Thanks for finding this, George!
On Aug 10, 2005, at 12:37 AM, George Bosilca wrote:
I run all the ex-Pallas test and the same error happens. We try to
malloc
0 bytes and we hang somewhere. Let me explain what I found. First of
all,
most of the tests seems to work perfectly (at least with t
Hi,
I got the latest code drop of 6791 today morning.
I have removed .ompi_ignore and .ompi_unignore files from
ompi/mca/mpool/mvapi directory. If I don't remove and build, the MPI
program fails with signal 11. After removing those hidden files from
that directory and building, signal 11 error d
George pointed out to me last night that there's a new version of
Valgrind that works on AMD (see http://www.valgrind.org/):
-
3.0.0 is a major overhaul of Valgrind. The most significant user
visible change is that Valgrind now supports architectures other than
x86. The new architectures
Hello Tim,
On Tue, Aug 09, 2005 at 10:22:34AM -0600, Timothy B. Prins wrote:
> If you have anyother ideas of how to do it please let us know.
>
>
I have to confess I don't like current pindown cache implementation much or
perhaps I don't understand it enough.
What I managed to understand from
On Wed, Aug 10, 2005 at 06:36:19AM -0400, Jeff Squyres wrote:
> George pointed out to me last night that there's a new version of
> Valgrind that works on AMD (see http://www.valgrind.org/):
>
> -
> 3.0.0 is a major overhaul of Valgrind. The most significant user
> visible change is that Va
On Aug 10, 2005, at 6:48 AM, Gleb Natapov wrote:
If you give an environment variable name to --log-file-qualifier, it
will use the value of that environment variable in the output file
name. Do we have a suitable environment variable which is globally
available for such a purpose? (in LAM, we
Hello Sridhar,
Can you try running the mpi-ping program again with:
export OMPI_MCA_oob_tcp_debug=10
orterun -np 2 ./mpi-ping
I'm thinking there may be a problem setting up an OOB connection
between the backend/frontend node.
Tim
Sridhar Chirravuri wrote:
Hi,
I got the latest code drop of
There is a variable for the entire name, but it follows the prefix
rule. I'd have to look in the code to give you the exact reference
(I'm not at a computer that has the code on it at the moment). It was
originally defined so that launchers could pass the proc name on to
the child process, but
That sounds perfect -- thanks!
On Aug 10, 2005, at 8:49 AM, Ralph H. Castain wrote:
There is a variable for the entire name, but it follows the prefix
rule. I'd have to look in the code to give you the exact reference
(I'm not at a computer that has the code on it at the moment). It was
origin
I just commit the collective fix. Now I'm able to run the ex-pallas
test on 2, 4 and 8 nodes without any errors. However, I'm unable to
test the IB BTL as I don't have access to any machine having such
hardware.
george.
PS: if one still notice the malloc with 0 bytes length, please sent
In Aug 10, 2005, at 5:28 AM, Sridhar Chirravuri wrote:
I got the latest code drop of 6791 today morning.
I have removed .ompi_ignore and .ompi_unignore files from
ompi/mca/mpool/mvapi directory. If I don't remove and build, the MPI
program fails with signal 11. After removing those hidden files
11 matches
Mail list logo
