GNU bug report logs - #66176
9.4 root test fails: tests/rm/ext3-perf tests/rm/ext3-perf.sh

Previous Next

To reply to this bug, email your comments to 66176 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: Jeffrey Cliff <jeffrey.cliff <at> gmail.com>
To: bug-coreutils <at> gnu.org
Subject: 9.4 root test fails: tests/rm/ext3-perf tests/rm/ext3-perf.sh
Date: Sat, 23 Sep 2023 10:35:29 -0600

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

Test suite told me to send this report in so here we are.

environment : custom LFS (GNM)
coreutils: 9.4
gcc: (GCC) 13.2.0

jeff cilff
--
------------------------------------------------------------------------------------------------
End the campaign to Cancel Richard Stallman - go to stallmansupport.org !
------------------------------------------------------------------------------------------------

[coreutils-test-suite.log (text/x-log, attachment)]

[coreutils-config.log (text/x-log, attachment)]

Message #8 received at 66176 <at> debbugs.gnu.org (full text, mbox):

From: Bernhard Voelker <mail <at> bernhard-voelker.de>
To: Jeffrey Cliff <jeffrey.cliff <at> gmail.com>, 66176 <at> debbugs.gnu.org
Subject: Re: bug#66176: 9.4 root test fails: tests/rm/ext3-perf
 tests/rm/ext3-perf.sh
Date: Sun, 24 Sep 2023 17:31:55 +0200

On 9/23/23 18:35, Jeffrey Cliff wrote:
> Test suite told me to send this report in so here we are.
> 
> environment : custom LFS (GNM)
> coreutils: 9.4
> gcc: (GCC) 13.2.0


> FAIL: tests/rm/ext3-perf
> ========================
> [...]
>
> ++ date +%s
> + start=1698065263
> + mkdir d
> + cd d
> + xargs touch
> + seq 400000
> + test -f 1
> + test -f 400000
> + cd ..
> + ok=1
> + test 1 = 1
> +++ date +%s
> ++ expr 1698065541 - 1698065263
> + setup_duration=278
> + echo creating a 400000-entry directory took 278 seconds
> creating a 400000-entry directory took 278 seconds
> + test 60 -lt 278
> + threshold_seconds=278
> ++ date +%s
> + start=1698065541
> + timeout 278s rm -rf d
> + err=124
> +++ date +%s
> ++ expr 1698065819 - 1698065541
> + duration=278
> + case $err in
> + fail=1
> + echo rm took longer than 278 seconds
> rm took longer than 278 seconds
> + echo removing a 400000-entry directory took 278 seconds
> removing a 400000-entry directory took 278 seconds
> + Exit 1

Hmm, that seems to be a rather slow system.
On my desktop system on ext4 here, it is:
  creating a 400000-entry directory took 6 seconds
  removing a 400000-entry directory took 4 seconds

The test measures the time for the creation, and uses it to verify
that the removal is faster.

Depending on other activities on the system during creation vs. that during
the deletion, the test might fail.

Is it always reproducible on that system, or was this just a one-off?

Have a nice day,
Berny

Message #11 received at 66176 <at> debbugs.gnu.org (full text, mbox):

From: Jeffrey Cliff <jeffrey.cliff <at> gmail.com>
To: Bernhard Voelker <mail <at> bernhard-voelker.de>, 66176 <at> debbugs.gnu.org
Subject: Re: bug#66176: 9.4 root test fails: tests/rm/ext3-perf
 tests/rm/ext3-perf.sh
Date: Mon, 25 Sep 2023 21:38:27 -0600

whoops didn't notice the debbugs email
reproduced 7 times now, same results.  usually 260-280 seconds.

On Mon, Sep 25, 2023 at 4:43 PM Jeffrey Cliff <jeffrey.cliff <at> gmail.com> wrote:
>
> Run the full test suite now at least 6 times, same results.  Would you
> like me to run them more?
>
> last couple of tests/rm/ext3-perf results ...266 seconds , 265
> seconds, 260 seconds ...
>
> On Sun, Sep 24, 2023 at 9:31 AM Bernhard Voelker
> <mail <at> bernhard-voelker.de> wrote:
> >
> > On 9/23/23 18:35, Jeffrey Cliff wrote:
> > > Test suite told me to send this report in so here we are.
> > >
> > > environment : custom LFS (GNM)
> > > coreutils: 9.4
> > > gcc: (GCC) 13.2.0
> >
> >
> >  > FAIL: tests/rm/ext3-perf
> >  > ========================
> >  > [...]
> >  >
> >  > ++ date +%s
> >  > + start=1698065263
> >  > + mkdir d
> >  > + cd d
> >  > + xargs touch
> >  > + seq 400000
> >  > + test -f 1
> >  > + test -f 400000
> >  > + cd ..
> >  > + ok=1
> >  > + test 1 = 1
> >  > +++ date +%s
> >  > ++ expr 1698065541 - 1698065263
> >  > + setup_duration=278
> >  > + echo creating a 400000-entry directory took 278 seconds
> >  > creating a 400000-entry directory took 278 seconds
> >  > + test 60 -lt 278
> >  > + threshold_seconds=278
> >  > ++ date +%s
> >  > + start=1698065541
> >  > + timeout 278s rm -rf d
> >  > + err=124
> >  > +++ date +%s
> >  > ++ expr 1698065819 - 1698065541
> >  > + duration=278
> >  > + case $err in
> >  > + fail=1
> >  > + echo rm took longer than 278 seconds
> >  > rm took longer than 278 seconds
> >  > + echo removing a 400000-entry directory took 278 seconds
> >  > removing a 400000-entry directory took 278 seconds
> >  > + Exit 1
> >
> > Hmm, that seems to be a rather slow system.
> > On my desktop system on ext4 here, it is:
> >    creating a 400000-entry directory took 6 seconds
> >    removing a 400000-entry directory took 4 seconds
> >
> > The test measures the time for the creation, and uses it to verify
> > that the removal is faster.
> >
> > Depending on other activities on the system during creation vs. that during
> > the deletion, the test might fail.
> >
> > Is it always reproducible on that system, or was this just a one-off?
> >
> > Have a nice day,
> > Berny
>
>
>
> --
> ------------------------------------------------------------------------------------------------
> End the campaign to Cancel Richard Stallman - go to stallmansupport.org !
> ------------------------------------------------------------------------------------------------



-- 
------------------------------------------------------------------------------------------------
End the campaign to Cancel Richard Stallman - go to stallmansupport.org !
------------------------------------------------------------------------------------------------

Previous Next

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