Replace non-breaking spaces.

[simgrid.git] / docs / source / app_smpi.rst

.. _SMPI_doc:

===============================
SMPI: Simulate MPI Applications
===============================

.. raw:: html

   <object id="TOC" data="graphical-toc.svg" type="image/svg+xml"></object>
   <script>
   window.onload=function() { // Wait for the SVG to be loaded before changing it
     var elem=document.querySelector("#TOC").contentDocument.getElementById("SMPIBox")
     elem.style="opacity:0.93999999;fill:#ff0000;fill-opacity:0.1";
     elem.style="opacity:0.93999999;fill:#ff0000;fill-opacity:0.1;stroke:#000000;stroke-width:0.35277778;stroke-linecap:round;stroke-linejoin:round;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1";
   }
   </script>
   <br/>
   <br/>

SMPI enables the study of MPI application by emulating them on top of
the SimGrid simulator. This is particularly interesting to study
existing MPI applications within the comfort of the simulator.

To get started with SMPI, you should head to :ref:`the SMPI tutorial
<usecase_smpi>`. You may also want to read the `SMPI reference
article <https://hal.inria.fr/hal-01415484>`_ or these `introductory
slides <http://simgrid.org/tutorials/simgrid-smpi-101.pdf>`_.  If you
are new to MPI, you should first take our online `SMPI CourseWare
<https://simgrid.github.io/SMPI_CourseWare/>`_. It consists in several
projects that progressively introduce the MPI concepts. It proposes to
use SimGrid and SMPI to run the experiments, but the learning
objectives are centered on MPI itself.

Our goal is to enable the study of **unmodified MPI applications**.
Some constructs and features are still missing, but we can probably
add them on demand.  If you already used MPI before, SMPI should sound
very familiar to you: Use smpicc instead of mpicc, and smpirun instead
of mpirun. The main difference is that smpirun takes a :ref:`simulated
platform <platform>` as an extra parameter.

For **further scalability**, you may modify your code to speed up your
studies or save memory space.  Maximal **simulation accuracy**
requires some specific care from you.

.. _SMPI_online:

-----------------
Using SMPI online
-----------------

In this mode, your application is actually executed. Every computation
occurs for real while every communication is simulated. In addition,
the executions are automatically benchmarked so that their timings can
be applied within the simulator.

SMPI can also go offline by replaying a trace. :ref:`Trace replay
<SMPI_offline>` is usually ways faster than online simulation (because
the computation are skipped), but it can only applied to applications
with constant execution and communication patterns (for the exact same
reason).

...................
Compiling your Code
...................

If your application is in C, then simply use ``smpicc`` as a
compiler just like you use mpicc with other MPI implementations. This
script still calls your default compiler (gcc, clang, ...) and adds
the right compilation flags along the way. If your application is in
C++, Fortran 77 or Fortran 90, use respectively ``smpicxx``,
``smpiff`` or ``smpif90``.

If you use cmake, set the variables ``MPI_C_COMPILER``, ``MPI_CXX_COMPILER`` and
``MPI_Fortran_COMPILER`` to the full path of smpicc, smpicxx and smpiff (or
smpif90), respectively. Example:

.. code-block:: shell

   cmake -DMPI_C_COMPILER=/opt/simgrid/bin/smpicc -DMPI_CXX_COMPILER=/opt/simgrid/bin/smpicxx -DMPI_Fortran_COMPILER=/opt/simgrid/bin/smpiff .

....................
Simulating your Code
....................

Use the ``smpirun`` script as follows:

.. code-block:: shell

   smpirun -hostfile my_hostfile.txt -platform my_platform.xml ./program -blah

- ``my_hostfile.txt`` is a classical MPI hostfile (that is, this file
  lists the machines on which the processes must be dispatched, one
  per line)
- ``my_platform.xml`` is a classical SimGrid platform file. Of course,
  the hosts of the hostfile must exist in the provided platform.
- ``./program`` is the MPI program to simulate, that you compiled with ``smpicc``
- ``-blah`` is a command-line parameter passed to this program.

``smpirun`` accepts other parameters, such as ``-np`` if you don't
want to use all the hosts defined in the hostfile, ``-map`` to display
on which host each rank gets mapped of ``-trace`` to activate the
tracing during the simulation. You can get the full list by running
``smpirun -help``

...............................
Debugging your Code within SMPI
...............................

If you want to explore the automatic platform and deployment files
that are generated by ``smpirun``, add ``-keep-temps`` to the command
line.

You can also run your simulation within valgrind or gdb using the
following commands. Once in GDB, each MPI ranks will be represented as
a regular thread, and you can explore the state of each of them as
usual.

.. code-block:: shell

   smpirun -wrapper valgrind ...other args...
   smpirun -wrapper "gdb --args" --cfg=contexts/factory:thread ...other args...

.. _SMPI_use_colls:

................................
Simulating Collective Operations
................................

MPI collective operations are crucial to the performance of MPI
applications and must be carefully optimized according to many
parameters. Every existing implementation provides several algorithms
for each collective operation, and selects by default the best suited
one, depending on the sizes sent, the number of nodes, the
communicator, or the communication library being used.  These
decisions are based on empirical results and theoretical complexity
estimation, and are very different between MPI implementations. In
most cases, the users can also manually tune the algorithm used for
each collective operation.

SMPI can simulate the behavior of several MPI implementations:
OpenMPI, MPICH, `STAR-MPI <http://star-mpi.sourceforge.net/>`_, and
MVAPICH2. For that, it provides 115 collective algorithms and several
selector algorithms, that were collected directly in the source code
of the targeted MPI implementations.

You can switch the automatic selector through the
``smpi/coll-selector`` configuration item. Possible values:

 - **ompi:** default selection logic of OpenMPI (version 3.1.2)
 - **mpich**: default selection logic of MPICH (version 3.3b)
 - **mvapich2**: selection logic of MVAPICH2 (version 1.9) tuned
   on the Stampede cluster
 - **impi**: preliminary version of an Intel MPI selector (version
   4.1.3, also tuned for the Stampede cluster). Due the closed source
   nature of Intel MPI, some of the algorithms described in the
   documentation are not available, and are replaced by mvapich ones.
 - **default**: legacy algorithms used in the earlier days of
   SimGrid. Do not use for serious perform performance studies.

.. todo:: default should not even exist.

....................
Available Algorithms
....................

You can also pick the algorithm used for each collective with the
corresponding configuration item. For example, to use the pairwise
alltoall algorithm, one should add ``--cfg=smpi/alltoall:pair`` to the
line. This will override the selector (if any) for this algorithm.  It
means that the selected algorithm will be used

.. Warning:: Some collective may require specific conditions to be
   executed correctly (for instance having a communicator with a power
   of two number of nodes only), which are currently not enforced by
   Simgrid. Some crashes can be expected while trying these algorithms
   with unusual sizes/parameters

MPI_Alltoall
^^^^^^^^^^^^

Most of these are best described in `STAR-MPI's white paper <https://doi.org/10.1145/1183401.1183431>`_.

 - default: naive one, by default
 - ompi: use openmpi selector for the alltoall operations
 - mpich: use mpich selector for the alltoall operations
 - mvapich2: use mvapich2 selector for the alltoall operations
 - impi: use intel mpi selector for the alltoall operations
 - automatic (experimental): use an automatic self-benchmarking algorithm
 - bruck: Described by Bruck et.al. in <a href="http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=642949">this paper</a>
 - 2dmesh: organizes the nodes as a two dimensional mesh, and perform allgather
   along the dimensions
 - 3dmesh: adds a third dimension to the previous algorithm
 - rdb: recursive doubling: extends the mesh to a nth dimension, each one
   containing two nodes
 - pair: pairwise exchange, only works for power of 2 procs, size-1 steps,
   each process sends and receives from the same process at each step
 - pair_light_barrier: same, with small barriers between steps to avoid
   contention
 - pair_mpi_barrier: same, with MPI_Barrier used
 - pair_one_barrier: only one barrier at the beginning
 - ring: size-1 steps, at each step a process send to process (n+i)%size, and receives from (n-i)%size
 - ring_light_barrier: same, with small barriers between some phases to avoid contention
 - ring_mpi_barrier: same, with MPI_Barrier used
 - ring_one_barrier: only one barrier at the beginning
 - basic_linear: posts all receives and all sends,
   starts the communications, and waits for all communication to finish
 - mvapich2_scatter_dest: isend/irecv with scattered destinations, posting only a few messages at the same time

MPI_Alltoallv
^^^^^^^^^^^^^
 - default: naive one, by default
 - ompi: use openmpi selector for the alltoallv operations
 - mpich: use mpich selector for the alltoallv operations
 - mvapich2: use mvapich2 selector for the alltoallv operations
 - impi: use intel mpi selector for the alltoallv operations
 - automatic (experimental): use an automatic self-benchmarking algorithm
 - bruck: same as alltoall
 - pair: same as alltoall
 - pair_light_barrier: same as alltoall
 - pair_mpi_barrier: same as alltoall
 - pair_one_barrier: same as alltoall
 - ring: same as alltoall
 - ring_light_barrier: same as alltoall
 - ring_mpi_barrier: same as alltoall
 - ring_one_barrier: same as alltoall
 - ompi_basic_linear: same as alltoall

MPI_Gather
^^^^^^^^^^

 - default: naive one, by default
 - ompi: use openmpi selector for the gather operations
 - mpich: use mpich selector for the gather operations
 - mvapich2: use mvapich2 selector for the gather operations
 - impi: use intel mpi selector for the gather operations
 - automatic (experimental): use an automatic self-benchmarking algorithm which will iterate over all implemented versions and output the best
 - ompi_basic_linear: basic linear algorithm from openmpi, each process sends to the root
 - ompi_binomial: binomial tree algorithm
 - ompi_linear_sync: same as basic linear, but with a synchronization at the
   beginning and message cut into two segments.
 - mvapich2_two_level: SMP-aware version from MVAPICH. Gather first intra-node (defaults to mpich's gather), and then exchange with only one process/node. Use mvapich2 selector to change these to tuned algorithms for Stampede cluster.

MPI_Barrier
^^^^^^^^^^^

 - default: naive one, by default
 - ompi: use openmpi selector for the barrier operations
 - mpich: use mpich selector for the barrier operations
 - mvapich2: use mvapich2 selector for the barrier operations
 - impi: use intel mpi selector for the barrier operations
 - automatic (experimental): use an automatic self-benchmarking algorithm
 - ompi_basic_linear: all processes send to root
 - ompi_two_procs: special case for two processes
 - ompi_bruck: nsteps = sqrt(size), at each step, exchange data with rank-2^k and rank+2^k
 - ompi_recursivedoubling: recursive doubling algorithm
 - ompi_tree: recursive doubling type algorithm, with tree structure
 - ompi_doublering: double ring algorithm
 - mvapich2_pair: pairwise algorithm
 - mpich_smp: barrier intra-node, then inter-node

MPI_Scatter
^^^^^^^^^^^

 - default: naive one, by default
 - ompi: use openmpi selector for the scatter operations
 - mpich: use mpich selector for the scatter operations
 - mvapich2: use mvapich2 selector for the scatter operations
 - impi: use intel mpi selector for the scatter operations
 - automatic (experimental): use an automatic self-benchmarking algorithm
 - ompi_basic_linear: basic linear scatter
 - ompi_binomial: binomial tree scatter
 - mvapich2_two_level_direct: SMP aware algorithm, with an intra-node stage (default set to mpich selector), and then a basic linear inter node stage. Use mvapich2 selector to change these to tuned algorithms for Stampede cluster. 
 - mvapich2_two_level_binomial: SMP aware algorithm, with an intra-node stage (default set to mpich selector), and then a binomial phase. Use mvapich2 selector to change these to tuned algorithms for Stampede cluster.

MPI_Reduce
^^^^^^^^^^

 - default: naive one, by default
 - ompi: use openmpi selector for the reduce operations
 - mpich: use mpich selector for the reduce operations
 - mvapich2: use mvapich2 selector for the reduce operations
 - impi: use intel mpi selector for the reduce operations
 - automatic (experimental): use an automatic self-benchmarking algorithm
 - arrival_pattern_aware: root exchanges with the first process to arrive
 - binomial: uses a binomial tree
 - flat_tree: uses a flat tree
 - NTSL: Non-topology-specific pipelined linear-bcast function
   0->1, 1->2 ,2->3, ....., ->last node: in a pipeline fashion, with segments
   of 8192 bytes
 - scatter_gather: scatter then gather
 - ompi_chain: openmpi reduce algorithms are built on the same basis, but the
   topology is generated differently for each flavor
   chain = chain with spacing of size/2, and segment size of 64KB
 - ompi_pipeline: same with pipeline (chain with spacing of 1), segment size
   depends on the communicator size and the message size
 - ompi_binary: same with binary tree, segment size of 32KB
 - ompi_in_order_binary: same with binary tree, enforcing order on the
   operations
 - ompi_binomial: same with binomial algo (redundant with default binomial
   one in most cases)
 - ompi_basic_linear: basic algorithm, each process sends to root
 - mvapich2_knomial: k-nomial algorithm. Default factor is 4 (mvapich2 selector adapts it through tuning)
 - mvapich2_two_level: SMP-aware reduce, with default set to mpich both for intra and inter communicators. Use mvapich2 selector to change these to tuned algorithms for Stampede cluster.
 - rab: `Rabenseifner <https://fs.hlrs.de/projects/par/mpi//myreduce.html>`_'s reduce algorithm

MPI_Allreduce
^^^^^^^^^^^^^

 - default: naive one, by default
 - ompi: use openmpi selector for the allreduce operations
 - mpich: use mpich selector for the allreduce operations
 - mvapich2: use mvapich2 selector for the allreduce operations
 - impi: use intel mpi selector for the allreduce operations
 - automatic (experimental): use an automatic self-benchmarking algorithm
 - lr: logical ring reduce-scatter then logical ring allgather
 - rab1: variations of the  <a href="https://fs.hlrs.de/projects/par/mpi//myreduce.html">Rabenseifner</a> algorithm: reduce_scatter then allgather
 - rab2: variations of the  <a href="https://fs.hlrs.de/projects/par/mpi//myreduce.html">Rabenseifner</a> algorithm: alltoall then allgather
 - rab_rsag: variation of the  <a href="https://fs.hlrs.de/projects/par/mpi//myreduce.html">Rabenseifner</a> algorithm: recursive doubling
   reduce_scatter then recursive doubling allgather
 - rdb: recursive doubling
 - smp_binomial: binomial tree with smp: binomial intra
   SMP reduce, inter reduce, inter broadcast then intra broadcast
 - smp_binomial_pipeline: same with segment size = 4096 bytes
 - smp_rdb: intra: binomial allreduce, inter: Recursive
   doubling allreduce, intra: binomial broadcast
 - smp_rsag: intra: binomial allreduce, inter: reduce-scatter,
   inter:allgather, intra: binomial broadcast
 - smp_rsag_lr: intra: binomial allreduce, inter: logical ring
   reduce-scatter, logical ring inter:allgather, intra: binomial broadcast
 - smp_rsag_rab: intra: binomial allreduce, inter: rab
   reduce-scatter, rab inter:allgather, intra: binomial broadcast
 - redbcast: reduce then broadcast, using default or tuned algorithms if specified
 - ompi_ring_segmented: ring algorithm used by OpenMPI
 - mvapich2_rs: rdb for small messages, reduce-scatter then allgather else
 - mvapich2_two_level: SMP-aware algorithm, with mpich as intra algorithm, and rdb as inter (Change this behavior by using mvapich2 selector to use tuned values)
 - rab: default `Rabenseifner <https://fs.hlrs.de/projects/par/mpi//myreduce.html>`_ implementation

MPI_Reduce_scatter
^^^^^^^^^^^^^^^^^^

 - default: naive one, by default
 - ompi: use openmpi selector for the reduce_scatter operations
 - mpich: use mpich selector for the reduce_scatter operations
 - mvapich2: use mvapich2 selector for the reduce_scatter operations
 - impi: use intel mpi selector for the reduce_scatter operations
 - automatic (experimental): use an automatic self-benchmarking algorithm
 - ompi_basic_recursivehalving: recursive halving version from OpenMPI
 - ompi_ring: ring version from OpenMPI
 - mpich_pair: pairwise exchange version from MPICH
 - mpich_rdb: recursive doubling version from MPICH
 - mpich_noncomm: only works for power of 2 procs, recursive doubling for noncommutative ops


MPI_Allgather
^^^^^^^^^^^^^

 - default: naive one, by default
 - ompi: use openmpi selector for the allgather operations
 - mpich: use mpich selector for the allgather operations
 - mvapich2: use mvapich2 selector for the allgather operations
 - impi: use intel mpi selector for the allgather operations
 - automatic (experimental): use an automatic self-benchmarking algorithm
 - 2dmesh: see alltoall
 - 3dmesh: see alltoall
 - bruck: Described by Bruck et.al. in <a href="http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=642949">
   Efficient algorithms for all-to-all communications in multiport message-passing systems</a>
 - GB: Gather - Broadcast (uses tuned version if specified)
 - loosely_lr: Logical Ring with grouping by core (hardcoded, default
   processes/node: 4)
 - NTSLR: Non Topology Specific Logical Ring
 - NTSLR_NB: Non Topology Specific Logical Ring, Non Blocking operations
 - pair: see alltoall
 - rdb: see alltoall
 - rhv: only power of 2 number of processes
 - ring: see alltoall
 - SMP_NTS: gather to root of each SMP, then every root of each SMP node
   post INTER-SMP Sendrecv, then do INTRA-SMP Bcast for each receiving message,
   using logical ring algorithm (hardcoded, default processes/SMP: 8)
 - smp_simple: gather to root of each SMP, then every root of each SMP node
   post INTER-SMP Sendrecv, then do INTRA-SMP Bcast for each receiving message,
   using simple algorithm (hardcoded, default processes/SMP: 8)
 - spreading_simple: from node i, order of communications is i -> i + 1, i ->
   i + 2, ..., i -> (i + p -1) % P
 - ompi_neighborexchange: Neighbor Exchange algorithm for allgather.
   Described by Chen et.al. in  `Performance Evaluation of Allgather
   Algorithms on Terascale Linux Cluster with Fast Ethernet <http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnumber=1592302>`_
 - mvapich2_smp: SMP aware algorithm, performing intra-node gather, inter-node allgather with one process/node, and bcast intra-node

MPI_Allgatherv
^^^^^^^^^^^^^^

 - default: naive one, by default
 - ompi: use openmpi selector for the allgatherv operations
 - mpich: use mpich selector for the allgatherv operations
 - mvapich2: use mvapich2 selector for the allgatherv operations
 - impi: use intel mpi selector for the allgatherv operations
 - automatic (experimental): use an automatic self-benchmarking algorithm
 - GB: Gatherv - Broadcast (uses tuned version if specified, but only for Bcast, gatherv is not tuned)
 - pair: see alltoall
 - ring: see alltoall
 - ompi_neighborexchange: see allgather
 - ompi_bruck: see allgather
 - mpich_rdb: recursive doubling algorithm from MPICH
 - mpich_ring: ring algorithm from MPICh - performs differently from the  one from STAR-MPI

MPI_Bcast
^^^^^^^^^

 - default: naive one, by default
 - ompi: use openmpi selector for the bcast operations
 - mpich: use mpich selector for the bcast operations
 - mvapich2: use mvapich2 selector for the bcast operations
 - impi: use intel mpi selector for the bcast operations
 - automatic (experimental): use an automatic self-benchmarking algorithm
 - arrival_pattern_aware: root exchanges with the first process to arrive
 - arrival_pattern_aware_wait: same with slight variation
 - binomial_tree: binomial tree exchange
 - flattree: flat tree exchange
 - flattree_pipeline: flat tree exchange, message split into 8192 bytes pieces
 - NTSB: Non-topology-specific pipelined binary tree with 8192 bytes pieces
 - NTSL: Non-topology-specific pipelined linear with 8192 bytes pieces
 - NTSL_Isend: Non-topology-specific pipelined linear with 8192 bytes pieces, asynchronous communications
 - scatter_LR_allgather: scatter followed by logical ring allgather
 - scatter_rdb_allgather: scatter followed by recursive doubling allgather
 - arrival_scatter: arrival pattern aware scatter-allgather
 - SMP_binary: binary tree algorithm with 8 cores/SMP
 - SMP_binomial: binomial tree algorithm with 8 cores/SMP
 - SMP_linear: linear algorithm with 8 cores/SMP
 - ompi_split_bintree: binary tree algorithm from OpenMPI, with message split in 8192 bytes pieces
 - ompi_pipeline: pipeline algorithm from OpenMPI, with message split in 128KB pieces
 - mvapich2_inter_node: Inter node default mvapich worker
 - mvapich2_intra_node: Intra node default mvapich worker
 - mvapich2_knomial_intra_node:  k-nomial intra node default mvapich worker. default factor is 4.

Automatic Evaluation
^^^^^^^^^^^^^^^^^^^^

.. warning:: This is still very experimental.

An automatic version is available for each collective (or even as a selector). This specific
version will loop over all other implemented algorithm for this particular collective, and apply
them while benchmarking the time taken for each process. It will then output the quickest for
each process, and the global quickest. This is still unstable, and a few algorithms which need
specific number of nodes may crash.

Adding an algorithm
^^^^^^^^^^^^^^^^^^^

To add a new algorithm, one should check in the src/smpi/colls folder
how other algorithms are coded. Using plain MPI code inside Simgrid
can't be done, so algorithms have to be changed to use smpi version of
the calls instead (MPI_Send will become smpi_mpi_send). Some functions
may have different signatures than their MPI counterpart, please check
the other algorithms or contact us using the `>SimGrid
developers mailing list <http://lists.gforge.inria.fr/mailman/listinfo/simgrid-devel>`_.

Example: adding a "pair" version of the Alltoall collective.

 - Implement it in a file called alltoall-pair.c in the src/smpi/colls folder. This file should include colls_private.hpp.

 - The name of the new algorithm function should be smpi_coll_tuned_alltoall_pair, with the same signature as MPI_Alltoall.

 - Once the adaptation to SMPI code is done, add a reference to the file ("src/smpi/colls/alltoall-pair.c") in the SMPI_SRC part of the DefinePackages.cmake file inside buildtools/cmake, to allow the file to be built and distributed.

 - To register the new version of the algorithm, simply add a line to the corresponding macro in src/smpi/colls/cools.h ( add a "COLL_APPLY(action, COLL_ALLTOALL_SIG, pair)" to the COLL_ALLTOALLS macro ). The algorithm should now be compiled and be selected when using --cfg=smpi/alltoall:pair at runtime.

 - To add a test for the algorithm inside Simgrid's test suite, juste add the new algorithm name in the ALLTOALL_COLL list found inside teshsuite/smpi/CMakeLists.txt . When running ctest, a test for the new algorithm should be generated and executed. If it does not pass, please check your code or contact us.

 - Please submit your patch for inclusion in SMPI, for example through a pull request on GitHub or directly per email.


Tracing of Internal Communications
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

By default, the collective operations are traced as a unique operation
because tracing all point-to-point communications composing them could
result in overloaded, hard to interpret traces. If you want to debug
and compare collective algorithms, you should set the
``tracing/smpi/internals`` configuration item to 1 instead of 0.

Here are examples of two alltoall collective algorithms runs on 16 nodes,
the first one with a ring algorithm, the second with a pairwise one.

.. image:: /img/smpi_simgrid_alltoall_ring_16.png
   :align: center

Alltoall on 16 Nodes with the Ring Algorithm.

.. image:: /img/smpi_simgrid_alltoall_pair_16.png
   :align: center

Alltoall on 16 Nodes with the Pairwise Algorithm.

-------------------------
What can run within SMPI?
-------------------------

You can run unmodified MPI applications (both C/C++ and Fortran) within
SMPI, provided that you only use MPI calls that we implemented. Global
variables should be handled correctly on Linux systems.

....................
MPI coverage of SMPI
....................

Our coverage of the interface is very decent, but still incomplete;
Given the size of the MPI standard, we may well never manage to
implement absolutely all existing primitives. Currently, we have
almost no support for I/O primitives, but we still pass a very large
amount of the MPICH coverage tests.

The full list of not yet implemented functions is documented in the
file `include/smpi/smpi.h
<https://framagit.org/simgrid/simgrid/tree/master/include/smpi/smpi.h>`_
in your version of SimGrid, between two lines containing the ``FIXME``
marker. If you really miss a feature, please get in touch with us: we
can guide you though the SimGrid code to help you implementing it, and
we'd be glad to integrate your contribution to the main project.

.. _SMPI_what_globals:

.................................
Privatization of global variables
.................................

Concerning the globals, the problem comes from the fact that usually,
MPI processes run as real UNIX processes while they are all folded
into threads of a unique system process in SMPI. Global variables are
usually private to each MPI process while they become shared between
the processes in SMPI.  The problem and some potential solutions are
discussed in this article: `Automatic Handling of Global Variables for
Multi-threaded MPI Programs
<http://charm.cs.illinois.edu/newPapers/11-23/paper.pdf>` (note that
this article does not deal with SMPI but with a competing solution
called AMPI that suffers of the same issue).  This point used to be
problematic in SimGrid, but the problem should now be handled
automatically on Linux.

Older versions of SimGrid came with a script that automatically
privatized the globals through static analysis of the source code. But
our implementation was not robust enough to be used in production, so
it was removed at some point. Currently, SMPI comes with two
privatization mechanisms that you can :ref:`select at runtime
<cfg=smpi/privatization>`.  The dlopen approach is used by
default as it is much faster and still very robust.  The mmap approach
is an older approach that proves to be slower.

With the **mmap approach**, SMPI duplicates and dynamically switch the
``.data`` and ``.bss`` segments of the ELF process when switching the
MPI ranks. This allows each ranks to have its own copy of the global
variables.  No copy actually occurs as this mechanism uses ``mmap()``
for efficiency. This mechanism is considered to be very robust on all
systems supporting ``mmap()`` (Linux and most BSDs). Its performance
is questionable since each context switch between MPI ranks induces
several syscalls to change the ``mmap`` that redirects the ``.data``
and ``.bss`` segments to the copies of the new rank. The code will
also be copied several times in memory, inducing a slight increase of
memory occupation.

Another limitation is that SMPI only accounts for global variables
defined in the executable. If the processes use external global
variables from dynamic libraries, they won't be switched
correctly. The easiest way to solve this is to statically link against
the library with these globals. This way, each MPI rank will get its
own copy of these libraries. Of course you should never statically
link against the SimGrid library itself.

With the **dlopen approach**, SMPI loads several copies of the same
executable in memory as if it were a library, so that the global
variables get naturally duplicated. It first requires the executable
to be compiled as a relocatable binary, which is less common for
programs than for libraries. But most distributions are now compiled
this way for security reason as it allows one to randomize the address
space layout. It should thus be safe to compile most (any?) program
this way.  The second trick is that the dynamic linker refuses to link
the exact same file several times, be it a library or a relocatable
executable. It makes perfectly sense in the general case, but we need
to circumvent this rule of thumb in our case. To that extend, the
binary is copied in a temporary file before being re-linked against.
``dlmopen()`` cannot be used as it only allows 256 contextes, and as it
would also duplicate simgrid itself.

This approach greatly speeds up the context switching, down to about
40 CPU cycles with our raw contextes, instead of requesting several
syscalls with the ``mmap()`` approach. Another advantage is that it
permits one to run the SMPI contexts in parallel, which is obviously not
possible with the ``mmap()`` approach. It was tricky to implement, but
we are not aware of any flaws, so smpirun activates it by default.

In the future, it may be possible to further reduce the memory and
disk consumption. It seems that we could `punch holes
<https://lwn.net/Articles/415889/>`_ in the files before dl-loading
them to remove the code and constants, and mmap these area onto a
unique copy. If done correctly, this would reduce the disk- and
memory- usage to the bare minimum, and would also reduce the pressure
on the CPU instruction cache. See the `relevant bug
<https://github.com/simgrid/simgrid/issues/137>`_ on github for
implementation leads.\n

Also, currently, only the binary is copied and dlopen-ed for each MPI
rank. We could probably extend this to external dependencies, but for
now, any external dependencies must be statically linked into your
application. As usual, simgrid itself shall never be statically linked
in your app. You don't want to give a copy of SimGrid to each MPI rank:
that's ways too much for them to deal with.

.. todo: speak of smpi/privatize-libs here

----------------------------------------------
Adapting your MPI code for further scalability
----------------------------------------------

As detailed in the `reference article
<http://hal.inria.fr/hal-01415484>`_, you may want to adapt your code
to improve the simulation performance. But these tricks may seriously
hinder the result quality (or even prevent the app to run) if used
wrongly. We assume that if you want to simulate an HPC application,
you know what you are doing. Don't prove us wrong!

..............................
Reducing your memory footprint
..............................

If you get short on memory (the whole app is executed on a single node when
simulated), you should have a look at the SMPI_SHARED_MALLOC and
SMPI_SHARED_FREE macros. It allows one to share memory areas between processes: The
purpose of these macro is that the same line malloc on each process will point
to the exact same memory area. So if you have a malloc of 2M and you have 16
processes, this macro will change your memory consumption from 2M*16 to 2M
only. Only one block for all processes.

If your program is ok with a block containing garbage value because all
processes write and read to the same place without any kind of coordination,
then this macro can dramatically shrink your memory consumption. For example,
that will be very beneficial to a matrix multiplication code, as all blocks will
be stored on the same area. Of course, the resulting computations will useless,
but you can still study the application behavior this way.

Naturally, this won't work if your code is data-dependent. For example, a Jacobi
iterative computation depends on the result computed by the code to detect
convergence conditions, so turning them into garbage by sharing the same memory
area between processes does not seem very wise. You cannot use the
SMPI_SHARED_MALLOC macro in this case, sorry.

This feature is demoed by the example file
`examples/smpi/NAS/dt.c <https://framagit.org/simgrid/simgrid/tree/master/examples/smpi/NAS/dt.c>`_

.. _SMPI_use_faster:

.........................
Toward Faster Simulations
.........................

If your application is too slow, try using SMPI_SAMPLE_LOCAL,
SMPI_SAMPLE_GLOBAL and friends to indicate which computation loops can
be sampled. Some of the loop iterations will be executed to measure
their duration, and this duration will be used for the subsequent
iterations. These samples are done per processor with
SMPI_SAMPLE_LOCAL, and shared between all processors with
SMPI_SAMPLE_GLOBAL. Of course, none of this will work if the execution
time of your loop iteration are not stable.

This feature is demoed by the example file
`examples/smpi/NAS/ep.c <https://framagit.org/simgrid/simgrid/tree/master/examples/smpi/NAS/ep.c>`_

.............................
Ensuring Accurate Simulations
.............................

Out of the box, SimGrid may give you fairly accurate results, but
there is a plenty of factors that could go wrong and make your results
inaccurate or even plainly wrong. Actually, you can only get accurate
results of a nicely built model, including both the system hardware
and your application. Such models are hard to pass over and reuse in
other settings, because elements that are not relevant to an
application (say, the latency of point-to-point communications,
collective operation implementation details or CPU-network
interaction) may be irrelevant to another application. The dream of
the perfect model, encompassing every aspects is only a chimera, as
the only perfect model of the reality is the reality. If you go for
simulation, then you have to ignore some irrelevant aspects of the
reality, but which aspects are irrelevant is actually
application-dependent...

The only way to assess whether your settings provide accurate results
is to double-check these results. If possible, you should first run
the same experiment in simulation and in real life, gathering as much
information as you can. Try to understand the discrepancies in the
results that you observe between both settings (visualization can be
precious for that). Then, try to modify your model (of the platform,
of the collective operations) to reduce the most preeminent differences.

If the discrepancies come from the computing time, try adapting the
``smpi/host-speed``: reduce it if your simulation runs faster than in
reality. If the error come from the communication, then you need to
fiddle with your platform file.

Be inventive in your modeling. Don't be afraid if the names given by
SimGrid does not match the real names: we got very good results by
modeling multicore/GPU machines with a set of separate hosts
interconnected with very fast networks (but don't trust your model
because it has the right names in the right place either).

Finally, you may want to check `this article
<https://hal.inria.fr/hal-00907887>`_ on the classical pitfalls in
modeling distributed systems.

-------------------------
Troubleshooting with SMPI
-------------------------

.................................
./configure refuses to use smpicc
.................................

If your ``./configure`` reports that the compiler is not
functional or that you are cross-compiling, try to define the
``SMPI_PRETEND_CC`` environment variable before running the
configuration.

.. code-block:: shell

   SMPI_PRETEND_CC=1 ./configure # here come the configure parameters
   make

Indeed, the programs compiled with ``smpicc`` cannot be executed
without ``smpirun`` (they are shared libraries and do weird things on
startup), while configure wants to test them directly.  With
``SMPI_PRETEND_CC`` smpicc does not compile as shared, and the SMPI
initialization stops and returns 0 before doing anything that would
fail without ``smpirun``.

.. warning::

  Make sure that SMPI_PRETEND_CC is only set when calling ./configure,
  not during the actual execution, or any program compiled with smpicc
  will stop before starting.

..............................................
./configure does not pick smpicc as a compiler
..............................................

In addition to the previous answers, some projects also need to be
explicitly told what compiler to use, as follows:

.. code-block:: shell

   SMPI_PRETEND_CC=1 ./configure CC=smpicc # here come the other configure parameters
   make

Maybe your configure is using another variable, such as ``cc`` (in
lower case) or similar. Just check the logs.

.....................................
error: unknown type name 'useconds_t'
.....................................

Try to add ``-D_GNU_SOURCE`` to your compilation line to get ride
of that error.

The reason is that SMPI provides its own version of ``usleep(3)``
to override it and to block in the simulation world, not in the real
one. It needs the ``useconds_t`` type for that, which is declared
only if you declare ``_GNU_SOURCE`` before including
``unistd.h``. If your project includes that header file before
SMPI, then you need to ensure that you pass the right configuration
defines as advised above.



.. _SMPI_offline:

-----------------------------
Trace Replay and Offline SMPI
-----------------------------

Although SMPI is often used for :ref:`online simulation
<SMPI_online>`, where the application is executed for real, you can
also go for offline simulation through trace replay.

SimGrid uses time-independent traces, in which each actor is given a
script of the actions to do sequentially. These trace files can
actually be captured with the online version of SMPI, as follows:

.. code-block:: shell

   $ smpirun -trace-ti --cfg=tracing/filename:LU.A.32 -np 32 -platform ../cluster_backbone.xml bin/lu.A.32

The produced trace is composed of a file ``LU.A.32`` and a folder
``LU.A.32_files``. The file names don't match with the MPI ranks, but
that's expected.

To replay this with SMPI, you need to first compile the provided
``smpi_replay.cpp`` file, that comes from
`simgrid/examples/smpi/replay
<https://framagit.org/simgrid/simgrid/tree/master/examples/smpi/replay>`_.

.. code-block:: shell

   $ smpicxx ../replay.cpp -O3 -o ../smpi_replay

Afterward, you can replay your trace in SMPI as follows:

   $ smpirun -np 32 -platform ../cluster_torus.xml -ext smpi_replay ../smpi_replay LU.A.32

All the outputs are gone, as the application is not really simulated
here. Its trace is simply replayed. But if you visualize the live
simulation and the replay, you will see that the behavior is
unchanged. The simulation does not run much faster on this very
example, but this becomes very interesting when your application
is computationally hungry.