Commit Graph

4 Commits

Author SHA1 Message Date
H.J. Lu
ef9c4cb6c7 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 11:09:59 -07:00
H.J. Lu
f43cb35c9b Require binutils 2.24 to build x86-64 glibc [BZ #20139]
If assembler doesn't support AVX512DQ, _dl_runtime_resolve_avx is used
to save the first 8 vector registers, which only saves the lower 256
bits of vector register, for lazy binding.  When it is called on AVX512
platform, the upper 256 bits of ZMM registers are clobbered.  Parameters
passed in ZMM registers will be wrong when the function is called the
first time.  This patch requires binutils 2.24, whose assembler can store
and load ZMM registers, to build x86-64 glibc.  Since mathvec library
needs assembler support for AVX512DQ,  we disable mathvec if assembler
doesn't support AVX512DQ.

	[BZ #20139]
	* config.h.in (HAVE_AVX512_ASM_SUPPORT): Renamed to ...
	(HAVE_AVX512DQ_ASM_SUPPORT): This.
	* sysdeps/x86_64/configure.ac: Require assembler from binutils
	2.24 or above.
	(HAVE_AVX512_ASM_SUPPORT): Removed.
	(HAVE_AVX512DQ_ASM_SUPPORT): New.
	* sysdeps/x86_64/configure: Regenerated.
	* sysdeps/x86_64/dl-trampoline.S: Make HAVE_AVX512_ASM_SUPPORT
	check unconditional.
	* sysdeps/x86_64/multiarch/ifunc-impl-list.c: Likewise.
	* sysdeps/x86_64/multiarch/memcpy.S: Likewise.
	* sysdeps/x86_64/multiarch/memcpy_chk.S: Likewise.
	* sysdeps/x86_64/multiarch/memmove-avx512-no-vzeroupper.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memmove-avx512-unaligned-erms.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memmove.S: Likewise.
	* sysdeps/x86_64/multiarch/memmove_chk.S: Likewise.
	* sysdeps/x86_64/multiarch/mempcpy.S: Likewise.
	* sysdeps/x86_64/multiarch/mempcpy_chk.S: Likewise.
	* sysdeps/x86_64/multiarch/memset-avx512-no-vzeroupper.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memset.S: Likewise.
	* sysdeps/x86_64/multiarch/memset_chk.S: Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_d_cos8_core_avx512.S: Check
	HAVE_AVX512DQ_ASM_SUPPORT instead of HAVE_AVX512_ASM_SUPPORT.
	* sysdeps/x86_64/fpu/multiarch/svml_d_exp8_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_d_log8_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_d_pow8_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_d_sin8_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_d_sincos8_core_avx512.:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_cosf16_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_expf16_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_powf16_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx51:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_sinf16_core_avx512.S:
	Likewise.
2016-07-01 06:03:05 -07:00
H.J. Lu
5cd7af016d Don't put SSE2/AVX/AVX512 memmove/memset in ld.so
Since memmove and memset in ld.so don't use IFUNC, don't put SSE2, AVX
and AVX512 memmove and memset in ld.so.

	* sysdeps/x86_64/multiarch/memmove-avx-unaligned-erms.S: Skip
	if not in libc.
	* sysdeps/x86_64/multiarch/memmove-avx512-unaligned-erms.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memset-avx2-unaligned-erms.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S:
	Likewise.
2016-04-03 14:35:38 -07:00
H.J. Lu
830566307f Add x86-64 memset with unaligned store and rep stosb
Implement x86-64 memset with unaligned store and rep movsb.  Support
16-byte, 32-byte and 64-byte vector register sizes.  A single file
provides 2 implementations of memset, one with rep stosb and the other
without rep stosb.  They share the same codes when size is between 2
times of vector register size and REP_STOSB_THRESHOLD which defaults
to 2KB.

Key features:

1. Use overlapping store to avoid branch.
2. For size <= 4 times of vector register size, fully unroll the loop.
3. For size > 4 times of vector register size, store 4 times of vector
register size at a time.

	[BZ #19881]
	* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
	memset-sse2-unaligned-erms, memset-avx2-unaligned-erms and
	memset-avx512-unaligned-erms.
	* sysdeps/x86_64/multiarch/ifunc-impl-list.c
	(__libc_ifunc_impl_list): Test __memset_chk_sse2_unaligned,
	__memset_chk_sse2_unaligned_erms, __memset_chk_avx2_unaligned,
	__memset_chk_avx2_unaligned_erms, __memset_chk_avx512_unaligned,
	__memset_chk_avx512_unaligned_erms, __memset_sse2_unaligned,
	__memset_sse2_unaligned_erms, __memset_erms,
	__memset_avx2_unaligned, __memset_avx2_unaligned_erms,
	__memset_avx512_unaligned_erms and __memset_avx512_unaligned.
	* sysdeps/x86_64/multiarch/memset-avx2-unaligned-erms.S: New
	file.
	* sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memset-sse2-unaligned-erms.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memset-vec-unaligned-erms.S:
	Likewise.
2016-03-31 10:06:07 -07:00