This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH, testsuite] Skip two tests for MIPS
- From: Janis Johnson <janis187 at us dot ibm dot com>
- To: Richard Sandiford <rdsandiford at googlemail dot com>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Mon, 19 Jan 2009 16:24:53 -0800
- Subject: Re: [PATCH, testsuite] Skip two tests for MIPS
- References: <8763kgi75s.fsf@firetop.home>
- Reply-to: janis187 at us dot ibm dot com
On Thu, 2009-01-15 at 21:35 +0000, Richard Sandiford wrote:
> gcc.dg/tree-ssa/ssa-store-ccp-3.c fails on traditional MIPS GNU/Linux
> targets[*] because code is PIC by default. In other words, __PIC__
> isn't a reliable test of whether shared-library semantics are in
> effect. The same is true for IRIX.
>
> We already skip the test for Darwin and HPPA for the same reason.
> Is it OK to do the same for MIPS?
OK with me, particularly since you're a MIPS maintainer.
> gcc.dg/tree-ssa/vrp47.c fails for MIPS targets because we define
> LOGICAL_OP_NON_SHORT_CIRCUIT to 0. This prevents the usual conversion
> of branches to setcc operations, so the VRP optimisation doesn't trigger.
>
> The only other target to define LOGICAL_OP_NON_SHORT_CIRCUIT is PowerPC,
> and gcc-testresults has similar failures there. So is it OK to skip
> this test for MIPS? (I'll leave PowerPC to someone who can test it.)
OK.
I added a link to PR middle-end/38219, which covers the test failure for
powerpc*-*-*.
Janis