GNU bug report logs - #71966
30.0.60; Intermittent failure in eglot-tests

Previous Next

To reply to this bug, email your comments to 71966 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: john muhl <jm <at> pub.pink>
To: bug-gnu-emacs <at> gnu.org
Subject: 30.0.60; Intermittent failure in eglot-tests
Date: Fri, 05 Jul 2024 22:42:33 -0500

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

The failing test is ‘eglot-test-auto-detect-running-server’. I
can’t reproduce it here by running the individual test in a loop
(well it didn’t fail during 5 hours of looping). However it has
failed 10 times in the last 62 runs of ‘make check’. It only
happens on the 32bit ARM test machine but happens across all
tested configurations there.

I was able to run ‘make check’ a dozen times without failure using
the patch below. Let’s see how it does for a week or two of builds.

  --- a/test/lisp/progmodes/eglot-tests.el
  +++ b/test/lisp/progmodes/eglot-tests.el
  @@ -326,7 +326,7 @@ eglot-tests--auto-detect-running-server-1
          ("anotherproject" . (("cena.c" . "bla"))))
        (with-current-buffer
            (eglot--find-file-noselect "project/coiso.c")
  -       (should (setq server (eglot--tests-connect)))
  +       (should (setq server (eglot--tests-connect 20)))
          (should (eglot-current-server)))
        (with-current-buffer
            (eglot--find-file-noselect "project/merdix.c")

[eglot-tests.log (text/plain, attachment)]

Message #8 received at 71966 <at> debbugs.gnu.org (full text, mbox):

From: Jeremy Bryant <jb <at> jeremybryant.net>
To: john muhl <jm <at> pub.pink>
Cc: 71966 <at> debbugs.gnu.org,
 João Távora <joaotavora <at> gmail.com>
Subject: Re: bug#71966: 30.0.60; Intermittent failure in eglot-tests
Date: Sat, 06 Jul 2024 23:32:37 +0100

john muhl <jm <at> pub.pink> writes:

> The failing test is ‘eglot-test-auto-detect-running-server’. I
> can’t reproduce it here by running the individual test in a loop
> (well it didn’t fail during 5 hours of looping). However it has
> failed 10 times in the last 62 runs of ‘make check’. It only
> happens on the 32bit ARM test machine but happens across all
> tested configurations there.
>
> I was able to run ‘make check’ a dozen times without failure using
> the patch below. Let’s see how it does for a week or two of builds.
>
>   --- a/test/lisp/progmodes/eglot-tests.el
>   +++ b/test/lisp/progmodes/eglot-tests.el
>   @@ -326,7 +326,7 @@ eglot-tests--auto-detect-running-server-1
>           ("anotherproject" . (("cena.c" . "bla"))))
>         (with-current-buffer
>             (eglot--find-file-noselect "project/coiso.c")
>   -       (should (setq server (eglot--tests-connect)))
>   +       (should (setq server (eglot--tests-connect 20)))
>           (should (eglot-current-server)))
>         (with-current-buffer
>             (eglot--find-file-noselect "project/merdix.c")
>
> [2. text/plain; eglot-tests.log]...

Adding João on this bug report

Message #11 received at 71966 <at> debbugs.gnu.org (full text, mbox):

From: João Távora <joaotavora <at> gmail.com>
To: Jeremy Bryant <jb <at> jeremybryant.net>
Cc: 71966 <at> debbugs.gnu.org, john muhl <jm <at> pub.pink>
Subject: Re: bug#71966: 30.0.60; Intermittent failure in eglot-tests
Date: Sun, 7 Jul 2024 00:30:11 +0100

Hmmm,
Seems like the clangd on that underpowered machine takes
longer than the default 10 seconds to respond to Eglot's
initiation.  I don't think it's worth bumping the timeouts for
such cases, but I don't mind it either.  Why are you running tests
on those ARM machines?

BTW clangd is usually very spiffy by comparison with other
 servers, so I can't imagine how the other tests will ever pass
(they're probably skipped).

João

On Sat, Jul 6, 2024 at 11:32 PM Jeremy Bryant <jb <at> jeremybryant.net> wrote:
>
> john muhl <jm <at> pub.pink> writes:
>
> > The failing test is ‘eglot-test-auto-detect-running-server’. I
> > can’t reproduce it here by running the individual test in a loop
> > (well it didn’t fail during 5 hours of looping). However it has
> > failed 10 times in the last 62 runs of ‘make check’. It only
> > happens on the 32bit ARM test machine but happens across all
> > tested configurations there.
> >
> > I was able to run ‘make check’ a dozen times without failure using
> > the patch below. Let’s see how it does for a week or two of builds.
> >
> >   --- a/test/lisp/progmodes/eglot-tests.el
> >   +++ b/test/lisp/progmodes/eglot-tests.el
> >   @@ -326,7 +326,7 @@ eglot-tests--auto-detect-running-server-1
> >           ("anotherproject" . (("cena.c" . "bla"))))
> >         (with-current-buffer
> >             (eglot--find-file-noselect "project/coiso.c")
> >   -       (should (setq server (eglot--tests-connect)))
> >   +       (should (setq server (eglot--tests-connect 20)))
> >           (should (eglot-current-server)))
> >         (with-current-buffer
> >             (eglot--find-file-noselect "project/merdix.c")
> >
> > [2. text/plain; eglot-tests.log]...
>
> Adding João on this bug report



-- 
João Távora

Previous Next

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