[PATCH] libstdc++: Ensure the test compiles without AVX512

[Andrea Pinski]

On Tue, Aug 18, 2026 at 1:06 AM Matthias Kretz <MatthiasKretz@gmx.net> wrote:
>
> OK for trunk and backports?
>
> ---------- 8< -----------
>
> From: Matthias Kretz <m.kretz@gsi.de>
>
> This test fails on my setup where I have a dejagnu board that sets
> -march=skylake-avx512. The intent of the test was to verify x86-64-v3
> optimizes as expected, therefore disable AVX512F unconditionally.
>
> libstdc++-v3/ChangeLog:
>
>         * testsuite/experimental/simd/pr114958.cc: Add -mno-avx512f to
>         dg-additional-options.
>
> Signed-off-by: Matthias Kretz <m.kretz@gsi.de>
> ---
>  libstdc++-v3/testsuite/experimental/simd/pr114958.cc | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/libstdc++-v3/testsuite/experimental/simd/pr114958.cc b/libstdc++-
> v3/testsuite/experimental/simd/pr114958.cc
> index 94c9e0a2d18..263a2d4bab3 100644
> --- a/libstdc++-v3/testsuite/experimental/simd/pr114958.cc
> +++ b/libstdc++-v3/testsuite/experimental/simd/pr114958.cc
> @@ -1,7 +1,7 @@
>  // { dg-options "-std=c++17" }
>  // { dg-do compile { target x86_64-*-* } }

This should be `{ target { x86_64-*-* i?86-*-* } && ! ia32 }`
Or does this not work for x32? Though I wonder if this works for -m32 too.

>  // { dg-require-effective-target c++17 }
> -// { dg-additional-options "-march=x86-64-v3" { target x86_64-*-* } }
> +// { dg-additional-options "-march=x86-64-v3 -mno-avx512f" { target x86_64-*-
> * } }
>  // { dg-require-cmath "" }
>  // { dg-final { scan-assembler-times "vperm(q|pd)\[\\t \]+\\\$144" 1 } }
>
> --
> ──────────────────────────────────────────────────────────────────────────
>  Dr. Matthias Kretz                           https://mattkretz.github.io
>  GSI Helmholtz Center for Heavy Ion Research               https://gsi.de
>  std::simd
> ──────────────────────────────────────────────────────────────────────────
>