Merge "Rewrite HORIZx4 and HORIZx8 in subpixel filter functions"

This commit is contained in:
Yunqing Wang 2013-10-03 12:17:47 -07:00 коммит произвёл Gerrit Code Review
Родитель 3aed95dbdb ed22179a82
Коммит 134dfea878
1 изменённых файлов: 111 добавлений и 92 удалений

Просмотреть файл

@ -534,6 +534,21 @@ sym(vp9_filter_block1d16_v8_avg_ssse3):
ret ret
;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
%macro HORIZx4_ROW 2
movdqa %2, %1
pshufb %1, [GLOBAL(shuf_t0t1)]
pshufb %2, [GLOBAL(shuf_t2t3)]
pmaddubsw %1, xmm6
pmaddubsw %2, xmm7
paddsw %1, %2
movdqa %2, %1
psrldq %2, 8
paddsw %1, %2
paddsw %1, xmm5
psraw %1, 7
packuswb %1, %1
%endm
%macro HORIZx4 1 %macro HORIZx4 1
mov rdx, arg(5) ;filter ptr mov rdx, arg(5) ;filter ptr
@ -544,64 +559,84 @@ sym(vp9_filter_block1d16_v8_avg_ssse3):
movdqa xmm4, [rdx] ;load filters movdqa xmm4, [rdx] ;load filters
movq xmm5, rcx movq xmm5, rcx
packsswb xmm4, xmm4 packsswb xmm4, xmm4
pshuflw xmm0, xmm4, 0b ;k0_k1 pshuflw xmm6, xmm4, 0b ;k0_k1
pshuflw xmm1, xmm4, 01010101b ;k2_k3 pshufhw xmm6, xmm6, 10101010b ;k0_k1_k4_k5
pshuflw xmm2, xmm4, 10101010b ;k4_k5 pshuflw xmm7, xmm4, 01010101b ;k2_k3
pshuflw xmm3, xmm4, 11111111b ;k6_k7 pshufhw xmm7, xmm7, 11111111b ;k2_k3_k6_k7
pshufd xmm5, xmm5, 0 ;rounding
punpcklqdq xmm0, xmm0
punpcklqdq xmm1, xmm1
punpcklqdq xmm2, xmm2
punpcklqdq xmm3, xmm3
movdqa k0k1, xmm0
movdqa k2k3, xmm1
pshufd xmm5, xmm5, 0
movdqa k4k5, xmm2
movdqa k6k7, xmm3
movdqa krd, xmm5
movsxd rax, dword ptr arg(1) ;src_pixels_per_line movsxd rax, dword ptr arg(1) ;src_pixels_per_line
movsxd rdx, dword ptr arg(3) ;output_pitch movsxd rdx, dword ptr arg(3) ;output_pitch
movsxd rcx, dword ptr arg(4) ;output_height movsxd rcx, dword ptr arg(4) ;output_height
shr rcx, 1
.loop: .loop:
movq xmm0, [rsi - 3] ; -3 -2 -1 0 1 2 3 4 ;Do two rows once
movq xmm0, [rsi - 3] ;load src
movq xmm1, [rsi + 5]
movq xmm2, [rsi + rax - 3]
movq xmm3, [rsi + rax + 5]
punpcklqdq xmm0, xmm1
punpcklqdq xmm2, xmm3
movq xmm3, [rsi + 5] ; 5 6 7 8 9 10 11 12 HORIZx4_ROW xmm0, xmm1
punpcklqdq xmm0, xmm3 HORIZx4_ROW xmm2, xmm3
%if %1
movd xmm1, [rdi]
pavgb xmm0, xmm1
movd xmm3, [rdi + rdx]
pavgb xmm2, xmm3
%endif
movd [rdi], xmm0
movd [rdi +rdx], xmm2
movdqa xmm1, xmm0 lea rsi, [rsi + rax]
pshufb xmm0, [GLOBAL(shuf_t0t1)] prefetcht0 [rsi + 4 * rax - 3]
pmaddubsw xmm0, k0k1 lea rsi, [rsi + rax]
lea rdi, [rdi + 2 * rdx]
prefetcht0 [rsi + 2 * rax - 3]
movdqa xmm2, xmm1 dec rcx
pshufb xmm1, [GLOBAL(shuf_t2t3)] jnz .loop
pmaddubsw xmm1, k2k3
movdqa xmm4, xmm2 ; Do last row if output_height is odd
pshufb xmm2, [GLOBAL(shuf_t4t5)] movsxd rcx, dword ptr arg(4) ;output_height
pmaddubsw xmm2, k4k5 and rcx, 1
je .done
pshufb xmm4, [GLOBAL(shuf_t6t7)] movq xmm0, [rsi - 3] ; load src
pmaddubsw xmm4, k6k7 movq xmm1, [rsi + 5]
punpcklqdq xmm0, xmm1
paddsw xmm0, xmm1 HORIZx4_ROW xmm0, xmm1
paddsw xmm0, xmm4
paddsw xmm0, xmm2
paddsw xmm0, krd
psraw xmm0, 7
packuswb xmm0, xmm0
%if %1 %if %1
movd xmm1, [rdi] movd xmm1, [rdi]
pavgb xmm0, xmm1 pavgb xmm0, xmm1
%endif %endif
lea rsi, [rsi + rax]
movd [rdi], xmm0 movd [rdi], xmm0
.done
%endm
lea rdi, [rdi + rdx] %macro HORIZx8_ROW 4
dec rcx movdqa %2, %1
jnz .loop movdqa %3, %1
movdqa %4, %1
pshufb %1, [GLOBAL(shuf_t0t1)]
pshufb %2, [GLOBAL(shuf_t2t3)]
pshufb %3, [GLOBAL(shuf_t4t5)]
pshufb %4, [GLOBAL(shuf_t6t7)]
pmaddubsw %1, k0k1
pmaddubsw %2, k2k3
pmaddubsw %3, k4k5
pmaddubsw %4, k6k7
paddsw %1, %2
paddsw %1, %4
paddsw %1, %3
paddsw %1, krd
psraw %1, 7
packuswb %1, %1
%endm %endm
%macro HORIZx8 1 %macro HORIZx8 1
@ -633,45 +668,51 @@ sym(vp9_filter_block1d16_v8_avg_ssse3):
movsxd rax, dword ptr arg(1) ;src_pixels_per_line movsxd rax, dword ptr arg(1) ;src_pixels_per_line
movsxd rdx, dword ptr arg(3) ;output_pitch movsxd rdx, dword ptr arg(3) ;output_pitch
movsxd rcx, dword ptr arg(4) ;output_height movsxd rcx, dword ptr arg(4) ;output_height
shr rcx, 1
.loop: .loop:
movq xmm0, [rsi - 3] ; -3 -2 -1 0 1 2 3 4 movq xmm0, [rsi - 3] ;load src
movq xmm3, [rsi + 5]
movq xmm4, [rsi + rax - 3]
movq xmm7, [rsi + rax + 5]
punpcklqdq xmm0, xmm3
punpcklqdq xmm4, xmm7
movq xmm3, [rsi + 5] ; 5 6 7 8 9 10 11 12 HORIZx8_ROW xmm0, xmm1, xmm2, xmm3
HORIZx8_ROW xmm4, xmm5, xmm6, xmm7
%if %1
movq xmm1, [rdi]
movq xmm2, [rdi + rdx]
pavgb xmm0, xmm1
pavgb xmm4, xmm2
%endif
movq [rdi], xmm0
movq [rdi + rdx], xmm4
lea rsi, [rsi + rax]
prefetcht0 [rsi + 4 * rax - 3]
lea rsi, [rsi + rax]
lea rdi, [rdi + 2 * rdx]
prefetcht0 [rsi + 2 * rax - 3]
dec rcx
jnz .loop
;Do last row if output_height is odd
movsxd rcx, dword ptr arg(4) ;output_height
and rcx, 1
je .done
movq xmm0, [rsi - 3]
movq xmm3, [rsi + 5]
punpcklqdq xmm0, xmm3 punpcklqdq xmm0, xmm3
movdqa xmm1, xmm0 HORIZx8_ROW xmm0, xmm1, xmm2, xmm3
pshufb xmm0, [GLOBAL(shuf_t0t1)]
pmaddubsw xmm0, k0k1
movdqa xmm2, xmm1
pshufb xmm1, [GLOBAL(shuf_t2t3)]
pmaddubsw xmm1, k2k3
movdqa xmm4, xmm2
pshufb xmm2, [GLOBAL(shuf_t4t5)]
pmaddubsw xmm2, k4k5
pshufb xmm4, [GLOBAL(shuf_t6t7)]
pmaddubsw xmm4, k6k7
paddsw xmm0, xmm1
paddsw xmm0, xmm4
paddsw xmm0, xmm2
paddsw xmm0, krd
psraw xmm0, 7
packuswb xmm0, xmm0
%if %1 %if %1
movq xmm1, [rdi] movq xmm1, [rdi]
pavgb xmm0, xmm1 pavgb xmm0, xmm1
%endif %endif
lea rsi, [rsi + rax]
movq [rdi], xmm0 movq [rdi], xmm0
.done
lea rdi, [rdi + rdx]
dec rcx
jnz .loop
%endm %endm
%macro HORIZx16 1 %macro HORIZx16 1
@ -785,19 +826,8 @@ sym(vp9_filter_block1d4_h8_ssse3):
push rdi push rdi
; end prolog ; end prolog
ALIGN_STACK 16, rax
sub rsp, 16*5
%define k0k1 [rsp + 16*0]
%define k2k3 [rsp + 16*1]
%define k4k5 [rsp + 16*2]
%define k6k7 [rsp + 16*3]
%define krd [rsp + 16*4]
HORIZx4 0 HORIZx4 0
add rsp, 16*5
pop rsp
; begin epilog ; begin epilog
pop rdi pop rdi
pop rsi pop rsi
@ -902,19 +932,8 @@ sym(vp9_filter_block1d4_h8_avg_ssse3):
push rdi push rdi
; end prolog ; end prolog
ALIGN_STACK 16, rax
sub rsp, 16*5
%define k0k1 [rsp + 16*0]
%define k2k3 [rsp + 16*1]
%define k4k5 [rsp + 16*2]
%define k6k7 [rsp + 16*3]
%define krd [rsp + 16*4]
HORIZx4 1 HORIZx4 1
add rsp, 16*5
pop rsp
; begin epilog ; begin epilog
pop rdi pop rdi
pop rsi pop rsi