Re: Kernel panic related to pipe

[Robert Elz <kre%munnari.OZ.AU@localhost>]

    Date:        Mon, 5 Oct 2026 14:21:48 -0000 (UTC)
    From:        mlelstv%serpens.de@localhost (Michael van Elst)
    Message-ID:  <11a0bps$3ql$1%serpens.de@localhost>

  | A regular kernel on the same machine also doesn't panic. The CPU is
  | an old dual core with hyperthreading.

No, as Taylor's reply to my (1.178) commit indicated, there is clearly
something else going on that we're yet to spot.   Most likely it takes
quite a lot of pipe activity (such as running a build of something) might
produce (with -jN for some big enough N) in order to trigger the bug.

I'd suggest that we (all) stop using the CVS repo to try out ideas however,
and only commit changes (to anything) when we have verified that they behave
as intended, at least for our own systems - no more guesswork (let alone not
even compile tested) commits.    If some other user has a problem and can
help testing a fix for something we can't personally see - send patches, and
await verification before committing.

kre