*****************************************
*DO NOT MIX 3.27 RESULTS WITH OLDER ONES*
*****************************************
- The order in which the different models (CPU, network, disk, etc) are
- solved internally has changed in this version. Although rare (no current
- test raises the problem), this change can result in slightest different simulations
- results. Take care when comparing simulations from different SimGrid's
- versions. Sorry for the inconvenience.
+ The order in which the different models (CPU, network, disk, etc) are
+ solved internally has changed in this version. Although rare (no current
+ test raises the problem), this change can result in slightest different simulations
+ results. Take care when comparing simulations from different SimGrid's
+ versions. Sorry for the inconvenience.
+ - SMPI API break from previous version: MPI global handles (predefined datatypes,
+ ops, COMM_WORLD) are now defined differently, and can be used in initializers,
+ avoiding "initializer element is not constant" errors with some codes. Please be
+ careful to recompile your MPI codes if switching to this new SimGrid version.
S4U:
- Define new template functions Mailbox::get_unique(), returning a std::unique_ptr.
xbt_str_split_quoted(), xbt_str_split_quoted_in_place().
- Deprecate xbt_dynar_to_array().
+SMPI:
+ - Handle leak checking (experimental): SMPI will now display a message if MPI
+ handles are still present at the end of a simulation. This may be nothing,
+ but if the number is high it could be a problem. To display more information
+ about the type of leaked handle, option --cfg=smpi/list-leaks:n. will display
+ the n first leaked pointers and their types. Known issue: MPI_Cancel usage
+ may trigger handle leaks in SMPI.
+ Using smpirun with -wrapper "valgrind --leak-check=full --track-origins=yes"
+ can provide more information to help locate precisely the leaks.
+ - Memory allocations analysis (experimental): MPI codes using malloc/calloc to
+ allocate memory, and not compiled with SMPI_NO_OVERRIDE_MALLOC, can use
+ smpi/display-analysis option to show at the end of execution the amount of
+ memory allocated through these calls, and locate the most expensive one.
+ This helps finding the targets for manual memory sharing, or the threshold
+ to use for smpi/auto-shared-malloc-thresh option.
+ - New implemented MPI calls: MPI_File_set_size, MPI_Aint_add, MPI_Aint_diff,
+ MPI_Type_get_extent_x, MPI_Type_get_true_extent_x, MPI_Status_set_elements_x
+ - Misc fixes :
+ - better handling of MPI_PROC_NULL
+ - MPI_Group_incl was broken in some cases
+ - PAPI tracing was fixed
+ - MPI_Errhandlers are properly handled for MPI_COMM_WORLD
+ - SMPI will now fail properly when receive size < sent size, instead of
+ just setting MPI_ERR_TRUNCATE and keep going.
+ - Proxy Apps ( https://framagit.org/simgrid/SMPI-proxy-apps ):
+ - Remove useless build patches and split build/run patches on the page
+ - Build in parallel when possible
+ - Test on yum-based systems
+ - New nightly tested Proxy-apps:
+ - OpenMPI mpi-test-suite
+ - Coral MCB, LSMS, Qbox
+ - ECP/Mantevo Cloverleaf3D
+
C binding and interface:
- Rename (lowercase) functions sg_actor_get_PID, sg_actor_get_PPID, sg_actor_by_PID.
