This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Unreviewed^2 testsuite patches


Richard Guenther <richard.guenther@gmail.com> writes:

> On Fri, Jul 23, 2010 at 10:34 AM, Rainer Orth
> <ro@cebitec.uni-bielefeld.de> wrote:
>> The following two patches have remained unreviewed for some time despite
>> a reminder:
>>
>> ? ? ? ?[build, doc, libjava, testsuite] Fully support TLS on Solaris 8 and 9
>> ? ? ? ?http://gcc.gnu.org/ml/gcc-patches/2010-07/msg01277.html
>>
>> Most of this patch is Solaris-specific, but I may need approval for the
>> testsuite parts. ?One might argue that they are obvious, though.

Any comments on the testsuite parts here?  Would they count as obvious?

>> ? ? ? ?[testsuite] Increase gcc.dg/pr43058.c timeout
>> ? ? ? ?http://gcc.gnu.org/ml/gcc-patches/2010-07/msg01162.html
>>
>> I'm going to install this as obvious soon (on the 4.5 branch only after
>> it reopens, of course).
>
> It builds in less than 60s for me, even with checking enabled and
> a compiler built with -O0.  So - is your machine really so slow?

It depends: as I wrote, on an idle 2.93 GHz Xeon X7350, it takes 78s,
but the test still can time out when run within make -j32 check.  On the
other hand, on an idle 1.2 GHz UltraSPARC-T2 it takes 296s (very close
to the regular 300s timeout), so it easily goes beyond.  Sure, the
machine is slow, but not unreasonably so.

> Increasing the timeout will make us fail notice compile-time regressions
> here.

Indeed: that's why I asked for approval rather than simply checking in
the patch as obvious.

> So the patch isn't obvious.

Ok: how do we proceed from here?

>> This is really getting annoying: work being blocked because nobody cares
>> to review the testsuite parts.
>
> Sorry - we're lacking a testsuite maintainer.

I though Janis' assignment had come through by now (I recall a message
to that effect).  One thing I wondered, btw.: even without an
assignment, couldn't she review and approve other's patches, just not
check in her own ones?  (I know, a rather absurd sitation, but better
than the current one).

Thanks.
	Rainer

-- 
-----------------------------------------------------------------------------
Rainer Orth, Center for Biotechnology, Bielefeld University


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]