@@ -2818,10 +2818,22 @@ int btrfs_inc_block_group_ro(struct btrfs_block_group *cache,
}
ret = inc_block_group_ro(cache, 0);
- if (!do_chunk_alloc || ret == -ETXTBSY)
- goto unlock_out;
if (!ret)
goto out;
+ if (ret == -ETXTBSY)
+ goto unlock_out;
+
+ /*
+ * Skip chunk alloc if the bg is SYSTEM, this is to avoid
+ * system chunk allocation storm to exhaust the system chunk
+ * array.
+ * Otherwise we still want to try our best to mark the
+ * block group read-only.
+ */
+ if (!do_chunk_alloc && ret == -ENOSPC &&
+ (cache->flags & BTRFS_BLOCK_GROUP_SYSTEM))
+ goto unlock_out;
+
alloc_flags = btrfs_get_alloc_profile(fs_info, cache->space_info->flags);
ret = btrfs_chunk_alloc(trans, alloc_flags, CHUNK_ALLOC_FORCE);
if (ret < 0)
@@ -2518,13 +2518,20 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
if (ret == 0) {
ro_set = 1;
- } else if (ret == -ENOSPC && !sctx->is_dev_replace) {
+ } else if (ret == -ENOSPC && !sctx->is_dev_replace &&
+ !(cache->flags & BTRFS_BLOCK_GROUP_RAID56_MASK)) {
/*
* btrfs_inc_block_group_ro return -ENOSPC when it
* failed in creating new chunk for metadata.
* It is not a problem for scrub, because
* metadata are always cowed, and our scrub paused
* commit_transactions.
+ *
+ * For RAID56 chunks, we have to mark them read-only
+ * for scrub, as later we would use our own cache
+ * out of RAID56 realm.
+ * Thus we want the RAID56 bg to br marked RO to
+ * prevent RMW from screwing up out cache.
*/
ro_set = 0;
} else if (ret == -ETXTBSY) {
Currently we allows a block group not to be marked read-only for scrub. But for RAID56 block groups if we require the block group to be read-only, then we're allowed to use cached content from scrub stripe to reduce unnecessary raid56 reads. So this patch would: - Make btrfs_inc_block_group_ro() to try harder During my tests, for cases like btrfs/061 and btrfs/064, we can hit -ENOSPC from btrfs_inc_block_group_ro() calls during scrub. The reason is if we only have one single data chunk, and trying to scrub it, we won't have any space left for any newer data writes. But this check should be done by the caller, especially for scrub cases we only temporarily mark the chunk read-only. And newer data writes would always try to allocate a new data chunk when needed. - Return error for scrub if we failed to mark a RAID56 chunk read-only Signed-off-by: Qu Wenruo <wqu@suse.com> --- fs/btrfs/block-group.c | 16 ++++++++++++++-- fs/btrfs/scrub.c | 9 ++++++++- 2 files changed, 22 insertions(+), 3 deletions(-)