1
1
Граф коммитов

4602 Коммитов

Автор SHA1 Сообщение Дата
Nathan Hjelm
84320f3815 btl/vader: fix compilation with SGI xpmem and add some debugging to
component_init.

cmr=v1.7.5:ticker=#4053

This commit was SVN r30535.
2014-02-03 17:42:40 +00:00
Nathan Hjelm
64321acc22 basesmuma: do not call MB directly
opal does not always define MB. It is recommended that opal_atomic_[rw]mb is
called instead. We will need to address the cases where these functions are
no-ops on weak-memory ordered cpus.

cmr=v1.7.5:ticket=trac:4158

This commit was SVN r30534.

The following Trac tickets were found above:
  Ticket 4158 --> https://svn.open-mpi.org/trac/ompi/ticket/4158
2014-02-03 17:01:57 +00:00
Nathan Hjelm
c2b061cc84 basesmuma: clean up code
Several changes are contained in this commit:

 - Clean up tabs and trailing whitespaces

 - Use consistent indentation in changed files

 - Remove unused code. None of the removed code will ever have been
   used in a trunk build.

 - Clean up the smcm code quite a bit

 - Do not fflush stderr and use opal_output instead of fprintf.

These changes have been tested on Cray XE-6 and PSM systems.

cmr=v1.7.5:ticket=trac:4158

This commit was SVN r30533.

The following Trac tickets were found above:
  Ticket 4158 --> https://svn.open-mpi.org/trac/ompi/ticket/4158
2014-02-03 17:01:46 +00:00
Christoph Niethammer
4f23d8214c Fixed incorrect calculation of reallocated memory in mca_bml_r2_del_btl.
This commit was SVN r30529.
2014-02-03 08:43:59 +00:00
Nathan Hjelm
1ae39753dc bcol/basesmuma: check the return code of bcol_basesmuma_smcm_allgather_connection.
Fixes a segmentation fault found by the bogus intercomm_create test.

cmr=v1.7.4:review=manjugv

This commit was SVN r30527.
2014-01-31 22:20:25 +00:00
Adrian Reber
7de34ea201 SNAPC/CRCP/SSTORE: remove compiler warnings
This commit was SVN r30488.
2014-01-29 20:52:00 +00:00
Adrian Reber
fa1036f38c SSTORE/CRCP: use ORTE_WAIT_FOR_COMPLETION with non-blocking receives
During the commits to make the C/R code compile again the
blocking receive calls were replaced by non-blocking
which broke the code. This patch uses ORTE_WAIT_FOR_COMPLETION()
to wait until the non-blocking calls have finished.

This commit was SVN r30486.
2014-01-29 20:30:35 +00:00
Hadi Montakhabi
7bf4c425ff Fix: making sure the file type is not overwritten by the last queried component
This commit was SVN r30478.
2014-01-29 19:21:03 +00:00
Nathan Hjelm
afae924e29 coll/ml: fix some warnings and the spelling of indices
This commit fixes one warning that should have caused coll/ml to segfault
on reduce. The fix should be correct but we will continue to investigate.

cmr=v1.7.5:ticket=trac:4158

This commit was SVN r30477.

The following Trac tickets were found above:
  Ticket 4158 --> https://svn.open-mpi.org/trac/ompi/ticket/4158
2014-01-29 18:44:21 +00:00
Nathan Hjelm
700e97cf6a btl/vader: add support for SGI's implementation of xpmem and add support
for 32-bit architectures.

This commit also modifies _OMPI_CHECK_HEADER to use AC_CHECK_HEADERS instead
of AC_CHECK_HEADER. This allows components to check for multiple headers
instead of just one. The new semantics of the header check in OMPI_CHECK_PACKAGE
are to return success if at least one of the specified headers exists. The new
semantics will not break current usage.

cmr=v1.7.5:ticket=trac:4053

This commit was SVN r30476.

The following Trac tickets were found above:
  Ticket 4053 --> https://svn.open-mpi.org/trac/ompi/ticket/4053
2014-01-29 18:35:47 +00:00
Jeff Squyres
3fa9d36aba Per http://www.open-mpi.org/community/lists/devel/2014/01/13938.php,
Orion Poplawski noticed that we should not be installing mpio.h.

cmr=v1.7.4:reviewer=hjelmn:subject=do not install mpio.h

This commit was SVN r30465.
2014-01-28 21:46:26 +00:00
George Bosilca
bde9619386 Various minor cleanups.
This commit was SVN r30431.
2014-01-26 17:27:12 +00:00
George Bosilca
d265981c55 Don't always retain the proc, do it only for new procs. This enforce a strict policy in the BML, it has one and only one ref on each proc.
This commit was SVN r30429.
2014-01-26 17:26:04 +00:00
Ralph Castain
b32556e6dc Fixes trac:4143
After IM with Nathan, apply patch from ticket after verification by Paul Hargrove that it fixes the problem on non-x86 32-bit platforms

Verified by Paul, RM-approved

cmr=v1.7.4:reviewer=ompi-gk1.7

This commit was SVN r30411.

The following Trac tickets were found above:
  Ticket 4143 --> https://svn.open-mpi.org/trac/ompi/ticket/4143
2014-01-24 17:56:52 +00:00
Nathan Hjelm
2435057a57 ignore the iboffload component for now.
This commit was SVN r30398.
2014-01-23 16:06:21 +00:00
Rolf vandeVaart
9f3bf4747d Provide option to have synchronous copy be asynchronous with a wait. For now,
this has to be selected at runtime.  Also fix up some error messages to have
node name in them.

This commit was SVN r30396.
2014-01-23 15:47:20 +00:00
Jeff Squyres
9fee7c2b4d According to a report from Adam Moody, there is a compile error with
ROMIO and Lustre 2.4.0.  It has been solved upstream already; here's
the ticket:

    http://trac.mpich.org/projects/mpich/ticket/1973

And here's the commit that fixed it:

    a0c4278f14

OMPI does not have the other code referred to in that git commit (in
ad_lustre_hints.c).

Thanks to Adam Moody for reporting the issue.

cmr=v1.7.4:reviewer=hjelmn:subject=Fix ROMIO compile error w/ Lustre 2.4

This commit was SVN r30393.
2014-01-23 14:15:35 +00:00
Christoph Niethammer
86776daf75 Fixed typo in opal output message.
This commit was SVN r30392.
2014-01-23 08:37:40 +00:00
Mike Dubman
071838bb0a HCOLL: call hcoll_finalize and hcoll progress unregister in case of hcoll module query failures
fixed by Elena, reviewed by Val/Miked
cmr=v1.7.4:reviewer=ompi-rm1.7

This commit was SVN r30390.
2014-01-23 07:29:23 +00:00
Ralph Castain
06e6a06f3e Cleanup a couple of abstraction breaks found by Thomas Naughton
This commit was SVN r30371.
2014-01-22 21:36:24 +00:00
Hadi Montakhabi
8af6b8b4e4 add support for PLFS filesystem
This commit was SVN r30370.
2014-01-22 21:16:15 +00:00
Nathan Hjelm
7ba8bd81fa coll/ml: remove debug fprintfs
cmr=v1.7.5:ticket=trac:4158

This commit was SVN r30367.

The following Trac tickets were found above:
  Ticket 4158 --> https://svn.open-mpi.org/trac/ompi/ticket/4158
2014-01-22 17:21:05 +00:00
Nathan Hjelm
82d996fb76 coll/ml: cleanup some merge related errors
cmr=v1.7.5:ticket=trac:4158

This commit was SVN r30366.

The following Trac tickets were found above:
  Ticket 4158 --> https://svn.open-mpi.org/trac/ompi/ticket/4158
2014-01-22 16:48:09 +00:00
Nathan Hjelm
ff4c9c808a btl/ugni: fix leak in new sendi function.
cmr=v1.7.5:ticket=trac:4151

This commit was SVN r30365.

The following Trac tickets were found above:
  Ticket 4151 --> https://svn.open-mpi.org/trac/ompi/ticket/4151
2014-01-22 16:32:07 +00:00
Nathan Hjelm
66b69da394 Fix a bug in the ob1 optimizations that can cause a segfault.
btl sendi functions currently can not handle the descriptor being NULL. The
send inline optimization was assuming (incorrectly) that NULL was ok.

cmr=v1.7.5:ticket=trac:4149

This commit was SVN r30364.

The following Trac tickets were found above:
  Ticket 4149 --> https://svn.open-mpi.org/trac/ompi/ticket/4149
2014-01-22 16:31:58 +00:00
Nathan Hjelm
1a021b8f2d coll/ml: add support for blocking and non-blocking allreduce, reduce, and
allgather.

The new collectives provide a signifigant performance increase over tuned for
small and medium messages. We are initially setting the priority lower than
tuned until this has had some time to soak in the trunk. Please set
coll_ml_priority to 90 for MTT runs.

Credit for this work goes to Manjunath Gorentla Venkata (ORNL), Pavel Shamis (ORNL),
and Nathan Hjelm (LANL).

Commit details (for reference):

Import ORNL's collectives for MPI_Allreduce, MPI_Reduce, and MPI_Allgather.

We need to take the basesmuma header into account when calculating the
ptpcoll small message thresholds. Add a define to bcol.h indicating the
maximum header size so we can take the header into account while not
making ptpcoll dependent on information from basesmuma.

This resolves an issue with allreduce where ptpcoll overwrites the
header of the next buffer in the basesmuma bank.

Fix reduce and make a sequential collective launcher in coll_ml_inlines.h

The root calculation for reduce was wrong for any root != 0. There are
four possibilities for the root:

 - The root is not the current process but is in the current hierarchy. In
   this case the root is the index of the global root as specified in the
   root vector.

 - The root is not the current process and is not in the next level of the
   hierarchy. In this case 0 must be the local root since this process will
   never communicate with the real root.

 - The root is not the current process but will be in next level of the
   hierarchy. In this case the current process must be the root.

 - I am the root. The root is my index.

Tested with IMB which rotates the root on every call to MPI_Reduce. Consider
IMB the reproducer for the issue this commit solves.

Make the bcast algorithm decision an enumerated variable

Resolve various asset failures when destructing coll ml requests.

Two issues:

 - Always reset the request to be invalid before returning it to the
   free list. This will avoid an asset in ompi_request_t's destructor.
   OMPI_REQUEST_FINI does this (and also releases the fortran handle
   index).

 - Never explicitly construct or destruct the superclass of an opal
   object. This screws up the class function tables and will cause
   either an assert failure or a segmentation fault when destructing
   coll ml requests.

Cleanup allgather.

I removed the duplicate non-blocking and blocking functions and modeled
the cleanup after what I found in allreduce. Also cleaned up the code
somewhat.

Don't bother copying from the send to the recieve buffer in
bcol_basesmuma_allreduce_intra_fanin_fanout if the pointers are the
same.

The eliminates a warning about memcpy and aliasing and avoids an
unnecessary call to memcpy.

Alwasy call CHECK_AND_RELEASE on memsync collectives.

There was a call to OBJ_RELEASE on the collective communicator but
because CHECK_AND_RECYLCE was never called there was not matching call
to OBJ_RELEASE. This caused coll ml to leak communicators.

Make allreduce use the sequential collective launcher in coll_ml_inlines.h

Just launch the next collective in the component progress.

I am a little unsure about this patch. There appears to be some sort
of race between collectives that causes buffer exhaustion in some cases
(IMB Allreduce is a reproducer). Changing progress to only launch the
next bcol seems to resolve the issue but might not be the best fix.

Note that I see little-no performance penalty for this change.

Fix allreduce when there are extra sources.

There was an issue with the buffer offset calculation when there are
extra sources. In the case of extra sources == 1 the offset was set
to buffer_size (just past the header of the next buffer). I adjusted
the buffer size to take into accoun the maximum header size (see the
earlier commit that added this) and simplified the offset calculation.

Make reduce/allreduce non-blocking. This is required for MPI_Comm_idup
to work correctly.

This has been tested with various layouts using the ibm testsuite and
imb and appears to have the same performance as the old blocking version.

Fix allgather for non-contiguous layouts and simplify parsing the
topology.

Some things in this patch:

 - There were several comments to the effect that level 0 of the
   hierarchy MUST contain all of the ranks. At least one function
   made this assumption but it was not true. I changed the sbgp
   components and the coll ml initization code to enforce this
   requirement.

 - Ensure that hierarchy level 0 has the ranks in the correct
   scatter gather order. This removes the need for a separate
   sort list and fixes the offset calculation for allgather.

 - There were several passes over the hierarchy to determine
   properties of the hierarchy. I eliminated these extra passes
   and the memory allocation associated with them and calculate the
   tree properties on the fly. The same DFS recursion also handles
   the re-order of level 0.

All these changes have been verified with MPI_Allreduce, MPI_Reduce, and
MPI_Allgather. All functions now pass all IBM/Open MPI, and IMB tests.

coll/ml: correct pointer usage for MPI_BOTTOM

Since contiguous datatypes are copied via memcpy (bypassing the convertor) we
need to adjust for the lb of the datatype. This corrects problems found testing
code that uses MPI_BOTTOM (NULL) as the send pointer.

Add fallback collectives for allreduce and reduce.

cmr=v1.7.5:reviewer=pasha

This commit was SVN r30363.
2014-01-22 15:39:19 +00:00
Nathan Hjelm
c9c335544e btl/ugni: fix a typo in r30353
cmr=v1.7.5:ticket=trac:4151

This commit was SVN r30354.

The following SVN revision numbers were found above:
  r30353 --> open-mpi/ompi@aa3fea55b2

The following Trac tickets were found above:
  Ticket 4151 --> https://svn.open-mpi.org/trac/ompi/ticket/4151
2014-01-21 21:02:28 +00:00
Nathan Hjelm
aa3fea55b2 btl/ugni: re-add a sendi function to exploit the new optimization in
ob1.

Also update LANL platform files to use the latest version of ugni.

cmr=v1.7.5:reviewer=manjugv

This commit was SVN r30353.
2014-01-21 20:53:35 +00:00
Nathan Hjelm
2b57f4227e ob1: optimize blocking send and receive paths
Per RFC. There are two optimizations in this commit:

 - Allocate requests for blocking sends and receives on the stack. This
   bypasses the request free list and saves two atomics on the critical path.
   This change improves the small message ping-pong by 50-200ns on both AMD
   and Intel CPUs.

 - For small messages try to use the btl sendi function before intializing a
   send request. If the sendi fails or the btl does not have a sendi function
   silently fallback on the standard send path.

cmr=v1.7.5:reviewer=brbarret

This commit was SVN r30343.
2014-01-21 15:16:21 +00:00
Mike Dubman
b8550a55a7 HCOLL: many fixes
Adds coll_hcoll_np mca parameter similar to that of fca component (defaults to 32). Those who use hcoll be aware that from now on the communicators less than 32 procs will run w/o hcoll by default. - Resolves fallback issue in case libhcoll runs out of allowed contexts. The solution is moving hcoll_context_create from comm_enable to comm_query. Shortly, comm_enable should never return OMPI_ERROR in the coll component with highest priority (hcoll). Otherwise the ompi coll_base_select will unselect the coll funtion pointers and module references leaving the communicator w/o coll pointer. This will cause the fail. Same behavior can be reproduced even with tuned if one would hardcore some "return OMPI_ERROR" into it's module_enable funtion. - Additionally, removed all the dead code under #if 0; removed unused variables (path for library, active_modules list) and classes (module list wrapper)

Fixed by Val, Reviewed by Devendar/Josh/Miked

cmr=v1.7.4:reviewer=ompi-rm1.7

This commit was SVN r30341.
2014-01-21 12:19:47 +00:00
Ralph Castain
2cf4862b49 Cleanup warnings for use of void* - requires intermediate cast to uintptr_t. Thanks to Paul Hargrove for reporting it
cmr=v1.7.4:reviewer=jsquyres

This commit was SVN r30333.
2014-01-20 15:44:45 +00:00
Edgar Gabriel
be5d5834c5 fix the problem identified by a user on the mailing list with MPI_MODE_EXCL
cmr=v1.7.4:reviewer=vvenkatesan:subject=fix a problem when opening a file with MODE_EXCL

This commit was SVN r30324.
2014-01-18 16:06:27 +00:00
Nathan Hjelm
c88626510c Fix a merge issues with new ROMIO and fix obvious ROMIO bug.
cmr=v1.7.4:reviewer=jsquyres

This commit was SVN r30319.
2014-01-18 00:29:16 +00:00
Hadi Montakhabi
8c14411289 f_cc_size is contiguous chunk size, not the stripe width. There is no stripe_width in the file handle structure.
This commit was SVN r30314.
2014-01-17 18:35:55 +00:00
Nathan Hjelm
f2a73fcdbd udreg: free huge page allocations correctly
This commit fixes an error path that occurs when huge page allocations are
enabled. In this case we allocate a huge page and try to register it but fail.
We then were calling free on the opal object. Fix this by calling the proper free
function.

cmr=v1.7.4:reviewer=rhc

This commit was SVN r30289.
2014-01-14 16:26:06 +00:00
Nathan Hjelm
f9d2032705 vader: ensure fast box data is aligned on 4-byte boundaries
This commit fixes a bus error on Solaris/Sparc.

Closes trac:4111

cmr=v1.7.5:ticket=trac:4053

This commit was SVN r30288.

The following Trac tickets were found above:
  Ticket 4053 --> https://svn.open-mpi.org/trac/ompi/ticket/4053
  Ticket 4111 --> https://svn.open-mpi.org/trac/ompi/ticket/4111
2014-01-14 16:04:52 +00:00
Rolf vandeVaart
e75afb2b82 Fix bug in distance computation code when deciding which devices to use on a NUMA node.
Also add a verbose flag so one can see what devices are selected as well as another flag to override
locality information and use all devices on the node.  

This commit was SVN r30287.
2014-01-14 15:41:56 +00:00
Nathan Hjelm
da1316ca6e vader: don't OBJ_RELEASE endpoint rcaches.
cmr=v1.7.4:reviewer=rhc

This commit was SVN r30284.
2014-01-13 23:44:34 +00:00
Jeff Squyres
20d6391734 Patch submitted by Paul Hargrove to fix NetBSD compile with -laio.
NetBSD puts the AIO functions in -lrt, vs. the usual libc.  So we
need the fbtl/posix configure.m4 to test for -lrt properly.

Reviewed by Jeff Squyres.

cmr=v1.7.4:reviewer=ompi-rm1.7:subject=Fix NetBSD use of -laio

This commit was SVN r30274.
2014-01-13 18:49:39 +00:00
Yossi Etigin
7564e2c13f Fix a recursion in mxm send flow which happens when mpi starts a new send from the context of send completion callback.
cmr=v1.7.5:reviewer=jsquyres

This commit was SVN r30265.
2014-01-12 17:47:03 +00:00
Yossi Etigin
9504969f7d fix communicator double-free from pt2pt component, caused by r29938.
cmr=v1.7.5:reviewer=brbarret

This commit was SVN r30264.

The following SVN revision numbers were found above:
  r29938 --> open-mpi/ompi@ecfb122c97
2014-01-12 17:38:14 +00:00
Ralph Castain
286ff6d552 For large scale systems, we would like to avoid doing a full modex during MPI_Init so that launch will scale a little better. At the moment, our options are somewhat limited as only a few BTLs don't immediately call modex_recv on all procs during startup. However, for those situations where someone can take advantage of it, add the ability to do a "modex on demand" retrieval of data from remote procs when we launch via mpirun.
NOTE: launch performance will be absolutely awful if you do this with BTLs that aren't configured to modex_recv on first message!

Even with "modex on demand", we still have to do a barrier in place of the modex - we simply don't move any data around, which does reduce the time impact. The barrier is required to ensure that the other proc has in fact registered all its BTL info and therefore is prepared to hand over a complete data package. Otherwise, you may not get the info you need. In addition, the shared memory BTL can fail to properly rendezvous as it expects the barrier to be in place.

This behavior will *only* take effect under the following conditions:

1. launched via mpirun

2. #procs is greater than ompi_hostname_cutoff, which defaults to UINT32_MAX

3. mca param rte_orte_direct_modex is set to 1. At the moment, we are having problems getting this param to register properly, so only the first two conditions are in effect. Still, the bottom line is you have to *want* this behavior to get it.

The planned next evolution of this will be to make the direct modex be non-blocking - this will require two fixes:

1. if the remote proc doesn't have the required info, then let it delay its response until it does. This means we need a way for the MPI layer to tell the RTE "I am done entering modex data".

2. adjust the SM rendezvous logic to loop until the required file has been created

Creating a placeholder to bring this over to 1.7.5 when ready.

cmr=v1.7.5:reviewer=hjelmn:subject=Enable direct modex at scale

This commit was SVN r30259.
2014-01-11 17:36:06 +00:00
Jeff Squyres
34ae50a0ed Fix int <--> pointer casting by adding intermediate cast through (intptr_t)
Reviewed by Dave Goodell

cmr=v1.7.4:reviewer=ompi-rm1.7:subject=Add intptr_t casting in usnic btl

This commit was SVN r30243.
2014-01-10 20:42:53 +00:00
Nathan Hjelm
5259ab213f Fix one more error path in udreg. In this case we hit the maximum size
of the udreg cache and get a different error code back.

cmr=v1.7.4:reviewer=rhc

This commit was SVN r30242.
2014-01-10 19:27:32 +00:00
Ralph Castain
9566650458 Per Marco, don't define a "min" function if one is already defined to avoid conflict with cygwin reserved word
This commit was SVN r30241.
2014-01-10 18:03:25 +00:00
Ralph Castain
880943dc10 Per Marco, rename "interface" to "tcp_interface" to avoid cygwin reserved word
This commit was SVN r30240.
2014-01-10 18:02:22 +00:00
Ralph Castain
c7a94a57d7 Per Marco, rename ERROR tags to exit_ERROR to avoid cygwin reserved name issues.
Refs trac:4085

This commit was SVN r30239.

The following Trac tickets were found above:
  Ticket 4085 --> https://svn.open-mpi.org/trac/ompi/ticket/4085
2014-01-10 18:00:49 +00:00
Jeff Squyres
350d989c00 Fix OpenBSD warnings where <malloc.h> is available and usable, but not
intended to be used and emits a compile-time warning.

Thanks to Paul Hargrove for identifying the issue.

cmr=v1.7.4:reviewer=hjelmn:subject=remove/replace malloc.h

This commit was SVN r30231.
2014-01-10 17:20:49 +00:00
Jeff Squyres
53a3defde9 s/CACHE_LINE_SIZE/BASESMUMA_CACHE_LINE_SIZE/g to avoid a system macro
name clash on some BSDs.

cmr=v1.7.4:reviewer=pasha

This commit was SVN r30230.
2014-01-10 16:48:43 +00:00
Edgar Gabriel
217e61e345 add proper typcasts to intptr_t to avoid warnings on 32bit systems.
This commit was SVN r30229.
2014-01-10 16:19:04 +00:00