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

VPX: remove step == 16 and filter[3] != 128 checks

Browse Source 
from FUN_CONV_1D and FUN_CONV_2D macros.  The functions
will not be called with these inputs.

Change-Id: I67ec75e4edafc0acee70190521a80ea85dfa521b
This commit is contained in:
Scott LaVarnway 2015-08-10 13:44:32 -07:00
parent 4e6b5079c6
commit a229dbc1f0
1 changed files with 87 additions and 88 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

175
vpx_dsp/x86/convolve.h
View File

@ -32,75 +32,74 @@ typedef void filter8_1dfunction (
const int16_t *filter_y, int y_step_q4, \ const int16_t *filter_y, int y_step_q4, \
int w, int h) { \ int w, int h) { \
assert(filter[3] != 128); \ assert(filter[3] != 128); \
if (step_q4 == 16 && filter[3] != 128) { \ assert(step_q4 == 16); \
if (filter[0] || filter[1] || filter[2]) { \ if (filter[0] || filter[1] || filter[2]) { \
while (w >= 16) { \ while (w >= 16) { \
vpx_filter_block1d16_##dir##8_##avg##opt(src_start, \ vpx_filter_block1d16_##dir##8_##avg##opt(src_start, \
src_stride, \ src_stride, \
dst, \ dst, \
dst_stride, \ dst_stride, \
h, \ h, \
filter); \ filter); \
src += 16; \ src += 16; \
dst += 16; \ dst += 16; \
w -= 16; \ w -= 16; \
} \ } \
while (w >= 8) { \ while (w >= 8) { \
vpx_filter_block1d8_##dir##8_##avg##opt(src_start, \ vpx_filter_block1d8_##dir##8_##avg##opt(src_start, \
src_stride, \ src_stride, \
dst, \ dst, \
dst_stride, \ dst_stride, \
h, \ h, \
filter); \ filter); \
src += 8; \ src += 8; \
dst += 8; \ dst += 8; \
w -= 8; \ w -= 8; \
} \ } \
while (w >= 4) { \ while (w >= 4) { \
vpx_filter_block1d4_##dir##8_##avg##opt(src_start, \ vpx_filter_block1d4_##dir##8_##avg##opt(src_start, \
src_stride, \ src_stride, \
dst, \ dst, \
dst_stride, \ dst_stride, \
h, \ h, \
filter); \ filter); \
src += 4; \ src += 4; \
dst += 4; \ dst += 4; \
w -= 4; \ w -= 4; \
} \ } \
} else { \ } else { \
while (w >= 16) { \ while (w >= 16) { \
vpx_filter_block1d16_##dir##2_##avg##opt(src, \ vpx_filter_block1d16_##dir##2_##avg##opt(src, \
src_stride, \ src_stride, \
dst, \ dst, \
dst_stride, \ dst_stride, \
h, \ h, \
filter); \ filter); \
src += 16; \ src += 16; \
dst += 16; \ dst += 16; \
w -= 16; \ w -= 16; \
} \ } \
while (w >= 8) { \ while (w >= 8) { \
vpx_filter_block1d8_##dir##2_##avg##opt(src, \ vpx_filter_block1d8_##dir##2_##avg##opt(src, \
src_stride, \ src_stride, \
dst, \ dst, \
dst_stride, \ dst_stride, \
h, \ h, \
filter); \ filter); \
src += 8; \ src += 8; \
dst += 8; \ dst += 8; \
w -= 8; \ w -= 8; \
} \ } \
while (w >= 4) { \ while (w >= 4) { \
vpx_filter_block1d4_##dir##2_##avg##opt(src, \ vpx_filter_block1d4_##dir##2_##avg##opt(src, \
src_stride, \ src_stride, \
dst, \ dst, \
dst_stride, \ dst_stride, \
h, \ h, \
filter); \ filter); \
src += 4; \ src += 4; \
dst += 4; \ dst += 4; \
w -= 4; \ w -= 4; \
} \
} \ } \
} \ } \
} }
@ -115,25 +114,25 @@ void vpx_convolve8_##avg##opt(const uint8_t *src, ptrdiff_t src_stride, \
assert(filter_y[3] != 128); \ assert(filter_y[3] != 128); \
assert(w <= 64); \ assert(w <= 64); \
assert(h <= 64); \ assert(h <= 64); \
if (x_step_q4 == 16 && y_step_q4 == 16) { \ assert(x_step_q4 == 16); \
if (filter_x[0] || filter_x[1] || filter_x[2] || filter_x[3] == 128 || \ assert(y_step_q4 == 16); \
filter_y[0] || filter_y[1] || filter_y[2] || filter_y[3] == 128) { \ if (filter_x[0] || filter_x[1] || filter_x[2]|| \
DECLARE_ALIGNED(16, uint8_t, fdata2[64 * 71]); \ filter_y[0] || filter_y[1] || filter_y[2]) { \
vpx_convolve8_horiz_##opt(src - 3 * src_stride, src_stride, fdata2, 64, \ DECLARE_ALIGNED(16, uint8_t, fdata2[64 * 71]); \
filter_x, x_step_q4, filter_y, y_step_q4, \ vpx_convolve8_horiz_##opt(src - 3 * src_stride, src_stride, fdata2, 64, \
w, h + 7); \ filter_x, x_step_q4, filter_y, y_step_q4, \
vpx_convolve8_##avg##vert_##opt(fdata2 + 3 * 64, 64, dst, dst_stride, \ w, h + 7); \
filter_x, x_step_q4, filter_y, \ vpx_convolve8_##avg##vert_##opt(fdata2 + 3 * 64, 64, dst, dst_stride, \
y_step_q4, w, h); \ filter_x, x_step_q4, filter_y, \
} else { \ y_step_q4, w, h); \
DECLARE_ALIGNED(16, uint8_t, fdata2[64 * 65]); \ } else { \
vpx_convolve8_horiz_##opt(src, src_stride, fdata2, 64, \ DECLARE_ALIGNED(16, uint8_t, fdata2[64 * 65]); \
filter_x, x_step_q4, filter_y, y_step_q4, \ vpx_convolve8_horiz_##opt(src, src_stride, fdata2, 64, \
w, h + 1); \ filter_x, x_step_q4, filter_y, y_step_q4, \
vpx_convolve8_##avg##vert_##opt(fdata2, 64, dst, dst_stride, \ w, h + 1); \
filter_x, x_step_q4, filter_y, \ vpx_convolve8_##avg##vert_##opt(fdata2, 64, dst, dst_stride, \
y_step_q4, w, h); \ filter_x, x_step_q4, filter_y, \
} \ y_step_q4, w, h); \
} \ } \
} }