diff options
| author | gxw <guxiwei-hf@loongson.cn> | 2019-08-07 17:52:00 +0800 |
|---|---|---|
| committer | Michael Niedermayer <michael@niedermayer.cc> | 2019-08-13 16:48:38 +0200 |
| commit | a3e572d96fd1dd6291f6b28e173db858c08ff8d8 (patch) | |
| tree | 85807f6ec1442cc362cf8946e67f564c92267e07 /libavcodec/mips/hevcpred_msa.c | |
| parent | 8f92eb05e063e6c4d6e36521020620d4e6e1c21d (diff) | |
avutil/mips: refine msa macros CLIP_*.
Changing details as following:
1. Remove the local variable 'out_m' in 'CLIP_SH' and store the result in
source vector.
2. Refine the implementation of macro 'CLIP_SH_0_255' and 'CLIP_SW_0_255'.
Performance of VP8 decoding has speed up about 1.1%(from 7.03x to 7.11x).
Performance of H264 decoding has speed up about 0.5%(from 4.35x to 4.37x).
Performance of Theora decoding has speed up about 0.7%(from 5.79x to 5.83x).
3. Remove redundant macro 'CLIP_SH/Wn_0_255_MAX_SATU' and use 'CLIP_SH/Wn_0_255'
instead, because there are no difference in the effect of this two macros.
Reviewed-by: Shiyou Yin <yinshiyou-hf@loongson.cn>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Diffstat (limited to 'libavcodec/mips/hevcpred_msa.c')
| -rw-r--r-- | libavcodec/mips/hevcpred_msa.c | 8 |
1 files changed, 4 insertions, 4 deletions
diff --git a/libavcodec/mips/hevcpred_msa.c b/libavcodec/mips/hevcpred_msa.c index b8df089e0c..930a89cb48 100644 --- a/libavcodec/mips/hevcpred_msa.c +++ b/libavcodec/mips/hevcpred_msa.c @@ -83,7 +83,7 @@ static void hevc_intra_pred_vert_4x4_msa(const uint8_t *src_top, vec2 -= vec0; vec2 >>= 1; vec2 += vec1; - vec2 = CLIP_SH_0_255(vec2); + CLIP_SH_0_255(vec2); for (col = 0; col < 4; col++) { dst[stride * col] = (uint8_t) vec2[col]; @@ -122,7 +122,7 @@ static void hevc_intra_pred_vert_8x8_msa(const uint8_t *src_top, vec2 -= vec0; vec2 >>= 1; vec2 += vec1; - vec2 = CLIP_SH_0_255(vec2); + CLIP_SH_0_255(vec2); val0 = vec2[0]; val1 = vec2[1]; @@ -214,7 +214,7 @@ static void hevc_intra_pred_horiz_4x4_msa(const uint8_t *src_top, src0_r -= src_top_val; src0_r >>= 1; src0_r += src_left_val; - src0_r = CLIP_SH_0_255(src0_r); + CLIP_SH_0_255(src0_r); src0 = __msa_pckev_b((v16i8) src0_r, (v16i8) src0_r); val0 = __msa_copy_s_w((v4i32) src0, 0); SW(val0, dst); @@ -254,7 +254,7 @@ static void hevc_intra_pred_horiz_8x8_msa(const uint8_t *src_top, src0_r -= src_top_val; src0_r >>= 1; src0_r += src_left_val; - src0_r = CLIP_SH_0_255(src0_r); + CLIP_SH_0_255(src0_r); src0 = __msa_pckev_b((v16i8) src0_r, (v16i8) src0_r); val0 = __msa_copy_s_d((v2i64) src0, 0); SD(val0, dst); |