On Tue, 20 Jan 2026, Alfie Richards wrote:

> On 20/01/2026 14:42, Alex Coplan wrote:
> > Hi both,
> > 
> > On 20/01/2026 12:59, Alfie Richards wrote:
> >> Hi Wilco and all,
> >>
> >> I agree with your feedback. I have moved it to the sve subdirectory and
> >> added
> >> the  ``/* { dg-do compile } */` for clarity.
> >>
> >> Okay for master?
> >> Alfie
> >>
> >> -- >8 --
> >>
> >>  PR middle-end/123697
> >>
> >> gcc/testsuite/ChangeLog:
> >>
> >>    * gcc.target/aarch64/sve/pr123697.c: New test.
> >> ---
> >>   gcc/testsuite/gcc.target/aarch64/sve/pr123697.c | 13 +++++++++++++
> >>   1 file changed, 13 insertions(+)
> >>   create mode 100644 gcc/testsuite/gcc.target/aarch64/sve/pr123697.c
> >>
> >> diff --git a/gcc/testsuite/gcc.target/aarch64/sve/pr123697.c
> >> b/gcc/testsuite/gcc.target/aarch64/sve/pr123697.c
> >> new file mode 100644
> >> index 00000000000..e4508512fc8
> >> --- /dev/null
> >> +++ b/gcc/testsuite/gcc.target/aarch64/sve/pr123697.c
> >> @@ -0,0 +1,13 @@
> >> +/* { dg-do compile } */
> >> +/* { dg-options "-msve-vector-bits=128 -mmax-vectorization -O3" } */
> > 
> > Is there a reason it isn't enough to add a line like the following:
> > 
> > /* { dg-additional-options "-march=armv9-a -msve-vector-bits=128
> > -mmax-vectorization" { target aarch64-*-* } } */
> > 
> > to the test that richi added, i.e.
> > gcc/testsuite/gcc.dg/vect/vect-pr123697.c?
> 
> Ah yeah that works too, hadn't occurred to me. I'm happy to do that instead.

I didn't do that because people tend to run vect.exp with all sorts
of flags ontop, and this would remove the ability of that to have
any effect.

Richard.

> Alfie
> 
> > 
> > Thanks,
> > Alex
> > 
> >> +
> >> +struct a {
> >> +  int c[2];
> >> +};
> >> +struct a d[3];
> >> +double f;
> >> +void g()
> >> +{
> >> +  for (int e = 0; e < 3; ++e)
> >> +    f += d[e].c[1];
> >> +}
> >> -- 
> >> 2.34.1
> >>
> 
> 
> 

-- 
Richard Biener <[email protected]>
SUSE Software Solutions Germany GmbH,
Frankenstrasse 146, 90461 Nuernberg, Germany;
GF: Jochen Jaser, Andrew McDonald, Werner Knoblich; (HRB 36809, AG Nuernberg)

Reply via email to