ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
3d32dbd793
openmpi/opal/mca/btl/openib
История
Nathan Hjelm 3d32dbd793 btl/openib: cuda: fix CUDA-aware support with async copy 
This commit should resolve an issue seen with CUDA-aware support. The
problem came in with BTL 3.0. Before 3.0 the size of the copy was
stored in the incoming segment's des_remote_count field. This field
does not exist in BTL 3.0 so I stored the value in the
des_segment_count field. This caused problems with the cuda support
code. To fix the issue the endpoint pointer is now stored in the in
fragment's endpoint pointer which free's up the segment's des_cbdata
pointer for storing the transfer size.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2015-03-10 14:38:12 -06:00
..
connect btl/openib: fix misc memory leaks 2015-03-05 14:06:18 +09:00
btl_openib_async.c cleanup dead code caused by the removal of the --with-threads configure option 2015-01-16 19:13:59 +09:00
btl_openib_async.h btl/openib: rename OPAL_HAVE_XRCD macro into OPAL_HAVE_CONNECTX_XRC_DOMAINS 2015-01-07 13:27:25 +09:00
btl_openib_atomic.c btl/openib: fix heterogeneous support 2015-03-04 13:53:05 +09:00
btl_openib_component.c btl/openib: cuda: fix CUDA-aware support with async copy 2015-03-10 14:38:12 -06:00
btl_openib_eager_rdma.h George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-) 2014-07-26 00:47:28 +00:00
btl_openib_endpoint.c Update code base to use the new opal_free_list_t 2015-02-24 10:05:45 -07:00
btl_openib_endpoint.h Update code base to use the new opal_free_list_t 2015-02-24 10:05:45 -07:00
btl_openib_failover.c btl/openib: update for BTL 3.0 interface 2015-02-13 11:46:36 -07:00
btl_openib_failover.h George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-) 2014-07-26 00:47:28 +00:00
btl_openib_fd.c cleanup dead code caused by the removal of the --with-threads configure option 2015-01-16 19:13:59 +09:00
btl_openib_fd.h George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-) 2014-07-26 00:47:28 +00:00
btl_openib_frag.c Update code base to use the new opal_free_list_t 2015-02-24 10:05:45 -07:00
btl_openib_frag.h Update code base to use the new opal_free_list_t 2015-02-24 10:05:45 -07:00
btl_openib_get.c btl/openib: fix heterogeneous support 2015-03-04 13:53:05 +09:00
btl_openib_ini.c btl/openib: cleanup duplicate code 2015-01-07 14:07:30 +09:00
btl_openib_ini.h George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-) 2014-07-26 00:47:28 +00:00
btl_openib_ip.c Cleanup some cruft resulting from the move of the btl's to opal. We had created the ability to delay modex operations, which included a need to delay retrieving hostname info for remote procs. This allowed us to not retrieve the modex info until first message unless required - the hostname is generally only required for debug and error messages. 2014-10-03 16:02:57 -06:00
btl_openib_ip.h George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-) 2014-07-26 00:47:28 +00:00
btl_openib_lex.h George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-) 2014-07-26 00:47:28 +00:00
btl_openib_lex.l George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-) 2014-07-26 00:47:28 +00:00
btl_openib_mca.c Merge pull request #441 from open-mpi/revert-438-topic/use_opal_common_verbs_want_fork_support 2015-03-04 10:13:00 +02:00
btl_openib_mca.h George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-) 2014-07-26 00:47:28 +00:00
btl_openib_proc.c Just because the openib BTL can't reach a process doesn't mean it is a job-ending error. If we have other methods for reaching the process (e.g., sm for a local proc), then that's okay. If there is no method for reaching a proc, then that's an error - but the BML will report that situation. 2014-09-10 17:02:16 +00:00
btl_openib_proc.h George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-) 2014-07-26 00:47:28 +00:00
btl_openib_put.c btl/openib: fix heterogeneous support 2015-03-04 13:53:05 +09:00
btl_openib_xrc.c btl/openib: rename OPAL_HAVE_XRCD macro into OPAL_HAVE_CONNECTX_XRC_DOMAINS 2015-01-07 13:27:25 +09:00
btl_openib_xrc.h btl/openib: add XRC support with OFED 3.12+ 2015-01-06 15:30:52 +09:00
btl_openib.c Update code base to use the new opal_free_list_t 2015-02-24 10:05:45 -07:00
btl_openib.h btl/openib: fix heterogeneous support 2015-03-04 13:53:05 +09:00
btl-openib-benchmark George did the work and deserves all the credit for it. Ralph did the merge, and deserves whatever blame results from errors in it :-) 2014-07-26 00:47:28 +00:00
configure.m4 btl/openib: add atomic operation support 2015-02-13 11:46:36 -07:00
help-mpi-btl-openib.txt Revert "create the opal_common_verbs_want_fork_support parameter." 2015-03-03 14:28:31 +02:00
Makefile.am btl/openib: add atomic operation support 2015-02-13 11:46:36 -07:00
mca-btl-openib-device-params.ini openib btl: add Soft iWARP device to the ini file 2014-11-04 14:48:43 -08:00
owner.txt add owner files to opa/ompi/orte mca directories 2015-02-22 15:10:23 -07:00