GNU bug report logs - #10848
automake-1.11.3 on OSX Lion with llvm-gcc - 1 of 829 tests failed

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 10848 in the body.
You can then email your comments to 10848 AT debbugs.gnu.org in the normal way.

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

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

From: Panther Martin <mrsmiley98 <at> lycos.com>
To: bug-automake <at> gnu.org
Subject: automake-1.11.3 on OSX Lion with llvm-gcc - 1 of 829 tests failed
Date: Mon, 20 Feb 2012 00:09:43 +0000 (GMT)

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

 
Hi I'm reporting 1 test-suite failure on 64bit OSX Lion 10.7.3 using

  llvm-gcc-4.2.1 (5658)
  autoconf-2.68
  automake-1.11.3
  gawk-4.0.0

I have attached the log.  Please let me know if I can provide more information.
Because Apple removed autoconf, automake, and libtool from their SDK
in February  2012, Automake will get built from source much more often.
Thank you for your help.  

[test-suite.log (application/octet-stream, attachment)]

Message #8 received at 10848 <at> debbugs.gnu.org (full text, mbox):

From: Stefano Lattarini <stefano.lattarini <at> gmail.com>
To: Panther Martin <mrsmiley98 <at> lycos.com>
Cc: 10848 <at> debbugs.gnu.org
Subject: Re: bug#10848: automake-1.11.3 on OSX Lion with llvm-gcc - 1 of 829
	tests failed
Date: Mon, 20 Feb 2012 14:24:03 +0100

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

Thanks for the report.

On 02/20/2012 01:09 AM, Panther Martin wrote:
>  
> Hi I'm reporting 1 test-suite failure on 64bit OSX Lion 10.7.3 using
> 
>   llvm-gcc-4.2.1 (5658)
>   autoconf-2.68
>   automake-1.11.3
>   gawk-4.0.0
> 
> FAIL: conffile-leading-dot.test (exit: 1)
> =========================================
>
> /private/tmp/homebrew-automake-1.11.3-vk9X/automake-1.11.3/tests:...
> conffile-leading-dot: running make --version -v | grep GNU
> GNU Make 3.81
> === Running test ./conffile-leading-dot.test
> ++ pwd
> /private/tmp/homebrew-automake-1.11.3-vk9X/automake-1.11.3/tests/conffile-leading-dot.dir
> + set -e
> + cat
> + echo foo = barbarbar
> + touch a.in b.in c.in
> + aclocal-1.11 -Werror
> + AUTOMAKE_fails -Wnone -Wunsupported
> + AUTOMAKE_run 1 -Wnone -Wunsupported
> + expected_exitcode=1
> + shift
> + exitcode=0
> + automake-1.11 --foreign -Werror -Wall -Wnone -Wunsupported
> + exitcode=1
> + cat stderr
> configure.in:3: omit leading './' from config file names such as './Makefile';
> configure.in:3: remake rules might be subtly broken otherwise
> configure.in:4: omit leading './' from config file names such as './foo';
> configure.in:4: remake rules might be subtly broken otherwise
> + cat stdout
> + test 1 = 1
> + grep '^configure\.in:3:.*'\''\./Makefile'\''' stderr
> configure.in:3: omit leading './' from config file names such as './Makefile';
> + grep '^configure\.in:3:.* omit leading '\''\./'\''' stderr
> configure.in:3: omit leading './' from config file names such as './Makefile';
> + grep '^configure\.in:3:.*remake rules might be subtly broken' stderr
> configure.in:3: remake rules might be subtly broken otherwise
> + grep '^configure\.in:4:.*'\''\./foo'\''' stderr
> configure.in:4: omit leading './' from config file names such as './foo';
> + grep '^configure\.in:4:.* omit leading '\''\./'\''' stderr
> configure.in:4: omit leading './' from config file names such as './foo';
> + grep '^configure\.in:4:.*remake rules might be subtly broken' stderr
> configure.in:4: remake rules might be subtly broken otherwise
> + autoconf -B /no/such/dir
> + automake-1.11 --foreign -Werror -Wall -Wall -Wno-unsupported
> + ./configure
> checking for a BSD-compatible install... /usr/bin/install -c
> checking whether build environment is sane... yes
> checking for a thread-safe mkdir -p... ./install-sh -c -d
> checking for gawk... gawk
> checking whether make sets $(MAKE)... yes
> configure: creating ./config.status
> config.status: creating ./Makefile
> config.status: creating ./foo
> + make
> make: Nothing to be done for `all'.
> + grep barbarbar Makefile
> foo = barbarbar
> + echo foo = bazbazbaz
> + make
>  cd . && /bin/sh /.../missing --run automake-1.11 --foreign ./Makefile
> make: Nothing to be done for `all'.
>
The failure is here.  I suspect this might be yet another timestamp issue.
Could you please answer the following questions?

  1. Is your system quite fast? (I'd assume yes, given the contents of
     config.log)

  2. Does the directory you are building Automake into have sub-second
     timestamp resolution?

  3. Does the reported failure happen consistently, or is it subject to
     a race?

If the answers to the questions above are respectively "yes", "no" and "yes
it's a racy failure", then I think the attached patch might fix the issue.

Regards,
  Stefano

[0001-tests-fix-a-timestamp-issue-and-other-minor-buglets.patch (text/x-diff, attachment)]

Message #11 received at 10848 <at> debbugs.gnu.org (full text, mbox):

From: Panther Martin <mrsmiley98 <at> lycos.com>
To: 10848 <at> debbugs.gnu.org
Subject: Re: bug#10848: automake-1.11.3 on OSX Lion with llvm-gcc - 1 of 829
	tests failed
Date: Mon, 20 Feb 2012 20:51:14 +0000 (GMT)

> On Feb 20, 2012, Stefano Lattarini <stefano.lattarini <at> gmail.com> wrote: 
  ...
 
> The failure is here.  I suspect this might be yet another timestamp issue.
> Could you please answer the following questions?
 

Yes I have a Core i5 8GB iMac.
No I there is no subsecond timestamp that I can see in touch or ls.
Yes it is repeatable 99% of the time.
Your patch has fixed the problem with make check and it passes
every test now with llvm-gcc-4.2.1 (5658) on Lion 10.7.3.  After
getting more deps for those tests, the final result was:

========================================================
All 898 tests behaved as expected (13 expected failures)
(40 tests were not run)
========================================================

Thanks.

Message #14 received at 10848 <at> debbugs.gnu.org (full text, mbox):

From: Stefano Lattarini <stefano.lattarini <at> gmail.com>
To: Panther Martin <mrsmiley98 <at> lycos.com>
Cc: 10848 <at> debbugs.gnu.org
Subject: Re: bug#10848: automake-1.11.3 on OSX Lion with llvm-gcc - 1 of 829
	tests failed
Date: Mon, 20 Feb 2012 23:06:14 +0100

severity 10848 minor
tags 10848 patch
close 10848
thanks

On 02/20/2012 09:51 PM, Panther Martin wrote:
> 
> 
>> On Feb 20, 2012, Stefano Lattarini <stefano.lattarini <at> gmail.com> wrote: 
>   ...
>  
>> The failure is here.  I suspect this might be yet another timestamp issue.
>> Could you please answer the following questions?
>  
> 
> Yes I have a Core i5 8GB iMac.
> No I there is no subsecond timestamp that I can see in touch or ls.
>
Neither with stat(1) I assume, right?

> Yes it is repeatable 99% of the time.
> Your patch has fixed the problem
>
Very good!  I'm closing this bug report then.

Thanks for your help,
  Stefano

Previous Next

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