about summary refs log tree commit diff
path: root/sysdeps/x86_64/multiarch/memset.S
diff options
context:
space:
mode:
authorH.J. Lu <hjl.tools@gmail.com>2014-07-14 07:58:27 -0700
committerH.J. Lu <hjl.tools@gmail.com>2014-07-14 07:58:27 -0700
commitf2fef657d8736c32fb600771949f59852558b11a (patch)
treea9e954223fa5cbd9e53f143e0f50c6f6276ebe0c /sysdeps/x86_64/multiarch/memset.S
parentf6c44d475104e931bab2b4ffa499961088de673c (diff)
downloadglibc-f2fef657d8736c32fb600771949f59852558b11a.tar.gz
glibc-f2fef657d8736c32fb600771949f59852558b11a.tar.xz
glibc-f2fef657d8736c32fb600771949f59852558b11a.zip
Enable AVX2 optimized memset only if -mavx2 works
	* config.h.in (HAVE_AVX2_SUPPORT): New #undef.
	* sysdeps/i386/configure.ac: Set HAVE_AVX2_SUPPORT and
	config-cflags-avx2.
	* sysdeps/x86_64/configure.ac: Likewise.
	* sysdeps/i386/configure: Regenerated.
	* sysdeps/x86_64/configure: Likewise.
	* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
	memset-avx2 only if config-cflags-avx2 is yes.
	* sysdeps/x86_64/multiarch/ifunc-impl-list.c (__libc_ifunc_impl_list):
	Tests for memset_chk and memset only if HAVE_AVX2_SUPPORT is
	defined.
	* sysdeps/x86_64/multiarch/memset.S: Define multiple versions
	only if HAVE_AVX2_SUPPORT is defined.
	* sysdeps/x86_64/multiarch/memset_chk.S: Likewise.
Diffstat (limited to 'sysdeps/x86_64/multiarch/memset.S')
-rw-r--r--sysdeps/x86_64/multiarch/memset.S24
1 files changed, 13 insertions, 11 deletions
diff --git a/sysdeps/x86_64/multiarch/memset.S b/sysdeps/x86_64/multiarch/memset.S
index 3113d1cbc0..00d46d12d0 100644
--- a/sysdeps/x86_64/multiarch/memset.S
+++ b/sysdeps/x86_64/multiarch/memset.S
@@ -17,12 +17,13 @@
    License along with the GNU C Library; if not, see
    <http://www.gnu.org/licenses/>.  */
 
+#ifdef HAVE_AVX2_SUPPORT
 #include <sysdep.h>
 #include <shlib-compat.h>
 #include <init-arch.h>
 
 /* Define multiple versions only for the definition in lib.  */
-#ifndef NOT_IN_libc
+# ifndef NOT_IN_libc
 ENTRY(memset)
 	.type	memset, @gnu_indirect_function
 	cmpl	$0, __cpu_features+KIND_OFFSET(%rip)
@@ -34,26 +35,27 @@ ENTRY(memset)
 	leaq	__memset_avx2(%rip), %rax
 2:	ret
 END(memset)
-#endif
+# endif
 
-#if !defined NOT_IN_libc
-# undef memset
-# define memset __memset_sse2
+# if !defined NOT_IN_libc
+#  undef memset
+#  define memset __memset_sse2
 
-# undef __memset_chk
-# define __memset_chk __memset_chk_sse2
+#  undef __memset_chk
+#  define __memset_chk __memset_chk_sse2
 
-# ifdef SHARED
+#  ifdef SHARED
 #  undef libc_hidden_builtin_def
 /* It doesn't make sense to send libc-internal memset calls through a PLT.
    The speedup we get from using GPR instruction is likely eaten away
    by the indirect call in the PLT.  */
 #  define libc_hidden_builtin_def(name) \
 	.globl __GI_memset; __GI_memset = __memset_sse2
-# endif
+#  endif
 
-# undef strong_alias
-# define strong_alias(original, alias)
+#  undef strong_alias
+#  define strong_alias(original, alias)
+# endif
 #endif
 
 #include "../memset.S"