The MSA optimization has been refined in commit 93218c2 and ce0a52e.
It is better than MMI version now.
Speed of decoding H264: 4.83x ==> 4.89x (tested on 3A4000).
---
libavcodec/mips/h264chroma_init_mips.c | 19 +++++++++----------
1 file changed, 9 insertions(+), 10 deletions(-)
diff --git a/libavcodec/mips/h264chroma_init_mips.c
b/libavcodec/mips/h264chroma_init_mips.c
index 6bb19d3..755cc04 100644
--- a/libavcodec/mips/h264chroma_init_mips.c
+++ b/libavcodec/mips/h264chroma_init_mips.c
@@ -28,7 +28,15 @@ av_cold void ff_h264chroma_init_mips(H264ChromaContext *c,
int bit_depth)
int cpu_flags = av_get_cpu_flags();
int high_bit_depth = bit_depth > 8;
- /* MMI apears to be faster than MSA here */
+ if (have_mmi(cpu_flags)) {
+ if (!high_bit_depth) {
+ c->put_h264_chroma_pixels_tab[0] = ff_put_h264_chroma_mc8_mmi;
+ c->avg_h264_chroma_pixels_tab[0] = ff_avg_h264_chroma_mc8_mmi;
+ c->put_h264_chroma_pixels_tab[1] = ff_put_h264_chroma_mc4_mmi;
+ c->avg_h264_chroma_pixels_tab[1] = ff_avg_h264_chroma_mc4_mmi;
+ }
+ }
+
if (have_msa(cpu_flags)) {
if (!high_bit_depth) {
c->put_h264_chroma_pixels_tab[0] = ff_put_h264_chroma_mc8_msa;
@@ -40,13 +48,4 @@ av_cold void ff_h264chroma_init_mips(H264ChromaContext *c,
int bit_depth)
c->avg_h264_chroma_pixels_tab[2] = ff_avg_h264_chroma_mc2_msa;
}
}
-
- if (have_mmi(cpu_flags)) {
- if (!high_bit_depth) {
- c->put_h264_chroma_pixels_tab[0] = ff_put_h264_chroma_mc8_mmi;
- c->avg_h264_chroma_pixels_tab[0] = ff_avg_h264_chroma_mc8_mmi;
- c->put_h264_chroma_pixels_tab[1] = ff_put_h264_chroma_mc4_mmi;
- c->avg_h264_chroma_pixels_tab[1] = ff_avg_h264_chroma_mc4_mmi;
- }
- }
}
--
2.1.0
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
