[FFmpeg-devel] [PATCH 2/9] fate: add 10bits YUV4:2:2 test

[Michael Niedermayer]

On Tue, Oct 13, 2015 at 04:41:50PM -0300, James Almer wrote:
> On 10/12/2015 11:04 PM, Michael Niedermayer wrote:
> > On Tue, Oct 13, 2015 at 12:00:07AM +0200, Michael Niedermayer wrote:
> >> On Mon, Oct 12, 2015 at 10:39:23PM +0200, Christophe Gisquet wrote:
> >>> Le 12 oct. 2015 21:37, "Michael Niedermayer" <michael at niedermayer.cc> a
> >>> écrit :
> >>>>
> >>>> On Mon, Oct 12, 2015 at 07:37:43PM +0200, Christophe Gisquet wrote:
> >>>>> It was useful to (accidentally?) spot an overflow in the column pass
> >>>>> of the x86 simple_idct10 implementation.
> >>>>> ---
> >>>>>  tests/fate/vcodec.mak                          | 6 +++++-
> >>>>>  tests/ref/vsynth/vsynth1-dnxhd-1080i-10bit     | 4 ++++
> >>>>>  tests/ref/vsynth/vsynth2-dnxhd-1080i-10bit     | 4 ++++
> >>>>>  tests/ref/vsynth/vsynth3-dnxhd-1080i-10bit     | 4 ++++
> >>>>>  tests/ref/vsynth/vsynth_lena-dnxhd-1080i-10bit | 4 ++++
> >>>>>  5 files changed, 21 insertions(+), 1 deletion(-)
> >>>>>  create mode 100644 tests/ref/vsynth/vsynth1-dnxhd-1080i-10bit
> >>>>>  create mode 100644 tests/ref/vsynth/vsynth2-dnxhd-1080i-10bit
> >>>>>  create mode 100644 tests/ref/vsynth/vsynth3-dnxhd-1080i-10bit
> >>>>>  create mode 100644 tests/ref/vsynth/vsynth_lena-dnxhd-1080i-10bit
> >>>>
> >>>> this alone and with patch 1 doesnt pass fate:
> >>>
> >>> I'm so bad at this codec stuff.
> >>
> >> no, you are not, this stuff is rather convoluted and the mpegvideo
> >> *dct stuff is not well documented.
> >>
> >> I think the patchset is mostly fine, ill retest things and will try to
> >> fix the remaining issues (if iam able to) and if successfull ill apply
> > 
> > 
> >> it i think this one patch here is fine as is if its applied later in
> > 
> > wasnt that easy, there was another difference, between 32 and 64bit
> > it may be float rounding in the scaler but its not dnxhd i worked
> > around it by adjusting the scaler parameters.
> 
> This test is failing on pretty much every fate client. Valgrind seems to
> complain about uninitialized values.
> http://fate.ffmpeg.org/report.cgi?time=20151013040721&slot=x86_64-archlinux-gcc-valgrindundef

this is different from the failure i saw
mine was vsynth3 and didnt show anything in valgrind on my box

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

There will always be a question for which you do not know the correct answer.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20151014/912db3a1/attachment.sig>