[petsc-dev] VecScatter problem

[Matthew Knepley]

On Thu, Aug 25, 2016 at 5:01 PM, Abbott, Stephen R. <abbottsr at ornl.gov>
wrote:

> >> If there was an error then you should have seen PETSc output to that
> effect.  Could you have missed that?
>
>
>
> I don’t see any error show up in the output, and the run continues happily
> even with the error checks in place, so I don’t think that’s it.
>
>
>
> >> You should definitely add these error check, but that gets rid of the
> error!
>
>
>
> I agree completely, they should be there. I would be much more comfortable
> if just putting them in didn’t solve the error I put them in to look for..
>
>
>
> >> Could this be a compiler error?
>
>
>
> We’re using the PGI compiler, and I’ve tried it with versions 16.5 (with
> your PETSc builds) and 15.10 (with the cray-petsc-3.7 build) and seen the
> same problem. But, yes, this could definitely be a compiler error. I’m
> rebuilding the code with optimizations disabled to run the memory checking,
> so I’ll have a chance to see if maybe the compiler is making a mistake
> there.
>
>
>
> Does PETSc’s logging count VecScatter calls and messages passed? Maybe I
> can look at the logs and compare counts for a working run to a broken run
> and see if those last two scatters are getting optimized out.
>

Yes, it does. However, I think its more likely that the compiler plays with
those array copies.

  Matt


> --steve
>
>
>
> On 8/25/16, 5:48 PM, "Mark Adams" <mfadams at lbl.gov> wrote:
>
>
>
>
>
>
>
> The troubling bit about this, which caused me to suggest that it was a
> race condition, was that if I modify the code a little bit I can get the
> correct result from the scatter. For example, we weren’t always checking
> the returned error flag.
>
>
>
> If there was an error then you should have seen PETSc output to that
> effect.  Could you have missed that?
>
>
>
> You should definitely add these error check, but that gets rid of the
> error!  Could this be a compiler error?
>
>
>
>


-- 
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/20160825/5509ad36/attachment.html>