[FFmpeg-devel] [PATCH 2/2] avcodec/ffv1enc: compactify unit table

[Michael Niedermayer]

This way the remap encode loop needs fewer iterations and it allows
some additional simplifications

Signed-off-by: Michael Niedermayer <michael at niedermayer.cc>
---
 libavcodec/ffv1enc.c | 31 ++++++++++++++++++-------------
 1 file changed, 18 insertions(+), 13 deletions(-)

diff --git a/libavcodec/ffv1enc.c b/libavcodec/ffv1enc.c
index 0c879fe5ce5..143c378c1b5 100644
--- a/libavcodec/ffv1enc.c
+++ b/libavcodec/ffv1enc.c
@@ -1310,23 +1310,28 @@ static int encode_float32_remap_segment(FFV1SliceContext *sc,
         s.run1final = 0;
     }
 
-    if (final) {
-        int64_t last_val = s.last_val;
-        for (int i = s.i; i<s.pixel_num; i++) {
-            int val = sc->unit[s.p][i].val;
-
-            if (last_val != val) {
-                s.compact_index ++;
-                last_val = val;
-            }
-            sc->bitmap[s.p][sc->unit[s.p][i].ndx] = s.compact_index;
+    int64_t last_val = s.last_val;
+    int j = 0;
+
+    for (int i = s.i; i<s.pixel_num; i++) {
+        int val = sc->unit[s.p][i].val;
+        int ndx = sc->unit[s.p][i].ndx;
+
+        if (last_val != val) {
+            s.compact_index ++;
+            last_val = val;
+            sc->unit[s.p][s.compact_index].val = val;
+            sc->unit[s.p][s.compact_index].ndx = ndx; //we dont need ndx but copy all should be simpler and faster
         }
+        if (final)
+            sc->bitmap[s.p][ndx] = s.compact_index;
     }
 
-    for (; s.i < s.pixel_num+1; s.i++) {
+    s.compact_index++;
+    for (; s.i < s.compact_index + 1; s.i++) {
         int current_mul = s.current_mul_index < 0 ? 1 : FFABS(s.mul[s.current_mul_index]);
         int64_t val;
-        if (s.i == s.pixel_num) {
+        if (s.i == s.compact_index) {
             if (s.last_val == 0xFFFFFFFF) {
                 break;
             } else {
@@ -1337,7 +1342,7 @@ static int encode_float32_remap_segment(FFV1SliceContext *sc,
         } else
             val = sc->unit[s.p][s.i].val;
 
-        if (s.last_val != val) {
+        {
             int64_t delta = val - s.last_val;
             int64_t step  = FFMAX(1, (delta + current_mul/2) / current_mul);
             av_assert2(s.last_val < val);
-- 
2.48.1