Re: [Libmesh-users] valgrind errors

[John P. <jwp...@gm...>]

On Tue, Nov 21, 2017 at 4:47 PM, Alexander Lindsay <ale...@gm...
> wrote:

> All,
>
> I get the following errors when running valgrind in parallel on a MOOSE
> test case. I could probably figure this out with time, but if someone
> smarter than I could quickly deduce whether these are false positives or
> something real then I would definitely not complain. (I've trimmed the
> output to 1 process; the errors for every other process were the same).
>

I think I would probably blame this on something in the MPI implementation?
It may be unavoidable, but it looks like the total leak is not very big:

==25211==    definitely lost: 2,928 bytes in 13 blocks

I wouldn't lose too much sleep over it...

-- 
John