[FFmpeg-devel] The state of FATE

Kostya kostya.shishkov
Sun Nov 1 12:22:34 CET 2009

On Sun, Nov 01, 2009 at 11:07:47AM +0100, Reimar D?ffinger wrote:
> On Sun, Nov 01, 2009 at 03:07:10AM +0000, Carl Eugen Hoyos wrote:
> > M?ns Rullg?rd <mans <at> mansr.com> writes:
> > 
> > > Linux, icc 10.1.022: 310/311 tests pass
> > > - Regression tests fail on mp2.
> > 
> > This was introduced while this configuration was not tested.
> > I did not yet find time to get the exact revision breaking the conformance test.
> Could be when I replaced exp with cbrtf, I'd test that one first if you get the time.
> Try using cbrt if it is the issue.
> > Note that there is still a bug in vc1 decoding for all 64 bit Intel compilers:
> > Intel issue 556938 concerning vc1_mc_4mv_chroma(). But since vc1 is not tested
> > by FATE, it does not currently matter.
> It matters because it means FATE should be changed to test VC1 :-P

It's the other way round. I won't mind changing it that way.
Also why FATE does not record execution times? Should be useful to
notice a sudden codec performance drop.

More information about the ffmpeg-devel mailing list