[petsc-dev] plans for testing examples?
Matthew Knepley
knepley at gmail.com
Mon Sep 17 10:02:21 CDT 2012
On Mon, Sep 17, 2012 at 10:01 AM, Jed Brown <jedbrown at mcs.anl.gov> wrote:
> On Mon, Sep 17, 2012 at 9:48 AM, Matthew Knepley <knepley at gmail.com>wrote:
>
>> On Mon, Sep 17, 2012 at 9:36 AM, Jed Brown <jedbrown at mcs.anl.gov> wrote:
>>
>>> On Mon, Sep 17, 2012 at 9:08 AM, Barry Smith <bsmith at mcs.anl.gov> wrote:
>>>
>>>> How do you map the loops in some of the shell scripts in the makefiles?
>>>>
>>>
>>> Not all the loop constructs are parsed now, but I can add that in an
>>> hour or so. It'll look like
>>>
>>> with Executable('ex19.c'):
>>> for mtype in 'aij baij sbaij'.split():
>>> for vecscatter in 'rsend ssend alltoall'.split():
>>> Test(id=('thename', mtype, vecscatter), args='-mat_type %s
>>> -vecscatter_%s' % (mtype,vecscatter), compare='ex19_thename')
>>>
>>
>> This cannot yet do what I need for ex62. I have a 'setup' directive that
>> executes code before the test. I like
>> this done for sets of tests to save time.
>>
>
> I would create a setup object and refer to it.
>
> with Executable('ex62.c'):
> quadrature = dict()
> for dim in 1 2 3:
> for order in 1 2 3 4:
> quadrature[(dim,order)] = GeneratedFile('ex62.h',
> command=[os.path.join(petsc_dir,'bin','pythonscripts','PetscGenerateFEMQuadrature.py'),
> ...])
> Test(..., requires=quadrature(2,3))
> # ... more tests, sharing one or more generated files
>
> Each test, or group of tests, will normally be run in its own private
> temporary directory. A sequence of tests sharing the same dependencies can
> reuse the generated file, but if you test with -j48, members of the same
> group may be executing concurrently.
>
Okay, I am fine with this. Where is the code so we can start doing it.
Matt
--
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20120917/db9614e0/attachment.html>
More information about the petsc-dev
mailing list