Re: [Libmesh-devel] Possible n^2 algorithm in mesh.write() ?

[Roy S. <roy...@ic...>]

On Mon, 26 Nov 2007, John Peterson wrote:

> I do remember it. I thought we had checked it in and so I figured
> that couldn't *still* be the problem.  Oh well.

It looked like there was one problem left over - I'm not sure how it
wasn't affecting your code, but on my problems Mesh::write started
hanging in parallel when processor 0 got to MeshTools::n_levels() and
the other processors weren't also there doing their part in that
parallel-only function.

I've just checked in a fix for legacy_xdr_io.C; you might want to
check it out and make sure I haven't re-broken anything for you.
---
Roy

> > On Nov 26, 2007 11:15 AM, John Peterson <pet...@cf...> wrote:
> > > OK, so it's not an n^2 algorithm, but XdrIO::write() currently gets
> > > called on all CPUs, and so all of them try to write to the same file
> > > at the same time.  Turns out this is a lot more noticeable on 128 CPUs
> > > than it is on 4...
> > >
> > > The fix should be to wrap the call to write in if (processor_id==0)
> > > as is done for GMVIO.  I'll try to check this fix in today.