SRARI_H2_SH(out0, out1, 7);
SAT_SH2_SH(out0, out1, 7);
out = PCKEV_XORI128_UB(out0, out1);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
}
static void common_hz_6t_4x8_msa(uint8_t *src, int32_t src_stride,
SRARI_H4_SH(out0, out1, out2, out3, 7);
SAT_SH4_SH(out0, out1, out2, out3, 7);
out = PCKEV_XORI128_UB(out0, out1);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
- dst += (4 * dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
out = PCKEV_XORI128_UB(out2, out3);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
}
void ff_put_vp8_epel4_h6_msa(uint8_t *dst, ptrdiff_t dst_stride,
SAT_SH4_SH(out0, out1, out2, out3, 7);
tmp0 = PCKEV_XORI128_UB(out0, out1);
tmp1 = PCKEV_XORI128_UB(out2, out3);
- ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+ ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
dst += (4 * dst_stride);
for (loop_cnt = (height >> 2) - 1; loop_cnt--;) {
SAT_SH4_SH(out0, out1, out2, out3, 7);
tmp0 = PCKEV_XORI128_UB(out0, out1);
tmp1 = PCKEV_XORI128_UB(out2, out3);
- ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+ ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
dst += (4 * dst_stride);
}
}
SRARI_H2_SH(out10, out32, 7);
SAT_SH2_SH(out10, out32, 7);
out = PCKEV_XORI128_UB(out10, out32);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
dst += (4 * dst_stride);
src2110 = src6554;
SAT_SH4_SH(out0_r, out1_r, out2_r, out3_r, 7);
tmp0 = PCKEV_XORI128_UB(out0_r, out1_r);
tmp1 = PCKEV_XORI128_UB(out2_r, out3_r);
- ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+ ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
dst += (4 * dst_stride);
src10_r = src76_r;
SRARI_H2_SH(tmp0, tmp1, 7);
SAT_SH2_SH(tmp0, tmp1, 7);
out = PCKEV_XORI128_UB(tmp0, tmp1);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
dst += (4 * dst_stride);
hz_out3 = hz_out7;
SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
vec0 = PCKEV_XORI128_UB(tmp0, tmp1);
vec1 = PCKEV_XORI128_UB(tmp2, tmp3);
- ST8x4_UB(vec0, vec1, dst, dst_stride);
+ ST_D4(vec0, vec1, 0, 1, 0, 1, dst, dst_stride);
dst += (4 * dst_stride);
hz_out4 = hz_out8;
SRARI_H2_SH(out0, out1, 7);
SAT_SH2_SH(out0, out1, 7);
out = PCKEV_XORI128_UB(out0, out1);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
}
static void common_hz_4t_4x8_msa(uint8_t *src, int32_t src_stride,
SRARI_H4_SH(out0, out1, out2, out3, 7);
SAT_SH4_SH(out0, out1, out2, out3, 7);
out = PCKEV_XORI128_UB(out0, out1);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
- dst += (4 * dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
out = PCKEV_XORI128_UB(out2, out3);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
}
static void common_hz_4t_4x16_msa(uint8_t *src, int32_t src_stride,
SRARI_H4_SH(out0, out1, out2, out3, 7);
SAT_SH4_SH(out0, out1, out2, out3, 7);
out = PCKEV_XORI128_UB(out0, out1);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
dst += (4 * dst_stride);
out = PCKEV_XORI128_UB(out2, out3);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
dst += (4 * dst_stride);
LD_SB8(src, src_stride, src0, src1, src2, src3, src4, src5, src6, src7);
SRARI_H4_SH(out0, out1, out2, out3, 7);
SAT_SH4_SH(out0, out1, out2, out3, 7);
out = PCKEV_XORI128_UB(out0, out1);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
dst += (4 * dst_stride);
out = PCKEV_XORI128_UB(out2, out3);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
}
void ff_put_vp8_epel4_h4_msa(uint8_t *dst, ptrdiff_t dst_stride,
SAT_SH4_SH(out0, out1, out2, out3, 7);
tmp0 = PCKEV_XORI128_UB(out0, out1);
tmp1 = PCKEV_XORI128_UB(out2, out3);
- ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+ ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
dst += (4 * dst_stride);
}
}
SRARI_H2_SH(out10, out32, 7);
SAT_SH2_SH(out10, out32, 7);
out = PCKEV_XORI128_UB(out10, out32);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
dst += (4 * dst_stride);
}
}
SAT_SH4_SH(out0_r, out1_r, out2_r, out3_r, 7);
tmp0 = PCKEV_XORI128_UB(out0_r, out1_r);
tmp1 = PCKEV_XORI128_UB(out2_r, out3_r);
- ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+ ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
dst += (4 * dst_stride);
src10_r = src98_r;
SRARI_H2_SH(tmp0, tmp1, 7);
SAT_SH2_SH(tmp0, tmp1, 7);
out = PCKEV_XORI128_UB(tmp0, tmp1);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
dst += (4 * dst_stride);
hz_out1 = hz_out5;
SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
out0 = PCKEV_XORI128_UB(tmp0, tmp1);
out1 = PCKEV_XORI128_UB(tmp2, tmp3);
- ST8x4_UB(out0, out1, dst, dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
dst += (4 * dst_stride);
vec0 = vec4;
SAT_SH2_SH(tmp0, tmp1, 7);
PCKEV_B2_UB(tmp0, tmp0, tmp1, tmp1, res0, res1);
XORI_B2_128_UB(res0, res1);
- ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
+ ST_W2(res0, 0, 1, dst, dst_stride);
+ ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
dst += (4 * dst_stride);
hz_out1 = hz_out5;
SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
out0 = PCKEV_XORI128_UB(tmp0, tmp1);
out1 = PCKEV_XORI128_UB(tmp2, tmp3);
- ST8x4_UB(out0, out1, dst, dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
dst += (4 * dst_stride);
}
}
SRARI_H2_SH(tmp0, tmp1, 7);
SAT_SH2_SH(tmp0, tmp1, 7);
out = PCKEV_XORI128_UB(tmp0, tmp1);
- ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
dst += (4 * dst_stride);
hz_out3 = hz_out7;
SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
vec0 = PCKEV_XORI128_UB(tmp0, tmp1);
vec1 = PCKEV_XORI128_UB(tmp2, tmp3);
- ST8x4_UB(vec0, vec1, dst, dst_stride);
+ ST_D4(vec0, vec1, 0, 1, 0, 1, dst, dst_stride);
dst += (4 * dst_stride);
hz_out4 = hz_out8;
DOTP_UB2_UH(vec0, vec1, filt0, filt0, vec2, vec3);
SRARI_H2_UH(vec2, vec3, 7);
PCKEV_B2_UB(vec2, vec2, vec3, vec3, res0, res1);
- ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
+ ST_W2(res0, 0, 1, dst, dst_stride);
+ ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
}
static void common_hz_2t_4x8_msa(uint8_t *src, int32_t src_stride,
SRARI_H4_UH(vec4, vec5, vec6, vec7, 7);
PCKEV_B4_SB(vec4, vec4, vec5, vec5, vec6, vec6, vec7, vec7,
res0, res1, res2, res3);
- ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
- dst += (4 * dst_stride);
- ST4x4_UB(res2, res3, 0, 1, 0, 1, dst, dst_stride);
+ ST_W2(res0, 0, 1, dst, dst_stride);
+ ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
+ ST_W2(res2, 0, 1, dst + 4 * dst_stride, dst_stride);
+ ST_W2(res3, 0, 1, dst + 6 * dst_stride, dst_stride);
}
void ff_put_vp8_bilinear4_h_msa(uint8_t *dst, ptrdiff_t dst_stride,
vec0, vec1, vec2, vec3);
SRARI_H4_UH(vec0, vec1, vec2, vec3, 7);
PCKEV_B2_SB(vec1, vec0, vec3, vec2, src0, src1);
- ST8x4_UB(src0, src1, dst, dst_stride);
+ ST_D4(src0, src1, 0, 1, 0, 1, dst, dst_stride);
}
static void common_hz_2t_8x8mult_msa(uint8_t *src, int32_t src_stride,
src += (4 * src_stride);
PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
- ST8x4_UB(out0, out1, dst, dst_stride);
- dst += (4 * dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
VSHF_B2_UH(src0, src0, src1, src1, mask, mask, vec0, vec1);
VSHF_B2_UH(src2, src2, src3, src3, mask, mask, vec2, vec3);
vec0, vec1, vec2, vec3);
SRARI_H4_UH(vec0, vec1, vec2, vec3, 7);
PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
- ST8x4_UB(out0, out1, dst, dst_stride);
- dst += (4 * dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
+ dst += (8 * dst_stride);
if (16 == height) {
LD_SB4(src, src_stride, src0, src1, src2, src3);
src += (4 * src_stride);
PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
- ST8x4_UB(out0, out1, dst, dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
VSHF_B2_UH(src0, src0, src1, src1, mask, mask, vec0, vec1);
VSHF_B2_UH(src2, src2, src3, src3, mask, mask, vec2, vec3);
vec0, vec1, vec2, vec3);
SRARI_H4_UH(vec0, vec1, vec2, vec3, 7);
PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
- ST8x4_UB(out0, out1, dst + 4 * dst_stride, dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
}
}
SRARI_H2_UH(tmp0, tmp1, 7);
SAT_UH2_UH(tmp0, tmp1, 7);
src2110 = __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
- ST4x4_UB(src2110, src2110, 0, 1, 2, 3, dst, dst_stride);
+ ST_W4(src2110, 0, 1, 2, 3, dst, dst_stride);
}
static void common_vt_2t_4x8_msa(uint8_t *src, int32_t src_stride,
SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, src2110, src4332);
- ST4x4_UB(src2110, src2110, 0, 1, 2, 3, dst, dst_stride);
- ST4x4_UB(src4332, src4332, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
+ ST_W8(src2110, src4332, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
}
void ff_put_vp8_bilinear4_v_msa(uint8_t *dst, ptrdiff_t dst_stride,
SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
- ST8x4_UB(out0, out1, dst, dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
}
static void common_vt_2t_8x8mult_msa(uint8_t *src, int32_t src_stride,
SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
- ST8x4_UB(out0, out1, dst, dst_stride);
- dst += (4 * dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
DOTP_UB4_UH(vec4, vec5, vec6, vec7, filt0, filt0, filt0, filt0,
tmp0, tmp1, tmp2, tmp3);
SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
- ST8x4_UB(out0, out1, dst, dst_stride);
- dst += (4 * dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
+ dst += (8 * dst_stride);
src0 = src8;
}
SRARI_H2_UH(tmp0, tmp1, 7);
SAT_UH2_UH(tmp0, tmp1, 7);
PCKEV_B2_UB(tmp0, tmp0, tmp1, tmp1, res0, res1);
- ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
+ ST_W2(res0, 0, 1, dst, dst_stride);
+ ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
}
static void common_hv_2ht_2vt_4x8_msa(uint8_t *src, int32_t src_stride,
SAT_UH4_UH(vec4, vec5, vec6, vec7, 7);
PCKEV_B4_SB(vec4, vec4, vec5, vec5, vec6, vec6, vec7, vec7,
res0, res1, res2, res3);
- ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
- dst += (4 * dst_stride);
- ST4x4_UB(res2, res3, 0, 1, 0, 1, dst, dst_stride);
+ ST_W2(res0, 0, 1, dst, dst_stride);
+ ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
+ ST_W2(res2, 0, 1, dst + 4 * dst_stride, dst_stride);
+ ST_W2(res3, 0, 1, dst + 6 * dst_stride, dst_stride);
}
void ff_put_vp8_bilinear4_hv_msa(uint8_t *dst, ptrdiff_t dst_stride,
SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
- ST8x4_UB(out0, out1, dst, dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
}
static void common_hv_2ht_2vt_8x8mult_msa(uint8_t *src, int32_t src_stride,
SRARI_H2_UH(tmp3, tmp4, 7);
SAT_UH2_UH(tmp3, tmp4, 7);
PCKEV_B2_SB(tmp2, tmp1, tmp4, tmp3, out0, out1);
- ST8x4_UB(out0, out1, dst, dst_stride);
- dst += (4 * dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
hz_out1 = HORIZ_2TAP_FILT_UH(src1, src1, mask, filt_hz, 7);
vec0 = (v16u8) __msa_ilvev_b((v16i8) hz_out1, (v16i8) hz_out0);
SRARI_H4_UH(tmp5, tmp6, tmp7, tmp8, 7);
SAT_UH4_UH(tmp5, tmp6, tmp7, tmp8, 7);
PCKEV_B2_SB(tmp6, tmp5, tmp8, tmp7, out0, out1);
- ST8x4_UB(out0, out1, dst, dst_stride);
- dst += (4 * dst_stride);
+ ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
+ dst += (8 * dst_stride);
}
}