Skip to content

F
ffmpeg.wasm-core
Commit e25be471 authored by Ronald S. Bultje's avatar Ronald S. Bultje
Browse files
Options

vp8: convert idct/mc x86 assembly to use cpuflags().

parent 8249a23f
Show whitespace changes
Inline Side-by-side
Showing with 124 additions and 116 deletions
libavcodec/x86/vp8dsp-init.c
View file @ e25be471
......@@ -29,16 +29,16 @@
/*
* MC functions
*/
extern void ff_put_vp8_epel4_h4_mmxext(uint8_t *dst, ptrdiff_t dststride,
extern void ff_put_vp8_epel4_h4_mmx2 (uint8_t *dst, ptrdiff_t dststride,
uint8_t *src, ptrdiff_t srcstride,
int height, int mx, int my);
extern void ff_put_vp8_epel4_h6_mmxext(uint8_t *dst, ptrdiff_t dststride,
extern void ff_put_vp8_epel4_h6_mmx2 (uint8_t *dst, ptrdiff_t dststride,
uint8_t *src, ptrdiff_t srcstride,
int height, int mx, int my);
extern void ff_put_vp8_epel4_v4_mmxext(uint8_t *dst, ptrdiff_t dststride,
extern void ff_put_vp8_epel4_v4_mmx2 (uint8_t *dst, ptrdiff_t dststride,
uint8_t *src, ptrdiff_t srcstride,
int height, int mx, int my);
extern void ff_put_vp8_epel4_v6_mmxext(uint8_t *dst, ptrdiff_t dststride,
extern void ff_put_vp8_epel4_v6_mmx2 (uint8_t *dst, ptrdiff_t dststride,
uint8_t *src, ptrdiff_t srcstride,
int height, int mx, int my);
......@@ -80,7 +80,7 @@ extern void ff_put_vp8_epel8_v6_ssse3 (uint8_t *dst, ptrdiff_t dststride,
uint8_t *src, ptrdiff_t srcstride,
int height, int mx, int my);
extern void ff_put_vp8_bilinear4_h_mmxext(uint8_t *dst, ptrdiff_t dststride,
extern void ff_put_vp8_bilinear4_h_mmx2 (uint8_t *dst, ptrdiff_t dststride,
uint8_t *src, ptrdiff_t srcstride,
int height, int mx, int my);
extern void ff_put_vp8_bilinear8_h_sse2 (uint8_t *dst, ptrdiff_t dststride,
......@@ -93,7 +93,7 @@ extern void ff_put_vp8_bilinear8_h_ssse3 (uint8_t *dst, ptrdiff_t dststride,
uint8_t *src, ptrdiff_t srcstride,
int height, int mx, int my);
extern void ff_put_vp8_bilinear4_v_mmxext(uint8_t *dst, ptrdiff_t dststride,
extern void ff_put_vp8_bilinear4_v_mmx2 (uint8_t *dst, ptrdiff_t dststride,
uint8_t *src, ptrdiff_t srcstride,
int height, int mx, int my);
extern void ff_put_vp8_bilinear8_v_sse2 (uint8_t *dst, ptrdiff_t dststride,
......@@ -139,16 +139,16 @@ static void ff_put_vp8_ ## FILTERTYPE ## 8_ ## TAPTYPE ## _ ## OPT( \
}
#if ARCH_X86_32
TAP_W8 (mmxext, epel, h4)
TAP_W8 (mmxext, epel, h6)
TAP_W16(mmxext, epel, h6)
TAP_W8 (mmxext, epel, v4)
TAP_W8 (mmxext, epel, v6)
TAP_W16(mmxext, epel, v6)
TAP_W8 (mmxext, bilinear, h)
TAP_W16(mmxext, bilinear, h)
TAP_W8 (mmxext, bilinear, v)
TAP_W16(mmxext, bilinear, v)
TAP_W8 (mmx2, epel, h4)
TAP_W8 (mmx2, epel, h6)
TAP_W16(mmx2, epel, h6)
TAP_W8 (mmx2, epel, v4)
TAP_W8 (mmx2, epel, v6)
TAP_W16(mmx2, epel, v6)
TAP_W8 (mmx2, bilinear, h)
TAP_W16(mmx2, bilinear, h)
TAP_W8 (mmx2, bilinear, v)
TAP_W16(mmx2, bilinear, v)
#endif
TAP_W16(sse2, epel, h6)
......@@ -177,13 +177,13 @@ static void ff_put_vp8_epel ## SIZE ## _h ## TAPNUMX ## v ## TAPNUMY ## _ ## OPT
#if ARCH_X86_32
#define HVTAPMMX(x, y) \
HVTAP(mmxext, 8, x, y, 4, 8) \
HVTAP(mmxext, 8, x, y, 8, 16)
HVTAP(mmx2, 8, x, y, 4, 8) \
HVTAP(mmx2, 8, x, y, 8, 16)
HVTAP(mmxext, 8, 6, 6, 16, 16)
HVTAP(mmx2, 8, 6, 6, 16, 16)
#else
#define HVTAPMMX(x, y) \
HVTAP(mmxext, 8, x, y, 4, 8)
HVTAP(mmx2, 8, x, y, 4, 8)
#endif
HVTAPMMX(4, 4)
......@@ -218,10 +218,10 @@ static void ff_put_vp8_bilinear ## SIZE ## _hv_ ## OPT( \
dst, dststride, tmp, SIZE, height, mx, my); \
}
HVBILIN(mmxext, 8, 4, 8)
HVBILIN(mmx2, 8, 4, 8)
#if ARCH_X86_32
HVBILIN(mmxext, 8, 8, 16)
HVBILIN(mmxext, 8, 16, 16)
HVBILIN(mmx2, 8, 8, 16)
HVBILIN(mmx2, 8, 16, 16)
#endif
HVBILIN(sse2, 8, 8, 16)
HVBILIN(sse2, 8, 16, 16)
......@@ -283,7 +283,7 @@ extern void ff_vp8_h_loop_filter8uv_mbedge_ ## NAME(uint8_t *dstU, \
int e, int i, int hvt);
DECLARE_LOOP_FILTER(mmx)
DECLARE_LOOP_FILTER(mmxext)
DECLARE_LOOP_FILTER(mmx2)
DECLARE_LOOP_FILTER(sse2)
DECLARE_LOOP_FILTER(ssse3)
DECLARE_LOOP_FILTER(sse4)
......@@ -351,26 +351,26 @@ av_cold void ff_vp8dsp_init_x86(VP8DSPContext* c)
/* note that 4-tap width=16 functions are missing because w=16
* is only used for luma, and luma is always a copy or sixtap. */
if (mm_flags & AV_CPU_FLAG_MMX2) {
VP8_MC_FUNC(2, 4, mmxext);
VP8_BILINEAR_MC_FUNC(2, 4, mmxext);
VP8_MC_FUNC(2, 4, mmx2);
VP8_BILINEAR_MC_FUNC(2, 4, mmx2);
#if ARCH_X86_32
VP8_LUMA_MC_FUNC(0, 16, mmxext);
VP8_MC_FUNC(1, 8, mmxext);
VP8_BILINEAR_MC_FUNC(0, 16, mmxext);
VP8_BILINEAR_MC_FUNC(1, 8, mmxext);
c->vp8_v_loop_filter_simple = ff_vp8_v_loop_filter_simple_mmxext;
c->vp8_h_loop_filter_simple = ff_vp8_h_loop_filter_simple_mmxext;
c->vp8_v_loop_filter16y_inner = ff_vp8_v_loop_filter16y_inner_mmxext;
c->vp8_h_loop_filter16y_inner = ff_vp8_h_loop_filter16y_inner_mmxext;
c->vp8_v_loop_filter8uv_inner = ff_vp8_v_loop_filter8uv_inner_mmxext;
c->vp8_h_loop_filter8uv_inner = ff_vp8_h_loop_filter8uv_inner_mmxext;
c->vp8_v_loop_filter16y = ff_vp8_v_loop_filter16y_mbedge_mmxext;
c->vp8_h_loop_filter16y = ff_vp8_h_loop_filter16y_mbedge_mmxext;
c->vp8_v_loop_filter8uv = ff_vp8_v_loop_filter8uv_mbedge_mmxext;
c->vp8_h_loop_filter8uv = ff_vp8_h_loop_filter8uv_mbedge_mmxext;
VP8_LUMA_MC_FUNC(0, 16, mmx2);
VP8_MC_FUNC(1, 8, mmx2);
VP8_BILINEAR_MC_FUNC(0, 16, mmx2);
VP8_BILINEAR_MC_FUNC(1, 8, mmx2);
c->vp8_v_loop_filter_simple = ff_vp8_v_loop_filter_simple_mmx2;
c->vp8_h_loop_filter_simple = ff_vp8_h_loop_filter_simple_mmx2;
c->vp8_v_loop_filter16y_inner = ff_vp8_v_loop_filter16y_inner_mmx2;
c->vp8_h_loop_filter16y_inner = ff_vp8_h_loop_filter16y_inner_mmx2;
c->vp8_v_loop_filter8uv_inner = ff_vp8_v_loop_filter8uv_inner_mmx2;
c->vp8_h_loop_filter8uv_inner = ff_vp8_h_loop_filter8uv_inner_mmx2;
c->vp8_v_loop_filter16y = ff_vp8_v_loop_filter16y_mbedge_mmx2;
c->vp8_h_loop_filter16y = ff_vp8_h_loop_filter16y_mbedge_mmx2;
c->vp8_v_loop_filter8uv = ff_vp8_v_loop_filter8uv_mbedge_mmx2;
c->vp8_h_loop_filter8uv = ff_vp8_h_loop_filter8uv_mbedge_mmx2;
#endif
}
......
libavcodec/x86/vp8dsp.asm
View file @ e25be471
......@@ -173,8 +173,8 @@ SECTION .text
; int height, int mx, int my);
;-----------------------------------------------------------------------------
%macro FILTER_SSSE3 3
cglobal put_vp8_epel%1_h6_ssse3, 6, 6, %2
%macro FILTER_SSSE3 1
cglobal put_vp8_epel%1_h6, 6, 6, 8
lea r5d, [r5*3]
mova m3, [filter_h6_shuf2]
mova m4, [filter_h6_shuf3]
......@@ -189,7 +189,7 @@ cglobal put_vp8_epel%1_h6_ssse3, 6, 6, %2
movu m0, [r2-2]
mova m1, m0
mova m2, m0
%ifidn %1, 4
%if mmsize == 8
; For epel4, we need 9 bytes, but only 8 get loaded; to compensate, do the
; shuffle with a memory operand
punpcklbw m0, [r2+3]
......@@ -215,7 +215,7 @@ cglobal put_vp8_epel%1_h6_ssse3, 6, 6, %2
jg .nextrow
REP_RET
cglobal put_vp8_epel%1_h4_ssse3, 6, 6, %3
cglobal put_vp8_epel%1_h4, 6, 6, 7
shl r5d, 4
mova m2, [pw_64]
mova m3, [filter_h2_shuf]
......@@ -246,7 +246,7 @@ cglobal put_vp8_epel%1_h4_ssse3, 6, 6, %3
jg .nextrow
REP_RET
cglobal put_vp8_epel%1_v4_ssse3, 7, 7, %2
cglobal put_vp8_epel%1_v4, 7, 7, 8
shl r6d, 4
%ifdef PIC
lea r11, [fourtap_filter_hb_m]
......@@ -285,7 +285,7 @@ cglobal put_vp8_epel%1_v4_ssse3, 7, 7, %2
jg .nextrow
REP_RET
cglobal put_vp8_epel%1_v6_ssse3, 7, 7, %2
cglobal put_vp8_epel%1_v6, 7, 7, 8
lea r6d, [r6*3]
%ifdef PIC
lea r11, [sixtap_filter_hb_m]
......@@ -333,13 +333,14 @@ cglobal put_vp8_epel%1_v6_ssse3, 7, 7, %2
REP_RET
%endmacro
INIT_MMX
FILTER_SSSE3 4, 0, 0
INIT_XMM
FILTER_SSSE3 8, 8, 7
INIT_MMX ssse3
FILTER_SSSE3 4
INIT_XMM ssse3
FILTER_SSSE3 8
; 4x4 block, H-only 4-tap filter
cglobal put_vp8_epel4_h4_mmxext, 6, 6
INIT_MMX mmx2
cglobal put_vp8_epel4_h4, 6, 6
shl r5d, 4
%ifdef PIC
lea r11, [fourtap_filter_hw_m]
......@@ -386,7 +387,8 @@ cglobal put_vp8_epel4_h4_mmxext, 6, 6
REP_RET
; 4x4 block, H-only 6-tap filter
cglobal put_vp8_epel4_h6_mmxext, 6, 6
INIT_MMX mmx2
cglobal put_vp8_epel4_h6, 6, 6
lea r5d, [r5*3]
%ifdef PIC
lea r11, [sixtap_filter_hw_m]
......@@ -442,8 +444,8 @@ cglobal put_vp8_epel4_h6_mmxext, 6, 6
jg .nextrow
REP_RET
INIT_XMM
cglobal put_vp8_epel8_h4_sse2, 6, 6, 10
INIT_XMM sse2
cglobal put_vp8_epel8_h4, 6, 6, 10
shl r5d, 5
%ifdef PIC
lea r11, [fourtap_filter_v_m]
......@@ -490,7 +492,8 @@ cglobal put_vp8_epel8_h4_sse2, 6, 6, 10
jg .nextrow
REP_RET
cglobal put_vp8_epel8_h6_sse2, 6, 6, 14
INIT_XMM sse2
cglobal put_vp8_epel8_h6, 6, 6, 14
lea r5d, [r5*3]
shl r5d, 4
%ifdef PIC
......@@ -552,9 +555,9 @@ cglobal put_vp8_epel8_h6_sse2, 6, 6, 14
jg .nextrow
REP_RET
%macro FILTER_V 3
%macro FILTER_V 1
; 4x4 block, V-only 4-tap filter
cglobal put_vp8_epel%2_v4_%1, 7, 7, %3
cglobal put_vp8_epel%1_v4, 7, 7, 8
shl r6d, 5
%ifdef PIC
lea r11, [fourtap_filter_v_m]
......@@ -607,7 +610,7 @@ cglobal put_vp8_epel%2_v4_%1, 7, 7, %3
; 4x4 block, V-only 6-tap filter
cglobal put_vp8_epel%2_v6_%1, 7, 7, %3
cglobal put_vp8_epel%1_v6, 7, 7, 8
shl r6d, 4
lea r6, [r6*3]
%ifdef PIC
......@@ -671,13 +674,13 @@ cglobal put_vp8_epel%2_v6_%1, 7, 7, %3
REP_RET
%endmacro
INIT_MMX
FILTER_V mmxext, 4, 0
INIT_XMM
FILTER_V sse2, 8, 8
INIT_MMX mmx2
FILTER_V 4
INIT_XMM sse2
FILTER_V 8
%macro FILTER_BILINEAR 3
cglobal put_vp8_bilinear%2_v_%1, 7,7,%3
%macro FILTER_BILINEAR 1
cglobal put_vp8_bilinear%1_v, 7, 7, 7
mov r5d, 8*16
shl r6d, 4
sub r5d, r6d
......@@ -705,7 +708,7 @@ cglobal put_vp8_bilinear%2_v_%1, 7,7,%3
psraw m2, 2
pavgw m0, m6
pavgw m2, m6
%ifidn %1, mmxext
%if mmsize == 8
packuswb m0, m0
packuswb m2, m2
movh [r0+r1*0], m0
......@@ -722,7 +725,7 @@ cglobal put_vp8_bilinear%2_v_%1, 7,7,%3
jg .nextrow
REP_RET
cglobal put_vp8_bilinear%2_h_%1, 7,7,%3
cglobal put_vp8_bilinear%1_h, 7, 7, 7
mov r6d, 8*16
shl r5d, 4
sub r6d, r5d
......@@ -751,7 +754,7 @@ cglobal put_vp8_bilinear%2_h_%1, 7,7,%3
psraw m2, 2
pavgw m0, m6
pavgw m2, m6
%ifidn %1, mmxext
%if mmsize == 8
packuswb m0, m0
packuswb m2, m2
movh [r0+r1*0], m0
......@@ -769,13 +772,13 @@ cglobal put_vp8_bilinear%2_h_%1, 7,7,%3
REP_RET
%endmacro
INIT_MMX
FILTER_BILINEAR mmxext, 4, 0
INIT_XMM
FILTER_BILINEAR sse2, 8, 7
INIT_MMX mmx2
FILTER_BILINEAR 4
INIT_XMM sse2
FILTER_BILINEAR 8
%macro FILTER_BILINEAR_SSSE3 1
cglobal put_vp8_bilinear%1_v_ssse3, 7,7
cglobal put_vp8_bilinear%1_v, 7, 7, 5
shl r6d, 4
%ifdef PIC
lea r11, [bilinear_filter_vb_m]
......@@ -811,7 +814,7 @@ cglobal put_vp8_bilinear%1_v_ssse3, 7,7
jg .nextrow
REP_RET
cglobal put_vp8_bilinear%1_h_ssse3, 7,7
cglobal put_vp8_bilinear%1_h, 7, 7, 5
shl r5d, 4
%ifdef PIC
lea r11, [bilinear_filter_vb_m]
......@@ -848,12 +851,13 @@ cglobal put_vp8_bilinear%1_h_ssse3, 7,7
REP_RET
%endmacro
INIT_MMX
INIT_MMX ssse3
FILTER_BILINEAR_SSSE3 4
INIT_XMM
INIT_XMM ssse3
FILTER_BILINEAR_SSSE3 8
cglobal put_vp8_pixels8_mmx, 5,5
INIT_MMX mmx
cglobal put_vp8_pixels8, 5,5
.nextrow:
movq mm0, [r2+r3*0]
movq mm1, [r2+r3*1]
......@@ -866,7 +870,8 @@ cglobal put_vp8_pixels8_mmx, 5,5
REP_RET
%if ARCH_X86_32
cglobal put_vp8_pixels16_mmx, 5,5
INIT_MMX mmx
cglobal put_vp8_pixels16, 5,5
.nextrow:
movq mm0, [r2+r3*0+0]
movq mm1, [r2+r3*0+8]
......@@ -883,7 +888,8 @@ cglobal put_vp8_pixels16_mmx, 5,5
REP_RET
%endif
cglobal put_vp8_pixels16_sse, 5,5,2
INIT_XMM sse
cglobal put_vp8_pixels16, 5,5,2
.nextrow:
movups xmm0, [r2+r3*0]
movups xmm1, [r2+r3*1]
......@@ -918,8 +924,8 @@ cglobal put_vp8_pixels16_sse, 5,5,2
%4 [r1+r2+%3], m5
%endmacro
INIT_MMX
cglobal vp8_idct_dc_add_mmx, 3, 3
INIT_MMX mmx
cglobal vp8_idct_dc_add, 3, 3
; load data
movd m0, [r1]
......@@ -941,8 +947,8 @@ cglobal vp8_idct_dc_add_mmx, 3, 3
ADD_DC m0, m1, 0, movh
RET
INIT_XMM
cglobal vp8_idct_dc_add_sse4, 3, 3, 6
INIT_XMM sse4
cglobal vp8_idct_dc_add, 3, 3, 6
; load data
movd m0, [r1]
pxor m1, m1
......@@ -976,8 +982,8 @@ cglobal vp8_idct_dc_add_sse4, 3, 3, 6
;-----------------------------------------------------------------------------
%if ARCH_X86_32
INIT_MMX
cglobal vp8_idct_dc_add4y_mmx, 3, 3
INIT_MMX mmx
cglobal vp8_idct_dc_add4y, 3, 3
; load data
movd m0, [r1+32*0] ; A
movd m1, [r1+32*2] ; C
......@@ -1012,8 +1018,8 @@ cglobal vp8_idct_dc_add4y_mmx, 3, 3
RET
%endif
INIT_XMM
cglobal vp8_idct_dc_add4y_sse2, 3, 3, 6
INIT_XMM sse2
cglobal vp8_idct_dc_add4y, 3, 3, 6
; load data
movd m0, [r1+32*0] ; A
movd m1, [r1+32*2] ; C
......@@ -1046,8 +1052,8 @@ cglobal vp8_idct_dc_add4y_sse2, 3, 3, 6
; void vp8_idct_dc_add4uv_<opt>(uint8_t *dst, DCTELEM block[4][16], int stride);
;-----------------------------------------------------------------------------
INIT_MMX
cglobal vp8_idct_dc_add4uv_mmx, 3, 3
INIT_MMX mmx
cglobal vp8_idct_dc_add4uv, 3, 3
; load data
movd m0, [r1+32*0] ; A
movd m1, [r1+32*2] ; C
......@@ -1118,9 +1124,8 @@ cglobal vp8_idct_dc_add4uv_mmx, 3, 3
SWAP %4, %3
%endmacro
INIT_MMX
%macro VP8_IDCT_ADD 1
cglobal vp8_idct_add_%1, 3, 3
%macro VP8_IDCT_ADD 0
cglobal vp8_idct_add, 3, 3
; load block data
movq m0, [r1+ 0]
movq m1, [r1+ 8]
......@@ -1128,7 +1133,7 @@ cglobal vp8_idct_add_%1, 3, 3
movq m3, [r1+24]
movq m6, [pw_20091]
movq m7, [pw_17734]
%ifidn %1, sse
%if cpuflag(sse)
xorps xmm0, xmm0
movaps [r1+ 0], xmm0
movaps [r1+16], xmm0
......@@ -1157,9 +1162,11 @@ cglobal vp8_idct_add_%1, 3, 3
%endmacro
%if ARCH_X86_32
VP8_IDCT_ADD mmx
INIT_MMX mmx
VP8_IDCT_ADD
%endif
VP8_IDCT_ADD sse
INIT_MMX sse
VP8_IDCT_ADD
;-----------------------------------------------------------------------------
; void vp8_luma_dc_wht_mmxext(DCTELEM block[4][4][16], DCTELEM dc[16])
......@@ -1192,13 +1199,13 @@ VP8_IDCT_ADD sse
SWAP %1, %4, %3
%endmacro
%macro VP8_DC_WHT 1
cglobal vp8_luma_dc_wht_%1, 2,3
%macro VP8_DC_WHT 0
cglobal vp8_luma_dc_wht, 2, 3
movq m0, [r1]
movq m1, [r1+8]
movq m2, [r1+16]
movq m3, [r1+24]
%ifidn %1, sse
%if cpuflag(sse)
xorps xmm0, xmm0
movaps [r1+ 0], xmm0
movaps [r1+16], xmm0
......@@ -1222,11 +1229,12 @@ cglobal vp8_luma_dc_wht_%1, 2,3
RET
%endmacro
INIT_MMX
%if ARCH_X86_32
VP8_DC_WHT mmx
INIT_MMX mmx
VP8_DC_WHT
%endif
VP8_DC_WHT sse
INIT_MMX sse
VP8_DC_WHT
;-----------------------------------------------------------------------------
; void vp8_h/v_loop_filter_simple_<opt>(uint8_t *dst, int stride, int flim);
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment