GNU bug report logs - #76188
erc-nicks-track-faces/prioritize failure on Fedora

Previous Next

Full log

View this message in rfc822 format

From: "J.P." <jp <at> neverwas.me>
To: Paul Eggert <eggert <at> cs.ucla.edu>
Cc: emacs-erc <at> gnu.org, 76188 <at> debbugs.gnu.org
Subject: bug#76188: erc-nicks-track-faces/prioritize failure on Fedora
Date: Mon, 10 Feb 2025 23:07:30 -0800

Hi Paul,

Paul Eggert <eggert <at> cs.ucla.edu> writes:

> On Fedora 41 with the current master (commit
> 4936a8d5acbfee2dee6d903400eba48cb2e3a6a7) I saw a failure with
> erc-nicks-track-faces/prioritize, as follows, when running "make -j5". I could
> not reproduce the problem when running just this test, so perhaps there's some
> conflict with some other test being run at about the same time. This is a
> machine with a slow CPU (AMD Phenom II X4 910e) so it's painful to run all the
> tests.

I believe I once saw a similar failure in that same family of tests but
dismissed it as a freak occurrence. This was likely with ERC's ELPA
package on Emacs 27 or 28 running in a cloud VM with nproc = 1 and
without any -j flag. If you're on Fedora 41, you likely have Make 4.4
installed, which doesn't suffer from the parallel -jN issue mentioned
here:

  https://lists.gnu.org/archive/html/bug-gnu-emacs/2022-04/msg00990.html

If that and my recollection are true, I think this test needs to be
rewritten, perhaps along with those like it. I've therefore tagged it as
:unstable for now.

Tests like this that are heavily mocked are difficult to debug without a
reliable recipe. So unless something glaring pops out that explains the
failure, I'll probably eventually rewrite it and related ones as higher
level "expensive" tests, which folks don't seem to like running at home
anyway.

Thanks,
J.P.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.