summaryrefslogtreecommitdiff
path: root/fs/btrfs/file-item.c
diff options
context:
space:
mode:
Diffstat (limited to 'fs/btrfs/file-item.c')
-rw-r--r--fs/btrfs/file-item.c1250
1 files changed, 833 insertions, 417 deletions
diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c
index 920bf3b4b0ef..14e5257f0f04 100644
--- a/fs/btrfs/file-item.c
+++ b/fs/btrfs/file-item.c
@@ -7,12 +7,18 @@
#include <linux/slab.h>
#include <linux/pagemap.h>
#include <linux/highmem.h>
+#include <linux/sched/mm.h>
+#include <crypto/hash.h>
+#include "messages.h"
#include "ctree.h"
#include "disk-io.h"
#include "transaction.h"
-#include "volumes.h"
-#include "print-tree.h"
+#include "bio.h"
#include "compression.h"
+#include "fs.h"
+#include "accessors.h"
+#include "file-item.h"
+#include "volumes.h"
#define __MAX_CSUM_ITEMS(r, size) ((unsigned long)(((BTRFS_LEAF_DATA_SIZE(r) - \
sizeof(struct btrfs_item) * 2) / \
@@ -21,53 +27,172 @@
#define MAX_CSUM_ITEMS(r, size) (min_t(u32, __MAX_CSUM_ITEMS(r, size), \
PAGE_SIZE))
-#define MAX_ORDERED_SUM_BYTES(fs_info) ((PAGE_SIZE - \
- sizeof(struct btrfs_ordered_sum)) / \
- sizeof(u32) * (fs_info)->sectorsize)
+/*
+ * Set inode's size according to filesystem options.
+ *
+ * @inode: inode we want to update the disk_i_size for
+ * @new_i_size: i_size we want to set to, 0 if we use i_size
+ *
+ * With NO_HOLES set this simply sets the disk_is_size to whatever i_size_read()
+ * returns as it is perfectly fine with a file that has holes without hole file
+ * extent items.
+ *
+ * However without NO_HOLES we need to only return the area that is contiguous
+ * from the 0 offset of the file. Otherwise we could end up adjust i_size up
+ * to an extent that has a gap in between.
+ *
+ * Finally new_i_size should only be set in the case of truncate where we're not
+ * ready to use i_size_read() as the limiter yet.
+ */
+void btrfs_inode_safe_disk_i_size_write(struct btrfs_inode *inode, u64 new_i_size)
+{
+ u64 start, end, i_size;
+ bool found;
+
+ spin_lock(&inode->lock);
+ i_size = new_i_size ?: i_size_read(&inode->vfs_inode);
+ if (!inode->file_extent_tree) {
+ inode->disk_i_size = i_size;
+ goto out_unlock;
+ }
+
+ found = btrfs_find_contiguous_extent_bit(inode->file_extent_tree, 0, &start,
+ &end, EXTENT_DIRTY);
+ if (found && start == 0)
+ i_size = min(i_size, end + 1);
+ else
+ i_size = 0;
+ inode->disk_i_size = i_size;
+out_unlock:
+ spin_unlock(&inode->lock);
+}
+
+/*
+ * Mark range within a file as having a new extent inserted.
+ *
+ * @inode: inode being modified
+ * @start: start file offset of the file extent we've inserted
+ * @len: logical length of the file extent item
+ *
+ * Call when we are inserting a new file extent where there was none before.
+ * Does not need to call this in the case where we're replacing an existing file
+ * extent, however if not sure it's fine to call this multiple times.
+ *
+ * The start and len must match the file extent item, so thus must be sectorsize
+ * aligned.
+ */
+int btrfs_inode_set_file_extent_range(struct btrfs_inode *inode, u64 start,
+ u64 len)
+{
+ if (!inode->file_extent_tree)
+ return 0;
+
+ if (len == 0)
+ return 0;
+
+ ASSERT(IS_ALIGNED(start + len, inode->root->fs_info->sectorsize));
+
+ return btrfs_set_extent_bit(inode->file_extent_tree, start, start + len - 1,
+ EXTENT_DIRTY, NULL);
+}
+
+/*
+ * Mark an inode range as not having a backing extent.
+ *
+ * @inode: inode being modified
+ * @start: start file offset of the file extent we've inserted
+ * @len: logical length of the file extent item
+ *
+ * Called when we drop a file extent, for example when we truncate. Doesn't
+ * need to be called for cases where we're replacing a file extent, like when
+ * we've COWed a file extent.
+ *
+ * The start and len must match the file extent item, so thus must be sectorsize
+ * aligned.
+ */
+int btrfs_inode_clear_file_extent_range(struct btrfs_inode *inode, u64 start,
+ u64 len)
+{
+ if (!inode->file_extent_tree)
+ return 0;
+
+ if (len == 0)
+ return 0;
+
+ ASSERT(IS_ALIGNED(start + len, inode->root->fs_info->sectorsize) ||
+ len == (u64)-1);
+
+ return btrfs_clear_extent_bit(inode->file_extent_tree, start,
+ start + len - 1, EXTENT_DIRTY, NULL);
+}
+
+static size_t bytes_to_csum_size(const struct btrfs_fs_info *fs_info, u32 bytes)
+{
+ ASSERT(IS_ALIGNED(bytes, fs_info->sectorsize));
+
+ return (bytes >> fs_info->sectorsize_bits) * fs_info->csum_size;
+}
+
+static size_t csum_size_to_bytes(const struct btrfs_fs_info *fs_info, u32 csum_size)
+{
+ ASSERT(IS_ALIGNED(csum_size, fs_info->csum_size));
+
+ return (csum_size / fs_info->csum_size) << fs_info->sectorsize_bits;
+}
+
+static inline u32 max_ordered_sum_bytes(const struct btrfs_fs_info *fs_info)
+{
+ u32 max_csum_size = round_down(PAGE_SIZE - sizeof(struct btrfs_ordered_sum),
+ fs_info->csum_size);
+
+ return csum_size_to_bytes(fs_info, max_csum_size);
+}
+
+/*
+ * Calculate the total size needed to allocate for an ordered sum structure
+ * spanning @bytes in the file.
+ */
+static int btrfs_ordered_sum_size(const struct btrfs_fs_info *fs_info, unsigned long bytes)
+{
+ return sizeof(struct btrfs_ordered_sum) + bytes_to_csum_size(fs_info, bytes);
+}
-int btrfs_insert_file_extent(struct btrfs_trans_handle *trans,
+int btrfs_insert_hole_extent(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
- u64 objectid, u64 pos,
- u64 disk_offset, u64 disk_num_bytes,
- u64 num_bytes, u64 offset, u64 ram_bytes,
- u8 compression, u8 encryption, u16 other_encoding)
+ u64 objectid, u64 pos, u64 num_bytes)
{
int ret = 0;
struct btrfs_file_extent_item *item;
struct btrfs_key file_key;
- struct btrfs_path *path;
+ BTRFS_PATH_AUTO_FREE(path);
struct extent_buffer *leaf;
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
+
file_key.objectid = objectid;
- file_key.offset = pos;
file_key.type = BTRFS_EXTENT_DATA_KEY;
+ file_key.offset = pos;
- path->leave_spinning = 1;
ret = btrfs_insert_empty_item(trans, root, path, &file_key,
sizeof(*item));
if (ret < 0)
- goto out;
- BUG_ON(ret); /* Can't happen */
+ return ret;
leaf = path->nodes[0];
item = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_file_extent_item);
- btrfs_set_file_extent_disk_bytenr(leaf, item, disk_offset);
- btrfs_set_file_extent_disk_num_bytes(leaf, item, disk_num_bytes);
- btrfs_set_file_extent_offset(leaf, item, offset);
+ btrfs_set_file_extent_disk_bytenr(leaf, item, 0);
+ btrfs_set_file_extent_disk_num_bytes(leaf, item, 0);
+ btrfs_set_file_extent_offset(leaf, item, 0);
btrfs_set_file_extent_num_bytes(leaf, item, num_bytes);
- btrfs_set_file_extent_ram_bytes(leaf, item, ram_bytes);
+ btrfs_set_file_extent_ram_bytes(leaf, item, num_bytes);
btrfs_set_file_extent_generation(leaf, item, trans->transid);
btrfs_set_file_extent_type(leaf, item, BTRFS_FILE_EXTENT_REG);
- btrfs_set_file_extent_compression(leaf, item, compression);
- btrfs_set_file_extent_encryption(leaf, item, encryption);
- btrfs_set_file_extent_other_encoding(leaf, item, other_encoding);
+ btrfs_set_file_extent_compression(leaf, item, 0);
+ btrfs_set_file_extent_encryption(leaf, item, 0);
+ btrfs_set_file_extent_other_encoding(leaf, item, 0);
- btrfs_mark_buffer_dirty(leaf);
-out:
- btrfs_free_path(path);
return ret;
}
@@ -84,12 +209,12 @@ btrfs_lookup_csum(struct btrfs_trans_handle *trans,
struct btrfs_csum_item *item;
struct extent_buffer *leaf;
u64 csum_offset = 0;
- u16 csum_size = btrfs_super_csum_size(fs_info->super_copy);
+ const u32 csum_size = fs_info->csum_size;
int csums_in_item;
file_key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
- file_key.offset = bytenr;
file_key.type = BTRFS_EXTENT_CSUM_KEY;
+ file_key.offset = bytenr;
ret = btrfs_search_slot(trans, root, &file_key, path, 0, cow);
if (ret < 0)
goto fail;
@@ -104,8 +229,8 @@ btrfs_lookup_csum(struct btrfs_trans_handle *trans,
goto fail;
csum_offset = (bytenr - found_key.offset) >>
- fs_info->sb->s_blocksize_bits;
- csums_in_item = btrfs_item_size_nr(leaf, path->slots[0]);
+ fs_info->sectorsize_bits;
+ csums_in_item = btrfs_item_size(leaf, path->slots[0]);
csums_in_item /= csum_size;
if (csum_offset == csums_in_item) {
@@ -130,61 +255,136 @@ int btrfs_lookup_file_extent(struct btrfs_trans_handle *trans,
struct btrfs_path *path, u64 objectid,
u64 offset, int mod)
{
- int ret;
struct btrfs_key file_key;
int ins_len = mod < 0 ? -1 : 0;
int cow = mod != 0;
file_key.objectid = objectid;
- file_key.offset = offset;
file_key.type = BTRFS_EXTENT_DATA_KEY;
- ret = btrfs_search_slot(trans, root, &file_key, path, ins_len, cow);
- return ret;
+ file_key.offset = offset;
+
+ return btrfs_search_slot(trans, root, &file_key, path, ins_len, cow);
}
-static blk_status_t __btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio,
- u64 logical_offset, u32 *dst, int dio)
+/*
+ * Find checksums for logical bytenr range [disk_bytenr, disk_bytenr + len) and
+ * store the result to @dst.
+ *
+ * Return >0 for the number of sectors we found.
+ * Return 0 for the range [disk_bytenr, disk_bytenr + sectorsize) has no csum
+ * for it. Caller may want to try next sector until one range is hit.
+ * Return <0 for fatal error.
+ */
+static int search_csum_tree(struct btrfs_fs_info *fs_info,
+ struct btrfs_path *path, u64 disk_bytenr,
+ u64 len, u8 *dst)
{
- struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
- struct bio_vec bvec;
- struct bvec_iter iter;
- struct btrfs_io_bio *btrfs_bio = btrfs_io_bio(bio);
+ struct btrfs_root *csum_root;
struct btrfs_csum_item *item = NULL;
- struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
- struct btrfs_path *path;
- u8 *csum;
- u64 offset = 0;
- u64 item_start_offset = 0;
- u64 item_last_offset = 0;
- u64 disk_bytenr;
- u64 page_bytes_left;
- u32 diff;
- int nblocks;
- int count = 0;
- u16 csum_size = btrfs_super_csum_size(fs_info->super_copy);
+ struct btrfs_key key;
+ const u32 sectorsize = fs_info->sectorsize;
+ const u32 csum_size = fs_info->csum_size;
+ u32 itemsize;
+ int ret;
+ u64 csum_start;
+ u64 csum_len;
+
+ ASSERT(IS_ALIGNED(disk_bytenr, sectorsize) &&
+ IS_ALIGNED(len, sectorsize));
+
+ /* Check if the current csum item covers disk_bytenr */
+ if (path->nodes[0]) {
+ item = btrfs_item_ptr(path->nodes[0], path->slots[0],
+ struct btrfs_csum_item);
+ btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
+ itemsize = btrfs_item_size(path->nodes[0], path->slots[0]);
+
+ csum_start = key.offset;
+ csum_len = (itemsize / csum_size) * sectorsize;
+
+ if (in_range(disk_bytenr, csum_start, csum_len))
+ goto found;
+ }
+
+ /* Current item doesn't contain the desired range, search again */
+ btrfs_release_path(path);
+ csum_root = btrfs_csum_root(fs_info, disk_bytenr);
+ item = btrfs_lookup_csum(NULL, csum_root, path, disk_bytenr, 0);
+ if (IS_ERR(item)) {
+ ret = PTR_ERR(item);
+ goto out;
+ }
+ btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
+ itemsize = btrfs_item_size(path->nodes[0], path->slots[0]);
+
+ csum_start = key.offset;
+ csum_len = (itemsize / csum_size) * sectorsize;
+ ASSERT(in_range(disk_bytenr, csum_start, csum_len));
+
+found:
+ ret = (min(csum_start + csum_len, disk_bytenr + len) -
+ disk_bytenr) >> fs_info->sectorsize_bits;
+ read_extent_buffer(path->nodes[0], dst, (unsigned long)item,
+ ret * csum_size);
+out:
+ if (ret == -ENOENT || ret == -EFBIG)
+ ret = 0;
+ return ret;
+}
+
+/*
+ * Lookup the checksum for the read bio in csum tree.
+ *
+ * Return: BLK_STS_RESOURCE if allocating memory fails, BLK_STS_OK otherwise.
+ */
+int btrfs_lookup_bio_sums(struct btrfs_bio *bbio)
+{
+ struct btrfs_inode *inode = bbio->inode;
+ struct btrfs_fs_info *fs_info = inode->root->fs_info;
+ struct bio *bio = &bbio->bio;
+ BTRFS_PATH_AUTO_FREE(path);
+ const u32 sectorsize = fs_info->sectorsize;
+ const u32 csum_size = fs_info->csum_size;
+ u32 orig_len = bio->bi_iter.bi_size;
+ u64 orig_disk_bytenr = bio->bi_iter.bi_sector << SECTOR_SHIFT;
+ const unsigned int nblocks = orig_len >> fs_info->sectorsize_bits;
+ int ret = 0;
+ u32 bio_offset = 0;
+
+ if ((inode->flags & BTRFS_INODE_NODATASUM) ||
+ test_bit(BTRFS_FS_STATE_NO_DATA_CSUMS, &fs_info->fs_state))
+ return 0;
+ /*
+ * This function is only called for read bio.
+ *
+ * This means two things:
+ * - All our csums should only be in csum tree
+ * No ordered extents csums, as ordered extents are only for write
+ * path.
+ * - No need to bother any other info from bvec
+ * Since we're looking up csums, the only important info is the
+ * disk_bytenr and the length, which can be extracted from bi_iter
+ * directly.
+ */
+ ASSERT(bio_op(bio) == REQ_OP_READ);
path = btrfs_alloc_path();
if (!path)
- return BLK_STS_RESOURCE;
-
- nblocks = bio->bi_iter.bi_size >> inode->i_sb->s_blocksize_bits;
- if (!dst) {
- if (nblocks * csum_size > BTRFS_BIO_INLINE_CSUM_SIZE) {
- btrfs_bio->csum = kmalloc_array(nblocks, csum_size,
- GFP_NOFS);
- if (!btrfs_bio->csum) {
- btrfs_free_path(path);
- return BLK_STS_RESOURCE;
- }
- } else {
- btrfs_bio->csum = btrfs_bio->csum_inline;
- }
- csum = btrfs_bio->csum;
+ return -ENOMEM;
+
+ if (nblocks * csum_size > BTRFS_BIO_INLINE_CSUM_SIZE) {
+ bbio->csum = kvcalloc(nblocks, csum_size, GFP_NOFS);
+ if (!bbio->csum)
+ return -ENOMEM;
} else {
- csum = (u8 *)dst;
+ bbio->csum = bbio->csum_inline;
}
- if (bio->bi_iter.bi_size > PAGE_SIZE * 8)
+ /*
+ * If requested number of sectors is larger than one leaf can contain,
+ * kick the readahead for csum tree.
+ */
+ if (nblocks > fs_info->csums_per_leaf)
path->reada = READA_FORWARD;
/*
@@ -193,107 +393,107 @@ static blk_status_t __btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio
* read from the commit root and sidestep a nasty deadlock
* between reading the free space cache and updating the csum tree.
*/
- if (btrfs_is_free_space_inode(BTRFS_I(inode))) {
- path->search_commit_root = 1;
- path->skip_locking = 1;
+ if (btrfs_is_free_space_inode(inode)) {
+ path->search_commit_root = true;
+ path->skip_locking = true;
}
- disk_bytenr = (u64)bio->bi_iter.bi_sector << 9;
- if (dio)
- offset = logical_offset;
-
- bio_for_each_segment(bvec, bio, iter) {
- page_bytes_left = bvec.bv_len;
- if (count)
- goto next;
-
- if (!dio)
- offset = page_offset(bvec.bv_page) + bvec.bv_offset;
- count = btrfs_find_ordered_sum(inode, offset, disk_bytenr,
- (u32 *)csum, nblocks);
- if (count)
- goto found;
+ /*
+ * If we are searching for a csum of an extent from a past
+ * transaction, we can search in the commit root and reduce
+ * lock contention on the csum tree extent buffers.
+ *
+ * This is important because that lock is an rwsem which gets
+ * pretty heavy write load under memory pressure and sustained
+ * csum overwrites, unlike the commit_root_sem. (Memory pressure
+ * makes us writeback the nodes multiple times per transaction,
+ * which makes us cow them each time, taking the write lock.)
+ *
+ * Due to how rwsem is implemented, there is a possible
+ * priority inversion where the readers holding the lock don't
+ * get scheduled (say they're in a cgroup stuck in heavy reclaim)
+ * which then blocks writers, including transaction commit. By
+ * using a semaphore with fewer writers (only a commit switching
+ * the roots), we make this issue less likely.
+ *
+ * Note that we don't rely on btrfs_search_slot to lock the
+ * commit root csum. We call search_slot multiple times, which would
+ * create a potential race where a commit comes in between searches
+ * while we are not holding the commit_root_sem, and we get csums
+ * from across transactions.
+ */
+ if (bbio->csum_search_commit_root) {
+ path->search_commit_root = true;
+ path->skip_locking = true;
+ down_read(&fs_info->commit_root_sem);
+ }
- if (!item || disk_bytenr < item_start_offset ||
- disk_bytenr >= item_last_offset) {
- struct btrfs_key found_key;
- u32 item_size;
-
- if (item)
- btrfs_release_path(path);
- item = btrfs_lookup_csum(NULL, fs_info->csum_root,
- path, disk_bytenr, 0);
- if (IS_ERR(item)) {
- count = 1;
- memset(csum, 0, csum_size);
- if (BTRFS_I(inode)->root->root_key.objectid ==
- BTRFS_DATA_RELOC_TREE_OBJECTID) {
- set_extent_bits(io_tree, offset,
- offset + fs_info->sectorsize - 1,
- EXTENT_NODATASUM);
- } else {
- btrfs_info_rl(fs_info,
- "no csum found for inode %llu start %llu",
- btrfs_ino(BTRFS_I(inode)), offset);
- }
- item = NULL;
- btrfs_release_path(path);
- goto found;
- }
- btrfs_item_key_to_cpu(path->nodes[0], &found_key,
- path->slots[0]);
-
- item_start_offset = found_key.offset;
- item_size = btrfs_item_size_nr(path->nodes[0],
- path->slots[0]);
- item_last_offset = item_start_offset +
- (item_size / csum_size) *
- fs_info->sectorsize;
- item = btrfs_item_ptr(path->nodes[0], path->slots[0],
- struct btrfs_csum_item);
+ while (bio_offset < orig_len) {
+ int count;
+ u64 cur_disk_bytenr = orig_disk_bytenr + bio_offset;
+ u8 *csum_dst = bbio->csum +
+ (bio_offset >> fs_info->sectorsize_bits) * csum_size;
+
+ count = search_csum_tree(fs_info, path, cur_disk_bytenr,
+ orig_len - bio_offset, csum_dst);
+ if (count < 0) {
+ ret = count;
+ if (bbio->csum != bbio->csum_inline)
+ kvfree(bbio->csum);
+ bbio->csum = NULL;
+ break;
}
+
/*
- * this byte range must be able to fit inside
- * a single leaf so it will also fit inside a u32
+ * We didn't find a csum for this range. We need to make sure
+ * we complain loudly about this, because we are not NODATASUM.
+ *
+ * However for the DATA_RELOC inode we could potentially be
+ * relocating data extents for a NODATASUM inode, so the inode
+ * itself won't be marked with NODATASUM, but the extent we're
+ * copying is in fact NODATASUM. If we don't find a csum we
+ * assume this is the case.
*/
- diff = disk_bytenr - item_start_offset;
- diff = diff / fs_info->sectorsize;
- diff = diff * csum_size;
- count = min_t(int, nblocks, (item_last_offset - disk_bytenr) >>
- inode->i_sb->s_blocksize_bits);
- read_extent_buffer(path->nodes[0], csum,
- ((unsigned long)item) + diff,
- csum_size * count);
-found:
- csum += count * csum_size;
- nblocks -= count;
-next:
- while (count--) {
- disk_bytenr += fs_info->sectorsize;
- offset += fs_info->sectorsize;
- page_bytes_left -= fs_info->sectorsize;
- if (!page_bytes_left)
- break; /* move to next bio */
+ if (count == 0) {
+ memset(csum_dst, 0, csum_size);
+ count = 1;
+
+ if (btrfs_is_data_reloc_root(inode->root)) {
+ u64 file_offset = bbio->file_offset + bio_offset;
+
+ btrfs_set_extent_bit(&inode->io_tree, file_offset,
+ file_offset + sectorsize - 1,
+ EXTENT_NODATASUM, NULL);
+ } else {
+ btrfs_warn_rl(fs_info,
+ "csum hole found for disk bytenr range [%llu, %llu)",
+ cur_disk_bytenr, cur_disk_bytenr + sectorsize);
+ }
}
+ bio_offset += count * sectorsize;
}
- WARN_ON_ONCE(count);
- btrfs_free_path(path);
- return 0;
-}
-
-blk_status_t btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio, u32 *dst)
-{
- return __btrfs_lookup_bio_sums(inode, bio, 0, dst, 0);
-}
-
-blk_status_t btrfs_lookup_bio_sums_dio(struct inode *inode, struct bio *bio, u64 offset)
-{
- return __btrfs_lookup_bio_sums(inode, bio, offset, NULL, 1);
+ if (bbio->csum_search_commit_root)
+ up_read(&fs_info->commit_root_sem);
+ return ret;
}
-int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
- struct list_head *list, int search_commit)
+/*
+ * Search for checksums for a given logical range.
+ *
+ * @root: The root where to look for checksums.
+ * @start: Logical address of target checksum range.
+ * @end: End offset (inclusive) of the target checksum range.
+ * @list: List for adding each checksum that was found.
+ * Can be NULL in case the caller only wants to check if
+ * there any checksums for the range.
+ * @nowait: Indicate if the search must be non-blocking or not.
+ *
+ * Return < 0 on error, 0 if no checksums were found, or 1 if checksums were
+ * found.
+ */
+int btrfs_lookup_csums_list(struct btrfs_root *root, u64 start, u64 end,
+ struct list_head *list, bool nowait)
{
struct btrfs_fs_info *fs_info = root->fs_info;
struct btrfs_key key;
@@ -301,12 +501,8 @@ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
struct extent_buffer *leaf;
struct btrfs_ordered_sum *sums;
struct btrfs_csum_item *item;
- LIST_HEAD(tmplist);
- unsigned long offset;
int ret;
- size_t size;
- u64 csum_end;
- u16 csum_size = btrfs_super_csum_size(fs_info->super_copy);
+ bool found_csums = false;
ASSERT(IS_ALIGNED(start, fs_info->sectorsize) &&
IS_ALIGNED(end + 1, fs_info->sectorsize));
@@ -315,38 +511,50 @@ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
if (!path)
return -ENOMEM;
- if (search_commit) {
- path->skip_locking = 1;
- path->reada = READA_FORWARD;
- path->search_commit_root = 1;
- }
+ path->nowait = nowait;
key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
- key.offset = start;
key.type = BTRFS_EXTENT_CSUM_KEY;
+ key.offset = start;
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
if (ret < 0)
- goto fail;
+ goto out;
if (ret > 0 && path->slots[0] > 0) {
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &key, path->slots[0] - 1);
+
+ /*
+ * There are two cases we can hit here for the previous csum
+ * item:
+ *
+ * |<- search range ->|
+ * |<- csum item ->|
+ *
+ * Or
+ * |<- search range ->|
+ * |<- csum item ->|
+ *
+ * Check if the previous csum item covers the leading part of
+ * the search range. If so we have to start from previous csum
+ * item.
+ */
if (key.objectid == BTRFS_EXTENT_CSUM_OBJECTID &&
key.type == BTRFS_EXTENT_CSUM_KEY) {
- offset = (start - key.offset) >>
- fs_info->sb->s_blocksize_bits;
- if (offset * csum_size <
- btrfs_item_size_nr(leaf, path->slots[0] - 1))
+ if (bytes_to_csum_size(fs_info, start - key.offset) <
+ btrfs_item_size(leaf, path->slots[0] - 1))
path->slots[0]--;
}
}
while (start <= end) {
+ u64 csum_end;
+
leaf = path->nodes[0];
if (path->slots[0] >= btrfs_header_nritems(leaf)) {
ret = btrfs_next_leaf(root, path);
if (ret < 0)
- goto fail;
+ goto out;
if (ret > 0)
break;
leaf = path->nodes[0];
@@ -361,178 +569,322 @@ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
if (key.offset > start)
start = key.offset;
- size = btrfs_item_size_nr(leaf, path->slots[0]);
- csum_end = key.offset + (size / csum_size) * fs_info->sectorsize;
+ csum_end = key.offset + csum_size_to_bytes(fs_info,
+ btrfs_item_size(leaf, path->slots[0]));
if (csum_end <= start) {
path->slots[0]++;
continue;
}
+ found_csums = true;
+ if (!list)
+ goto out;
+
csum_end = min(csum_end, end + 1);
item = btrfs_item_ptr(path->nodes[0], path->slots[0],
struct btrfs_csum_item);
while (start < csum_end) {
+ unsigned long offset;
+ size_t size;
+
size = min_t(size_t, csum_end - start,
- MAX_ORDERED_SUM_BYTES(fs_info));
+ max_ordered_sum_bytes(fs_info));
sums = kzalloc(btrfs_ordered_sum_size(fs_info, size),
GFP_NOFS);
if (!sums) {
ret = -ENOMEM;
- goto fail;
+ goto out;
}
- sums->bytenr = start;
- sums->len = (int)size;
+ sums->logical = start;
+ sums->len = size;
- offset = (start - key.offset) >>
- fs_info->sb->s_blocksize_bits;
- offset *= csum_size;
- size >>= fs_info->sb->s_blocksize_bits;
+ offset = bytes_to_csum_size(fs_info, start - key.offset);
read_extent_buffer(path->nodes[0],
sums->sums,
((unsigned long)item) + offset,
- csum_size * size);
+ bytes_to_csum_size(fs_info, size));
- start += fs_info->sectorsize * size;
- list_add_tail(&sums->list, &tmplist);
+ start += size;
+ list_add_tail(&sums->list, list);
}
path->slots[0]++;
}
- ret = 0;
-fail:
- while (ret < 0 && !list_empty(&tmplist)) {
- sums = list_entry(tmplist.next, struct btrfs_ordered_sum, list);
- list_del(&sums->list);
- kfree(sums);
+out:
+ btrfs_free_path(path);
+ if (ret < 0) {
+ if (list) {
+ struct btrfs_ordered_sum *tmp_sums;
+
+ list_for_each_entry_safe(sums, tmp_sums, list, list)
+ kfree(sums);
+ }
+
+ return ret;
}
- list_splice_tail(&tmplist, list);
- btrfs_free_path(path);
- return ret;
+ return found_csums ? 1 : 0;
}
-blk_status_t btrfs_csum_one_bio(struct inode *inode, struct bio *bio,
- u64 file_start, int contig)
+/*
+ * Do the same work as btrfs_lookup_csums_list(), the difference is in how
+ * we return the result.
+ *
+ * This version will set the corresponding bits in @csum_bitmap to represent
+ * that there is a csum found.
+ * Each bit represents a sector. Thus caller should ensure @csum_buf passed
+ * in is large enough to contain all csums.
+ */
+int btrfs_lookup_csums_bitmap(struct btrfs_root *root, struct btrfs_path *path,
+ u64 start, u64 end, u8 *csum_buf,
+ unsigned long *csum_bitmap)
{
- struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
- struct btrfs_ordered_sum *sums;
- struct btrfs_ordered_extent *ordered = NULL;
- char *data;
- struct bvec_iter iter;
- struct bio_vec bvec;
- int index;
- int nr_sectors;
- unsigned long total_bytes = 0;
- unsigned long this_sum_bytes = 0;
- int i;
- u64 offset;
-
- sums = kzalloc(btrfs_ordered_sum_size(fs_info, bio->bi_iter.bi_size),
- GFP_NOFS);
- if (!sums)
- return BLK_STS_RESOURCE;
+ struct btrfs_fs_info *fs_info = root->fs_info;
+ struct btrfs_key key;
+ struct extent_buffer *leaf;
+ struct btrfs_csum_item *item;
+ const u64 orig_start = start;
+ bool free_path = false;
+ int ret;
- sums->len = bio->bi_iter.bi_size;
- INIT_LIST_HEAD(&sums->list);
+ ASSERT(IS_ALIGNED(start, fs_info->sectorsize) &&
+ IS_ALIGNED(end + 1, fs_info->sectorsize));
- if (contig)
- offset = file_start;
- else
- offset = 0; /* shut up gcc */
+ if (!path) {
+ path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
+ free_path = true;
+ }
+
+ /* Check if we can reuse the previous path. */
+ if (path->nodes[0]) {
+ btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
+
+ if (key.objectid == BTRFS_EXTENT_CSUM_OBJECTID &&
+ key.type == BTRFS_EXTENT_CSUM_KEY &&
+ key.offset <= start)
+ goto search_forward;
+ btrfs_release_path(path);
+ }
- sums->bytenr = (u64)bio->bi_iter.bi_sector << 9;
- index = 0;
+ key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
+ key.type = BTRFS_EXTENT_CSUM_KEY;
+ key.offset = start;
- bio_for_each_segment(bvec, bio, iter) {
- if (!contig)
- offset = page_offset(bvec.bv_page) + bvec.bv_offset;
+ ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+ if (ret < 0)
+ goto fail;
+ if (ret > 0 && path->slots[0] > 0) {
+ leaf = path->nodes[0];
+ btrfs_item_key_to_cpu(leaf, &key, path->slots[0] - 1);
- if (!ordered) {
- ordered = btrfs_lookup_ordered_extent(inode, offset);
- BUG_ON(!ordered); /* Logic error */
+ /*
+ * There are two cases we can hit here for the previous csum
+ * item:
+ *
+ * |<- search range ->|
+ * |<- csum item ->|
+ *
+ * Or
+ * |<- search range ->|
+ * |<- csum item ->|
+ *
+ * Check if the previous csum item covers the leading part of
+ * the search range. If so we have to start from previous csum
+ * item.
+ */
+ if (key.objectid == BTRFS_EXTENT_CSUM_OBJECTID &&
+ key.type == BTRFS_EXTENT_CSUM_KEY) {
+ if (bytes_to_csum_size(fs_info, start - key.offset) <
+ btrfs_item_size(leaf, path->slots[0] - 1))
+ path->slots[0]--;
}
+ }
- data = kmap_atomic(bvec.bv_page);
-
- nr_sectors = BTRFS_BYTES_TO_BLKS(fs_info,
- bvec.bv_len + fs_info->sectorsize
- - 1);
-
- for (i = 0; i < nr_sectors; i++) {
- if (offset >= ordered->file_offset + ordered->len ||
- offset < ordered->file_offset) {
- unsigned long bytes_left;
-
- kunmap_atomic(data);
- sums->len = this_sum_bytes;
- this_sum_bytes = 0;
- btrfs_add_ordered_sum(inode, ordered, sums);
- btrfs_put_ordered_extent(ordered);
-
- bytes_left = bio->bi_iter.bi_size - total_bytes;
-
- sums = kzalloc(btrfs_ordered_sum_size(fs_info, bytes_left),
- GFP_NOFS);
- BUG_ON(!sums); /* -ENOMEM */
- sums->len = bytes_left;
- ordered = btrfs_lookup_ordered_extent(inode,
- offset);
- ASSERT(ordered); /* Logic error */
- sums->bytenr = ((u64)bio->bi_iter.bi_sector << 9)
- + total_bytes;
- index = 0;
-
- data = kmap_atomic(bvec.bv_page);
- }
+search_forward:
+ while (start <= end) {
+ u64 csum_end;
+
+ leaf = path->nodes[0];
+ if (path->slots[0] >= btrfs_header_nritems(leaf)) {
+ ret = btrfs_next_leaf(root, path);
+ if (ret < 0)
+ goto fail;
+ if (ret > 0)
+ break;
+ leaf = path->nodes[0];
+ }
+
+ btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
+ if (key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
+ key.type != BTRFS_EXTENT_CSUM_KEY ||
+ key.offset > end)
+ break;
+
+ if (key.offset > start)
+ start = key.offset;
+
+ csum_end = key.offset + csum_size_to_bytes(fs_info,
+ btrfs_item_size(leaf, path->slots[0]));
+ if (csum_end <= start) {
+ path->slots[0]++;
+ continue;
+ }
+
+ csum_end = min(csum_end, end + 1);
+ item = btrfs_item_ptr(path->nodes[0], path->slots[0],
+ struct btrfs_csum_item);
+ while (start < csum_end) {
+ unsigned long offset;
+ size_t size;
+ u8 *csum_dest = csum_buf + bytes_to_csum_size(fs_info,
+ start - orig_start);
+
+ size = min_t(size_t, csum_end - start, end + 1 - start);
+
+ offset = bytes_to_csum_size(fs_info, start - key.offset);
- sums->sums[index] = ~(u32)0;
- sums->sums[index]
- = btrfs_csum_data(data + bvec.bv_offset
- + (i * fs_info->sectorsize),
- sums->sums[index],
- fs_info->sectorsize);
- btrfs_csum_final(sums->sums[index],
- (char *)(sums->sums + index));
- index++;
- offset += fs_info->sectorsize;
- this_sum_bytes += fs_info->sectorsize;
- total_bytes += fs_info->sectorsize;
+ read_extent_buffer(path->nodes[0], csum_dest,
+ ((unsigned long)item) + offset,
+ bytes_to_csum_size(fs_info, size));
+
+ bitmap_set(csum_bitmap,
+ (start - orig_start) >> fs_info->sectorsize_bits,
+ size >> fs_info->sectorsize_bits);
+
+ start += size;
}
+ path->slots[0]++;
+ }
+ ret = 0;
+fail:
+ if (free_path)
+ btrfs_free_path(path);
+ return ret;
+}
- kunmap_atomic(data);
+static void csum_one_bio(struct btrfs_bio *bbio, struct bvec_iter *src)
+{
+ struct btrfs_inode *inode = bbio->inode;
+ struct btrfs_fs_info *fs_info = inode->root->fs_info;
+ SHASH_DESC_ON_STACK(shash, fs_info->csum_shash);
+ struct bio *bio = &bbio->bio;
+ struct btrfs_ordered_sum *sums = bbio->sums;
+ struct bvec_iter iter = *src;
+ phys_addr_t paddr;
+ const u32 blocksize = fs_info->sectorsize;
+ const u32 step = min(blocksize, PAGE_SIZE);
+ const u32 nr_steps = blocksize / step;
+ phys_addr_t paddrs[BTRFS_MAX_BLOCKSIZE / PAGE_SIZE];
+ u32 offset = 0;
+ int index = 0;
+
+ shash->tfm = fs_info->csum_shash;
+
+ btrfs_bio_for_each_block(paddr, bio, &iter, step) {
+ paddrs[(offset / step) % nr_steps] = paddr;
+ offset += step;
+
+ if (IS_ALIGNED(offset, blocksize)) {
+ btrfs_calculate_block_csum_pages(fs_info, paddrs, sums->sums + index);
+ index += fs_info->csum_size;
+ }
}
- this_sum_bytes = 0;
- btrfs_add_ordered_sum(inode, ordered, sums);
- btrfs_put_ordered_extent(ordered);
+}
+
+static void csum_one_bio_work(struct work_struct *work)
+{
+ struct btrfs_bio *bbio = container_of(work, struct btrfs_bio, csum_work);
+
+ ASSERT(btrfs_op(&bbio->bio) == BTRFS_MAP_WRITE);
+ ASSERT(bbio->async_csum == true);
+ csum_one_bio(bbio, &bbio->csum_saved_iter);
+ complete(&bbio->csum_done);
+}
+
+/*
+ * Calculate checksums of the data contained inside a bio.
+ */
+int btrfs_csum_one_bio(struct btrfs_bio *bbio, bool async)
+{
+ struct btrfs_ordered_extent *ordered = bbio->ordered;
+ struct btrfs_inode *inode = bbio->inode;
+ struct btrfs_fs_info *fs_info = inode->root->fs_info;
+ struct bio *bio = &bbio->bio;
+ struct btrfs_ordered_sum *sums;
+ unsigned nofs_flag;
+
+ nofs_flag = memalloc_nofs_save();
+ sums = kvzalloc(btrfs_ordered_sum_size(fs_info, bio->bi_iter.bi_size),
+ GFP_KERNEL);
+ memalloc_nofs_restore(nofs_flag);
+
+ if (!sums)
+ return -ENOMEM;
+
+ sums->logical = bbio->orig_logical;
+ sums->len = bio->bi_iter.bi_size;
+ INIT_LIST_HEAD(&sums->list);
+ bbio->sums = sums;
+ btrfs_add_ordered_sum(ordered, sums);
+
+ if (!async) {
+ csum_one_bio(bbio, &bbio->bio.bi_iter);
+ return 0;
+ }
+ init_completion(&bbio->csum_done);
+ bbio->async_csum = true;
+ bbio->csum_saved_iter = bbio->bio.bi_iter;
+ INIT_WORK(&bbio->csum_work, csum_one_bio_work);
+ schedule_work(&bbio->csum_work);
+ return 0;
+}
+
+/*
+ * Nodatasum I/O on zoned file systems still requires an btrfs_ordered_sum to
+ * record the updated logical address on Zone Append completion.
+ * Allocate just the structure with an empty sums array here for that case.
+ */
+int btrfs_alloc_dummy_sum(struct btrfs_bio *bbio)
+{
+ bbio->sums = kmalloc(sizeof(*bbio->sums), GFP_NOFS);
+ if (!bbio->sums)
+ return -ENOMEM;
+ bbio->sums->len = bbio->bio.bi_iter.bi_size;
+ bbio->sums->logical = bbio->bio.bi_iter.bi_sector << SECTOR_SHIFT;
+ btrfs_add_ordered_sum(bbio->ordered, bbio->sums);
return 0;
}
/*
- * helper function for csum removal, this expects the
- * key to describe the csum pointed to by the path, and it expects
- * the csum to overlap the range [bytenr, len]
+ * Remove one checksum overlapping a range.
+ *
+ * This expects the key to describe the csum pointed to by the path, and it
+ * expects the csum to overlap the range [bytenr, len]
*
- * The csum should not be entirely contained in the range and the
- * range should not be entirely contained in the csum.
+ * The csum should not be entirely contained in the range and the range should
+ * not be entirely contained in the csum.
*
- * This calls btrfs_truncate_item with the correct args based on the
- * overlap, and fixes up the key as required.
+ * This calls btrfs_truncate_item with the correct args based on the overlap,
+ * and fixes up the key as required.
*/
-static noinline void truncate_one_csum(struct btrfs_fs_info *fs_info,
+static noinline void truncate_one_csum(struct btrfs_trans_handle *trans,
struct btrfs_path *path,
struct btrfs_key *key,
u64 bytenr, u64 len)
{
+ struct btrfs_fs_info *fs_info = trans->fs_info;
struct extent_buffer *leaf;
- u16 csum_size = btrfs_super_csum_size(fs_info->super_copy);
+ const u32 csum_size = fs_info->csum_size;
u64 csum_end;
u64 end_byte = bytenr + len;
- u32 blocksize_bits = fs_info->sb->s_blocksize_bits;
+ u32 blocksize_bits = fs_info->sectorsize_bits;
leaf = path->nodes[0];
- csum_end = btrfs_item_size_nr(leaf, path->slots[0]) / csum_size;
- csum_end <<= fs_info->sb->s_blocksize_bits;
+ csum_end = btrfs_item_size(leaf, path->slots[0]) / csum_size;
+ csum_end <<= blocksize_bits;
csum_end += key->offset;
if (key->offset < bytenr && csum_end <= end_byte) {
@@ -544,7 +896,7 @@ static noinline void truncate_one_csum(struct btrfs_fs_info *fs_info,
*/
u32 new_size = (bytenr - key->offset) >> blocksize_bits;
new_size *= csum_size;
- btrfs_truncate_item(fs_info, path, new_size, 1);
+ btrfs_truncate_item(trans, path, new_size, 1);
} else if (key->offset >= bytenr && csum_end > end_byte &&
end_byte > key->offset) {
/*
@@ -556,31 +908,33 @@ static noinline void truncate_one_csum(struct btrfs_fs_info *fs_info,
u32 new_size = (csum_end - end_byte) >> blocksize_bits;
new_size *= csum_size;
- btrfs_truncate_item(fs_info, path, new_size, 0);
+ btrfs_truncate_item(trans, path, new_size, 0);
key->offset = end_byte;
- btrfs_set_item_key_safe(fs_info, path, key);
+ btrfs_set_item_key_safe(trans, path, key);
} else {
BUG();
}
}
/*
- * deletes the csum items from the csum tree for a given
- * range of bytes.
+ * Delete the csum items from the csum tree for a given range of bytes.
*/
int btrfs_del_csums(struct btrfs_trans_handle *trans,
- struct btrfs_fs_info *fs_info, u64 bytenr, u64 len)
+ struct btrfs_root *root, u64 bytenr, u64 len)
{
- struct btrfs_root *root = fs_info->csum_root;
- struct btrfs_path *path;
+ struct btrfs_fs_info *fs_info = trans->fs_info;
+ BTRFS_PATH_AUTO_FREE(path);
struct btrfs_key key;
u64 end_byte = bytenr + len;
u64 csum_end;
struct extent_buffer *leaf;
- int ret;
- u16 csum_size = btrfs_super_csum_size(fs_info->super_copy);
- int blocksize_bits = fs_info->sb->s_blocksize_bits;
+ int ret = 0;
+ const u32 csum_size = fs_info->csum_size;
+ u32 blocksize_bits = fs_info->sectorsize_bits;
+
+ ASSERT(btrfs_root_id(root) == BTRFS_CSUM_TREE_OBJECTID ||
+ btrfs_root_id(root) == BTRFS_TREE_LOG_OBJECTID);
path = btrfs_alloc_path();
if (!path)
@@ -588,12 +942,12 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans,
while (1) {
key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
- key.offset = end_byte - 1;
key.type = BTRFS_EXTENT_CSUM_KEY;
+ key.offset = end_byte - 1;
- path->leave_spinning = 1;
ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
if (ret > 0) {
+ ret = 0;
if (path->slots[0] == 0)
break;
path->slots[0]--;
@@ -612,7 +966,7 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans,
if (key.offset >= end_byte)
break;
- csum_end = btrfs_item_size_nr(leaf, path->slots[0]) / csum_size;
+ csum_end = btrfs_item_size(leaf, path->slots[0]) / csum_size;
csum_end <<= blocksize_bits;
csum_end += key.offset;
@@ -650,7 +1004,7 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans,
ret = btrfs_del_items(trans, root, path,
path->slots[0], del_nr);
if (ret)
- goto out;
+ break;
if (key.offset == bytenr)
break;
} else if (key.offset < bytenr && csum_end > end_byte) {
@@ -692,25 +1046,54 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans,
* item changed size or key
*/
ret = btrfs_split_item(trans, root, path, &key, offset);
- if (ret && ret != -EAGAIN) {
+ if (unlikely(ret && ret != -EAGAIN)) {
btrfs_abort_transaction(trans, ret);
- goto out;
+ break;
}
+ ret = 0;
key.offset = end_byte - 1;
} else {
- truncate_one_csum(fs_info, path, &key, bytenr, len);
+ truncate_one_csum(trans, path, &key, bytenr, len);
if (key.offset < bytenr)
break;
}
btrfs_release_path(path);
}
- ret = 0;
-out:
- btrfs_free_path(path);
return ret;
}
+static int find_next_csum_offset(struct btrfs_root *root,
+ struct btrfs_path *path,
+ u64 *next_offset)
+{
+ const u32 nritems = btrfs_header_nritems(path->nodes[0]);
+ struct btrfs_key found_key;
+ int slot = path->slots[0] + 1;
+ int ret;
+
+ if (nritems == 0 || slot >= nritems) {
+ ret = btrfs_next_leaf(root, path);
+ if (ret < 0) {
+ return ret;
+ } else if (ret > 0) {
+ *next_offset = (u64)-1;
+ return 0;
+ }
+ slot = path->slots[0];
+ }
+
+ btrfs_item_key_to_cpu(path->nodes[0], &found_key, slot);
+
+ if (found_key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
+ found_key.type != BTRFS_EXTENT_CSUM_KEY)
+ *next_offset = (u64)-1;
+ else
+ *next_offset = found_key.offset;
+
+ return 0;
+}
+
int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct btrfs_ordered_sum *sums)
@@ -718,7 +1101,7 @@ int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans,
struct btrfs_fs_info *fs_info = root->fs_info;
struct btrfs_key file_key;
struct btrfs_key found_key;
- struct btrfs_path *path;
+ BTRFS_PATH_AUTO_FREE(path);
struct btrfs_csum_item *item;
struct btrfs_csum_item *item_end;
struct extent_buffer *leaf = NULL;
@@ -726,12 +1109,11 @@ int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans,
u64 total_bytes = 0;
u64 csum_offset;
u64 bytenr;
- u32 nritems;
u32 ins_size;
int index = 0;
int found_next;
int ret;
- u16 csum_size = btrfs_super_csum_size(fs_info->super_copy);
+ const u32 csum_size = fs_info->csum_size;
path = btrfs_alloc_path();
if (!path)
@@ -739,10 +1121,10 @@ int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans,
again:
next_offset = (u64)-1;
found_next = 0;
- bytenr = sums->bytenr + total_bytes;
+ bytenr = sums->logical + total_bytes;
file_key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
- file_key.offset = bytenr;
file_key.type = BTRFS_EXTENT_CSUM_KEY;
+ file_key.offset = bytenr;
item = btrfs_lookup_csum(trans, root, path, bytenr, 1);
if (!IS_ERR(item)) {
@@ -751,55 +1133,56 @@ again:
item_end = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_csum_item);
item_end = (struct btrfs_csum_item *)((char *)item_end +
- btrfs_item_size_nr(leaf, path->slots[0]));
+ btrfs_item_size(leaf, path->slots[0]));
goto found;
}
ret = PTR_ERR(item);
if (ret != -EFBIG && ret != -ENOENT)
- goto fail_unlock;
+ goto out;
if (ret == -EFBIG) {
u32 item_size;
/* we found one, but it isn't big enough yet */
leaf = path->nodes[0];
- item_size = btrfs_item_size_nr(leaf, path->slots[0]);
+ item_size = btrfs_item_size(leaf, path->slots[0]);
if ((item_size / csum_size) >=
MAX_CSUM_ITEMS(fs_info, csum_size)) {
/* already at max size, make a new one */
goto insert;
}
} else {
- int slot = path->slots[0] + 1;
- /* we didn't find a csum item, insert one */
- nritems = btrfs_header_nritems(path->nodes[0]);
- if (!nritems || (path->slots[0] >= nritems - 1)) {
- ret = btrfs_next_leaf(root, path);
- if (ret == 1)
- found_next = 1;
- if (ret != 0)
- goto insert;
- slot = path->slots[0];
- }
- btrfs_item_key_to_cpu(path->nodes[0], &found_key, slot);
- if (found_key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
- found_key.type != BTRFS_EXTENT_CSUM_KEY) {
- found_next = 1;
- goto insert;
- }
- next_offset = found_key.offset;
+ /* We didn't find a csum item, insert one. */
+ ret = find_next_csum_offset(root, path, &next_offset);
+ if (ret < 0)
+ goto out;
found_next = 1;
goto insert;
}
/*
- * at this point, we know the tree has an item, but it isn't big
- * enough yet to put our csum in. Grow it
+ * At this point, we know the tree has a checksum item that ends at an
+ * offset matching the start of the checksum range we want to insert.
+ * We try to extend that item as much as possible and then add as many
+ * checksums to it as they fit.
+ *
+ * First check if the leaf has enough free space for at least one
+ * checksum. If it has go directly to the item extension code, otherwise
+ * release the path and do a search for insertion before the extension.
*/
+ if (btrfs_leaf_free_space(leaf) >= csum_size) {
+ btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
+ csum_offset = (bytenr - found_key.offset) >>
+ fs_info->sectorsize_bits;
+ goto extend_csum;
+ }
+
btrfs_release_path(path);
+ path->search_for_extension = true;
ret = btrfs_search_slot(trans, root, &file_key, path,
csum_size, 1);
+ path->search_for_extension = false;
if (ret < 0)
- goto fail_unlock;
+ goto out;
if (ret > 0) {
if (path->slots[0] == 0)
@@ -809,8 +1192,7 @@ again:
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
- csum_offset = (bytenr - found_key.offset) >>
- fs_info->sb->s_blocksize_bits;
+ csum_offset = (bytenr - found_key.offset) >> fs_info->sectorsize_bits;
if (found_key.type != BTRFS_EXTENT_CSUM_KEY ||
found_key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
@@ -818,34 +1200,68 @@ again:
goto insert;
}
- if (csum_offset == btrfs_item_size_nr(leaf, path->slots[0]) /
+extend_csum:
+ if (csum_offset == btrfs_item_size(leaf, path->slots[0]) /
csum_size) {
int extend_nr;
u64 tmp;
u32 diff;
- u32 free_space;
-
- if (btrfs_leaf_free_space(fs_info, leaf) <
- sizeof(struct btrfs_item) + csum_size * 2)
- goto insert;
- free_space = btrfs_leaf_free_space(fs_info, leaf) -
- sizeof(struct btrfs_item) - csum_size;
tmp = sums->len - total_bytes;
- tmp >>= fs_info->sb->s_blocksize_bits;
+ tmp >>= fs_info->sectorsize_bits;
WARN_ON(tmp < 1);
+ extend_nr = max_t(int, 1, tmp);
+
+ /*
+ * A log tree can already have checksum items with a subset of
+ * the checksums we are trying to log. This can happen after
+ * doing a sequence of partial writes into prealloc extents and
+ * fsyncs in between, with a full fsync logging a larger subrange
+ * of an extent for which a previous fast fsync logged a smaller
+ * subrange. And this happens in particular due to merging file
+ * extent items when we complete an ordered extent for a range
+ * covered by a prealloc extent - this is done at
+ * btrfs_mark_extent_written().
+ *
+ * So if we try to extend the previous checksum item, which has
+ * a range that ends at the start of the range we want to insert,
+ * make sure we don't extend beyond the start offset of the next
+ * checksum item. If we are at the last item in the leaf, then
+ * forget the optimization of extending and add a new checksum
+ * item - it is not worth the complexity of releasing the path,
+ * getting the first key for the next leaf, repeat the btree
+ * search, etc, because log trees are temporary anyway and it
+ * would only save a few bytes of leaf space.
+ */
+ if (btrfs_root_id(root) == BTRFS_TREE_LOG_OBJECTID) {
+ if (path->slots[0] + 1 >=
+ btrfs_header_nritems(path->nodes[0])) {
+ ret = find_next_csum_offset(root, path, &next_offset);
+ if (ret < 0)
+ goto out;
+ found_next = 1;
+ goto insert;
+ }
+
+ ret = find_next_csum_offset(root, path, &next_offset);
+ if (ret < 0)
+ goto out;
+
+ tmp = (next_offset - bytenr) >> fs_info->sectorsize_bits;
+ if (tmp <= INT_MAX)
+ extend_nr = min_t(int, extend_nr, tmp);
+ }
- extend_nr = max_t(int, 1, (int)tmp);
diff = (csum_offset + extend_nr) * csum_size;
diff = min(diff,
MAX_CSUM_ITEMS(fs_info, csum_size) * csum_size);
- diff = diff - btrfs_item_size_nr(leaf, path->slots[0]);
- diff = min(free_space, diff);
+ diff = diff - btrfs_item_size(leaf, path->slots[0]);
+ diff = min_t(u32, btrfs_leaf_free_space(leaf), diff);
diff /= csum_size;
diff *= csum_size;
- btrfs_extend_item(fs_info, path, diff);
+ btrfs_extend_item(trans, path, diff);
ret = 0;
goto csum;
}
@@ -857,9 +1273,9 @@ insert:
u64 tmp;
tmp = sums->len - total_bytes;
- tmp >>= fs_info->sb->s_blocksize_bits;
+ tmp >>= fs_info->sectorsize_bits;
tmp = min(tmp, (next_offset - file_key.offset) >>
- fs_info->sb->s_blocksize_bits);
+ fs_info->sectorsize_bits);
tmp = max_t(u64, 1, tmp);
tmp = min_t(u64, tmp, MAX_CSUM_ITEMS(fs_info, csum_size));
@@ -867,52 +1283,41 @@ insert:
} else {
ins_size = csum_size;
}
- path->leave_spinning = 1;
ret = btrfs_insert_empty_item(trans, root, path, &file_key,
ins_size);
- path->leave_spinning = 0;
if (ret < 0)
- goto fail_unlock;
- if (WARN_ON(ret != 0))
- goto fail_unlock;
+ goto out;
leaf = path->nodes[0];
csum:
item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_csum_item);
item_end = (struct btrfs_csum_item *)((unsigned char *)item +
- btrfs_item_size_nr(leaf, path->slots[0]));
+ btrfs_item_size(leaf, path->slots[0]));
item = (struct btrfs_csum_item *)((unsigned char *)item +
csum_offset * csum_size);
found:
- ins_size = (u32)(sums->len - total_bytes) >>
- fs_info->sb->s_blocksize_bits;
+ ins_size = (u32)(sums->len - total_bytes) >> fs_info->sectorsize_bits;
ins_size *= csum_size;
ins_size = min_t(u32, (unsigned long)item_end - (unsigned long)item,
ins_size);
write_extent_buffer(leaf, sums->sums + index, (unsigned long)item,
ins_size);
+ index += ins_size;
ins_size /= csum_size;
total_bytes += ins_size * fs_info->sectorsize;
- index += ins_size;
- btrfs_mark_buffer_dirty(path->nodes[0]);
if (total_bytes < sums->len) {
btrfs_release_path(path);
cond_resched();
goto again;
}
out:
- btrfs_free_path(path);
return ret;
-
-fail_unlock:
- goto out;
}
void btrfs_extent_item_to_extent_map(struct btrfs_inode *inode,
const struct btrfs_path *path,
- struct btrfs_file_extent_item *fi,
- const bool new_inline,
+ const struct btrfs_file_extent_item *fi,
struct extent_map *em)
{
struct btrfs_fs_info *fs_info = inode->root->fs_info;
@@ -920,69 +1325,80 @@ void btrfs_extent_item_to_extent_map(struct btrfs_inode *inode,
struct extent_buffer *leaf = path->nodes[0];
const int slot = path->slots[0];
struct btrfs_key key;
- u64 extent_start, extent_end;
- u64 bytenr;
+ u64 extent_start;
u8 type = btrfs_file_extent_type(leaf, fi);
int compress_type = btrfs_file_extent_compression(leaf, fi);
- em->bdev = fs_info->fs_devices->latest_bdev;
btrfs_item_key_to_cpu(leaf, &key, slot);
extent_start = key.offset;
-
- if (type == BTRFS_FILE_EXTENT_REG ||
- type == BTRFS_FILE_EXTENT_PREALLOC) {
- extent_end = extent_start +
- btrfs_file_extent_num_bytes(leaf, fi);
- } else if (type == BTRFS_FILE_EXTENT_INLINE) {
- size_t size;
- size = btrfs_file_extent_ram_bytes(leaf, fi);
- extent_end = ALIGN(extent_start + size,
- fs_info->sectorsize);
- }
-
em->ram_bytes = btrfs_file_extent_ram_bytes(leaf, fi);
+ em->generation = btrfs_file_extent_generation(leaf, fi);
if (type == BTRFS_FILE_EXTENT_REG ||
type == BTRFS_FILE_EXTENT_PREALLOC) {
+ const u64 disk_bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
+
em->start = extent_start;
- em->len = extent_end - extent_start;
- em->orig_start = extent_start -
- btrfs_file_extent_offset(leaf, fi);
- em->orig_block_len = btrfs_file_extent_disk_num_bytes(leaf, fi);
- bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
- if (bytenr == 0) {
- em->block_start = EXTENT_MAP_HOLE;
+ em->len = btrfs_file_extent_end(path) - extent_start;
+ if (disk_bytenr == 0) {
+ em->disk_bytenr = EXTENT_MAP_HOLE;
+ em->disk_num_bytes = 0;
+ em->offset = 0;
return;
}
+ em->disk_bytenr = disk_bytenr;
+ em->disk_num_bytes = btrfs_file_extent_disk_num_bytes(leaf, fi);
+ em->offset = btrfs_file_extent_offset(leaf, fi);
if (compress_type != BTRFS_COMPRESS_NONE) {
- set_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
- em->compress_type = compress_type;
- em->block_start = bytenr;
- em->block_len = em->orig_block_len;
+ btrfs_extent_map_set_compression(em, compress_type);
} else {
- bytenr += btrfs_file_extent_offset(leaf, fi);
- em->block_start = bytenr;
- em->block_len = em->len;
+ /*
+ * Older kernels can create regular non-hole data
+ * extents with ram_bytes smaller than disk_num_bytes.
+ * Not a big deal, just always use disk_num_bytes
+ * for ram_bytes.
+ */
+ em->ram_bytes = em->disk_num_bytes;
if (type == BTRFS_FILE_EXTENT_PREALLOC)
- set_bit(EXTENT_FLAG_PREALLOC, &em->flags);
+ em->flags |= EXTENT_FLAG_PREALLOC;
}
} else if (type == BTRFS_FILE_EXTENT_INLINE) {
- em->block_start = EXTENT_MAP_INLINE;
- em->start = extent_start;
- em->len = extent_end - extent_start;
- /*
- * Initialize orig_start and block_len with the same values
- * as in inode.c:btrfs_get_extent().
- */
- em->orig_start = EXTENT_MAP_HOLE;
- em->block_len = (u64)-1;
- if (!new_inline && compress_type != BTRFS_COMPRESS_NONE) {
- set_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
- em->compress_type = compress_type;
- }
+ /* Tree-checker has ensured this. */
+ ASSERT(extent_start == 0);
+
+ em->disk_bytenr = EXTENT_MAP_INLINE;
+ em->start = 0;
+ em->len = fs_info->sectorsize;
+ em->offset = 0;
+ btrfs_extent_map_set_compression(em, compress_type);
} else {
btrfs_err(fs_info,
"unknown file extent item type %d, inode %llu, offset %llu, "
"root %llu", type, btrfs_ino(inode), extent_start,
- root->root_key.objectid);
+ btrfs_root_id(root));
}
}
+
+/*
+ * Returns the end offset (non inclusive) of the file extent item the given path
+ * points to. If it points to an inline extent, the returned offset is rounded
+ * up to the sector size.
+ */
+u64 btrfs_file_extent_end(const struct btrfs_path *path)
+{
+ const struct extent_buffer *leaf = path->nodes[0];
+ const int slot = path->slots[0];
+ struct btrfs_file_extent_item *fi;
+ struct btrfs_key key;
+ u64 end;
+
+ btrfs_item_key_to_cpu(leaf, &key, slot);
+ ASSERT(key.type == BTRFS_EXTENT_DATA_KEY);
+ fi = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);
+
+ if (btrfs_file_extent_type(leaf, fi) == BTRFS_FILE_EXTENT_INLINE)
+ end = leaf->fs_info->sectorsize;
+ else
+ end = key.offset + btrfs_file_extent_num_bytes(leaf, fi);
+
+ return end;
+}
generated by cgit (git 2.34.1) at 2025-12-24 21:58:39 +0000