GNU bug report logs - #80166
process-test-stderr-buffer failure in Emacs master 2026-01-09

Previous Next

To reply to this bug, email your comments to 80166 AT debbugs.gnu.org.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Paul Eggert <eggert <at> cs.ucla.edu>
To: Emacs bug reports and feature requests <bug-gnu-emacs <at> gnu.org>
Subject: process-test-stderr-buffer failure in Emacs master 2026-01-09
Date: Fri, 9 Jan 2026 12:51:17 -0800

[Message part 1 (text/plain, inline)]

This is Emacs master commit 65090ec691d9bd82a9ba3bd25756227e0d43bd80 
dated today, running on Fedora 42 x86-64 on an old, slow CPU (an AMD 
Phenom II X4 910e). The failure symptom with "make -j5 check" is in 
test/src/process-tests.log, a copy of which I am attaching.

[process-tests.log (text/x-log, attachment)]

Message #8 received at 80166 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Paul Eggert <eggert <at> cs.ucla.edu>
Cc: 80166 <at> debbugs.gnu.org
Subject: Re: bug#80166: process-test-stderr-buffer failure in Emacs master
 2026-01-09
Date: Sat, 10 Jan 2026 12:00:35 +0200

> Date: Fri, 9 Jan 2026 12:51:17 -0800
> From: Paul Eggert <eggert <at> cs.ucla.edu>
> 
> This is Emacs master commit 65090ec691d9bd82a9ba3bd25756227e0d43bd80 
> dated today, running on Fedora 42 x86-64 on an old, slow CPU (an AMD 
> Phenom II X4 910e). The failure symptom with "make -j5 check" is in 
> test/src/process-tests.log, a copy of which I am attaching.

Any idea why it fails?  What is actually at point-min when it fails?

Also, did you try enlarging the value of
process-test-sentinel-wait-timeout, and if so, which larger values did
you try?

That CPU has 4 cores, AFAICT, so maybe running "make -j5" incurs
additional delays.  Does -j3 or -j4 (or without -jN at all) succeed?
What happens if you try something like -j8 or -j10?

Message #11 received at 80166 <at> debbugs.gnu.org (full text, mbox):

From: Paul Eggert <eggert <at> cs.ucla.edu>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: 80166 <at> debbugs.gnu.org
Subject: Re: bug#80166: process-test-stderr-buffer failure in Emacs master
 2026-01-09
Date: Sat, 10 Jan 2026 13:25:27 -0800

[Message part 1 (text/plain, inline)]

On 2026-01-10 02:00, Eli Zaretskii wrote:
>> Date: Fri, 9 Jan 2026 12:51:17 -0800
>> From: Paul Eggert <eggert <at> cs.ucla.edu>
>>
>> This is Emacs master commit 65090ec691d9bd82a9ba3bd25756227e0d43bd80
>> dated today, running on Fedora 42 x86-64 on an old, slow CPU (an AMD
>> Phenom II X4 910e). The failure symptom with "make -j5 check" is in
>> test/src/process-tests.log, a copy of which I am attaching.
> 
> Any idea why it fails?  What is actually at point-min when it fails?

I expect it failed because the system is loaded. How would I determine 
what is actually at point-min? (But also see below.)


> Also, did you try enlarging the value of
> process-test-sentinel-wait-timeout, and if so, which larger values did
> you try?

I did not try that.


> That CPU has 4 cores, AFAICT, so maybe running "make -j5" incurs
> additional delays.  Does -j3 or -j4 (or without -jN at all) succeed?
> What happens if you try something like -j8 or -j10?

Yes, it has 4 cores.

I tried running make -j5 check several times. Sometimes it works, 
sometimes it fails. Once lisp/eshell/esh-proc-tests also failed, with 
the attached symptoms.

I tried with -j8 and with -j10. The test worked in both cases. However, 
the -j10 build caused proced-tests to fail with the attached symptoms.

By the way, I am in Emacs M-x shell.

make -j3 check and make -j4 check seem to work, though I would not be 
surprised if this were intermittent.

[esh-proc-tests.txt (text/plain, attachment)]

[proced-tests.log (text/x-log, attachment)]

Message #14 received at 80166 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Paul Eggert <eggert <at> cs.ucla.edu>
Cc: 80166 <at> debbugs.gnu.org
Subject: Re: bug#80166: process-test-stderr-buffer failure in Emacs master
 2026-01-09
Date: Sun, 11 Jan 2026 08:47:36 +0200

> Date: Sat, 10 Jan 2026 13:25:27 -0800
> From: Paul Eggert <eggert <at> cs.ucla.edu>
> Cc: 80166 <at> debbugs.gnu.org
> 
> > Any idea why it fails?  What is actually at point-min when it fails?
> 
> I expect it failed because the system is loaded. How would I determine 
> what is actually at point-min?

Edit the test to add something like

  (message (buffer-string))

just before the call to looking-at, and then see what it prints while
running the test when the test fails.

Or add something like

  (should (string-equal (buffer-substring
                         (point-min) (+ (point-min) 13))
			"hello stdout!"))

before the call to looking-at, and see in the log what was actually
the buffer substring there.

> > Also, did you try enlarging the value of
> > process-test-sentinel-wait-timeout, and if so, which larger values did
> > you try?
> 
> I did not try that.

Please try that.  If the problem is a loaded system with too few
execution cores, that could be the problem.  The current value of the
timeout is just some heuristic.

> I tried running make -j5 check several times. Sometimes it works, 
> sometimes it fails. Once lisp/eshell/esh-proc-tests also failed, with 
> the attached symptoms.
> 
> I tried with -j8 and with -j10. The test worked in both cases. However, 
> the -j10 build caused proced-tests to fail with the attached symptoms.

I think those should be separate bug reports.

Previous Next

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