Comment # 14 on bug 1206681 from Jiri Slaby

(In reply to Michal Suchanek from comment #12)
> Did not notice a problem with 6.1.1-1-default so far, immediately triggered
> with 6.1.4-1-default, definitely not a problem with 6.0.12-1-default

git log --oneline  v6.1.1..v6.1.4 fs/btrfs/|sed 's@^@> @'
> a94b90ac1f25 btrfs: fix resolving backrefs for inline extent followed by prealloc
> 169a4cf46882 btrfs: fix extent map use-after-free when handling missing device in read_one_chunk
> 712b0938166a btrfs: fix uninitialized parent in insert_state
> 1baf3370e2dc btrfs: do not BUG_ON() on ENOMEM when dropping extent items for a range
> ef7cf3c92382 btrfs: do not panic if we can't allocate a prealloc extent state

Maybe it rings a bell?