is enabled and fix a bug in the handling of watermark performance
variables.
cmr=v1.7.3:ticket=trac:3725:reviewer=jsquyres
This commit was SVN r29068.
The following Trac tickets were found above:
Ticket 3725 --> https://svn.open-mpi.org/trac/ompi/ticket/3725
*** THIS RFC INCLUDES A MINOR CHANGE TO THE MPI-RTE INTERFACE ***
Note: during the course of this work, it was necessary to completely separate the MPI and RTE progress engines. There were multiple places in the MPI layer where ORTE_WAIT_FOR_COMPLETION was being used. A new OMPI_WAIT_FOR_COMPLETION macro was created (defined in ompi/mca/rte/rte.h) that simply cycles across opal_progress until the provided flag becomes false. Places where the MPI layer blocked waiting for RTE to complete an event have been modified to use this macro.
***************************************************************************************
I am reissuing this RFC because of the time that has passed since its original release. Since its initial release and review, I have debugged it further to ensure it fully supports tests like loop_spawn. It therefore seems ready for merge back to the trunk. Given its prior review, I have set the timeout for one week.
The code is in https://bitbucket.org/rhc/ompi-oob2
WHAT: Rewrite of ORTE OOB
WHY: Support asynchronous progress and a host of other features
WHEN: Wed, August 21
SYNOPSIS:
The current OOB has served us well, but a number of limitations have been identified over the years. Specifically:
* it is only progressed when called via opal_progress, which can lead to hangs or recursive calls into libevent (which is not supported by that code)
* we've had issues when multiple NICs are available as the code doesn't "shift" messages between transports - thus, all nodes had to be available via the same TCP interface.
* the OOB "unloads" incoming opal_buffer_t objects during the transmission, thus preventing use of OBJ_RETAIN in the code when repeatedly sending the same message to multiple recipients
* there is no failover mechanism across NICs - if the selected NIC (or its attached switch) fails, we are forced to abort
* only one transport (i.e., component) can be "active"
The revised OOB resolves these problems:
* async progress is used for all application processes, with the progress thread blocking in the event library
* each available TCP NIC is supported by its own TCP module. The ability to asynchronously progress each module independently is provided, but not enabled by default (a runtime MCA parameter turns it "on")
* multi-address TCP NICs (e.g., a NIC with both an IPv4 and IPv6 address, or with virtual interfaces) are supported - reachability is determined by comparing the contact info for a peer against all addresses within the range covered by the address/mask pairs for the NIC.
* a message that arrives on one TCP NIC is automatically shifted to whatever NIC that is connected to the next "hop" if that peer cannot be reached by the incoming NIC. If no TCP module will reach the peer, then the OOB attempts to send the message via all other available components - if none can reach the peer, then an "error" is reported back to the RML, which then calls the errmgr for instructions.
* opal_buffer_t now conforms to standard object rules re OBJ_RETAIN as we no longer "unload" the incoming object
* NIC failure is reported to the TCP component, which then tries to resend the message across any other available TCP NIC. If that doesn't work, then the message is given back to the OOB base to try using other components. If all that fails, then the error is reported to the RML, which reports to the errmgr for instructions
* obviously from the above, multiple OOB components (e.g., TCP and UD) can be active in parallel
* the matching code has been moved to the RML (and out of the OOB/TCP component) so it is independent of transport
* routing is done by the individual OOB modules (as opposed to the RML). Thus, both routed and non-routed transports can simultaneously be active
* all blocking send/recv APIs have been removed. Everything operates asynchronously.
KNOWN LIMITATIONS:
* although provision is made for component failover as described above, the code for doing so has not been fully implemented yet. At the moment, if all connections for a given peer fail, the errmgr is notified of a "lost connection", which by default results in termination of the job if it was a lifeline
* the IPv6 code is present and compiles, but is not complete. Since the current IPv6 support in the OOB doesn't work anyway, I don't consider this a blocker
* routing is performed at the individual module level, yet the active routed component is selected on a global basis. We probably should update that to reflect that different transports may need/choose to route in different ways
* obviously, not every error path has been tested nor necessarily covered
* determining abnormal termination is more challenging than in the old code as we now potentially have multiple ways of connecting to a process. Ideally, we would declare "connection failed" when *all* transports can no longer reach the process, but that requires some additional (possibly complex) code. For now, the code replicates the old behavior only somewhat modified - i.e., if a module sees its connection fail, it checks to see if it is a lifeline. If so, it notifies the errmgr that the lifeline is lost - otherwise, it notifies the errmgr that a non-lifeline connection was lost.
* reachability is determined solely on the basis of a shared subnet address/mask - more sophisticated algorithms (e.g., the one used in the tcp btl) are required to handle routing via gateways
* the RML needs to assign sequence numbers to each message on a per-peer basis. The receiving RML will then deliver messages in order, thus preventing out-of-order messaging in the case where messages travel across different transports or a message needs to be redirected/resent due to failure of a NIC
This commit was SVN r29058.
This creates a really bad scaling behavior. Users have found a nearly 20% launch time differential between mpirun and PMI, with PMI being the slower method. Some of the problem is attributable to poor exchange algorithms in RM's like Slurm and Alps, but we make things worse by calling "get" so many times.
Nathan (with a tad advice from me) has attempted to alleviate this problem by reducing the number of "get" calls. This required the following changes:
* upon first request for data, have the OPAL db pmi component fetch and decode *all* the info from a given remote proc. It turned out we weren't caching the info, so we would continually request it and only decode the piece we needed for the immediate request. We now decode all the info and push it into the db hash component for local storage - and then all subsequent retrievals are fulfilled locally
* reduced the amount of data by eliminating the exchange of the OMPI_ARCH value if heterogeneity is not enabled. This was used solely as a check so we would error out if the system wasn't actually homogeneous, which was fine when we thought there was no cost in doing the check. Unfortunately, at large scale and with direct launch, there is a non-zero cost of making this test. We are open to finding a compromise (perhaps turning the test off if requested?), if people feel strongly about performing the test
* reduced the amount of RTE data being automatically fetched, and fetched the rest only upon request. In particular, we no longer immediately fetch the hostname (which is only used for error reporting), but instead get it when needed. Likewise for the RML uri as that info is only required for some (not all) environments. In addition, we no longer fetch the locality unless required, relying instead on the PMI clique info to tell us who is on our local node (if additional info is required, the fetch is performed when a modex_recv is issued).
Again, all this only impacts direct launch - all the info is provided when launched via mpirun as there is no added cost to getting it
Barring objections, we may move this (plus any required other pieces) to the 1.7 branch once it soaks for an appropriate time.
This commit was SVN r29040.
This commit reintroduces key compression into the pmi db. This feature
compresses the keys stored into the component into a small number of
PMI keys by serializing the data and base64 encoding the result. This
will avoid issues with Cray PMI which restricts us to ~ 3 PMI keys per
rank.
This commit was SVN r28993.
Add support for MPI_Count type and MPI_COUNT datatype and add the required
MPI-3 functions MPI_Get_elements_x, MPI_Status_set_elements_x,
MPI_Type_get_extent_x, MPI_Type_get_true_extent_x, and MPI_Type_size_x.
This commit adds only the C bindings. Fortran bindins will be added in
another commit. For now the MPI_Count type is define to have the same size
as MPI_Offset. The type is required to be at least as large as MPI_Offset
and MPI_Aint. The type was initially intended to be a ssize_t (if it was
the same size as a long long) but there were issues compiling romio with
that definition (despite the inclusion of stddef.h).
I updated the datatype engine to use size_t instead of uint32_t to support
large datatypes. This will require some review to make sure that 1) the
changes are beneficial, 2) nothing was broken by the change (I doubt
anything was), and 3) there are no performance regressions due to this
change.
Increase the maximum number of predifined datatypes to support MPI_Count
Put common get_elements code to ompi/datatype/ompi_datatype_get_elements.c
Update MPI_Get_count to reflect changes in MPI-3 (return MPI_UNDEFINED when the count is too large for an int)
This commit was SVN r28932.
This commit adds an API for registering and querying performance
variables (mca_base_pvar) in the MCA base. The existing MCA variable
system API has been updated to reflect the new API: MCA variable
groups have performance variables, and new types have been added (double,
unsigned long long) to reflect what is required by the MPI_T
interface. Additionally, the MCA variable group code has been split
into its own set of files: mca_base_var_group.[ch].
Details of the new API can be found in doxygen comments in the header:
mca_base_pvar.h.
Other changes to the variable system:
- Use an opal_hash_table to speed up variable/group lookup.
- Clean up code associated with MCA variable types.
- Registered performance variables are printed by ompi_info -a. In the
future an option should be added to control this behavior.
Changes to OMPI:
- Added full support for the MPI_T performance variable interface.
This commit was SVN r28800.
(since this is used in the fast path) for when you ''know'' that there
will be a room available:
* Don't do the last_unoccupied_room check
* Return void
This commit was SVN r28757.
Add an option to ompi_info (-l, --level) that takes a number in the
interval (1,9). Only MCA variables up to this level will be printed.
The default level is 1.
Print the level as part of both the parsable and readable output.
This commit was SVN r28750.
sandbox team has informed me that they are getting rid of SANDBOX_PID
in the future and that using SANDBOX_ON would be preferred.
This commit was SVN r28708.
To resolve this situation, add the ability to specify a backend topology file that mpirun shall use for its mapping operations. Create a new "set_topology" function in opal hwloc to support it.
This commit was SVN r28682.
Chose the crc32 function present in opal/util/crc.c as the hash function. The
performance should be sufficient for most cases. If not we can always change
the function again.
This commit was SVN r28629.
of the original datatype are allowed (not in type map nor extent). Make it
clear in the code.
Allow 0-count cases to the contiguous memory check.
This commit was SVN r28568.
some relevant updates/new functionality in the opal/mca/hwloc and
orte/mca/rmaps bases. This work was mainly developed by Mellanox,
with a bunch of advice from Ralph Castain, and some minor advice from
Brice Goglin and Jeff Squyres.
Even though this is mainly Mellanox's work, Jeff is committing only
for logistical reasons (he holds the hg+svn combo tree, and can
therefore commit it directly back to SVN).
-----
Implemented distance-based mapping algorithm as a new "mindist"
component in the rmaps framework. It allows mapping processes by NUMA
due to PCI locality information as reported by the BIOS - from the
closest to device to furthest.
To use this algorithm, specify:
{{{mpirun --map-by dist:<device_name>}}}
where <device_name> can be mlx5_0, ib0, etc.
There are two modes provided:
1. bynode: load-balancing across nodes
1. byslot: go through slots sequentially (i.e., the first nodes are
more loaded)
These options are regulated by the optional ''span'' modifier; the
command line parameter looks like:
{{{mpirun --map-by dist:<device_name>,span}}}
So, for example, if there are 2 nodes, each with 8 cores, and we'd
like to run 10 processes, the mindist algorithm will place 8 processes
to the first node and 2 to the second by default. But if you want to
place 5 processes to each node, you can add a span modifier in your
command line to do that.
If there are two NUMA nodes on the node, each with 4 cores, and we run
6 processes, the mindist algorithm will try to find the NUMA closest
to the specified device, and if successful, it will place 4 processes
on that NUMA but leaving the remaining two to the next NUMA node.
You can also specify the number of cpus per MPI process. This option
is handled so that we map as many processes to the closest NUMA as we
can (number of available processors at the NUMA divided by number of
cpus per rank) and then go on with the next closest NUMA.
The default binding option for this mapping is bind-to-numa. It works
if you don't specify any binding policy. But if you specified binding
level that was "lower" than NUMA (i.e hwthread, core, socket) it would
bind to whatever level you specify.
This commit was SVN r28552.
in generated executables on systems that support it. Use
--disable-wrapper-rpath to disable this behavior. See text in
README about --disable-wrapper-rpath for more details.
This commit was SVN r28479.
The following Trac tickets were found above:
Ticket 376 --> https://svn.open-mpi.org/trac/ompi/ticket/376
added by hwloc's embedding so that it doesn't appear in
libhwloc_embedded.la (and therefore propogate all the way up to
libmpi.la).
Committed upstream in hwloc SVN r5588.
This commit was SVN r28457.
The following SVN revisions from the original message are invalid or
inconsistent and therefore were not cross-referenced:
r5588
http://www.open-mpi.org/community/lists/users/2013/04/21689.php, the
assert in opal_datatype_is_contiguous_memory_layout() is not always
correct -- he supplied a test case where it was not valid,
essentially:
1. Call MPI_Type_create_indexed_block(0, ..., &newtype) and commit newtype
1. Call MPI_Type_create_resized(newtype, 0, nonzero_value, &resized) and commit resized
1. Call MPI_File_set_view with resized
This will eventually call opal_datatype_is_contiguous_memory_layout(),
and the assert will fail. After some consultation with George, it was
determined that the assert() is basically good, but it needs to also
check for (count != 0).
This commit was SVN r28398.
everything out before using it.
This is not in response to any known bug, but rather just a
pre-emptive, defensive move to help prevent bugs in code that forgets
to initialize a field.
This commit was SVN r28343.
representation is not correctly optimized (it is off by the extend).
During the data representation process, if the opportunity to merge several
items appear, we replace them with the new merged element. However, if one
of the components of this merged element was comming from a "loop representation"
then the new first element of this loop must have a displacement moved by the
extent of the loop.
This commit was SVN r28319.
Provide some nice error messages if we fail to set the limits. Since the user had to specifically request we set the limit, treat failure as an error-out situation.
This commit was SVN r28288.
Notes:
- This commit also eliminates the need for an available components list in use
in several frameworks. None of the code in question was making use of the
priority field of the priority component list item so these extra lists were
removed.
- Cleaned up selection code in several frameworks to sort lists using opal_list_sort.
- Cleans up the ompi/orte-info functions. Expose the functions that construct the
list of params so they can be used elsewhere.
patches for mtl/portals4 from brian
missed a few output variables in openib
This commit was SVN r28241.
Other changes:
- Added a flag to the MCA variable system to indicate a variable should go away
when its group does. Both mca_base_framework_var_register() and
mca_base_component_var_register() set this flag.
Notes:
- mca_base_components_open is deprecated. It will be removed in a future commit.
- All frameworks should use MCA_BASE_FRAMEWORK_DECLARE to declare their
framework structure.
- All calls to framework open/close functions should be changed to use the
mca_base_framework_* functions.
- Instead of special-casing installdirs a flag was added to prevent calling
into the variable system when opening a framework.
- Ralph: Clarify the functional definition of the "register" function in the
MCA framework object - it had the same name as another function that does a
totally different thing.
- As per discussion with Ralph the behavior of mca_base_framework_register()
is to always call mca_base_framework_components_register() if the framework's
register function was successful. This removed the need for frameworks to
have to call this function directly.
This commit was SVN r28237.
Features:
- Support for an override parameter file (openmpi-mca-param-override.conf).
Variable values in this file can not be overridden by any file or environment
value.
- Support for boolean, unsigned, and unsigned long long variables.
- Support for true/false values.
- Support for enumerations on integer variables.
- Support for MPIT scope, verbosity, and binding.
- Support for command line source.
- Support for setting variable source via the environment using
OMPI_MCA_SOURCE_<var name>=source (either command or file:filename)
- Cleaner API.
- Support for variable groups (equivalent to MPIT categories).
Notes:
- Variables must be created with a backing store (char **, int *, or bool *)
that must live at least as long as the variable.
- Creating a variable with the MCA_BASE_VAR_FLAG_SETTABLE enables the use of
mca_base_var_set_value() to change the value.
- String values are duplicated when the variable is registered. It is up to
the caller to free the original value if necessary. The new value will be
freed by the mca_base_var system and must not be freed by the user.
- Variables with constant scope may not be settable.
- Variable groups (and all associated variables) are deregistered when the
component is closed or the component repository item is freed. This
prevents a segmentation fault from accessing a variable after its component
is unloaded.
- After some discussion we decided we should remove the automatic registration
of component priority variables. Few component actually made use of this
feature.
- The enumerator interface was updated to be general enough to handle
future uses of the interface.
- The code to generate ompi_info output has been moved into the MCA variable
system. See mca_base_var_dump().
opal: update core and components to mca_base_var system
orte: update core and components to mca_base_var system
ompi: update core and components to mca_base_var system
This commit also modifies the rmaps framework. The following variables were
moved from ppr and lama: rmaps_base_pernode, rmaps_base_n_pernode,
rmaps_base_n_persocket. Both lama and ppr create synonyms for these variables.
This commit was SVN r28236.
binding. This fix was included in the upstream 1.6 series, but not
the upstream 1.5 series, and was therefore missed when we brought
1.5.2 to OMPI.
This commit was SVN r28212.
The following SVN revision numbers were found above:
r28040 --> open-mpi/ompi@3d44f97572
At the same time, fix a minor issue where the init hook was being called twice, once by the libc malloc and once by our malloc by removing the call from our malloc.
This commit was SVN r28202.
library to multiple libraries that are implicitly sucked into the executable
as a dependency of libmpi. The initialize hook isn't visible to libc on some
linux distributions when it's in libopal and libopal isn't explicity linked
into the executable. The fix is to have a duplicate initialize hook in
libmpi as well as libopal. *sigh*.
This commit was SVN r28164.
A few changes were required to support this move:
1. the PMI component used to identify rte-related data (e.g., host name, bind level) and package them as a unit to reduce the number of PMI keys. This code was moved up to the ORTE layer as the OPAL layer has no understanding of these concepts. In addition, the component locally stored data based on process jobid/vpid - this could no longer be supported (see below for the solution).
2. the hash component was updated to use the new opal_identifier_t instead of orte_process_name_t as its index for storing data in the hash tables. Previously, we did a hash on the vpid and stored the data in a 32-bit hash table. In the revised system, we don't see a separate "vpid" field - we only have a 64-bit opaque value. The orte_process_name_t hash turned out to do nothing useful, so we now store the data in a 64-bit hash table. Preliminary tests didn't show any identifiable change in behavior or performance, but we'll have to see if a move back to the 32-bit table is required at some later time.
3. the db framework was a "select one" system. However, since the PMI component could no longer use its internal storage system, the framework has now been changed to a "select many" mode of operation. This allows the hash component to handle all internal storage, while the PMI component only handles pushing/pulling things from the PMI system. This was something we had planned for some time - when fetching data, we first check internal storage to see if we already have it, and then automatically go to the global system to look for it if we don't. Accordingly, the framework was provided with a custom query function used during "select" that lets you seperately specify the "store" and "fetch" ordering.
4. the ORTE grpcomm and ess/pmi components, and the nidmap code, were updated to work with the new db framework and to specify internal/global storage options.
No changes were made to the MPI layer, except for modifying the ORTE component of the OMPI/rte framework to support the new db framework.
This commit was SVN r28112.
* Clean up ${includedir} and ${libdir} for script wrapper compilers
* Update script wrapper compilers to work like the C wrapper compilers w.r.t static and dynamic linking
* Remove the ORTE script wrapper compilers since they didn't support the ${includedir} stuff and Ralph said they weren't used anymore.
This commit was SVN r28052.
