On Thu, Oct 12, 2006, Gustavo Seabra wrote:
>
> I tried removing the "-std95" flag and re-compiling, but the problem
> is still there. ALL test cases fail, and fail horribly! (see the
> attached TEST_FAILURES.DIFF).
Doesn't happen for me. Things in test.sander.BASIC.MPI work, except for
the always-problematic "targetted MD with changing target and restraints";
everything in test.sander.GB.MPI works for me, as do the tests in
test.sander.REM. Tests under test.sander.QMMM.MPI appear to fail, but it
just looks like formatting changes that are not reflected in the *save
files.
This is for ifort 9.1.033 on emt64, and openmpi-1.1.1. That's only one data
point, and Ross is apparently seeing problems too (although his email didn't
give any details.)
[I'm assuming that you've confirmed that amber9 is fine with your setup
and mpi installation. I don't remember that you have explicitly stated this
or not.]
...dac
Received on Thu Oct 12 2006 - 20:36:15 PDT
