[FFmpeg-cvslog] ffv1enc: consider 2pass float rounding, fix loop

[Michael Niedermayer]

ffmpeg | branch: master | Michael Niedermayer <michaelni at gmx.at> | Fri Nov 16 16:55:48 2012 +0100| [c2cbc80ae9e3f00e21235915a80f81a2aeb7db0a] | committer: Michael Niedermayer

ffv1enc: consider 2pass float rounding, fix loop

Found-by: "Peter B." <pb at das-werkstatt.com>
Signed-off-by: Michael Niedermayer <michaelni at gmx.at>

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=c2cbc80ae9e3f00e21235915a80f81a2aeb7db0a
---

 libavcodec/ffv1enc.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/libavcodec/ffv1enc.c b/libavcodec/ffv1enc.c
index 41ace3f..b3879fc 100644
--- a/libavcodec/ffv1enc.c
+++ b/libavcodec/ffv1enc.c
@@ -604,7 +604,7 @@ static int sort_stt(FFV1Context *s, uint8_t stt[256])
 
                 double size0 = COST2(i,  i) + COST2(i2, i2);
                 double sizeX = COST2(i, i2) + COST2(i2, i);
-                if (sizeX < size0 && i != 128 && i2 != 128) {
+                if (size0 - sizeX > size0*(1e-14) && i != 128 && i2 != 128) {
                     int j;
                     FFSWAP(int, stt[i], stt[i2]);
                     FFSWAP(int, s->rc_stat[i][0], s->rc_stat[i2][0]);