-
Notifications
You must be signed in to change notification settings - Fork 68
v2.x mpi ext updates #345
v2.x mpi ext updates #345
Conversation
* affinity: will build unless disabled. * cr: will build if FT is enabled, unless it is disabled. It will also complain/abort if you --with-mpi-ext=cr, but FT is disabled. * example: will only build if --with-mpi-ext=example (and .ompi_ignore is removed) (cherry picked from commit fca193d)
(cherry picked from commit open-mpi/ompi@2fa8167)
that allow this extension to compile if configure'd with --without-hwloc (cherry picked from commit 6994d74)
No real functional changes: * Reduce #if's a little -- have a single "no hwloc" and "hwloc" section. * Make a common subroutine (no_hwloc_support()) for when we don't have any hwloc support (cherry picked from commit f6882a8)
Refer to this link for build results (access rights to CI server needed): |
Test FAILed. |
@hppritcha This looks like a false failure on the distcheck jenkins:
Can you confirm? |
@hppritcha BTW, why do all the LANL CI checks say "No test results found"? |
This is a false failure on the jenkins distcheck. The no test results found is a default from the ghprb - trying to figure out how to turn this off. There's a lot of churn in the options for the ghprb lately. |
bot:retest |
Test FAILed. |
okay try again - consequence of temporarily moving distcheck to hopper. bot:retest |
@jsquyres you need to cherry-pick open-mpi/ompi@a4cc83f as well (incomplete prototype) :bot🏷️pushed-back |
(cherry picked from commit open-mpi/ompi@a4cc83f)
@ggouaillardet Got it; thanks. |
Refer to this link for build results (access rights to CI server needed): |
@rhc54 suggested off list that the affinity extension should not be built at all without helicopter support any thought ? |
my whirlybird actually suggests that it is better to build as you have done - that way, a user's code can just call the extension without worrying that it might not be there. It just can't tell them anything useful as we cannot bind without helicopter |
+1 |
sanity check since IU jenkins ghprb plugin was upgraded to 1.24.4 and there were errors processing the ompi-release webhooks on the server. bot:retest |
I'm assuming @ggouaillardet's "+1" was meant to be a 👍 |
These are pulled straight from master. I did not include @rolfv's new cuda MPI extension; that can be a separate PR.
@ggouaillardet please review