generic-library/vpx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Code refactoring to remove duplication in masked sub pixel var fns

Browse Source 
Change-Id: Idcc93ff94ed2c68367225b6bebfcae0b7952f074
This commit is contained in:
Julia Robson 2015-12-03 17:57:35 +00:00
parent 9227a372d3
commit 997595e8a6
1 changed files with 53 additions and 273 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

326
vp9/encoder/x86/vp9_masked_variance_intrin_ssse3.c
View File

@ -346,6 +346,15 @@ MASKED_VAR4XH(8)
// Sub pixel versions
//////////////////////////////////////////////////////////////////////////////
typedef __m128i (*filter_fn_t)(__m128i v_a_b, __m128i v_b_b,
__m128i v_filter_b);
static INLINE __m128i apply_filter8(const __m128i v_a_b, const __m128i v_b_b,
const __m128i v_filter_b) {
(void) v_filter_b;
return _mm_avg_epu8(v_a_b, v_b_b);
}
static INLINE __m128i apply_filter(const __m128i v_a_b, const __m128i v_b_b,
const __m128i v_filter_b) {
const __m128i v_rounding_w = _mm_set1_epi16(1 << (FILTER_BITS - 1));
@ -438,51 +447,10 @@ static INLINE int calc_masked_variance(__m128i v_sum_d, __m128i v_sse_q,
// Functions for width (W) >= 16
unsigned int vp9_masked_subpel_varWxH_xzero_yhalf(
const uint8_t *src, int src_stride,
const uint8_t *dst, int dst_stride, const uint8_t *msk, int msk_stride,
unsigned int *sse, int w, int h) {
int i, j;
__m128i v_src0_b, v_src1_b, v_res_b, v_dst_b, v_msk_b;
__m128i v_sum_d = _mm_setzero_si128();
__m128i v_sse_q = _mm_setzero_si128();
for (j = 0; j < w; j += 16) {
// Load the first row ready
v_src0_b = _mm_loadu_si128((const __m128i*)(src + j));
// Process 2 rows at a time
for (i = 0; i < h; i += 2) {
// Load the next row apply the filter
v_src1_b = _mm_loadu_si128((const __m128i*)(src + j + src_stride));
v_res_b = _mm_avg_epu8(v_src0_b, v_src1_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + j));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + j));
sum_and_sse(v_res_b, v_dst_b, v_msk_b, &v_sum_d, &v_sse_q);
// Load the next row apply the filter
v_src0_b = _mm_loadu_si128((const __m128i*)(src + j + src_stride * 2));
v_res_b = _mm_avg_epu8(v_src1_b, v_src0_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + j + dst_stride));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + j + msk_stride));
sum_and_sse(v_res_b, v_dst_b, v_msk_b, &v_sum_d, &v_sse_q);
// Move onto the next block of rows
src += src_stride * 2;
dst += dst_stride * 2;
msk += msk_stride * 2;
}
// Reset to the top of the block
src -= src_stride * h;
dst -= dst_stride * h;
msk -= msk_stride * h;
}
return calc_masked_variance(v_sum_d, v_sse_q, sse, w, h);
}
unsigned int vp9_masked_subpel_varWxH_xzero_yother(
unsigned int vp9_masked_subpel_varWxH_xzero(
const uint8_t *src, int src_stride, int yoffset,
const uint8_t *dst, int dst_stride, const uint8_t *msk, int msk_stride,
unsigned int *sse, int w, int h) {
unsigned int *sse, int w, int h, filter_fn_t filter_fn) {
int i, j;
__m128i v_src0_b, v_src1_b, v_res_b, v_dst_b, v_msk_b;
__m128i v_sum_d = _mm_setzero_si128();
@ -497,7 +465,7 @@ unsigned int vp9_masked_subpel_varWxH_xzero_yother(
for (i = 0; i < h; i += 2) {
// Load the next row apply the filter
v_src1_b = _mm_loadu_si128((const __m128i*)(src + j + src_stride));
v_res_b = apply_filter(v_src0_b, v_src1_b, v_filter_b);
v_res_b = filter_fn(v_src0_b, v_src1_b, v_filter_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + j));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + j));
@ -505,7 +473,7 @@ unsigned int vp9_masked_subpel_varWxH_xzero_yother(
// Load the next row apply the filter
v_src0_b = _mm_loadu_si128((const __m128i*)(src + j + src_stride * 2));
v_res_b = apply_filter(v_src1_b, v_src0_b, v_filter_b);
v_res_b = filter_fn(v_src1_b, v_src0_b, v_filter_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + j + dst_stride));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + j + msk_stride));
@ -522,138 +490,10 @@ unsigned int vp9_masked_subpel_varWxH_xzero_yother(
}
return calc_masked_variance(v_sum_d, v_sse_q, sse, w, h);
}
unsigned int vp9_masked_subpel_varWxH_xhalf_yzero(
const uint8_t *src, int src_stride,
const uint8_t *dst, int dst_stride, const uint8_t *msk, int msk_stride,
unsigned int *sse, int w, int h) {
int i, j;
__m128i v_src0_b, v_src1_b, v_res_b, v_dst_b, v_msk_b;
__m128i v_sum_d = _mm_setzero_si128();
__m128i v_sse_q = _mm_setzero_si128();
for (i = 0; i < h; i++) {
for (j = 0; j < w; j += 16) {
// Load this row & apply the filter to them
v_src0_b = _mm_loadu_si128((const __m128i*)(src + j));
v_src1_b = _mm_loadu_si128((const __m128i*)(src + j + 1));
v_res_b = _mm_avg_epu8(v_src0_b, v_src1_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + j));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + j));
sum_and_sse(v_res_b, v_dst_b, v_msk_b, &v_sum_d, &v_sse_q);
}
src += src_stride;
dst += dst_stride;
msk += msk_stride;
}
return calc_masked_variance(v_sum_d, v_sse_q, sse, w, h);
}
unsigned int vp9_masked_subpel_varWxH_xhalf_yhalf(
const uint8_t *src, int src_stride,
const uint8_t *dst, int dst_stride, const uint8_t *msk, int msk_stride,
unsigned int *sse, int w, int h) {
int i, j;
__m128i v_src0_b, v_src1_b, v_src2_b, v_src3_b, v_avg0_b, v_avg1_b;
__m128i v_res_b, v_dst_b, v_msk_b;
__m128i v_sum_d = _mm_setzero_si128();
__m128i v_sse_q = _mm_setzero_si128();
for (j = 0; j < w; j += 16) {
// Load the first row ready
v_src0_b = _mm_loadu_si128((const __m128i*)(src + j));
v_src1_b = _mm_loadu_si128((const __m128i*)(src + j + 1));
v_avg0_b = _mm_avg_epu8(v_src0_b, v_src1_b);
// Process 2 rows at a time
for (i = 0; i < h; i += 2) {
// Load the next row & apply the filter
v_src2_b = _mm_loadu_si128((const __m128i*)(src + src_stride + j));
v_src3_b = _mm_loadu_si128((const __m128i*)(src + src_stride + j + 1));
v_avg1_b = _mm_avg_epu8(v_src2_b, v_src3_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + j));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + j));
// Complete the calculation for this row and add it to the running total
v_res_b = _mm_avg_epu8(v_avg0_b, v_avg1_b);
sum_and_sse(v_res_b, v_dst_b, v_msk_b, &v_sum_d, &v_sse_q);
// Load the next row & apply the filter
v_src0_b = _mm_loadu_si128((const __m128i*)(src + src_stride * 2 + j));
v_src1_b = _mm_loadu_si128((const __m128i*)(src + src_stride * 2 +
j + 1));
v_avg0_b = _mm_avg_epu8(v_src0_b, v_src1_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + dst_stride + j));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + msk_stride + j));
// Complete the calculation for this row and add it to the running total
v_res_b = _mm_avg_epu8(v_avg1_b, v_avg0_b);
sum_and_sse(v_res_b, v_dst_b, v_msk_b, &v_sum_d, &v_sse_q);
// Move onto the next block of rows
src += src_stride * 2;
dst += dst_stride * 2;
msk += msk_stride * 2;
}
// Reset to the top of the block
src -= src_stride * h;
dst -= dst_stride * h;
msk -= msk_stride * h;
}
return calc_masked_variance(v_sum_d, v_sse_q, sse, w, h);
}
unsigned int vp9_masked_subpel_varWxH_xhalf_yother(
const uint8_t *src, int src_stride, int yoffset,
const uint8_t *dst, int dst_stride, const uint8_t *msk, int msk_stride,
unsigned int *sse, int w, int h) {
int i, j;
__m128i v_src0_b, v_src1_b, v_src2_b, v_src3_b, v_avg0_b, v_avg1_b;
__m128i v_res_b, v_dst_b, v_msk_b;
__m128i v_sum_d = _mm_setzero_si128();
__m128i v_sse_q = _mm_setzero_si128();
const __m128i v_filter_b = _mm_set1_epi16((
BILINEAR_FILTERS_2TAP(yoffset)[1] << 8) +
BILINEAR_FILTERS_2TAP(yoffset)[0]);
for (j = 0; j < w; j += 16) {
// Load the first row ready
v_src0_b = _mm_loadu_si128((const __m128i*)(src + j));
v_src1_b = _mm_loadu_si128((const __m128i*)(src + j + 1));
v_avg0_b = _mm_avg_epu8(v_src0_b, v_src1_b);
// Process 2 rows at a time
for (i = 0; i < h; i += 2) {
// Load the next row & apply the filter
v_src2_b = _mm_loadu_si128((const __m128i*)(src + src_stride + j));
v_src3_b = _mm_loadu_si128((const __m128i*)(src + src_stride + j + 1));
v_avg1_b = _mm_avg_epu8(v_src2_b, v_src3_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + j));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + j));
// Complete the calculation for this row and add it to the running total
v_res_b = apply_filter(v_avg0_b, v_avg1_b, v_filter_b);
sum_and_sse(v_res_b, v_dst_b, v_msk_b, &v_sum_d, &v_sse_q);
// Load the next row & apply the filter
v_src0_b = _mm_loadu_si128((const __m128i*)(src + src_stride * 2 + j));
v_src1_b = _mm_loadu_si128((const __m128i*)(src + src_stride * 2 +
j + 1));
v_avg0_b = _mm_avg_epu8(v_src0_b, v_src1_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + dst_stride + j));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + msk_stride + j));
// Complete the calculation for this row and add it to the running total
v_res_b = apply_filter(v_avg1_b, v_avg0_b, v_filter_b);
sum_and_sse(v_res_b, v_dst_b, v_msk_b, &v_sum_d, &v_sse_q);
// Move onto the next block of rows
src += src_stride * 2;
dst += dst_stride * 2;
msk += msk_stride * 2;
}
// Reset to the top of the block
src -= src_stride * h;
dst -= dst_stride * h;
msk -= msk_stride * h;
}
return calc_masked_variance(v_sum_d, v_sse_q, sse, w, h);
}
unsigned int vp9_masked_subpel_varWxH_xother_yzero(
unsigned int vp9_masked_subpel_varWxH_yzero(
const uint8_t *src, int src_stride, int xoffset,
const uint8_t *dst, int dst_stride, const uint8_t *msk, int msk_stride,
unsigned int *sse, int w, int h) {
unsigned int *sse, int w, int h, filter_fn_t filter_fn) {
int i, j;
__m128i v_src0_b, v_src1_b, v_res_b, v_dst_b, v_msk_b;
__m128i v_sum_d = _mm_setzero_si128();
@ -666,7 +506,7 @@ unsigned int vp9_masked_subpel_varWxH_xother_yzero(
// Load this row and one below & apply the filter to them
v_src0_b = _mm_loadu_si128((const __m128i*)(src + j));
v_src1_b = _mm_loadu_si128((const __m128i*)(src + j + 1));
v_res_b = apply_filter(v_src0_b, v_src1_b, v_filter_b);
v_res_b = filter_fn(v_src0_b, v_src1_b, v_filter_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + j));
@ -679,63 +519,11 @@ unsigned int vp9_masked_subpel_varWxH_xother_yzero(
}
return calc_masked_variance(v_sum_d, v_sse_q, sse, w, h);
}
unsigned int vp9_masked_subpel_varWxH_xother_yhalf(
const uint8_t *src, int src_stride, int xoffset,
const uint8_t *dst, int dst_stride, const uint8_t *msk, int msk_stride,
unsigned int *sse, int w, int h) {
int i, j;
__m128i v_src0_b, v_src1_b, v_src2_b, v_src3_b;
__m128i v_filtered0_b, v_filtered1_b, v_res_b, v_dst_b, v_msk_b;
__m128i v_sum_d = _mm_setzero_si128();
__m128i v_sse_q = _mm_setzero_si128();
const __m128i v_filter_b = _mm_set1_epi16((
BILINEAR_FILTERS_2TAP(xoffset)[1] << 8) +
BILINEAR_FILTERS_2TAP(xoffset)[0]);
for (j = 0; j < w; j += 16) {
// Load the first row ready
v_src0_b = _mm_loadu_si128((const __m128i*)(src + j));
v_src1_b = _mm_loadu_si128((const __m128i*)(src + j + 1));
v_filtered0_b = apply_filter(v_src0_b, v_src1_b, v_filter_b);
// Process 2 rows at a time
for (i = 0; i < h; i += 2) {
// Load the next row & apply the filter
v_src2_b = _mm_loadu_si128((const __m128i*)(src + src_stride + j));
v_src3_b = _mm_loadu_si128((const __m128i*)(src + src_stride + j + 1));
v_filtered1_b = apply_filter(v_src2_b, v_src3_b, v_filter_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + j));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + j));
// Complete the calculation for this row and add it to the running total
v_res_b = _mm_avg_epu8(v_filtered0_b, v_filtered1_b);
sum_and_sse(v_res_b, v_dst_b, v_msk_b, &v_sum_d, &v_sse_q);
// Load the next row & apply the filter
v_src0_b = _mm_loadu_si128((const __m128i*)(src + src_stride * 2 + j));
v_src1_b = _mm_loadu_si128((const __m128i*)(src + src_stride * 2 +
j + 1));
v_filtered0_b = apply_filter(v_src0_b, v_src1_b, v_filter_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + dst_stride + j));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + msk_stride + j));
// Complete the calculation for this row and add it to the running total
v_res_b = _mm_avg_epu8(v_filtered1_b, v_filtered0_b);
sum_and_sse(v_res_b, v_dst_b, v_msk_b, &v_sum_d, &v_sse_q);
// Move onto the next block of rows
src += src_stride * 2;
dst += dst_stride * 2;
msk += msk_stride * 2;
}
// Reset to the top of the block
src -= src_stride * h;
dst -= dst_stride * h;
msk -= msk_stride * h;
}
return calc_masked_variance(v_sum_d, v_sse_q, sse, w, h);
}
unsigned int vp9_masked_subpel_varWxH_xother_yother(
unsigned int vp9_masked_subpel_varWxH_xnonzero_ynonzero(
const uint8_t *src, int src_stride, int xoffset, int yoffset,
const uint8_t *dst, int dst_stride, const uint8_t *msk, int msk_stride,
unsigned int *sse, int w, int h) {
unsigned int *sse, int w, int h, filter_fn_t xfilter_fn,
filter_fn_t yfilter_fn) {
int i, j;
__m128i v_src0_b, v_src1_b, v_src2_b, v_src3_b;
__m128i v_filtered0_b, v_filtered1_b, v_res_b, v_dst_b, v_msk_b;
@ -751,30 +539,30 @@ unsigned int vp9_masked_subpel_varWxH_xother_yother(
// Load the first row ready
v_src0_b = _mm_loadu_si128((const __m128i*)(src + j));
v_src1_b = _mm_loadu_si128((const __m128i*)(src + j + 1));
v_filtered0_b = apply_filter(v_src0_b, v_src1_b, v_filterx_b);
v_filtered0_b = xfilter_fn(v_src0_b, v_src1_b, v_filterx_b);
// Process 2 rows at a time
for (i = 0; i < h; i += 2) {
// Load the next row & apply the filter
v_src2_b = _mm_loadu_si128((const __m128i*)(src + src_stride + j));
v_src3_b = _mm_loadu_si128((const __m128i*)(src + src_stride + j + 1));
v_filtered1_b = apply_filter(v_src2_b, v_src3_b, v_filterx_b);
v_filtered1_b = xfilter_fn(v_src2_b, v_src3_b, v_filterx_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + j));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + j));
// Complete the calculation for this row and add it to the running total
v_res_b = apply_filter(v_filtered0_b, v_filtered1_b, v_filtery_b);
v_res_b = yfilter_fn(v_filtered0_b, v_filtered1_b, v_filtery_b);
sum_and_sse(v_res_b, v_dst_b, v_msk_b, &v_sum_d, &v_sse_q);
// Load the next row & apply the filter
v_src0_b = _mm_loadu_si128((const __m128i*)(src + src_stride * 2 + j));
v_src1_b = _mm_loadu_si128((const __m128i*)(src + src_stride * 2 +
j + 1));
v_filtered0_b = apply_filter(v_src0_b, v_src1_b, v_filterx_b);
v_filtered0_b = xfilter_fn(v_src0_b, v_src1_b, v_filterx_b);
// Load the dst and msk for the variance calculation
v_dst_b = _mm_loadu_si128((const __m128i*)(dst + dst_stride + j));
v_msk_b = _mm_loadu_si128((const __m128i*)(msk + msk_stride + j));
// Complete the calculation for this row and add it to the running total
v_res_b = apply_filter(v_filtered1_b, v_filtered0_b, v_filtery_b);
v_res_b = yfilter_fn(v_filtered1_b, v_filtered0_b, v_filtery_b);
sum_and_sse(v_res_b, v_dst_b, v_msk_b, &v_sum_d, &v_sse_q);
// Move onto the next block of rows
src += src_stride * 2;
@ -1246,48 +1034,40 @@ unsigned int vp9_masked_sub_pixel_variance##W##x##H##_ssse3( \
dst, dst_stride, \
msk, msk_stride, sse); \
else if (yoffset == 8) \
return vp9_masked_subpel_varWxH_xzero_yhalf(src, src_stride, \
dst, dst_stride, \
msk, msk_stride, \
sse, W, H); \
return vp9_masked_subpel_varWxH_xzero(src, src_stride, 8, \
dst, dst_stride, msk, msk_stride, \
sse, W, H, apply_filter8); \
else \
return vp9_masked_subpel_varWxH_xzero_yother(src, src_stride, yoffset, \
dst, dst_stride, \
msk, msk_stride, \
sse, W, H); \
return vp9_masked_subpel_varWxH_xzero(src, src_stride, yoffset, \
dst, dst_stride, msk, msk_stride, \
sse, W, H, apply_filter); \
} else if (yoffset == 0) { \
if (xoffset == 8) \
return vp9_masked_subpel_varWxH_yzero(src, src_stride, 8, \
dst, dst_stride, msk, msk_stride, \
sse, W, H, apply_filter8); \
else \
return vp9_masked_subpel_varWxH_yzero(src, src_stride, xoffset, \
dst, dst_stride, msk, msk_stride, \
sse, W, H, apply_filter); \
} else if (xoffset == 8) { \
if (yoffset == 0) \
return vp9_masked_subpel_varWxH_xhalf_yzero(src, src_stride, \
dst, dst_stride, \
msk, msk_stride, \
sse, W, H); \
else if (yoffset == 8) \
return vp9_masked_subpel_varWxH_xhalf_yhalf(src, src_stride, \
dst, dst_stride, \
msk, msk_stride, \
sse, W, H); \
if (yoffset == 8) \
return vp9_masked_subpel_varWxH_xnonzero_ynonzero(src, src_stride, \
8, 8, dst, dst_stride, msk, msk_stride, sse, W, H, \
apply_filter8, apply_filter8); \
else \
return vp9_masked_subpel_varWxH_xhalf_yother(src, src_stride, yoffset, \
dst, dst_stride, \
msk, msk_stride, \
sse, W, H); \
return vp9_masked_subpel_varWxH_xnonzero_ynonzero(src, src_stride, \
8, yoffset, dst, dst_stride, msk, msk_stride, sse, W, H, \
apply_filter8, apply_filter); \
} else { \
if (yoffset == 0) \
return vp9_masked_subpel_varWxH_xother_yzero(src, src_stride, xoffset, \
dst, dst_stride, \
msk, msk_stride, \
sse, W, H); \
else if (yoffset == 8) \
return vp9_masked_subpel_varWxH_xother_yhalf(src, src_stride, xoffset, \
dst, dst_stride, \
msk, msk_stride, \
sse, W, H); \
if (yoffset == 8) \
return vp9_masked_subpel_varWxH_xnonzero_ynonzero(src, src_stride, \
xoffset, 8, dst, dst_stride, msk, msk_stride, sse, W, H, \
apply_filter, apply_filter8); \
else \
return vp9_masked_subpel_varWxH_xother_yother(src, src_stride, \
xoffset, yoffset, \
dst, dst_stride, \
msk, msk_stride, \
sse, W, H); \
return vp9_masked_subpel_varWxH_xnonzero_ynonzero(src, src_stride, \
xoffset, yoffset, dst, dst_stride, msk, msk_stride, sse, W, H, \
apply_filter, apply_filter); \
} \
}