virtualenv/FFmpeg
1
0
Fork 0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-12 19:18:44 +02:00
Code Issues Releases Activity
ad4d647591
FFmpeg/libavcodec/x86/h264_chromamc.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

657 lines
17 KiB
NASM
Raw Blame History

;******************************************************************************
;* MMX/SSSE3-optimized functions for H.264 chroma MC
;* Copyright (c) 2005 Zoltan Hidvegi <hzoli -a- hzoli -d- com>,
;* 2005-2008 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
rnd_rv40_2d_tbl: times 4 dw 0
times 4 dw 16
times 4 dw 32
times 4 dw 16
times 4 dw 32
times 4 dw 28
times 4 dw 32
times 4 dw 28
times 4 dw 0
times 4 dw 32
times 4 dw 16
times 4 dw 32
times 4 dw 32
times 4 dw 28
times 4 dw 32
times 4 dw 28
rnd_rv40_1d_tbl: times 4 dw 0
times 4 dw 2
times 4 dw 4
times 4 dw 2
times 4 dw 4
times 4 dw 3
times 4 dw 4
times 4 dw 3
times 4 dw 0
times 4 dw 4
times 4 dw 2
times 4 dw 4
times 4 dw 4
times 4 dw 3
times 4 dw 4
times 4 dw 3
cextern pw_3
cextern pw_4
cextern pw_8
pw_28: times 8 dw 28
cextern pw_32
cextern pw_64
SECTION .text
%macro mv0_pixels_mc8 0
lea r4, [r2*2 ]
.next4rows:
movq mm0, [r1 ]
movq mm1, [r1+r2]
add r1, r4
CHROMAMC_AVG mm0, [r0 ]
CHROMAMC_AVG mm1, [r0+r2]
movq [r0 ], mm0
movq [r0+r2], mm1
add r0, r4
movq mm0, [r1 ]
movq mm1, [r1+r2]
add r1, r4
CHROMAMC_AVG mm0, [r0 ]
CHROMAMC_AVG mm1, [r0+r2]
movq [r0 ], mm0
movq [r0+r2], mm1
add r0, r4
sub r3d, 4
jne .next4rows
%endmacro
%macro chroma_mc8_mmx_func 2-3
%ifidn %2, rv40
%ifdef PIC
%define rnd_1d_rv40 r8
%define rnd_2d_rv40 r8
%define extra_regs 2
%else ; no-PIC
%define rnd_1d_rv40 rnd_rv40_1d_tbl
%define rnd_2d_rv40 rnd_rv40_2d_tbl
%define extra_regs 1
%endif ; PIC
%else
%define extra_regs 0
%endif ; rv40
; void ff_put/avg_h264_chroma_mc8_*(uint8_t *dst /* align 8 */,
; const uint8_t *src /* align 1 */,
; ptrdiff_t stride, int h, int mx, int my)
cglobal %1_%2_chroma_mc8%3, 6, 7 + extra_regs, 0
mov r6d, r5d
or r6d, r4d
jne .at_least_one_non_zero
; mx == 0 AND my == 0 - no filter needed
mv0_pixels_mc8
RET
.at_least_one_non_zero:
%ifidn %2, rv40
%if ARCH_X86_64
mov r7, r5
and r7, 6 ; &~1 for mx/my=[0,7]
lea r7, [r7*4+r4]
sar r7d, 1
%define rnd_bias r7
%define dest_reg r0
%else ; x86-32
mov r0, r5
and r0, 6 ; &~1 for mx/my=[0,7]
lea r0, [r0*4+r4]
sar r0d, 1
%define rnd_bias r0
%define dest_reg r5
%endif
%else ; vc1, h264
%define rnd_bias 0
%define dest_reg r0
%endif
test r5d, r5d
mov r6, 1
je .my_is_zero
test r4d, r4d
mov r6, r2 ; dxy = x ? 1 : stride
jne .both_non_zero
.my_is_zero:
; mx == 0 XOR my == 0 - 1 dimensional filter only
or r4d, r5d ; x + y
%ifidn %2, rv40
%ifdef PIC
lea r8, [rnd_rv40_1d_tbl]
%endif
%if ARCH_X86_64 == 0
mov r5, r0m
%endif
%endif
movd m5, r4d
movq m4, [pw_8]
movq m6, [rnd_1d_%2+rnd_bias*8] ; mm6 = rnd >> 3
punpcklwd m5, m5
punpckldq m5, m5 ; mm5 = B = x
pxor m7, m7
psubw m4, m5 ; mm4 = A = 8-x
.next1drow:
movq m0, [r1 ] ; mm0 = src[0..7]
movq m2, [r1+r6] ; mm1 = src[1..8]
movq m1, m0
movq m3, m2
punpcklbw m0, m7
punpckhbw m1, m7
punpcklbw m2, m7
punpckhbw m3, m7
pmullw m0, m4 ; [mm0,mm1] = A * src[0..7]
pmullw m1, m4
pmullw m2, m5 ; [mm2,mm3] = B * src[1..8]
pmullw m3, m5
paddw m0, m6
paddw m1, m6
paddw m0, m2
paddw m1, m3
psrlw m0, 3
psrlw m1, 3
packuswb m0, m1
CHROMAMC_AVG m0, [dest_reg]
movq [dest_reg], m0 ; dst[0..7] = (A * src[0..7] + B * src[1..8] + (rnd >> 3)) >> 3
add dest_reg, r2
add r1, r2
dec r3d
jne .next1drow
RET
.both_non_zero: ; general case, bilinear
movd m4, r4d ; x
movd m6, r5d ; y
%ifidn %2, rv40
%ifdef PIC
lea r8, [rnd_rv40_2d_tbl]
%endif
%if ARCH_X86_64 == 0
mov r5, r0m
%endif
%endif
mov r6, rsp ; backup stack pointer
and rsp, ~(mmsize-1) ; align stack
sub rsp, 16 ; AA and DD
punpcklwd m4, m4
punpcklwd m6, m6
punpckldq m4, m4 ; mm4 = x words
punpckldq m6, m6 ; mm6 = y words
movq m5, m4
pmullw m4, m6 ; mm4 = x * y
psllw m5, 3
psllw m6, 3
movq m7, m5
paddw m7, m6
movq [rsp+8], m4 ; DD = x * y
psubw m5, m4 ; mm5 = B = 8x - xy
psubw m6, m4 ; mm6 = C = 8y - xy
paddw m4, [pw_64]
psubw m4, m7 ; mm4 = A = xy - (8x+8y) + 64
pxor m7, m7
movq [rsp ], m4
movq m0, [r1 ] ; mm0 = src[0..7]
movq m1, [r1+1] ; mm1 = src[1..8]
.next2drow:
add r1, r2
movq m2, m0
movq m3, m1
punpckhbw m0, m7
punpcklbw m1, m7
punpcklbw m2, m7
punpckhbw m3, m7
pmullw m0, [rsp]
pmullw m2, [rsp]
pmullw m1, m5
pmullw m3, m5
paddw m2, m1 ; mm2 = A * src[0..3] + B * src[1..4]
paddw m3, m0 ; mm3 = A * src[4..7] + B * src[5..8]
movq m0, [r1]
movq m1, m0
punpcklbw m0, m7
punpckhbw m1, m7
pmullw m0, m6
pmullw m1, m6
paddw m2, m0
paddw m3, m1 ; [mm2,mm3] += C * src[0..7]
movq m1, [r1+1]
movq m0, m1
movq m4, m1
punpcklbw m0, m7
punpckhbw m4, m7
pmullw m0, [rsp+8]
pmullw m4, [rsp+8]
paddw m2, m0
paddw m3, m4 ; [mm2,mm3] += D * src[1..8]
movq m0, [r1]
paddw m2, [rnd_2d_%2+rnd_bias*8]
paddw m3, [rnd_2d_%2+rnd_bias*8]
psrlw m2, 6
psrlw m3, 6
packuswb m2, m3
CHROMAMC_AVG m2, [dest_reg]
movq [dest_reg], m2 ; dst[0..7] = ([mm2,mm3] + rnd) >> 6
add dest_reg, r2
dec r3d
jne .next2drow
mov rsp, r6 ; restore stack pointer
RET
%endmacro
%macro chroma_mc4_mmx_func 2
%define extra_regs 0
%ifidn %2, rv40
%ifdef PIC
%define extra_regs 1
%endif ; PIC
%endif ; rv40
cglobal %1_%2_chroma_mc4, 6, 6 + extra_regs, 0
pxor m7, m7
movd m2, r4d ; x
movd m3, r5d ; y
movq m4, [pw_8]
movq m5, [pw_8]
punpcklwd m2, m2
punpcklwd m3, m3
punpcklwd m2, m2
punpcklwd m3, m3
psubw m4, m2
psubw m5, m3
%ifidn %2, rv40
%ifdef PIC
lea r6, [rnd_rv40_2d_tbl]
%define rnd_2d_rv40 r6
%else
%define rnd_2d_rv40 rnd_rv40_2d_tbl
%endif
and r5, 6 ; &~1 for mx/my=[0,7]
lea r5, [r5*4+r4]
sar r5d, 1
%define rnd_bias r5
%else ; vc1, h264
%define rnd_bias 0
%endif
movd m0, [r1 ]
movd m6, [r1+1]
add r1, r2
punpcklbw m0, m7
punpcklbw m6, m7
pmullw m0, m4
pmullw m6, m2
paddw m6, m0
.next2rows:
movd m0, [r1 ]
movd m1, [r1+1]
add r1, r2
punpcklbw m0, m7
punpcklbw m1, m7
pmullw m0, m4
pmullw m1, m2
paddw m1, m0
movq m0, m1
pmullw m6, m5
pmullw m1, m3
paddw m6, [rnd_2d_%2+rnd_bias*8]
paddw m1, m6
psrlw m1, 6
packuswb m1, m1
CHROMAMC_AVG4 m1, m6, [r0]
movd [r0], m1
add r0, r2
movd m6, [r1 ]
movd m1, [r1+1]
add r1, r2
punpcklbw m6, m7
punpcklbw m1, m7
pmullw m6, m4
pmullw m1, m2
paddw m1, m6
movq m6, m1
pmullw m0, m5
pmullw m1, m3
paddw m0, [rnd_2d_%2+rnd_bias*8]
paddw m1, m0
psrlw m1, 6
packuswb m1, m1
CHROMAMC_AVG4 m1, m0, [r0]
movd [r0], m1
add r0, r2
sub r3d, 2
jnz .next2rows
RET
%endmacro
%macro chroma_mc2_mmx_func 2
cglobal %1_%2_chroma_mc2, 6, 7, 0
mov r6d, r4d
shl r4d, 16
sub r4d, r6d
add r4d, 8
imul r5d, r4d ; x*y<<16 | y*(8-x)
shl r4d, 3
sub r4d, r5d ; x*(8-y)<<16 | (8-x)*(8-y)
movd m5, r4d
movd m6, r5d
punpckldq m5, m5 ; mm5 = {A,B,A,B}
punpckldq m6, m6 ; mm6 = {C,D,C,D}
pxor m7, m7
movd m2, [r1]
punpcklbw m2, m7
pshufw m2, m2, 0x94 ; mm0 = src[0,1,1,2]
.nextrow:
add r1, r2
movq m1, m2
pmaddwd m1, m5 ; mm1 = A * src[0,1] + B * src[1,2]
movd m0, [r1]
punpcklbw m0, m7
pshufw m0, m0, 0x94 ; mm0 = src[0,1,1,2]
movq m2, m0
pmaddwd m0, m6
paddw m1, [rnd_2d_%2]
paddw m1, m0 ; mm1 += C * src[0,1] + D * src[1,2]
psrlw m1, 6
packssdw m1, m7
packuswb m1, m7
CHROMAMC_AVG4 m1, m3, [r0]
movd r5d, m1
mov [r0], r5w
add r0, r2
sub r3d, 1
jnz .nextrow
RET
%endmacro
%define rnd_1d_h264 pw_4
%define rnd_2d_h264 pw_32
%define rnd_1d_vc1 pw_3
%define rnd_2d_vc1 pw_28
%macro NOTHING 2-3
%endmacro
%macro DIRECT_AVG 2
PAVGB %1, %2
%endmacro
%macro COPY_AVG 3
movd %2, %3
PAVGB %1, %2
%endmacro
INIT_MMX mmx
%define CHROMAMC_AVG NOTHING
%define CHROMAMC_AVG4 NOTHING
chroma_mc8_mmx_func put, h264, _rnd
chroma_mc8_mmx_func put, vc1, _nornd
chroma_mc8_mmx_func put, rv40
chroma_mc4_mmx_func put, h264
chroma_mc4_mmx_func put, rv40
INIT_MMX mmxext
chroma_mc2_mmx_func put, h264
%define CHROMAMC_AVG DIRECT_AVG
%define CHROMAMC_AVG4 COPY_AVG
chroma_mc8_mmx_func avg, h264, _rnd
chroma_mc8_mmx_func avg, vc1, _nornd
chroma_mc8_mmx_func avg, rv40
chroma_mc4_mmx_func avg, h264
chroma_mc4_mmx_func avg, rv40
chroma_mc2_mmx_func avg, h264
%macro chroma_mc8_ssse3_func 2-3
cglobal %1_%2_chroma_mc8%3, 6, 7, 8
mov r6d, r5d
or r6d, r4d
jne .at_least_one_non_zero
; mx == 0 AND my == 0 - no filter needed
mv0_pixels_mc8
RET
.at_least_one_non_zero:
test r5d, r5d
je .my_is_zero
test r4d, r4d
je .mx_is_zero
; general case, bilinear
mov r6d, r4d
shl r4d, 8
sub r4, r6
mov r6, 8
add r4, 8 ; x*288+8 = x<<8 | (8-x)
sub r6d, r5d
imul r6, r4 ; (8-y)*(x*255+8) = (8-y)*x<<8 | (8-y)*(8-x)
imul r4d, r5d ; y *(x*255+8) = y *x<<8 | y *(8-x)
movd m7, r6d
movd m6, r4d
movdqa m5, [rnd_2d_%2]
movq m0, [r1 ]
movq m1, [r1+1]
pshuflw m7, m7, 0
pshuflw m6, m6, 0
punpcklbw m0, m1
movlhps m7, m7
movlhps m6, m6
.next2rows:
movq m1, [r1+r2*1 ]
movq m2, [r1+r2*1+1]
movq m3, [r1+r2*2 ]
movq m4, [r1+r2*2+1]
lea r1, [r1+r2*2]
punpcklbw m1, m2
movdqa m2, m1
punpcklbw m3, m4
movdqa m4, m3
pmaddubsw m0, m7
pmaddubsw m1, m6
pmaddubsw m2, m7
pmaddubsw m3, m6
paddw m0, m5
paddw m2, m5
paddw m1, m0
paddw m3, m2
psrlw m1, 6
movdqa m0, m4
psrlw m3, 6
%ifidn %1, avg
movq m2, [r0 ]
movhps m2, [r0+r2]
%endif
packuswb m1, m3
CHROMAMC_AVG m1, m2
movq [r0 ], m1
movhps [r0+r2], m1
sub r3d, 2
lea r0, [r0+r2*2]
jg .next2rows
RET
.my_is_zero:
mov r5d, r4d
shl r4d, 8
add r4, 8
sub r4, r5 ; 255*x+8 = x<<8 | (8-x)
movd m7, r4d
movdqa m6, [rnd_1d_%2]
pshuflw m7, m7, 0
movlhps m7, m7
.next2xrows:
movq m0, [r1 ]
movq m1, [r1 +1]
movq m2, [r1+r2 ]
movq m3, [r1+r2+1]
punpcklbw m0, m1
punpcklbw m2, m3
pmaddubsw m0, m7
pmaddubsw m2, m7
%ifidn %1, avg
movq m4, [r0 ]
movhps m4, [r0+r2]
%endif
paddw m0, m6
paddw m2, m6
psrlw m0, 3
psrlw m2, 3
packuswb m0, m2
CHROMAMC_AVG m0, m4
movq [r0 ], m0
movhps [r0+r2], m0
sub r3d, 2
lea r0, [r0+r2*2]
lea r1, [r1+r2*2]
jg .next2xrows
RET
.mx_is_zero:
mov r4d, r5d
shl r5d, 8
add r5, 8
sub r5, r4 ; 255*y+8 = y<<8 | (8-y)
movd m7, r5d
movdqa m6, [rnd_1d_%2]
pshuflw m7, m7, 0
movlhps m7, m7
.next2yrows:
movq m0, [r1 ]
movq m1, [r1+r2 ]
movdqa m2, m1
movq m3, [r1+r2*2]
lea r1, [r1+r2*2]
punpcklbw m0, m1
punpcklbw m2, m3
pmaddubsw m0, m7
pmaddubsw m2, m7
%ifidn %1, avg
movq m4, [r0 ]
movhps m4, [r0+r2]
%endif
paddw m0, m6
paddw m2, m6
psrlw m0, 3
psrlw m2, 3
packuswb m0, m2
CHROMAMC_AVG m0, m4
movq [r0 ], m0
movhps [r0+r2], m0
sub r3d, 2
lea r0, [r0+r2*2]
jg .next2yrows
RET
%endmacro
%macro chroma_mc4_ssse3_func 2
cglobal %1_%2_chroma_mc4, 6, 7, 0
mov r6, r4
shl r4d, 8
sub r4d, r6d
mov r6, 8
add r4d, 8 ; x*288+8
sub r6d, r5d
imul r6d, r4d ; (8-y)*(x*255+8) = (8-y)*x<<8 | (8-y)*(8-x)
imul r4d, r5d ; y *(x*255+8) = y *x<<8 | y *(8-x)
movd m7, r6d
movd m6, r4d
movq m5, [pw_32]
movd m0, [r1 ]
pshufw m7, m7, 0
punpcklbw m0, [r1+1]
pshufw m6, m6, 0
.next2rows:
movd m1, [r1+r2*1 ]
movd m3, [r1+r2*2 ]
punpcklbw m1, [r1+r2*1+1]
punpcklbw m3, [r1+r2*2+1]
lea r1, [r1+r2*2]
movq m2, m1
movq m4, m3
pmaddubsw m0, m7
pmaddubsw m1, m6
pmaddubsw m2, m7
pmaddubsw m3, m6
paddw m0, m5
paddw m2, m5
paddw m1, m0
paddw m3, m2
psrlw m1, 6
movq m0, m4
psrlw m3, 6
packuswb m1, m1
packuswb m3, m3
CHROMAMC_AVG m1, [r0 ]
CHROMAMC_AVG m3, [r0+r2]
movd [r0 ], m1
movd [r0+r2], m3
sub r3d, 2
lea r0, [r0+r2*2]
jg .next2rows
RET
%endmacro
%define CHROMAMC_AVG NOTHING
INIT_XMM ssse3
chroma_mc8_ssse3_func put, h264, _rnd
chroma_mc8_ssse3_func put, vc1, _nornd
INIT_MMX ssse3
chroma_mc4_ssse3_func put, h264
%define CHROMAMC_AVG DIRECT_AVG
INIT_XMM ssse3
chroma_mc8_ssse3_func avg, h264, _rnd
chroma_mc8_ssse3_func avg, vc1, _nornd
INIT_MMX ssse3
chroma_mc4_ssse3_func avg, h264
Reference in New Issue View Git Blame Copy Permalink