[PATCH] math: Fix acosf when building with gcc <= 11

[Paul Zimmermann]

looks good to me.
Thanks,
Paul

> From: Adhemerval Zanella <adhemerval.zanella@linaro.org>
> Cc: Paul Zimmermann <Paul.Zimmermann@inria.fr>,
> 	Florian Weimer <fweimer@redhat.com>
> Date: Thu,  2 Jan 2025 11:34:21 -0300
> 
> GCC <= 11 wrongly assumes the rounding is to nearest and performs a
> constant folding where it should evaluate since the result is not
> exact [1].
> 
> [1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=57245
> ---
>  sysdeps/aarch64/libm-test-ulps                 |  2 --
>  sysdeps/i386/fpu/libm-test-ulps                |  2 --
>  sysdeps/i386/i686/fpu/multiarch/libm-test-ulps |  2 --
>  sysdeps/ieee754/flt-32/e_acosf.c               | 10 +++++++---
>  sysdeps/powerpc/fpu/libm-test-ulps             |  2 --
>  sysdeps/x86_64/fpu/libm-test-ulps              |  2 --
>  6 files changed, 7 insertions(+), 13 deletions(-)
> 
> diff --git a/sysdeps/aarch64/libm-test-ulps b/sysdeps/aarch64/libm-test-ulps
> index f1ab172f27..190685f893 100644
> --- a/sysdeps/aarch64/libm-test-ulps
> +++ b/sysdeps/aarch64/libm-test-ulps
> @@ -11,7 +11,6 @@ float: 1
>  
>  Function: "acos_downward":
>  double: 1
> -float: 1
>  ldouble: 1
>  
>  Function: "acos_sve":
> @@ -20,7 +19,6 @@ float: 1
>  
>  Function: "acos_towardzero":
>  double: 1
> -float: 1
>  ldouble: 1
>  
>  Function: "acos_upward":
> diff --git a/sysdeps/i386/fpu/libm-test-ulps b/sysdeps/i386/fpu/libm-test-ulps
> index 27f0f71b0a..423061e12e 100644
> --- a/sysdeps/i386/fpu/libm-test-ulps
> +++ b/sysdeps/i386/fpu/libm-test-ulps
> @@ -7,12 +7,10 @@ float128: 1
>  ldouble: 2
>  
>  Function: "acos_downward":
> -float: 1
>  float128: 1
>  ldouble: 2
>  
>  Function: "acos_towardzero":
> -float: 1
>  float128: 1
>  ldouble: 2
>  
> diff --git a/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps b/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
> index 66733c2d1d..466e68dae0 100644
> --- a/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
> +++ b/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
> @@ -7,12 +7,10 @@ float128: 1
>  ldouble: 2
>  
>  Function: "acos_downward":
> -float: 1
>  float128: 1
>  ldouble: 2
>  
>  Function: "acos_towardzero":
> -float: 1
>  float128: 1
>  ldouble: 2
>  
> diff --git a/sysdeps/ieee754/flt-32/e_acosf.c b/sysdeps/ieee754/flt-32/e_acosf.c
> index cba01221dc..a5a4de4fc2 100644
> --- a/sysdeps/ieee754/flt-32/e_acosf.c
> +++ b/sysdeps/ieee754/flt-32/e_acosf.c
> @@ -3,7 +3,7 @@
>  Copyright (c) 2023-2024 Alexei Sibidanov.
>  
>  The original version of this file was copied from the CORE-MATH
> -project (file src/binary32/acos/acosf.c, revision 61d7bef).
> +project (file src/binary32/acos/acosf.c, revision 56dd347).
>  
>  Permission is hereby granted, free of charge, to any person obtaining a copy
>  of this software and associated documentation files (the "Software"), to deal
> @@ -28,6 +28,7 @@ SOFTWARE.
>  #include <math.h>
>  #include <math_private.h>
>  #include <libm-alias-finite.h>
> +#include <math-barriers.h>
>  #include "math_config.h"
>  
>  static __attribute__ ((noinline)) float
> @@ -66,7 +67,7 @@ poly12 (double z, const double *c)
>  float
>  __ieee754_acosf (float x)
>  {
> -  const double pi2 = 0x1.921fb54442d18p+0;
> +  double pi2 = 0x1.921fb54442d18p+0;
>    static const double o[] = { 0, 0x1.921fb54442d18p+1 };
>    double xs = x;
>    double r;
> @@ -87,7 +88,10 @@ __ieee754_acosf (float x)
>  	};
>        /* Avoid spurious underflow exception.  */
>        if (__glibc_unlikely (ax <= 0x40000000u)) /* |x| < 2^-63 */
> -	return (float) pi2;
> +	/* GCC <= 11 wrongly assumes the rounding is to nearest and
> +	   performs a constant folding here:
> +	   https://gcc.gnu.org/bugzilla/show_bug.cgi?id=57245 */
> +	return math_opt_barrier (pi2);
>        double z = xs;
>        double z2 = z * z;
>        double z4 = z2 * z2;
> diff --git a/sysdeps/powerpc/fpu/libm-test-ulps b/sysdeps/powerpc/fpu/libm-test-ulps
> index 89dcef3603..07b91f8534 100644
> --- a/sysdeps/powerpc/fpu/libm-test-ulps
> +++ b/sysdeps/powerpc/fpu/libm-test-ulps
> @@ -8,13 +8,11 @@ ldouble: 1
>  
>  Function: "acos_downward":
>  double: 1
> -float: 1
>  float128: 1
>  ldouble: 3
>  
>  Function: "acos_towardzero":
>  double: 1
> -float: 1
>  float128: 1
>  ldouble: 3
>  
> diff --git a/sysdeps/x86_64/fpu/libm-test-ulps b/sysdeps/x86_64/fpu/libm-test-ulps
> index 624186f3f6..ad936ad32f 100644
> --- a/sysdeps/x86_64/fpu/libm-test-ulps
> +++ b/sysdeps/x86_64/fpu/libm-test-ulps
> @@ -8,13 +8,11 @@ ldouble: 2
>  
>  Function: "acos_downward":
>  double: 1
> -float: 1
>  float128: 1
>  ldouble: 2
>  
>  Function: "acos_towardzero":
>  double: 1
> -float: 1
>  float128: 1
>  ldouble: 2
>  
> -- 
> 2.43.0
> 
>