[patch] XFAIL two testcases in gcc testsuite.
Janis Johnson
janis187@us.ibm.com
Tue May 22 20:14:00 GMT 2007
On Tue, May 22, 2007 at 12:23:10PM -0700, Kazu Hirata wrote:
> Hi,
>
> Attached is a patch to XFAIL two testcases in gcc testsuite.
>
> Currently, gcc.c-torture/execute/ieee/20000320-1.c and
> gcc.c-torture/execute/ieee/mul-subnormal-single-1.c are XFAILed when
> they are run on m68k-linux-gnu, and we know that we are testing on a
> ColdFire with an FPU.
>
> It turns out that these testcases also fail on m68k-elf (i.e. bare
> metal) if they are run on a ColdFire with an FPU. That makes sense
> because these testcases check that the denormal numbers are correctly
> generated, but ColdFire FPUs do not generate denormal numbers. (They
> do give you an option of generating denormal numbers by software via
> an interrupt hook.)
>
> This patch widens the XFAILing a little bit. Specifically, with this
> patch, the two testcases are XFAILed on any m68k-*-* as long as they
> are run on a ColdFire with an FPU.
>
> Tested individual testcases on m68k-elf. OK to apply?
Do you really want them to XFAIL, or do you want to skip them? To skip
them return 1 if the condition is true.
OK either way.
Janis
More information about the Gcc-patches
mailing list