har0ke/FFmpeg
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
FFmpeg/libavutil/x86/float_dsp.asm
Lynne bbe95f7353
x86: replace explicit REP_RETs with RETs 
From x86inc:
> On AMD cpus <=K10, an ordinary ret is slow if it immediately follows either
> a branch or a branch target. So switch to a 2-byte form of ret in that case.
> We can automatically detect "follows a branch", but not a branch target.
> (SSSE3 is a sufficient condition to know that your cpu doesn't have this problem.)

x86inc can automatically determine whether to use REP_RET rather than
REP in most of these cases, so impact is minimal. Additionally, a few
REP_RETs were used unnecessary, despite the return being nowhere near a
branch.

The only CPUs affected were AMD K10s, made between 2007 and 2011, 16
years ago and 12 years ago, respectively.

In the future, everyone involved with x86inc should consider dropping
REP_RETs altogether.
2023-02-01 04:23:55 +01:00

589 lines
15 KiB
NASM
Raw Blame History

;*****************************************************************************
;* x86-optimized Float DSP functions
;*
;* Copyright 2006 Loren Merritt
;*
;* This file is part of FFmpeg.
;*
;* FFmpeg 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.
;*
;* FFmpeg 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 FFmpeg; if not, write to the Free Software
;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
;******************************************************************************
%include "libavutil/x86/x86util.asm"
SECTION_RODATA 32
pd_reverse: dd 7, 6, 5, 4, 3, 2, 1, 0
SECTION .text
;-----------------------------------------------------------------------------
; void vector_fmul(float *dst, const float *src0, const float *src1, int len)
;-----------------------------------------------------------------------------
%macro VECTOR_FMUL 0
cglobal vector_fmul, 4,4,2, dst, src0, src1, len
lea lenq, [lend*4 - 64]
ALIGN 16
.loop:
%assign a 0
%rep 32/mmsize
mova m0, [src0q + lenq + (a+0)*mmsize]
mova m1, [src0q + lenq + (a+1)*mmsize]
mulps m0, m0, [src1q + lenq + (a+0)*mmsize]
mulps m1, m1, [src1q + lenq + (a+1)*mmsize]
mova [dstq + lenq + (a+0)*mmsize], m0
mova [dstq + lenq + (a+1)*mmsize], m1
%assign a a+2
%endrep
sub lenq, 64
jge .loop
RET
%endmacro
INIT_XMM sse
VECTOR_FMUL
%if HAVE_AVX_EXTERNAL
INIT_YMM avx
VECTOR_FMUL
%endif
;-----------------------------------------------------------------------------
; void vector_dmul(double *dst, const double *src0, const double *src1, int len)
;-----------------------------------------------------------------------------
%macro VECTOR_DMUL 0
cglobal vector_dmul, 4,4,4, dst, src0, src1, len
lea lend, [lenq*8 - mmsize*4]
ALIGN 16
.loop:
movaps m0, [src0q + lenq + 0*mmsize]
movaps m1, [src0q + lenq + 1*mmsize]
movaps m2, [src0q + lenq + 2*mmsize]
movaps m3, [src0q + lenq + 3*mmsize]
mulpd m0, m0, [src1q + lenq + 0*mmsize]
mulpd m1, m1, [src1q + lenq + 1*mmsize]
mulpd m2, m2, [src1q + lenq + 2*mmsize]
mulpd m3, m3, [src1q + lenq + 3*mmsize]
movaps [dstq + lenq + 0*mmsize], m0
movaps [dstq + lenq + 1*mmsize], m1
movaps [dstq + lenq + 2*mmsize], m2
movaps [dstq + lenq + 3*mmsize], m3
sub lenq, mmsize*4
jge .loop
RET
%endmacro
INIT_XMM sse2
VECTOR_DMUL
%if HAVE_AVX_EXTERNAL
INIT_YMM avx
VECTOR_DMUL
%endif
;------------------------------------------------------------------------------
; void ff_vector_fmac_scalar(float *dst, const float *src, float mul, int len)
;------------------------------------------------------------------------------
%macro VECTOR_FMAC_SCALAR 0
%if UNIX64
cglobal vector_fmac_scalar, 3,3,5, dst, src, len
%else
cglobal vector_fmac_scalar, 4,4,5, dst, src, mul, len
%endif
%if ARCH_X86_32
VBROADCASTSS m0, mulm
%else
%if WIN64
SWAP 0, 2
%endif
shufps xm0, xm0, 0
%if cpuflag(avx)
vinsertf128 m0, m0, xm0, 1
%endif
%endif
lea lenq, [lend*4-64]
.loop:
%if cpuflag(fma3)
mova m1, [dstq+lenq]
mova m2, [dstq+lenq+1*mmsize]
fmaddps m1, m0, [srcq+lenq], m1
fmaddps m2, m0, [srcq+lenq+1*mmsize], m2
%else ; cpuflag
mulps m1, m0, [srcq+lenq]
mulps m2, m0, [srcq+lenq+1*mmsize]
%if mmsize < 32
mulps m3, m0, [srcq+lenq+2*mmsize]
mulps m4, m0, [srcq+lenq+3*mmsize]
%endif ; mmsize
addps m1, m1, [dstq+lenq]
addps m2, m2, [dstq+lenq+1*mmsize]
%if mmsize < 32
addps m3, m3, [dstq+lenq+2*mmsize]
addps m4, m4, [dstq+lenq+3*mmsize]
%endif ; mmsize
%endif ; cpuflag
mova [dstq+lenq], m1
mova [dstq+lenq+1*mmsize], m2
%if mmsize < 32
mova [dstq+lenq+2*mmsize], m3
mova [dstq+lenq+3*mmsize], m4
%endif ; mmsize
sub lenq, 64
jge .loop
RET
%endmacro
INIT_XMM sse
VECTOR_FMAC_SCALAR
%if HAVE_AVX_EXTERNAL
INIT_YMM avx
VECTOR_FMAC_SCALAR
%endif
%if HAVE_FMA3_EXTERNAL
INIT_YMM fma3
VECTOR_FMAC_SCALAR
%endif
;------------------------------------------------------------------------------
; void ff_vector_fmul_scalar(float *dst, const float *src, float mul, int len)
;------------------------------------------------------------------------------
%macro VECTOR_FMUL_SCALAR 0
%if UNIX64
cglobal vector_fmul_scalar, 3,3,2, dst, src, len
%else
cglobal vector_fmul_scalar, 4,4,3, dst, src, mul, len
%endif
%if ARCH_X86_32
movss m0, mulm
%elif WIN64
SWAP 0, 2
%endif
shufps m0, m0, 0
lea lenq, [lend*4-mmsize]
.loop:
mova m1, [srcq+lenq]
mulps m1, m0
mova [dstq+lenq], m1
sub lenq, mmsize
jge .loop
RET
%endmacro
INIT_XMM sse
VECTOR_FMUL_SCALAR
;------------------------------------------------------------------------------
; void ff_vector_dmac_scalar(double *dst, const double *src, double mul,
; int len)
;------------------------------------------------------------------------------
%macro VECTOR_DMAC_SCALAR 0
%if ARCH_X86_32
cglobal vector_dmac_scalar, 2,4,5, dst, src, mul, len, lenaddr
mov lenq, lenaddrm
VBROADCASTSD m0, mulm
%else
%if UNIX64
cglobal vector_dmac_scalar, 3,3,5, dst, src, len
%else
cglobal vector_dmac_scalar, 4,4,5, dst, src, mul, len
SWAP 0, 2
%endif
movlhps xm0, xm0
%if cpuflag(avx)
vinsertf128 m0, m0, xm0, 1
%endif
%endif
lea lenq, [lend*8-mmsize*4]
.loop:
%if cpuflag(fma3)
movaps m1, [dstq+lenq]
movaps m2, [dstq+lenq+1*mmsize]
movaps m3, [dstq+lenq+2*mmsize]
movaps m4, [dstq+lenq+3*mmsize]
fmaddpd m1, m0, [srcq+lenq], m1
fmaddpd m2, m0, [srcq+lenq+1*mmsize], m2
fmaddpd m3, m0, [srcq+lenq+2*mmsize], m3
fmaddpd m4, m0, [srcq+lenq+3*mmsize], m4
%else ; cpuflag
mulpd m1, m0, [srcq+lenq]
mulpd m2, m0, [srcq+lenq+1*mmsize]
mulpd m3, m0, [srcq+lenq+2*mmsize]
mulpd m4, m0, [srcq+lenq+3*mmsize]
addpd m1, m1, [dstq+lenq]
addpd m2, m2, [dstq+lenq+1*mmsize]
addpd m3, m3, [dstq+lenq+2*mmsize]
addpd m4, m4, [dstq+lenq+3*mmsize]
%endif ; cpuflag
movaps [dstq+lenq], m1
movaps [dstq+lenq+1*mmsize], m2
movaps [dstq+lenq+2*mmsize], m3
movaps [dstq+lenq+3*mmsize], m4
sub lenq, mmsize*4
jge .loop
RET
%endmacro
INIT_XMM sse2
VECTOR_DMAC_SCALAR
%if HAVE_AVX_EXTERNAL
INIT_YMM avx
VECTOR_DMAC_SCALAR
%endif
%if HAVE_FMA3_EXTERNAL
INIT_YMM fma3
VECTOR_DMAC_SCALAR
%endif
;------------------------------------------------------------------------------
; void ff_vector_dmul_scalar(double *dst, const double *src, double mul,
; int len)
;------------------------------------------------------------------------------
%macro VECTOR_DMUL_SCALAR 0
%if ARCH_X86_32
cglobal vector_dmul_scalar, 3,4,3, dst, src, mul, len, lenaddr
mov lenq, lenaddrm
%elif UNIX64
cglobal vector_dmul_scalar, 3,3,3, dst, src, len
%else
cglobal vector_dmul_scalar, 4,4,3, dst, src, mul, len
%endif
%if ARCH_X86_32
VBROADCASTSD m0, mulm
%else
%if WIN64
SWAP 0, 2
%endif
movlhps xm0, xm0
%if cpuflag(avx)
vinsertf128 ym0, ym0, xm0, 1
%endif
%endif
lea lenq, [lend*8-2*mmsize]
.loop:
mulpd m1, m0, [srcq+lenq ]
mulpd m2, m0, [srcq+lenq+mmsize]
movaps [dstq+lenq ], m1
movaps [dstq+lenq+mmsize], m2
sub lenq, 2*mmsize
jge .loop
RET
%endmacro
INIT_XMM sse2
VECTOR_DMUL_SCALAR
%if HAVE_AVX_EXTERNAL
INIT_YMM avx
VECTOR_DMUL_SCALAR
%endif
;-----------------------------------------------------------------------------
; vector_fmul_window(float *dst, const float *src0,
; const float *src1, const float *win, int len);
;-----------------------------------------------------------------------------
INIT_XMM sse
cglobal vector_fmul_window, 5, 6, 6, dst, src0, src1, win, len, len1
shl lend, 2
lea len1q, [lenq - mmsize]
add src0q, lenq
add dstq, lenq
add winq, lenq
neg lenq
.loop:
mova m0, [winq + lenq]
mova m4, [src0q + lenq]
mova m1, [winq + len1q]
mova m5, [src1q + len1q]
shufps m1, m1, 0x1b
shufps m5, m5, 0x1b
mova m2, m0
mova m3, m1
mulps m2, m4
mulps m3, m5
mulps m1, m4
mulps m0, m5
addps m2, m3
subps m1, m0
shufps m2, m2, 0x1b
mova [dstq + lenq], m1
mova [dstq + len1q], m2
sub len1q, mmsize
add lenq, mmsize
jl .loop
RET
;-----------------------------------------------------------------------------
; vector_fmul_add(float *dst, const float *src0, const float *src1,
; const float *src2, int len)
;-----------------------------------------------------------------------------
%macro VECTOR_FMUL_ADD 0
cglobal vector_fmul_add, 5,5,4, dst, src0, src1, src2, len
lea lenq, [lend*4 - 2*mmsize]
ALIGN 16
.loop:
mova m0, [src0q + lenq]
mova m1, [src0q + lenq + mmsize]
%if cpuflag(fma3)
mova m2, [src2q + lenq]
mova m3, [src2q + lenq + mmsize]
fmaddps m0, m0, [src1q + lenq], m2
fmaddps m1, m1, [src1q + lenq + mmsize], m3
%else
mulps m0, m0, [src1q + lenq]
mulps m1, m1, [src1q + lenq + mmsize]
addps m0, m0, [src2q + lenq]
addps m1, m1, [src2q + lenq + mmsize]
%endif
mova [dstq + lenq], m0
mova [dstq + lenq + mmsize], m1
sub lenq, 2*mmsize
jge .loop
RET
%endmacro
INIT_XMM sse
VECTOR_FMUL_ADD
%if HAVE_AVX_EXTERNAL
INIT_YMM avx
VECTOR_FMUL_ADD
%endif
%if HAVE_FMA3_EXTERNAL
INIT_YMM fma3
VECTOR_FMUL_ADD
%endif
;-----------------------------------------------------------------------------
; void vector_fmul_reverse(float *dst, const float *src0, const float *src1,
; int len)
;-----------------------------------------------------------------------------
%macro VECTOR_FMUL_REVERSE 0
cglobal vector_fmul_reverse, 4,4,2, dst, src0, src1, len
%if cpuflag(avx2)
movaps m2, [pd_reverse]
%endif
lea lenq, [lend*4 - 2*mmsize]
ALIGN 16
.loop:
%if cpuflag(avx2)
vpermps m0, m2, [src1q]
vpermps m1, m2, [src1q+mmsize]
%elif cpuflag(avx)
vmovaps xmm0, [src1q + 16]
vinsertf128 m0, m0, [src1q], 1
vshufps m0, m0, m0, q0123
vmovaps xmm1, [src1q + mmsize + 16]
vinsertf128 m1, m1, [src1q + mmsize], 1
vshufps m1, m1, m1, q0123
%else
mova m0, [src1q]
mova m1, [src1q + mmsize]
shufps m0, m0, q0123
shufps m1, m1, q0123
%endif
mulps m0, m0, [src0q + lenq + mmsize]
mulps m1, m1, [src0q + lenq]
movaps [dstq + lenq + mmsize], m0
movaps [dstq + lenq], m1
add src1q, 2*mmsize
sub lenq, 2*mmsize
jge .loop
RET
%endmacro
INIT_XMM sse
VECTOR_FMUL_REVERSE
%if HAVE_AVX_EXTERNAL
INIT_YMM avx
VECTOR_FMUL_REVERSE
%endif
%if HAVE_AVX2_EXTERNAL
INIT_YMM avx2
VECTOR_FMUL_REVERSE
%endif
; float scalarproduct_float_sse(const float *v1, const float *v2, int len)
INIT_XMM sse
cglobal scalarproduct_float, 3,3,2, v1, v2, offset
shl offsetd, 2
add v1q, offsetq
add v2q, offsetq
neg offsetq
xorps xmm0, xmm0
.loop:
movaps xmm1, [v1q+offsetq]
mulps xmm1, [v2q+offsetq]
addps xmm0, xmm1
add offsetq, 16
js .loop
movhlps xmm1, xmm0
addps xmm0, xmm1
movss xmm1, xmm0
shufps xmm0, xmm0, 1
addss xmm0, xmm1
%if ARCH_X86_64 == 0
movss r0m, xmm0
fld dword r0m
%endif
RET
INIT_YMM fma3
cglobal scalarproduct_float, 3,5,8, v1, v2, size, len, offset
xor offsetq, offsetq
xorps m0, m0, m0
shl sized, 2
mov lenq, sizeq
cmp lenq, 32
jl .l16
cmp lenq, 64
jl .l32
xorps m1, m1, m1
cmp lenq, 128
jl .l64
and lenq, ~127
xorps m2, m2, m2
xorps m3, m3, m3
.loop128:
movups m4, [v1q+offsetq]
movups m5, [v1q+offsetq + 32]
movups m6, [v1q+offsetq + 64]
movups m7, [v1q+offsetq + 96]
fmaddps m0, m4, [v2q+offsetq ], m0
fmaddps m1, m5, [v2q+offsetq + 32], m1
fmaddps m2, m6, [v2q+offsetq + 64], m2
fmaddps m3, m7, [v2q+offsetq + 96], m3
add offsetq, 128
cmp offsetq, lenq
jl .loop128
addps m0, m0, m2
addps m1, m1, m3
mov lenq, sizeq
and lenq, 127
cmp lenq, 64
jge .l64
addps m0, m0, m1
cmp lenq, 32
jge .l32
vextractf128 xmm2, m0, 1
addps xmm0, xmm2
cmp lenq, 16
jge .l16
movhlps xmm1, xmm0
addps xmm0, xmm1
movss xmm1, xmm0
shufps xmm0, xmm0, 1
addss xmm0, xmm1
%if ARCH_X86_64 == 0
movss r0m, xm0
fld dword r0m
%endif
RET
.l64:
and lenq, ~63
add lenq, offsetq
.loop64:
movups m4, [v1q+offsetq]
movups m5, [v1q+offsetq + 32]
fmaddps m0, m4, [v2q+offsetq], m0
fmaddps m1, m5, [v2q+offsetq + 32], m1
add offsetq, 64
cmp offsetq, lenq
jl .loop64
addps m0, m0, m1
mov lenq, sizeq
and lenq, 63
cmp lenq, 32
jge .l32
vextractf128 xmm2, m0, 1
addps xmm0, xmm2
cmp lenq, 16
jge .l16
movhlps xmm1, xmm0
addps xmm0, xmm1
movss xmm1, xmm0
shufps xmm0, xmm0, 1
addss xmm0, xmm1
%if ARCH_X86_64 == 0
movss r0m, xm0
fld dword r0m
%endif
RET
.l32:
and lenq, ~31
add lenq, offsetq
.loop32:
movups m4, [v1q+offsetq]
fmaddps m0, m4, [v2q+offsetq], m0
add offsetq, 32
cmp offsetq, lenq
jl .loop32
vextractf128 xmm2, m0, 1
addps xmm0, xmm2
mov lenq, sizeq
and lenq, 31
cmp lenq, 16
jge .l16
movhlps xmm1, xmm0
addps xmm0, xmm1
movss xmm1, xmm0
shufps xmm0, xmm0, 1
addss xmm0, xmm1
%if ARCH_X86_64 == 0
movss r0m, xm0
fld dword r0m
%endif
RET
.l16:
and lenq, ~15
add lenq, offsetq
.loop16:
movaps xmm1, [v1q+offsetq]
mulps xmm1, [v2q+offsetq]
addps xmm0, xmm1
add offsetq, 16
cmp offsetq, lenq
jl .loop16
movhlps xmm1, xmm0
addps xmm0, xmm1
movss xmm1, xmm0
shufps xmm0, xmm0, 1
addss xmm0, xmm1
%if ARCH_X86_64 == 0
movss r0m, xm0
fld dword r0m
%endif
RET
;-----------------------------------------------------------------------------
; void ff_butterflies_float(float *src0, float *src1, int len);
;-----------------------------------------------------------------------------
INIT_XMM sse
cglobal butterflies_float, 3,3,3, src0, src1, len
shl lend, 2
add src0q, lenq
add src1q, lenq
neg lenq
.loop:
mova m0, [src0q + lenq]
mova m1, [src1q + lenq]
subps m2, m0, m1
addps m0, m0, m1
mova [src1q + lenq], m2
mova [src0q + lenq], m0
add lenq, mmsize
jl .loop
RET
Reference in New Issue View Git Blame Copy Permalink