diff options
Diffstat (limited to 'libavcodec/sh4')
-rw-r--r-- | libavcodec/sh4/dsputil_align.c | 89 | ||||
-rw-r--r-- | libavcodec/sh4/qpel.c | 148 |
2 files changed, 119 insertions, 118 deletions
diff --git a/libavcodec/sh4/dsputil_align.c b/libavcodec/sh4/dsputil_align.c index d9286e1cea..7350b40f60 100644 --- a/libavcodec/sh4/dsputil_align.c +++ b/libavcodec/sh4/dsputil_align.c @@ -26,6 +26,7 @@ #define LP(p) *(uint32_t*)(p) +#define LPC(p) *(const uint32_t*)(p) #define UNPACK(ph,pl,tt0,tt1) do { \ @@ -51,14 +52,14 @@ #define OP_C4(ofs) \ ref-=ofs; \ do { \ - OP(LP(dest),MERGE1(LP(ref),LP(ref+4),ofs)); \ + OP(LP(dest),MERGE1(LPC(ref),LPC(ref+4),ofs)); \ ref+=stride; \ dest+=stride; \ } while(--height) #define OP_C40() \ do { \ - OP(LP(dest),LP(ref)); \ + OP(LP(dest),LPC(ref)); \ ref+=stride; \ dest+=stride; \ } while(--height) @@ -96,15 +97,15 @@ static void avg_pixels4_c(uint8_t *dest,const uint8_t *ref, const int stride,int ref-=ofs; \ do { \ uint32_t t0,t1; \ - t0 = LP(ref+0); \ - t1 = LP(ref+4); \ + t0 = LPC(ref+0); \ + t1 = LPC(ref+4); \ OP(LP(dest+0), MERGE1(t0,t1,ofs)); \ - t0 = LP(ref+8); \ + t0 = LPC(ref+8); \ OP(LP(dest+4), MERGE1(t1,t0,ofs)); \ if (sz==16) { \ - t1 = LP(ref+12); \ + t1 = LPC(ref+12); \ OP(LP(dest+8), MERGE1(t0,t1,ofs)); \ - t0 = LP(ref+16); \ + t0 = LPC(ref+16); \ OP(LP(dest+12), MERGE1(t1,t0,ofs)); \ } \ ref+=stride; \ @@ -116,11 +117,11 @@ if (sz==16) { \ #define OP_C0(sz,avg2) \ { \ do { \ - OP(LP(dest+0), LP(ref+0)); \ - OP(LP(dest+4), LP(ref+4)); \ + OP(LP(dest+0), LPC(ref+0)); \ + OP(LP(dest+4), LPC(ref+4)); \ if (sz==16) { \ - OP(LP(dest+8), LP(ref+8)); \ - OP(LP(dest+12), LP(ref+12)); \ + OP(LP(dest+8), LPC(ref+8)); \ + OP(LP(dest+12), LPC(ref+12)); \ } \ ref+=stride; \ dest+= stride; \ @@ -132,15 +133,15 @@ if (sz==16) { \ ref-=ofs; \ do { \ uint32_t t0,t1; \ - t0 = LP(ref+0); \ - t1 = LP(ref+4); \ + t0 = LPC(ref+0); \ + t1 = LPC(ref+4); \ OP(LP(dest+0), avg2(MERGE1(t0,t1,ofs),MERGE2(t0,t1,ofs))); \ - t0 = LP(ref+8); \ + t0 = LPC(ref+8); \ OP(LP(dest+4), avg2(MERGE1(t1,t0,ofs),MERGE2(t1,t0,ofs))); \ if (sz==16) { \ - t1 = LP(ref+12); \ + t1 = LPC(ref+12); \ OP(LP(dest+8), avg2(MERGE1(t0,t1,ofs),MERGE2(t0,t1,ofs))); \ - t0 = LP(ref+16); \ + t0 = LPC(ref+16); \ OP(LP(dest+12), avg2(MERGE1(t1,t0,ofs),MERGE2(t1,t0,ofs))); \ } \ ref+=stride; \ @@ -153,23 +154,23 @@ if (sz==16) { \ { \ uint32_t t0,t1,t2,t3,t; \ \ - t0 = LP(ref+0); \ - t1 = LP(ref+4); \ + t0 = LPC(ref+0); \ + t1 = LPC(ref+4); \ if (sz==16) { \ - t2 = LP(ref+8); \ - t3 = LP(ref+12); \ + t2 = LPC(ref+8); \ + t3 = LPC(ref+12); \ } \ do { \ ref += stride; \ \ - t = LP(ref+0); \ + t = LPC(ref+0); \ OP(LP(dest+0), avg2(t0,t)); t0 = t; \ - t = LP(ref+4); \ + t = LPC(ref+4); \ OP(LP(dest+4), avg2(t1,t)); t1 = t; \ if (sz==16) { \ - t = LP(ref+8); \ + t = LPC(ref+8); \ OP(LP(dest+8), avg2(t2,t)); t2 = t; \ - t = LP(ref+12); \ + t = LPC(ref+12); \ OP(LP(dest+12), avg2(t3,t)); t3 = t; \ } \ dest+= stride; \ @@ -181,32 +182,32 @@ if (sz==16) { \ uint32_t t0,t1,t2,t3,t,w0,w1; \ \ ref-=ofs; \ - w0 = LP(ref+0); \ - w1 = LP(ref+4); \ + w0 = LPC(ref+0); \ + w1 = LPC(ref+4); \ t0 = MERGE1(w0,w1,ofs); \ - w0 = LP(ref+8); \ + w0 = LPC(ref+8); \ t1 = MERGE1(w1,w0,ofs); \ if (sz==16) { \ - w1 = LP(ref+12); \ + w1 = LPC(ref+12); \ t2 = MERGE1(w0,w1,ofs); \ - w0 = LP(ref+16); \ + w0 = LPC(ref+16); \ t3 = MERGE1(w1,w0,ofs); \ } \ do { \ ref += stride; \ \ - w0 = LP(ref+0); \ - w1 = LP(ref+4); \ + w0 = LPC(ref+0); \ + w1 = LPC(ref+4); \ t = MERGE1(w0,w1,ofs); \ OP(LP(dest+0), avg2(t0,t)); t0 = t; \ - w0 = LP(ref+8); \ + w0 = LPC(ref+8); \ t = MERGE1(w1,w0,ofs); \ OP(LP(dest+4), avg2(t1,t)); t1 = t; \ if (sz==16) { \ - w1 = LP(ref+12); \ + w1 = LPC(ref+12); \ t = MERGE1(w0,w1,ofs); \ OP(LP(dest+8), avg2(t2,t)); t2 = t; \ - w0 = LP(ref+16); \ + w0 = LPC(ref+16); \ t = MERGE1(w1,w0,ofs); \ OP(LP(dest+12), avg2(t3,t)); t3 = t; \ } \ @@ -222,34 +223,34 @@ if (sz==16) { \ uint32_t a0,a1,a2,a3,a4,a5,a6,a7; \ \ ref -= ofs; \ - w0 = LP(ref+0); \ - w1 = LP(ref+4); \ + w0 = LPC(ref+0); \ + w1 = LPC(ref+4); \ UNPACK(a0,a1,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ - w0 = LP(ref+8); \ + w0 = LPC(ref+8); \ UNPACK(a2,a3,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ if (sz==16) { \ - w1 = LP(ref+12); \ + w1 = LPC(ref+12); \ UNPACK(a4,a5,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ - w0 = LP(ref+16); \ + w0 = LPC(ref+16); \ UNPACK(a6,a7,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ } \ do { \ ref+=stride; \ - w0 = LP(ref+0); \ - w1 = LP(ref+4); \ + w0 = LPC(ref+0); \ + w1 = LPC(ref+4); \ UNPACK(t2,t3,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ OP(LP(dest+0),PACK(a0,a1,t2,t3)); \ a0 = t2; a1 = t3; \ - w0 = LP(ref+8); \ + w0 = LPC(ref+8); \ UNPACK(t2,t3,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ OP(LP(dest+4),PACK(a2,a3,t2,t3)); \ a2 = t2; a3 = t3; \ if (sz==16) { \ - w1 = LP(ref+12); \ + w1 = LPC(ref+12); \ UNPACK(t2,t3,MERGE1(w0,w1,ofs),MERGE2(w0,w1,ofs)); \ OP(LP(dest+8),PACK(a4,a5,t2,t3)); \ a4 = t2; a5 = t3; \ - w0 = LP(ref+16); \ + w0 = LPC(ref+16); \ UNPACK(t2,t3,MERGE1(w1,w0,ofs),MERGE2(w1,w0,ofs)); \ OP(LP(dest+12),PACK(a6,a7,t2,t3)); \ a6 = t2; a7 = t3; \ diff --git a/libavcodec/sh4/qpel.c b/libavcodec/sh4/qpel.c index 2069bd38c1..917068d0c6 100644 --- a/libavcodec/sh4/qpel.c +++ b/libavcodec/sh4/qpel.c @@ -26,7 +26,7 @@ static inline void OPNAME ## _pixels4_l2_aligned(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ do {\ - OP(LP(dst ),rnd_avg32(LP(src1 ),LP(src2 )) ); \ + OP(LP(dst ),rnd_avg32(LPC(src1 ),LPC(src2 )) ); \ src1+=src_stride1; \ src2+=src_stride2; \ dst+=dst_stride; \ @@ -36,7 +36,7 @@ static inline void OPNAME ## _pixels4_l2_aligned(uint8_t *dst, const uint8_t *sr static inline void OPNAME ## _pixels4_l2_aligned2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ do {\ - OP(LP(dst ),rnd_avg32(AV_RN32(src1 ),LP(src2 )) ); \ + OP(LP(dst ),rnd_avg32(AV_RN32(src1 ),LPC(src2 )) ); \ src1+=src_stride1; \ src2+=src_stride2; \ dst+=dst_stride; \ @@ -46,10 +46,10 @@ static inline void OPNAME ## _pixels4_l2_aligned2(uint8_t *dst, const uint8_t *s static inline void OPNAME ## _no_rnd_pixels16_l2_aligned2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ do {\ - OP(LP(dst ),no_rnd_avg32(AV_RN32(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),no_rnd_avg32(AV_RN32(src1+4),LP(src2+4)) ); \ - OP(LP(dst+8),no_rnd_avg32(AV_RN32(src1+8),LP(src2+8)) ); \ - OP(LP(dst+12),no_rnd_avg32(AV_RN32(src1+12),LP(src2+12)) ); \ + OP(LP(dst ),no_rnd_avg32(AV_RN32(src1 ),LPC(src2 )) ); \ + OP(LP(dst+4),no_rnd_avg32(AV_RN32(src1+4),LPC(src2+4)) ); \ + OP(LP(dst+8),no_rnd_avg32(AV_RN32(src1+8),LPC(src2+8)) ); \ + OP(LP(dst+12),no_rnd_avg32(AV_RN32(src1+12),LPC(src2+12)) ); \ src1+=src_stride1; \ src2+=src_stride2; \ dst+=dst_stride; \ @@ -59,10 +59,10 @@ static inline void OPNAME ## _no_rnd_pixels16_l2_aligned2(uint8_t *dst, const ui static inline void OPNAME ## _pixels16_l2_aligned2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ do {\ - OP(LP(dst ),rnd_avg32(AV_RN32(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),rnd_avg32(AV_RN32(src1+4),LP(src2+4)) ); \ - OP(LP(dst+8),rnd_avg32(AV_RN32(src1+8),LP(src2+8)) ); \ - OP(LP(dst+12),rnd_avg32(AV_RN32(src1+12),LP(src2+12)) ); \ + OP(LP(dst ),rnd_avg32(AV_RN32(src1 ),LPC(src2 )) ); \ + OP(LP(dst+4),rnd_avg32(AV_RN32(src1+4),LPC(src2+4)) ); \ + OP(LP(dst+8),rnd_avg32(AV_RN32(src1+8),LPC(src2+8)) ); \ + OP(LP(dst+12),rnd_avg32(AV_RN32(src1+12),LPC(src2+12)) ); \ src1+=src_stride1; \ src2+=src_stride2; \ dst+=dst_stride; \ @@ -72,8 +72,8 @@ static inline void OPNAME ## _pixels16_l2_aligned2(uint8_t *dst, const uint8_t * static inline void OPNAME ## _no_rnd_pixels8_l2_aligned2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ do { /* onlye src2 aligned */\ - OP(LP(dst ),no_rnd_avg32(AV_RN32(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),no_rnd_avg32(AV_RN32(src1+4),LP(src2+4)) ); \ + OP(LP(dst ),no_rnd_avg32(AV_RN32(src1 ),LPC(src2 )) ); \ + OP(LP(dst+4),no_rnd_avg32(AV_RN32(src1+4),LPC(src2+4)) ); \ src1+=src_stride1; \ src2+=src_stride2; \ dst+=dst_stride; \ @@ -83,8 +83,8 @@ static inline void OPNAME ## _no_rnd_pixels8_l2_aligned2(uint8_t *dst, const uin static inline void OPNAME ## _pixels8_l2_aligned2(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ do {\ - OP(LP(dst ),rnd_avg32(AV_RN32(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),rnd_avg32(AV_RN32(src1+4),LP(src2+4)) ); \ + OP(LP(dst ),rnd_avg32(AV_RN32(src1 ),LPC(src2 )) ); \ + OP(LP(dst+4),rnd_avg32(AV_RN32(src1+4),LPC(src2+4)) ); \ src1+=src_stride1; \ src2+=src_stride2; \ dst+=dst_stride; \ @@ -94,8 +94,8 @@ static inline void OPNAME ## _pixels8_l2_aligned2(uint8_t *dst, const uint8_t *s static inline void OPNAME ## _no_rnd_pixels8_l2_aligned(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ do {\ - OP(LP(dst ),no_rnd_avg32(LP(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),no_rnd_avg32(LP(src1+4),LP(src2+4)) ); \ + OP(LP(dst ),no_rnd_avg32(LPC(src1 ),LPC(src2 )) ); \ + OP(LP(dst+4),no_rnd_avg32(LPC(src1+4),LPC(src2+4)) ); \ src1+=src_stride1; \ src2+=src_stride2; \ dst+=dst_stride; \ @@ -105,8 +105,8 @@ static inline void OPNAME ## _no_rnd_pixels8_l2_aligned(uint8_t *dst, const uint static inline void OPNAME ## _pixels8_l2_aligned(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ do {\ - OP(LP(dst ),rnd_avg32(LP(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),rnd_avg32(LP(src1+4),LP(src2+4)) ); \ + OP(LP(dst ),rnd_avg32(LPC(src1 ),LPC(src2 )) ); \ + OP(LP(dst+4),rnd_avg32(LPC(src1+4),LPC(src2+4)) ); \ src1+=src_stride1; \ src2+=src_stride2; \ dst+=dst_stride; \ @@ -116,10 +116,10 @@ static inline void OPNAME ## _pixels8_l2_aligned(uint8_t *dst, const uint8_t *sr static inline void OPNAME ## _no_rnd_pixels16_l2_aligned(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ do {\ - OP(LP(dst ),no_rnd_avg32(LP(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),no_rnd_avg32(LP(src1+4),LP(src2+4)) ); \ - OP(LP(dst+8),no_rnd_avg32(LP(src1+8),LP(src2+8)) ); \ - OP(LP(dst+12),no_rnd_avg32(LP(src1+12),LP(src2+12)) ); \ + OP(LP(dst ),no_rnd_avg32(LPC(src1 ),LPC(src2 )) ); \ + OP(LP(dst+4),no_rnd_avg32(LPC(src1+4),LPC(src2+4)) ); \ + OP(LP(dst+8),no_rnd_avg32(LPC(src1+8),LPC(src2+8)) ); \ + OP(LP(dst+12),no_rnd_avg32(LPC(src1+12),LPC(src2+12)) ); \ src1+=src_stride1; \ src2+=src_stride2; \ dst+=dst_stride; \ @@ -129,10 +129,10 @@ static inline void OPNAME ## _no_rnd_pixels16_l2_aligned(uint8_t *dst, const uin static inline void OPNAME ## _pixels16_l2_aligned(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, int src_stride1, int src_stride2, int h) \ {\ do {\ - OP(LP(dst ),rnd_avg32(LP(src1 ),LP(src2 )) ); \ - OP(LP(dst+4),rnd_avg32(LP(src1+4),LP(src2+4)) ); \ - OP(LP(dst+8),rnd_avg32(LP(src1+8),LP(src2+8)) ); \ - OP(LP(dst+12),rnd_avg32(LP(src1+12),LP(src2+12)) ); \ + OP(LP(dst ),rnd_avg32(LPC(src1 ),LPC(src2 )) ); \ + OP(LP(dst+4),rnd_avg32(LPC(src1+4),LPC(src2+4)) ); \ + OP(LP(dst+8),rnd_avg32(LPC(src1+8),LPC(src2+8)) ); \ + OP(LP(dst+12),rnd_avg32(LPC(src1+12),LPC(src2+12)) ); \ src1+=src_stride1; \ src2+=src_stride2; \ dst+=dst_stride; \ @@ -154,11 +154,11 @@ static inline void OPNAME ## _pixels8_l2_aligned1(uint8_t *dst, const uint8_t *s static inline void OPNAME ## _pixels8_l4_aligned(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ do { \ uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,LP(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ + UNPACK(a0,a1,LPC(src1),LPC(src2)); \ + UNPACK(a2,a3,LPC(src3),LPC(src4)); \ OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + UNPACK(a0,a1,LPC(src1+4),LPC(src2+4)); \ + UNPACK(a2,a3,LPC(src3+4),LPC(src4+4)); \ OP(LP(dst+4),rnd_PACK(a0,a1,a2,a3)); \ src1+=src_stride1;\ src2+=src_stride2;\ @@ -171,11 +171,11 @@ static inline void OPNAME ## _pixels8_l4_aligned(uint8_t *dst, const uint8_t *sr static inline void OPNAME ## _no_rnd_pixels8_l4_aligned(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ do { \ uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,LP(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ + UNPACK(a0,a1,LPC(src1),LPC(src2)); \ + UNPACK(a2,a3,LPC(src3),LPC(src4)); \ OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + UNPACK(a0,a1,LPC(src1+4),LPC(src2+4)); \ + UNPACK(a2,a3,LPC(src3+4),LPC(src4+4)); \ OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ src1+=src_stride1;\ src2+=src_stride2;\ @@ -188,11 +188,11 @@ static inline void OPNAME ## _no_rnd_pixels8_l4_aligned(uint8_t *dst, const uint static inline void OPNAME ## _pixels8_l4_aligned0(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ do { \ uint32_t a0,a1,a2,a3; /* src1 only not aligned */\ - UNPACK(a0,a1,AV_RN32(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ + UNPACK(a0,a1,AV_RN32(src1),LPC(src2)); \ + UNPACK(a2,a3,LPC(src3),LPC(src4)); \ OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,AV_RN32(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + UNPACK(a0,a1,AV_RN32(src1+4),LPC(src2+4)); \ + UNPACK(a2,a3,LPC(src3+4),LPC(src4+4)); \ OP(LP(dst+4),rnd_PACK(a0,a1,a2,a3)); \ src1+=src_stride1;\ src2+=src_stride2;\ @@ -205,11 +205,11 @@ static inline void OPNAME ## _pixels8_l4_aligned0(uint8_t *dst, const uint8_t *s static inline void OPNAME ## _no_rnd_pixels8_l4_aligned0(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ do { \ uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,AV_RN32(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ + UNPACK(a0,a1,AV_RN32(src1),LPC(src2)); \ + UNPACK(a2,a3,LPC(src3),LPC(src4)); \ OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,AV_RN32(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + UNPACK(a0,a1,AV_RN32(src1+4),LPC(src2+4)); \ + UNPACK(a2,a3,LPC(src3+4),LPC(src4+4)); \ OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ src1+=src_stride1;\ src2+=src_stride2;\ @@ -222,17 +222,17 @@ static inline void OPNAME ## _no_rnd_pixels8_l4_aligned0(uint8_t *dst, const uin static inline void OPNAME ## _pixels16_l4_aligned(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ do { \ uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,LP(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ + UNPACK(a0,a1,LPC(src1),LPC(src2)); \ + UNPACK(a2,a3,LPC(src3),LPC(src4)); \ OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + UNPACK(a0,a1,LPC(src1+4),LPC(src2+4)); \ + UNPACK(a2,a3,LPC(src3+4),LPC(src4+4)); \ OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+8),LP(src2+8)); \ - UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ + UNPACK(a0,a1,LPC(src1+8),LPC(src2+8)); \ + UNPACK(a2,a3,LPC(src3+8),LPC(src4+8)); \ OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+12),LP(src2+12)); \ - UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ + UNPACK(a0,a1,LPC(src1+12),LPC(src2+12)); \ + UNPACK(a2,a3,LPC(src3+12),LPC(src4+12)); \ OP(LP(dst+12),rnd_PACK(a0,a1,a2,a3)); \ src1+=src_stride1;\ src2+=src_stride2;\ @@ -245,17 +245,17 @@ static inline void OPNAME ## _pixels16_l4_aligned(uint8_t *dst, const uint8_t *s static inline void OPNAME ## _no_rnd_pixels16_l4_aligned(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ do { \ uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,LP(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ + UNPACK(a0,a1,LPC(src1),LPC(src2)); \ + UNPACK(a2,a3,LPC(src3),LPC(src4)); \ OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + UNPACK(a0,a1,LPC(src1+4),LPC(src2+4)); \ + UNPACK(a2,a3,LPC(src3+4),LPC(src4+4)); \ OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+8),LP(src2+8)); \ - UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ + UNPACK(a0,a1,LPC(src1+8),LPC(src2+8)); \ + UNPACK(a2,a3,LPC(src3+8),LPC(src4+8)); \ OP(LP(dst+8),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,LP(src1+12),LP(src2+12)); \ - UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ + UNPACK(a0,a1,LPC(src1+12),LPC(src2+12)); \ + UNPACK(a2,a3,LPC(src3+12),LPC(src4+12)); \ OP(LP(dst+12),no_rnd_PACK(a0,a1,a2,a3)); \ src1+=src_stride1;\ src2+=src_stride2;\ @@ -268,17 +268,17 @@ static inline void OPNAME ## _no_rnd_pixels16_l4_aligned(uint8_t *dst, const uin static inline void OPNAME ## _pixels16_l4_aligned0(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ do { /* src1 is unaligned */\ uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,AV_RN32(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ + UNPACK(a0,a1,AV_RN32(src1),LPC(src2)); \ + UNPACK(a2,a3,LPC(src3),LPC(src4)); \ OP(LP(dst),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,AV_RN32(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + UNPACK(a0,a1,AV_RN32(src1+4),LPC(src2+4)); \ + UNPACK(a2,a3,LPC(src3+4),LPC(src4+4)); \ OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,AV_RN32(src1+8),LP(src2+8)); \ - UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ + UNPACK(a0,a1,AV_RN32(src1+8),LPC(src2+8)); \ + UNPACK(a2,a3,LPC(src3+8),LPC(src4+8)); \ OP(LP(dst+8),rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,AV_RN32(src1+12),LP(src2+12)); \ - UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ + UNPACK(a0,a1,AV_RN32(src1+12),LPC(src2+12)); \ + UNPACK(a2,a3,LPC(src3+12),LPC(src4+12)); \ OP(LP(dst+12),rnd_PACK(a0,a1,a2,a3)); \ src1+=src_stride1;\ src2+=src_stride2;\ @@ -291,17 +291,17 @@ static inline void OPNAME ## _pixels16_l4_aligned0(uint8_t *dst, const uint8_t * static inline void OPNAME ## _no_rnd_pixels16_l4_aligned0(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\ do { \ uint32_t a0,a1,a2,a3; \ - UNPACK(a0,a1,AV_RN32(src1),LP(src2)); \ - UNPACK(a2,a3,LP(src3),LP(src4)); \ + UNPACK(a0,a1,AV_RN32(src1),LPC(src2)); \ + UNPACK(a2,a3,LPC(src3),LPC(src4)); \ OP(LP(dst),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,AV_RN32(src1+4),LP(src2+4)); \ - UNPACK(a2,a3,LP(src3+4),LP(src4+4)); \ + UNPACK(a0,a1,AV_RN32(src1+4),LPC(src2+4)); \ + UNPACK(a2,a3,LPC(src3+4),LPC(src4+4)); \ OP(LP(dst+4),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,AV_RN32(src1+8),LP(src2+8)); \ - UNPACK(a2,a3,LP(src3+8),LP(src4+8)); \ + UNPACK(a0,a1,AV_RN32(src1+8),LPC(src2+8)); \ + UNPACK(a2,a3,LPC(src3+8),LPC(src4+8)); \ OP(LP(dst+8),no_rnd_PACK(a0,a1,a2,a3)); \ - UNPACK(a0,a1,AV_RN32(src1+12),LP(src2+12)); \ - UNPACK(a2,a3,LP(src3+12),LP(src4+12)); \ + UNPACK(a0,a1,AV_RN32(src1+12),LPC(src2+12)); \ + UNPACK(a2,a3,LPC(src3+12),LPC(src4+12)); \ OP(LP(dst+12),no_rnd_PACK(a0,a1,a2,a3)); \ src1+=src_stride1;\ src2+=src_stride2;\ |