[PATCH v2 1/5] x86 long double: Support pseudo numbers in fpclassifyl

[Adhemerval Zanella]

On 23/12/2020 08:55, Siddhesh Poyarekar wrote:
> Also move sysdeps/i386/fpu/s_fpclassifyl.c to
> sysdeps/x86/fpu/s_fpclassifyl.c and remove
> sysdeps/x86_64/fpu/s_fpclassifyl.c

LGTM, thanks.

Reviewed-by: Adhemerval Zanella  <adhemerval.zanella@linaro.org>

> ---
>  sysdeps/{i386 => x86}/fpu/s_fpclassifyl.c | 4 ++++
>  sysdeps/x86_64/fpu/s_fpclassifyl.c        | 2 --
>  2 files changed, 4 insertions(+), 2 deletions(-)
>  rename sysdeps/{i386 => x86}/fpu/s_fpclassifyl.c (87%)
>  delete mode 100644 sysdeps/x86_64/fpu/s_fpclassifyl.c
> 
> diff --git a/sysdeps/i386/fpu/s_fpclassifyl.c b/sysdeps/x86/fpu/s_fpclassifyl.c
> similarity index 87%
> rename from sysdeps/i386/fpu/s_fpclassifyl.c
> rename to sysdeps/x86/fpu/s_fpclassifyl.c
> index 501312f51e..a48b7685c4 100644
> --- a/sysdeps/i386/fpu/s_fpclassifyl.c
> +++ b/sysdeps/x86/fpu/s_fpclassifyl.c
> @@ -34,6 +34,10 @@ __fpclassifyl (long double x)
>      retval = FP_ZERO;
>    else if (ex == 0 && (hx & 0x80000000) == 0)
>      retval = FP_SUBNORMAL;
> +  /* Pseudo-normals, i.e. pseudo-zero, pseudo-infinity and un-normals.  They
> +     behave like NaNs, so categorize them as such.  */
> +  else if ((hx & 0x80000000) == 0)
> +    retval = FP_NAN;
>    else if (ex == 0x7fff)
>      retval = ((hx & 0x7fffffff) | lx) != 0 ? FP_NAN : FP_INFINITE;
>  
> diff --git a/sysdeps/x86_64/fpu/s_fpclassifyl.c b/sysdeps/x86_64/fpu/s_fpclassifyl.c
> deleted file mode 100644
> index 856854b0f5..0000000000
> --- a/sysdeps/x86_64/fpu/s_fpclassifyl.c
> +++ /dev/null
> @@ -1,2 +0,0 @@
> -#include <sysdeps/i386/fpu/s_fpclassifyl.c>
> -
>