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

fix lossless_neon.c

Browse Source 
* some extra {xx , 0 } in initializers
* replaced by vget_lane_u32() where appropriate

Change-Id: Iabcd8ec34d7c853920491fb147a10d4472280a36
This commit is contained in:
skal 2014-04-14 14:25:52 +02:00
parent 8ff96a027a
commit 41c6efbdc5
1 changed files with 9 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

29
src/dsp/lossless_neon.c
View File

@ -154,9 +154,7 @@ static WEBP_INLINE uint32_t Average2(const uint32_t* const a,
const uint8x8_t a1 = vreinterpret_u8_u64(a0); const uint8x8_t a1 = vreinterpret_u8_u64(a0);
const uint8x8_t b1 = vreinterpret_u8_u64(b0); const uint8x8_t b1 = vreinterpret_u8_u64(b0);
const uint8x8_t avg = vhadd_u8(a1, b1); const uint8x8_t avg = vhadd_u8(a1, b1);
uint32_t ret; return vget_lane_u32(vreinterpret_u32_u8(avg), 0);
vst1_lane_u32(&ret, vreinterpret_u32_u8(avg), 0);
return ret;
} }
static WEBP_INLINE uint32_t Average3(const uint32_t* const a, static WEBP_INLINE uint32_t Average3(const uint32_t* const a,
@ -168,9 +166,7 @@ static WEBP_INLINE uint32_t Average3(const uint32_t* const a,
const uint8x8_t c1 = vreinterpret_u8_u64(c0); const uint8x8_t c1 = vreinterpret_u8_u64(c0);
const uint8x8_t avg1 = vhadd_u8(a1, c1); const uint8x8_t avg1 = vhadd_u8(a1, c1);
const uint8x8_t avg2 = vhadd_u8(avg1, b1); const uint8x8_t avg2 = vhadd_u8(avg1, b1);
uint32_t ret; return vget_lane_u32(vreinterpret_u32_u8(avg2), 0);
vst1_lane_u32(&ret, vreinterpret_u32_u8(avg2), 0);
return ret;
} }
static WEBP_INLINE uint32_t Average4(const uint32_t* const a, static WEBP_INLINE uint32_t Average4(const uint32_t* const a,
@ -185,9 +181,7 @@ static WEBP_INLINE uint32_t Average4(const uint32_t* const a,
const uint8x8_t avg1 = vhadd_u8(a1, b1); const uint8x8_t avg1 = vhadd_u8(a1, b1);
const uint8x8_t avg2 = vhadd_u8(c1, d1); const uint8x8_t avg2 = vhadd_u8(c1, d1);
const uint8x8_t avg3 = vhadd_u8(avg1, avg2); const uint8x8_t avg3 = vhadd_u8(avg1, avg2);
uint32_t ret; return vget_lane_u32(vreinterpret_u32_u8(avg3), 0);
vst1_lane_u32(&ret, vreinterpret_u32_u8(avg3), 0);
return ret;
} }
static uint32_t Predictor5(uint32_t left, const uint32_t* const top) { static uint32_t Predictor5(uint32_t left, const uint32_t* const top) {
@ -221,7 +215,7 @@ static uint32_t Predictor10(uint32_t left, const uint32_t* const top) {
static WEBP_INLINE uint32_t Select(const uint32_t* const c0, static WEBP_INLINE uint32_t Select(const uint32_t* const c0,
const uint32_t* const c1, const uint32_t* const c1,
const uint32_t* const c2) { const uint32_t* const c2) {
const uint64x1_t C0 = { *c0, 0 }, C1 = { *c1, 0 }, C2 = { *c2, 0 }; const uint64x1_t C0 = { *c0 }, C1 = { *c1 }, C2 = { *c2 };
const uint8x8_t p0 = vreinterpret_u8_u64(C0); const uint8x8_t p0 = vreinterpret_u8_u64(C0);
const uint8x8_t p1 = vreinterpret_u8_u64(C1); const uint8x8_t p1 = vreinterpret_u8_u64(C1);
const uint8x8_t p2 = vreinterpret_u8_u64(C2); const uint8x8_t p2 = vreinterpret_u8_u64(C2);
@ -230,8 +224,7 @@ static WEBP_INLINE uint32_t Select(const uint32_t* const c0,
const int16x4_t sum_bc = vreinterpret_s16_u16(vpaddl_u8(bc)); const int16x4_t sum_bc = vreinterpret_s16_u16(vpaddl_u8(bc));
const int16x4_t sum_ac = vreinterpret_s16_u16(vpaddl_u8(ac)); const int16x4_t sum_ac = vreinterpret_s16_u16(vpaddl_u8(ac));
const int32x2_t diff = vpaddl_s16(vsub_s16(sum_bc, sum_ac)); const int32x2_t diff = vpaddl_s16(vsub_s16(sum_bc, sum_ac));
int32_t pa_minus_pb; const int32_t pa_minus_pb = vget_lane_s32(diff, 0);
vst1_lane_s32(&pa_minus_pb, diff, 0);
return (pa_minus_pb <= 0) ? *c0 : *c1; return (pa_minus_pb <= 0) ? *c0 : *c1;
} }
@ -242,16 +235,14 @@ static uint32_t Predictor11(uint32_t left, const uint32_t* const top) {
static WEBP_INLINE uint32_t ClampedAddSubtractFull(const uint32_t* const c0, static WEBP_INLINE uint32_t ClampedAddSubtractFull(const uint32_t* const c0,
const uint32_t* const c1, const uint32_t* const c1,
const uint32_t* const c2) { const uint32_t* const c2) {
const uint64x1_t C0 = { *c0, 0 }, C1 = { *c1, 0 }, C2 = { *c2, 0 }; const uint64x1_t C0 = { *c0 }, C1 = { *c1 }, C2 = { *c2 };
const uint8x8_t p0 = vreinterpret_u8_u64(C0); const uint8x8_t p0 = vreinterpret_u8_u64(C0);
const uint8x8_t p1 = vreinterpret_u8_u64(C1); const uint8x8_t p1 = vreinterpret_u8_u64(C1);
const uint8x8_t p2 = vreinterpret_u8_u64(C2); const uint8x8_t p2 = vreinterpret_u8_u64(C2);
const uint16x8_t sum0 = vaddl_u8(p0, p1); // add and widen const uint16x8_t sum0 = vaddl_u8(p0, p1); // add and widen
const uint16x8_t sum1 = vqsubq_u16(sum0, vmovl_u8(p2)); // widen and subtract const uint16x8_t sum1 = vqsubq_u16(sum0, vmovl_u8(p2)); // widen and subtract
const uint8x8_t out = vqmovn_u16(sum1); // narrow and clamp const uint8x8_t out = vqmovn_u16(sum1); // narrow and clamp
uint32_t ret; return vget_lane_u32(vreinterpret_u32_u8(out), 0);
vst1_lane_u32(&ret, vreinterpret_u32_u8(out), 0);
return ret;
} }
static uint32_t Predictor12(uint32_t left, const uint32_t* const top) { static uint32_t Predictor12(uint32_t left, const uint32_t* const top) {
@ -261,7 +252,7 @@ static uint32_t Predictor12(uint32_t left, const uint32_t* const top) {
static WEBP_INLINE uint32_t ClampedAddSubtractHalf(const uint32_t* const c0, static WEBP_INLINE uint32_t ClampedAddSubtractHalf(const uint32_t* const c0,
const uint32_t* const c1, const uint32_t* const c1,
const uint32_t* const c2) { const uint32_t* const c2) {
const uint64x1_t C0 = { *c0, 0 }, C1 = { *c1, 0 }, C2 = { *c2, 0 }; const uint64x1_t C0 = { *c0 }, C1 = { *c1 }, C2 = { *c2 };
const uint8x8_t p0 = vreinterpret_u8_u64(C0); const uint8x8_t p0 = vreinterpret_u8_u64(C0);
const uint8x8_t p1 = vreinterpret_u8_u64(C1); const uint8x8_t p1 = vreinterpret_u8_u64(C1);
const uint8x8_t p2 = vreinterpret_u8_u64(C2); const uint8x8_t p2 = vreinterpret_u8_u64(C2);
@ -269,9 +260,7 @@ static WEBP_INLINE uint32_t ClampedAddSubtractHalf(const uint32_t* const c0,
const uint8x8_t ab = vshr_n_u8(vqsub_u8(avg, p2), 1); // (a-b)>>1 saturated const uint8x8_t ab = vshr_n_u8(vqsub_u8(avg, p2), 1); // (a-b)>>1 saturated
const uint8x8_t ba = vshr_n_u8(vqsub_u8(p2, avg), 1); // (b-a)>>1 saturated const uint8x8_t ba = vshr_n_u8(vqsub_u8(p2, avg), 1); // (b-a)>>1 saturated
const uint8x8_t out = vqsub_u8(vqadd_u8(avg, ab), ba); const uint8x8_t out = vqsub_u8(vqadd_u8(avg, ab), ba);
uint32_t ret; return vget_lane_u32(vreinterpret_u32_u8(out), 0);
vst1_lane_u32(&ret, vreinterpret_u32_u8(out), 0);
return ret;
} }
static uint32_t Predictor13(uint32_t left, const uint32_t* const top) { static uint32_t Predictor13(uint32_t left, const uint32_t* const top) {