[visit-developers] test status summary

Mark Miller miller86 at llnl.gov
Thu May 7 17:05:18 EDT 2009


Hi All and Hank,

Ok, well what you are missing is a realization of the fact that Mark was
looking at results from a manual test on a poorly configured trunk that
he ran last night shortly after the PNG commit. But, he got confused and
concluded those were the results of the actual nightly test ;)

So, here is test suite status as I see it now after several updates from
this morning aimed at fixing it...

CGNS.py - awaiting a look at streamline alg.
NASTRAN.py - 'fixed' by update 7236. will still fail subtley.
itaps.py - mine. will probably add to skips later today
nek3d.py - failed all but 2 tests. Is anyone on this?
netcdf.py - streamline problem. Anyone on this?
silo.py - fixed by reinstalling largefile.silo
expressions.py - fixed by update 7233
parcoords.py - needs rebasline. I'll do it.
queries/database.py - may need rebaseline. Need to know why though.
scalable2.py - fixed by update 7237


Mark

On Thu, 2009-05-07 at 13:12, Hank Childs wrote:
> Good info.  Thanks Gunther!
> 
> Also: looking at the tests on hoth (the legacy LLNL version of the  
> test suite), it looks like h5part isn't failing at all.  Mark, what  
> am I missing?
> 
> -Hank
> 
> On May 7, 2009, at 12:55 PM, Gunther Weber wrote:
> 
> > Hi Hank,
> >
> > actually, the regular H5Part reader which should be used for this  
> > test has not changed at all. However, the most recent version of  
> > H5Part is now in the trunk thirdparty directory.
> >
> > Gunther
> >
> >
> > Hank Childs wrote:
> >>
> >> H5Part may be due to a new version of the lib.  I can try to put  
> >> the latest version out there on hoth.  I believe Gunther did this  
> >> on davinci.
> >>
> >> -Hank
> >>
> >> On May 7, 2009, at 10:11 AM, Mark Miller wrote:
> >>
> >>> CGNS.py -- change in streamline behavior. baseline looks 'better' in
> >>> that there is a streamline that tracks right through the gap in the
> >>> wing.
> >>>
> >>> H5Part -- everything failing I am investigating.
> >>>
> >>> metadata.py -- missing 'parsec' for zUnits on "Mesh2D" in noise.silo
> >>>
> >>> nek3d -- plots everything fine, cores or something during exit
> >>>
> >>> ITAPS.py -- This is mine.
> >>>
> >>> silo.py -- think its a problem with >2Gig files. I am investigating.
> >>>
> >>> expressions.py -- Kathleen just fixed this.
> >>>
> >>> parcoords.py -- rebaseline?
> >>>
> >>> queries/database.py - rebaseline? But lets confirm we know why it
> >>> changed
> >>>
> >>> scalable2.py -- need to change baseline text due to 'png' replacing
> >>> 'tif'
> >>>
> >>>
> >>> -- 
> >>> Mark C. Miller, Lawrence Livermore National Laboratory
> >>> email: mailto:miller86 at llnl.gov
> >>> (M/T/W) (925)-423-5901 (!!LLNL BUSINESS ONLY!!)
> >>> (Th/F)  (530)-753-8511 (!!LLNL BUSINESS ONLY!!)
> >>>
> >>
> >
-- 
Mark C. Miller, Lawrence Livermore National Laboratory
email: mailto:miller86 at llnl.gov
(M/T/W) (925)-423-5901 (!!LLNL BUSINESS ONLY!!)
(Th/F)  (530)-753-8511 (!!LLNL BUSINESS ONLY!!)



More information about the visit-developers mailing list