On 12/15/22 18:39, Joseph Salisbury wrote:
On 12/15/22 13:39, Joseph Salisbury wrote:
Hello,
A bug has been reported[0] against a linux-5.15.y based Ubuntu kernel
that has the 5.15.79-rt54 real-time patch set applied.
This bug is causing a system hang when running disk tests with the
stress-ng[1] test suite. A 'hung task' call trace is being
generated[2] just prior to the system hanging. It appears the hang
is happening during a call to schedule.
To see if this was a regression, I also tested prior real-time patch
sets, going as far back to rt38 and all versions exhibit this bug. I
will test back further to see if this is in fact a regression and a
kernel bisect can be done. I also confirmed the same kernel without
the rt54 patch set applied does not exhibit this bug.
I will continue to debug this issue, but I wanted to send this
message upstream in case anyone else runs into it. Any specific
tests, tracing or debugging steps would be greatly appreciated.
Otherwise, I'll report back to this thread with any findings.
Thanks,
Joe
[0] https://bugs.launchpad.net/ubuntu-realtime/+bug/1998536
[1] https://github.com/ColinIanKing/stress-ng
[2] https://launchpadlibrarian.net/637058129/dmesg_dl110_rt_kernel.txt
I should have also mentioned, I will also test the latest available
patch set (6.1-rc7-rt5) to see if it also exhibits the bug.
This bug/system hang does still happen with the v6.1-rc7 kernel and the
rt5 real-time patch. I also confirmed the bug/system hang does not
happen with just v6.1-rc7 without the real-time patch applied.
I will bisect through the rt patches to see if I can find the specific
patch(s) that causes the bug. I will use the v5.15 kernel and it's rt
patches, since v5.15 has the patches broken out individually, and they
can be applied one at a time with 'git am'.
I'll provide further updates during the bisect.
Thanks,
Joe
