* Remove paffinity, maffinity, and carto frameworks -- they've been
wholly replaced by hwloc.
* Move ompi_mpi_init() affinity-setting/checking code down to ORTE.
* Update sm, smcuda, wv, and openib components to no longer use carto.
Instead, use hwloc data. There are still optimizations possible in
the sm/smcuda BTLs (i.e., making multiple mpools). Also, the old
carto-based code found out how many NUMA nodes were ''available''
-- not how many were used ''in this job''. The new hwloc-using
code computes the same value -- it was not updated to calculate how
many NUMA nodes are used ''by this job.''
* Note that I cannot compile the smcuda and wv BTLs -- I ''think''
they're right, but they need to be verified by their owners.
* The openib component now does a bunch of stuff to figure out where
"near" OpenFabrics devices are. '''THIS IS A CHANGE IN DEFAULT
BEHAVIOR!!''' and still needs to be verified by OpenFabrics vendors
(I do not have a NUMA machine with an OpenFabrics device that is a
non-uniform distance from multiple different NUMA nodes).
* Completely rewrite the OMPI_Affinity_str() routine from the
"affinity" mpiext extension. This extension now understands
hyperthreads; the output format of it has changed a bit to reflect
this new information.
* Bunches of minor changes around the code base to update names/types
from maffinity/paffinity-based names to hwloc-based names.
* Add some helper functions into the hwloc base, mainly having to do
with the fact that we have the hwloc data reporting ''all''
topology information, but sometimes you really only want the
(online | available) data.
This commit was SVN r26391.
is zero. The routine assumes that at least one process is available in the
group, which lead to a segfault when creating communicators with GROUP_EMPTY.
Fixes trac:2752
This commit was SVN r24595.
The following Trac tickets were found above:
Ticket 2752 --> https://svn.open-mpi.org/trac/ompi/ticket/2752
Mostly TAB to spaces changes, though a couple style fixes were included as well.
The tab/space issue was causing problems with off-trunk branch merging.
This commit was SVN r23827.
INTERNAL to EXTRA_RETAIN, because not all "internal" communicators
have this flag set (only internal communicators with CIDs less than
their parent). Hence, what this flag ''really'' means is that there
was an extra RETAIN performed on it. So name the flag just that --
EXTRA_RETAIN -- indicating that an extra RETAIN has occurred.
This commit was SVN r22690.
The following SVN revision numbers were found above:
r22671 --> open-mpi/ompi@61dee816db
communicator that we created has a lower CID than the parent comm. This can
happen when using the hierarch collective communication module or for
inter-communicators (since we make a duplicate of the original communicator).
This is not a problem as long as the user calls MPI_Comm_free on the parent
communicator. However, if the communicators are not freed by the user but
released by Open MPI in MPI_Finalize, we walk through the list of still
available communicators and free them one by one. Thus, local_comm is freed
before the actual inter-communicator. However, the local_comm pointer in the
inter communicator will still contain the 'previous' address of the local_comm
and thus this will lead to a segmentation violation. In order to prevent that
from happening, we increase the reference counter local_comm by one if its CID
is lower than the parent. We cannot increase however its reference counter if
the CID of local_comm is larger than the CID of the inter communicators, since
a regular MPI_Comm_free would leave in that the case the local_comm hanging
around and thus we would not recycle CID's properly, which was the reason and
the cause for this trouble.
This commit fixes tickets 2094 and 2166. Note however, that I want to close
them manually, since a slightly different patch is required for the 1.4
series. This commit will have to be applied for the 1.5 series. And I will
need a volunteer to review it.
This commit was SVN r22671.
in the v1.2 series the cid's could never go above the max. allowed for a
particular pml. Because of that, pml_add_comm never checked for the cid, and
in fact pml_add_comm was called in comm_set, which is *before* we knew the
cid.
in the v1.3 series (and trunk) we check now the cid to detect overflow, and
because of that pml_add_comm has been moved *after* the cid allocation
routine, namely into the comm_activate routine.
in the v1.2 series, the comm_activate contained a synchronization step of the
old communicator in order to prevent incoming fragments on the new
communicator, with the main problem being that the allreduce in the
communicator allocation finished at different times on different processes,
and thus, this scenario could and did really occur.
in the v1.3 series, the comm_activate does not contain the synchronization
step anymore, since we introduced the new queue for fragments with unknown
cid. The problem is however, that whether a fragment is known or not is
decided by using ompi_comm_lookup(), which will return something useful as
soon as the cid allocation finished, even before pml_add_comm has been
called. So there is a small time gap where we will not post a message into
queue for unknown cid's, but we can also not look up the process structure
belonging to the rank in that comm ( that is in pml_ob1_match_recv_frag or
something like that).
The current fix reintroduces the synchronization step in comm_activate, and
ensures that no fragment can be received for a new communicator before the
synchronization occurs , and thus comm_nextcid() and pml_add_comm has been
called. It seems to be the safest and easiest way for now. Welcome back, v1.2.
This commit was SVN r21970.
happens when hierarch is used. . Two major items:
- modify the comm_activate step to take an additional argument, indicating
whether the new communicatio has to go through the collective selection
step. This is not required sometimes (e.g. when a process calls
MPI_COMM_SPLIT with color=MPI_UNDEFINED), and contributed significantly to
the exhaustion of cids.
- when freeing a communicator, check whether we can reuse the block of cids
assigned to that comm. This only works if the current front of the cid
assignment (cid_block_start) is right ater the block of cids assigned to this
comm.
Fixes trac:1904
Fixes trac:1926
This commit was SVN r21296.
The following Trac tickets were found above:
Ticket 1904 --> https://svn.open-mpi.org/trac/ompi/ticket/1904
Ticket 1926 --> https://svn.open-mpi.org/trac/ompi/ticket/1926
- Delete unnecessary header files using
contrib/check_unnecessary_headers.sh after applying
patches, that include headers, being "lost" due to
inclusion in one of the now deleted headers...
In total 817 files are touched.
In ompi/mpi/c/ header files are moved up into the actual c-file,
where necessary (these are the only additional #include),
otherwise it is only deletions of #include (apart from the above
additions required due to notifier...)
- To get different MCAs (OpenIB, TM, ALPS), an earlier version was
successfully compiled (yesterday) on:
Linux locally using intel-11, gcc-4.3.2 and gcc-SVN + warnings enabled
Smoky cluster (x86-64 running Linux) using PGI-8.0.2 + warnings enabled
Lens cluster (x86-64 running Linux) using Pathscale-3.2 + warnings enabled
This commit was SVN r21096.
Often, orte/util/show_help.h is included, although no functionality
is required -- instead, most often opal_output.h, or
orte/mca/rml/rml_types.h
Please see orte_show_help_replacement.sh commited next.
- Local compilation (Linux/x86_64) w/ -Wimplicit-function-declaration
actually showed two *missing* #include "orte/util/show_help.h"
in orte/mca/odls/base/odls_base_default_fns.c and
in orte/tools/orte-top/orte-top.c
Manually added these.
Let's have MTT the last word.
This commit was SVN r20557.
I'm unable to split it in two parts, my patch and Edgar's one. So I just update
copyright information for both of us.
What this patch do:
- it use the unexpected queue create by commit r19562 to dispatch the
unexpected message to the right communicator (once this communicator
is created and initialized).
- delay the PML comm_add until we have the context_id for the new communicator.
- only do the PML comm_add on processes that really belong to the new
communicator. Please read the lengthy comment in the source code for the
reason behind this.
This commit was SVN r19929.
The following SVN revision numbers were found above:
r19562 --> open-mpi/ompi@acd3406aa7
* Various changes to enable 0-dimensional cartesian communicators:
* Set various mtc_* members to NULL when there are 0 dimensions (and
don't bother trying to memcpy these arrays when duplicating the
communicator -- because they're NULL)
* adjust topo_base_cart_sub to correctly handle 0 dimensions
(simplified it a bit)
* adjust a few error codes to return ERR_OUT_OF_RESOURCE
* adjust error checking of CART_CREATE, CART_RANK
* Allow MPI_GRAPH_CREATE to accept 0 == nnodes.
* Bump reported MPI version in mpi.h to 2.1
This commit was SVN r19461.
The following Trac tickets were found above:
Ticket 1236 --> https://svn.open-mpi.org/trac/ompi/ticket/1236
The optimization that was introduced a year ago for saving a collective
synchronization step for certain communicator creation functions has to be
disabled for now. The bug has been exposed by the hierarch module, but could
appear as well for inter-communicator creations. The problem is, that within a
communicator creation step we invoke a comm_dup (for intercomm_create) or
other collective operations (in case of hierarch) before all processes have
been synchronized. This lead to the "Dropped message for non-existant
communicators" error. This commit disables the optimization without removing
it from the code base. In theory, it can be enabled again as soon as we have
the unexpected message queues for unknown cid's, which were required if I
remember right anyway for the multi-threaded scenarios and potentially for
fault tolerance.
Before moving the patch to 1.3 I would like to let it soak for a couple of
days on trunk. Please note, taht my 2nd comment on ticket #1408 was
semi-correct, since the order of activation of the communicator and quering
the collective module have already been changed earlier.
This commit was SVN r19139.
The following Trac tickets were found above:
Ticket 1408 --> https://svn.open-mpi.org/trac/ompi/ticket/1408
* add "register" function to mca_base_component_t
* converted coll:basic and paffinity:linux and paffinity:solaris to
use this function
* we'll convert the rest over time (I'll file a ticket once all
this is committed)
* add 32 bytes of "reserved" space to the end of mca_base_component_t
and mca_base_component_data_2_0_0_t to make future upgrades
[slightly] easier
* new mca_base_component_t size: 196 bytes
* new mca_base_component_data_2_0_0_t size: 36 bytes
* MCA base version bumped to v2.0
* '''We now refuse to load components that are not MCA v2.0.x'''
* all MCA frameworks versions bumped to v2.0
* be a little more explicit about version numbers in the MCA base
* add big comment in mca.h about versioning philosophy
This commit was SVN r19073.
The following Trac tickets were found above:
Ticket 1392 --> https://svn.open-mpi.org/trac/ompi/ticket/1392
The issue is that the field mca_topo_base_comm_t->mtc_periods_or_edges
has a different length, depending on whether the communicator is a
graph or a cart. One of the comm dup functions always assumed that it
was the length required by graph comms, which could lead to badness in
some cases. This commit makes the legnth of that field on a comm dup
be the proper length and copies the data over appropriately.
I also changed the syntax of the ompi_comm_copy_topo() function to use
shorter pointer notation; it made the code much easier to read and
fix.
This commit was SVN r18752.
The following Trac tickets were found above:
Ticket 1345 --> https://svn.open-mpi.org/trac/ompi/ticket/1345
After much work by Jeff and myself, and quite a lot of discussion, it has become clear that we simply cannot resolve the infinite loops caused by RML-involved subsystems calling orte_output. The original rationale for the change to orte_output has also been reduced by shifting the output of XML-formatted vs human readable messages to an alternative approach.
I have globally replaced the orte_output/ORTE_OUTPUT calls in the code base, as well as the corresponding .h file name. I have test compiled and run this on the various environments within my reach, so hopefully this will prove minimally disruptive.
This commit was SVN r18619.
such, the commit message back to the master SVN repository is fairly
long.
= ORTE Job-Level Output Messages =
Add two new interfaces that should be used for all new code throughout
the ORTE and OMPI layers (we already make the search-and-replace on
the existing ORTE / OMPI layers):
* orte_output(): (and corresponding friends ORTE_OUTPUT,
orte_output_verbose, etc.) This function sends the output directly
to the HNP for processing as part of a job-specific output
channel. It supports all the same outputs as opal_output()
(syslog, file, stdout, stderr), but for stdout/stderr, the output
is sent to the HNP for processing and output. More on this below.
* orte_show_help(): This function is a drop-in-replacement for
opal_show_help(), with two differences in functionality:
1. the rendered text help message output is sent to the HNP for
display (rather than outputting directly into the process' stderr
stream)
1. the HNP detects duplicate help messages and does not display them
(so that you don't see the same error message N times, once from
each of your N MPI processes); instead, it counts "new" instances
of the help message and displays a message every ~5 seconds when
there are new ones ("I got X new copies of the help message...")
opal_show_help and opal_output still exist, but they only output in
the current process. The intent for the new orte_* functions is that
they can apply job-level intelligence to the output. As such, we
recommend that all new ORTE and OMPI code use the new orte_*
functions, not thei opal_* functions.
=== New code ===
For ORTE and OMPI programmers, here's what you need to do differently
in new code:
* Do not include opal/util/show_help.h or opal/util/output.h.
Instead, include orte/util/output.h (this one header file has
declarations for both the orte_output() series of functions and
orte_show_help()).
* Effectively s/opal_output/orte_output/gi throughout your code.
Note that orte_output_open() takes a slightly different argument
list (as a way to pass data to the filtering stream -- see below),
so you if explicitly call opal_output_open(), you'll need to
slightly adapt to the new signature of orte_output_open().
* Literally s/opal_show_help/orte_show_help/. The function signature
is identical.
=== Notes ===
* orte_output'ing to stream 0 will do similar to what
opal_output'ing did, so leaving a hard-coded "0" as the first
argument is safe.
* For systems that do not use ORTE's RML or the HNP, the effect of
orte_output_* and orte_show_help will be identical to their opal
counterparts (the additional information passed to
orte_output_open() will be lost!). Indeed, the orte_* functions
simply become trivial wrappers to their opal_* counterparts. Note
that we have not tested this; the code is simple but it is quite
possible that we mucked something up.
= Filter Framework =
Messages sent view the new orte_* functions described above and
messages output via the IOF on the HNP will now optionally be passed
through a new "filter" framework before being output to
stdout/stderr. The "filter" OPAL MCA framework is intended to allow
preprocessing to messages before they are sent to their final
destinations. The first component that was written in the filter
framework was to create an XML stream, segregating all the messages
into different XML tags, etc. This will allow 3rd party tools to read
the stdout/stderr from the HNP and be able to know exactly what each
text message is (e.g., a help message, another OMPI infrastructure
message, stdout from the user process, stderr from the user process,
etc.).
Filtering is not active by default. Filter components must be
specifically requested, such as:
{{{
$ mpirun --mca filter xml ...
}}}
There can only be one filter component active.
= New MCA Parameters =
The new functionality described above introduces two new MCA
parameters:
* '''orte_base_help_aggregate''': Defaults to 1 (true), meaning that
help messages will be aggregated, as described above. If set to 0,
all help messages will be displayed, even if they are duplicates
(i.e., the original behavior).
* '''orte_base_show_output_recursions''': An MCA parameter to help
debug one of the known issues, described below. It is likely that
this MCA parameter will disappear before v1.3 final.
= Known Issues =
* The XML filter component is not complete. The current output from
this component is preliminary and not real XML. A bit more work
needs to be done to configure.m4 search for an appropriate XML
library/link it in/use it at run time.
* There are possible recursion loops in the orte_output() and
orte_show_help() functions -- e.g., if RML send calls orte_output()
or orte_show_help(). We have some ideas how to fix these, but
figured that it was ok to commit before feature freeze with known
issues. The code currently contains sub-optimal workarounds so
that this will not be a problem, but it would be good to actually
solve the problem rather than have hackish workarounds before v1.3 final.
This commit was SVN r18434.
* New/improved bootstrapping technique for DLLs
* First cut of the MPI handle debugging interface. It is still
evolving, but the interface is getting more stable.
* Some minor bugs were fixed in the unity topo component (brought to
light because of the new MPI handle debugging stuff).
Fixes trac:1209.
This commit was SVN r17730.
The following Trac tickets were found above:
Ticket 1209 --> https://svn.open-mpi.org/trac/ompi/ticket/1209
The commit has been tested for C/R and Cray operations, and on Odin (SLURM, rsh) and RoadRunner (TM). I tried to update all environments, but obviously could not test them. I know that Windows needs some work, and have highlighted what is know to be needed in the odls process component.
This represents a lot of work by Brian, Tim P, Josh, and myself, with much advice from Jeff and others. For posterity, I have appended a copy of the email describing the work that was done:
As we have repeatedly noted, the modex operation in MPI_Init is the single greatest consumer of time during startup. To-date, we have executed that operation as an ORTE stage gate that held the process until a startup message containing all required modex (and OOB contact info - see #3 below) info could be sent to it. Each process would send its data to the HNP's registry, which assembled and sent the message when all processes had reported in.
In addition, ORTE had taken responsibility for monitoring process status as it progressed through a series of "stage gates". The process reported its status at each gate, and ORTE would then send a "release" message once all procs had reported in.
The incoming changes revamp these procedures in three ways:
1. eliminating the ORTE stage gate system and cleanly delineating responsibility between the OMPI and ORTE layers for MPI init/finalize. The modex stage gate (STG1) has been replaced by a collective operation in the modex itself that performs an allgather on the required modex info. The allgather is implemented using the orte_grpcomm framework since the BTL's are not active at that point. At the moment, the grpcomm framework only has a "basic" component analogous to OMPI's "basic" coll framework - I would recommend that the MPI team create additional, more advanced components to improve performance of this step.
The other stage gates have been replaced by orte_grpcomm barrier functions. We tried to use MPI barriers instead (since the BTL's are active at that point), but - as we discussed on the telecon - these are not currently true barriers so the job would hang when we fell through while messages were still in process. Note that the grpcomm barrier doesn't actually resolve that problem, but Brian has pointed out that we are unlikely to ever see it violated. Again, you might want to spend a little time on an advanced barrier algorithm as the one in "basic" is very simplistic.
Summarizing this change: ORTE no longer tracks process state nor has direct responsibility for synchronizing jobs. This is now done via collective operations within the MPI layer, albeit using ORTE collective communication services. I -strongly- urge the MPI team to implement advanced collective algorithms to improve the performance of this critical procedure.
2. reducing the volume of data exchanged during modex. Data in the modex consisted of the process name, the name of the node where that process is located (expressed as a string), plus a string representation of all contact info. The nodename was required in order for the modex to determine if the process was local or not - in addition, some people like to have it to print pretty error messages when a connection failed.
The size of this data has been reduced in three ways:
(a) reducing the size of the process name itself. The process name consisted of two 32-bit fields for the jobid and vpid. This is far larger than any current system, or system likely to exist in the near future, can support. Accordingly, the default size of these fields has been reduced to 16-bits, which means you can have 32k procs in each of 32k jobs. Since the daemons must have a vpid, and we require one daemon/node, this also restricts the default configuration to 32k nodes.
To support any future "mega-clusters", a configuration option --enable-jumbo-apps has been added. This option increases the jobid and vpid field sizes to 32-bits. Someday, if necessary, someone can add yet another option to increase them to 64-bits, I suppose.
(b) replacing the string nodename with an integer nodeid. Since we have one daemon/node, the nodeid corresponds to the local daemon's vpid. This replaces an often lengthy string with only 2 (or at most 4) bytes, a substantial reduction.
(c) when the mca param requesting that nodenames be sent to support pretty error messages, a second mca param is now used to request FQDN - otherwise, the domain name is stripped (by default) from the message to save space. If someone wants to combine those into a single param somehow (perhaps with an argument?), they are welcome to do so - I didn't want to alter what people are already using.
While these may seem like small savings, they actually amount to a significant impact when aggregated across the entire modex operation. Since every proc must receive the modex data regardless of the collective used to send it, just reducing the size of the process name removes nearly 400MBytes of communication from a 32k proc job (admittedly, much of this comm may occur in parallel). So it does add up pretty quickly.
3. routing RML messages to reduce connections. The default messaging system remains point-to-point - i.e., each proc opens a socket to every proc it communicates with and sends its messages directly. A new option uses the orteds as routers - i.e., each proc only opens a single socket to its local orted. All messages are sent from the proc to the orted, which forwards the message to the orted on the node where the intended recipient proc is located - that orted then forwards the message to its local proc (the recipient). This greatly reduces the connection storm we have encountered during startup.
It also has the benefit of removing the sharing of every proc's OOB contact with every other proc. The orted routing tables are populated during launch since every orted gets a map of where every proc is being placed. Each proc, therefore, only needs to know the contact info for its local daemon, which is passed in via the environment when the proc is fork/exec'd by the daemon. This alone removes ~50 bytes/process of communication that was in the current STG1 startup message - so for our 32k proc job, this saves us roughly 32k*50 = 1.6MBytes sent to 32k procs = 51GBytes of messaging.
Note that you can use the new routing method by specifying -mca routed tree - if you so desire. This mode will become the default at some point in the future.
There are a few minor additional changes in the commit that I'll just note in passing:
* propagation of command line mca params to the orteds - fixes ticket #1073. See note there for details.
* requiring of "finalize" prior to "exit" for MPI procs - fixes ticket #1144. See note there for details.
* cleanup of some stale header files
This commit was SVN r16364.
used at nce (up to one unique collective module per collective function).
Matches r15795:15921 of the tmp/bwb-coll-select branch
This commit was SVN r15924.
The following SVN revisions from the original message are invalid or
inconsistent and therefore were not cross-referenced:
r15795
r15921
Cleanup ALL instances of output involving the printing of orte_process_name_t structures using the ORTE_NAME_ARGS macro so that the number of fields and type of data match. Replace those values with a new macro/function pair ORTE_NAME_PRINT that outputs a string (using the new thread safe data capability) so that any future changes to the printing of those structures can be accomplished with a change to a single point.
Note that I could not possibly find outputs that directly print the orte_process_name_t fields, but only dealt with those that used ORTE_NAME_ARGS. Hence, you may still have a few outputs that bark during compilation. Also, I could only verify those that fall within environments I can compile on, so other environments may yield some minor warnings.
This commit was SVN r15517.
* Do not add new procs to the global list during modex callback or
when sharing orte names during accept/connect. For modex, we
cache the modex info for later, in case that proc ever does get
added to the global proc list. For accept/connect orte name
exchange between the roots, we only need the orte name, so no
need to add a proc structure anyway. The procs will be added
to the global process list during the proc exchange later in
the wireup process
* Rename proc_get_namebuf and proc_get_proclist to proc_pack
and proc_unpack and extend them to include all information
needed to build that proc struct on a remote node (which
includes ORTE name, architecture, and hostname). Change
unpack to call pml_add_procs for the entire list of new
procs at once, rather than one at a time.
* Remove ompi_proc_find_and_add from the public proc
interface and make it a private function. This function
would add a half-created proc to the global proc list, so
making it harder to call is a good thing.
This means that there's only two ways to add new procs into the global proc list at this time: During MPI_INIT via the call to ompi_proc_init, where my job is added to the list and via ompi_proc_unpack using a buffer from a packed proc list sent to us by someone else. Currently, this is enough to implement MPI semantics. We can extend the interface more if we like, but that may require HNP communication to get the remote proc information and I wanted to avoid that if at all possible.
Refs trac:564
This commit was SVN r12798.
The following Trac tickets were found above:
Ticket 564 --> https://svn.open-mpi.org/trac/ompi/ticket/564
* don't load data into a buffer until we have the data, as
the data contains some header information needed to
properly load the data
This commit was SVN r12792.
Accordingly, there are new APIs to the name service to support the ability to get a job's parent, root, immediate children, and all its descendants. In addition, the terminate_job, terminate_orted, and signal_job APIs for the PLS have been modified to accept attributes that define the extent of their actions. For example, doing a "terminate_job" with an attribute of ORTE_NS_INCLUDE_DESCENDANTS will terminate the given jobid AND all jobs that descended from it.
I have tested this capability on a MacBook under rsh, Odin under SLURM, and LANL's Flash (bproc). It worked successfully on non-MPI jobs (both simple and including a spawn), and MPI jobs (again, both simple and with a spawn).
This commit was SVN r12597.
Clean up the remainder of the size_t references in the runtime itself. Convert to orte_std_cntr_t wherever it makes sense (only avoid those places where the actual memory size is referenced).
Remove the obsolete oob barrier function (we actually obsoleted it a long time ago - just never bothered to clean it up).
I have done my best to go through all the components and catch everything, even if I couldn't test compile them since I wasn't on that type of system. Still, I cannot guarantee that problems won't show up when you test this on specific systems. Usually, these will just show as "warning: comparison between signed and unsigned" notes which are easily fixed (just change a size_t to orte_std_cntr_t).
In some places, people didn't use size_t, but instead used some other variant (e.g., I found several places with uint32_t). I tried to catch all of them, but...
Once we get all the instances caught and fixed, this should once and for all resolve many of the heterogeneity problems.
This commit was SVN r11204.
