On Wed, Feb 4, 2026 at 5:51 PM Alexandre Oliva <[email protected]> wrote:
>
>
> The execution testcase requires sve2 and 128-bit sve hardware, but it
> doesn't state those requiremnets. I'm think the latter is implied by
> the former, but I'm not entirely sure, so I'm requiring both
> explicitly.
>
> I've tested this with aarch64-elf on trunk and gcc-15.
> Ok for trunk? Ok for gcc-15?
Ok for both.
Note gcc.target/aarch64/sve/pr122103_1.c has the same issue (that is
recorded as PR 123936 which is assigned to me but you can take it
over; the patch for that testcase is pre-approved).
>
> PS: It looks like we might need to check for armv9-a hw as well, but...
> we don't seem to have an effective target to use to that end, do we?
I don't think we really need armv9-a check here since armv9-a is
really an alias for armv8.5-a+sve2 at this stage. So checking for sve2
is enough.
Thanks,
Andrew
>
>
> for gcc/testsuite/ChangeLog
>
> PR middle-end/123775
> * gcc.target/aarch64/sve2/pr123775.c: Add sve128 and sve2 hw
> requirements.
> ---
> gcc/testsuite/gcc.target/aarch64/sve2/pr123775.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/gcc/testsuite/gcc.target/aarch64/sve2/pr123775.c
> b/gcc/testsuite/gcc.target/aarch64/sve2/pr123775.c
> index ba292045a1857..001c3e2cf71bc 100644
> --- a/gcc/testsuite/gcc.target/aarch64/sve2/pr123775.c
> +++ b/gcc/testsuite/gcc.target/aarch64/sve2/pr123775.c
> @@ -1,4 +1,4 @@
> -/* { dg-do run } */
> +/* { dg-do run { target { aarch64_sve128_hw && aarch64_sve2_hw } } } */
> /* { dg-additional-options "-O3 -march=armv9-a+sve2 -msve-vector-bits=128
> --param aarch64-autovec-preference=sve-only" } */
>
> int main(int argc, char *argv[]) {
>
>
> --
> Alexandre Oliva, happy hacker https://blog.lx.oliva.nom.br/
> Free Software Activist FSFLA co-founder GNU Toolchain Engineer
> More tolerance and less prejudice are key for inclusion and diversity.
> Excluding neuro-others for not behaving ""normal"" is *not* inclusive!
