On Fri, Oct 25, 2024 at 05:52:57PM +0800, Lai, Yi wrote: > On Wed, Oct 23, 2024 at 09:57:53PM -0400, Kent Overstreet wrote: > > On Mon, Oct 21, 2024 at 06:27:25PM +0800, Lai, Yi wrote: > > > Hi Kent Overstreet, > > > > > > Greetings! > > > > > > I used Syzkaller and found that there is INFO: task hung in __rq_qos_throttle in v6.12-rc2 > > > > > > After bisection and the first bad commit is: > > > " > > > 63332394c7e1 bcachefs: Move snapshot table size to struct snapshot_table > > > > You sure...? > > > > Look at the patch, that's a pretty unlikely culprit; we would've seen > > something from kasan, and anyways there's guards on the new memory > > accesses/array derefs. > > > > I've been seeing that bug too, but it's very intermittent. How did you > > get it to trigger reliably enough for a bisect? > > Look into my local bisection log. You are right, that the bug is intermittent > and takes a very long time to reproduce the issue. > > I didn't observe similar issues during following v6.12-rcx kernel > fuzzing. I will keep monitoring. yeah, this is one for Jens...