Skip to content

F
ffmpeg
Commit e641c941 authored by Martin Vignali's avatar Martin Vignali
Browse files
Options

avcodec/huffyuvdsp : add add_int16 AVX2 func

parent 6955e884
Hide whitespace changes
Inline Side-by-side
Showing with 12 additions and 2 deletions
libavcodec/x86/huffyuvdsp.asm
View file @ e641c941
...@@ -53,6 +53,11 @@ ADD_INT16 ...@@ -53,6 +53,11 @@ ADD_INT16
INIT_XMM sse2 INIT_XMM sse2
ADD_INT16 ADD_INT16
%if HAVE_AVX2_EXTERNAL
INIT_YMM avx2
ADD_INT16
%endif
; void add_hfyu_left_pred_bgr32(uint8_t *dst, const uint8_t *src, ; void add_hfyu_left_pred_bgr32(uint8_t *dst, const uint8_t *src,
; intptr_t w, uint8_t *left) ; intptr_t w, uint8_t *left)
%macro LEFT_BGR32 0 %macro LEFT_BGR32 0
......
libavcodec/x86/huffyuvdsp_init.c
View file @ e641c941
...@@ -28,6 +28,7 @@ ...@@ -28,6 +28,7 @@
void ff_add_int16_mmx(uint16_t *dst, const uint16_t *src, unsigned mask, int w); void ff_add_int16_mmx(uint16_t *dst, const uint16_t *src, unsigned mask, int w);
void ff_add_int16_sse2(uint16_t *dst, const uint16_t *src, unsigned mask, int w); void ff_add_int16_sse2(uint16_t *dst, const uint16_t *src, unsigned mask, int w);
void ff_add_int16_avx2(uint16_t *dst, const uint16_t *src, unsigned mask, int w);
void ff_add_hfyu_left_pred_bgr32_mmx(uint8_t *dst, const uint8_t *src, void ff_add_hfyu_left_pred_bgr32_mmx(uint8_t *dst, const uint8_t *src,
intptr_t w, uint8_t *left); intptr_t w, uint8_t *left);
...@@ -53,4 +54,8 @@ av_cold void ff_huffyuvdsp_init_x86(HuffYUVDSPContext *c, enum AVPixelFormat pix ...@@ -53,4 +54,8 @@ av_cold void ff_huffyuvdsp_init_x86(HuffYUVDSPContext *c, enum AVPixelFormat pix
c->add_int16 = ff_add_int16_sse2; c->add_int16 = ff_add_int16_sse2;
c->add_hfyu_left_pred_bgr32 = ff_add_hfyu_left_pred_bgr32_sse2; c->add_hfyu_left_pred_bgr32 = ff_add_hfyu_left_pred_bgr32_sse2;
} }
if (EXTERNAL_AVX2_FAST(cpu_flags)) {
c->add_int16 = ff_add_int16_avx2;
}
} }
libavcodec/x86/huffyuvdsp_template.asm
View file @ e641c941
...@@ -21,8 +21,8 @@ ...@@ -21,8 +21,8 @@
;****************************************************************************** ;******************************************************************************
%macro INT16_LOOP 2 ; %1 = a/u (aligned/unaligned), %2 = add/sub %macro INT16_LOOP 2 ; %1 = a/u (aligned/unaligned), %2 = add/sub
movd m4, maskd movd xm4, maskd
SPLATW m4, m4 SPLATW m4, xm4
add wd, wd add wd, wd
test wq, 2*mmsize - 1 test wq, 2*mmsize - 1
jz %%.tomainloop jz %%.tomainloop
......
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