This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH][BZ #18740] Fix for SSE vector math tests run on non SSE4.1 machines
- From: Andrew Senkevich <andrew dot n dot senkevich at gmail dot com>
- To: "H.J. Lu" <hjl dot tools at gmail dot com>
- Cc: libc-alpha <libc-alpha at sourceware dot org>
- Date: Thu, 30 Jul 2015 17:51:24 +0300
- Subject: Re: [PATCH][BZ #18740] Fix for SSE vector math tests run on non SSE4.1 machines
- Authentication-results: sourceware.org; auth=none
- References: <CAMXFM3sTR3rng4bYsb0rPi=-YDxHrJdzhzV2CHBgtZw6=4oV1w at mail dot gmail dot com> <CAMe9rOre+A+Y5D5PjvP5qo_1nzZ+ZRPudQPKMzO3Cf7kHzXBcw at mail dot gmail dot com>
2015-07-30 17:42 GMT+03:00 H.J. Lu <hjl.tools@gmail.com>:
> On Thu, Jul 30, 2015 at 7:27 AM, Andrew Senkevich
> <andrew.n.senkevich@gmail.com> wrote:
>> Hi,
>>
>> we can build SSE vector math tests with -msse2 because ifunc in vector
>> function itself selects SSE2 or SSE4 implementation to test.
>>
>> 2015-07-30 Andrew Senkevich <andrew.senkevich@intel.com>
>>
>> [BZ #18740]
>> * sysdeps/x86_64/fpu/Makefile (double-vlen2-arch-ext-cflags,
>> float-vlen4-arch-ext-cflags): Use -msee2 instead of -msse4.
>>
>> diff --git a/ChangeLog b/ChangeLog
>> index bcb14cf..b5c21b5 100644
>> --- a/ChangeLog
>> +++ b/ChangeLog
>> @@ -1,3 +1,9 @@
>> +2015-07-30 Andrew Senkevich <andrew.senkevich@intel.com>
>> +
>> + [BZ #18740]
>> + * sysdeps/x86_64/fpu/Makefile (double-vlen2-arch-ext-cflags,
>> + float-vlen4-arch-ext-cflags): Use -msee2 instead of -msse4.
>> +
>> 2015-07-30 Mike Frysinger <vapier@gentoo.org>
>>
>> * sysdeps/hppa/nptl/bits/pthreadtypes.h (pthread_mutex_t): Change
>> diff --git a/sysdeps/x86_64/fpu/Makefile b/sysdeps/x86_64/fpu/Makefile
>> index e270b8a..28802dd 100644
>> --- a/sysdeps/x86_64/fpu/Makefile
>> +++ b/sysdeps/x86_64/fpu/Makefile
>> @@ -33,12 +33,12 @@ ifeq (yes,$(config-cflags-avx512))
>> libmvec-tests += double-vlen8 float-vlen16
>> endif
>>
>> -double-vlen2-arch-ext-cflags = -msse4
>> +double-vlen2-arch-ext-cflags = -msse2
>> double-vlen4-arch-ext-cflags = -mavx
>> double-vlen4-arch-ext2-cflags = -mavx2
>> double-vlen8-arch-ext-cflags = -mavx512f
>>
>> -float-vlen4-arch-ext-cflags = -msse4
>> +float-vlen4-arch-ext-cflags = -msse2
>> float-vlen8-arch-ext-cflags = -mavx
>> float-vlen8-arch-ext2-cflags = -mavx2
>> float-vlen16-arch-ext-cflags = -mavx512f
>>
>
> SSE2 is the default for x86-64. Please just remove the line.
Here is updated patch.
diff --git a/ChangeLog b/ChangeLog
index bcb14cf..b4321fa 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,11 @@
+2015-07-30 Andrew Senkevich <andrew.senkevich@intel.com>
+
+ [BZ #18740]
+ * sysdeps/x86_64/fpu/Makefile (double-vlen2-arch-ext-cflags,
+ float-vlen4-arch-ext-cflags): Removed.
+ * math/Makefile (CFLAGS-test-double-vlen2-wrappers.c,
+ CFLAGS-test-float-vlen4-wrappers.c): Likewise.
+
2015-07-30 Mike Frysinger <vapier@gentoo.org>
* sysdeps/hppa/nptl/bits/pthreadtypes.h (pthread_mutex_t): Change
diff --git a/math/Makefile b/math/Makefile
index 2c78dd0..6388bae 100644
--- a/math/Makefile
+++ b/math/Makefile
@@ -151,7 +151,6 @@ libm-test-vec-cflags =
$(libm-test-fast-math-cflags) -fno-inline \
-ffloat-store -D_OPENMP=201307 -Wno-unknown-pragmas
CFLAGS-test-double-vlen2.c = $(libm-test-vec-cflags)
-CFLAGS-test-double-vlen2-wrappers.c = $(double-vlen2-arch-ext-cflags)
CFLAGS-test-double-vlen4.c = $(libm-test-vec-cflags)
CFLAGS-test-double-vlen4-wrappers.c = $(double-vlen4-arch-ext-cflags)
@@ -160,7 +159,6 @@ CFLAGS-test-double-vlen8.c = $(libm-test-vec-cflags)
CFLAGS-test-double-vlen8-wrappers.c = $(double-vlen8-arch-ext-cflags)
CFLAGS-test-float-vlen4.c = $(libm-test-vec-cflags)
-CFLAGS-test-float-vlen4-wrappers.c = $(float-vlen4-arch-ext-cflags)
CFLAGS-test-float-vlen8.c = $(libm-test-vec-cflags)
CFLAGS-test-float-vlen8-wrappers.c = $(float-vlen8-arch-ext-cflags)
diff --git a/sysdeps/x86_64/fpu/Makefile b/sysdeps/x86_64/fpu/Makefile
index e270b8a..1ebe511 100644
--- a/sysdeps/x86_64/fpu/Makefile
+++ b/sysdeps/x86_64/fpu/Makefile
@@ -33,12 +33,10 @@ ifeq (yes,$(config-cflags-avx512))
libmvec-tests += double-vlen8 float-vlen16
endif
-double-vlen2-arch-ext-cflags = -msse4
double-vlen4-arch-ext-cflags = -mavx
double-vlen4-arch-ext2-cflags = -mavx2
double-vlen8-arch-ext-cflags = -mavx512f
-float-vlen4-arch-ext-cflags = -msse4
float-vlen8-arch-ext-cflags = -mavx
float-vlen8-arch-ext2-cflags = -mavx2
float-vlen16-arch-ext-cflags = -mavx512f
--
WBR,
Andrew