[v2,0/3] Optimize CORE-MATH internal tables

Message ID	20250213193737.3784065-1-adhemerval.zanella@linaro.org
Headers	show Delivered-To: patch@linaro.org Received-SPF: pass (google.com: domain of libc-alpha-bounces~patch=linaro.org@sourceware.org designates 8.43.85.97 as permitted sender) client-ip=8.43.85.97; DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org B56C23858D20 From: Adhemerval Zanella <adhemerval.zanella@linaro.org> To: libc-alpha@sourceware.org Cc: "Andreas K . Huettel" <dilfridge@gentoo.org>, Paul Zimmermann <Paul.Zimmermann@inria.fr>, Alexei Sibidanov <sibid@uvic.ca>, DJ Delorie <dj@redhat.com> Subject: [PATCH v2 0/3] Optimize CORE-MATH internal tables Date: Thu, 13 Feb 2025 16:35:42 -0300 Message-ID: <20250213193737.3784065-1-adhemerval.zanella@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Precedence: list Errors-To: libc-alpha-bounces~patch=linaro.org@sourceware.org
Series	Optimize CORE-MATH internal tables \| expand [v2,0/3] Optimize CORE-MATH internal tables [v2,1/3] math: Consolidate cospif and sinpif internal tables [v2,2/3] math: Consolidate acospif and asinpif internal tables [v2,3/3] math: Consolidate acosf and asinf internal tables

Message ID

20250213193737.3784065-1-adhemerval.zanella@linaro.org

Headers

Received-SPF: pass (google.com: domain of
 libc-alpha-bounces~patch=linaro.org@sourceware.org designates 8.43.85.97 as
 permitted sender) client-ip=8.43.85.97;
DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org B56C23858D20
From: Adhemerval Zanella <adhemerval.zanella@linaro.org>
To: libc-alpha@sourceware.org
Cc: "Andreas K . Huettel" <dilfridge@gentoo.org>,
 Paul Zimmermann <Paul.Zimmermann@inria.fr>,
 Alexei Sibidanov <sibid@uvic.ca>, DJ Delorie <dj@redhat.com>
Subject: [PATCH v2 0/3] Optimize CORE-MATH internal tables
Date: Thu, 13 Feb 2025 16:35:42 -0300
Message-ID: <20250213193737.3784065-1-adhemerval.zanella@linaro.org>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: list
Errors-To: libc-alpha-bounces~patch=linaro.org@sourceware.org

Series

Optimize CORE-MATH internal tables | expand

Message

Adhemerval Zanella Netto Feb. 13, 2025, 7:35 p.m. UTC

Some implementation share internal table, which can be optimized
to avoid duplicate data.  The are small code or no changes for x86_64
and powerpc64le, which do not affect performance; but on aarch64 with
gcc-14 I see a slight better code generation due the usage of ldq for
floating point constant loading.

Besides glibc own testsuite, I also checked with CORE-MATH tests
(which covers all inputs for all rounding modes).

Adhemerval Zanella (3):
  math: Consolidate cospif and sinpif internal tables
  math: Consolidate acospif and asinpif internal tables
  math: Consolidate acosf and asinf internal tables

 math/Makefile                              |  3 +
 sysdeps/ieee754/flt-32/e_acosf.c           | 19 +----
 sysdeps/ieee754/flt-32/e_asinf.c           | 19 +----
 sysdeps/ieee754/flt-32/s_acospif.c         | 53 +-------------
 sysdeps/ieee754/flt-32/s_asincosf_data.c   | 43 ++++++++++++
 sysdeps/ieee754/flt-32/s_asincosf_data.h   | 35 ++++++++++
 sysdeps/ieee754/flt-32/s_asincospif_data.c | 81 ++++++++++++++++++++++
 sysdeps/ieee754/flt-32/s_asincospif_data.h | 34 +++++++++
 sysdeps/ieee754/flt-32/s_asinpif.c         | 55 +--------------
 sysdeps/ieee754/flt-32/s_cospif.c          | 61 +---------------
 sysdeps/ieee754/flt-32/s_sincospif_data.c  | 80 +++++++++++++++++++++
 sysdeps/ieee754/flt-32/s_sincospif_data.h  | 37 ++++++++++
 sysdeps/ieee754/flt-32/s_sinpif.c          | 60 +---------------
 13 files changed, 329 insertions(+), 251 deletions(-)
 create mode 100644 sysdeps/ieee754/flt-32/s_asincosf_data.c
 create mode 100644 sysdeps/ieee754/flt-32/s_asincosf_data.h
 create mode 100644 sysdeps/ieee754/flt-32/s_asincospif_data.c
 create mode 100644 sysdeps/ieee754/flt-32/s_asincospif_data.h
 create mode 100644 sysdeps/ieee754/flt-32/s_sincospif_data.c
 create mode 100644 sysdeps/ieee754/flt-32/s_sincospif_data.h