Hi! On Fri, Feb 07, 2020 at 06:22:56PM -0600, Peter Bergner wrote: > On 1/9/20 6:29 PM, Peter Bergner wrote: > > On 1/9/20 4:51 PM, Segher Boessenkool wrote: > >> Splitting out separate functions in the testcase shouldn't be so much > >> work? Or am I too optimistic :-) > >> > >> This should make the test a good deal less prone to random changes in > >> output caused by the lunar cycle. > > > > Ok, let me take a stab at rewriting the tests to be more similar to the > > pr92923-[12].c and see how much work that is. I do agree that it would > > be nice not having the insn counts be so fragile. > > Sorry for taking so long to get back to this. I split the functions into > smaller chunks, but didn't need to go all the way to one function per > builtin call. Does this look better?
> PR target/93136 > * gcc.dg/vmx/ops.c: Add -flax-vector-conversions to dg-options. Eww. Well, I guess we want to test that flag as well ;-) > * gcc.target/powerpc/vsx-vector-6.h: Split tests into smaller functions. > * gcc.target/powerpc/vsx-vector-6.p7.c: Adjust scan-assembler-times > regex directives. Adjust expected instruction counts. > * gcc.target/powerpc/vsx-vector-6.p8.c: Likewise. > * gcc.target/powerpc/vsx-vector-6.p9.c: Likewise. > --- a/gcc/testsuite/gcc.target/powerpc/vsx-vector-6.p7.c > +++ b/gcc/testsuite/gcc.target/powerpc/vsx-vector-6.p7.c > @@ -1,41 +1,43 @@ > -/* { dg-do compile { target { lp64 && be } } } */ > +/* { dg-do compile { target lp64 } } */ This worked for LE as well? Nice :-) That isn't supported of course, but there is no reason to explicitly not test it (or break it more than it is) :-) > +/* { dg-final { scan-assembler-times {\mvperm[r]?\M} 1 } } */ You can write this without the square brackets, fwiw. Okay for trunk. Thank you! Segher