[PATCH V4 5/5] ext4: flush background discard kwork when retry allocation
From: Wang Jianchao <hidden>
Date: 2021-08-30 07:54:27
Also in:
lkml
Subsystem:
ext4 file system, filesystems (vfs and infrastructure), the rest · Maintainers:
"Theodore Ts'o", Alexander Viro, Christian Brauner, Linus Torvalds
From: Wang Jianchao <redacted> The background discard kwork tries to mark blocks used and issue discard. This can make filesystem suffer from NOSPC error, xfstest generic/371 can fail due to it. Fix it by flushing discard kwork in ext4_should_retry_alloc. At the same time, give up discard at the moment. Signed-off-by: Wang Jianchao <redacted> --- fs/ext4/balloc.c | 8 +++++++- fs/ext4/ext4.h | 1 + fs/ext4/mballoc.c | 7 +++++-- 3 files changed, 13 insertions(+), 3 deletions(-)
diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index 9dc6e74b265c..a0fb0c4bdc7c 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c@@ -652,8 +652,14 @@ int ext4_should_retry_alloc(struct super_block *sb, int *retries) * possible we just missed a transaction commit that did so */ smp_mb(); - if (sbi->s_mb_free_pending == 0) + if (sbi->s_mb_free_pending == 0) { + if (test_opt(sb, DISCARD)) { + atomic_inc(&sbi->s_retry_alloc_pending); + flush_work(&sbi->s_discard_work); + atomic_dec(&sbi->s_retry_alloc_pending); + } return ext4_has_free_clusters(sbi, 1, 0); + } /* * it's possible we've just missed a transaction commit here,
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 6b678b968d84..d71dcac3b97f 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h@@ -1538,6 +1538,7 @@ struct ext4_sb_info { after commit completed */ struct list_head s_discard_list; struct work_struct s_discard_work; + atomic_t s_retry_alloc_pending; struct rb_root s_mb_avg_fragment_size_root; rwlock_t s_mb_rb_lock; struct list_head *s_mb_largest_free_orders;
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 259822fc0ae9..c0dea52a7124 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c@@ -3331,9 +3331,11 @@ static void ext4_discard_work(struct work_struct *work) load_grp = UINT_MAX; list_for_each_entry_safe(fd, nfd, &discard_list, efd_list) { /* - * If filesystem is umounting or no memory, give up the discard + * If filesystem is umounting or no memory or suffering + * from no space, give up the discard */ - if ((sb->s_flags & SB_ACTIVE) && !err) { + if ((sb->s_flags & SB_ACTIVE) && !err && + !atomic_read(&sbi->s_retry_alloc_pending)) { grp = fd->efd_group; if (grp != load_grp) { if (load_grp != UINT_MAX)
@@ -3431,6 +3433,7 @@ int ext4_mb_init(struct super_block *sb) INIT_LIST_HEAD(&sbi->s_freed_data_list); INIT_LIST_HEAD(&sbi->s_discard_list); INIT_WORK(&sbi->s_discard_work, ext4_discard_work); + atomic_set(&sbi->s_retry_alloc_pending, 0); sbi->s_mb_max_to_scan = MB_DEFAULT_MAX_TO_SCAN; sbi->s_mb_min_to_scan = MB_DEFAULT_MIN_TO_SCAN;
--
2.17.1