXTENT_CSUM_KEY)
1028		*next_offset = (u64)-1;
1029	else
1030		*next_offset = found_key.offset;
1031
1032	return 0;
1033}
1034
1035int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans,
1036			   struct btrfs_root *root,
1037			   struct btrfs_ordered_sum *sums)
1038{
1039	struct btrfs_fs_info *fs_info = root->fs_info;
1040	struct btrfs_key file_key;
1041	struct btrfs_key found_key;
1042	struct btrfs_path *path;
1043	struct btrfs_csum_item *item;
1044	struct btrfs_csum_item *item_end;
1045	struct extent_buffer *leaf = NULL;
1046	u64 next_offset;
1047	u64 total_bytes = 0;
1048	u64 csum_offset;
1049	u64 bytenr;
 
1050	u32 ins_size;
1051	int index = 0;
1052	int found_next;
1053	int ret;
1054	const u32 csum_size = fs_info->csum_size;
1055
1056	path = btrfs_alloc_path();
1057	if (!path)
1058		return -ENOMEM;
1059again:
1060	next_offset = (u64)-1;
1061	found_next = 0;
1062	bytenr = sums->logical + total_bytes;
1063	file_key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
1064	file_key.offset = bytenr;
1065	file_key.type = BTRFS_EXTENT_CSUM_KEY;
1066
1067	item = btrfs_lookup_csum(trans, root, path, bytenr, 1);
1068	if (!IS_ERR(item)) {
1069		ret = 0;
1070		leaf = path->nodes[0];
1071		item_end = btrfs_item_ptr(leaf, path->slots[0],
1072					  struct btrfs_csum_item);
1073		item_end = (struct btrfs_csum_item *)((char *)item_end +
1074			   btrfs_item_size(leaf, path->slots[0]));
1075		goto found;
1076	}
1077	ret = PTR_ERR(item);
1078	if (ret != -EFBIG && ret != -ENOENT)
1079		goto out;
1080
1081	if (ret == -EFBIG) {
1082		u32 item_size;
1083		/* we found one, but it isn't big enough yet */
1084		leaf = path->nodes[0];
1085		item_size = btrfs_item_size(leaf, path->slots[0]);
1086		if ((item_size / csum_size) >=
1087		    MAX_CSUM_ITEMS(fs_info, csum_size)) {
1088			/* already at max size, make a new one */
1089			goto insert;
1090		}
1091	} else {
1092		/* We didn't find a csum item, insert one. */
1093		ret = find_next_csum_offset(root, path, &next_offset);
1094		if (ret < 0)
1095			goto out;
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1096		found_next = 1;
1097		goto insert;
1098	}
1099
1100	/*
1101	 * At this point, we know the tree has a checksum item that ends at an
1102	 * offset matching the start of the checksum range we want to insert.
1103	 * We try to extend that item as much as possible and then add as many
1104	 * checksums to it as they fit.
1105	 *
1106	 * First check if the leaf has enough free space for at least one
1107	 * checksum. If it has go directly to the item extension code, otherwise
1108	 * release the path and do a search for insertion before the extension.
1109	 */
1110	if (btrfs_leaf_free_space(leaf) >= csum_size) {
1111		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
1112		csum_offset = (bytenr - found_key.offset) >>
1113			fs_info->sectorsize_bits;
1114		goto extend_csum;
1115	}
1116
1117	btrfs_release_path(path);
1118	path->search_for_extension = 1;
1119	ret = btrfs_search_slot(trans, root, &file_key, path,
1120				csum_size, 1);
1121	path->search_for_extension = 0;
1122	if (ret < 0)
1123		goto out;
1124
1125	if (ret > 0) {
1126		if (path->slots[0] == 0)
1127			goto insert;
1128		path->slots[0]--;
1129	}
1130
1131	leaf = path->nodes[0];
1132	btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
1133	csum_offset = (bytenr - found_key.offset) >> fs_info->sectorsize_bits;
 
1134
1135	if (found_key.type != BTRFS_EXTENT_CSUM_KEY ||
1136	    found_key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
1137	    csum_offset >= MAX_CSUM_ITEMS(fs_info, csum_size)) {
1138		goto insert;
1139	}
1140
1141extend_csum:
1142	if (csum_offset == btrfs_item_size(leaf, path->slots[0]) /
1143	    csum_size) {
1144		int extend_nr;
1145		u64 tmp;
1146		u32 diff;
 
1147
 
 
 
 
 
 
1148		tmp = sums->len - total_bytes;
1149		tmp >>= fs_info->sectorsize_bits;
1150		WARN_ON(tmp < 1);
1151		extend_nr = max_t(int, 1, tmp);
1152
1153		/*
1154		 * A log tree can already have checksum items with a subset of
1155		 * the checksums we are trying to log. This can happen after
1156		 * doing a sequence of partial writes into prealloc extents and
1157		 * fsyncs in between, with a full fsync logging a larger subrange
1158		 * of an extent for which a previous fast fsync logged a smaller
1159		 * subrange. And this happens in particular due to merging file
1160		 * extent items when we complete an ordered extent for a range
1161		 * covered by a prealloc extent - this is done at
1162		 * btrfs_mark_extent_written().
1163		 *
1164		 * So if we try to extend the previous checksum item, which has
1165		 * a range that ends at the start of the range we want to insert,
1166		 * make sure we don't extend beyond the start offset of the next
1167		 * checksum item. If we are at the last item in the leaf, then
1168		 * forget the optimization of extending and add a new checksum
1169		 * item - it is not worth the complexity of releasing the path,
1170		 * getting the first key for the next leaf, repeat the btree
1171		 * search, etc, because log trees are temporary anyway and it
1172		 * would only save a few bytes of leaf space.
1173		 */
1174		if (root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID) {
1175			if (path->slots[0] + 1 >=
1176			    btrfs_header_nritems(path->nodes[0])) {
1177				ret = find_next_csum_offset(root, path, &next_offset);
1178				if (ret < 0)
1179					goto out;
1180				found_next = 1;
1181				goto insert;
1182			}
1183
1184			ret = find_next_csum_offset(root, path, &next_offset);
1185			if (ret < 0)
1186				goto out;
1187
1188			tmp = (next_offset - bytenr) >> fs_info->sectorsize_bits;
1189			if (tmp <= INT_MAX)
1190				extend_nr = min_t(int, extend_nr, tmp);
1191		}
1192
 
1193		diff = (csum_offset + extend_nr) * csum_size;
1194		diff = min(diff,
1195			   MAX_CSUM_ITEMS(fs_info, csum_size) * csum_size);
1196
1197		diff = diff - btrfs_item_size(leaf, path->slots[0]);
1198		diff = min_t(u32, btrfs_leaf_free_space(leaf), diff);
1199		diff /= csum_size;
1200		diff *= csum_size;
1201
1202		btrfs_extend_item(trans, path, diff);
1203		ret = 0;
1204		goto csum;
1205	}
1206
1207insert:
1208	btrfs_release_path(path);
1209	csum_offset = 0;
1210	if (found_next) {
1211		u64 tmp;
1212
1213		tmp = sums->len - total_bytes;
1214		tmp >>= fs_info->sectorsize_bits;
1215		tmp = min(tmp, (next_offset - file_key.offset) >>
1216					 fs_info->sectorsize_bits);
1217
1218		tmp = max_t(u64, 1, tmp);
1219		tmp = min_t(u64, tmp, MAX_CSUM_ITEMS(fs_info, csum_size));
1220		ins_size = csum_size * tmp;
1221	} else {
1222		ins_size = csum_size;
1223	}
 
1224	ret = btrfs_insert_empty_item(trans, root, path, &file_key,
1225				      ins_size);
 
1226	if (ret < 0)
1227		goto out;
 
 
1228	leaf = path->nodes[0];
1229csum:
1230	item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_csum_item);
1231	item_end = (struct btrfs_csum_item *)((unsigned char *)item +
1232				      btrfs_item_size(leaf, path->slots[0]));
1233	item = (struct btrfs_csum_item *)((unsigned char *)item +
1234					  csum_offset * csum_size);
1235found:
1236	ins_size = (u32)(sums->len - total_bytes) >> fs_info->sectorsize_bits;
 
1237	ins_size *= csum_size;
1238	ins_size = min_t(u32, (unsigned long)item_end - (unsigned long)item,
1239			      ins_size);
1240	write_extent_buffer(leaf, sums->sums + index, (unsigned long)item,
1241			    ins_size);
1242
 
 
1243	index += ins_size;
1244	ins_size /= csum_size;
1245	total_bytes += ins_size * fs_info->sectorsize;
1246
1247	btrfs_mark_buffer_dirty(trans, path->nodes[0]);
1248	if (total_bytes < sums->len) {
1249		btrfs_release_path(path);
1250		cond_resched();
1251		goto again;
1252	}
1253out:
1254	btrfs_free_path(path);
1255	return ret;
 
 
 
1256}
1257
1258void btrfs_extent_item_to_extent_map(struct btrfs_inode *inode,
1259				     const struct btrfs_path *path,
1260				     struct btrfs_file_extent_item *fi,
 
1261				     struct extent_map *em)
1262{
1263	struct btrfs_fs_info *fs_info = inode->root->fs_info;
1264	struct btrfs_root *root = inode->root;
1265	struct extent_buffer *leaf = path->nodes[0];
1266	const int slot = path->slots[0];
1267	struct btrfs_key key;
1268	u64 extent_start, extent_end;
1269	u64 bytenr;
1270	u8 type = btrfs_file_extent_type(leaf, fi);
1271	int compress_type = btrfs_file_extent_compression(leaf, fi);
1272
 
1273	btrfs_item_key_to_cpu(leaf, &key, slot);
1274	extent_start = key.offset;
1275	extent_end = btrfs_file_extent_end(path);
 
 
 
 
 
 
 
 
 
 
1276	em->ram_bytes = btrfs_file_extent_ram_bytes(leaf, fi);
1277	em->generation = btrfs_file_extent_generation(leaf, fi);
1278	if (type == BTRFS_FILE_EXTENT_REG ||
1279	    type == BTRFS_FILE_EXTENT_PREALLOC) {
1280		em->start = extent_start;
1281		em->len = extent_end - extent_start;
1282		em->orig_start = extent_start -
1283			btrfs_file_extent_offset(leaf, fi);
1284		em->orig_block_len = btrfs_file_extent_disk_num_bytes(leaf, fi);
1285		bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
1286		if (bytenr == 0) {
1287			em->block_start = EXTENT_MAP_HOLE;
1288			return;
1289		}
1290		if (compress_type != BTRFS_COMPRESS_NONE) {
1291			extent_map_set_compression(em, compress_type);
 
1292			em->block_start = bytenr;
1293			em->block_len = em->orig_block_len;
1294		} else {
1295			bytenr += btrfs_file_extent_offset(leaf, fi);
1296			em->block_start = bytenr;
1297			em->block_len = em->len;
1298			if (type == BTRFS_FILE_EXTENT_PREALLOC)
1299				em->flags |= EXTENT_FLAG_PREALLOC;
1300		}
1301	} else if (type == BTRFS_FILE_EXTENT_INLINE) {
1302		em->block_start = EXTENT_MAP_INLINE;
1303		em->start = extent_start;
1304		em->len = extent_end - extent_start;
1305		/*
1306		 * Initialize orig_start and block_len with the same values
1307		 * as in inode.c:btrfs_get_extent().
1308		 */
1309		em->orig_start = EXTENT_MAP_HOLE;
1310		em->block_len = (u64)-1;
1311		extent_map_set_compression(em, compress_type);
 
 
 
1312	} else {
1313		btrfs_err(fs_info,
1314			  "unknown file extent item type %d, inode %llu, offset %llu, "
1315			  "root %llu", type, btrfs_ino(inode), extent_start,
1316			  root->root_key.objectid);
1317	}
1318}
1319
1320/*
1321 * Returns the end offset (non inclusive) of the file extent item the given path
1322 * points to. If it points to an inline extent, the returned offset is rounded
1323 * up to the sector size.
1324 */
1325u64 btrfs_file_extent_end(const struct btrfs_path *path)
1326{
1327	const struct extent_buffer *leaf = path->nodes[0];
1328	const int slot = path->slots[0];
1329	struct btrfs_file_extent_item *fi;
1330	struct btrfs_key key;
1331	u64 end;
1332
1333	btrfs_item_key_to_cpu(leaf, &key, slot);
1334	ASSERT(key.type == BTRFS_EXTENT_DATA_KEY);
1335	fi = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);
1336
1337	if (btrfs_file_extent_type(leaf, fi) == BTRFS_FILE_EXTENT_INLINE) {
1338		end = btrfs_file_extent_ram_bytes(leaf, fi);
1339		end = ALIGN(key.offset + end, leaf->fs_info->sectorsize);
1340	} else {
1341		end = key.offset + btrfs_file_extent_num_bytes(leaf, fi);
1342	}
1343
1344	return end;
1345}