TrellisNode *t = node_buf + frontier*(i&1);
TrellisNode **u;
int sample = samples[i*stride];
+ int heap_pos = 0;
memset(nodes_next, 0, frontier*sizeof(TrellisNode*));
for(j=0; j<frontier && nodes[j]; j++) {
// higher j have higher ssd already, so they're unlikely to use a suboptimal next sample too
#define STORE_NODE(NAME, STEP_INDEX)\
int d;\
uint32_t ssd;\
+ int pos;\
+ TrellisNode *u;\
dec_sample = av_clip_int16(dec_sample);\
d = sample - dec_sample;\
ssd = nodes[j]->ssd + d*d;\
- if(nodes_next[frontier-1] && ssd >= nodes_next[frontier-1]->ssd)\
- continue;\
/* Collapse any two states with the same previous sample value. \
* One could also distinguish states by step and by 2nd to last
* sample, but the effects of that are negligible. */\
goto next_##NAME;\
}\
}\
- for(k=0; k<frontier; k++) {\
- if(!nodes_next[k] || ssd < nodes_next[k]->ssd) {\
- TrellisNode *u = nodes_next[frontier-1];\
+ if (heap_pos < frontier) {\
+ pos = heap_pos++;\
+ } else {\
+ /* Find the largest node in the heap, which is one \
+ * of the leaf nodes. */\
+ int maxpos = 0;\
+ uint32_t max_ssd = 0;\
+ for (k = frontier >> 1; k < frontier; k++) {\
+ if (nodes_next[k]->ssd > max_ssd) {\
+ maxpos = k;\
+ max_ssd = nodes_next[k]->ssd;\
+ }\
+ }\
+ pos = maxpos;\
+ if (ssd > max_ssd)\
+ goto next_##NAME;\
+ }\
+ u = nodes_next[pos];\
if(!u) {\
assert(pathn < FREEZE_INTERVAL<<avctx->trellis);\
u = t++;\
+ nodes_next[pos] = u;\
u->path = pathn++;\
}\
u->ssd = ssd;\
u->sample1 = dec_sample;\
paths[u->path].nibble = nibble;\
paths[u->path].prev = nodes[j]->path;\
- memmove(&nodes_next[k+1], &nodes_next[k], (frontier-k-1)*sizeof(TrellisNode*));\
- nodes_next[k] = u;\
+ /* Sift the newly inserted node down in the heap to \
+ * restore the heap property. */\
+ while (pos > 0) {\
+ int parent = (pos - 1) >> 1;\
+ if (nodes_next[parent]->ssd <= ssd)\
break;\
- }\
+ FFSWAP(TrellisNode*, nodes_next[parent], nodes_next[pos]);\
+ pos = parent;\
}\
next_##NAME:;
STORE_NODE(ms, FFMAX(16, (AdaptationTable[nibble] * step) >> 8));