On 13/05/2020 14:17, Johannes Thumshirn wrote:
[...]

quoted

quoted

This doesn't make sense.

That line doesn't even call read_tree_block() nor even any function call.

This looks really strange.

Indeed, it does. I have no clue what's going on here.

I just retested without KASAN enabled (to rule out any problems in that area) and 
the test (btrfs/028) passes. 

Turning KASAN back on fails the test again. So I /suspect/ it's a race somewhere and 
changed timing (due to KASAN) triggers it.