Logo AND Algorithmique Numérique Distribuée

Public GIT Repository
On the fly parsing of the actions file. One idle process can still force the whole...
[simgrid.git] / src / smpi /
2010-01-28 piniBetter fix for r7026.
2010-01-27 piniFix errors with relative includes with smpicc
2010-01-27 piniIn smpi_mpi.c:
2010-01-20 piniIn smpi_mpi.c:
2010-01-19 piniAdded flag for printing more debug info
2010-01-19 piniFix stripping of first argument.
2009-12-11 mquinsonallow smpicc to survive to non-functionnal readlink...
2009-10-05 mquinsonImprovements to the MSG port on top of SIMIX network...
2009-08-17 mquinsonSMPI: Fix the prototype of SMPI_MPI_Type_get_extent...
2009-08-01 genaudAlltoallv
2009-07-31 genaudstarted alltoallv(). Still some bugs.
2009-07-30 genaudalltoall implemented (almost opmpi algorithms)
2009-07-30 genaudthought this one would fix my long-standing wait_all...
2009-07-21 genaud* minimum of datatype handling for alltoall tuned
2009-07-20 genaud... untested ...
2009-07-16 genaud* MPI_Sendrecv user+internal levels
2009-07-16 genaudupdate test
2009-07-16 genaudseparate user-level and internal-level for SMPI_MPI_...
2009-07-15 mquinsonYet more debug
2009-07-15 mquinsontodo++
2009-07-15 mquinsondo not call smpi_bench_* from SMPI_MPI_Allreduce since...
2009-07-15 mquinsonMore debug
2009-07-15 mquinsonOnly sleep to handle real computation if the time to...
2009-07-15 genaudexamples to demonstrate bugs in sendrecv
2009-07-13 genaud+ MPI_Sendrecv()
2009-07-13 mquinsonsimdata cleanup from smx_action_t. Now all SIMIX's...
2009-07-10 genaud MPI_Scatter() ok
2009-07-09 genaud- MPI_Scatter() /* untested */
2009-07-07 genaudheader to explain
2009-07-07 genaudooops. Forgot this new one.
2009-07-03 genaud- MPI_Barrier() as a collective
2009-07-02 genaud* added support for optimized collectives:
2009-07-01 genaudadaptations for MacOSX:
2009-06-30 mquinsonfurther polishing on the merge of all model types ...
2009-06-30 genaudMPI_Allreduce()
2009-06-30 mquinsonSURF: Unify the types of models in a uniq s_surf_model_...
2009-06-30 genaud/home/genaud/Documents/svn/simgrid/simgrid/trunk/src...
2009-06-30 mquinsonPlug some easy memleaks
2009-06-30 mquinsonMore debug
2009-06-30 genaudMPI_Reduce() is ok except memleak
2009-06-30 genaud- corrected test on -map
2009-06-29 genaudstill some mem cleanup to finish
2009-06-29 genaudMPI_Reduce using waitany(). Buggy.
2009-06-29 mquinsonMake Waitall and Waitany visible from user side (and...
2009-06-29 mquinsonImplement MPI_Waitany and MPI_Waitall
2009-06-27 genaudsome preliminary additions to implement more collectives
2009-06-26 mquinsonuse readlink instead of realpath because it's more...
2009-06-26 mquinsonSMPI: Add an option to display the timing at the end...
2009-06-26 mquinsonAdd a flag to keep the temp files around, and pass...
2009-06-26 mquinsonSMPI: Change the reference speed to a command line...
2009-06-26 genaudadded MPI_Wtime()
2009-06-25 mquinsonThe examples moved, and speak of LD_LIBRARY_PATH
2009-06-25 mquinsonGet smpirun displaying the real process rank
2009-06-25 mquinsonSet in release mode: remove temps and less debug
2009-06-25 mquinsonCompile about every existing SMPI examples (reduce...
2009-06-25 genaudmore in README
2009-06-25 mquinsonSMPI: Use the amount of processes to malloc the process...
2009-06-25 mquinsonChange host data into process data; count processes...
2009-06-24 genaudcheck that specified platform and hostfile exist.
2009-06-24 mquinsonSMPI: Pass the rank from smpirun to processes through...
2009-06-24 mquinsonSMPI: kill dead code
2009-06-24 mquinsonSMPI: move the last queue (received_message_queue)...
2009-06-24 mquinsonSMPI: move pending_send_request_queue from global to...
2009-06-24 mquinsonSMPI: kill some free() from the smpi_process_finalize...
2009-06-24 mquinsonSMPI: Change the way senders and receivers are stopped...
2009-06-24 mquinsonSMPI: Kill the global list of senders and receivers
2009-06-24 mquinsonSMPI: Add a global table of processes, which one day...
2009-06-24 mquinsonMove pending_recv_request_queues into host_data (ie...
2009-06-23 mquinsonSMPI: get the main process launch the sender and receiv...
2009-06-23 mquinsoncosmetics (use xbt_fifo_foreach where relevant)
2009-06-23 mquinsoncosmetics (function renaming)
2009-06-23 mquinsonKill another useless synchronization source: no one...
2009-06-23 mquinsonSMPI: Kill a whole bunch of unneeded synchronization...
2009-06-23 mquinsonSMPI: Remove the initialization barrier now that it...
2009-06-23 mquinsonSMPI: cleanup the initialization process by moving...
2009-06-23 genaudquick notes to get started compiling an smpi example.
2009-06-23 genaudextended smpirun accepting -platform and -hostfile
2009-06-23 mquinsonBug fix: add the compiler name to the command line
2009-05-26 mquinsonReindent everything (possibly breaking all branches...
2008-09-28 marklscleaned up mpi_comm_split code some more.
2008-09-26 marklsjust changed some comments.
2008-09-26 marklschange spmirun and smpicc to use bash instead of sh...
2008-09-26 marklsminor changes to comm_split code.
2008-05-15 mquinsonRemove a crude hack where first call to surf_solve...
2008-04-25 marklsfixed bugs in SMPI_MPI_Comm_split and added back SMPI_M...
2008-03-07 mquinsoncosmetics
2008-03-07 marklsmoved XBT_LOG_CONNECT calls into smpi_global so that...
2008-03-06 marklsfixed nasty border condition bug.
2008-03-06 mquinsonConnect correctly the log channels manually for windows...
2008-03-06 mquinsonDo connect all log channel manually to parent using...
2008-02-28 marklsadded simple usage message to smpirun and removed nonwo...
2007-12-24 marklssome of the action checking code didn't work right...
2007-12-23 marklsaccidently tried to compare action to state. now fixed.
2007-12-23 marklsblah
2007-11-30 marklsI just realized that the changes I made to the cond...
2007-11-30 marklschaned global execute_mutex and execute_cond to host...
2007-11-27 marklsrenamed MPI_ functions to avoid namespace collisions...
2007-11-12 marklsrenamed DO_ONCE to SMPI_DO_ONCE.
2007-11-12 marklsmade some changes to allow more than one DO_ONCE block...
2007-11-10 marklscleaning up and refactoring some of the code to create...
next