aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorMichael Niedermayer <michaelni@gmx.at>2013-12-20 23:33:10 +0100
committerMichael Niedermayer <michaelni@gmx.at>2013-12-20 23:33:15 +0100
commit4fa91b88c6180957781036614fa94a6fc3b7046f (patch)
tree8efb706aacdccbbfce90ea8dfb6b9f50b6102b7f
parent69278d94c4823f5ea988751e34cb2b8e33426b01 (diff)
parentb254490bdabb21bd517c05b1a68717f9952ac8c4 (diff)
downloadffmpeg-4fa91b88c6180957781036614fa94a6fc3b7046f.tar.gz
Merge commit 'b254490bdabb21bd517c05b1a68717f9952ac8c4'
* commit 'b254490bdabb21bd517c05b1a68717f9952ac8c4': vc1: arm: Add NEON no_rnd chroma MC Merged-by: Michael Niedermayer <michaelni@gmx.at>
-rw-r--r--libavcodec/arm/h264cmc_neon.S13
-rw-r--r--libavcodec/arm/vc1dsp_init_neon.c14
2 files changed, 27 insertions, 0 deletions
diff --git a/libavcodec/arm/h264cmc_neon.S b/libavcodec/arm/h264cmc_neon.S
index 112d5d6f86..0bcae11204 100644
--- a/libavcodec/arm/h264cmc_neon.S
+++ b/libavcodec/arm/h264cmc_neon.S
@@ -39,6 +39,9 @@ function ff_\type\()_\codec\()_chroma_mc8_neon, export=1
add r6, r6, r7, lsl #1
vld1.16 {d22[],d23[]}, [r6,:16]
.endif
+ .ifc \codec,vc1
+ vmov.u16 q11, #28
+ .endif
A muls r7, r4, r5
T mul r7, r4, r5
@@ -183,6 +186,9 @@ function ff_\type\()_\codec\()_chroma_mc4_neon, export=1
add r6, r6, r7, lsl #1
vld1.16 {d22[],d23[]}, [r6,:16]
.endif
+ .ifc \codec,vc1
+ vmov.u16 q11, #28
+ .endif
A muls r7, r4, r5
T mul r7, r4, r5
@@ -396,3 +402,10 @@ endconst
h264_chroma_mc4 put, rv40
h264_chroma_mc4 avg, rv40
#endif
+
+#if CONFIG_VC1_DECODER
+ h264_chroma_mc8 put, vc1
+ h264_chroma_mc8 avg, vc1
+ h264_chroma_mc4 put, vc1
+ h264_chroma_mc4 avg, vc1
+#endif
diff --git a/libavcodec/arm/vc1dsp_init_neon.c b/libavcodec/arm/vc1dsp_init_neon.c
index b9276b5bd0..f562344b85 100644
--- a/libavcodec/arm/vc1dsp_init_neon.c
+++ b/libavcodec/arm/vc1dsp_init_neon.c
@@ -70,6 +70,15 @@ void ff_put_vc1_mspel_mc32_neon(uint8_t *dst, const uint8_t *src,
void ff_put_vc1_mspel_mc33_neon(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
+void ff_put_vc1_chroma_mc8_neon(uint8_t *dst, uint8_t *src, int stride, int h,
+ int x, int y);
+void ff_avg_vc1_chroma_mc8_neon(uint8_t *dst, uint8_t *src, int stride, int h,
+ int x, int y);
+void ff_put_vc1_chroma_mc4_neon(uint8_t *dst, uint8_t *src, int stride, int h,
+ int x, int y);
+void ff_avg_vc1_chroma_mc4_neon(uint8_t *dst, uint8_t *src, int stride, int h,
+ int x, int y);
+
av_cold void ff_vc1dsp_init_neon(VC1DSPContext *dsp)
{
dsp->vc1_inv_trans_8x8 = ff_vc1_inv_trans_8x8_neon;
@@ -97,4 +106,9 @@ av_cold void ff_vc1dsp_init_neon(VC1DSPContext *dsp)
dsp->put_vc1_mspel_pixels_tab[13] = ff_put_vc1_mspel_mc13_neon;
dsp->put_vc1_mspel_pixels_tab[14] = ff_put_vc1_mspel_mc23_neon;
dsp->put_vc1_mspel_pixels_tab[15] = ff_put_vc1_mspel_mc33_neon;
+
+ dsp->put_no_rnd_vc1_chroma_pixels_tab[0] = ff_put_vc1_chroma_mc8_neon;
+ dsp->avg_no_rnd_vc1_chroma_pixels_tab[0] = ff_avg_vc1_chroma_mc8_neon;
+ dsp->put_no_rnd_vc1_chroma_pixels_tab[1] = ff_put_vc1_chroma_mc4_neon;
+ dsp->avg_no_rnd_vc1_chroma_pixels_tab[1] = ff_avg_vc1_chroma_mc4_neon;
}