unsigned int vpx_variance64x32_avx2(const uint8_t *src, int src_stride,
const uint8_t *ref, int ref_stride,
unsigned int *sse) {
int sum;
variance_avx2(src, src_stride, ref, ref_stride, 64, 32,
sse, &sum, vpx_get32x32var_avx2, 32);
return *sse - (((int64_t)sum * sum) >> 11);
}
unsigned int vpx_variance16x16_avx2(const uint8_t *src, int src_stride,
const uint8_t *ref, int ref_stride,
unsigned int *sse) {
int sum;
variance_avx2(src, src_stride, ref, ref_stride, 16, 16,
sse, &sum, vpx_get16x16var_avx2, 16);
return *sse - (((unsigned int)sum * sum) >> 8);
}
unsigned int aom_variance32x16_avx2(const uint8_t *src, int src_stride,
const uint8_t *ref, int ref_stride,
unsigned int *sse) {
int sum;
variance_avx2(src, src_stride, ref, ref_stride, 32, 16, sse, &sum,
aom_get32x32var_avx2, 32);
return *sse - (((int64_t)sum * sum) >> 9);
}
unsigned int aom_variance16x16_avx2(const uint8_t *src, int src_stride,
const uint8_t *ref, int ref_stride,
unsigned int *sse) {
int sum;
variance_avx2(src, src_stride, ref, ref_stride, 16, 16, sse, &sum,
aom_get16x16var_avx2, 16);
return *sse - (((uint32_t)((int64_t)sum * sum)) >> 8);
}
unsigned int vp9_variance32x32_avx2(const uint8_t *src_ptr,
int source_stride,
const uint8_t *ref_ptr,
int recon_stride,
unsigned int *sse) {
unsigned int var;
int avg;
// processing 32 elements vertically in parallel
variance_avx2(src_ptr, source_stride, ref_ptr, recon_stride, 32, 32,
&var, &avg, vp9_get32x32var_avx2, 32);
*sse = var;
return (var - (((int64_t)avg * avg) >> 10));
}
unsigned int vp9_variance16x16_avx2
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse) {
unsigned int var;
int avg;
variance_avx2(src_ptr, source_stride, ref_ptr, recon_stride, 16, 16,
&var, &avg, vp9_get16x16var_avx2, 16);
*sse = var;
return (var - (((unsigned int)avg * avg) >> 8));
}
