All of lore.kernel.org
 help / color / mirror / Atom feed
From: Adhemerval Zanella <adhemerval.zanella@linaro.org>
To: Vineet Gupta <Vineet.Gupta1@synopsys.com>, libc-alpha@sourceware.org
Cc: linux-snps-arc@lists.infradead.org
Subject: Re: [PATCH 5/5] powerpc/fpu: use generic fma functions
Date: Mon, 1 Jun 2020 14:10:44 -0300	[thread overview]
Message-ID: <8eb04259-7529-cbb4-81b8-d5b0f56eb6b8@linaro.org> (raw)
In-Reply-To: <20200530020047.5490-6-vgupta@synopsys.com>



On 29/05/2020 23:00, Vineet Gupta wrote:

LGTM, thanks.

This new addition, along with the aarch64 one, makes me wonder if
it would be better to decompose the USE_* defined in multiple files
so the architecture adds only the required ones (instead of copy/paste
the whole file with the generic ones).

Also for powerpc, maybe a future cleanup would to move the logic
to select the builtin to math-use-builtins.h. Something like:

  #ifdef _ARCH_PWR5X
  #  define USE_FLOOR_BUILTIN 1
  #  define USE_FLOORF_BUILTIN 1
  #else
  #  define USE_FLOOR_BUILTIN 0
  #  define USE_FLOORF_BUILTIN 0
  #endif
  #define USE_FLOORL_BUILTIN 0
  #define USE_FLOORF128_BUILTIN 0

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

> ---
>  sysdeps/powerpc/fpu/math-use-builtins.h | 70 +++++++++++++++++++++++++
>  sysdeps/powerpc/fpu/s_fma.c             | 27 ----------
>  sysdeps/powerpc/fpu/s_fmaf.c            | 27 ----------
>  3 files changed, 70 insertions(+), 54 deletions(-)
>  create mode 100644 sysdeps/powerpc/fpu/math-use-builtins.h
>  delete mode 100644 sysdeps/powerpc/fpu/s_fma.c
>  delete mode 100644 sysdeps/powerpc/fpu/s_fmaf.c
> 
> diff --git a/sysdeps/powerpc/fpu/math-use-builtins.h b/sysdeps/powerpc/fpu/math-use-builtins.h
> new file mode 100644
> index 000000000000..9bdde66cf1de
> --- /dev/null
> +++ b/sysdeps/powerpc/fpu/math-use-builtins.h
> @@ -0,0 +1,70 @@
> +/* Using math gcc builtins instead of generic implementation.  PowerPC version.
> +   Copyright (C) 2019-2020 Free Software Foundation, Inc.
> +   This file is part of the GNU C Library.
> +
> +   The GNU C Library is free software; you can redistribute it and/or
> +   modify it under the terms of the GNU Lesser General Public
> +   License as published by the Free Software Foundation; either
> +   version 2.1 of the License, or (at your option) any later version.
> +
> +   The GNU C Library is distributed in the hope that it will be useful,
> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> +   Lesser General Public License for more details.
> +
> +   You should have received a copy of the GNU Lesser General Public
> +   License along with the GNU C Library; if not, see
> +   <https://www.gnu.org/licenses/>.  */
> +
> +#ifndef MATH_USE_BUILTINS_H
> +#define MATH_USE_BUILTINS_H	1
> +
> +#include <features.h> /* For __GNUC_PREREQ.  */
> +
> +/* Define these macros to 1 to use __builtin_xyz instead of the
> +   generic implementation.  */
> +#define USE_NEARBYINT_BUILTIN 0
> +#define USE_NEARBYINTF_BUILTIN 0
> +#define USE_NEARBYINTL_BUILTIN 0
> +#define USE_NEARBYINTF128_BUILTIN 0
> +
> +#define USE_RINT_BUILTIN 0
> +#define USE_RINTF_BUILTIN 0
> +#define USE_RINTL_BUILTIN 0
> +#define USE_RINTF128_BUILTIN 0
> +
> +#define USE_FLOOR_BUILTIN 0
> +#define USE_FLOORF_BUILTIN 0
> +#define USE_FLOORL_BUILTIN 0
> +#define USE_FLOORF128_BUILTIN 0
> +
> +#define USE_CEIL_BUILTIN 0
> +#define USE_CEILF_BUILTIN 0
> +#define USE_CEILL_BUILTIN 0
> +#define USE_CEILF128_BUILTIN 0
> +
> +#define USE_TRUNC_BUILTIN 0
> +#define USE_TRUNCF_BUILTIN 0
> +#define USE_TRUNCL_BUILTIN 0
> +#define USE_TRUNCF128_BUILTIN 0
> +
> +#define USE_ROUND_BUILTIN 0
> +#define USE_ROUNDF_BUILTIN 0
> +#define USE_ROUNDL_BUILTIN 0
> +#define USE_ROUNDF128_BUILTIN 0
> +
> +#define USE_COPYSIGNL_BUILTIN 1
> +#if __GNUC_PREREQ (7, 0)
> +# define USE_COPYSIGNF128_BUILTIN 1
> +#else
> +# define USE_COPYSIGNF128_BUILTIN 0
> +#endif
> +
> +#define USE_SQRT_BUILTIN 0
> +#define USE_SQRTF_BUILTIN 0
> +
> +#define USE_FMA_BUILTIN 1
> +#define USE_FMAF_BUILTIN 1
> +#define USE_FMAL_BUILTIN 0
> +
> +#endif /* math-use-builtins.h */

Ok.

> diff --git a/sysdeps/powerpc/fpu/s_fma.c b/sysdeps/powerpc/fpu/s_fma.c
> deleted file mode 100644
> index 9ddd13253485..000000000000
> --- a/sysdeps/powerpc/fpu/s_fma.c
> +++ /dev/null
> @@ -1,27 +0,0 @@
> -/* Compute x * y + z as ternary operation.  PowerPC version.
> -   Copyright (C) 2019-2020 Free Software Foundation, Inc.
> -   This file is part of the GNU C Library.
> -
> -   The GNU C Library is free software; you can redistribute it and/or
> -   modify it under the terms of the GNU Lesser General Public
> -   License as published by the Free Software Foundation; either
> -   version 2.1 of the License, or (at your option) any later version.
> -
> -   The GNU C Library is distributed in the hope that it will be useful,
> -   but WITHOUT ANY WARRANTY; without even the implied warranty of
> -   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> -   Lesser General Public License for more details.
> -
> -   You should have received a copy of the GNU Lesser General Public
> -   License along with the GNU C Library; if not, see
> -   <https://www.gnu.org/licenses/>.  */
> -
> -#include <sysdep.h>
> -#include <libm-alias-double.h>
> -
> -double
> -__fma (double x, double y, double z)
> -{
> -  return __builtin_fma (x, y, z);
> -}
> -libm_alias_double (__fma, fma)

Ok.

> diff --git a/sysdeps/powerpc/fpu/s_fmaf.c b/sysdeps/powerpc/fpu/s_fmaf.c
> deleted file mode 100644
> index d132c20125bb..000000000000
> --- a/sysdeps/powerpc/fpu/s_fmaf.c
> +++ /dev/null
> @@ -1,27 +0,0 @@
> -/* Compute x * y + z as ternary operation.  PowerPC version.
> -   Copyright (C) 2019-2020 Free Software Foundation, Inc.
> -   This file is part of the GNU C Library.
> -
> -   The GNU C Library is free software; you can redistribute it and/or
> -   modify it under the terms of the GNU Lesser General Public
> -   License as published by the Free Software Foundation; either
> -   version 2.1 of the License, or (at your option) any later version.
> -
> -   The GNU C Library is distributed in the hope that it will be useful,
> -   but WITHOUT ANY WARRANTY; without even the implied warranty of
> -   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> -   Lesser General Public License for more details.
> -
> -   You should have received a copy of the GNU Lesser General Public
> -   License along with the GNU C Library; if not, see
> -   <https://www.gnu.org/licenses/>.  */
> -
> -#include <sysdep.h>
> -#include <libm-alias-float.h>
> -
> -float
> -__fmaf (float x, float y, float z)
> -{
> -  return __builtin_fmaf (x, y, z);
> -}
> -libm_alias_float (__fma, fma)
> 

Ok.

_______________________________________________
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc

      reply	other threads:[~2020-06-01 17:10 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-30  2:00 [PATCH 0/5] use gcc builtins for sqrt and fma functions Vineet Gupta
2020-05-30  2:00 ` [PATCH 1/5] ARC/dl-runtime helper macros Vineet Gupta
2020-05-30  2:17   ` Vineet Gupta
2020-05-30  2:00 ` [PATCH 2/5] iee754: prvoide gcc builtins based generic sqrt functions Vineet Gupta
2020-06-01 14:13   ` Adhemerval Zanella
2020-06-01 14:42     ` Adhemerval Zanella
2020-06-01 18:12       ` Vineet Gupta
2020-06-01 18:32         ` Adhemerval Zanella
2020-06-01 14:18   ` Florian Weimer
2020-06-01 18:12     ` Vineet Gupta
2020-05-30  2:00 ` [PATCH 3/5] iee754: prvoide gcc builtins based generic fma functions Vineet Gupta
2020-06-01 14:20   ` Adhemerval Zanella
2020-06-01 16:13   ` Florian Weimer
2020-05-30  2:00 ` [PATCH 4/5] aarch/fpu: use generic sqrt, " Vineet Gupta
2020-06-01 16:38   ` Adhemerval Zanella
2020-06-01 21:44     ` Vineet Gupta
2020-06-01 22:08       ` Ramana Radhakrishnan
2020-06-02 18:12       ` Joseph Myers
2020-05-30  2:00 ` [PATCH 5/5] powerpc/fpu: use generic " Vineet Gupta
2020-06-01 17:10   ` Adhemerval Zanella [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=8eb04259-7529-cbb4-81b8-d5b0f56eb6b8@linaro.org \
    --to=adhemerval.zanella@linaro.org \
    --cc=Vineet.Gupta1@synopsys.com \
    --cc=libc-alpha@sourceware.org \
    --cc=linux-snps-arc@lists.infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.