Change Summary for MPICH2 for week of Jul 29, 2007

Information on all weekly change summaries is available.

List of revisions

• (jayesh) 1) Changes related to generalized requests - Async IO on windows (2) Thread safe IO error messages on windows
• (gropp) Add code to try and determine the Fortran name mangling when C cannot be used as the linking language (e.g., when using the PG Fortran 90 compiler as Fortran)
• (robl) add two aggregation tests, exercising the "force cb_writes" path
• (chan) added MPE changes
• (chan) added MPE_KIND_MSG_INIT into allowed_mask, so MPI_Send_init()/MPI_Recv_init + MPI_Startall() will work on persistent requests
• (balaji) Added a mention for the memory leaks.
• (balaji) Added a union to ensure alignment of data payload.
• (gropp) Added additional check to connection private data to avoid SEGV during disconnect
• (chan) added an OpenMP version of pthread_sendrecv.c
• (gropp) Added comment for eventual hook for using libatomic
• (chan) added cross-compilation instruction for XT4(jaguar) catamount
• (balaji) Added datatype and MPI_Gather changes to the CHANGES file.
• (gropp) Added include <string.h>
• (gropp) Added missing prototypes
• (jayesh) Added section on Performance Analysis using MPE
• (gropp) Added some interfaces for external routines, required by picky Fortran 90 compilers
• (gropp) Added some of the mpi-2 debugger information
• (gropp) Added temp stub for pmi2, this probably should be completely rewritten, but must be checked in before the move to svn
• (gropp) Added the second test program from Mathworks
• (thakur) changed version number, updates to readme and release notes
• (rbutler) Check existence of key before using.
• (robl) clean up small amount of leaked memory. reported by Lisandro Dalcin(MPICH2 req #3570)
• (chan) corrected typo pointed out by Rajeev.
• (ganesh) First cut implementation of the Infinipath PSM module within nemesis
• (balaji) Fix for alignment errors on 32-bit systems. We need a better fix for this.
• (gropp) fix for pg_destroy after finalize
• (gropp) Fix some errors in setting the include paths for ssm builddll (we need a better system for this)
• (gropp) fix typo
• (thakur) fixed LaTeX bugs
• (balaji) Fixed a bug in the calculation in the amount of space left.
• (thakur) fixed small typo
• (buntinas) fixed x86 high resolution timer issue re rdtsc serialization. Configure now tries rctscp, then cpuid;rdtsc and finally rdtsc.
• (thakur) further minor updates
• (robl) give up on NFS asynchronous I/O operations
• (gropp) Improve Dims_create
• (gropp) initinfo has no PMPI versions and is ponly part of the libmpich, not libpmpich
• (robl) just dropped in a comment so i wouldn't forget an idea
• (gropp) Keep the ps check from reporting non numeric pids (that means that it failed to properly parse the ps output)
• (jayesh) Latex formatting update
• (penoff) MPICH2 1.0.6 ch3:sctp documentation updates.
• (jayesh) MPICH2_VERSION & MPICH2_NUMVERSION now supported on windows
• (balaji) MPID_nem_cell_t and MPID_nem_abs_cell_t weren't synchronized causing some of the nightly tests to fail.
• (chan) made openmp_sendrecv.c and pthread_sendrecv.c comparable
• (gropp) Make sure that special f90 type routines are included in the shared library version of libmpich.so
• (buntinas) Minor changes to README re nemesis
• (thakur) minor language and style updates
• (jayesh) Modified section names, added location of cpilog.c
• (ganesh) Modifying encode_buffer and decode_buffer methods to make them non-static so that they can be used within other sections. (Used withing PSM module)
• (gropp) More fixes for Fortran name mangling detection
• (balaji) More optimizations to the vector copy code.
• (gropp) Pick the first channel name as the default to load (instead of always using sock)
• (gropp) Provide more details about the reason for failing when the F77 and F90 compilers are judged incompatible
• (gropp) Provide prototypes for the param-from-env functions
• (robl) quiet "may be used uninitialized" warning
• (balaji) Reduced the number of iterations of sendflood. Nemesis takes too long to finish with overprovisioning.
• (jayesh) Removed references to unavailable sections and features
• (jayesh) Removing developer libs from x32 installable
• (balaji) Reordered the elements of a structure to maintain alignment.
• (balaji) Results seem to be very noisy. Trying increasing the number of elements in the vector to reduce the percentage variation.
• (robl) revert the John Bent fix from March 2007, which broke a different accesspattern. Commit this version, which seems to do the right thing, and even if anew situation comes up where it does the wrong thing, will still get the rightbehavior.
• (ganesh) Reverting back to static methods for now. Will need to convert them to non-static methods eventually.
• (balaji) Reverting the alignment issue for nemesis. We need a better solution for this.
• (balaji) The configure help message for selecting the process manager mentions forker; should be gforker.
• (jayesh) Update version & Author list
• (thakur) updated
• (robl) use the MPICH2 error messages correctly
• (robl) we weren't setting status correctly with generalized requests. rectify.

Files for each revision