Skip to content

F
ffmpeg.wasm-core
Commit 8f6c398d authored by Kaustubh Raste's avatar Kaustubh Raste Committed by Michael Niedermayer
Browse files
Options

avcodec/mips: Improve hevc bi wgt 4 tap hz and vt mc msa functions 

Use global mask buffer for appropriate mask load.
Signed-off-by: 's avatarKaustubh Raste <kaustubh.raste@imgtec.com>
Reviewed-by: 's avatarManojkumar Bhosale <Manojkumar.Bhosale@imgtec.com>
Signed-off-by: 's avatarMichael Niedermayer <michael@niedermayer.cc>
parent e1555eb7
Show whitespace changes
Inline Side-by-side
Showing with 247 additions and 340 deletions
libavcodec/mips/hevc_mc_biw_msa.c
View file @ 8f6c398d
...@@ -2633,22 +2633,21 @@ static void hevc_hz_biwgt_4t_4x2_msa(uint8_t *src0_ptr, ...@@ -2633,22 +2633,21 @@ static void hevc_hz_biwgt_4t_4x2_msa(uint8_t *src0_ptr,
uint8_t *dst, uint8_t *dst,
int32_t dst_stride, int32_t dst_stride,
const int8_t *filter, const int8_t *filter,
int32_t height,
int32_t weight0, int32_t weight0,
int32_t weight1, int32_t weight1,
int32_t offset0, int32_t offset0,
int32_t offset1, int32_t offset1,
int32_t rnd_val) int32_t rnd_val)
{ {
int32_t offset, weight; int32_t offset, weight, constant;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v16i8 src0, src1; v16i8 src0, src1;
v8i16 in0, in1; v8i16 in0, in1;
v16i8 mask0 = { 0, 1, 1, 2, 2, 3, 3, 4, 16, 17, 17, 18, 18, 19, 19, 20 }; v16i8 mask0 = LD_SB(&ff_hevc_mask_arr[16]);
v16i8 mask1, vec0, vec1; v16i8 mask1, vec0, vec1;
v8i16 dst0; v8i16 dst0;
v4i32 dst0_r, dst0_l; v4i32 dst0_r, dst0_l;
v8i16 filter_vec, const_vec; v8i16 out0, filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= 1; src0_ptr -= 1;
...@@ -2661,9 +2660,10 @@ static void hevc_hz_biwgt_4t_4x2_msa(uint8_t *src0_ptr, ...@@ -2661,9 +2660,10 @@ static void hevc_hz_biwgt_4t_4x2_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -2674,18 +2674,16 @@ static void hevc_hz_biwgt_4t_4x2_msa(uint8_t *src0_ptr, ...@@ -2674,18 +2674,16 @@ static void hevc_hz_biwgt_4t_4x2_msa(uint8_t *src0_ptr,
XORI_B2_128_SB(src0, src1); XORI_B2_128_SB(src0, src1);
VSHF_B2_SB(src0, src1, src0, src1, mask0, mask1, vec0, vec1); VSHF_B2_SB(src0, src1, src0, src1, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
ILVRL_H2_SW(dst0, in0, dst0_r, dst0_l); ILVRL_H2_SW(dst0, in0, dst0_r, dst0_l);
dst0_r = __msa_dpadd_s_w(offset_vec, (v8i16) dst0_r, (v8i16) weight_vec); dst0_r = __msa_dpadd_s_w(offset_vec, (v8i16) dst0_r, (v8i16) weight_vec);
dst0_l = __msa_dpadd_s_w(offset_vec, (v8i16) dst0_l, (v8i16) weight_vec); dst0_l = __msa_dpadd_s_w(offset_vec, (v8i16) dst0_l, (v8i16) weight_vec);
SRAR_W2_SW(dst0_r, dst0_l, rnd_vec); SRAR_W2_SW(dst0_r, dst0_l, rnd_vec);
dst0_r = CLIP_SW_0_255(dst0_r); dst0_r = (v4i32) __msa_pckev_h((v8i16) dst0_l, (v8i16) dst0_r);
dst0_l = CLIP_SW_0_255(dst0_l); out0 = CLIP_SH_0_255(dst0_r);
out0 = (v8i16) __msa_pckev_b((v16i8) out0, (v16i8) out0);
HEVC_PCK_SW_SB2(dst0_l, dst0_r, dst0_r); ST4x2_UB(out0, dst, dst_stride);
ST4x2_UB(dst0_r, dst, dst_stride);
} }
static void hevc_hz_biwgt_4t_4x4_msa(uint8_t *src0_ptr, static void hevc_hz_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
...@@ -2695,22 +2693,21 @@ static void hevc_hz_biwgt_4t_4x4_msa(uint8_t *src0_ptr, ...@@ -2695,22 +2693,21 @@ static void hevc_hz_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
uint8_t *dst, uint8_t *dst,
int32_t dst_stride, int32_t dst_stride,
const int8_t *filter, const int8_t *filter,
int32_t height,
int32_t weight0, int32_t weight0,
int32_t weight1, int32_t weight1,
int32_t offset0, int32_t offset0,
int32_t offset1, int32_t offset1,
int32_t rnd_val) int32_t rnd_val)
{ {
int32_t offset, weight; int32_t offset, weight, constant;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v16i8 src0, src1, src2, src3; v16i8 src0, src1, src2, src3;
v16i8 mask0 = { 0, 1, 1, 2, 2, 3, 3, 4, 16, 17, 17, 18, 18, 19, 19, 20 }; v16i8 mask0 = LD_SB(&ff_hevc_mask_arr[16]);
v16i8 mask1; v16i8 mask1;
v8i16 dst0, dst1; v8i16 dst0, dst1;
v16i8 vec0, vec1; v16i8 vec0, vec1;
v8i16 in0, in1, in2, in3; v8i16 in0, in1, in2, in3;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= 1; src0_ptr -= 1;
...@@ -2724,9 +2721,10 @@ static void hevc_hz_biwgt_4t_4x4_msa(uint8_t *src0_ptr, ...@@ -2724,9 +2721,10 @@ static void hevc_hz_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -2737,11 +2735,9 @@ static void hevc_hz_biwgt_4t_4x4_msa(uint8_t *src0_ptr, ...@@ -2737,11 +2735,9 @@ static void hevc_hz_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
ILVR_D2_SH(in1, in0, in3, in2, in0, in1); ILVR_D2_SH(in1, in0, in3, in2, in0, in1);
VSHF_B2_SB(src0, src1, src0, src1, mask0, mask1, vec0, vec1); VSHF_B2_SB(src0, src1, src0, src1, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
VSHF_B2_SB(src2, src3, src2, src3, mask0, mask1, vec0, vec1); VSHF_B2_SB(src2, src3, src2, src3, mask0, mask1, vec0, vec1);
dst1 = const_vec; dst1 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
HEVC_BIW_RND_CLIP2(dst0, dst1, in0, in1, HEVC_BIW_RND_CLIP2(dst0, dst1, in0, in1,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
dst0, dst1); dst0, dst1);
...@@ -2765,15 +2761,15 @@ static void hevc_hz_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr, ...@@ -2765,15 +2761,15 @@ static void hevc_hz_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t weight, offset; int32_t weight, offset, constant;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v16i8 src0, src1, src2, src3, src4, src5, src6, src7; v16i8 src0, src1, src2, src3, src4, src5, src6, src7;
v16i8 mask0 = { 0, 1, 1, 2, 2, 3, 3, 4, 16, 17, 17, 18, 18, 19, 19, 20 }; v16i8 mask0 = LD_SB(&ff_hevc_mask_arr[16]);
v16i8 mask1; v16i8 mask1;
v16i8 vec0, vec1; v16i8 vec0, vec1;
v8i16 dst0, dst1, dst2, dst3; v8i16 dst0, dst1, dst2, dst3;
v8i16 in0, in1, in2, in3, in4, in5, in6, in7; v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= 1; src0_ptr -= 1;
...@@ -2784,9 +2780,10 @@ static void hevc_hz_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr, ...@@ -2784,9 +2780,10 @@ static void hevc_hz_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -2806,17 +2803,13 @@ static void hevc_hz_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr, ...@@ -2806,17 +2803,13 @@ static void hevc_hz_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr,
XORI_B8_128_SB(src0, src1, src2, src3, src4, src5, src6, src7); XORI_B8_128_SB(src0, src1, src2, src3, src4, src5, src6, src7);
VSHF_B2_SB(src0, src1, src0, src1, mask0, mask1, vec0, vec1); VSHF_B2_SB(src0, src1, src0, src1, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
VSHF_B2_SB(src2, src3, src2, src3, mask0, mask1, vec0, vec1); VSHF_B2_SB(src2, src3, src2, src3, mask0, mask1, vec0, vec1);
dst1 = const_vec; dst1 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
VSHF_B2_SB(src4, src5, src4, src5, mask0, mask1, vec0, vec1); VSHF_B2_SB(src4, src5, src4, src5, mask0, mask1, vec0, vec1);
dst2 = const_vec; dst2 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst2, dst2);
VSHF_B2_SB(src6, src7, src6, src7, mask0, mask1, vec0, vec1); VSHF_B2_SB(src6, src7, src6, src7, mask0, mask1, vec0, vec1);
dst3 = const_vec; dst3 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst3, dst3);
HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3, HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -2844,11 +2837,11 @@ static void hevc_hz_biwgt_4t_4w_msa(uint8_t *src0_ptr, ...@@ -2844,11 +2837,11 @@ static void hevc_hz_biwgt_4t_4w_msa(uint8_t *src0_ptr,
{ {
if (2 == height) { if (2 == height) {
hevc_hz_biwgt_4t_4x2_msa(src0_ptr, src_stride, src1_ptr, src2_stride, hevc_hz_biwgt_4t_4x2_msa(src0_ptr, src_stride, src1_ptr, src2_stride,
dst, dst_stride, filter, height, dst, dst_stride, filter,
weight0, weight1, offset0, offset1, rnd_val); weight0, weight1, offset0, offset1, rnd_val);
} else if (4 == height) { } else if (4 == height) {
hevc_hz_biwgt_4t_4x4_msa(src0_ptr, src_stride, src1_ptr, src2_stride, hevc_hz_biwgt_4t_4x4_msa(src0_ptr, src_stride, src1_ptr, src2_stride,
dst, dst_stride, filter, height, dst, dst_stride, filter,
weight0, weight1, offset0, offset1, rnd_val); weight0, weight1, offset0, offset1, rnd_val);
} else if (0 == (height % 8)) { } else if (0 == (height % 8)) {
hevc_hz_biwgt_4t_4x8multiple_msa(src0_ptr, src_stride, hevc_hz_biwgt_4t_4x8multiple_msa(src0_ptr, src_stride,
...@@ -2874,15 +2867,15 @@ static void hevc_hz_biwgt_4t_6w_msa(uint8_t *src0_ptr, ...@@ -2874,15 +2867,15 @@ static void hevc_hz_biwgt_4t_6w_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t offset, weight; int32_t offset, weight, constant;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v16i8 src0, src1, src2, src3; v16i8 src0, src1, src2, src3;
v16i8 mask0 = { 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8 }; v16i8 mask0 = LD_SB(&ff_hevc_mask_arr[0]);
v16i8 mask1; v16i8 mask1;
v16i8 vec0, vec1; v16i8 vec0, vec1;
v8i16 in0, in1, in2, in3; v8i16 in0, in1, in2, in3;
v8i16 dst0, dst1, dst2, dst3; v8i16 dst0, dst1, dst2, dst3;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= 1; src0_ptr -= 1;
...@@ -2893,16 +2886,17 @@ static void hevc_hz_biwgt_4t_6w_msa(uint8_t *src0_ptr, ...@@ -2893,16 +2886,17 @@ static void hevc_hz_biwgt_4t_6w_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
mask1 = mask0 + 2; mask1 = mask0 + 2;
for (loop_cnt = (height >> 2); loop_cnt--;) { for (loop_cnt = 2; loop_cnt--;) {
LD_SB4(src0_ptr, src_stride, src0, src1, src2, src3); LD_SB4(src0_ptr, src_stride, src0, src1, src2, src3);
src0_ptr += (4 * src_stride); src0_ptr += (4 * src_stride);
LD_SH4(src1_ptr, src2_stride, in0, in1, in2, in3); LD_SH4(src1_ptr, src2_stride, in0, in1, in2, in3);
...@@ -2910,17 +2904,13 @@ static void hevc_hz_biwgt_4t_6w_msa(uint8_t *src0_ptr, ...@@ -2910,17 +2904,13 @@ static void hevc_hz_biwgt_4t_6w_msa(uint8_t *src0_ptr,
XORI_B4_128_SB(src0, src1, src2, src3); XORI_B4_128_SB(src0, src1, src2, src3);
VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1); VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1); VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1);
dst1 = const_vec; dst1 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1); VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1);
dst2 = const_vec; dst2 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst2, dst2);
VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1); VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1);
dst3 = const_vec; dst3 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst3, dst3);
HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3, HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3,
in0, in1, in2, in3, in0, in1, in2, in3,
...@@ -2940,21 +2930,20 @@ static void hevc_hz_biwgt_4t_8x2_msa(uint8_t *src0_ptr, ...@@ -2940,21 +2930,20 @@ static void hevc_hz_biwgt_4t_8x2_msa(uint8_t *src0_ptr,
uint8_t *dst, uint8_t *dst,
int32_t dst_stride, int32_t dst_stride,
const int8_t *filter, const int8_t *filter,
int32_t height,
int32_t weight0, int32_t weight0,
int32_t weight1, int32_t weight1,
int32_t offset0, int32_t offset0,
int32_t offset1, int32_t offset1,
int32_t rnd_val) int32_t rnd_val)
{ {
int32_t offset, weight; int32_t offset, weight, constant;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v16i8 src0, src1; v16i8 src0, src1;
v8i16 in0, in1; v8i16 in0, in1;
v16i8 mask0 = { 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8 }; v16i8 mask0 = LD_SB(&ff_hevc_mask_arr[0]);
v16i8 mask1, vec0, vec1; v16i8 mask1, vec0, vec1;
v8i16 dst0, dst1; v8i16 dst0, dst1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= 1; src0_ptr -= 1;
...@@ -2965,9 +2954,10 @@ static void hevc_hz_biwgt_4t_8x2_msa(uint8_t *src0_ptr, ...@@ -2965,9 +2954,10 @@ static void hevc_hz_biwgt_4t_8x2_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -2978,11 +2968,9 @@ static void hevc_hz_biwgt_4t_8x2_msa(uint8_t *src0_ptr, ...@@ -2978,11 +2968,9 @@ static void hevc_hz_biwgt_4t_8x2_msa(uint8_t *src0_ptr,
LD_SH2(src1_ptr, src2_stride, in0, in1); LD_SH2(src1_ptr, src2_stride, in0, in1);
XORI_B2_128_SB(src0, src1); XORI_B2_128_SB(src0, src1);
VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1); VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1); VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1);
dst1 = const_vec; dst1 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
HEVC_BIW_RND_CLIP2(dst0, dst1, in0, in1, HEVC_BIW_RND_CLIP2(dst0, dst1, in0, in1,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
dst0, dst1); dst0, dst1);
...@@ -2998,22 +2986,21 @@ static void hevc_hz_biwgt_4t_8x6_msa(uint8_t *src0_ptr, ...@@ -2998,22 +2986,21 @@ static void hevc_hz_biwgt_4t_8x6_msa(uint8_t *src0_ptr,
uint8_t *dst, uint8_t *dst,
int32_t dst_stride, int32_t dst_stride,
const int8_t *filter, const int8_t *filter,
int32_t height,
int32_t weight0, int32_t weight0,
int32_t weight1, int32_t weight1,
int32_t offset0, int32_t offset0,
int32_t offset1, int32_t offset1,
int32_t rnd_val) int32_t rnd_val)
{ {
int32_t weight, offset; int32_t weight, offset, constant;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v16i8 src0, src1, src2, src3, src4, src5; v16i8 src0, src1, src2, src3, src4, src5;
v8i16 in0, in1, in2, in3, in4, in5; v8i16 in0, in1, in2, in3, in4, in5;
v16i8 mask0 = { 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8 }; v16i8 mask0 = LD_SB(&ff_hevc_mask_arr[0]);
v16i8 mask1; v16i8 mask1;
v16i8 vec0, vec1; v16i8 vec0, vec1;
v8i16 dst0, dst1, dst2, dst3, dst4, dst5; v8i16 dst0, dst1, dst2, dst3, dst4, dst5;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= 1; src0_ptr -= 1;
...@@ -3024,9 +3011,10 @@ static void hevc_hz_biwgt_4t_8x6_msa(uint8_t *src0_ptr, ...@@ -3024,9 +3011,10 @@ static void hevc_hz_biwgt_4t_8x6_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3040,23 +3028,17 @@ static void hevc_hz_biwgt_4t_8x6_msa(uint8_t *src0_ptr, ...@@ -3040,23 +3028,17 @@ static void hevc_hz_biwgt_4t_8x6_msa(uint8_t *src0_ptr,
LD_SH2(src1_ptr, src2_stride, in4, in5); LD_SH2(src1_ptr, src2_stride, in4, in5);
XORI_B6_128_SB(src0, src1, src2, src3, src4, src5); XORI_B6_128_SB(src0, src1, src2, src3, src4, src5);
VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1); VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1); VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1);
dst1 = const_vec; dst1 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1); VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1);
dst2 = const_vec; dst2 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst2, dst2);
VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1); VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1);
dst3 = const_vec; dst3 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst3, dst3);
VSHF_B2_SB(src4, src4, src4, src4, mask0, mask1, vec0, vec1); VSHF_B2_SB(src4, src4, src4, src4, mask0, mask1, vec0, vec1);
dst4 = const_vec; dst4 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst4, dst4);
VSHF_B2_SB(src5, src5, src5, src5, mask0, mask1, vec0, vec1); VSHF_B2_SB(src5, src5, src5, src5, mask0, mask1, vec0, vec1);
dst5 = const_vec; dst5 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst5, dst5);
HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3, HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -3087,15 +3069,15 @@ static void hevc_hz_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr, ...@@ -3087,15 +3069,15 @@ static void hevc_hz_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t offset, weight; int32_t offset, weight, constant;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v16i8 src0, src1, src2, src3; v16i8 src0, src1, src2, src3;
v16i8 mask0 = { 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8 }; v16i8 mask0 = LD_SB(ff_hevc_mask_arr);
v16i8 mask1; v16i8 mask1;
v16i8 vec0, vec1; v16i8 vec0, vec1;
v8i16 in0, in1, in2, in3; v8i16 in0, in1, in2, in3;
v8i16 dst0, dst1, dst2, dst3; v8i16 dst0, dst1, dst2, dst3;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= 1; src0_ptr -= 1;
...@@ -3106,9 +3088,10 @@ static void hevc_hz_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr, ...@@ -3106,9 +3088,10 @@ static void hevc_hz_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3123,17 +3106,13 @@ static void hevc_hz_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr, ...@@ -3123,17 +3106,13 @@ static void hevc_hz_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr,
XORI_B4_128_SB(src0, src1, src2, src3); XORI_B4_128_SB(src0, src1, src2, src3);
VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1); VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1); VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1);
dst1 = const_vec; dst1 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1); VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1);
dst2 = const_vec; dst2 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst2, dst2);
VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1); VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1);
dst3 = const_vec; dst3 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst3, dst3);
HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3, HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -3161,11 +3140,11 @@ static void hevc_hz_biwgt_4t_8w_msa(uint8_t *src0_ptr, ...@@ -3161,11 +3140,11 @@ static void hevc_hz_biwgt_4t_8w_msa(uint8_t *src0_ptr,
{ {
if (2 == height) { if (2 == height) {
hevc_hz_biwgt_4t_8x2_msa(src0_ptr, src_stride, src1_ptr, src2_stride, hevc_hz_biwgt_4t_8x2_msa(src0_ptr, src_stride, src1_ptr, src2_stride,
dst, dst_stride, filter, height, dst, dst_stride, filter,
weight0, weight1, offset0, offset1, rnd_val); weight0, weight1, offset0, offset1, rnd_val);
} else if (6 == height) { } else if (6 == height) {
hevc_hz_biwgt_4t_8x6_msa(src0_ptr, src_stride, src1_ptr, src2_stride, hevc_hz_biwgt_4t_8x6_msa(src0_ptr, src_stride, src1_ptr, src2_stride,
dst, dst_stride, filter, height, dst, dst_stride, filter,
weight0, weight1, offset0, offset1, rnd_val); weight0, weight1, offset0, offset1, rnd_val);
} else if (0 == (height % 4)) { } else if (0 == (height % 4)) {
hevc_hz_biwgt_4t_8x4multiple_msa(src0_ptr, src_stride, hevc_hz_biwgt_4t_8x4multiple_msa(src0_ptr, src_stride,
...@@ -3191,18 +3170,18 @@ static void hevc_hz_biwgt_4t_12w_msa(uint8_t *src0_ptr, ...@@ -3191,18 +3170,18 @@ static void hevc_hz_biwgt_4t_12w_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t offset, weight; int32_t offset, weight, constant;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v16i8 src0, src1, src2, src3; v16i8 src0, src1, src2, src3;
v8i16 in0, in1, in2, in3, in4, in5, in6, in7; v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
v16i8 mask0 = { 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8 }; v16i8 mask0 = LD_SB(&ff_hevc_mask_arr[0]);
v16i8 mask2 = { v16i8 mask2 = {
8, 9, 9, 10, 10, 11, 11, 12, 24, 25, 25, 26, 26, 27, 27, 28 8, 9, 9, 10, 10, 11, 11, 12, 24, 25, 25, 26, 26, 27, 27, 28
}; };
v16i8 mask1, mask3; v16i8 mask1, mask3;
v16i8 vec0, vec1; v16i8 vec0, vec1;
v8i16 dst0, dst1, dst2, dst3, dst4, dst5; v8i16 dst0, dst1, dst2, dst3, dst4, dst5;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= 1; src0_ptr -= 1;
...@@ -3213,9 +3192,10 @@ static void hevc_hz_biwgt_4t_12w_msa(uint8_t *src0_ptr, ...@@ -3213,9 +3192,10 @@ static void hevc_hz_biwgt_4t_12w_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3223,7 +3203,7 @@ static void hevc_hz_biwgt_4t_12w_msa(uint8_t *src0_ptr, ...@@ -3223,7 +3203,7 @@ static void hevc_hz_biwgt_4t_12w_msa(uint8_t *src0_ptr,
mask1 = mask0 + 2; mask1 = mask0 + 2;
mask3 = mask2 + 2; mask3 = mask2 + 2;
for (loop_cnt = (height >> 2); loop_cnt--;) { for (loop_cnt = 4; loop_cnt--;) {
LD_SB4(src0_ptr, src_stride, src0, src1, src2, src3); LD_SB4(src0_ptr, src_stride, src0, src1, src2, src3);
src0_ptr += (4 * src_stride); src0_ptr += (4 * src_stride);
LD_SH4(src1_ptr, src2_stride, in0, in1, in2, in3); LD_SH4(src1_ptr, src2_stride, in0, in1, in2, in3);
...@@ -3233,23 +3213,17 @@ static void hevc_hz_biwgt_4t_12w_msa(uint8_t *src0_ptr, ...@@ -3233,23 +3213,17 @@ static void hevc_hz_biwgt_4t_12w_msa(uint8_t *src0_ptr,
XORI_B4_128_SB(src0, src1, src2, src3); XORI_B4_128_SB(src0, src1, src2, src3);
VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1); VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1); VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1);
dst1 = const_vec; dst1 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1); VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1);
dst2 = const_vec; dst2 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst2, dst2);
VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1); VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1);
dst3 = const_vec; dst3 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst3, dst3);
VSHF_B2_SB(src0, src1, src0, src1, mask2, mask3, vec0, vec1); VSHF_B2_SB(src0, src1, src0, src1, mask2, mask3, vec0, vec1);
dst4 = const_vec; dst4 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst4, dst4);
VSHF_B2_SB(src2, src3, src2, src3, mask2, mask3, vec0, vec1); VSHF_B2_SB(src2, src3, src2, src3, mask2, mask3, vec0, vec1);
dst5 = const_vec; dst5 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst5, dst5);
HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3, HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3,
in0, in1, in2, in3, in0, in1, in2, in3,
...@@ -3281,15 +3255,15 @@ static void hevc_hz_biwgt_4t_16w_msa(uint8_t *src0_ptr, ...@@ -3281,15 +3255,15 @@ static void hevc_hz_biwgt_4t_16w_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t offset, weight; int32_t offset, weight, constant;
v16i8 src0, src1, src2, src3, src4, src5, src6, src7; v16i8 src0, src1, src2, src3, src4, src5, src6, src7;
v8i16 in0, in1, in2, in3, in4, in5, in6, in7; v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v16i8 mask0 = { 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8 }; v16i8 mask0 = LD_SB(&ff_hevc_mask_arr[0]);
v16i8 mask1; v16i8 mask1;
v8i16 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; v8i16 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7;
v16i8 vec0, vec1; v16i8 vec0, vec1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= 1; src0_ptr -= 1;
...@@ -3300,9 +3274,10 @@ static void hevc_hz_biwgt_4t_16w_msa(uint8_t *src0_ptr, ...@@ -3300,9 +3274,10 @@ static void hevc_hz_biwgt_4t_16w_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3319,29 +3294,21 @@ static void hevc_hz_biwgt_4t_16w_msa(uint8_t *src0_ptr, ...@@ -3319,29 +3294,21 @@ static void hevc_hz_biwgt_4t_16w_msa(uint8_t *src0_ptr,
XORI_B8_128_SB(src0, src1, src2, src3, src4, src5, src6, src7); XORI_B8_128_SB(src0, src1, src2, src3, src4, src5, src6, src7);
VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1); VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1); VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1);
dst1 = const_vec; dst1 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1); VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1);
dst2 = const_vec; dst2 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst2, dst2);
VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1); VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1);
dst3 = const_vec; dst3 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst3, dst3);
VSHF_B2_SB(src4, src4, src4, src4, mask0, mask1, vec0, vec1); VSHF_B2_SB(src4, src4, src4, src4, mask0, mask1, vec0, vec1);
dst4 = const_vec; dst4 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst4, dst4);
VSHF_B2_SB(src5, src5, src5, src5, mask0, mask1, vec0, vec1); VSHF_B2_SB(src5, src5, src5, src5, mask0, mask1, vec0, vec1);
dst5 = const_vec; dst5 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst5, dst5);
VSHF_B2_SB(src6, src6, src6, src6, mask0, mask1, vec0, vec1); VSHF_B2_SB(src6, src6, src6, src6, mask0, mask1, vec0, vec1);
dst6 = const_vec; dst6 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst6, dst6);
VSHF_B2_SB(src7, src7, src7, src7, mask0, mask1, vec0, vec1); VSHF_B2_SB(src7, src7, src7, src7, mask0, mask1, vec0, vec1);
dst7 = const_vec; dst7 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst7, dst7);
HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3, HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -3377,16 +3344,15 @@ static void hevc_hz_biwgt_4t_24w_msa(uint8_t *src0_ptr, ...@@ -3377,16 +3344,15 @@ static void hevc_hz_biwgt_4t_24w_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t offset, weight; int32_t offset, weight, constant;
uint8_t *dst_tmp = dst + 16;
v16i8 src0, src1, src2, src3; v16i8 src0, src1, src2, src3;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v16i8 mask0 = { 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8 }; v16i8 mask0 = LD_SB(&ff_hevc_mask_arr[0]);
v16i8 mask1, mask2, mask3; v16i8 mask1, mask2, mask3;
v16i8 vec0, vec1; v16i8 vec0, vec1;
v8i16 dst0, dst1, dst2, dst3; v8i16 dst0, dst1, dst2, dst3;
v8i16 in0, in1, in2, in3, in4, in5; v8i16 in0, in1, in2, in3, in4, in5;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= 1; src0_ptr -= 1;
...@@ -3397,9 +3363,10 @@ static void hevc_hz_biwgt_4t_24w_msa(uint8_t *src0_ptr, ...@@ -3397,9 +3363,10 @@ static void hevc_hz_biwgt_4t_24w_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3408,7 +3375,7 @@ static void hevc_hz_biwgt_4t_24w_msa(uint8_t *src0_ptr, ...@@ -3408,7 +3375,7 @@ static void hevc_hz_biwgt_4t_24w_msa(uint8_t *src0_ptr,
mask2 = mask0 + 8; mask2 = mask0 + 8;
mask3 = mask0 + 10; mask3 = mask0 + 10;
for (loop_cnt = (height >> 1); loop_cnt--;) { for (loop_cnt = 16; loop_cnt--;) {
LD_SB2(src0_ptr, src_stride, src0, src2); LD_SB2(src0_ptr, src_stride, src0, src2);
LD_SB2(src0_ptr + 16, src_stride, src1, src3); LD_SB2(src0_ptr + 16, src_stride, src1, src3);
src0_ptr += (2 * src_stride); src0_ptr += (2 * src_stride);
...@@ -3419,17 +3386,13 @@ static void hevc_hz_biwgt_4t_24w_msa(uint8_t *src0_ptr, ...@@ -3419,17 +3386,13 @@ static void hevc_hz_biwgt_4t_24w_msa(uint8_t *src0_ptr,
XORI_B4_128_SB(src0, src1, src2, src3); XORI_B4_128_SB(src0, src1, src2, src3);
VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1); VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
VSHF_B2_SB(src0, src1, src0, src1, mask2, mask3, vec0, vec1); VSHF_B2_SB(src0, src1, src0, src1, mask2, mask3, vec0, vec1);
dst1 = const_vec; dst1 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1); VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1);
dst2 = const_vec; dst2 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst2, dst2);
VSHF_B2_SB(src2, src3, src2, src3, mask2, mask3, vec0, vec1); VSHF_B2_SB(src2, src3, src2, src3, mask2, mask3, vec0, vec1);
dst3 = const_vec; dst3 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst3, dst3);
HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3, HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -3437,21 +3400,19 @@ static void hevc_hz_biwgt_4t_24w_msa(uint8_t *src0_ptr, ...@@ -3437,21 +3400,19 @@ static void hevc_hz_biwgt_4t_24w_msa(uint8_t *src0_ptr,
PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1); PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
ST_SH2(dst0, dst1, dst, dst_stride); ST_SH2(dst0, dst1, dst, dst_stride);
dst += (2 * dst_stride);
/* 8 width */ /* 8 width */
VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1); VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1); VSHF_B2_SB(src3, src3, src3, src3, mask0, mask1, vec0, vec1);
dst1 = const_vec; dst1 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
HEVC_BIW_RND_CLIP2(dst0, dst1, in4, in5, HEVC_BIW_RND_CLIP2(dst0, dst1, in4, in5,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
dst0, dst1); dst0, dst1);
dst0 = (v8i16) __msa_pckev_b((v16i8) dst1, (v16i8) dst0); dst0 = (v8i16) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
ST8x2_UB(dst0, dst_tmp, dst_stride); ST8x2_UB(dst0, (dst + 16), dst_stride);
dst_tmp += (2 * dst_stride); dst += (2 * dst_stride);
} }
} }
...@@ -3470,15 +3431,15 @@ static void hevc_hz_biwgt_4t_32w_msa(uint8_t *src0_ptr, ...@@ -3470,15 +3431,15 @@ static void hevc_hz_biwgt_4t_32w_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t offset, weight; int32_t offset, weight, constant;
v16i8 src0, src1, src2; v16i8 src0, src1, src2;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v16i8 mask0 = { 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8 }; v16i8 mask0 = LD_SB(&ff_hevc_mask_arr[0]);
v16i8 mask1, mask2, mask3; v16i8 mask1, mask2, mask3;
v8i16 dst0, dst1, dst2, dst3; v8i16 dst0, dst1, dst2, dst3;
v16i8 vec0, vec1; v16i8 vec0, vec1;
v8i16 in0, in1, in2, in3; v8i16 in0, in1, in2, in3;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= 1; src0_ptr -= 1;
...@@ -3489,9 +3450,10 @@ static void hevc_hz_biwgt_4t_32w_msa(uint8_t *src0_ptr, ...@@ -3489,9 +3450,10 @@ static void hevc_hz_biwgt_4t_32w_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3509,17 +3471,13 @@ static void hevc_hz_biwgt_4t_32w_msa(uint8_t *src0_ptr, ...@@ -3509,17 +3471,13 @@ static void hevc_hz_biwgt_4t_32w_msa(uint8_t *src0_ptr,
XORI_B3_128_SB(src0, src1, src2); XORI_B3_128_SB(src0, src1, src2);
VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1); VSHF_B2_SB(src0, src0, src0, src0, mask0, mask1, vec0, vec1);
dst0 = const_vec; dst0 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
VSHF_B2_SB(src0, src1, src0, src1, mask2, mask3, vec0, vec1); VSHF_B2_SB(src0, src1, src0, src1, mask2, mask3, vec0, vec1);
dst1 = const_vec; dst1 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1); VSHF_B2_SB(src1, src1, src1, src1, mask0, mask1, vec0, vec1);
dst2 = const_vec; dst2 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst2, dst2);
VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1); VSHF_B2_SB(src2, src2, src2, src2, mask0, mask1, vec0, vec1);
dst3 = const_vec; dst3 = HEVC_FILT_4TAP_SH(vec0, vec1, filt0, filt1);
DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst3, dst3);
HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3, HEVC_BIW_RND_CLIP4(dst0, dst1, dst2, dst3,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -3538,20 +3496,19 @@ static void hevc_vt_biwgt_4t_4x2_msa(uint8_t *src0_ptr, ...@@ -3538,20 +3496,19 @@ static void hevc_vt_biwgt_4t_4x2_msa(uint8_t *src0_ptr,
uint8_t *dst, uint8_t *dst,
int32_t dst_stride, int32_t dst_stride,
const int8_t *filter, const int8_t *filter,
int32_t height,
int32_t weight0, int32_t weight0,
int32_t weight1, int32_t weight1,
int32_t offset0, int32_t offset0,
int32_t offset1, int32_t offset1,
int32_t rnd_val) int32_t rnd_val)
{ {
int32_t weight, offset; int32_t weight, offset, constant;
v16i8 src0, src1, src2, src3, src4; v16i8 src0, src1, src2, src3, src4;
v8i16 in0, in1, dst10; v8i16 in0, in1, dst10;
v16i8 src10_r, src32_r, src21_r, src43_r, src2110, src4332; v16i8 src10_r, src32_r, src21_r, src43_r, src2110, src4332;
v4i32 dst10_r, dst10_l; v4i32 dst10_r, dst10_l;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v8i16 filter_vec, const_vec; v8i16 filter_vec, out;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= src_stride; src0_ptr -= src_stride;
...@@ -3559,9 +3516,10 @@ static void hevc_vt_biwgt_4t_4x2_msa(uint8_t *src0_ptr, ...@@ -3559,9 +3516,10 @@ static void hevc_vt_biwgt_4t_4x2_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3584,18 +3542,16 @@ static void hevc_vt_biwgt_4t_4x2_msa(uint8_t *src0_ptr, ...@@ -3584,18 +3542,16 @@ static void hevc_vt_biwgt_4t_4x2_msa(uint8_t *src0_ptr,
src4332 = (v16i8) __msa_ilvr_d((v2i64) src43_r, (v2i64) src32_r); src4332 = (v16i8) __msa_ilvr_d((v2i64) src43_r, (v2i64) src32_r);
src4332 = (v16i8) __msa_xori_b((v16u8) src4332, 128); src4332 = (v16i8) __msa_xori_b((v16u8) src4332, 128);
dst10 = const_vec; dst10 = HEVC_FILT_4TAP_SH(src2110, src4332, filt0, filt1);
DPADD_SB2_SH(src2110, src4332, filt0, filt1, dst10, dst10);
ILVRL_H2_SW(dst10, in0, dst10_r, dst10_l); ILVRL_H2_SW(dst10, in0, dst10_r, dst10_l);
dst10_r = __msa_dpadd_s_w(offset_vec, (v8i16) dst10_r, (v8i16) weight_vec); dst10_r = __msa_dpadd_s_w(offset_vec, (v8i16) dst10_r, (v8i16) weight_vec);
dst10_l = __msa_dpadd_s_w(offset_vec, (v8i16) dst10_l, (v8i16) weight_vec); dst10_l = __msa_dpadd_s_w(offset_vec, (v8i16) dst10_l, (v8i16) weight_vec);
SRAR_W2_SW(dst10_r, dst10_l, rnd_vec); SRAR_W2_SW(dst10_r, dst10_l, rnd_vec);
dst10_r = CLIP_SW_0_255(dst10_r); dst10_r = (v4i32) __msa_pckev_h((v8i16) dst10_l, (v8i16) dst10_r);
dst10_l = CLIP_SW_0_255(dst10_l); out = CLIP_SH_0_255(dst10_r);
out = (v8i16) __msa_pckev_b((v16i8) out, (v16i8) out);
HEVC_PCK_SW_SB2(dst10_l, dst10_r, dst10_r); ST4x2_UB(out, dst, dst_stride);
ST4x2_UB(dst10_r, dst, dst_stride);
} }
static void hevc_vt_biwgt_4t_4x4_msa(uint8_t *src0_ptr, static void hevc_vt_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
...@@ -3605,21 +3561,20 @@ static void hevc_vt_biwgt_4t_4x4_msa(uint8_t *src0_ptr, ...@@ -3605,21 +3561,20 @@ static void hevc_vt_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
uint8_t *dst, uint8_t *dst,
int32_t dst_stride, int32_t dst_stride,
const int8_t *filter, const int8_t *filter,
int32_t height,
int32_t weight0, int32_t weight0,
int32_t weight1, int32_t weight1,
int32_t offset0, int32_t offset0,
int32_t offset1, int32_t offset1,
int32_t rnd_val) int32_t rnd_val)
{ {
int32_t weight, offset; int32_t weight, offset, constant;
v16i8 src0, src1, src2, src3, src4, src5, src6; v16i8 src0, src1, src2, src3, src4, src5, src6;
v8i16 in0, in1, in2, in3; v8i16 in0, in1, in2, in3;
v16i8 src10_r, src32_r, src54_r, src21_r, src43_r, src65_r; v16i8 src10_r, src32_r, src54_r, src21_r, src43_r, src65_r;
v16i8 src2110, src4332, src6554; v16i8 src2110, src4332, src6554;
v8i16 dst10, dst32; v8i16 dst10, dst32;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= src_stride; src0_ptr -= src_stride;
...@@ -3627,9 +3582,10 @@ static void hevc_vt_biwgt_4t_4x4_msa(uint8_t *src0_ptr, ...@@ -3627,9 +3582,10 @@ static void hevc_vt_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3653,10 +3609,8 @@ static void hevc_vt_biwgt_4t_4x4_msa(uint8_t *src0_ptr, ...@@ -3653,10 +3609,8 @@ static void hevc_vt_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
ILVR_D2_SB(src43_r, src32_r, src65_r, src54_r, src4332, src6554); ILVR_D2_SB(src43_r, src32_r, src65_r, src54_r, src4332, src6554);
XORI_B2_128_SB(src4332, src6554); XORI_B2_128_SB(src4332, src6554);
dst10 = const_vec; dst10 = HEVC_FILT_4TAP_SH(src2110, src4332, filt0, filt1);
DPADD_SB2_SH(src2110, src4332, filt0, filt1, dst10, dst10); dst32 = HEVC_FILT_4TAP_SH(src4332, src6554, filt0, filt1);
dst32 = const_vec;
DPADD_SB2_SH(src4332, src6554, filt0, filt1, dst32, dst32);
HEVC_BIW_RND_CLIP2(dst10, dst32, in0, in1, HEVC_BIW_RND_CLIP2(dst10, dst32, in0, in1,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -3682,7 +3636,7 @@ static void hevc_vt_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr, ...@@ -3682,7 +3636,7 @@ static void hevc_vt_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t weight, offset; int32_t weight, offset, constant;
v16i8 src0, src1, src2, src3, src4, src5, src6, src7, src8, src9; v16i8 src0, src1, src2, src3, src4, src5, src6, src7, src8, src9;
v8i16 in0, in1, in2, in3, in4, in5, in6, in7; v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
v16i8 src10_r, src32_r, src54_r, src76_r, src98_r; v16i8 src10_r, src32_r, src54_r, src76_r, src98_r;
...@@ -3690,7 +3644,7 @@ static void hevc_vt_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr, ...@@ -3690,7 +3644,7 @@ static void hevc_vt_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr,
v16i8 src2110, src4332, src6554, src8776; v16i8 src2110, src4332, src6554, src8776;
v8i16 dst10, dst32, dst54, dst76; v8i16 dst10, dst32, dst54, dst76;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= src_stride; src0_ptr -= src_stride;
...@@ -3698,9 +3652,10 @@ static void hevc_vt_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr, ...@@ -3698,9 +3652,10 @@ static void hevc_vt_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3730,12 +3685,9 @@ static void hevc_vt_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr, ...@@ -3730,12 +3685,9 @@ static void hevc_vt_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr,
src4332, src6554, src8776); src4332, src6554, src8776);
XORI_B3_128_SB(src4332, src6554, src8776); XORI_B3_128_SB(src4332, src6554, src8776);
dst10 = const_vec; dst10 = HEVC_FILT_4TAP_SH(src2110, src4332, filt0, filt1);
DPADD_SB2_SH(src2110, src4332, filt0, filt1, dst10, dst10); dst32 = HEVC_FILT_4TAP_SH(src4332, src6554, filt0, filt1);
dst32 = const_vec; dst54 = HEVC_FILT_4TAP_SH(src6554, src8776, filt0, filt1);
DPADD_SB2_SH(src4332, src6554, filt0, filt1, dst32, dst32);
dst54 = const_vec;
DPADD_SB2_SH(src6554, src8776, filt0, filt1, dst54, dst54);
LD_SB2(src0_ptr, src_stride, src9, src2); LD_SB2(src0_ptr, src_stride, src9, src2);
src0_ptr += (2 * src_stride); src0_ptr += (2 * src_stride);
...@@ -3743,8 +3695,7 @@ static void hevc_vt_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr, ...@@ -3743,8 +3695,7 @@ static void hevc_vt_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr,
src2110 = (v16i8) __msa_ilvr_d((v2i64) src109_r, (v2i64) src98_r); src2110 = (v16i8) __msa_ilvr_d((v2i64) src109_r, (v2i64) src98_r);
src2110 = (v16i8) __msa_xori_b((v16u8) src2110, 128); src2110 = (v16i8) __msa_xori_b((v16u8) src2110, 128);
dst76 = const_vec; dst76 = HEVC_FILT_4TAP_SH(src8776, src2110, filt0, filt1);
DPADD_SB2_SH(src8776, src2110, filt0, filt1, dst76, dst76);
HEVC_BIW_RND_CLIP4(dst10, dst32, dst54, dst76, HEVC_BIW_RND_CLIP4(dst10, dst32, dst54, dst76,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -3772,11 +3723,11 @@ static void hevc_vt_biwgt_4t_4w_msa(uint8_t *src0_ptr, ...@@ -3772,11 +3723,11 @@ static void hevc_vt_biwgt_4t_4w_msa(uint8_t *src0_ptr,
{ {
if (2 == height) { if (2 == height) {
hevc_vt_biwgt_4t_4x2_msa(src0_ptr, src_stride, src1_ptr, src2_stride, hevc_vt_biwgt_4t_4x2_msa(src0_ptr, src_stride, src1_ptr, src2_stride,
dst, dst_stride, filter, height, dst, dst_stride, filter,
weight0, weight1, offset0, offset1, rnd_val); weight0, weight1, offset0, offset1, rnd_val);
} else if (4 == height) { } else if (4 == height) {
hevc_vt_biwgt_4t_4x4_msa(src0_ptr, src_stride, src1_ptr, src2_stride, hevc_vt_biwgt_4t_4x4_msa(src0_ptr, src_stride, src1_ptr, src2_stride,
dst, dst_stride, filter, height, dst, dst_stride, filter,
weight0, weight1, offset0, offset1, rnd_val); weight0, weight1, offset0, offset1, rnd_val);
} else if (0 == (height % 8)) { } else if (0 == (height % 8)) {
hevc_vt_biwgt_4t_4x8multiple_msa(src0_ptr, src_stride, hevc_vt_biwgt_4t_4x8multiple_msa(src0_ptr, src_stride,
...@@ -3802,13 +3753,13 @@ static void hevc_vt_biwgt_4t_6w_msa(uint8_t *src0_ptr, ...@@ -3802,13 +3753,13 @@ static void hevc_vt_biwgt_4t_6w_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t offset, weight; int32_t offset, weight, constant;
v16i8 src0, src1, src2, src3, src4; v16i8 src0, src1, src2, src3, src4;
v8i16 in0, in1, in2, in3; v8i16 in0, in1, in2, in3;
v16i8 src10_r, src32_r, src21_r, src43_r; v16i8 src10_r, src32_r, src21_r, src43_r;
v8i16 tmp0, tmp1, tmp2, tmp3; v8i16 tmp0, tmp1, tmp2, tmp3;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= src_stride; src0_ptr -= src_stride;
...@@ -3816,9 +3767,10 @@ static void hevc_vt_biwgt_4t_6w_msa(uint8_t *src0_ptr, ...@@ -3816,9 +3767,10 @@ static void hevc_vt_biwgt_4t_6w_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3839,20 +3791,16 @@ static void hevc_vt_biwgt_4t_6w_msa(uint8_t *src0_ptr, ...@@ -3839,20 +3791,16 @@ static void hevc_vt_biwgt_4t_6w_msa(uint8_t *src0_ptr,
XORI_B2_128_SB(src3, src4); XORI_B2_128_SB(src3, src4);
ILVR_B2_SB(src3, src2, src4, src3, src32_r, src43_r); ILVR_B2_SB(src3, src2, src4, src3, src32_r, src43_r);
tmp0 = const_vec; tmp0 = HEVC_FILT_4TAP_SH(src10_r, src32_r, filt0, filt1);
DPADD_SB2_SH(src10_r, src32_r, filt0, filt1, tmp0, tmp0); tmp1 = HEVC_FILT_4TAP_SH(src21_r, src43_r, filt0, filt1);
tmp1 = const_vec;
DPADD_SB2_SH(src21_r, src43_r, filt0, filt1, tmp1, tmp1);
LD_SB2(src0_ptr, src_stride, src1, src2); LD_SB2(src0_ptr, src_stride, src1, src2);
src0_ptr += (2 * src_stride); src0_ptr += (2 * src_stride);
XORI_B2_128_SB(src1, src2); XORI_B2_128_SB(src1, src2);
ILVR_B2_SB(src1, src4, src2, src1, src10_r, src21_r); ILVR_B2_SB(src1, src4, src2, src1, src10_r, src21_r);
tmp2 = const_vec; tmp2 = HEVC_FILT_4TAP_SH(src32_r, src10_r, filt0, filt1);
DPADD_SB2_SH(src32_r, src10_r, filt0, filt1, tmp2, tmp2); tmp3 = HEVC_FILT_4TAP_SH(src43_r, src21_r, filt0, filt1);
tmp3 = const_vec;
DPADD_SB2_SH(src43_r, src21_r, filt0, filt1, tmp3, tmp3);
HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3, HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -3871,19 +3819,18 @@ static void hevc_vt_biwgt_4t_8x2_msa(uint8_t *src0_ptr, ...@@ -3871,19 +3819,18 @@ static void hevc_vt_biwgt_4t_8x2_msa(uint8_t *src0_ptr,
uint8_t *dst, uint8_t *dst,
int32_t dst_stride, int32_t dst_stride,
const int8_t *filter, const int8_t *filter,
int32_t height,
int32_t weight0, int32_t weight0,
int32_t weight1, int32_t weight1,
int32_t offset0, int32_t offset0,
int32_t offset1, int32_t offset1,
int32_t rnd_val) int32_t rnd_val)
{ {
int32_t offset, weight; int32_t offset, weight, constant;
v16i8 src0, src1, src2, src3, src4; v16i8 src0, src1, src2, src3, src4;
v8i16 in0, in1, tmp0, tmp1; v8i16 in0, in1, tmp0, tmp1;
v16i8 src10_r, src32_r, src21_r, src43_r; v16i8 src10_r, src32_r, src21_r, src43_r;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= src_stride; src0_ptr -= src_stride;
...@@ -3891,9 +3838,10 @@ static void hevc_vt_biwgt_4t_8x2_msa(uint8_t *src0_ptr, ...@@ -3891,9 +3838,10 @@ static void hevc_vt_biwgt_4t_8x2_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3911,10 +3859,8 @@ static void hevc_vt_biwgt_4t_8x2_msa(uint8_t *src0_ptr, ...@@ -3911,10 +3859,8 @@ static void hevc_vt_biwgt_4t_8x2_msa(uint8_t *src0_ptr,
XORI_B2_128_SB(src3, src4); XORI_B2_128_SB(src3, src4);
ILVR_B2_SB(src3, src2, src4, src3, src32_r, src43_r); ILVR_B2_SB(src3, src2, src4, src3, src32_r, src43_r);
tmp0 = const_vec; tmp0 = HEVC_FILT_4TAP_SH(src10_r, src32_r, filt0, filt1);
DPADD_SB2_SH(src10_r, src32_r, filt0, filt1, tmp0, tmp0); tmp1 = HEVC_FILT_4TAP_SH(src21_r, src43_r, filt0, filt1);
tmp1 = const_vec;
DPADD_SB2_SH(src21_r, src43_r, filt0, filt1, tmp1, tmp1);
HEVC_BIW_RND_CLIP2(tmp0, tmp1, in0, in1, HEVC_BIW_RND_CLIP2(tmp0, tmp1, in0, in1,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
tmp0, tmp1); tmp0, tmp1);
...@@ -3930,21 +3876,20 @@ static void hevc_vt_biwgt_4t_8x6_msa(uint8_t *src0_ptr, ...@@ -3930,21 +3876,20 @@ static void hevc_vt_biwgt_4t_8x6_msa(uint8_t *src0_ptr,
uint8_t *dst, uint8_t *dst,
int32_t dst_stride, int32_t dst_stride,
const int8_t *filter, const int8_t *filter,
int32_t height,
int32_t weight0, int32_t weight0,
int32_t weight1, int32_t weight1,
int32_t offset0, int32_t offset0,
int32_t offset1, int32_t offset1,
int32_t rnd_val) int32_t rnd_val)
{ {
int32_t offset, weight; int32_t offset, weight, constant;
v16i8 src0, src1, src2, src3, src4, src5, src6, src7, src8; v16i8 src0, src1, src2, src3, src4, src5, src6, src7, src8;
v8i16 in0, in1, in2, in3, in4, in5; v8i16 in0, in1, in2, in3, in4, in5;
v16i8 src10_r, src32_r, src54_r, src76_r; v16i8 src10_r, src32_r, src54_r, src76_r;
v16i8 src21_r, src43_r, src65_r, src87_r; v16i8 src21_r, src43_r, src65_r, src87_r;
v8i16 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5; v8i16 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= src_stride; src0_ptr -= src_stride;
...@@ -3952,9 +3897,10 @@ static void hevc_vt_biwgt_4t_8x6_msa(uint8_t *src0_ptr, ...@@ -3952,9 +3897,10 @@ static void hevc_vt_biwgt_4t_8x6_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -3974,18 +3920,12 @@ static void hevc_vt_biwgt_4t_8x6_msa(uint8_t *src0_ptr, ...@@ -3974,18 +3920,12 @@ static void hevc_vt_biwgt_4t_8x6_msa(uint8_t *src0_ptr,
src32_r, src43_r, src54_r, src65_r); src32_r, src43_r, src54_r, src65_r);
ILVR_B2_SB(src7, src6, src8, src7, src76_r, src87_r); ILVR_B2_SB(src7, src6, src8, src7, src76_r, src87_r);
tmp0 = const_vec; tmp0 = HEVC_FILT_4TAP_SH(src10_r, src32_r, filt0, filt1);
DPADD_SB2_SH(src10_r, src32_r, filt0, filt1, tmp0, tmp0); tmp1 = HEVC_FILT_4TAP_SH(src21_r, src43_r, filt0, filt1);
tmp1 = const_vec; tmp2 = HEVC_FILT_4TAP_SH(src32_r, src54_r, filt0, filt1);
DPADD_SB2_SH(src21_r, src43_r, filt0, filt1, tmp1, tmp1); tmp3 = HEVC_FILT_4TAP_SH(src43_r, src65_r, filt0, filt1);
tmp2 = const_vec; tmp4 = HEVC_FILT_4TAP_SH(src54_r, src76_r, filt0, filt1);
DPADD_SB2_SH(src32_r, src54_r, filt0, filt1, tmp2, tmp2); tmp5 = HEVC_FILT_4TAP_SH(src65_r, src87_r, filt0, filt1);
tmp3 = const_vec;
DPADD_SB2_SH(src43_r, src65_r, filt0, filt1, tmp3, tmp3);
tmp4 = const_vec;
DPADD_SB2_SH(src54_r, src76_r, filt0, filt1, tmp4, tmp4);
tmp5 = const_vec;
DPADD_SB2_SH(src65_r, src87_r, filt0, filt1, tmp5, tmp5);
HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3, HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -4016,13 +3956,13 @@ static void hevc_vt_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr, ...@@ -4016,13 +3956,13 @@ static void hevc_vt_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t offset, weight; int32_t offset, weight, constant;
v16i8 src0, src1, src2, src3, src4; v16i8 src0, src1, src2, src3, src4;
v8i16 in0, in1, in2, in3; v8i16 in0, in1, in2, in3;
v16i8 src10_r, src32_r, src21_r, src43_r; v16i8 src10_r, src32_r, src21_r, src43_r;
v8i16 tmp0, tmp1, tmp2, tmp3; v8i16 tmp0, tmp1, tmp2, tmp3;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= src_stride; src0_ptr -= src_stride;
...@@ -4030,9 +3970,10 @@ static void hevc_vt_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr, ...@@ -4030,9 +3970,10 @@ static void hevc_vt_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -4053,20 +3994,16 @@ static void hevc_vt_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr, ...@@ -4053,20 +3994,16 @@ static void hevc_vt_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr,
XORI_B2_128_SB(src3, src4); XORI_B2_128_SB(src3, src4);
ILVR_B2_SB(src3, src2, src4, src3, src32_r, src43_r); ILVR_B2_SB(src3, src2, src4, src3, src32_r, src43_r);
tmp0 = const_vec; tmp0 = HEVC_FILT_4TAP_SH(src10_r, src32_r, filt0, filt1);
DPADD_SB2_SH(src10_r, src32_r, filt0, filt1, tmp0, tmp0); tmp1 = HEVC_FILT_4TAP_SH(src21_r, src43_r, filt0, filt1);
tmp1 = const_vec;
DPADD_SB2_SH(src21_r, src43_r, filt0, filt1, tmp1, tmp1);
LD_SB2(src0_ptr, src_stride, src1, src2); LD_SB2(src0_ptr, src_stride, src1, src2);
src0_ptr += (2 * src_stride); src0_ptr += (2 * src_stride);
XORI_B2_128_SB(src1, src2); XORI_B2_128_SB(src1, src2);
ILVR_B2_SB(src1, src4, src2, src1, src10_r, src21_r); ILVR_B2_SB(src1, src4, src2, src1, src10_r, src21_r);
tmp2 = const_vec; tmp2 = HEVC_FILT_4TAP_SH(src32_r, src10_r, filt0, filt1);
DPADD_SB2_SH(src32_r, src10_r, filt0, filt1, tmp2, tmp2); tmp3 = HEVC_FILT_4TAP_SH(src43_r, src21_r, filt0, filt1);
tmp3 = const_vec;
DPADD_SB2_SH(src43_r, src21_r, filt0, filt1, tmp3, tmp3);
HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3, HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -4094,11 +4031,11 @@ static void hevc_vt_biwgt_4t_8w_msa(uint8_t *src0_ptr, ...@@ -4094,11 +4031,11 @@ static void hevc_vt_biwgt_4t_8w_msa(uint8_t *src0_ptr,
{ {
if (2 == height) { if (2 == height) {
hevc_vt_biwgt_4t_8x2_msa(src0_ptr, src_stride, src1_ptr, src2_stride, hevc_vt_biwgt_4t_8x2_msa(src0_ptr, src_stride, src1_ptr, src2_stride,
dst, dst_stride, filter, height, dst, dst_stride, filter,
weight0, weight1, offset0, offset1, rnd_val); weight0, weight1, offset0, offset1, rnd_val);
} else if (6 == height) { } else if (6 == height) {
hevc_vt_biwgt_4t_8x6_msa(src0_ptr, src_stride, src1_ptr, src2_stride, hevc_vt_biwgt_4t_8x6_msa(src0_ptr, src_stride, src1_ptr, src2_stride,
dst, dst_stride, filter, height, dst, dst_stride, filter,
weight0, weight1, offset0, offset1, rnd_val); weight0, weight1, offset0, offset1, rnd_val);
} else { } else {
hevc_vt_biwgt_4t_8x4multiple_msa(src0_ptr, src_stride, hevc_vt_biwgt_4t_8x4multiple_msa(src0_ptr, src_stride,
...@@ -4124,7 +4061,7 @@ static void hevc_vt_biwgt_4t_12w_msa(uint8_t *src0_ptr, ...@@ -4124,7 +4061,7 @@ static void hevc_vt_biwgt_4t_12w_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t offset, weight; int32_t offset, weight, constant;
v16i8 src0, src1, src2, src3, src4, src5; v16i8 src0, src1, src2, src3, src4, src5;
v8i16 in0, in1, in2, in3, in4, in5, in6, in7; v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
v16i8 src10_r, src32_r, src21_r, src43_r; v16i8 src10_r, src32_r, src21_r, src43_r;
...@@ -4132,7 +4069,7 @@ static void hevc_vt_biwgt_4t_12w_msa(uint8_t *src0_ptr, ...@@ -4132,7 +4069,7 @@ static void hevc_vt_biwgt_4t_12w_msa(uint8_t *src0_ptr,
v16i8 src10_l, src32_l, src54_l, src21_l, src43_l, src65_l; v16i8 src10_l, src32_l, src54_l, src21_l, src43_l, src65_l;
v16i8 src2110, src4332; v16i8 src2110, src4332;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= (1 * src_stride); src0_ptr -= (1 * src_stride);
...@@ -4140,9 +4077,10 @@ static void hevc_vt_biwgt_4t_12w_msa(uint8_t *src0_ptr, ...@@ -4140,9 +4077,10 @@ static void hevc_vt_biwgt_4t_12w_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -4170,12 +4108,9 @@ static void hevc_vt_biwgt_4t_12w_msa(uint8_t *src0_ptr, ...@@ -4170,12 +4108,9 @@ static void hevc_vt_biwgt_4t_12w_msa(uint8_t *src0_ptr,
ILVL_B2_SB(src3, src2, src4, src3, src32_l, src43_l); ILVL_B2_SB(src3, src2, src4, src3, src32_l, src43_l);
src4332 = (v16i8) __msa_ilvr_d((v2i64) src43_l, (v2i64) src32_l); src4332 = (v16i8) __msa_ilvr_d((v2i64) src43_l, (v2i64) src32_l);
tmp0 = const_vec; tmp0 = HEVC_FILT_4TAP_SH(src10_r, src32_r, filt0, filt1);
DPADD_SB2_SH(src10_r, src32_r, filt0, filt1, tmp0, tmp0); tmp1 = HEVC_FILT_4TAP_SH(src21_r, src43_r, filt0, filt1);
tmp1 = const_vec; tmp4 = HEVC_FILT_4TAP_SH(src2110, src4332, filt0, filt1);
DPADD_SB2_SH(src21_r, src43_r, filt0, filt1, tmp1, tmp1);
tmp4 = const_vec;
DPADD_SB2_SH(src2110, src4332, filt0, filt1, tmp4, tmp4);
LD_SB2(src0_ptr, src_stride, src5, src2); LD_SB2(src0_ptr, src_stride, src5, src2);
src0_ptr += (2 * src_stride); src0_ptr += (2 * src_stride);
...@@ -4184,12 +4119,9 @@ static void hevc_vt_biwgt_4t_12w_msa(uint8_t *src0_ptr, ...@@ -4184,12 +4119,9 @@ static void hevc_vt_biwgt_4t_12w_msa(uint8_t *src0_ptr,
ILVL_B2_SB(src5, src4, src2, src5, src54_l, src65_l); ILVL_B2_SB(src5, src4, src2, src5, src54_l, src65_l);
src2110 = (v16i8) __msa_ilvr_d((v2i64) src65_l, (v2i64) src54_l); src2110 = (v16i8) __msa_ilvr_d((v2i64) src65_l, (v2i64) src54_l);
tmp2 = const_vec; tmp2 = HEVC_FILT_4TAP_SH(src32_r, src10_r, filt0, filt1);
DPADD_SB2_SH(src32_r, src10_r, filt0, filt1, tmp2, tmp2); tmp3 = HEVC_FILT_4TAP_SH(src43_r, src21_r, filt0, filt1);
tmp3 = const_vec; tmp5 = HEVC_FILT_4TAP_SH(src4332, src2110, filt0, filt1);
DPADD_SB2_SH(src43_r, src21_r, filt0, filt1, tmp3, tmp3);
tmp5 = const_vec;
DPADD_SB2_SH(src4332, src2110, filt0, filt1, tmp5, tmp5);
HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3, HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -4220,14 +4152,14 @@ static void hevc_vt_biwgt_4t_16w_msa(uint8_t *src0_ptr, ...@@ -4220,14 +4152,14 @@ static void hevc_vt_biwgt_4t_16w_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t offset, weight; int32_t offset, weight, constant;
v16i8 src0, src1, src2, src3, src4, src5; v16i8 src0, src1, src2, src3, src4, src5;
v8i16 in0, in1, in2, in3; v8i16 in0, in1, in2, in3;
v16i8 src10_r, src32_r, src21_r, src43_r; v16i8 src10_r, src32_r, src21_r, src43_r;
v16i8 src10_l, src32_l, src21_l, src43_l; v16i8 src10_l, src32_l, src21_l, src43_l;
v8i16 tmp0, tmp1, tmp2, tmp3; v8i16 tmp0, tmp1, tmp2, tmp3;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= src_stride; src0_ptr -= src_stride;
...@@ -4235,9 +4167,10 @@ static void hevc_vt_biwgt_4t_16w_msa(uint8_t *src0_ptr, ...@@ -4235,9 +4167,10 @@ static void hevc_vt_biwgt_4t_16w_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -4261,14 +4194,10 @@ static void hevc_vt_biwgt_4t_16w_msa(uint8_t *src0_ptr, ...@@ -4261,14 +4194,10 @@ static void hevc_vt_biwgt_4t_16w_msa(uint8_t *src0_ptr,
ILVR_B2_SB(src3, src2, src4, src3, src32_r, src43_r); ILVR_B2_SB(src3, src2, src4, src3, src32_r, src43_r);
ILVL_B2_SB(src3, src2, src4, src3, src32_l, src43_l); ILVL_B2_SB(src3, src2, src4, src3, src32_l, src43_l);
tmp0 = const_vec; tmp0 = HEVC_FILT_4TAP_SH(src10_r, src32_r, filt0, filt1);
DPADD_SB2_SH(src10_r, src32_r, filt0, filt1, tmp0, tmp0); tmp1 = HEVC_FILT_4TAP_SH(src21_r, src43_r, filt0, filt1);
tmp1 = const_vec; tmp2 = HEVC_FILT_4TAP_SH(src10_l, src32_l, filt0, filt1);
DPADD_SB2_SH(src21_r, src43_r, filt0, filt1, tmp1, tmp1); tmp3 = HEVC_FILT_4TAP_SH(src21_l, src43_l, filt0, filt1);
tmp2 = const_vec;
DPADD_SB2_SH(src10_l, src32_l, filt0, filt1, tmp2, tmp2);
tmp3 = const_vec;
DPADD_SB2_SH(src21_l, src43_l, filt0, filt1, tmp3, tmp3);
HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3, HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3,
in0, in1, in2, in3, in0, in1, in2, in3,
...@@ -4287,14 +4216,10 @@ static void hevc_vt_biwgt_4t_16w_msa(uint8_t *src0_ptr, ...@@ -4287,14 +4216,10 @@ static void hevc_vt_biwgt_4t_16w_msa(uint8_t *src0_ptr,
ILVR_B2_SB(src5, src4, src2, src5, src10_r, src21_r); ILVR_B2_SB(src5, src4, src2, src5, src10_r, src21_r);
ILVL_B2_SB(src5, src4, src2, src5, src10_l, src21_l); ILVL_B2_SB(src5, src4, src2, src5, src10_l, src21_l);
tmp0 = const_vec; tmp0 = HEVC_FILT_4TAP_SH(src32_r, src10_r, filt0, filt1);
DPADD_SB2_SH(src32_r, src10_r, filt0, filt1, tmp0, tmp0); tmp1 = HEVC_FILT_4TAP_SH(src43_r, src21_r, filt0, filt1);
tmp1 = const_vec; tmp2 = HEVC_FILT_4TAP_SH(src32_l, src10_l, filt0, filt1);
DPADD_SB2_SH(src43_r, src21_r, filt0, filt1, tmp1, tmp1); tmp3 = HEVC_FILT_4TAP_SH(src43_l, src21_l, filt0, filt1);
tmp2 = const_vec;
DPADD_SB2_SH(src32_l, src10_l, filt0, filt1, tmp2, tmp2);
tmp3 = const_vec;
DPADD_SB2_SH(src43_l, src21_l, filt0, filt1, tmp3, tmp3);
HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3, HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp2, tmp3,
in0, in1, in2, in3, in0, in1, in2, in3,
weight_vec, rnd_vec, offset_vec, weight_vec, rnd_vec, offset_vec,
...@@ -4321,7 +4246,7 @@ static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr, ...@@ -4321,7 +4246,7 @@ static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr,
int32_t rnd_val) int32_t rnd_val)
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
int32_t offset, weight; int32_t offset, weight, constant;
v16i8 src0, src1, src2, src3, src4, src5; v16i8 src0, src1, src2, src3, src4, src5;
v16i8 src6, src7, src8, src9, src10, src11; v16i8 src6, src7, src8, src9, src10, src11;
v8i16 in0, in1, in2, in3, in4, in5; v8i16 in0, in1, in2, in3, in4, in5;
...@@ -4330,7 +4255,7 @@ static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr, ...@@ -4330,7 +4255,7 @@ static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr,
v16i8 src21_r, src43_r, src87_r, src109_r; v16i8 src21_r, src43_r, src87_r, src109_r;
v8i16 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5; v8i16 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= src_stride; src0_ptr -= src_stride;
...@@ -4338,9 +4263,10 @@ static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr, ...@@ -4338,9 +4263,10 @@ static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -4376,19 +4302,13 @@ static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr, ...@@ -4376,19 +4302,13 @@ static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr,
XORI_B2_128_SB(src9, src10); XORI_B2_128_SB(src9, src10);
ILVR_B2_SB(src9, src8, src10, src9, src98_r, src109_r); ILVR_B2_SB(src9, src8, src10, src9, src98_r, src109_r);
/* 16width */ /* 16width */
tmp0 = const_vec; tmp0 = HEVC_FILT_4TAP_SH(src10_r, src32_r, filt0, filt1);
DPADD_SB2_SH(src10_r, src32_r, filt0, filt1, tmp0, tmp0); tmp4 = HEVC_FILT_4TAP_SH(src10_l, src32_l, filt0, filt1);
tmp4 = const_vec; tmp1 = HEVC_FILT_4TAP_SH(src21_r, src43_r, filt0, filt1);
DPADD_SB2_SH(src10_l, src32_l, filt0, filt1, tmp4, tmp4); tmp5 = HEVC_FILT_4TAP_SH(src21_l, src43_l, filt0, filt1);
tmp1 = const_vec;
DPADD_SB2_SH(src21_r, src43_r, filt0, filt1, tmp1, tmp1);
tmp5 = const_vec;
DPADD_SB2_SH(src21_l, src43_l, filt0, filt1, tmp5, tmp5);
/* 8width */ /* 8width */
tmp2 = const_vec; tmp2 = HEVC_FILT_4TAP_SH(src76_r, src98_r, filt0, filt1);
DPADD_SB2_SH(src76_r, src98_r, filt0, filt1, tmp2, tmp2); tmp3 = HEVC_FILT_4TAP_SH(src87_r, src109_r, filt0, filt1);
tmp3 = const_vec;
DPADD_SB2_SH(src87_r, src109_r, filt0, filt1, tmp3, tmp3);
/* 16width */ /* 16width */
HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp4, tmp5, HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp4, tmp5,
in0, in1, in2, in3, in0, in1, in2, in3,
...@@ -4421,19 +4341,13 @@ static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr, ...@@ -4421,19 +4341,13 @@ static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr,
XORI_B2_128_SB(src11, src8); XORI_B2_128_SB(src11, src8);
ILVR_B2_SB(src11, src10, src8, src11, src76_r, src87_r); ILVR_B2_SB(src11, src10, src8, src11, src76_r, src87_r);
/* 16width */ /* 16width */
tmp0 = const_vec; tmp0 = HEVC_FILT_4TAP_SH(src32_r, src10_r, filt0, filt1);
DPADD_SB2_SH(src32_r, src10_r, filt0, filt1, tmp0, tmp0); tmp4 = HEVC_FILT_4TAP_SH(src32_l, src10_l, filt0, filt1);
tmp4 = const_vec; tmp1 = HEVC_FILT_4TAP_SH(src43_r, src21_r, filt0, filt1);
DPADD_SB2_SH(src32_l, src10_l, filt0, filt1, tmp4, tmp4); tmp5 = HEVC_FILT_4TAP_SH(src43_l, src21_l, filt0, filt1);
tmp1 = const_vec;
DPADD_SB2_SH(src43_r, src21_r, filt0, filt1, tmp1, tmp1);
tmp5 = const_vec;
DPADD_SB2_SH(src43_l, src21_l, filt0, filt1, tmp5, tmp5);
/* 8width */ /* 8width */
tmp2 = const_vec; tmp2 = HEVC_FILT_4TAP_SH(src98_r, src76_r, filt0, filt1);
DPADD_SB2_SH(src98_r, src76_r, filt0, filt1, tmp2, tmp2); tmp3 = HEVC_FILT_4TAP_SH(src109_r, src87_r, filt0, filt1);
tmp3 = const_vec;
DPADD_SB2_SH(src109_r, src87_r, filt0, filt1, tmp3, tmp3);
/* 16width */ /* 16width */
HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp4, tmp5, HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp4, tmp5,
in0, in1, in2, in3, in0, in1, in2, in3,
...@@ -4470,7 +4384,7 @@ static void hevc_vt_biwgt_4t_32w_msa(uint8_t *src0_ptr, ...@@ -4470,7 +4384,7 @@ static void hevc_vt_biwgt_4t_32w_msa(uint8_t *src0_ptr,
{ {
uint32_t loop_cnt; uint32_t loop_cnt;
uint8_t *dst_tmp = dst + 16; uint8_t *dst_tmp = dst + 16;
int32_t offset, weight; int32_t offset, weight, constant;
v16i8 src0, src1, src2, src3, src4, src6, src7, src8, src9, src10; v16i8 src0, src1, src2, src3, src4, src6, src7, src8, src9, src10;
v8i16 in0, in1, in2, in3, in4, in5, in6, in7; v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
v16i8 src10_r, src32_r, src76_r, src98_r; v16i8 src10_r, src32_r, src76_r, src98_r;
...@@ -4479,7 +4393,7 @@ static void hevc_vt_biwgt_4t_32w_msa(uint8_t *src0_ptr, ...@@ -4479,7 +4393,7 @@ static void hevc_vt_biwgt_4t_32w_msa(uint8_t *src0_ptr,
v16i8 src10_l, src32_l, src76_l, src98_l; v16i8 src10_l, src32_l, src76_l, src98_l;
v16i8 src21_l, src43_l, src87_l, src109_l; v16i8 src21_l, src43_l, src87_l, src109_l;
v8i16 filt0, filt1; v8i16 filt0, filt1;
v8i16 filter_vec, const_vec; v8i16 filter_vec;
v4i32 weight_vec, offset_vec, rnd_vec; v4i32 weight_vec, offset_vec, rnd_vec;
src0_ptr -= src_stride; src0_ptr -= src_stride;
...@@ -4487,9 +4401,10 @@ static void hevc_vt_biwgt_4t_32w_msa(uint8_t *src0_ptr, ...@@ -4487,9 +4401,10 @@ static void hevc_vt_biwgt_4t_32w_msa(uint8_t *src0_ptr,
offset = (offset0 + offset1) << rnd_val; offset = (offset0 + offset1) << rnd_val;
weight0 = weight0 & 0x0000FFFF; weight0 = weight0 & 0x0000FFFF;
weight = weight0 | (weight1 << 16); weight = weight0 | (weight1 << 16);
constant = 128 * weight1;
constant <<= 6;
offset += constant;
const_vec = __msa_ldi_h(128);
const_vec <<= 6;
offset_vec = __msa_fill_w(offset); offset_vec = __msa_fill_w(offset);
weight_vec = __msa_fill_w(weight); weight_vec = __msa_fill_w(weight);
rnd_vec = __msa_fill_w(rnd_val + 1); rnd_vec = __msa_fill_w(rnd_val + 1);
...@@ -4519,14 +4434,10 @@ static void hevc_vt_biwgt_4t_32w_msa(uint8_t *src0_ptr, ...@@ -4519,14 +4434,10 @@ static void hevc_vt_biwgt_4t_32w_msa(uint8_t *src0_ptr,
ILVL_B2_SB(src3, src2, src4, src3, src32_l, src43_l); ILVL_B2_SB(src3, src2, src4, src3, src32_l, src43_l);
/* 16width */ /* 16width */
tmp0 = const_vec; tmp0 = HEVC_FILT_4TAP_SH(src10_r, src32_r, filt0, filt1);
DPADD_SB2_SH(src10_r, src32_r, filt0, filt1, tmp0, tmp0); tmp4 = HEVC_FILT_4TAP_SH(src10_l, src32_l, filt0, filt1);
tmp4 = const_vec; tmp1 = HEVC_FILT_4TAP_SH(src21_r, src43_r, filt0, filt1);
DPADD_SB2_SH(src10_l, src32_l, filt0, filt1, tmp4, tmp4); tmp5 = HEVC_FILT_4TAP_SH(src21_l, src43_l, filt0, filt1);
tmp1 = const_vec;
DPADD_SB2_SH(src21_r, src43_r, filt0, filt1, tmp1, tmp1);
tmp5 = const_vec;
DPADD_SB2_SH(src21_l, src43_l, filt0, filt1, tmp5, tmp5);
/* 16width */ /* 16width */
HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp4, tmp5, HEVC_BIW_RND_CLIP4(tmp0, tmp1, tmp4, tmp5,
in0, in1, in2, in3, in0, in1, in2, in3,
...@@ -4553,14 +4464,10 @@ static void hevc_vt_biwgt_4t_32w_msa(uint8_t *src0_ptr, ...@@ -4553,14 +4464,10 @@ static void hevc_vt_biwgt_4t_32w_msa(uint8_t *src0_ptr,
ILVR_B2_SB(src9, src8, src10, src9, src98_r, src109_r); ILVR_B2_SB(src9, src8, src10, src9, src98_r, src109_r);
ILVL_B2_SB(src9, src8, src10, src9, src98_l, src109_l); ILVL_B2_SB(src9, src8, src10, src9, src98_l, src109_l);
/* next 16width */ /* next 16width */
tmp2 = const_vec; tmp2 = HEVC_FILT_4TAP_SH(src76_r, src98_r, filt0, filt1);
DPADD_SB2_SH(src76_r, src98_r, filt0, filt1, tmp2, tmp2); tmp6 = HEVC_FILT_4TAP_SH(src76_l, src98_l, filt0, filt1);
tmp6 = const_vec; tmp3 = HEVC_FILT_4TAP_SH(src87_r, src109_r, filt0, filt1);
DPADD_SB2_SH(src76_l, src98_l, filt0, filt1, tmp6, tmp6); tmp7 = HEVC_FILT_4TAP_SH(src87_l, src109_l, filt0, filt1);
tmp3 = const_vec;
DPADD_SB2_SH(src87_r, src109_r, filt0, filt1, tmp3, tmp3);
tmp7 = const_vec;
DPADD_SB2_SH(src87_l, src109_l, filt0, filt1, tmp7, tmp7);
/* next 16width */ /* next 16width */
HEVC_BIW_RND_CLIP4(tmp2, tmp3, tmp6, tmp7, HEVC_BIW_RND_CLIP4(tmp2, tmp3, tmp6, tmp7,
in4, in5, in6, in7, in4, in5, in6, in7,
......
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