diff options
| author | H.J. Lu <hjl.tools@gmail.com> | 2017-10-22 08:09:08 -0700 |
|---|---|---|
| committer | H.J. Lu <hjl.tools@gmail.com> | 2017-10-22 08:10:18 -0700 |
| commit | 6089a3ee24cede17e9443aef0aa72fa1a0ba1548 (patch) | |
| tree | c160e567478cc34bf93e845362968562317a41f2 /sysdeps/i386/fpu/libm-test-ulps | |
| parent | 80bb5935632faf8a8e94daf976340e549be9d4c5 (diff) | |
| download | glibc-6089a3ee24cede17e9443aef0aa72fa1a0ba1548.tar.gz glibc-6089a3ee24cede17e9443aef0aa72fa1a0ba1548.tar.xz glibc-6089a3ee24cede17e9443aef0aa72fa1a0ba1548.zip | |
i386: Replace assembly versions of e_log2f with generic e_log2f.c
This patch replaces i386 assembly versions of e_log2f with generic
e_log2f.c. For workload-spec2017.wrf, on Nehalem, it improves
performance by:
Before After Improvement
reciprocal-throughput 92.3845 30.8752 199%
latency 112.855 54.8645 105%
On Skylake, it improves performance by:
Before After Improvement
reciprocal-throughput 98.7488 22.7507 334%
latency 118.01 51.6083 128%
On IvyBridge with --disable-multi-arch, it improves performance by:
Before After Improvement
reciprocal-throughput 106.635 28.8596 269%
latency 129.888 56.9187 128%
* sysdeps/i386/fpu/e_log2f.S: Removed.
* sysdeps/i386/fpu/e_log2f_data.c: Likewise.
* sysdeps/i386/fpu/w_log2f.c: Likewise.
* sysdeps/i386/fpu/libm-test-ulps: Updated for generic e_log2f.c.
* sysdeps/i386/i686/fpu/multiarch/libm-test-ulps: Likewise.
* sysdeps/i386/i686/fpu/multiarch/Makefile (libm-sysdep_routines):
Add e_log2f-sse2.
(CFLAGS-e_log2f-sse2.c): New.
* sysdeps/i386/i686/fpu/multiarch/e_log2f-sse2.c: New file.
* sysdeps/i386/i686/fpu/multiarch/e_log2f.c: Likewise.
Diffstat (limited to 'sysdeps/i386/fpu/libm-test-ulps')
| -rw-r--r-- | sysdeps/i386/fpu/libm-test-ulps | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/sysdeps/i386/fpu/libm-test-ulps b/sysdeps/i386/fpu/libm-test-ulps index ba9b9ec00f..64cac565f2 100644 --- a/sysdeps/i386/fpu/libm-test-ulps +++ b/sysdeps/i386/fpu/libm-test-ulps @@ -2300,8 +2300,10 @@ ldouble: 3 Function: "log2": double: 1 +float: 1 float128: 2 idouble: 1 +ifloat: 1 ifloat128: 2 ildouble: 1 ldouble: 1 |