Correctly handle the corner case in MPI_Alltoallv when
some tasks have no data to transfer and some other tasks
do have data to transfer.
This test case is covered in ibm/collective/alltoallv_somezeros
from the ompi-tests repo.
cmr=v1.8.2:reviewer=bosilca
This commit was SVN r31985.
check for possibility of heap2heap copy was incorrect
in case when shared heaps have different virtual
addresses on same host.
It seems that ibv_exp_reg_mr() on CIB cards may return
different VAs for heap on same node. On CX3 addresses are
the same.
reviewed by miked
cmr=v1.8.2:reviewer=ompi-rm1.8
This commit was SVN r31969.
* allow users to specify just a modifier for map-by instead of requiring that they also specify a policy. Thus, we now accept --map-by :pe=3 as indicating that we should use the default mapping policy, but bind 3 cpus/proc.
* if users specify a pe's/proc but no policy, default to --map-by NUMA to ensure we have access to multiple cpus for the request. This won't guarantee we have access to enough to meet the request, but gives us a chance. In addition, we know that binding a proc to multiple cpus will work best if those cpus are all in the same NUMA, so this provides some degree of optimized behavior.
Per a request from Jeff, define "oversubscribe" for binding as a synonym for the "overload" modifier.
cmr=v1.8.2:reviewer=rhc
This commit was SVN r31967.
Few components had wrong mca variables registration procedure
List of them:
- atomic basic and mxm
- spml yoda and ikrit
Two mca variables as runtime_api_verbose and runtime_lock_recursive change
names to oshmem_api_verbose and oshmem_lock_recursive otherwise they
were not shown by oshmem_info tool.
fixed by Igor, reviewed by Miked
cmr=v1.8.2:reviewer=ompi-rm1.8
This commit was SVN r31962.
This should eliminate the connectivity issues that have been reported, and will make maintenance of this component much easier.
cmr=v1.8.2:reviewer=jsquyres:subject=simplify the OOB/TCP component
This commit was SVN r31956.
is really special as the weights can be one of the following three
values (NULL, EMPTY or some legal value). As such, we need a complex
if to correctly convert the Fortran value to the corresponding C
value. Thus, always defining the c_ array is the simplest and most
straighforward approach.
cmr=v1.8.2:reviewer=jsquyres
This commit was SVN r31955.
Issue noted by Walter Spector on the user's mailing list.
Throwing to Craig Rasmussen for review.
cmr=v1.8.2:reviewer=jsquyres
This commit was SVN r31933.
This would be a really, really weird case if it ever happens (i.e.,
you have usnics but the agent process failed somewhere in MPI_INIT
such that the agent never appears), but having an infinite loop
doesn't seem like a good idea.
(does not need to go to v1.8 because v1.8 still uses RML for
communication for the connectivity checker)
This commit was SVN r31932.
This conservative fixes tries to fetch info from both
opal_dstore_nonpeer and opal_dstore_peer.
This is required is task A spawns tasks B and C.
B was previously unable to find info from C, this caused locality
info not being set and a hang in coll/ml init.
no CMR is required since v1.8 uses a unique dstore
This commit was SVN r31923.
if eager rdma is used, endpoint reference_count is greater than one.
this commit is a temporary fix that OBJ_RELEASE the endpoint as much as needed.
thought this is likely correct, it can be suboptimal and hence needs to be reviewed
cmr=v1.8.2:reviewer=hjelmn
This commit was SVN r31922.
