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

Merge changes Ie15e301e,Ib070c79b

Browse Source 
* changes:
  vp9_reconintra_neon: cosmetics: reindent
  vp9_reconintra_neon: cosmetics: drop unneeded returns
This commit is contained in:
James Zern 2015-05-23 17:47:51 +00:00 committed by Gerrit Code Review
commit 02fda6582c
1 changed files with 369 additions and 417 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

96
vp9/common/arm/neon/vp9_reconintra_neon.c
View File

@ -11,11 +11,8 @@
#include <stddef.h> #include <stddef.h>
#include <arm_neon.h> #include <arm_neon.h>
void vp9_v_predictor_4x4_neon( void vp9_v_predictor_4x4_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
int i; int i;
uint32x2_t d0u32 = vdup_n_u32(0); uint32x2_t d0u32 = vdup_n_u32(0);
(void)left; (void)left;
@ -23,14 +20,10 @@ void vp9_v_predictor_4x4_neon(
d0u32 = vld1_lane_u32((const uint32_t *)above, d0u32, 0); d0u32 = vld1_lane_u32((const uint32_t *)above, d0u32, 0);
for (i = 0; i < 4; i++, dst += y_stride) for (i = 0; i < 4; i++, dst += y_stride)
vst1_lane_u32((uint32_t *)dst, d0u32, 0); vst1_lane_u32((uint32_t *)dst, d0u32, 0);
return;
} }
void vp9_v_predictor_8x8_neon( void vp9_v_predictor_8x8_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
int i; int i;
uint8x8_t d0u8 = vdup_n_u8(0); uint8x8_t d0u8 = vdup_n_u8(0);
(void)left; (void)left;
@ -38,14 +31,10 @@ void vp9_v_predictor_8x8_neon(
d0u8 = vld1_u8(above); d0u8 = vld1_u8(above);
for (i = 0; i < 8; i++, dst += y_stride) for (i = 0; i < 8; i++, dst += y_stride)
vst1_u8(dst, d0u8); vst1_u8(dst, d0u8);
return;
} }
void vp9_v_predictor_16x16_neon( void vp9_v_predictor_16x16_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
int i; int i;
uint8x16_t q0u8 = vdupq_n_u8(0); uint8x16_t q0u8 = vdupq_n_u8(0);
(void)left; (void)left;
@ -53,14 +42,10 @@ void vp9_v_predictor_16x16_neon(
q0u8 = vld1q_u8(above); q0u8 = vld1q_u8(above);
for (i = 0; i < 16; i++, dst += y_stride) for (i = 0; i < 16; i++, dst += y_stride)
vst1q_u8(dst, q0u8); vst1q_u8(dst, q0u8);
return;
} }
void vp9_v_predictor_32x32_neon( void vp9_v_predictor_32x32_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
int i; int i;
uint8x16_t q0u8 = vdupq_n_u8(0); uint8x16_t q0u8 = vdupq_n_u8(0);
uint8x16_t q1u8 = vdupq_n_u8(0); uint8x16_t q1u8 = vdupq_n_u8(0);
@ -72,14 +57,10 @@ void vp9_v_predictor_32x32_neon(
vst1q_u8(dst, q0u8); vst1q_u8(dst, q0u8);
vst1q_u8(dst + 16, q1u8); vst1q_u8(dst + 16, q1u8);
} }
return;
} }
void vp9_h_predictor_4x4_neon( void vp9_h_predictor_4x4_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
uint8x8_t d0u8 = vdup_n_u8(0); uint8x8_t d0u8 = vdup_n_u8(0);
uint32x2_t d1u32 = vdup_n_u32(0); uint32x2_t d1u32 = vdup_n_u32(0);
(void)above; (void)above;
@ -97,14 +78,10 @@ void vp9_h_predictor_4x4_neon(
dst += y_stride; dst += y_stride;
d0u8 = vdup_lane_u8(vreinterpret_u8_u32(d1u32), 3); d0u8 = vdup_lane_u8(vreinterpret_u8_u32(d1u32), 3);
vst1_lane_u32((uint32_t *)dst, vreinterpret_u32_u8(d0u8), 0); vst1_lane_u32((uint32_t *)dst, vreinterpret_u32_u8(d0u8), 0);
return;
} }
void vp9_h_predictor_8x8_neon( void vp9_h_predictor_8x8_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
uint8x8_t d0u8 = vdup_n_u8(0); uint8x8_t d0u8 = vdup_n_u8(0);
uint64x1_t d1u64 = vdup_n_u64(0); uint64x1_t d1u64 = vdup_n_u64(0);
(void)above; (void)above;
@ -134,14 +111,10 @@ void vp9_h_predictor_8x8_neon(
dst += y_stride; dst += y_stride;
d0u8 = vdup_lane_u8(vreinterpret_u8_u64(d1u64), 7); d0u8 = vdup_lane_u8(vreinterpret_u8_u64(d1u64), 7);
vst1_u8(dst, d0u8); vst1_u8(dst, d0u8);
return;
} }
void vp9_h_predictor_16x16_neon( void vp9_h_predictor_16x16_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
int j; int j;
uint8x8_t d2u8 = vdup_n_u8(0); uint8x8_t d2u8 = vdup_n_u8(0);
uint8x16_t q0u8 = vdupq_n_u8(0); uint8x16_t q0u8 = vdupq_n_u8(0);
@ -176,14 +149,10 @@ void vp9_h_predictor_16x16_neon(
vst1q_u8(dst, q0u8); vst1q_u8(dst, q0u8);
dst += y_stride; dst += y_stride;
} }
return;
} }
void vp9_h_predictor_32x32_neon( void vp9_h_predictor_32x32_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
int j, k; int j, k;
uint8x8_t d2u8 = vdup_n_u8(0); uint8x8_t d2u8 = vdup_n_u8(0);
uint8x16_t q0u8 = vdupq_n_u8(0); uint8x16_t q0u8 = vdupq_n_u8(0);
@ -228,14 +197,10 @@ void vp9_h_predictor_32x32_neon(
dst += y_stride; dst += y_stride;
} }
} }
return;
} }
void vp9_tm_predictor_4x4_neon( void vp9_tm_predictor_4x4_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
int i; int i;
uint16x8_t q1u16, q3u16; uint16x8_t q1u16, q3u16;
int16x8_t q1s16; int16x8_t q1s16;
@ -252,14 +217,10 @@ void vp9_tm_predictor_4x4_neon(
d0u8 = vqmovun_s16(q1s16); d0u8 = vqmovun_s16(q1s16);
vst1_lane_u32((uint32_t *)dst, vreinterpret_u32_u8(d0u8), 0); vst1_lane_u32((uint32_t *)dst, vreinterpret_u32_u8(d0u8), 0);
} }
return;
} }
void vp9_tm_predictor_8x8_neon( void vp9_tm_predictor_8x8_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
int j; int j;
uint16x8_t q0u16, q3u16, q10u16; uint16x8_t q0u16, q3u16, q10u16;
int16x8_t q0s16; int16x8_t q0s16;
@ -298,14 +259,10 @@ void vp9_tm_predictor_8x8_neon(
vst1_u64((uint64_t *)dst, vreinterpret_u64_u8(d0u8)); vst1_u64((uint64_t *)dst, vreinterpret_u64_u8(d0u8));
dst += y_stride; dst += y_stride;
} }
return;
} }
void vp9_tm_predictor_16x16_neon( void vp9_tm_predictor_16x16_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
int j, k; int j, k;
uint16x8_t q0u16, q2u16, q3u16, q8u16, q10u16; uint16x8_t q0u16, q2u16, q3u16, q8u16, q10u16;
uint8x16_t q0u8, q1u8; uint8x16_t q0u8, q1u8;
@ -365,14 +322,10 @@ void vp9_tm_predictor_16x16_neon(
dst += y_stride; dst += y_stride;
} }
} }
return;
} }
void vp9_tm_predictor_32x32_neon( void vp9_tm_predictor_32x32_neon(uint8_t *dst, ptrdiff_t y_stride,
uint8_t *dst, const uint8_t *above, const uint8_t *left) {
ptrdiff_t y_stride,
const uint8_t *above,
const uint8_t *left) {
int j, k; int j, k;
uint16x8_t q0u16, q3u16, q8u16, q9u16, q10u16, q11u16; uint16x8_t q0u16, q3u16, q8u16, q9u16, q10u16, q11u16;
uint8x16_t q0u8, q1u8, q2u8; uint8x16_t q0u8, q1u8, q2u8;
@ -469,5 +422,4 @@ void vp9_tm_predictor_32x32_neon(
dst += y_stride; dst += y_stride;
} }
} }
return;
} }