On Sunday, April 5, 2020 12:44:38 PM CEST Ankur Sinha wrote:
On Sat, Apr 04, 2020 21:37:05 -0000, Artem Tim wrote:
That does look like it. I'll go through the comments and see if I can
help with more info.
Note also this bug  where's kernel traceback in bfq kernel code.
I have never reproduced the problem on my box (ssd nvme disk), so I can't
tell.. but those are possible workarounds for those who use bfq scheduler:
- per Zbyszek, switching back to deadline
- per Vitaly, systemd.unified_cgroup_hierarchy=0 kernel parameter