On 2016-11-23 11:58:38 +0100, Diego Biurrun wrote:
> On Tue, Nov 22, 2016 at 08:39:26PM +0100, Janne Grunau wrote:
> > On 2016-11-22 16:43:50 +0100, Diego Biurrun wrote:
> > > On Mon, Nov 21, 2016 at 12:31:24AM +0100, Janne Grunau wrote:
> > > > --- a/configure
> > > > +++ b/configure
> > > > @@ -348,9 +348,8 @@ Developer options (useful when working on Libav
> > > > itself):
> > > > - --ignore-tests=TESTS whitespace-separated list (without "fate-"
> > > > prefix
> > > > - in the name) of tests not to run as part of
> > > > the
> > > > - fate target
> > > > + --ignore-tests=TESTS comma-separated list (without "fate-" prefix
> > > > + in the name) of tests which result is
> > > > ignored
> > >
> > > whose result
> > >
> > > Why comma-separated instead of whitespace-separated?
> >
> > to be more in line with the other lists configure consumes and I find
> > whitespace separated lists as configure parameters midly annoying
>
> Am I overlooking such a list apart from --{en|dis}able-random?
> Because that one is a fringe feature that was never used for
> anything..
yes, all the --{enable,disable}-{encoder,decoder,demuxer,...} options
> > > > --- a/tests/fate-run.sh
> > > > +++ b/tests/fate-run.sh
> > > > @@ -212,13 +213,16 @@ if test -e "$ref" || test $cmp = "oneline" ; then
> > > > esac
> > > > cmperr=$?
> > > > test $err = 0 && err=$cmperr
> > > > - test $err = 0 || cat $cmpfile
> > > > + case $report_type in
> > > > + ignore) test $err = 0 || echo "ignoring fate-${test}" && err=0
> > > > ;;
> > > > + *) test $err = 0 || cat $cmpfile ;;
> > > > + esac
> > >
> > > Do you plan to add more cases here? Otherwise a simple if-check seems
> > > simpler. Not a big deal of course...
> >
> > I had initially the idea to add a know broken report type to turn tests
> > yellow if they run successfully. I'm not sure if that's has advantages
> > though.
>
> I'm not following 100% - what exactly did you have in mind?
The main idea was to notice if a test which is known to be broken in a
certain config runs successfully. But just inverting test result is
obviously a bad idea. I can't think of anything atm which would require
different report types.
Janne
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
