This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: [build, testsuite, v3] Increase gcc, g++, gfortran and libstdc++-v3 testsuite parallelism
- From: Benjamin Kosnik <bkoz at redhat dot com>
- To: Rainer Orth <ro at CeBiTec dot Uni-Bielefeld dot DE>
- Cc: gcc-patches at gcc dot gnu dot org, libstdc++ at gcc dot gnu dot org, Janis Johnson <janis dot marie dot johnson at gmail dot com>, Paolo Bonzini <bonzini at gnu dot org>, Jakub Jelinek <jakub at redhat dot com>
- Date: Wed, 21 Jul 2010 12:36:02 -0700
- Subject: Re: [build, testsuite, v3] Increase gcc, g++, gfortran and libstdc++-v3 testsuite parallelism
- References: <yddhbjt40ty.fsf@manam.CeBiTec.Uni-Bielefeld.DE>
On Wed, 21 Jul 2010 11:54:17 +0200
Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE> wrote:
> There's one partially related change included: even before this
> change, there could be a strange double summary line in
> mail-report.log:
>
> === libstdc++ Summary ===
>
> # of expected passes 13934
> # of unexpected failures 8
> # of unexpected successes 2
> # of expected failures 118
> # of unsupported tests 732
>
> === libstdc++ Summary ===
>
> I could trace this to the fact that currently no libstdc++ ABI
> baselines exist for Solaris 2 (being worked on), thus runtest abi.exp
> produces
>
> [...]
> === libstdc++ Summary for unix ===
>
> Running target unix/-m64
> Running /vol/gcc/src/hg/trunk/local/libstdc++-v3/testsuite/libstdc++-abi/abi.exp ...
>
> === libstdc++ Summary for unix/-m64 ===
>
>
> === libstdc++ Summary ===
>
>
> This situation is likely to occur more often in the future when
> runtest is run with a .exp file that doesn't match any tests on a
> given platform (like i386.exp on SPARC). The fix is trivial: only
> emit the summary line if any of the test counts is non-zero.
Sounds great to me. Feel free to check this part in as-is.
-benjamin