x86: Remove str{p}{n}cpy-ssse3

With SSE2, SSE4.1, AVX2, and EVEX versions very few targets prefer
SSSE3. As a result it is no longer worth it to keep the SSSE3
versions given the code size cost.
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>

6 files changed, 0 insertions, 3572 deletions

diff --git a/sysdeps/x86_64/multiarch/Makefile b/sysdeps/x86_64/multiarch/Makefile
index 2b3c625ea2..5b02ec8de5 100644
--- a/sysdeps/x86_64/multiarch/Makefile
+++ b/sysdeps/x86_64/multiarch/Makefile
@@ -46,13 +46,11 @@ sysdep_routines += \
   stpcpy-evex \
   stpcpy-sse2 \
   stpcpy-sse2-unaligned \
-  stpcpy-ssse3 \
   stpncpy-avx2 \
   stpncpy-avx2-rtm \
   stpncpy-c \
   stpncpy-evex \
   stpncpy-sse2-unaligned \
-  stpncpy-ssse3 \
   strcasecmp_l-avx2 \
   strcasecmp_l-avx2-rtm \
   strcasecmp_l-evex \
@@ -83,7 +81,6 @@ sysdep_routines += \
   strcpy-evex \
   strcpy-sse2 \
   strcpy-sse2-unaligned \
-  strcpy-ssse3 \
   strcspn-c \
   strcspn-sse2 \
   strlen-avx2 \
@@ -110,7 +107,6 @@ sysdep_routines += \
   strncpy-c \
   strncpy-evex \
   strncpy-sse2-unaligned \
-  strncpy-ssse3 \
   strnlen-avx2 \
   strnlen-avx2-rtm \
   strnlen-evex \
diff --git a/sysdeps/x86_64/multiarch/ifunc-impl-list.c b/sysdeps/x86_64/multiarch/ifunc-impl-list.c
index 41a04621ad..49ce6860d0 100644
--- a/sysdeps/x86_64/multiarch/ifunc-impl-list.c
+++ b/sysdeps/x86_64/multiarch/ifunc-impl-list.c
@@ -399,8 +399,6 @@ __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 
   /* Support sysdeps/x86_64/multiarch/stpncpy.c.  */
   IFUNC_IMPL (i, name, stpncpy,
-	      IFUNC_IMPL_ADD (array, i, stpncpy, CPU_FEATURE_USABLE (SSSE3),
-			      __stpncpy_ssse3)
 	      IFUNC_IMPL_ADD (array, i, stpncpy, CPU_FEATURE_USABLE (AVX2),
 			      __stpncpy_avx2)
 	      IFUNC_IMPL_ADD (array, i, stpncpy,
@@ -417,8 +415,6 @@ __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 
   /* Support sysdeps/x86_64/multiarch/stpcpy.c.  */
   IFUNC_IMPL (i, name, stpcpy,
-	      IFUNC_IMPL_ADD (array, i, stpcpy, CPU_FEATURE_USABLE (SSSE3),
-			      __stpcpy_ssse3)
 	      IFUNC_IMPL_ADD (array, i, stpcpy, CPU_FEATURE_USABLE (AVX2),
 			      __stpcpy_avx2)
 	      IFUNC_IMPL_ADD (array, i, stpcpy,
@@ -567,8 +563,6 @@ __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 			      (CPU_FEATURE_USABLE (AVX512VL)
 			       && CPU_FEATURE_USABLE (AVX512BW)),
 			      __strcpy_evex)
-	      IFUNC_IMPL_ADD (array, i, strcpy, CPU_FEATURE_USABLE (SSSE3),
-			      __strcpy_ssse3)
 	      IFUNC_IMPL_ADD (array, i, strcpy, 1, __strcpy_sse2_unaligned)
 	      IFUNC_IMPL_ADD (array, i, strcpy, 1, __strcpy_sse2))
 
@@ -644,8 +638,6 @@ __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 			      (CPU_FEATURE_USABLE (AVX512VL)
 			       && CPU_FEATURE_USABLE (AVX512BW)),
 			      __strncpy_evex)
-	      IFUNC_IMPL_ADD (array, i, strncpy, CPU_FEATURE_USABLE (SSSE3),
-			      __strncpy_ssse3)
 	      IFUNC_IMPL_ADD (array, i, strncpy, 1,
 			      __strncpy_sse2_unaligned)
 	      IFUNC_IMPL_ADD (array, i, strncpy, 1, __strncpy_sse2))
diff --git a/sysdeps/x86_64/multiarch/stpcpy-ssse3.S b/sysdeps/x86_64/multiarch/stpcpy-ssse3.S
deleted file mode 100644
index d971c2da38..0000000000
--- a/sysdeps/x86_64/multiarch/stpcpy-ssse3.S
+++ /dev/null
@@ -1,3 +0,0 @@
-#define USE_AS_STPCPY
-#define STRCPY __stpcpy_ssse3
-#include "strcpy-ssse3.S"
diff --git a/sysdeps/x86_64/multiarch/stpncpy-ssse3.S b/sysdeps/x86_64/multiarch/stpncpy-ssse3.S
deleted file mode 100644
index 14ed16f6b5..0000000000
--- a/sysdeps/x86_64/multiarch/stpncpy-ssse3.S
+++ /dev/null
@@ -1,4 +0,0 @@
-#define USE_AS_STPCPY
-#define USE_AS_STRNCPY
-#define STRCPY __stpncpy_ssse3
-#include "strcpy-ssse3.S"
diff --git a/sysdeps/x86_64/multiarch/strcpy-ssse3.S b/sysdeps/x86_64/multiarch/strcpy-ssse3.S
deleted file mode 100644
index f617a535cf..0000000000
--- a/sysdeps/x86_64/multiarch/strcpy-ssse3.S
+++ /dev/null
@@ -1,3550 +0,0 @@
-/* strcpy with SSSE3
-   Copyright (C) 2011-2022 Free Software Foundation, Inc.
-   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, see
-   <https://www.gnu.org/licenses/>.  */
-
-#if IS_IN (libc)
-
-# ifndef USE_AS_STRCAT
-#  include <sysdep.h>
-
-#  ifndef STRCPY
-#   define STRCPY  __strcpy_ssse3
-#  endif
-
-	.section .text.ssse3,"ax",@progbits
-ENTRY (STRCPY)
-
-	mov	%rsi, %rcx
-#  ifdef USE_AS_STRNCPY
-	mov	%RDX_LP, %R8_LP
-#  endif
-	mov	%rdi, %rdx
-#  ifdef USE_AS_STRNCPY
-	test	%R8_LP, %R8_LP
-	jz	L(Exit0)
-	cmp	$8, %R8_LP
-	jbe	L(StrncpyExit8Bytes)
-# endif
-	cmpb	$0, (%rcx)
-	jz	L(Exit1)
-	cmpb	$0, 1(%rcx)
-	jz	L(Exit2)
-	cmpb	$0, 2(%rcx)
-	jz	L(Exit3)
-	cmpb	$0, 3(%rcx)
-	jz	L(Exit4)
-	cmpb	$0, 4(%rcx)
-	jz	L(Exit5)
-	cmpb	$0, 5(%rcx)
-	jz	L(Exit6)
-	cmpb	$0, 6(%rcx)
-	jz	L(Exit7)
-	cmpb	$0, 7(%rcx)
-	jz	L(Exit8)
-# ifdef USE_AS_STRNCPY
-	cmp	$16, %r8
-	jb	L(StrncpyExit15Bytes)
-# endif
-	cmpb	$0, 8(%rcx)
-	jz	L(Exit9)
-	cmpb	$0, 9(%rcx)
-	jz	L(Exit10)
-	cmpb	$0, 10(%rcx)
-	jz	L(Exit11)
-	cmpb	$0, 11(%rcx)
-	jz	L(Exit12)
-	cmpb	$0, 12(%rcx)
-	jz	L(Exit13)
-	cmpb	$0, 13(%rcx)
-	jz	L(Exit14)
-	cmpb	$0, 14(%rcx)
-	jz	L(Exit15)
-# ifdef USE_AS_STRNCPY
-	cmp	$16, %r8
-	je	L(Exit16)
-# endif
-	cmpb	$0, 15(%rcx)
-	jz	L(Exit16)
-# endif
-
-# ifdef USE_AS_STRNCPY
-	mov	%rcx, %rsi
-	sub	$16, %r8
-	and	$0xf, %rsi
-
-/* add 16 bytes rcx_offset to r8 */
-
-	add	%rsi, %r8
-# endif
-	lea	16(%rcx), %rsi
-	and	$-16, %rsi
-	pxor	%xmm0, %xmm0
-	mov	(%rcx), %r9
-	mov	%r9, (%rdx)
-	pcmpeqb	(%rsi), %xmm0
-	mov	8(%rcx), %r9
-	mov	%r9, 8(%rdx)
-
-/* convert byte mask in xmm0 to bit mask */
-
-	pmovmskb %xmm0, %rax
-	sub	%rcx, %rsi
-
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(CopyFrom1To16BytesCase2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(CopyFrom1To16Bytes)
-
-	mov	%rdx, %rax
-	lea	16(%rdx), %rdx
-	and	$-16, %rdx
-	sub	%rdx, %rax
-
-# ifdef USE_AS_STRNCPY
-	add	%rax, %rsi
-	lea	-1(%rsi), %rsi
-	and	$1<<31, %esi
-	test	%rsi, %rsi
-	jnz	L(ContinueCopy)
-	lea	16(%r8), %r8
-
-L(ContinueCopy):
-# endif
-	sub	%rax, %rcx
-	mov	%rcx, %rax
-	and	$0xf, %rax
-	mov	$0, %rsi
-
-/* case: rcx_offset == rdx_offset */
-
-	jz	L(Align16Both)
-
-	cmp	$8, %rax
-	jae	L(ShlHigh8)
-	cmp	$1, %rax
-	je	L(Shl1)
-	cmp	$2, %rax
-	je	L(Shl2)
-	cmp	$3, %rax
-	je	L(Shl3)
-	cmp	$4, %rax
-	je	L(Shl4)
-	cmp	$5, %rax
-	je	L(Shl5)
-	cmp	$6, %rax
-	je	L(Shl6)
-	jmp	L(Shl7)
-
-L(ShlHigh8):
-	je	L(Shl8)
-	cmp	$9, %rax
-	je	L(Shl9)
-	cmp	$10, %rax
-	je	L(Shl10)
-	cmp	$11, %rax
-	je	L(Shl11)
-	cmp	$12, %rax
-	je	L(Shl12)
-	cmp	$13, %rax
-	je	L(Shl13)
-	cmp	$14, %rax
-	je	L(Shl14)
-	jmp	L(Shl15)
-
-L(Align16Both):
-	movaps	(%rcx), %xmm1
-	movaps	16(%rcx), %xmm2
-	movaps	%xmm1, (%rdx)
-	pcmpeqb	%xmm2, %xmm0
-	pmovmskb %xmm0, %rax
-	lea	16(%rsi), %rsi
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(CopyFrom1To16BytesCase2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(CopyFrom1To16Bytes)
-
-	movaps	16(%rcx, %rsi), %xmm3
-	movaps	%xmm2, (%rdx, %rsi)
-	pcmpeqb	%xmm3, %xmm0
-	pmovmskb %xmm0, %rax
-	lea	16(%rsi), %rsi
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(CopyFrom1To16BytesCase2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(CopyFrom1To16Bytes)
-
-	movaps	16(%rcx, %rsi), %xmm4
-	movaps	%xmm3, (%rdx, %rsi)
-	pcmpeqb	%xmm4, %xmm0
-	pmovmskb %xmm0, %rax
-	lea	16(%rsi), %rsi
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(CopyFrom1To16BytesCase2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(CopyFrom1To16Bytes)
-
-	movaps	16(%rcx, %rsi), %xmm1
-	movaps	%xmm4, (%rdx, %rsi)
-	pcmpeqb	%xmm1, %xmm0
-	pmovmskb %xmm0, %rax
-	lea	16(%rsi), %rsi
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(CopyFrom1To16BytesCase2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(CopyFrom1To16Bytes)
-
-	movaps	16(%rcx, %rsi), %xmm2
-	movaps	%xmm1, (%rdx, %rsi)
-	pcmpeqb	%xmm2, %xmm0
-	pmovmskb %xmm0, %rax
-	lea	16(%rsi), %rsi
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(CopyFrom1To16BytesCase2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(CopyFrom1To16Bytes)
-
-	movaps	16(%rcx, %rsi), %xmm3
-	movaps	%xmm2, (%rdx, %rsi)
-	pcmpeqb	%xmm3, %xmm0
-	pmovmskb %xmm0, %rax
-	lea	16(%rsi), %rsi
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(CopyFrom1To16BytesCase2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(CopyFrom1To16Bytes)
-
-	movaps	%xmm3, (%rdx, %rsi)
-	mov	%rcx, %rax
-	lea	16(%rcx, %rsi), %rcx
-	and	$-0x40, %rcx
-	sub	%rcx, %rax
-	sub	%rax, %rdx
-# ifdef USE_AS_STRNCPY
-	lea	112(%r8, %rax), %r8
-# endif
-	mov	$-0x40, %rsi
-
-	.p2align 4
-L(Aligned64Loop):
-	movaps	(%rcx), %xmm2
-	movaps	%xmm2, %xmm4
-	movaps	16(%rcx), %xmm5
-	movaps	32(%rcx), %xmm3
-	movaps	%xmm3, %xmm6
-	movaps	48(%rcx), %xmm7
-	pminub	%xmm5, %xmm2
-	pminub	%xmm7, %xmm3
-	pminub	%xmm2, %xmm3
-	pcmpeqb	%xmm0, %xmm3
-	pmovmskb %xmm3, %rax
-	lea	64(%rdx), %rdx
-	lea	64(%rcx), %rcx
-# ifdef USE_AS_STRNCPY
-	sub	$64, %r8
-	jbe	L(StrncpyLeaveCase2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Aligned64Leave)
-	movaps	%xmm4, -64(%rdx)
-	movaps	%xmm5, -48(%rdx)
-	movaps	%xmm6, -32(%rdx)
-	movaps	%xmm7, -16(%rdx)
-	jmp	L(Aligned64Loop)
-
-L(Aligned64Leave):
-# ifdef USE_AS_STRNCPY
-	lea	48(%r8), %r8
-# endif
-	pcmpeqb	%xmm4, %xmm0
-	pmovmskb %xmm0, %rax
-	test	%rax, %rax
-	jnz	L(CopyFrom1To16Bytes)
-
-	pcmpeqb	%xmm5, %xmm0
-# ifdef USE_AS_STRNCPY
-	lea	-16(%r8), %r8
-# endif
-	pmovmskb %xmm0, %rax
-	movaps	%xmm4, -64(%rdx)
-	test	%rax, %rax
-	lea	16(%rsi), %rsi
-	jnz	L(CopyFrom1To16Bytes)
-
-	pcmpeqb	%xmm6, %xmm0
-# ifdef USE_AS_STRNCPY
-	lea	-16(%r8), %r8
-# endif
-	pmovmskb %xmm0, %rax
-	movaps	%xmm5, -48(%rdx)
-	test	%rax, %rax
-	lea	16(%rsi), %rsi
-	jnz	L(CopyFrom1To16Bytes)
-
-	movaps	%xmm6, -32(%rdx)
-	pcmpeqb	%xmm7, %xmm0
-# ifdef USE_AS_STRNCPY
-	lea	-16(%r8), %r8
-# endif
-	pmovmskb %xmm0, %rax
-	lea	16(%rsi), %rsi
-	jmp	L(CopyFrom1To16Bytes)
-
-	.p2align 4
-L(Shl1):
-	movaps	-1(%rcx), %xmm1
-	movaps	15(%rcx), %xmm2
-L(Shl1Start):
-	pcmpeqb	%xmm2, %xmm0
-	pmovmskb %xmm0, %rax
-	movaps	%xmm2, %xmm3
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit1Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl1LoopExit)
-
-	palignr	$1, %xmm1, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	31(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-	movaps	%xmm2, %xmm1
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit1Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl1LoopExit)
-
-	palignr	$1, %xmm3, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	31(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-	movaps	%xmm2, %xmm3
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit1Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl1LoopExit)
-
-	palignr	$1, %xmm1, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	31(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit1Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl1LoopExit)
-
-	palignr	$1, %xmm3, %xmm2
-	movaps	%xmm2, (%rdx)
-	lea	31(%rcx), %rcx
-	lea	16(%rdx), %rdx
-
-	mov	%rcx, %rax
-	and	$-0x40, %rcx
-	sub	%rcx, %rax
-	lea	-15(%rcx), %rcx
-	sub	%rax, %rdx
-# ifdef USE_AS_STRNCPY
-	add	%rax, %r8
-# endif
-	movaps	-1(%rcx), %xmm1
-
-/* 64 bytes loop */
-	.p2align 4
-L(Shl1LoopStart):
-	movaps	15(%rcx), %xmm2
-	movaps	31(%rcx), %xmm3
-	movaps	%xmm3, %xmm6
-	movaps	47(%rcx), %xmm4
-	movaps	%xmm4, %xmm7
-	movaps	63(%rcx), %xmm5
-	pminub	%xmm2, %xmm6
-	pminub	%xmm5, %xmm7
-	pminub	%xmm6, %xmm7
-	pcmpeqb	%xmm0, %xmm7
-	pmovmskb %xmm7, %rax
-	movaps	%xmm5, %xmm7
-	palignr	$1, %xmm4, %xmm5
-	test	%rax, %rax
-	palignr	$1, %xmm3, %xmm4
-	jnz	L(Shl1Start)
-# ifdef USE_AS_STRNCPY
-	sub	$64, %r8
-	jbe	L(StrncpyLeave1)
-# endif
-	palignr	$1, %xmm2, %xmm3
-	lea	64(%rcx), %rcx
-	palignr	$1, %xmm1, %xmm2
-	movaps	%xmm7, %xmm1
-	movaps	%xmm5, 48(%rdx)
-	movaps	%xmm4, 32(%rdx)
-	movaps	%xmm3, 16(%rdx)
-	movaps	%xmm2, (%rdx)
-	lea	64(%rdx), %rdx
-	jmp	L(Shl1LoopStart)
-
-L(Shl1LoopExit):
-	movdqu	-1(%rcx), %xmm1
-	mov	$15, %rsi
-	movdqu	%xmm1, -1(%rdx)
-	jmp	L(CopyFrom1To16Bytes)
-
-	.p2align 4
-L(Shl2):
-	movaps	-2(%rcx), %xmm1
-	movaps	14(%rcx), %xmm2
-L(Shl2Start):
-	pcmpeqb	%xmm2, %xmm0
-	pmovmskb %xmm0, %rax
-	movaps	%xmm2, %xmm3
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit2Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl2LoopExit)
-
-	palignr	$2, %xmm1, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	30(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-	movaps	%xmm2, %xmm1
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit2Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl2LoopExit)
-
-	palignr	$2, %xmm3, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	30(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-	movaps	%xmm2, %xmm3
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit2Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl2LoopExit)
-
-	palignr	$2, %xmm1, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	30(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit2Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl2LoopExit)
-
-	palignr	$2, %xmm3, %xmm2
-	movaps	%xmm2, (%rdx)
-	lea	30(%rcx), %rcx
-	lea	16(%rdx), %rdx
-
-	mov	%rcx, %rax
-	and	$-0x40, %rcx
-	sub	%rcx, %rax
-	lea	-14(%rcx), %rcx
-	sub	%rax, %rdx
-# ifdef USE_AS_STRNCPY
-	add	%rax, %r8
-# endif
-	movaps	-2(%rcx), %xmm1
-
-/* 64 bytes loop */
-	.p2align 4
-L(Shl2LoopStart):
-	movaps	14(%rcx), %xmm2
-	movaps	30(%rcx), %xmm3
-	movaps	%xmm3, %xmm6
-	movaps	46(%rcx), %xmm4
-	movaps	%xmm4, %xmm7
-	movaps	62(%rcx), %xmm5
-	pminub	%xmm2, %xmm6
-	pminub	%xmm5, %xmm7
-	pminub	%xmm6, %xmm7
-	pcmpeqb	%xmm0, %xmm7
-	pmovmskb %xmm7, %rax
-	movaps	%xmm5, %xmm7
-	palignr	$2, %xmm4, %xmm5
-	test	%rax, %rax
-	palignr	$2, %xmm3, %xmm4
-	jnz	L(Shl2Start)
-# ifdef USE_AS_STRNCPY
-	sub	$64, %r8
-	jbe	L(StrncpyLeave2)
-# endif
-	palignr	$2, %xmm2, %xmm3
-	lea	64(%rcx), %rcx
-	palignr	$2, %xmm1, %xmm2
-	movaps	%xmm7, %xmm1
-	movaps	%xmm5, 48(%rdx)
-	movaps	%xmm4, 32(%rdx)
-	movaps	%xmm3, 16(%rdx)
-	movaps	%xmm2, (%rdx)
-	lea	64(%rdx), %rdx
-	jmp	L(Shl2LoopStart)
-
-L(Shl2LoopExit):
-	movdqu	-2(%rcx), %xmm1
-	mov	$14, %rsi
-	movdqu	%xmm1, -2(%rdx)
-	jmp	L(CopyFrom1To16Bytes)
-
-	.p2align 4
-L(Shl3):
-	movaps	-3(%rcx), %xmm1
-	movaps	13(%rcx), %xmm2
-L(Shl3Start):
-	pcmpeqb	%xmm2, %xmm0
-	pmovmskb %xmm0, %rax
-	movaps	%xmm2, %xmm3
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit3Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl3LoopExit)
-
-	palignr	$3, %xmm1, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	29(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-	movaps	%xmm2, %xmm1
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit3Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl3LoopExit)
-
-	palignr	$3, %xmm3, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	29(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-	movaps	%xmm2, %xmm3
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit3Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl3LoopExit)
-
-	palignr	$3, %xmm1, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	29(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit3Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl3LoopExit)
-
-	palignr	$3, %xmm3, %xmm2
-	movaps	%xmm2, (%rdx)
-	lea	29(%rcx), %rcx
-	lea	16(%rdx), %rdx
-
-	mov	%rcx, %rax
-	and	$-0x40, %rcx
-	sub	%rcx, %rax
-	lea	-13(%rcx), %rcx
-	sub	%rax, %rdx
-# ifdef USE_AS_STRNCPY
-	add	%rax, %r8
-# endif
-	movaps	-3(%rcx), %xmm1
-
-/* 64 bytes loop */
-	.p2align 4
-L(Shl3LoopStart):
-	movaps	13(%rcx), %xmm2
-	movaps	29(%rcx), %xmm3
-	movaps	%xmm3, %xmm6
-	movaps	45(%rcx), %xmm4
-	movaps	%xmm4, %xmm7
-	movaps	61(%rcx), %xmm5
-	pminub	%xmm2, %xmm6
-	pminub	%xmm5, %xmm7
-	pminub	%xmm6, %xmm7
-	pcmpeqb	%xmm0, %xmm7
-	pmovmskb %xmm7, %rax
-	movaps	%xmm5, %xmm7
-	palignr	$3, %xmm4, %xmm5
-	test	%rax, %rax
-	palignr	$3, %xmm3, %xmm4
-	jnz	L(Shl3Start)
-# ifdef USE_AS_STRNCPY
-	sub	$64, %r8
-	jbe	L(StrncpyLeave3)
-# endif
-	palignr	$3, %xmm2, %xmm3
-	lea	64(%rcx), %rcx
-	palignr	$3, %xmm1, %xmm2
-	movaps	%xmm7, %xmm1
-	movaps	%xmm5, 48(%rdx)
-	movaps	%xmm4, 32(%rdx)
-	movaps	%xmm3, 16(%rdx)
-	movaps	%xmm2, (%rdx)
-	lea	64(%rdx), %rdx
-	jmp	L(Shl3LoopStart)
-
-L(Shl3LoopExit):
-	movdqu	-3(%rcx), %xmm1
-	mov	$13, %rsi
-	movdqu	%xmm1, -3(%rdx)
-	jmp	L(CopyFrom1To16Bytes)
-
-	.p2align 4
-L(Shl4):
-	movaps	-4(%rcx), %xmm1
-	movaps	12(%rcx), %xmm2
-L(Shl4Start):
-	pcmpeqb	%xmm2, %xmm0
-	pmovmskb %xmm0, %rax
-	movaps	%xmm2, %xmm3
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit4Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl4LoopExit)
-
-	palignr	$4, %xmm1, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	28(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-	movaps	%xmm2, %xmm1
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit4Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl4LoopExit)
-
-	palignr	$4, %xmm3, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	28(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-	movaps	%xmm2, %xmm3
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit4Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl4LoopExit)
-
-	palignr	$4, %xmm1, %xmm2
-	movaps	%xmm2, (%rdx)
-	movaps	28(%rcx), %xmm2
-
-	pcmpeqb	%xmm2, %xmm0
-	lea	16(%rdx), %rdx
-	pmovmskb %xmm0, %rax
-	lea	16(%rcx), %rcx
-# ifdef USE_AS_STRNCPY
-	sub	$16, %r8
-	jbe	L(StrncpyExit4Case2OrCase3)
-# endif
-	test	%rax, %rax
-	jnz	L(Shl4LoopExit)
-
-	palignr	$4, %xmm3, %xmm2
-	movaps	%xmm2, (%rdx)
-	lea	28(%rcx), %rcx
-	lea	16(%rdx), %rdx
-