add_proto qw/unsigned int vp9_variance16x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
specialize qw/vp9_variance16x16 mmx/, "$sse2_x86inc", "$avx2_x86inc";
+add_proto qw/void vp9_get_sse_sum_16x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum";
+specialize qw/vp9_get_sse_sum_16x16 sse2/;
+$vp9_get_sse_sum_16x16_sse2=vp9_get16x16var_sse2;
+
add_proto qw/unsigned int vp9_variance16x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
specialize qw/vp9_variance16x8 mmx/, "$sse2_x86inc";
add_proto qw/unsigned int vp9_variance8x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
specialize qw/vp9_variance8x8 mmx/, "$sse2_x86inc";
+add_proto qw/void vp9_get_sse_sum_8x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum";
+specialize qw/vp9_get_sse_sum_8x8 sse2/;
+$vp9_get_sse_sum_8x8_sse2=vp9_get8x8var_sse2;
+
add_proto qw/unsigned int vp9_variance8x4/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
specialize qw/vp9_variance8x4/, "$sse2_x86inc";
unsigned int var;
} diff;
-static void get_sse_sum_8x8(const uint8_t *src, int src_stride,
- const uint8_t *ref, int ref_stride,
- unsigned int *sse, int *sum) {
- variance(src, src_stride, ref, ref_stride, 8, 8, sse, sum);
-}
-
-static void get_sse_sum_16x16(const uint8_t *src, int src_stride,
- const uint8_t *ref, int ref_stride,
- unsigned int *sse, int *sum) {
- variance(src, src_stride, ref, ref_stride, 16, 16, sse, sum);
-}
-
static unsigned int get_sby_perpixel_variance(VP9_COMP *cpi,
const struct buf_2d *ref,
BLOCK_SIZE bs) {
unsigned int sse = 0;
int sum = 0;
if (x_idx < pixels_wide && y_idx < pixels_high)
- get_sse_sum_8x8(s + y_idx * sp + x_idx, sp,
- d + y_idx * dp + x_idx, dp, &sse, &sum);
+ vp9_get_sse_sum_8x8(s + y_idx * sp + x_idx, sp,
+ d + y_idx * dp + x_idx, dp, &sse, &sum);
fill_variance(sse, sum, 64, &vst->split[k].part_variances.none);
}
}
int b_offset = b_mi_row * MI_SIZE * src_stride +
b_mi_col * MI_SIZE;
- get_sse_sum_16x16(src + b_offset, src_stride,
- pre_src + b_offset, pre_stride,
- &d16[j].sse, &d16[j].sum);
+ vp9_get_sse_sum_16x16(src + b_offset, src_stride,
+ pre_src + b_offset, pre_stride,
+ &d16[j].sse, &d16[j].sum);
d16[j].var = d16[j].sse -
(((uint32_t)d16[j].sum * d16[j].sum) >> 8);
return vp9_variance##W##x##H##_c(temp3, W, dst, dst_stride, sse); \
}
+
+void vp9_get_sse_sum_16x16_c(const uint8_t *src_ptr, int source_stride,
+ const uint8_t *ref_ptr, int ref_stride,
+ unsigned int *sse, int *sum) {
+ variance(src_ptr, source_stride, ref_ptr, ref_stride, 16, 16, sse, sum);
+}
+
+void vp9_get_sse_sum_8x8_c(const uint8_t *src_ptr, int source_stride,
+ const uint8_t *ref_ptr, int ref_stride,
+ unsigned int *sse, int *sum) {
+ variance(src_ptr, source_stride, ref_ptr, ref_stride, 8, 8, sse, sum);
+}
+
unsigned int vp9_mse16x16_c(const uint8_t *src, int src_stride,
const uint8_t *ref, int ref_stride,
unsigned int *sse) {