2014-10-29, 17:09
I am building a stripped down xbmc like -
./configure --disable-mysql --disable-samba --disable-dbus --disable-sdl --disable-texturepacker --disable-ssh --disable-optical-drive --disable-joystick --disable-pulse --prefix=/usr
On an LFS setup where Intel drivers/mesa/kernel are recent gits/
While testing for vaapi deinterlacers I noticed that the file burosch1.mpg behaved strangely - debug osd showing 36 fps (should be 50) with any deint other than bob, scaling is bilinear
This does not happen with any other file I tried including burosch2.mpg and also only happens with current master and not FernetMenta master.
logs below, obvious diff is that failing master shows some pullup detected. I renamed burosch1.mpg Letters.mpg years ago - I have just confirmed md5sum matches with
https://samples.libav.org/MPEG2/interlaced/
Working log
http://xbmclogs.com/show.php?id=333764
Failing log
http://xbmclogs.com/show.php?id=333776
./configure --disable-mysql --disable-samba --disable-dbus --disable-sdl --disable-texturepacker --disable-ssh --disable-optical-drive --disable-joystick --disable-pulse --prefix=/usr
On an LFS setup where Intel drivers/mesa/kernel are recent gits/
While testing for vaapi deinterlacers I noticed that the file burosch1.mpg behaved strangely - debug osd showing 36 fps (should be 50) with any deint other than bob, scaling is bilinear
This does not happen with any other file I tried including burosch2.mpg and also only happens with current master and not FernetMenta master.
logs below, obvious diff is that failing master shows some pullup detected. I renamed burosch1.mpg Letters.mpg years ago - I have just confirmed md5sum matches with
https://samples.libav.org/MPEG2/interlaced/
Working log
http://xbmclogs.com/show.php?id=333764
Failing log
http://xbmclogs.com/show.php?id=333776