Optimized strchr and strrchr with SSE2 on x86-32

9 files changed, 1671 insertions, 1 deletions

diff --git a/ChangeLog b/ChangeLog
index 3d10ec2..b0a79e1 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,15 @@
+2011-07-20  Liubov Dmitrieva  <liubov.dmitrieva@intel.com>
+
+	* sysdeps/i386/i686/multiarch/Makefile (sysdep_routines): Add
+	strchr-sse2 strrchr-sse2 strchr-sse2-bsf
+	strrchr-sse2-bsf
+	* sysdeps/i386/i686/multiarch/strchr.S: New file.
+	* sysdeps/i386/i686/multiarch/strrchr.S: New file.
+	* sysdeps/i386/i686/multiarch/strchr-sse2.S: New file.
+	* sysdeps/i386/i686/multiarch/strchr-sse2-bsf.S: New file.
+	* sysdeps/i386/i686/multiarch/strrchr-sse2.S: New file.
+	* sysdeps/i386/i686/multiarch/strrchr-sse2-bsf.S: New file.
+
 2011-08-29  Liubov Dmitrieva  <liubov.dmitrieva@gmail.com>
 
 	* sysdeps/x86_64/wcscmp.S: New file.
diff --git a/NEWS b/NEWS
index 5e407f6..e38c57f 100644
--- a/NEWS
+++ b/NEWS
@@ -29,6 +29,9 @@ Version 2.15
   x86-64.
   Contributed by Liubov Dmitrieva.
 
+* Optimized strchrm strrchr for SSE on x86-32.
+  Contributed by Liubov Dmitrieva.
+
 * New interfaces: scandirat, scandirat64
   Implemented by Ulrich Drepper.
 
diff --git a/sysdeps/i386/i686/multiarch/Makefile b/sysdeps/i386/i686/multiarch/Makefile
index 1449d91..c89ae92 100644
--- a/sysdeps/i386/i686/multiarch/Makefile
+++ b/sysdeps/i386/i686/multiarch/Makefile
@@ -14,7 +14,8 @@ sysdep_routines += bzero-sse2 memset-sse2 memcpy-ssse3 mempcpy-ssse3 \
 		   strncpy-ssse3 stpcpy-ssse3 stpncpy-ssse3 strcpy-sse2 \
 		   strncpy-sse2 stpcpy-sse2 stpncpy-sse2 strcat-ssse3 \
 		   strcat-sse2 strncat-ssse3 strncat-sse2 strncat-c \
-			wcscmp-sse2 wcscmp-c
+		   strchr-sse2 strrchr-sse2 strchr-sse2-bsf strrchr-sse2-bsf \
+		   wcscmp-sse2 wcscmp-c
 ifeq (yes,$(config-cflags-sse4))
 sysdep_routines += strcspn-c strpbrk-c strspn-c strstr-c strcasestr-c
 CFLAGS-varshift.c += -msse4
diff --git a/sysdeps/i386/i686/multiarch/strchr-sse2-bsf.S b/sysdeps/i386/i686/multiarch/strchr-sse2-bsf.S
new file mode 100644
index 0000000..5a19ba2
--- /dev/null
+++ b/sysdeps/i386/i686/multiarch/strchr-sse2-bsf.S
@@ -0,0 +1,159 @@
+/* strchr with SSE2 with bsf
+   Copyright (C) 2011 Free Software Foundation, Inc.
+   Contributed by Intel Corporation.
+   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, write to the Free
+   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+   02111-1307 USA.  */
+
+#ifndef NOT_IN_libc
+
+# include <sysdep.h>
+
+# define CFI_PUSH(REG)	\
+	cfi_adjust_cfa_offset (4);	\
+	cfi_rel_offset (REG, 0)
+
+# define CFI_POP(REG)	\
+	cfi_adjust_cfa_offset (-4);	\
+	cfi_restore (REG)
+
+# define PUSH(REG) pushl REG; CFI_PUSH (REG)
+# define POP(REG) popl REG; CFI_POP (REG)
+
+# define PARMS  8
+# define ENTRANCE PUSH(%edi)
+# define RETURN  POP(%edi); ret; CFI_PUSH(%edi);
+
+# define STR1  PARMS
+# define STR2  STR1+4
+
+	.text
+ENTRY (__strchr_sse2_bsf)
+
+	ENTRANCE
+	mov	STR1(%esp), %ecx
+	movd	STR2(%esp), %xmm1
+
+	pxor	%xmm2, %xmm2
+	mov	%ecx, %edi
+	punpcklbw %xmm1, %xmm1
+	punpcklbw %xmm1, %xmm1
+	/* ECX has OFFSET. */
+	and	$15, %ecx
+	pshufd	$0, %xmm1, %xmm1
+	je	L(loop)
+
+/* Handle unaligned string.  */
+	and	$-16, %edi
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	pcmpeqb	%xmm1, %xmm0
+	/* Find where NULL is.  */
+	pmovmskb %xmm2, %edx
+	/* Check if there is a match.  */
+	pmovmskb %xmm0, %eax
+	/* Remove the leading bytes.  */
+	sarl	%cl, %edx
+	sarl	%cl, %eax
+	test	%eax, %eax
+	je	L(unaligned_no_match)
+	/* Check which byte is a match.  */
+	bsf	%eax, %eax
+	/* Is there a NULL? */
+	test	%edx, %edx
+	je	L(unaligned_match)
+	bsf	%edx, %edx
+	cmpl	%edx, %eax
+	/* Return NULL if NULL comes first.  */
+	ja	L(return_null)
+L(unaligned_match):
+	add	%edi, %eax
+	add	%ecx, %eax
+	RETURN
+
+	.p2align 4
+L(unaligned_no_match):
+	test	%edx, %edx
+	jne	L(return_null)
+	pxor	%xmm2, %xmm2
+
+	add	$16, %edi
+
+	.p2align 4
+/* Loop start on aligned string.  */
+L(loop):
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %edx
+	pmovmskb %xmm0, %eax
+	or	%eax, %edx
+	jnz	L(matches)
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %edx
+	pmovmskb %xmm0, %eax
+	or	%eax, %edx
+	jnz	L(matches)
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %edx
+	pmovmskb %xmm0, %eax
+	or	%eax, %edx
+	jnz	L(matches)
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %edx
+	pmovmskb %xmm0, %eax
+	or	%eax, %edx
+	jnz	L(matches)
+	jmp	L(loop)
+
+L(matches):
+	pmovmskb %xmm2, %edx
+	test	%eax, %eax
+	jz	L(return_null)
+	bsf	%eax, %eax
+	/* There is a match.  First find where NULL is.  */
+	test	%edx, %edx
+	je	L(match)
+	bsf	%edx, %ecx
+	/* Check if NULL comes first.  */
+	cmpl	%ecx, %eax
+	ja	L(return_null)
+L(match):
+	sub	$16, %edi
+	add	%edi, %eax
+	RETURN
+
+/* Return NULL.  */
+	.p2align 4
+L(return_null):
+	xor	%eax, %eax
+	RETURN
+
+END (__strchr_sse2_bsf)
+#endif
diff --git a/sysdeps/i386/i686/multiarch/strchr-sse2.S b/sysdeps/i386/i686/multiarch/strchr-sse2.S
new file mode 100644
index 0000000..7702210
--- /dev/null
+++ b/sysdeps/i386/i686/multiarch/strchr-sse2.S
@@ -0,0 +1,350 @@
+/* strchr SSE2 without bsf
+   Copyright (C) 2011 Free Software Foundation, Inc.
+   Contributed by Intel Corporation.
+   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, write to the Free
+   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+   02111-1307 USA.  */
+
+#ifndef NOT_IN_libc
+
+# include <sysdep.h>
+
+# define CFI_PUSH(REG)	\
+	cfi_adjust_cfa_offset (4);	\
+	cfi_rel_offset (REG, 0)
+
+# define CFI_POP(REG)	\
+	cfi_adjust_cfa_offset (-4);	\
+	cfi_restore (REG)
+
+# define PUSH(REG) pushl REG; CFI_PUSH (REG)
+# define POP(REG) popl REG; CFI_POP (REG)
+
+# define PARMS  8
+# define ENTRANCE PUSH(%edi)
+# define RETURN  POP(%edi); ret; CFI_PUSH(%edi);
+
+# define STR1  PARMS
+# define STR2  STR1+4
+
+	.text
+ENTRY (__strchr_sse2)
+
+	ENTRANCE
+	mov	STR1(%esp), %ecx
+	movd	STR2(%esp), %xmm1
+
+	pxor	%xmm2, %xmm2
+	mov	%ecx, %edi
+	punpcklbw %xmm1, %xmm1
+	punpcklbw %xmm1, %xmm1
+	/* ECX has OFFSET. */
+	and	$15, %ecx
+	pshufd	$0, %xmm1, %xmm1
+	je	L(loop)
+
+/* Handle unaligned string.  */
+	and	$-16, %edi
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	pcmpeqb	%xmm1, %xmm0
+	/* Find where NULL is.  */
+	pmovmskb %xmm2, %edx
+	/* Check if there is a match.  */
+	pmovmskb %xmm0, %eax
+	/* Remove the leading bytes.  */
+	sarl	%cl, %edx
+	sarl	%cl, %eax
+	test	%eax, %eax
+	jz	L(unaligned_no_match)
+	/* Check which byte is a match.  */
+	/* Is there a NULL? */
+	add	%ecx, %edi
+	test	%edx, %edx
+	jz	L(match_case1)
+	jmp	L(match_case2)
+
+	.p2align 4
+L(unaligned_no_match):
+	test	%edx, %edx
+	jne	L(return_null)
+
+	pxor	%xmm2, %xmm2
+	add	$16, %edi
+
+	.p2align 4
+/* Loop start on aligned string.  */
+L(loop):
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %edx
+	pmovmskb %xmm0, %eax
+	test	%eax, %eax
+	jnz	L(matches)
+	test	%edx, %edx
+	jnz	L(return_null)
+	add	$16, %edi
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %edx
+	pmovmskb %xmm0, %eax
+	test	%eax, %eax
+	jnz	L(matches)
+	test	%edx, %edx
+	jnz	L(return_null)
+	add	$16, %edi
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %edx
+	pmovmskb %xmm0, %eax
+	test	%eax, %eax
+	jnz	L(matches)
+	test	%edx, %edx
+	jnz	L(return_null)
+	add	$16, %edi
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %edx
+	pmovmskb %xmm0, %eax
+	test	%eax, %eax
+	jnz	L(matches)
+	test	%edx, %edx
+	jnz	L(return_null)
+	add	$16, %edi
+	jmp	L(loop)
+
+L(matches):
+	/* There is a match.  First find where NULL is.  */
+	test	%edx, %edx
+	jz	L(match_case1)
+
+	.p2align 4
+L(match_case2):
+	test	%al, %al
+	jz	L(match_higth_case2)
+
+	mov	%al, %cl
+	and	$15, %cl
+	jnz	L(match_case2_4)
+
+	mov	%dl, %ch
+	and	$15, %ch
+	jnz	L(return_null)
+
+	test	$0x10, %al
+	jnz	L(Exit5)
+	test	$0x10, %dl
+	jnz	L(return_null)
+	test	$0x20, %al
+	jnz	L(Exit6)
+	test	$0x20, %dl
+	jnz	L(return_null)
+	test	$0x40, %al
+	jnz	L(Exit7)
+	test	$0x40, %dl
+	jnz	L(return_null)
+	lea	7(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(match_case2_4):
+	test	$0x01, %al
+	jnz	L(Exit1)
+	test	$0x01, %dl
+	jnz	L(return_null)
+	test	$0x02, %al
+	jnz	L(Exit2)
+	test	$0x02, %dl
+	jnz	L(return_null)
+	test	$0x04, %al
+	jnz	L(Exit3)
+	test	$0x04, %dl
+	jnz	L(return_null)
+	lea	3(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(match_higth_case2):
+	test	%dl, %dl
+	jnz	L(return_null)
+
+	mov	%ah, %cl
+	and	$15, %cl
+	jnz	L(match_case2_12)
+
+	mov	%dh, %ch
+	and	$15, %ch
+	jnz	L(return_null)
+
+	test	$0x10, %ah
+	jnz	L(Exit13)
+	test	$0x10, %dh
+	jnz	L(return_null)
+	test	$0x20, %ah
+	jnz	L(Exit14)
+	test	$0x20, %dh
+	jnz	L(return_null)
+	test	$0x40, %ah
+	jnz	L(Exit15)
+	test	$0x40, %dh
+	jnz	L(return_null)
+	lea	15(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(match_case2_12):
+	test	$0x01, %ah
+	jnz	L(Exit9)
+	test	$0x01, %dh
+	jnz	L(return_null)
+	test	$0x02, %ah
+	jnz	L(Exit10)
+	test	$0x02, %dh
+	jnz	L(return_null)
+	test	$0x04, %ah
+	jnz	L(Exit11)
+	test	$0x04, %dh
+	jnz	L(return_null)
+	lea	11(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(match_case1):
+	test	%al, %al
+	jz	L(match_higth_case1)
+
+	test	$0x01, %al
+	jnz	L(Exit1)
+	test	$0x02, %al
+	jnz	L(Exit2)
+	test	$0x04, %al
+	jnz	L(Exit3)
+	test	$0x08, %al
+	jnz	L(Exit4)
+	test	$0x10, %al
+	jnz	L(Exit5)
+	test	$0x20, %al
+	jnz	L(Exit6)
+	test	$0x40, %al
+	jnz	L(Exit7)
+	lea	7(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(match_higth_case1):
+	test	$0x01, %ah
+	jnz	L(Exit9)
+	test	$0x02, %ah
+	jnz	L(Exit10)
+	test	$0x04, %ah
+	jnz	L(Exit11)
+	test	$0x08, %ah
+	jnz	L(Exit12)
+	test	$0x10, %ah
+	jnz	L(Exit13)
+	test	$0x20, %ah
+	jnz	L(Exit14)
+	test	$0x40, %ah
+	jnz	L(Exit15)
+	lea	15(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit1):
+	lea	(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit2):
+	lea	1(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit3):
+	lea	2(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit4):
+	lea	3(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit5):
+	lea	4(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit6):
+	lea	5(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit7):
+	lea	6(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit9):
+	lea	8(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit10):
+	lea	9(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit11):
+	lea	10(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit12):
+	lea	11(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit13):
+	lea	12(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit14):
+	lea	13(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit15):
+	lea	14(%edi), %eax
+	RETURN
+
+/* Return NULL.  */
+	.p2align 4
+L(return_null):
+	xor	%eax, %eax
+	RETURN
+
+END (__strchr_sse2)
+#endif
+
diff --git a/sysdeps/i386/i686/multiarch/strchr.S b/sysdeps/i386/i686/multiarch/strchr.S
new file mode 100644
index 0000000..aed967c
--- /dev/null
+++ b/sysdeps/i386/i686/multiarch/strchr.S
@@ -0,0 +1,76 @@
+/* Multiple versions of strchr
+   Copyright (C) 2011 Free Software Foundation, Inc.
+   Contributed by Intel Corporation.
+   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, write to the Free
+   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+   02111-1307 USA.  */
+
+#include <sysdep.h>
+#include <init-arch.h>
+
+#ifndef NOT_IN_libc
+	.section	.gnu.linkonce.t.__i686.get_pc_thunk.bx,"ax",@progbits
+	.globl	__i686.get_pc_thunk.bx
+	.hidden	__i686.get_pc_thunk.bx
+	.p2align 4
+	.type	__i686.get_pc_thunk.bx,@function
+__i686.get_pc_thunk.bx:
+	movl	(%esp), %ebx
+	ret
+
+	.text
+ENTRY(strchr)
+	.type	strchr, @gnu_indirect_function
+	pushl	%ebx
+	cfi_adjust_cfa_offset (4)
+	cfi_rel_offset (ebx, 0)
+	call	__i686.get_pc_thunk.bx
+	addl	$_GLOBAL_OFFSET_TABLE_, %ebx
+	cmpl	$0, KIND_OFFSET+__cpu_features@GOTOFF(%ebx)
+	jne	1f
+	call	__init_cpu_features
+1:	leal	__strchr_ia32@GOTOFF(%ebx), %eax
+	testl	$bit_SSE2, CPUID_OFFSET+index_SSE2+__cpu_features@GOTOFF(%ebx)
+	jz	2f
+	leal	__strchr_sse2_bsf@GOTOFF(%ebx), %eax
+	testl	$bit_Slow_BSF, FEATURE_OFFSET+index_Slow_BSF+__cpu_features@GOTOFF(%ebx)
+	jz	2f
+	leal	__strchr_sse2@GOTOFF(%ebx), %eax
+2:	popl	%ebx
+	cfi_adjust_cfa_offset (-4);
+	cfi_restore (ebx)
+	ret
+END(strchr)
+
+# undef ENTRY
+# define ENTRY(name) \
+	.type __strchr_ia32, @function; \
+	.globl __strchr_ia32; \
+	.p2align 4; \
+	__strchr_ia32: cfi_startproc; \
+	CALL_MCOUNT
+# undef END
+# define END(name) \
+	cfi_endproc; .size __strchr_ia32, .-__strchr_ia32
+# undef libc_hidden_builtin_def
+/* IFUNC doesn't work with the hidden functions in shared library since
+   they will be called without setting up EBX needed for PLT which is
+   used by IFUNC.  */
+# define libc_hidden_builtin_def(name) \
+	.globl __GI_strchr; __GI_strchr = __strchr_ia32
+#endif
+
+#include "../../i586/strchr.S"
diff --git a/sysdeps/i386/i686/multiarch/strrchr-sse2-bsf.S b/sysdeps/i386/i686/multiarch/strrchr-sse2-bsf.S
new file mode 100644
index 0000000..f1bcb78
--- /dev/null
+++ b/sysdeps/i386/i686/multiarch/strrchr-sse2-bsf.S
@@ -0,0 +1,284 @@
+/* strrchr with SSE2 with bsf and bsr
+   Copyright (C) 2011 Free Software Foundation, Inc.
+   Contributed by Intel Corporation.
+   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, write to the Free
+   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+   02111-1307 USA.  */
+
+#ifndef NOT_IN_libc
+
+# include <sysdep.h>
+
+# define CFI_PUSH(REG)	\
+	cfi_adjust_cfa_offset (4);	\
+	cfi_rel_offset (REG, 0)
+
+# define CFI_POP(REG)	\
+	cfi_adjust_cfa_offset (-4);	\
+	cfi_restore (REG)
+
+# define PUSH(REG) pushl REG; CFI_PUSH (REG)
+# define POP(REG) popl REG; CFI_POP (REG)
+
+# define PARMS  4
+# define STR1  PARMS
+# define STR2  STR1+4
+
+	.text
+ENTRY (__strrchr_sse2_bsf)
+
+	mov	STR1(%esp), %ecx
+	movd	STR2(%esp), %xmm1
+
+	PUSH	(%edi)
+	pxor	%xmm2, %xmm2
+	mov	%ecx, %edi
+	punpcklbw %xmm1, %xmm1
+	punpcklbw %xmm1, %xmm1
+	/* ECX has OFFSET. */
+	and	$63, %ecx
+	cmp	$48, %ecx
+	pshufd	$0, %xmm1, %xmm1
+	ja	L(crosscashe)
+
+/* unaligned string. */
+	movdqu	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	pcmpeqb	%xmm1, %xmm0
+	/* Find where NULL is.  */
+	pmovmskb %xmm2, %edx
+	/* Check if there is a match.  */
+	pmovmskb %xmm0, %eax
+
+	test	%eax, %eax
+	jnz	L(unaligned_match1)
+
+	test	%edx, %edx
+	jnz	L(return_null)
+
+	and	$-16, %edi
+	add	$16, %edi
+
+	PUSH	(%esi)
+	PUSH	(%ebx)
+
+	xor	%ebx, %ebx
+	jmp	L(loop)
+
+	CFI_POP	(%esi)
+	CFI_POP	(%ebx)
+
+	.p2align 4
+L(unaligned_return_value1):
+	bsf	%edx, %ecx
+	mov	$2, %edx
+	shl	%cl, %edx
+	sub	$1, %edx
+	and	%edx, %eax
+	jz	L(return_null)
+	bsr	%eax, %eax
+	add	%edi, %eax
+	POP	(%edi)
+	ret
+	CFI_PUSH	(%edi)
+
+	.p2align 4
+L(unaligned_match1):
+	test	%edx, %edx
+	jnz	L(unaligned_return_value1)
+
+	PUSH	(%esi)
+	PUSH	(%ebx)
+
+	mov	%eax, %ebx
+	lea	16(%edi), %esi
+	and	$-16, %edi
+	add	$16, %edi
+	jmp	L(loop)
+
+	CFI_POP	(%esi)
+	CFI_POP	(%ebx)
+
+	.p2align 4
+	L(crosscashe):
+/* Hancle unaligned string.  */
+	and	$15, %ecx
+	and	$-16, %edi
+	pxor	%xmm3, %xmm3
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm3
+	pcmpeqb	%xmm1, %xmm0
+	/* Find where NULL is.  */
+	pmovmskb %xmm3, %edx
+	/* Check if there is a match.  */
+	pmovmskb %xmm0, %eax
+	/* Remove the leading bytes.  */
+	shr	%cl, %edx
+	shr	%cl, %eax
+
+	test	%eax, %eax
+	jnz	L(unaligned_match)
+
+	test	%edx, %edx
+	jnz	L(return_null)
+
+	add	$16, %edi
+
+	PUSH	(%esi)
+	PUSH	(%ebx)
+
+	xor	%ebx, %ebx
+	jmp	L(loop)
+
+	CFI_POP	(%esi)
+	CFI_POP	(%ebx)
+
+	.p2align 4
+L(unaligned_return_value):
+	add	%ecx, %edi
+	bsf	%edx, %ecx
+	mov	$2, %edx
+	shl	%cl, %edx
+	sub	$1, %edx
+	and	%edx, %eax
+	jz	L(return_null)
+	bsr	%eax, %eax
+	add	%edi, %eax
+	POP	(%edi)
+	ret
+	CFI_PUSH	(%edi)
+
+	.p2align 4
+L(unaligned_match):
+	test	%edx, %edx
+	jnz	L(unaligned_return_value)
+
+	PUSH	(%esi)
+	PUSH	(%ebx)
+
+	mov	%eax, %ebx
+	add	$16, %edi
+	lea	(%edi, %ecx), %esi
+
+/* Loop start on aligned string.  */
+	.p2align 4
+L(loop):
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %ecx
+	pmovmskb %xmm0, %eax
+	or	%eax, %ecx
+	jnz	L(matches)
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %ecx
+	pmovmskb %xmm0, %eax
+	or	%eax, %ecx
+	jnz	L(matches)
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %ecx
+	pmovmskb %xmm0, %eax
+	or	%eax, %ecx
+	jnz	L(matches)
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %ecx
+	pmovmskb %xmm0, %eax
+	or	%eax, %ecx
+	jz	L(loop)
+
+L(matches):
+	test	%eax, %eax
+	jnz	L(match)
+L(return_value):
+	test	%ebx, %ebx
+	jz	L(return_null_1)
+	bsr	%ebx, %eax
+	add	%esi, %eax
+
+	POP	(%ebx)
+	POP	(%esi)
+
+	sub	$16, %eax
+	POP	(%edi)
+	ret
+
+	CFI_PUSH	(%edi)
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(match):
+	pmovmskb %xmm2, %ecx
+	test	%ecx, %ecx
+	jnz	L(return_value_1)
+	mov	%eax, %ebx
+	mov	%edi, %esi
+	jmp	L(loop)
+
+	.p2align 4
+L(return_value_1):
+	bsf	%ecx, %ecx
+	mov	$2, %edx
+	shl	%cl, %edx
+	sub	$1, %edx
+	and	%edx, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+
+	bsr	%eax, %eax
+	add	%edi, %eax
+	sub	$16, %eax
+	POP	(%edi)
+	ret
+
+	CFI_PUSH	(%edi)
+/* Return NULL.  */
+	.p2align 4
+L(return_null):
+	xor	%eax, %eax
+	POP	(%edi)
+	ret
+
+	CFI_PUSH	(%edi)
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+/* Return NULL.  */
+	.p2align 4
+L(return_null_1):
+	POP	(%ebx)
+	POP	(%esi)
+	POP	(%edi)
+	xor	%eax, %eax
+	ret
+
+END (__strrchr_sse2_bsf)
+#endif
+
diff --git a/sysdeps/i386/i686/multiarch/strrchr-sse2.S b/sysdeps/i386/i686/multiarch/strrchr-sse2.S
new file mode 100644
index 0000000..71cc69d
--- /dev/null
+++ b/sysdeps/i386/i686/multiarch/strrchr-sse2.S
@@ -0,0 +1,709 @@
+/* strrchr SSE2 without bsf and bsr
+   Copyright (C) 2011 Free Software Foundation, Inc.
+   Contributed by Intel Corporation.
+   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, write to the Free
+   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+   02111-1307 USA.  */
+
+#ifndef NOT_IN_libc
+
+# include <sysdep.h>
+
+# define CFI_PUSH(REG)	\
+	cfi_adjust_cfa_offset (4);	\
+	cfi_rel_offset (REG, 0)
+
+# define CFI_POP(REG)	\
+	cfi_adjust_cfa_offset (-4);	\
+	cfi_restore (REG)
+
+# define PUSH(REG) pushl REG; CFI_PUSH (REG)
+# define POP(REG) popl REG; CFI_POP (REG)
+
+# define PARMS  8
+# define ENTRANCE PUSH(%edi);
+# define RETURN  POP(%edi); ret; CFI_PUSH(%edi);
+
+# define STR1  PARMS
+# define STR2  STR1+4
+
+	.text
+ENTRY (__strrchr_sse2)
+
+	ENTRANCE
+	mov	STR1(%esp), %ecx
+	movd	STR2(%esp), %xmm1
+
+	pxor	%xmm2, %xmm2
+	mov	%ecx, %edi
+	punpcklbw %xmm1, %xmm1
+	punpcklbw %xmm1, %xmm1
+	/* ECX has OFFSET. */
+	and	$63, %ecx
+	cmp	$48, %ecx
+	pshufd	$0, %xmm1, %xmm1
+	ja	L(crosscache)
+
+/* unaligned string. */
+	movdqu	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	pcmpeqb	%xmm1, %xmm0
+	/* Find where NULL is.  */
+	pmovmskb %xmm2, %ecx
+	/* Check if there is a match.  */
+	pmovmskb %xmm0, %eax
+	add	$16, %edi
+
+	test	%eax, %eax
+	jnz	L(unaligned_match1)
+
+	test	%ecx, %ecx
+	jnz	L(return_null)
+
+	and	$-16, %edi
+
+	PUSH	(%esi)
+	PUSH	(%ebx)
+
+	xor	%ebx, %ebx
+	jmp	L(loop)
+
+	CFI_POP	(%esi)
+	CFI_POP	(%ebx)
+
+	.p2align 4
+L(unaligned_match1):
+	test	%ecx, %ecx
+	jnz	L(prolog_find_zero_1)
+
+	PUSH	(%esi)
+	PUSH	(%ebx)
+
+	mov	%eax, %ebx
+	mov	%edi, %esi
+	and	$-16, %edi
+	jmp	L(loop)
+
+	CFI_POP	(%esi)
+	CFI_POP	(%ebx)
+
+	.p2align 4
+L(crosscache):
+/* Hancle unaligned string.  */
+	and	$15, %ecx
+	and	$-16, %edi
+	pxor	%xmm3, %xmm3
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm3
+	pcmpeqb	%xmm1, %xmm0
+	/* Find where NULL is.  */
+	pmovmskb %xmm3, %edx
+	/* Check if there is a match.  */
+	pmovmskb %xmm0, %eax
+	/* Remove the leading bytes.  */
+	shr	%cl, %edx
+	shr	%cl, %eax
+	add	$16, %edi
+
+	test	%eax, %eax
+	jnz	L(unaligned_match)
+
+	test	%edx, %edx
+	jnz	L(return_null)
+
+	PUSH	(%esi)
+	PUSH	(%ebx)
+
+	xor	%ebx, %ebx
+	jmp	L(loop)
+
+	CFI_POP	(%esi)
+	CFI_POP	(%ebx)
+
+	.p2align 4
+L(unaligned_match):
+	test	%edx, %edx
+	jnz	L(prolog_find_zero)
+
+	PUSH	(%esi)
+	PUSH	(%ebx)
+
+	mov	%eax, %ebx
+	lea	(%edi, %ecx), %esi
+
+/* Loop start on aligned string.  */
+	.p2align 4
+L(loop):
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %ecx
+	pmovmskb %xmm0, %eax
+	or	%eax, %ecx
+	jnz	L(matches)
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %ecx
+	pmovmskb %xmm0, %eax
+	or	%eax, %ecx
+	jnz	L(matches)
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %ecx
+	pmovmskb %xmm0, %eax
+	or	%eax, %ecx
+	jnz	L(matches)
+
+	movdqa	(%edi), %xmm0
+	pcmpeqb	%xmm0, %xmm2
+	add	$16, %edi
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm2, %ecx
+	pmovmskb %xmm0, %eax
+	or	%eax, %ecx
+	jz	L(loop)
+
+L(matches):
+	test	%eax, %eax
+	jnz	L(match)
+L(return_value):
+	test	%ebx, %ebx
+	jz	L(return_null_1)
+	mov	%ebx, %eax
+	mov	%esi, %edi
+
+	POP	(%ebx)
+	POP	(%esi)
+
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(return_null_1):
+	POP	(%ebx)
+	POP	(%esi)
+
+	xor	%eax, %eax
+	RETURN
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(match):
+	pmovmskb %xmm2, %ecx
+	test	%ecx, %ecx
+	jnz	L(find_zero)
+	mov	%eax, %ebx
+	mov	%edi, %esi
+	jmp	L(loop)
+
+	.p2align 4
+L(find_zero):
+	test	%cl, %cl
+	jz	L(find_zero_high)
+	mov	%cl, %dl
+	and	$15, %dl
+	jz	L(find_zero_8)
+	test	$0x01, %cl
+	jnz	L(FindZeroExit1)
+	test	$0x02, %cl
+	jnz	L(FindZeroExit2)
+	test	$0x04, %cl
+	jnz	L(FindZeroExit3)
+	and	$1 << 4 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(find_zero_8):
+	test	$0x10, %cl
+	jnz	L(FindZeroExit5)
+	test	$0x20, %cl
+	jnz	L(FindZeroExit6)
+	test	$0x40, %cl
+	jnz	L(FindZeroExit7)
+	and	$1 << 8 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(find_zero_high):
+	mov	%ch, %dh
+	and	$15, %dh
+	jz	L(find_zero_high_8)
+	test	$0x01, %ch
+	jnz	L(FindZeroExit9)
+	test	$0x02, %ch
+	jnz	L(FindZeroExit10)
+	test	$0x04, %ch
+	jnz	L(FindZeroExit11)
+	and	$1 << 12 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(find_zero_high_8):
+	test	$0x10, %ch
+	jnz	L(FindZeroExit13)
+	test	$0x20, %ch
+	jnz	L(FindZeroExit14)
+	test	$0x40, %ch
+	jnz	L(FindZeroExit15)
+	and	$1 << 16 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit1):
+	and	$1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit2):
+	and	$1 << 2 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit3):
+	and	$1 << 3 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit5):
+	and	$1 << 5 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit6):
+	and	$1 << 6 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit7):
+	and	$1 << 7 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit9):
+	and	$1 << 9 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit10):
+	and	$1 << 10 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit11):
+	and	$1 << 11 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit13):
+	and	$1 << 13 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit14):
+	and	$1 << 14 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+	jmp	L(match_exit)
+
+	CFI_PUSH	(%ebx)
+	CFI_PUSH	(%esi)
+
+	.p2align 4
+L(FindZeroExit15):
+	and	$1 << 15 - 1, %eax
+	jz	L(return_value)
+
+	POP	(%ebx)
+	POP	(%esi)
+
+	.p2align 4
+L(match_exit):
+	test	%ah, %ah
+	jnz	L(match_exit_high)
+	mov	%al, %dl
+	and	$15 << 4, %dl
+	jnz	L(match_exit_8)
+	test	$0x08, %al
+	jnz	L(Exit4)
+	test	$0x04, %al
+	jnz	L(Exit3)
+	test	$0x02, %al
+	jnz	L(Exit2)
+	lea	-16(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(match_exit_8):
+	test	$0x80, %al
+	jnz	L(Exit8)
+	test	$0x40, %al
+	jnz	L(Exit7)
+	test	$0x20, %al
+	jnz	L(Exit6)
+	lea	-12(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(match_exit_high):
+	mov	%ah, %dh
+	and	$15 << 4, %dh
+	jnz	L(match_exit_high_8)
+	test	$0x08, %ah
+	jnz	L(Exit12)
+	test	$0x04, %ah
+	jnz	L(Exit11)
+	test	$0x02, %ah
+	jnz	L(Exit10)
+	lea	-8(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(match_exit_high_8):
+	test	$0x80, %ah
+	jnz	L(Exit16)
+	test	$0x40, %ah
+	jnz	L(Exit15)
+	test	$0x20, %ah
+	jnz	L(Exit14)
+	lea	-4(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit2):
+	lea	-15(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit3):
+	lea	-14(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit4):
+	lea	-13(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit6):
+	lea	-11(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit7):
+	lea	-10(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit8):
+	lea	-9(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit10):
+	lea	-7(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit11):
+	lea	-6(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit12):
+	lea	-5(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit14):
+	lea	-3(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit15):
+	lea	-2(%edi), %eax
+	RETURN
+
+	.p2align 4
+L(Exit16):
+	lea	-1(%edi), %eax
+	RETURN
+
+/* Return NULL.  */
+	.p2align 4
+L(return_null):
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(prolog_find_zero):
+	add	%ecx, %edi
+	mov     %edx, %ecx
+L(prolog_find_zero_1):
+	test	%cl, %cl
+	jz	L(prolog_find_zero_high)
+	mov	%cl, %dl
+	and	$15, %dl
+	jz	L(prolog_find_zero_8)
+	test	$0x01, %cl
+	jnz	L(PrologFindZeroExit1)
+	test	$0x02, %cl
+	jnz	L(PrologFindZeroExit2)
+	test	$0x04, %cl
+	jnz	L(PrologFindZeroExit3)
+	and	$1 << 4 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(prolog_find_zero_8):
+	test	$0x10, %cl
+	jnz	L(PrologFindZeroExit5)
+	test	$0x20, %cl
+	jnz	L(PrologFindZeroExit6)
+	test	$0x40, %cl
+	jnz	L(PrologFindZeroExit7)
+	and	$1 << 8 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(prolog_find_zero_high):
+	mov	%ch, %dh
+	and	$15, %dh
+	jz	L(prolog_find_zero_high_8)
+	test	$0x01, %ch
+	jnz	L(PrologFindZeroExit9)
+	test	$0x02, %ch
+	jnz	L(PrologFindZeroExit10)
+	test	$0x04, %ch
+	jnz	L(PrologFindZeroExit11)
+	and	$1 << 12 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(prolog_find_zero_high_8):
+	test	$0x10, %ch
+	jnz	L(PrologFindZeroExit13)
+	test	$0x20, %ch
+	jnz	L(PrologFindZeroExit14)
+	test	$0x40, %ch
+	jnz	L(PrologFindZeroExit15)
+	and	$1 << 16 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit1):
+	and	$1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit2):
+	and	$1 << 2 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit3):
+	and	$1 << 3 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit5):
+	and	$1 << 5 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit6):
+	and	$1 << 6 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit7):
+	and	$1 << 7 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit9):
+	and	$1 << 9 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit10):
+	and	$1 << 10 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit11):
+	and	$1 << 11 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit13):
+	and	$1 << 13 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit14):
+	and	$1 << 14 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+	.p2align 4
+L(PrologFindZeroExit15):
+	and	$1 << 15 - 1, %eax
+	jnz	L(match_exit)
+	xor	%eax, %eax
+	RETURN
+
+END (__strrchr_sse2)
+#endif
diff --git a/sysdeps/i386/i686/multiarch/strrchr.S b/sysdeps/i386/i686/multiarch/strrchr.S
new file mode 100644
index 0000000..866cac0
--- /dev/null
+++ b/sysdeps/i386/i686/multiarch/strrchr.S
@@ -0,0 +1,76 @@
+/* Multiple versions of strrchr
+   Copyright (C) 2011 Free Software Foundation, Inc.
+   Contributed by Intel Corporation.
+   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, write to the Free
+   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+   02111-1307 USA.  */
+
+#include <sysdep.h>
+#include <init-arch.h>
+
+#ifndef NOT_IN_libc
+	.section	.gnu.linkonce.t.__i686.get_pc_thunk.bx,"ax",@progbits
+	.globl	__i686.get_pc_thunk.bx
+	.hidden	__i686.get_pc_thunk.bx
+	.p2align 4
+	.type	__i686.get_pc_thunk.bx,@function
+__i686.get_pc_thunk.bx:
+	movl	(%esp), %ebx
+	ret
+
+	.text
+ENTRY(strrchr)
+	.type	strrchr, @gnu_indirect_function
+	pushl	%ebx
+	cfi_adjust_cfa_offset (4)
+	cfi_rel_offset (ebx, 0)
+	call	__i686.get_pc_thunk.bx
+	addl	$_GLOBAL_OFFSET_TABLE_, %ebx
+	cmpl	$0, KIND_OFFSET+__cpu_features@GOTOFF(%ebx)
+	jne	1f
+	call	__init_cpu_features
+1:	leal	__strrchr_ia32@GOTOFF(%ebx), %eax
+	testl	$bit_SSE2, CPUID_OFFSET+index_SSE2+__cpu_features@GOTOFF(%ebx)
+	jz	2f
+	leal	__strrchr_sse2_bsf@GOTOFF(%ebx), %eax
+	testl	$bit_Slow_BSF, FEATURE_OFFSET+index_Slow_BSF+__cpu_features@GOTOFF(%ebx)
+	jz	2f
+	leal	__strrchr_sse2@GOTOFF(%ebx), %eax
+2:	popl	%ebx
+	cfi_adjust_cfa_offset (-4);
+	cfi_restore (ebx)
+	ret
+END(strrchr)
+
+# undef ENTRY
+# define ENTRY(name) \
+	.type __strrchr_ia32, @function; \
+	.globl __strrchr_ia32; \
+	.p2align 4; \
+	__strrchr_ia32: cfi_startproc; \
+	CALL_MCOUNT
+# undef END
+# define END(name) \
+	cfi_endproc; .size __strrchr_ia32, .-__strrchr_ia32
+# undef libc_hidden_builtin_def
+/* IFUNC doesn't work with the hidden functions in shared library since
+   they will be called without setting up EBX needed for PLT which is
+   used by IFUNC.  */
+# define libc_hidden_builtin_def(name) \
+	.globl __GI_strrchr; __GI_strrchr = __strrchr_ia32
+#endif
+
+#include "../../strrchr.S"