ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
23 048 коммитов 3 Ветки 0 Теги 130 MiB
4341639a66
Граф коммитов

3518 Коммитов

Автор SHA1 Сообщение Дата
Jeff Squyres
4341639a66 Revert "configury: fix (again) XRC detection on OFED < 3.12" 
@ggouaillardet is likely offline for the weekend, but master is broken
on RHEL 6.5 systems that do not have MOFED installed.  So I'm taking
the liberty of revering this commit; I'm guessing Gilles will fixup
and re-commit next week.

This reverts commit 77f8282d51.
 2015-07-10 06:45:33 -07:00
Gilles Gouaillardet
77f8282d51 configury: fix (again) XRC detection on OFED < 3.12 
since ibv_create_xrc_rcv_qp is now deprecated, and in order to
be "future-proof", we have to consider the case in which only XRC Domains are supported.

Thanks Paul Hargrove for the detailled report.
 2015-07-10 15:31:45 +09:00
Rolf vandeVaart
ae0f3cfee7 Make explicit call to initalize MCA parameters in common CUDA code. This allows us to view them with ompi_info and possibly modify with tools interface 2015-07-09 12:51:55 -04:00
Rolf vandeVaart
cdffa4724d Force smcuda BTL to use CUDA IPC path for all GPU buffers where possible 2015-07-08 17:11:25 -04:00
Ralph Castain
ed93154e43 Fix hetero operations. An error in the hwloc utilities only allocated memory for the first display of a binding map, and then assumed that all nodes had the same number of cores in them. This resulted in memory corruption whenever someone displayed a binding pattern for a hetero cluster, and a smaller node was first in line. 2015-07-07 12:52:16 -07:00
Gilles Gouaillardet
9f171de412 btl/openib: queue pending fragments once only when running out of credit 
Fixes open-mpi/ompi#640
 2015-07-06 09:45:01 +09:00
bosilca
77367ca02c Merge pull request #687 from rolfv/pr/fix-smcuda-perfprob 
Add the ability use different size buffers for host and CUDA buffers
 2015-07-02 18:42:41 -04:00
Jeff Squyres
4e7d979f8d Merge pull request #686 from jsquyres/pr/autogen-no-ompi-bool-fixes 
bool: use SIZEOF__BOOL, not SIZEOF_BOOL
 2015-07-02 12:19:07 -04:00
Rolf vandeVaart
30a872b478 Add the ability to send host buffers through one sized staging buffers and CUDA buffers through different sized buffers. Fixes performance issues 2015-07-02 11:11:15 -04:00
Jeff Squyres
f1353947ff libfabric: fix wrappers for static builds 
Need to set the WRAPPER_EXTRA flags so that the wrappers for static
builds pull in -lfabric.

Also update/fix some comments.
 2015-07-02 07:58:16 -07:00
Jeff Squyres
cd5751c217 bool: use SIZEOF__BOOL, not SIZEOF_BOOL 
When you "autogen.pl --no-ompi", the AC_SIZEOF(bool) test is not run.
But we *do* run AC_SIZEOF(_Bool), which is the equivalent.  So switch
the uses of SIZEOF_BOOL in the code base to be SIZEOF__BOOL, and it's
all good.
 2015-07-02 07:32:02 -07:00
Ralph Castain
861fe1d9dd This is the third time I am fixing this - I have no idea who or why this is being reset. 2015-07-02 08:39:48 -05:00
Alina Sklarevich
27797654db openib btl: added a new vendor_part_id for Mellanox ConnectX4-LX. 2015-06-29 13:50:43 +03:00
Ralph Castain
75ceec663a Now that it has been officially released, update the embedded HWLOC to 1.11.0 2015-06-28 14:07:45 -07:00
bureddy
c78b8e9b8e Merge pull request #664 from bureddy/master 
powerpc: update mem barrier instructions
 2015-06-25 14:09:49 -07:00
Jeff Squyres
a172bd161e usnic: switch to use the new libfabric common library 
The usnic BTL configure.m4 no longer needs to OPAL_CHECK_LIBFABRIC; it
just uses the results from opal/mca/common/libfabric's configure.m4.
We also now don't need to link against libfabric -- they just link
against the opal_common_libfabric library.
 2015-06-25 13:33:15 -07:00
Ralph Castain
8d128fe090 Remove the non-null attributes from the cmd_line parser as this isn't something we can guarantee, and the optimization isn't worth the potential for error 2015-06-25 13:26:20 -07:00
Ralph Castain
ea0e21bb06 Add a common/libfabric component to the opal layer where we can place common functions 2015-06-25 11:04:00 -07:00
Nathan Hjelm
ee36d813dc Merge pull request #657 from hjelmn/c99 
more c99 updates
 2015-06-25 11:21:09 -06:00
Howard Pritchard
f45914db9b Merge pull request #670 from hppritcha/topic/ownership_update 
ownership: update ownership files
 2015-06-25 11:02:45 -06:00
Nathan Hjelm
4d92c9989e more c99 updates 
This commit does two things. It removes checks for C99 required
headers (stdlib.h, string.h, signal.h, etc). Additionally it removes
definitions for required C99 types (intptr_t, int64_t, int32_t, etc).

Signed-off-by: Nathan Hjelm <hjelmn@me.com>
 2015-06-25 10:14:13 -06:00
rhc54
1a767ed47c Merge pull request #654 from rhc54/topic/config 
Remove internal bool type definitions
 2015-06-25 09:10:21 -07:00
Howard Pritchard
e49a37c034 ownership: update ownership files 
per discussions at OMPI devel workshop

Signed-off-by: Howard Pritchard <howardp@lanl.gov>
 2015-06-25 10:04:42 -06:00
Devendar Bureddy
ed406b05cb powerpc: update mem barrier instructions 
- added isync interface.
       - define opal_atomic_wmb() to lwsync as it is recommend over eieio
         on cache enabled storage.
        (http://www.ibm.com/developerworks/systems/articles/powerpc.html).
 2015-06-25 10:54:44 +03:00
Nathan Hjelm
4552afff06 Fix definition of MPI_T_pvar_get_index 
The definition of MPI_T_pvar_get_index was incorrect. This commit
fixes the definition and adds a missing return code.

Signed-off-by: Nathan Hjelm <hjelmn@me.com>
 2015-06-24 17:31:26 -06:00
Ralph Castain
869041f770 Purge whitespace from the repo 2015-06-23 20:59:57 -07:00
Ralph Castain
a809902c0a Now that we require C99, and stdbool.h is part of C99, we no longer need to define our own bool types. Since bool is commonly used in a lot of places, just include stdbool.h in opal_config_bottom.h 2015-06-23 11:31:48 -07:00
Ralph Castain
cc9b416ab3 Ensure we properly commit suicide if/when we lose connection to the daemon. There are multiple paths by which a lost daemon can be reported, and so a race condition exists in the pmix support. Our MPI layer wants the ability to determine the response to the failure, and so it will call down to the RTE with any abort request. This comes down to the pmix layer as a "pmix_abort" command, which involves communicating the request to the daemon - who is gone. Sadly, the pmix component may not know that just yet, and so we hang. 
So add a brief timer event to kick us out of the communication. The precise amount of time we should wait is somewhat TBD, but set something short for now and we can adjust.
 2015-06-18 09:45:52 -07:00
Jeff Squyres
8ab2b11f88 btl_openib.c: fix another compiler warning 
Remove this unused variable
 2015-06-17 09:00:12 -07:00
Jeff Squyres
f688289aaf btl_openib.c: fix compiler warning 
This return code is not used; tell the compiler we're not going to
use it.
 2015-06-17 08:56:56 -07:00
Jeff Squyres
097b48d521 mca_base_component_respository.c: fix compiler warning 
This function is only used in the DL case -- it can be #if'ed out if
we're not compiling with DL support to avoid a compiler warning about
defined-but-not-used.
 2015-06-17 08:54:59 -07:00
Jeff Squyres
dfa36197ea usnic/Makefile.am: ensure static builds include -lfabric 2015-06-17 08:15:29 -07:00
Gilles Gouaillardet
2cef2d0fe6 opal/memory: silence a warning 
as reported by Coverity with CID 71663
 2015-06-17 11:17:55 +09:00
Gilles Gouaillardet
58d1b3f4d0 opal_os_dirpath_create: fix TOCTOU 
as reported by Coverity with CID 70396
 2015-06-17 11:17:54 +09:00
Gilles Gouaillardet
de66447ebb opal_cmd_line_get_usage_msg: silence warning 
as reported by Coverity with CID 1269967
 2015-06-17 11:17:54 +09:00
Gilles Gouaillardet
f2f66e6e63 opal_daemon_init: silence warning 
as reported by Coverity with CID 710642
 2015-06-17 11:17:53 +09:00
Gilles Gouaillardet
8427e87ee9 opal_argv_delete: silence warning 
as reported by Coverity with CID 71914
 2015-06-17 11:17:53 +09:00
Gilles Gouaillardet
d9c490cf9f refactor opal_bitmap_get_string 
make it more efficient and fix CID 71992 (dead code)
 2015-06-17 11:17:53 +09:00
Jeff Squyres
44e7646de9 usnic/configure.m4: convert to use external libfabric 
Use the new OPAL_CHECK_LIBFABRIC macro.
 2015-06-15 15:17:06 -07:00
Jeff Squyres
3e1b85ceb3 libfabric: remove embedded libfabric 
OMPI now only builds against external libfabric installations.
 2015-06-15 15:17:05 -07:00
Jeff Squyres
c74ab51dd4 opal/mca/dl/dl.h: fix the #ifndef/#define name 
Thanks to Scott Atchley for noticing the name mismatch.
 2015-06-15 13:08:57 -07:00
rhc54
adbff46a13 Merge pull request #642 from rhc54/topic/hwloc 
Update hwloc to 1.11.0
 2015-06-13 12:09:58 -07:00
Ralph Castain
ff92781ec4 Replace hwloc191 with hwloc1110 
Fix hwloc compile. Ignore LAMA mapper due to deprecated hwloc functions
 2015-06-13 10:11:45 -07:00
Jeff Squyres
4384131e65 openib: minor style and defensive programming fixes 
Minor comment/whitespace fixes.  Also some minor logic changes that
are mainly for defensive programming purposes (i.e., ensure to always
set malloc_hook_set to true or false, and then check it before we try
to actually invoke it).
 2015-06-12 20:11:47 -07:00
Jeff Squyres
2f137ff151 openib: reset memalign threshhold properly 
Now that open-mpi/ompi#638 is fixed, reset the openib BTL memalign
threshhold properly.

This effectively re-instates commit
open-mpi/ompi@ce915b5757.
 2015-06-12 20:11:47 -07:00
Jeff Squyres
88c13adc8c openib: only set the memory hook if it is enabled 
Instead of unconditionally setting the memory hook, only set it when
the memory hooks are both available and have been enabled (e.g.,
opal/mca/memory/linux has decided that it *can* be enabled, and when
the mpi_leave_pinned MCA param is set to 1, or is set to -1 and some
component requested the memory hooks be enabled).

If we set the memory hook when memory hooks are not enabled,
__malloc_hook will be NULL, which will cause problems when
btl_openib_malloc_hook() tries to invoke it.

Fixes open-mpi/ompi#638.
 2015-06-12 20:11:47 -07:00
Ralph Castain
12d3c9ca22 Revert "Fix a typo that incorrectly set the alignment threshold in the openib BTL." 
This reverts commit ce915b5757.
 2015-06-10 14:02:49 -07:00
Gilles Gouaillardet
8885b34637 mca/base: fix a misc memory leak 
as reported by Coverity with CID 1294415
 2015-06-10 15:10:57 +09:00
Gilles Gouaillardet
9e278a21ce opal/crs: fix a string overflow 
and revamp out of resource handling
fixes resource leak as reported by Coverity with CID 1304752
 2015-06-10 14:23:25 +09:00
Nathan Hjelm
6772d32b85 opal/crs: silence clang warnings introduced by coverity fixes 
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
 2015-06-08 09:16:13 -06:00