Dave and Pavan,<br><br><b>1)</b><br>The most subtle problem I found related to the use of MPID_Segment_pack by MPIR_Localcopy in src/mpi/coll/helper_fns.c. It passes the variable "last" of type "MPIDI_msg_sz_t" as "&last" to the pack/unpack functions, which are expecting an MPI_Aint*. On our system, <b>MPIDI_msg_sz_t is an unsigned</b>, and <b>MPI_Aint is a long long</b>. This means that MPID_Segment_pack clobbers too much memory. To fix this, I changed "last" to be an MPI_Aint, and it works fine. I think that means that<br>
A) "last" should be an MPI_Aint, or<br>B) Segment_*pack() should take a MPIDI_msg_sz_t*, or<br>C) MPIDI_msg_sz_t should be the same as MPI_Aint.<br><br>I don't really know what the answer is. However, I caught this thanks to a compiler warning, and our other uses of MPIDI_msg_sz_t do not generate warnings. I think that means that MPIDI_msg_sz_t is the correct size.<br>
<br><br><b>2)</b><br>I'm still having issues with "MPIU_Find_local_and_external". I am getting the fall-back version (judging from the line numbers), but it prints a lot of these messages:<blockquote style="border: medium none ; margin: 0pt 0pt 0pt 40px; padding: 0px;">
stderr[2]: Internal Error: invalid error code 1312d10 (Ring Index out of range) in MPIU_Find_local_and_external:213<br>stderr[0]: Internal Error: invalid error code 1312d10 (Ring Index out of range) in MPIU_Find_local_and_external:213<br>
stderr[3]: Internal Error: invalid error code 1312d10 (Ring Index out of range) in MPIU_Find_local_and_external:213<br>stderr[1]: Internal Error: invalid error code 1312d10 (Ring Index out of range) in MPIU_Find_local_and_external:213</blockquote>
<div><div><br>Changing the whole of that function to "return MPI_ERR_UNKNOWN;" still seems to work fine. Is there a way to be sure that the "MPIU_ERR_SETANDJUMP" doesn't print anything?<br><br><br>
<br>Thanks,<br>Joe Ratterman<br><a href="mailto:jratt@us.ibm.com">jratt@us.ibm.com</a><br></div><div><br></div><br><div class="gmail_quote">On Wed, Mar 18, 2009 at 2:09 PM, Pavan Balaji <span dir="ltr"><<a href="mailto:balaji@mcs.anl.gov" target="_blank">balaji@mcs.anl.gov</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Hi all,<br>
<br>
A new release of MPICH2 (1.1b1) is available from <a href="http://www.mcs.anl.gov/research/projects/mpich2/downloads/index.php?s=downloads" target="_blank">http://www.mcs.anl.gov/research/projects/mpich2/downloads/index.php?s=downloads</a><br>
<br>
This is a preview release in the MPICH2-1.1 series. MPICH2 developers, testers and other advanced users are encouraged to try this release out and let us know if you encounter any problems.<br>
<br>
The major additions in the 1.1b1 release are highlighted below.<br>
<br>
(Windows binaries are also available on the project website: <a href="http://www.mcs.anl.gov/research/projects/mpich2" target="_blank">http://www.mcs.anl.gov/research/projects/mpich2</a>).<br>
<br>
Regards,<br>
The MPICH2 Team<br>
<br>
--------------------------------------------------------------------------<br>
Main changes in the 1.1 release<br>
===============================<br>
<br>
# OVERALL: Added MPI 2.1 support.<br>
<br>
# OVERALL: Nemesis is now the default configuration channel with a completely new TCP communication module.<br>
<br>
# OVERALL: Windows support for nemesis.<br>
<br>
# OVERALL: Added a new Myrinet MX network module for nemesis.<br>
<br>
# OVERALL: Initial support for shared-memory aware collective communication operations. Currently MPI_Bcast, MPI_Reduce, MPI_Allreduce, and MPI_Scan.<br>
<br>
# OVERALL: Improved handling of MPI Attributes.<br>
<br>
# OVERALL: Support for BlueGene/P through the DCMF library (thanks to IBM for the patch).<br>
<br>
# OVERALL: Added dynamic processes support for Nemesis.<br>
<br>
# OVERALL: Added automatic as well as statically runtime configurable receive timeout variation for MPD (thanks to OSU for the patch).<br>
<br>
# OVERALL: Improved performance for MPI_Allgatherv and MPI_Gatherv.<br>
<br>
# PM/PMI: Initial support for the new Hydra process management framework (current support is for ssh and fork).<br>
<br>
# ROMIO: Added support for MPI_Type_create_resized and MPI_Type_create_indexed_block datatypes in ROMIO.<br>
<br>
# ROMIO: Optimized Lustre ADIO driver (thanks to Weikuan Yu for<br>
initial work and Sun for further improvements).<br>
<br>
# Many other bug fixes, memory leak fixes and code cleanup. A full list of changes is available using:<br>
<br>
svn log -r813:HEAD <a href="https://svn.mcs.anl.gov/repos/mpi/mpich2/tags/release/mpich2-1.1b1" target="_blank">https://svn.mcs.anl.gov/repos/mpi/mpich2/tags/release/mpich2-1.1b1</a><br>
<br>
... or at the following link:<br>
<br>
<a href="https://trac.mcs.anl.gov/projects/mpich2/log/mpich2/tags/release/mpich2-1.1b1?action=follow_copy&rev=HEAD&stop_rev=813&mode=follow_copy" target="_blank">https://trac.mcs.anl.gov/projects/mpich2/log/mpich2/tags/release/mpich2-1.1b1?action=follow_copy&rev=HEAD&stop_rev=813&mode=follow_copy</a><br>
<font color="#888888">
<br>
-- <br>
Pavan Balaji<br>
<a href="http://www.mcs.anl.gov/%7Ebalaji" target="_blank">http://www.mcs.anl.gov/~balaji</a><br>
</font></blockquote></div><br></div>