1
1
Граф коммитов

25240 Коммитов

Автор SHA1 Сообщение Дата
Gilles Gouaillardet
80e362de52 coll/base: fix memory free in ompi_coll_base_allreduce_intra_recursivedoubling err handler
Fix CID 1362630

Fixes open-mpi/ompi@0e393195d9
2016-06-09 13:12:25 +09:00
Gilles Gouaillardet
1f651d17c1 opal/util/ethtool: fix (infamous) strncpy usage
the infamous strncpy does not NULL terminate the destination when the buffer is truncated
do it ourself !

fix CID 1362576
2016-06-09 09:54:50 +09:00
Gilles Gouaillardet
ead7efef3f coll/basic: silence CID 1362614 in mca_coll_basic_allreduce_inter() 2016-06-09 09:40:19 +09:00
Gilles Gouaillardet
ad2e1a5ae9 coll/base: silence CID 1362613 in ompi_coll_base_alltoall_intra_basic_linear() 2016-06-09 09:40:05 +09:00
Gilles Gouaillardet
80b267af1c coll/base: silence CID 1362601 in ompi_coll_base_sendrecv_zero() 2016-06-09 09:37:31 +09:00
rhc54
84e1425d32 Merge pull request #1772 from rhc54/topic/strnlen
Abstract the strnlen function for environments that do not have it (e.g., Solaris 10)
2016-06-08 12:19:17 -07:00
Ralph Castain
8fa935534b Abstract the strnlen function for environments that do not have it (e.g., Solaris 10) 2016-06-08 10:12:43 -07:00
Jeff Squyres
95ecae8688 coverity: add --enable-debug to nightly Coverity builds
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
2016-06-08 08:01:47 -07:00
Gilles Gouaillardet
0e393195d9 coll/base: fix [all]reduce with non zero lower bound datatypes
Offset temporary buffer when a non zero lower bound datatype is used.

Thanks Hristo Iliev for the report
2016-06-08 16:48:00 +09:00
Nathan Hjelm
f8957f24af Merge pull request #1768 from hjelmn/cq_fix
btl/openib: fix cq resize calculation
2016-06-07 21:34:36 -06:00
Nathan Hjelm
dd519c55b1 btl/openib: fix cq resize calculation
Before dynamic add_procs the openib_btl_size_queues was called exactly
once for non-dynamic jobs. Now the function is called on each new
connection so the calculation was wrong. Re-wrote the function to
correctly calculate the CQ size and only attempt to adjust the CQ if
the requested size has changed. This fixes a bug when using the openib
btl on psm2 hardware that is caused by the time needed to resize a
CQ. The overhead was causing udcm to timeout and fail.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2016-06-07 16:05:56 -06:00
Nathan Hjelm
97c1643216 Merge pull request #1766 from hjelmn/req_fix
ompi/request: fix loop conditional
2016-06-07 12:11:56 -06:00
Nathan Hjelm
3ddf3ccbf3 Merge pull request #1758 from hjelmn/ob1_fixes
pml/ob1: bug fixes
2016-06-07 11:18:55 -06:00
Nathan Hjelm
5a4adb866d ompi/request: fix loop conditional
This commit fixes a bug in waitany that causes the code to go past the
beginning of the request array. The loop conditional i >= 0 is invalid
since i is unsigned. Changed to loop to check (i+1) > 0.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2016-06-07 10:28:46 -06:00
rhc54
0efb1b5d1f Merge pull request #1761 from hjelmn/progress_warnings
opal/progress: fix warnings
2016-06-07 06:35:09 -07:00
Nathan Hjelm
e082ed752a opal/progress: fix warnings
This commit fixes several warning introduced by
open-mpi/ompi@fc26d9c69f .

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2016-06-06 22:18:24 -06:00
Todd Kordenbrock
9671d6af47 Merge pull request #1689 from francois-wellenreiter/remove_trig_rdv_portals4
MTL portals4 : remove the triggered rendez-vous protocol
2016-06-06 21:55:01 -05:00
Nathan Hjelm
5d0b4679ea pml/ob1: bug fixes
This commit fixes two bugs in pml/ob1:

 - Do not called MCA_PML_OB1_PROGRESS_PENDING from
   mca_pml_ob1_send_request_start_copy as this may lead to a recursive
   call to mca_pml_ob1_send_request_process_pending.

 - In mca_pml_ob1_send_request_start_rdma return the rdma frag object
   if a btl fragment can not be allocated. This fixes a leak
   identified by @abouteiller and @bosilca.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2016-06-06 17:54:55 -06:00
rhc54
c2a02ab06c Merge pull request #1756 from rhc54/topic/hangs
Fix rare hangs observed on OS-X by properly thread-shifting upcalls from the PMIx server into ORTE
2016-06-06 07:41:58 -07:00
Joshua Ladd
db70852d31 Merge pull request #1757 from alinask/topic/revert_master_mlnx_opt_rdmacm_ibaddr
Revert "mellanox/optimized: set enable_openib_rdmacm_ibaddr=yes in the mellanox/optimized file."
2016-06-06 09:55:16 -04:00
Alina Sklarevich
a2be17ec14 Revert "mellanox/optimized: set enable_openib_rdmacm_ibaddr=yes in the mellanox/optimized file."
This reverts commit 6cd7282631.
2016-06-06 11:26:11 +03:00
Gilles Gouaillardet
01591626b3 Merge pull request #1295 from ggouaillardet/poc/nag_configury
configury: add support for NAG compilers
2016-06-06 13:42:49 +09:00
Ralph Castain
dd0f843843 Fix rare hangs observed on OS-X by properly thread-shifting upcalls from the PMIx server into ORTE 2016-06-05 21:39:44 -07:00
Gilles Gouaillardet
dc5adc5a91 configury: pass -Wl,-Wl,,--enable-new-dtags when NAG compiler is used for linking
Thanks Paul Hargrove for the report
2016-06-06 11:54:25 +09:00
Gilles Gouaillardet
11b3bc962b configury: add the nagfor NAG compiler to the default Fortran compilers 2016-06-06 11:54:25 +09:00
Gilles Gouaillardet
1ce5393fa4 configury: add the -mismatch flag to NAG compiler
NAG compiler is too picky about naming convention and cannot build
OpenMPI unless the -mismatch flag is used
2016-06-06 11:54:25 +09:00
Gilles Gouaillardet
20bfc6b3d1 autogen: patch config/ltmain.sh in order to make NAG compiler pass the -pthread option to the linker 2016-06-06 11:54:24 +09:00
Gilles Gouaillardet
544a2f1631 configury: fix mpifort and oshmemfort wrapper data
NAG compiler use gcc (and not ld) as a linker, so in order to pass an option to the linker,
the flag is -Wl,-Wl,,<option> and not -Wl,<option>

Thanks Paul Hargrove for the report
2016-06-06 11:54:12 +09:00
Gilles Gouaillardet
bbed1d4a5f configury: append LIBS to OMPI_WRAPPER_EXTRA_LIBS
This is required so NAG compiler can build static OpenMPI

Thanks Paul Hargrove for the report.
2016-06-06 11:53:42 +09:00
Gilles Gouaillardet
c976559877 coll/basic: fix log basic bcast
The log basic bcast was completely broken. The rank 0 gets the
hibit set to -1, so it always returned an error.
2016-06-06 11:01:51 +09:00
Gilles Gouaillardet
b707d138fe pmix114/pmix1_client: fix misc memory leaks
Fixes CID 1325146-1325149
2016-06-06 09:33:35 +09:00
Gilles Gouaillardet
99fedcb7a3 fs/base: silence a memory leak in mca_fs_base_get_fstype()
Fixes CID 1351211
2016-06-06 09:20:14 +09:00
George Bosilca
9376b0340b Fix the basic barrier.
The log basic barrier was completely broken. The rank 0 gets the
hibit set to 0, so it always returned an error.
2016-06-03 23:46:25 -04:00
Jeff Squyres
c0fc0d12e4 Merge pull request #1581 from jsquyres/pr/AUTHORS-update
RFC: AUTHORS: reformat and include all git log email addresses
2016-06-03 18:18:02 -07:00
Edgar Gabriel
bbdd6bd9e8 Merge pull request #1750 from edgargabriel/pr/get_byte_offset_fix
io/ompio: fix the get_byte_offset code
2016-06-03 13:32:06 -05:00
Edgar Gabriel
d6af5444a6 fix the get_byte_offset code 2016-06-03 11:36:53 -05:00
Josh Hursey
9f9f70ee50 Merge pull request #1746 from jjhursey/topic/op-init
ompi/op: Provide a default value for type/flags
2016-06-03 07:56:29 -05:00
rhc54
45fbfa2b4d Merge pull request #1748 from rhc54/topic/abrt2
Cleanup the forced termination a bit by restoring the delay before is…
2016-06-02 21:09:26 -07:00
Nathan Hjelm
ef12c13842 Merge pull request #1749 from hjelmn/enum_fix
btl: adjust values of new atomic flags
2016-06-02 21:39:15 -06:00
Nathan Hjelm
6169d03ea3 btl: adjust values of new atomic flags
Signed-off-by: Nathan Hjelm <hjelmn@me.com>
2016-06-02 19:21:34 -06:00
Ralph Castain
0ba9572f9f Cleanup the forced termination a bit by restoring the delay before issuing the sigkill, and eliminating the large time loss spent checking if the proc died. The latter is responsible for a large number of test timeouts in MTT
Update alps component
2016-06-02 17:48:21 -07:00
Nathan Hjelm
9f43b23725 Merge pull request #1710 from hjelmn/ugni_atomics
Additional ugni atomics
2016-06-02 18:25:49 -06:00
Nathan Hjelm
e968ddfe64 start bug fixes (#1729)
* mpi/start: fix bugs in cm and ob1 start functions

There were several problems with the implementation of start in Open
MPI:

 - There are no checks whatsoever on the state of the request(s)
   provided to MPI_Start/MPI_Start_all. It is erroneous to provide an
   active request to either of these calls. Since we are already
   looping over the provided requests there is little overhead in
   verifying that the request can be started.

 - Both ob1 and cm were always throwing away the request on the
   initial call to start and start_all with a particular
   request. Subsequent calls would see that the request was
   pml_complete and reuse it. This introduced a leak as the initial
   request was never freed. Since the only pml request that can
   be mpi complete but not pml complete is a buffered send the
   code to reallocate the request has been moved. To detect that
   a request is indeed mpi complete but not pml complete isend_init
   in both cm and ob1 now marks the new request as pml complete.

 - If a new request was needed the callbacks on the original request
   were not copied over to the new request. This can cause osc/pt2pt
   to hang as the incoming message callback is never called.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>

* osc/pt2pt: add request for gc after starting a new request

Starting a new receive may cause a recursive call into the pt2pt
frag receive function. If this happens and the prior request is
on the garbage collection list it could cause problems. This commit
moves the gc insert until after the new request has been posted.

Signed-off-by: Nathan Hjelm <hjelmn@me.com>
2016-06-02 20:22:40 -04:00
George Bosilca
e1c6b0e4a7 Some compilers are more than picky. 2016-06-03 09:04:34 +09:00
Matias Cabral
a15b648e65 Merge pull request #1747 from matcabral/master
Adding owner file for PSM2 MTL.
2016-06-02 16:32:26 -07:00
Matias A Cabral
29ab28f4f6 Adding owner.txt file for PSM2 MTL. 2016-06-02 16:26:16 -07:00
Nathan Hjelm
d9fc855955 Merge pull request #1743 from hjelmn/gcc_atomics_fix
atomic/gcc: add check for 128-bit CAS being lock-free
2016-06-02 16:55:31 -06:00
Nathan Hjelm
d86e41ea13 atomic/gcc: add check for 128-bit CAS being lock-free
Compiler implementations are free to include support for atomics that
use locks. Unfortunately lock-free and lock atomics do not mix. Older
versions of llvm on OS X use locks to provide
__atomic_compare_exchange on 128-bit values but are lock-free on
64-bit values. This screws up our lifo implementation which mixes
64-bit and 128-bit atomics on the same values to improve
performance. This commit adds a configure-time check if 128-bit
atomics are lock free. If they are not then the 128-bit __atomic CAS
is disabled and we check for the __sync version as a fallback.

Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
2016-06-02 15:59:05 -06:00
Nathan Hjelm
5aab4b2d51 Merge pull request #1662 from ggouaillardet/topic/amd64_atomic
amd64/atomic: silence warnings
2016-06-02 14:10:20 -06:00
Joshua Hursey
a776d78f2d ompi/op: Provide a default value for type/flags
* User defined ops leave the op_type unset which can confuse logic
   in a collective component that is trying to convert the op to the
   approprate local function.
2016-06-02 13:59:04 -05:00