Revert "OPAL: drop dead with core on bad flow. rarely happens with helloworld on large scale."
This reverts commit 86f1d5af3e.
Will be reconsidered via RFC as it represents a significant change in behavior
Use experimental verbs to allocate memory at fixed base
virtual address.
verbs will disqualify itself if shared_mr is disabled
or not supported and it is impossible to allocate memory
starting at fixed base virtual address.
verbs contig pages allocator did not guarantee fixed va, now it does.
(cherry picked from commit fd77ebd452)
Apply Jeff`s comments
Update with Jeff commits
(cherry picked from commit open-mpi/ompi-release@4dc487fc3d)
of preregistered buffers
Before this change eager gets we retried on each progress loop. This commit
modifies the protocol to only retry eager gets when another eager get has
completed. This commit also cleans up some callback code that is no longer
needed.
The GNI_RDMAMODE_FENCE bit was a left over from
async progress work that is not needed at this point
in the gni BTL. Removing the bit also allows
for the removal of the GNI_CDM_MODE_BTE_SINGLE_CHANNEL
bit from the GNI_CdmCreate call.
1. It's actually hashing now, whereas the old OPAL hash table was not. Thus, it is a bug fix for and, as such, should be included in the 1.8 series.
2. It is dynamic and can grow and shrink the number of buckets in accordance with job size, whereas the old OPAL hash table had a fixed number of buckets which resulted in poor retrieval performance at large scale.
This scheme has been deployed in the field on very large H.P./Mellanox systems and has been demonstrated to significantly decrease job start-up time (~ 20% improvement) when launching applications directly with srun in SLURM environments. However, neither SLURM nor direct launch are prerequisites to take advantage of this change as any entity that utilizes OPAL hash table objects can benefit (at least partially) from this contribution.
This commit adds initial ugni thread safety support.
With this commit, sun thread tests (excepting MPI-2 RMA)
pass with various process counts and threads/process.
Also osu_latency_mt passes.
osc/rdma uses counters to determine if all messages have been received
before exiting synchronization calls. The problem is that the active
target counter is always increasing (never zeroed). If over 2^31-1
messages are sent this causes the counter to overflow (in itself this
isn't an error). This causes test/wait to return before the communication
is complete. There is an additional error in the use of the fragment
flush function. If PSCW synchronization is in use this function CAN NOT
be called unless a post message has arrived.
Relevant mailing list thread: http://www.open-mpi.org/community/lists/devel/2014/10/16016.php
This commit fixes both issues. Tested against MTT and issue reproducer.
Closes#224.
With --enable-memchecker builds, use calloc(3) for OBJ_NEW instead of
malloc(3). This cuts down on a lot of valgrind/memory checker false
positive output.
Also make a minor change in the valgrind configure.m4; have it assign
0xf to a char. The prior assignment (of 0xff) was warning about an
overflow. This didn't really matter, but we might as well make the
test not have a gratuitious warning in it.
It looks like using MAP_PRIVATE instead of MAP_SHARED greatly
speeds up infiniband memory registration.
Change-Id: Id7089f58458ef8fff4034a2c4707d31f7e8b6694
Use experimental verbs to allocate memory at fixed base
virtual address.
verbs will disqualify itself if shared_mr is disabled
or not supported and it is impossible to allocate memory
starting at fixed base virtual address.
verbs contig pages allocator did not guarantee fixed va, now it does.
Added use_hp flag in sshmem/sysv variable to control huge page usage;
Added shared_mr sshmem/verbs;
Both paraemetes are set in auto.
Fix help messages
fixed by Igor, reviewed by @miked-mellanox and @alex-mikheev
Lock server pe computation was incorrect in cases when:
lock virtual address is signed long. In this case negative pe
value was returned.
In case when lock has different virtual adresses on different pes.
It can happen when memheap or static segment have different base
addresses. Use offset instead of absolute virtual address to
compute server pe
Fixed by @alex-mikheev, reviewed by @miked-mellanox
