[TLDR: This mail in primarily relevant for Linux kernel regression
tracking. See link in footer if these mails annoy you.]

On 30.01.23 12:57, Linux kernel regression tracking (#adding) wrote:

On 30.01.23 10:34, Harshit Mogalapalli wrote:

quoted

Since kernel 6.1-rc6 rasdaemon fails to update the summary of the records.

When we inject MCE errors, generally ras-mc-ctl --summary should be able
to read new errors, but starting from 6.1-rc6 the summary(count on
number of MCE records) doesnot udpate when we inject new mce errors.

This started happening after this commit
42fb0a1e84ff525ebe560e2baf9451ab69127e2b ("tracing/ring-buffer: Have
polling block on watermark") -- Commit landed first in 6.1-rc6, 6.1-rc5
kernel doesnot have this problem.

On reverting this commit, rasdaemon works good(i.e It is able to read
the new mce records).

This continues to happen on latest kernel(v6.2-rc6) as well.

Thanks for the report. To be sure the issue doesn't fall through the
cracks unnoticed, I'm adding it to regzbot, the Linux kernel regression
tracking bot:

#regzbot ^introduced 42fb0a1e84ff525
#regzbot title tracing/ring-buffer: rasdaemon does not report new records
#regzbot ignore-activity

For the record, I in
https://lore.kernel.org/lkml/1e759e44f5e64b4e99096afd9e89b6dc@huawei.com/ (local)
was told that the mainline commit 3e46d910d8ac ("tracing: Fix poll() and
select() do not work on per_cpu trace_pipe and trace_pipe_raw") fixes
this on the kernel side. There apparently is a rasdaemon change to
improve things, too.

#regzbot fix: 3e46d910d8ac

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)
--
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
That page also explains what to do if mails like this annoy you.

#regzbot ignore-activity