reada background works is not designed to finish all jobs
completely, it will break in following case:
1: When a device reachs workload limit(MAX_IN_FLIGHT)
2: Total reads reachs max limit(10000)
3: All devices haven't queue more jobs, often happened in DUP case
And if all background works exit with remain jobs,
btrfs_reada_wait() will enter into infinite wait.
Above problem is rarely happened in old code, because:
1: Every works queues 2x new works
So many works reduced odds of undone jobs.
2: One work will continue 10000 times loop in case of no-jobs
It reduced no-thread window time.
But after we fixed above case, the "undone reada extents" frequently
happened.
Fix:
Check to ensure we have at least one thread if there are undone jobs
in btrfs_reada_wait().
Changelog v1->v2:
1: RFC->PATCH
2: Some cleanup and wait_time adjust.
Signed-off-by: Zhao Lei <zhaolei@xxxxxxxxxxxxxx>
---
fs/btrfs/reada.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/fs/btrfs/reada.c b/fs/btrfs/reada.c
index 245a961..12a9c9c 100644
--- a/fs/btrfs/reada.c
+++ b/fs/btrfs/reada.c
@@ -953,8 +953,11 @@ struct reada_control *btrfs_reada_add(struct btrfs_root *root,
int btrfs_reada_wait(void *handle)
{
struct reada_control *rc = handle;
+ struct btrfs_fs_info *fs_info = rc->root->fs_info;
while (atomic_read(&rc->elems)) {
+ if (!atomic_read(&fs_info->reada_works_cnt))
+ reada_start_machine(fs_info);
wait_event_timeout(rc->wait, atomic_read(&rc->elems) == 0,
5 * HZ);
dump_devs(rc->root->fs_info,
@@ -971,9 +974,13 @@ int btrfs_reada_wait(void *handle)
int btrfs_reada_wait(void *handle)
{
struct reada_control *rc = handle;
+ struct btrfs_fs_info *fs_info = rc->root->fs_info;
while (atomic_read(&rc->elems)) {
- wait_event(rc->wait, atomic_read(&rc->elems) == 0);
+ if (!atomic_read(&fs_info->reada_works_cnt))
+ reada_start_machine(fs_info);
+ wait_event_timeout(rc->wait, atomic_read(&rc->elems) == 0,
+ (HZ + 9) / 10);
}
kref_put(&rc->refcnt, reada_control_release);
--
1.8.5.1
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
