2014-06-19 20:00:56 +00:00
|
|
|
/* Multiple versions of memset
|
|
|
|
All versions must be listed in ifunc-impl-list.c.
|
2017-01-01 00:14:16 +00:00
|
|
|
Copyright (C) 2014-2017 Free Software Foundation, Inc.
|
2014-06-19 20:00:56 +00:00
|
|
|
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
|
|
|
|
<http://www.gnu.org/licenses/>. */
|
|
|
|
|
|
|
|
#include <sysdep.h>
|
|
|
|
#include <shlib-compat.h>
|
|
|
|
#include <init-arch.h>
|
|
|
|
|
|
|
|
/* Define multiple versions only for the definition in lib. */
|
2015-10-28 13:29:03 +00:00
|
|
|
#if IS_IN (libc)
|
2014-06-19 20:00:56 +00:00
|
|
|
ENTRY(memset)
|
|
|
|
.type memset, @gnu_indirect_function
|
2015-08-13 10:38:47 +00:00
|
|
|
LOAD_RTLD_GLOBAL_RO_RDX
|
2016-06-30 14:57:07 +00:00
|
|
|
lea __memset_erms(%rip), %RAX_LP
|
|
|
|
HAS_ARCH_FEATURE (Prefer_ERMS)
|
|
|
|
jnz 2f
|
2016-06-08 20:55:45 +00:00
|
|
|
lea __memset_sse2_unaligned_erms(%rip), %RAX_LP
|
|
|
|
HAS_CPU_FEATURE (ERMS)
|
|
|
|
jnz 1f
|
|
|
|
lea __memset_sse2_unaligned(%rip), %RAX_LP
|
|
|
|
1:
|
2015-08-13 10:38:47 +00:00
|
|
|
HAS_ARCH_FEATURE (AVX2_Usable)
|
2014-06-19 20:00:56 +00:00
|
|
|
jz 2f
|
2016-06-08 20:55:45 +00:00
|
|
|
lea __memset_avx2_unaligned_erms(%rip), %RAX_LP
|
|
|
|
HAS_CPU_FEATURE (ERMS)
|
|
|
|
jnz L(AVX512F)
|
|
|
|
lea __memset_avx2_unaligned(%rip), %RAX_LP
|
|
|
|
L(AVX512F):
|
2017-04-18 21:01:45 +00:00
|
|
|
HAS_ARCH_FEATURE (Prefer_No_AVX512)
|
|
|
|
jnz 2f
|
2015-12-18 23:47:28 +00:00
|
|
|
HAS_ARCH_FEATURE (AVX512F_Usable)
|
|
|
|
jz 2f
|
2016-06-08 20:55:45 +00:00
|
|
|
lea __memset_avx512_no_vzeroupper(%rip), %RAX_LP
|
2015-12-18 23:47:28 +00:00
|
|
|
HAS_ARCH_FEATURE (Prefer_No_VZEROUPPER)
|
2016-06-08 20:55:45 +00:00
|
|
|
jnz 2f
|
|
|
|
lea __memset_avx512_unaligned_erms(%rip), %RAX_LP
|
|
|
|
HAS_CPU_FEATURE (ERMS)
|
|
|
|
jnz 2f
|
|
|
|
lea __memset_avx512_unaligned(%rip), %RAX_LP
|
2014-06-19 20:00:56 +00:00
|
|
|
2: ret
|
|
|
|
END(memset)
|
2015-10-28 13:29:03 +00:00
|
|
|
#endif
|
2014-06-19 20:00:56 +00:00
|
|
|
|
2015-10-28 13:29:03 +00:00
|
|
|
#if IS_IN (libc)
|
2016-06-08 20:55:45 +00:00
|
|
|
# define MEMSET_SYMBOL(p,s) p##_sse2_##s
|
x86-64: Optimize wmemset with SSE2/AVX2/AVX512
The difference between memset and wmemset is byte vs int. Add stubs
to SSE2/AVX2/AVX512 memset for wmemset with updated constant and size:
SSE2 wmemset:
shl $0x2,%rdx
movd %esi,%xmm0
mov %rdi,%rax
pshufd $0x0,%xmm0,%xmm0
jmp entry_from_wmemset
SSE2 memset:
movd %esi,%xmm0
mov %rdi,%rax
punpcklbw %xmm0,%xmm0
punpcklwd %xmm0,%xmm0
pshufd $0x0,%xmm0,%xmm0
entry_from_wmemset:
Since the ERMS versions of wmemset requires "rep stosl" instead of
"rep stosb", only the vector store stubs of SSE2/AVX2/AVX512 wmemset
are added. The SSE2 wmemset is about 3X faster and the AVX2 wmemset
is about 6X faster on Haswell.
* include/wchar.h (__wmemset_chk): New.
* sysdeps/x86_64/memset.S (VDUP_TO_VEC0_AND_SET_RETURN): Renamed
to MEMSET_VDUP_TO_VEC0_AND_SET_RETURN.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_CHK_SYMBOL): Likewise.
(WMEMSET_SYMBOL): Likewise.
(__wmemset): Add hidden definition.
(wmemset): Add weak hidden definition.
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
wmemset_chk-nonshared.
* sysdeps/x86_64/multiarch/ifunc-impl-list.c
(__libc_ifunc_impl_list): Add __wmemset_sse2_unaligned,
__wmemset_avx2_unaligned, __wmemset_avx512_unaligned,
__wmemset_chk_sse2_unaligned, __wmemset_chk_avx2_unaligned
and __wmemset_chk_avx512_unaligned.
* sysdeps/x86_64/multiarch/memset-avx2-unaligned-erms.S
(VDUP_TO_VEC0_AND_SET_RETURN): Renamed to ...
(MEMSET_VDUP_TO_VEC0_AND_SET_RETURN): This.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_SYMBOL): Likewise.
* sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S
(VDUP_TO_VEC0_AND_SET_RETURN): Renamed to ...
(MEMSET_VDUP_TO_VEC0_AND_SET_RETURN): This.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_SYMBOL): Likewise.
* sysdeps/x86_64/multiarch/memset-vec-unaligned-erms.S: Updated.
(WMEMSET_CHK_SYMBOL): New.
(WMEMSET_CHK_SYMBOL (__wmemset_chk, unaligned)): Likewise.
(WMEMSET_SYMBOL (__wmemset, unaligned)): Likewise.
* sysdeps/x86_64/multiarch/memset.S (WMEMSET_SYMBOL): New.
(libc_hidden_builtin_def): Also define __GI_wmemset and
__GI___wmemset.
(weak_alias): New.
* sysdeps/x86_64/multiarch/wmemset.c: New file.
* sysdeps/x86_64/multiarch/wmemset.h: Likewise.
* sysdeps/x86_64/multiarch/wmemset_chk-nonshared.S: Likewise.
* sysdeps/x86_64/multiarch/wmemset_chk.c: Likewise.
* sysdeps/x86_64/wmemset.c: Likewise.
* sysdeps/x86_64/wmemset_chk.c: Likewise.
2017-06-05 18:09:48 +00:00
|
|
|
# define WMEMSET_SYMBOL(p,s) p##_sse2_##s
|
2014-06-19 20:00:56 +00:00
|
|
|
|
2015-10-28 13:29:03 +00:00
|
|
|
# ifdef SHARED
|
x86-64: Optimize wmemset with SSE2/AVX2/AVX512
The difference between memset and wmemset is byte vs int. Add stubs
to SSE2/AVX2/AVX512 memset for wmemset with updated constant and size:
SSE2 wmemset:
shl $0x2,%rdx
movd %esi,%xmm0
mov %rdi,%rax
pshufd $0x0,%xmm0,%xmm0
jmp entry_from_wmemset
SSE2 memset:
movd %esi,%xmm0
mov %rdi,%rax
punpcklbw %xmm0,%xmm0
punpcklwd %xmm0,%xmm0
pshufd $0x0,%xmm0,%xmm0
entry_from_wmemset:
Since the ERMS versions of wmemset requires "rep stosl" instead of
"rep stosb", only the vector store stubs of SSE2/AVX2/AVX512 wmemset
are added. The SSE2 wmemset is about 3X faster and the AVX2 wmemset
is about 6X faster on Haswell.
* include/wchar.h (__wmemset_chk): New.
* sysdeps/x86_64/memset.S (VDUP_TO_VEC0_AND_SET_RETURN): Renamed
to MEMSET_VDUP_TO_VEC0_AND_SET_RETURN.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_CHK_SYMBOL): Likewise.
(WMEMSET_SYMBOL): Likewise.
(__wmemset): Add hidden definition.
(wmemset): Add weak hidden definition.
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
wmemset_chk-nonshared.
* sysdeps/x86_64/multiarch/ifunc-impl-list.c
(__libc_ifunc_impl_list): Add __wmemset_sse2_unaligned,
__wmemset_avx2_unaligned, __wmemset_avx512_unaligned,
__wmemset_chk_sse2_unaligned, __wmemset_chk_avx2_unaligned
and __wmemset_chk_avx512_unaligned.
* sysdeps/x86_64/multiarch/memset-avx2-unaligned-erms.S
(VDUP_TO_VEC0_AND_SET_RETURN): Renamed to ...
(MEMSET_VDUP_TO_VEC0_AND_SET_RETURN): This.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_SYMBOL): Likewise.
* sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S
(VDUP_TO_VEC0_AND_SET_RETURN): Renamed to ...
(MEMSET_VDUP_TO_VEC0_AND_SET_RETURN): This.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_SYMBOL): Likewise.
* sysdeps/x86_64/multiarch/memset-vec-unaligned-erms.S: Updated.
(WMEMSET_CHK_SYMBOL): New.
(WMEMSET_CHK_SYMBOL (__wmemset_chk, unaligned)): Likewise.
(WMEMSET_SYMBOL (__wmemset, unaligned)): Likewise.
* sysdeps/x86_64/multiarch/memset.S (WMEMSET_SYMBOL): New.
(libc_hidden_builtin_def): Also define __GI_wmemset and
__GI___wmemset.
(weak_alias): New.
* sysdeps/x86_64/multiarch/wmemset.c: New file.
* sysdeps/x86_64/multiarch/wmemset.h: Likewise.
* sysdeps/x86_64/multiarch/wmemset_chk-nonshared.S: Likewise.
* sysdeps/x86_64/multiarch/wmemset_chk.c: Likewise.
* sysdeps/x86_64/wmemset.c: Likewise.
* sysdeps/x86_64/wmemset_chk.c: Likewise.
2017-06-05 18:09:48 +00:00
|
|
|
# undef libc_hidden_builtin_def
|
2014-06-19 20:00:56 +00:00
|
|
|
/* It doesn't make sense to send libc-internal memset calls through a PLT.
|
2016-06-08 20:55:45 +00:00
|
|
|
The speedup we get from using SSE2 instructions is likely eaten away
|
2014-06-19 20:00:56 +00:00
|
|
|
by the indirect call in the PLT. */
|
x86-64: Optimize wmemset with SSE2/AVX2/AVX512
The difference between memset and wmemset is byte vs int. Add stubs
to SSE2/AVX2/AVX512 memset for wmemset with updated constant and size:
SSE2 wmemset:
shl $0x2,%rdx
movd %esi,%xmm0
mov %rdi,%rax
pshufd $0x0,%xmm0,%xmm0
jmp entry_from_wmemset
SSE2 memset:
movd %esi,%xmm0
mov %rdi,%rax
punpcklbw %xmm0,%xmm0
punpcklwd %xmm0,%xmm0
pshufd $0x0,%xmm0,%xmm0
entry_from_wmemset:
Since the ERMS versions of wmemset requires "rep stosl" instead of
"rep stosb", only the vector store stubs of SSE2/AVX2/AVX512 wmemset
are added. The SSE2 wmemset is about 3X faster and the AVX2 wmemset
is about 6X faster on Haswell.
* include/wchar.h (__wmemset_chk): New.
* sysdeps/x86_64/memset.S (VDUP_TO_VEC0_AND_SET_RETURN): Renamed
to MEMSET_VDUP_TO_VEC0_AND_SET_RETURN.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_CHK_SYMBOL): Likewise.
(WMEMSET_SYMBOL): Likewise.
(__wmemset): Add hidden definition.
(wmemset): Add weak hidden definition.
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
wmemset_chk-nonshared.
* sysdeps/x86_64/multiarch/ifunc-impl-list.c
(__libc_ifunc_impl_list): Add __wmemset_sse2_unaligned,
__wmemset_avx2_unaligned, __wmemset_avx512_unaligned,
__wmemset_chk_sse2_unaligned, __wmemset_chk_avx2_unaligned
and __wmemset_chk_avx512_unaligned.
* sysdeps/x86_64/multiarch/memset-avx2-unaligned-erms.S
(VDUP_TO_VEC0_AND_SET_RETURN): Renamed to ...
(MEMSET_VDUP_TO_VEC0_AND_SET_RETURN): This.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_SYMBOL): Likewise.
* sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S
(VDUP_TO_VEC0_AND_SET_RETURN): Renamed to ...
(MEMSET_VDUP_TO_VEC0_AND_SET_RETURN): This.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_SYMBOL): Likewise.
* sysdeps/x86_64/multiarch/memset-vec-unaligned-erms.S: Updated.
(WMEMSET_CHK_SYMBOL): New.
(WMEMSET_CHK_SYMBOL (__wmemset_chk, unaligned)): Likewise.
(WMEMSET_SYMBOL (__wmemset, unaligned)): Likewise.
* sysdeps/x86_64/multiarch/memset.S (WMEMSET_SYMBOL): New.
(libc_hidden_builtin_def): Also define __GI_wmemset and
__GI___wmemset.
(weak_alias): New.
* sysdeps/x86_64/multiarch/wmemset.c: New file.
* sysdeps/x86_64/multiarch/wmemset.h: Likewise.
* sysdeps/x86_64/multiarch/wmemset_chk-nonshared.S: Likewise.
* sysdeps/x86_64/multiarch/wmemset_chk.c: Likewise.
* sysdeps/x86_64/wmemset.c: Likewise.
* sysdeps/x86_64/wmemset_chk.c: Likewise.
2017-06-05 18:09:48 +00:00
|
|
|
# define libc_hidden_builtin_def(name) \
|
|
|
|
.globl __GI_memset; __GI_memset = __memset_sse2_unaligned; \
|
|
|
|
.globl __GI_wmemset; __GI_wmemset = __wmemset_sse2_unaligned; \
|
|
|
|
.globl __GI___wmemset; __GI___wmemset = __wmemset_sse2_unaligned
|
2014-07-14 14:58:27 +00:00
|
|
|
# endif
|
2015-10-28 13:29:03 +00:00
|
|
|
|
x86-64: Optimize wmemset with SSE2/AVX2/AVX512
The difference between memset and wmemset is byte vs int. Add stubs
to SSE2/AVX2/AVX512 memset for wmemset with updated constant and size:
SSE2 wmemset:
shl $0x2,%rdx
movd %esi,%xmm0
mov %rdi,%rax
pshufd $0x0,%xmm0,%xmm0
jmp entry_from_wmemset
SSE2 memset:
movd %esi,%xmm0
mov %rdi,%rax
punpcklbw %xmm0,%xmm0
punpcklwd %xmm0,%xmm0
pshufd $0x0,%xmm0,%xmm0
entry_from_wmemset:
Since the ERMS versions of wmemset requires "rep stosl" instead of
"rep stosb", only the vector store stubs of SSE2/AVX2/AVX512 wmemset
are added. The SSE2 wmemset is about 3X faster and the AVX2 wmemset
is about 6X faster on Haswell.
* include/wchar.h (__wmemset_chk): New.
* sysdeps/x86_64/memset.S (VDUP_TO_VEC0_AND_SET_RETURN): Renamed
to MEMSET_VDUP_TO_VEC0_AND_SET_RETURN.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_CHK_SYMBOL): Likewise.
(WMEMSET_SYMBOL): Likewise.
(__wmemset): Add hidden definition.
(wmemset): Add weak hidden definition.
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
wmemset_chk-nonshared.
* sysdeps/x86_64/multiarch/ifunc-impl-list.c
(__libc_ifunc_impl_list): Add __wmemset_sse2_unaligned,
__wmemset_avx2_unaligned, __wmemset_avx512_unaligned,
__wmemset_chk_sse2_unaligned, __wmemset_chk_avx2_unaligned
and __wmemset_chk_avx512_unaligned.
* sysdeps/x86_64/multiarch/memset-avx2-unaligned-erms.S
(VDUP_TO_VEC0_AND_SET_RETURN): Renamed to ...
(MEMSET_VDUP_TO_VEC0_AND_SET_RETURN): This.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_SYMBOL): Likewise.
* sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S
(VDUP_TO_VEC0_AND_SET_RETURN): Renamed to ...
(MEMSET_VDUP_TO_VEC0_AND_SET_RETURN): This.
(WMEMSET_VDUP_TO_VEC0_AND_SET_RETURN): New.
(WMEMSET_SYMBOL): Likewise.
* sysdeps/x86_64/multiarch/memset-vec-unaligned-erms.S: Updated.
(WMEMSET_CHK_SYMBOL): New.
(WMEMSET_CHK_SYMBOL (__wmemset_chk, unaligned)): Likewise.
(WMEMSET_SYMBOL (__wmemset, unaligned)): Likewise.
* sysdeps/x86_64/multiarch/memset.S (WMEMSET_SYMBOL): New.
(libc_hidden_builtin_def): Also define __GI_wmemset and
__GI___wmemset.
(weak_alias): New.
* sysdeps/x86_64/multiarch/wmemset.c: New file.
* sysdeps/x86_64/multiarch/wmemset.h: Likewise.
* sysdeps/x86_64/multiarch/wmemset_chk-nonshared.S: Likewise.
* sysdeps/x86_64/multiarch/wmemset_chk.c: Likewise.
* sysdeps/x86_64/wmemset.c: Likewise.
* sysdeps/x86_64/wmemset_chk.c: Likewise.
2017-06-05 18:09:48 +00:00
|
|
|
# undef weak_alias
|
|
|
|
# define weak_alias(original, alias) \
|
|
|
|
.weak bzero; bzero = __bzero
|
|
|
|
|
2015-10-28 13:29:03 +00:00
|
|
|
# undef strong_alias
|
|
|
|
# define strong_alias(original, alias)
|
2014-06-19 20:00:56 +00:00
|
|
|
#endif
|
|
|
|
|
|
|
|
#include "../memset.S"
|