I lied -- I did get those errors. :-)
Turns out that it was the victim of several things:
- silly mistakes in CVS
- we had some half-committed changes to the MPI startup protocols
There may still be some mixed-endian issues in CVS that should get fixed
in a week or two, but CVS should now [finally] run the whole test suite
properly.
On Mon, 7 Apr 2003, Jeff Squyres wrote:
> On Sun, 6 Apr 2003, Nicholas Henke wrote:
>
> > Ok -- attached is gzipp-ed log from the whole make -k check MODES="tcp
> > lamd sysv ussyv"
>
> Hmm. I don't get these errors. I might have to login to your cluster
> and have a look. I assume that this is a CVS HEAD copy as of 6 April?
--
{+} Jeff Squyres
{+} jsquyres_at_[hidden]
{+} http://www.lam-mpi.org/
|
