::testing::Values(
make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
&vpx_highbd_lpf_horizontal_4_c, 8, 1),
- make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
- &vpx_highbd_lpf_vertical_4_c, 8, 1),
+ make_tuple(&wrapper_nc<vpx_highbd_lpf_vertical_4_sse2>,
+ &wrapper_nc<vpx_highbd_lpf_vertical_4_c>, 8, 1),
make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
&vpx_highbd_lpf_horizontal_8_c, 8, 1),
make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
&wrapper_nc<vpx_highbd_lpf_vertical_16_c>, 8, 1),
make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
&vpx_highbd_lpf_horizontal_4_c, 10, 1),
- make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
- &vpx_highbd_lpf_vertical_4_c, 10, 1),
+ make_tuple(&wrapper_nc<vpx_highbd_lpf_vertical_4_sse2>,
+ &wrapper_nc<vpx_highbd_lpf_vertical_4_c>, 10, 1),
make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
&vpx_highbd_lpf_horizontal_8_c, 10, 1),
make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
&wrapper_nc<vpx_highbd_lpf_vertical_16_c>, 10, 1),
make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
&vpx_highbd_lpf_horizontal_4_c, 12, 1),
- make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
- &vpx_highbd_lpf_vertical_4_c, 12, 1),
+ make_tuple(&wrapper_nc<vpx_highbd_lpf_vertical_4_sse2>,
+ &wrapper_nc<vpx_highbd_lpf_vertical_4_c>, 12, 1),
make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
&vpx_highbd_lpf_horizontal_8_c, 12, 1),
make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
lfi1->hev_thr, bd);
} else if (mask_4x4_0 & 1) {
vpx_highbd_lpf_vertical_4(s, pitch, lfi0->mblim, lfi0->lim,
- lfi0->hev_thr, 1, bd);
+ lfi0->hev_thr, bd);
} else {
vpx_highbd_lpf_vertical_4(s + 8 * pitch, pitch, lfi1->mblim,
- lfi1->lim, lfi1->hev_thr, 1, bd);
+ lfi1->lim, lfi1->hev_thr, bd);
}
}
lfi1->hev_thr, bd);
} else if (mask_4x4_int_0 & 1) {
vpx_highbd_lpf_vertical_4(s + 4, pitch, lfi0->mblim, lfi0->lim,
- lfi0->hev_thr, 1, bd);
+ lfi0->hev_thr, bd);
} else {
vpx_highbd_lpf_vertical_4(s + 8 * pitch + 4, pitch, lfi1->mblim,
- lfi1->lim, lfi1->hev_thr, 1, bd);
+ lfi1->lim, lfi1->hev_thr, bd);
}
}
}
lfi->hev_thr, bd);
} else if (mask_4x4 & 1) {
vpx_highbd_lpf_vertical_4(s, pitch, lfi->mblim, lfi->lim,
- lfi->hev_thr, 1, bd);
+ lfi->hev_thr, bd);
}
}
if (mask_4x4_int & 1)
vpx_highbd_lpf_vertical_4(s + 4, pitch, lfi->mblim, lfi->lim,
- lfi->hev_thr, 1, bd);
+ lfi->hev_thr, bd);
s += 8;
lfl += 1;
mask_16x16 >>= 1;
lfi1->hev_thr, bd);
} else if (mask_4x4_0 & 1) {
vpx_highbd_lpf_vertical_4(s, pitch, lfi0->mblim, lfi0->lim,
- lfi0->hev_thr, 1, bd);
+ lfi0->hev_thr, bd);
} else {
vpx_highbd_lpf_vertical_4(s + 8 * pitch, pitch, lfi1->mblim,
- lfi1->lim, lfi1->hev_thr, 1, bd);
+ lfi1->lim, lfi1->hev_thr, bd);
}
}
lfi1->hev_thr, bd);
} else if (mask_4x4_int_0 & 1) {
vpx_highbd_lpf_vertical_4(s + 4, pitch, lfi0->mblim, lfi0->lim,
- lfi0->hev_thr, 1, bd);
+ lfi0->hev_thr, bd);
} else {
vpx_highbd_lpf_vertical_4(s + 8 * pitch + 4, pitch, lfi1->mblim,
- lfi1->lim, lfi1->hev_thr, 1, bd);
+ lfi1->lim, lfi1->hev_thr, bd);
}
}
}
lfi->hev_thr, bd);
} else if (mask_4x4 & 1) {
vpx_highbd_lpf_vertical_4(s, pitch, lfi->mblim, lfi->lim,
- lfi->hev_thr, 1, bd);
+ lfi->hev_thr, bd);
}
}
if (mask_4x4_int & 1)
vpx_highbd_lpf_vertical_4(s + 4, pitch, lfi->mblim, lfi->lim,
- lfi->hev_thr, 1, bd);
+ lfi->hev_thr, bd);
s += 8;
lfl += 1;
mask_16x16 >>= 1;
void vpx_highbd_lpf_vertical_4_c(uint16_t *s, int pitch, const uint8_t *blimit,
const uint8_t *limit, const uint8_t *thresh,
- int count, int bd) {
+ int bd) {
int i;
// loop filter designed to work using chars so that we can make maximum use
// of 8 bit simd instructions.
- for (i = 0; i < 8 * count; ++i) {
+ for (i = 0; i < 8; ++i) {
const uint16_t p3 = s[-4], p2 = s[-3], p1 = s[-2], p0 = s[-1];
const uint16_t q0 = s[0], q1 = s[1], q2 = s[2], q3 = s[3];
const int8_t mask = highbd_filter_mask(*limit, *blimit,
const uint8_t *limit1,
const uint8_t *thresh1,
int bd) {
- vpx_highbd_lpf_vertical_4_c(s, pitch, blimit0, limit0, thresh0, 1, bd);
+ vpx_highbd_lpf_vertical_4_c(s, pitch, blimit0, limit0, thresh0, bd);
vpx_highbd_lpf_vertical_4_c(s + 8 * pitch, pitch, blimit1, limit1,
- thresh1, 1, bd);
+ thresh1, bd);
}
static INLINE void highbd_filter8(int8_t mask, uint8_t thresh, uint8_t flat,
add_proto qw/void vpx_highbd_lpf_vertical_8_dual/, "uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1, int bd";
specialize qw/vpx_highbd_lpf_vertical_8_dual sse2/;
- add_proto qw/void vpx_highbd_lpf_vertical_4/, "uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count, int bd";
+ add_proto qw/void vpx_highbd_lpf_vertical_4/, "uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd";
specialize qw/vpx_highbd_lpf_vertical_4 sse2/;
add_proto qw/void vpx_highbd_lpf_vertical_4_dual/, "uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1, int bd";
const uint8_t *blimit,
const uint8_t *limit,
const uint8_t *thresh,
- int count, int bd) {
+ int bd) {
DECLARE_ALIGNED(16, uint16_t, t_dst[8 * 8]);
uint16_t *src[1];
uint16_t *dst[1];
- (void)count;
// Transpose 8x8
src[0] = s - 4;