btrfs: zoned: limit max extent size to max zone append
cover btrfs: zoned: limit max extent size to max zone append
The first patch factors out retrieving BTRFS_MAX_EXTENT_SIZE into a helper and the second patch fixes the actual problem for zoned btrfs.
Johannes Thumshirn (2):
btrfs: add a helper for retrieving BTRFS_MAX_EXTENT_SIZE
btrfs: zoned: limit ordered extent to zoned append size
fs/btrfs/ctree.h | 26 ++++++++++++++++++-------- fs/btrfs/delalloc-space.c | 6 +++--- fs/btrfs/extent_io.c | 2 +- fs/btrfs/inode.c | 22 ++++++++++++---------- fs/btrfs/tests/extent-io-tests.c | 2 +- fs/btrfs/tests/inode-tests.c | 31 ++++++++++++++++--------------- 6 files changed, 51 insertions(+), 38 deletions(-)
Signature · 1 line
2.31.1
PATCH 1/2 btrfs: add a helper for retrieving BTRFS_MAX_EXTENT_SIZE
Add a helper for retrieving BTRFS_MAX_EXTENT_SIZE as a preparation for a fix in this area.
This patch has no functional changes.
--- fs/btrfs/ctree.h | 23 +++++++++++++++-------- fs/btrfs/delalloc-space.c | 6 +++--- fs/btrfs/extent_io.c | 2 +- fs/btrfs/inode.c | 22 ++++++++++++---------- fs/btrfs/tests/extent-io-tests.c | 2 +- fs/btrfs/tests/inode-tests.c | 31 ++++++++++++++++--------------- 6 files changed, 48 insertions(+), 38 deletions(-)
fs/btrfs/ctree.h +15 -8
@@ -105,14 +105,6 @@ struct btrfs_ref;#define BTRFS_STAT_CURR 0#define BTRFS_STAT_PREV 1 -/* -* Count how many BTRFS_MAX_EXTENT_SIZE cover the @size -*/ -static inline u32 count_max_extents (u64 size) -{ -return div_u64 (size +BTRFS_MAX_EXTENT_SIZE -1 ,BTRFS_MAX_EXTENT_SIZE ); -} -static inline unsigned long btrfs_chunk_item_size (int num_stripes) {BUG_ON (num_stripes ==0 );
@@ -1379,6 +1371,21 @@ static inline u32 BTRFS_MAX_XATTR_SIZE(const struct btrfs_fs_info *info)return BTRFS_MAX_ITEM_SIZE (info) -sizeof (struct btrfs_dir_item ); } +static inline u64 btrfs_get_max_extent_size (struct btrfs_fs_info *fs_info) +{ +return BTRFS_MAX_EXTENT_SIZE ; +} + +/* +* Count how many BTRFS_MAX_EXTENT_SIZE cover the @size +*/ +static inline u32 count_max_extents (struct btrfs_fs_info *fs_info,u64 size) +{ +u64 max_extent_size =btrfs_get_max_extent_size (fs_info); + +return div_u64 (size + max_extent_size -1 , max_extent_size); +} +/* * Flags for mount options. *
fs/btrfs/delalloc-space.c +3 -3
@@ -270,7 +270,7 @@ static void calc_inode_reservations(struct btrfs_fs_info *fs_info, u64 num_bytes, u64 *meta_reserve, u64 *qgroup_reserve) { - u64 nr_extents = count_max_extents(num_bytes); + u64 nr_extents = count_max_extents(fs_info, num_bytes);u64 csum_leaves =btrfs_csum_bytes_to_leaves (fs_info, num_bytes);u64 inode_update =btrfs_calc_metadata_size (fs_info,1 );
@@ -344,7 +344,7 @@ int btrfs_delalloc_reserve_metadata(struct btrfs_inode *inode, u64 num_bytes)* needs to free the reservation we just made. */ spin_lock (&inode->lock); - nr_extents =count_max_extents (num_bytes); + nr_extents =count_max_extents (fs_info, num_bytes);btrfs_mod_outstanding_extents (inode, nr_extents); inode->csum_bytes += num_bytes;btrfs_calculate_inode_block_rsv_size (fs_info, inode);
@@ -407,7 +407,7 @@ void btrfs_delalloc_release_extents(struct btrfs_inode *inode, u64 num_bytes)unsigned num_extents;spin_lock (&inode->lock); - num_extents =count_max_extents (num_bytes); + num_extents =count_max_extents (fs_info, num_bytes);btrfs_mod_outstanding_extents (inode, -num_extents);btrfs_calculate_inode_block_rsv_size (fs_info, inode);spin_unlock (&inode->lock);
fs/btrfs/extent_io.c +1 -1
@@ -1863,7 +1863,7 @@ noinline_for_stack bool find_lock_delalloc_range(struct inode *inode, u64 *end) {struct extent_io_tree *tree = &BTRFS_I (inode)->io_tree; -u64 max_bytes =BTRFS_MAX_EXTENT_SIZE ; +u64 max_bytes =btrfs_get_max_extent_size (btrfs_sb (inode->i_sb));u64 delalloc_start;u64 delalloc_end;bool found;
fs/btrfs/inode.c +12 -10
@@ -1936,6 +1936,7 @@ int btrfs_run_delalloc_range(struct btrfs_inode *inode, struct page *locked_pagevoid btrfs_split_delalloc_extent (struct inode *inode,struct extent_state *orig,u64 split) { +struct btrfs_fs_info *fs_info =btrfs_sb (inode->i_sb);u64 size;/* not delalloc, ignore it */
@@ -1943,7 +1944,7 @@ void btrfs_split_delalloc_extent(struct inode *inode,return ; size = orig->end - orig->start +1 ; -if (size >BTRFS_MAX_EXTENT_SIZE ) { +if (size >btrfs_get_max_extent_size (fs_info)) {u32 num_extents;u64 new_size;
@@ -1952,10 +1953,10 @@ void btrfs_split_delalloc_extent(struct inode *inode,* applies here, just in reverse. */ new_size = orig->end - split +1 ; - num_extents =count_max_extents (new_size); + num_extents =count_max_extents (fs_info, new_size); new_size = split - orig->start; - num_extents +=count_max_extents (new_size); -if (count_max_extents (size) >= num_extents) + num_extents +=count_max_extents (fs_info, new_size); +if (count_max_extents (fs_info, size) >= num_extents)return ; }
@@ -1972,6 +1973,7 @@ void btrfs_split_delalloc_extent(struct inode *inode,void btrfs_merge_delalloc_extent (struct inode *inode,struct extent_state *new,struct extent_state *other) { +struct btrfs_fs_info *fs_info =btrfs_sb (inode->i_sb);u64 new_size, old_size;u32 num_extents;
@@ -1985,7 +1987,7 @@ void btrfs_merge_delalloc_extent(struct inode *inode, struct extent_state *new, new_size = other->end - new->start +1 ;/* we're not bigger than the max, unreserve the space and go */ -if (new_size <=BTRFS_MAX_EXTENT_SIZE ) { +if (new_size <=btrfs_get_max_extent_size (fs_info)) {spin_lock (&BTRFS_I (inode)->lock);btrfs_mod_outstanding_extents (BTRFS_I (inode),-1 );spin_unlock (&BTRFS_I (inode)->lock);
@@ -2011,10 +2013,10 @@ void btrfs_merge_delalloc_extent(struct inode *inode, struct extent_state *new,* this case. */ old_size = other->end - other->start +1 ; - num_extents =count_max_extents (old_size); + num_extents =count_max_extents (fs_info, old_size); old_size = new->end - new->start +1 ; - num_extents +=count_max_extents (old_size); -if (count_max_extents (new_size) >= num_extents) + num_extents +=count_max_extents (fs_info, old_size); +if (count_max_extents (fs_info, new_size) >= num_extents)return ;spin_lock (&BTRFS_I (inode)->lock);
@@ -2093,7 +2095,7 @@ void btrfs_set_delalloc_extent(struct inode *inode, struct extent_state *state,if (!(state->state &EXTENT_DELALLOC ) && (*bits &EXTENT_DELALLOC )) {struct btrfs_root *root =BTRFS_I (inode)->root;u64 len = state->end +1 - state->start; -u32 num_extents =count_max_extents (len); +u32 num_extents =count_max_extents (fs_info, len);bool do_list = !btrfs_is_free_space_inode (BTRFS_I (inode));spin_lock (&BTRFS_I (inode)->lock);
@@ -2135,7 +2137,7 @@ void btrfs_clear_delalloc_extent(struct inode *vfs_inode,struct btrfs_inode *inode =BTRFS_I (vfs_inode);struct btrfs_fs_info *fs_info =btrfs_sb (vfs_inode->i_sb);u64 len = state->end +1 - state->start; -u32 num_extents =count_max_extents (len); +u32 num_extents =count_max_extents (fs_info, len);if ((state->state &EXTENT_DEFRAG ) && (*bits &EXTENT_DEFRAG )) {spin_lock (&inode->lock);
fs/btrfs/tests/extent-io-tests.c +1 -1
@@ -64,7 +64,7 @@ static int test_find_delalloc(u32 sectorsize)struct page *locked_page =NULL ;unsigned long index =0 ;/* In this test we need at least 2 file extents at its maximum size */ -u64 max_bytes =BTRFS_MAX_EXTENT_SIZE ; +u64 max_bytes =btrfs_get_max_extent_size (NULL );u64 total_dirty =2 * max_bytes;u64 start, end, test_start;bool found;
fs/btrfs/tests/inode-tests.c +16 -15
@@ -917,6 +917,7 @@ static int test_extent_accounting(u32 sectorsize, u32 nodesize)struct btrfs_fs_info *fs_info =NULL ;struct inode *inode =NULL ;struct btrfs_root *root =NULL ; +u64 max_extent_size =btrfs_get_max_extent_size (NULL );int ret = -ENOMEM ;test_msg ("running outstanding_extents tests" );
@@ -943,7 +944,7 @@ static int test_extent_accounting(u32 sectorsize, u32 nodesize)/* [BTRFS_MAX_EXTENT_SIZE] */ ret =btrfs_set_extent_delalloc (BTRFS_I (inode),0 , -BTRFS_MAX_EXTENT_SIZE -1 ,0 ,NULL ); + max_extent_size -1 ,0 ,NULL );if (ret) {test_err ("btrfs_set_extent_delalloc returned %d" , ret); goto out;
@@ -956,8 +957,8 @@ static int test_extent_accounting(u32 sectorsize, u32 nodesize) }/* [BTRFS_MAX_EXTENT_SIZE][sectorsize] */ - ret =btrfs_set_extent_delalloc (BTRFS_I (inode),BTRFS_MAX_EXTENT_SIZE , -BTRFS_MAX_EXTENT_SIZE + sectorsize -1 , + ret =btrfs_set_extent_delalloc (BTRFS_I (inode), max_extent_size, + max_extent_size + sectorsize -1 ,0 ,NULL );if (ret) {test_err ("btrfs_set_extent_delalloc returned %d" , ret);
@@ -972,8 +973,8 @@ static int test_extent_accounting(u32 sectorsize, u32 nodesize)/* [BTRFS_MAX_EXTENT_SIZE/2][sectorsize HOLE][the rest] */ ret =clear_extent_bit (&BTRFS_I (inode)->io_tree, -BTRFS_MAX_EXTENT_SIZE >>1 , - (BTRFS_MAX_EXTENT_SIZE >>1 ) + sectorsize -1 , + max_extent_size >>1 , + (max_extent_size >>1 ) + sectorsize -1 ,EXTENT_DELALLOC |EXTENT_DELALLOC_NEW |EXTENT_UPTODATE ,0 ,0 ,NULL );if (ret) {
@@ -988,8 +989,8 @@ static int test_extent_accounting(u32 sectorsize, u32 nodesize) }/* [BTRFS_MAX_EXTENT_SIZE][sectorsize] */ - ret =btrfs_set_extent_delalloc (BTRFS_I (inode),BTRFS_MAX_EXTENT_SIZE >>1 , - (BTRFS_MAX_EXTENT_SIZE >>1 ) + ret =btrfs_set_extent_delalloc (BTRFS_I (inode), max_extent_size >>1 , + (max_extent_size >>1 ) + sectorsize -1 ,0 ,NULL );if (ret) {
@@ -1007,8 +1008,8 @@ static int test_extent_accounting(u32 sectorsize, u32 nodesize)* [BTRFS_MAX_EXTENT_SIZE+sectorsize][sectorsize HOLE][BTRFS_MAX_EXTENT_SIZE+sectorsize] */ ret =btrfs_set_extent_delalloc (BTRFS_I (inode), -BTRFS_MAX_EXTENT_SIZE +2 * sectorsize, - (BTRFS_MAX_EXTENT_SIZE <<1 ) +3 * sectorsize -1 , + max_extent_size +2 * sectorsize, + (max_extent_size <<1 ) +3 * sectorsize -1 ,0 ,NULL );if (ret) {test_err ("btrfs_set_extent_delalloc returned %d" , ret);
@@ -1025,8 +1026,8 @@ static int test_extent_accounting(u32 sectorsize, u32 nodesize)* [BTRFS_MAX_EXTENT_SIZE+sectorsize][sectorsize][BTRFS_MAX_EXTENT_SIZE+sectorsize] */ ret =btrfs_set_extent_delalloc (BTRFS_I (inode), -BTRFS_MAX_EXTENT_SIZE + sectorsize, -BTRFS_MAX_EXTENT_SIZE +2 * sectorsize -1 ,0 ,NULL ); + max_extent_size + sectorsize, + max_extent_size +2 * sectorsize -1 ,0 ,NULL );if (ret) {test_err ("btrfs_set_extent_delalloc returned %d" , ret); goto out;
@@ -1040,8 +1041,8 @@ static int test_extent_accounting(u32 sectorsize, u32 nodesize)/* [BTRFS_MAX_EXTENT_SIZE+4k][4K HOLE][BTRFS_MAX_EXTENT_SIZE+4k] */ ret =clear_extent_bit (&BTRFS_I (inode)->io_tree, -BTRFS_MAX_EXTENT_SIZE + sectorsize, -BTRFS_MAX_EXTENT_SIZE +2 * sectorsize -1 , + max_extent_size + sectorsize, + max_extent_size +2 * sectorsize -1 ,EXTENT_DELALLOC |EXTENT_DELALLOC_NEW |EXTENT_UPTODATE ,0 ,0 ,NULL );if (ret) {
@@ -1060,8 +1061,8 @@ static int test_extent_accounting(u32 sectorsize, u32 nodesize)* might fail and I'd rather satisfy my paranoia at this point. */ ret =btrfs_set_extent_delalloc (BTRFS_I (inode), -BTRFS_MAX_EXTENT_SIZE + sectorsize, -BTRFS_MAX_EXTENT_SIZE +2 * sectorsize -1 ,0 ,NULL ); + max_extent_size + sectorsize, + max_extent_size +2 * sectorsize -1 ,0 ,NULL );if (ret) {test_err ("btrfs_set_extent_delalloc returned %d" , ret); goto out;
Signature · 1 line
2.31.1
PATCH 2/2 btrfs: zoned: limit ordered extent to zoned append size
In reply to Johannes ThumshirnDamien reported a test failure with btrfs/209. The test itself ran fine, but the fsck run afterwards reported a corrupted filesystem.
The filesystem corruption happens because we're splitting an extent and then writing the extent twice. We have to split the extent though, because we're creating too large extents for a REQ_OP_ZONE_APPEND operation.
When dumping the extent tree, we can see two EXTENT_ITEMs at the same start address but different lengths.
$ btrfs inspect dump-tree /dev/nullb1 -t extent
...
item 19 key (269484032 EXTENT_ITEM 126976) itemoff 15470 itemsize 53
refs 1 gen 7 flags DATA
extent data backref root FS_TREE objectid 257 offset 786432 count 1
item 20 key (269484032 EXTENT_ITEM 262144) itemoff 15417 itemsize 53
refs 1 gen 7 flags DATA
extent data backref root FS_TREE objectid 257 offset 786432 count 1On a zoned filesystem, limit the size of an ordered extent to the maximum size that can be issued as a single REQ_OP_ZONE_APPEND operation.
Note: This patch breaks fstests btrfs/079, as it increases the number of on-disk extents from 80 to 83 per 10M write.
--- fs/btrfs/ctree.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
fs/btrfs/ctree.h +4 -1
@@ -1373,7 +1373,10 @@ static inline u32 BTRFS_MAX_XATTR_SIZE(const struct btrfs_fs_info *info)static inline u64 btrfs_get_max_extent_size (struct btrfs_fs_info *fs_info) { -return BTRFS_MAX_EXTENT_SIZE ; +if (!fs_info || !fs_info->max_zone_append_size) +return BTRFS_MAX_EXTENT_SIZE ; +return min_t (u64,BTRFS_MAX_EXTENT_SIZE , +ALIGN_DOWN (fs_info->max_zone_append_size,PAGE_SIZE )); }/*
Signature · 1 line
2.31.1
Re: [PATCH 2/2] btrfs: zoned: limit ordered extent to zoned append size
Quoted 36 lines · Johannes Thumshirn
Damien reported a test failure with btrfs/209. The test itself ran fine, but the fsck run afterwards reported a corrupted filesystem.
The filesystem corruption happens because we're splitting an extent and then writing the extent twice. We have to split the extent though, because we're creating too large extents for a
REQ_OP_ZONE_APPENDoperation.When dumping the extent tree, we can see two
EXTENT_ITEMsat the same start address but different lengths.$ btrfs inspect dump-tree /dev/nullb1 -t extent ... item 19 key (269484032 EXTENT_ITEM 126976) itemoff 15470 itemsize 53 refs 1 gen 7 flags DATA extent data backref root FS_TREE objectid 257 offset 786432 count 1 item 20 key (269484032 EXTENT_ITEM 262144) itemoff 15417 itemsize 53 refs 1 gen 7 flags DATA extent data backref root FS_TREE objectid 257 offset 786432 count 1On a zoned filesystem, limit the size of an ordered extent to the maximum size that can be issued as a single
REQ_OP_ZONE_APPENDoperation.Note: This patch breaks fstests btrfs/079, as it increases the number of on-disk extents from 80 to 83 per 10M write.
Reported-by Damien Le Moal <[email protected]>Signed-off-by Johannes Thumshirn <[email protected]>--- fs/btrfs/ctree.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)fs/btrfs/ctree.h +0
-0
fs/btrfs/ctree.h static inline u32 BTRFS_MAX_XATTR_SIZE(const struct btrfs_fs_info *info) · line 1379
Quoted 8 lines
fs/btrfs/ctree.h +4
-1@@ -1373,7 +1373,10 @@ static inline u32 BTRFS_MAX_XATTR_SIZE(const struct btrfs_fs_info *info)static inline u64 btrfs_get_max_extent_size (struct btrfs_fs_info *fs_info) { -return BTRFS_MAX_EXTENT_SIZE ; +if (!fs_info || !fs_info->max_zone_append_size) +return BTRFS_MAX_EXTENT_SIZE ; +return min_t (u64,BTRFS_MAX_EXTENT_SIZE , +ALIGN_DOWN (fs_info->max_zone_append_size,PAGE_SIZE ));
Should this be set only once in btrfs_check_zoned_mode ?
Quoted 6 lines
fs/btrfs/ctree.h +0
-0}/* Signature · 1 line
2.31.1
Re: [PATCH 2/2] btrfs: zoned: limit ordered extent to zoned append size
Quoted 10 lines · David Sterba
Quoted 10 lines
--- fs/btrfs/ctree.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)fs/btrfs/ctree.h +0
-0@@ -1373,7 +1373,10 @@ static inline u32 BTRFS_MAX_XATTR_SIZE(const struct btrfs_fs_info *info)
Quoted 9 lines
fs/btrfs/ctree.h static inline u32 BTRFS_MAX_XATTR_SIZE(const struct btrfs_fs_info *info) · line 1379
Quoted 7 lines
fs/btrfs/ctree.h +4
-1static inline u64 btrfs_get_max_extent_size (struct btrfs_fs_info *fs_info) { -return BTRFS_MAX_EXTENT_SIZE ; +if (!fs_info || !fs_info->max_zone_append_size) +return BTRFS_MAX_EXTENT_SIZE ; +return min_t (u64,BTRFS_MAX_EXTENT_SIZE , +ALIGN_DOWN (fs_info->max_zone_append_size,PAGE_SIZE ));Should this be set only once in
btrfs_check_zoned_mode?
Do you mean adding a fs_info->max_extent_size member or fs_info->max_zone_append_size = min(BTRFS_MAX_EXTENT_SIZE, queue_max_append_size())?
I'd opt for #1
Re: [PATCH 2/2] btrfs: zoned: limit ordered extent to zoned append size
Quoted 21 lines · Johannes Thumshirn
Quoted 10 lines · David Sterba
Quoted 10 lines
--- fs/btrfs/ctree.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)fs/btrfs/ctree.h +0
-0@@ -1373,7 +1373,10 @@ static inline u32 BTRFS_MAX_XATTR_SIZE(const struct btrfs_fs_info *info)Quoted 9 lines
fs/btrfs/ctree.h static inline u32 BTRFS_MAX_XATTR_SIZE(const struct btrfs_fs_info *info) · line 1379
Quoted 7 lines
fs/btrfs/ctree.h +4
-1static inline u64 btrfs_get_max_extent_size (struct btrfs_fs_info *fs_info) { -return BTRFS_MAX_EXTENT_SIZE ; +if (!fs_info || !fs_info->max_zone_append_size) +return BTRFS_MAX_EXTENT_SIZE ; +return min_t (u64,BTRFS_MAX_EXTENT_SIZE , +ALIGN_DOWN (fs_info->max_zone_append_size,PAGE_SIZE ));Should this be set only once in
btrfs_check_zoned_mode?
Quoted 4 lines
Do you mean adding a
fs_info->max_extent_sizemember orfs_info->max_zone_append_size=min(BTRFS_MAX_EXTENT_SIZE, queue_max_append_size())?I'd opt for #1
Yeah, replace BTRFS_MAX_EXTENT_SIZE by a fs_info member and then adjust it for max append zone.