Sorry, it's due to the makefile and not a problem of libav.
If I compile using one invocation of gcc, like below, the return value is wrong.
gcc avutil-test.c -lavutil -o avutil-test
However, if I compile it in the following way, the return value is correct.
gcc -c avutil-test.c -o
The way I found this bug is that when I was using gstreamer0.10-ffmpeg on
Ubuntu Precise amd64 to demux MP3 files, if you enable the plugin logs by
setting ffmepg:5, you will see the PTS in the log changed from a few seconds
to a very large value in a sudden.
Then I try to recompile the
Public bug reported:
I found a bug with the function av_rescale_q() in the latest versions on
Precise LTS Lucid LTS on amd64 processor.
Below is my test code (to be linked against libavutilXX):
uint64_t av_rescale_q_test(uint64_t a, AVRational b, AVRational c)
{
uint64_t num = b.num *
Public bug reported:
I was using the ffdemux_mp3 plugin to stream MP3 files on Lucid and
everything worked pretty fine! But after I upgraded to Precise, I can't
find ffdemux_mp3 in gstreamer0.10-ffmpeg any more.
I did a little checking and found that:
ffdemux_mp3 is in the Lucid package build
