Ok, I will take a look today and see if I can determine why vader is
hanging in 32-bit builds.
-Nathan
On Fri, Mar 27, 2015 at 11:26:36AM -0600, Orion Poplawski wrote:
> On 03/25/2015 01:46 PM, Nathan Hjelm wrote:
> >
> > Can you please retest both make check and vader with the following patch
On 03/25/2015 01:46 PM, Nathan Hjelm wrote:
>
> Can you please retest both make check and vader with the following patch
> applied? It fixes the constraint modifiers for opal_atomic_add_32 and
> opal_atomic_sub_32 and adds a native opal_atomic_swap_32.
>
> -Nathan
It does not appear to affect th
Can you please retest both make check and vader with the following patch
applied? It fixes the constraint modifiers for opal_atomic_add_32 and
opal_atomic_sub_32 and adds a native opal_atomic_swap_32.
-Nathan
On Fri, Mar 13, 2015 at 12:20:13PM -0600, Orion Poplawski wrote:
> We currently have op
As mentioned on the ticket I cannot reproduce this on master (same
version of vader) with netcdf master, hdf 1.8.15, and gcc 4.8.2. The
test runs to completion with both ompio and romio with both xpmem and no
single copy support. This could be a romio bug as the version in 1.8.4
lags behind master
https://github.com/open-mpi/ompi/issues/473 filed.
> On Mar 13, 2015, at 4:28 PM, Orion Poplawski wrote:
>
> That does appear to make it work. So I guess the issue is in the vader btl
> somewhere. FWIW I don't see any warning compiling the vader btl code.
>
> On 03/13/2015 01:08 PM, George B
That does appear to make it work. So I guess the issue is in the vader btl
somewhere. FWIW I don't see any warning compiling the vader btl code.
On 03/13/2015 01:08 PM, George Bosilca wrote:
> Do you have the same behavior when you disable the vader BTL ? (--mca btl
> ^vader).
>
> George.
>
Do you have the same behavior when you disable the vader BTL ? (--mca btl
^vader).
George.
On Fri, Mar 13, 2015 at 2:20 PM, Orion Poplawski
wrote:
> We currently have openmpi-1.8.4-99-20150228 built in Fedora Rawhide. I'm
> now
> seeing crashes/hangs when running the netcdf test suite on i6
We currently have openmpi-1.8.4-99-20150228 built in Fedora Rawhide. I'm now
seeing crashes/hangs when running the netcdf test suite on i686. Crashes
include:
[mock1:23702] *** An error occurred in MPI_Allreduce
[mock1:23702] *** reported by process [3653173249,1]
[mock1:23702] *** on communic
