[PATCH] x86: Disable AVX Fast Unaligned Load on Hygon 1/2/3 due to performance regression

[H.J. Lu]

On Mon, Oct 20, 2025 at 5:57 PM litenglong <litenglong@kylinos.cn> wrote:
>
> - Performance testing revealed significant memcpy performance degradation
>   when bit_arch_AVX_Fast_Unaligned_Load is enabled on Hygon 3.
> - Hygon confirmed AVX performance issues in certain memory functions.
> - Glibc benchmarks show SSE outperforms AVX for
>   memcpy/memmove/memset/strcmp/strcpy/strlen and so on.
> - Hardware differences primarily in floating-point operations don't justify
>   AVX usage for memory operations.
>
> Reviewed-by: gaoxiang <gaoxiang@kylinos.cn>
> Signed-off-by: litenglong <litenglong@kylinos.cn>
> ---
>  sysdeps/x86/cpu-features.c | 5 +++++
>  1 file changed, 5 insertions(+)
>
> diff --git a/sysdeps/x86/cpu-features.c b/sysdeps/x86/cpu-features.c
> index b67ef541dd..286cbfb1e2 100644
> --- a/sysdeps/x86/cpu-features.c
> +++ b/sysdeps/x86/cpu-features.c
> @@ -1123,6 +1123,11 @@ disable_tsx:
>         hardware.  */
>        cpu_features->preferred[index_arch_Avoid_Non_Temporal_Memset]
>             &= ~bit_arch_Avoid_Non_Temporal_Memset;
> +      if (model < 0x4) {
> +        /*  Unaligned AVX loads are slower.  */
> +        cpu_features->preferred[index_arch_AVX_Fast_Unaligned_Load]
> +                 &= ~bit_arch_AVX_Fast_Unaligned_Load;
> +      }
>      }
>    else
>      {
> --
> 2.25.1
>

The commit subject is too long.  Just use

x86: Disable AVX Fast Unaligned Load on Hygon 1/2/3

-- 
H.J.