Merge commit 'fca3c3b61952aacc45e9ca54d86a762946c21942'

[ffmpeg] / libavcodec / lossless_videodsp.c

diff --git a/libavcodec/lossless_videodsp.c b/libavcodec/lossless_videodsp.c
index 3491621dc481a8876f8fe1a5a8db9a8bd1a6e544..1de288b24641bafe1ce3548220f65f0fdbc5cc18 100644 (file)
--- a/libavcodec/lossless_videodsp.c
+++ b/libavcodec/lossless_videodsp.c
@@ -21,55 +21,36 @@
 #include "lossless_videodsp.h"
 #include "libavcodec/mathops.h"
 
-static void add_int16_c(uint16_t *dst, const uint16_t *src, unsigned mask, int w){
-    long i;
-    unsigned long pw_lsb = (mask >> 1) * 0x0001000100010001ULL;
-    unsigned long pw_msb = pw_lsb +  0x0001000100010001ULL;
-    for (i = 0; i <= w - (int)sizeof(long)/2; i += sizeof(long)/2) {
-        long a = *(long*)(src+i);
-        long b = *(long*)(dst+i);
-        *(long*)(dst+i) = ((a&pw_lsb) + (b&pw_lsb)) ^ ((a^b)&pw_msb);
-    }
-    for(; i<w; i++)
-        dst[i] = (dst[i] + src[i]) & mask;
-}
+// 0x7f7f7f7f or 0x7f7f7f7f7f7f7f7f or whatever, depending on the cpu's native arithmetic size
+#define pb_7f (~0UL / 255 * 0x7f)
+#define pb_80 (~0UL / 255 * 0x80)
 
-static void diff_int16_c(uint16_t *dst, const uint16_t *src1, const uint16_t *src2, unsigned mask, int w){
+static void add_bytes_c(uint8_t *dst, uint8_t *src, intptr_t w)
+{
     long i;
-#if !HAVE_FAST_UNALIGNED
-    if((long)src2 & (sizeof(long)-1)){
-        for(i=0; i+3<w; i+=4){
-            dst[i+0] = (src1[i+0]-src2[i+0]) & mask;
-            dst[i+1] = (src1[i+1]-src2[i+1]) & mask;
-            dst[i+2] = (src1[i+2]-src2[i+2]) & mask;
-            dst[i+3] = (src1[i+3]-src2[i+3]) & mask;
-        }
-    }else
-#endif
-    {
-        unsigned long pw_lsb = (mask >> 1) * 0x0001000100010001ULL;
-        unsigned long pw_msb = pw_lsb +  0x0001000100010001ULL;
-
-        for (i = 0; i <= w - (int)sizeof(long)/2; i += sizeof(long)/2) {
-            long a = *(long*)(src1+i);
-            long b = *(long*)(src2+i);
-            *(long*)(dst+i) = ((a|pw_msb) - (b&pw_lsb)) ^ ((a^b^pw_msb)&pw_msb);
-        }
+
+    for (i = 0; i <= w - (int) sizeof(long); i += sizeof(long)) {
+        long a = *(long *) (src + i);
+        long b = *(long *) (dst + i);
+        *(long *) (dst + i) = ((a & pb_7f) + (b & pb_7f)) ^ ((a ^ b) & pb_80);
     }
-    for (; i<w; i++)
-        dst[i] = (src1[i] - src2[i]) & mask;
+    for (; i < w; i++)
+        dst[i + 0] += src[i + 0];
 }
 
-static void add_hfyu_median_pred_int16_c(uint16_t *dst, const uint16_t *src, const uint16_t *diff, unsigned mask, int w, int *left, int *left_top){
+static void add_median_pred_c(uint8_t *dst, const uint8_t *src1,
+                              const uint8_t *diff, intptr_t w,
+                              int *left, int *left_top)
+{
     int i;
-    uint16_t l, lt;
+    uint8_t l, lt;
 
     l  = *left;
     lt = *left_top;
 
-    for(i=0; i<w; i++){
-        l  = (mid_pred(l, src[i], (l + src[i] - lt) & mask) + diff[i]) & mask;
-        lt = src[i];
+    for (i = 0; i < w; i++) {
+        l      = mid_pred(l, src1[i], (l + src1[i] - lt) & 0xFF) + diff[i];
+        lt     = src1[i];
         dst[i] = l;
     }
 
@@ -77,52 +58,57 @@ static void add_hfyu_median_pred_int16_c(uint16_t *dst, const uint16_t *src, con
     *left_top = lt;
 }
 
-static void sub_hfyu_median_pred_int16_c(uint16_t *dst, const uint16_t *src1, const uint16_t *src2, unsigned mask, int w, int *left, int *left_top){
+static int add_left_pred_c(uint8_t *dst, const uint8_t *src, intptr_t w,
+                           int acc)
+{
     int i;
-    uint16_t l, lt;
 
-    l  = *left;
-    lt = *left_top;
+    for (i = 0; i < w - 1; i++) {
+        acc   += src[i];
+        dst[i] = acc;
+        i++;
+        acc   += src[i];
+        dst[i] = acc;
+    }
 
-    for(i=0; i<w; i++){
-        const int pred = mid_pred(l, src1[i], (l + src1[i] - lt) & mask);
-        lt = src1[i];
-        l  = src2[i];
-        dst[i] = (l - pred) & mask;
+    for (; i < w; i++) {
+        acc   += src[i];
+        dst[i] = acc;
     }
 
-    *left     = l;
-    *left_top = lt;
+    return acc;
 }
 
-static int add_hfyu_left_pred_int16_c(uint16_t *dst, const uint16_t *src, unsigned mask, int w, unsigned acc){
+static int add_left_pred_int16_c(uint16_t *dst, const uint16_t *src, unsigned mask, int w, unsigned acc){
     int i;
 
     for(i=0; i<w-1; i++){
         acc+= src[i];
-        dst[i]= acc & mask;
+        dst[i]= acc &= mask;
         i++;
         acc+= src[i];
-        dst[i]= acc & mask;
+        dst[i]= acc &= mask;
     }
 
     for(; i<w; i++){
         acc+= src[i];
-        dst[i]= acc & mask;
+        dst[i]= acc &= mask;
     }
 
     return acc;
 }
 
 
-void ff_llviddsp_init(LLVidDSPContext *c, AVCodecContext *avctx)
+void ff_llviddsp_init(LLVidDSPContext *c)
 {
-    c->add_int16 = add_int16_c;
-    c->diff_int16= diff_int16_c;
-    c->add_hfyu_left_pred_int16   = add_hfyu_left_pred_int16_c;
-    c->add_hfyu_median_pred_int16 = add_hfyu_median_pred_int16_c;
-    c->sub_hfyu_median_pred_int16 = sub_hfyu_median_pred_int16_c;
+    c->add_bytes                  = add_bytes_c;
+    c->add_median_pred            = add_median_pred_c;
+    c->add_left_pred              = add_left_pred_c;
+
+    c->add_left_pred_int16        = add_left_pred_int16_c;
 
+    if (ARCH_PPC)
+        ff_llviddsp_init_ppc(c);
     if (ARCH_X86)
-        ff_llviddsp_init_x86(c, avctx);
+        ff_llviddsp_init_x86(c);
 }