Charm++ parallel programming system

["Buch, Ronak Akshay" <rabuch2 AT illinois.edu>]

Hello everyone!

On behalf of the Charm++ development team and contributors, I'm pleased to announce a beta release of Charm++ in advance of the upcoming version 6.10.0. We ask that users take this opportunity to test the latest code with their applications and report any issues encountered.

The code for this release can be obtained via git by

git clone https://github.com/UIUC-PPL/charm.git

git checkout v6.10.0-beta1

Or downloaded directly from: https://github.com/UIUC-PPL/charm/releases/tag/v6.10.0-beta1

There are a few known issues in this beta, visible at https://github.com/UIUC-PPL/charm/milestone/16.

537 commits have been made since the release of version 6.9.0, some of the larger and more exciting improvements in the system include (for a full list, please see the current release notes at https://github.com/UIUC-PPL/charm/pull/2208):

- Updated the license to clarify the restriction on commercial use of the software

   in the academic distribution.

- We have moved away from .tex in favor of .rst files to make building the
  documentation more portable.

- We have moved bug/issue tracking from Redmine to GitHub, and code
  review from Gerrit to GitHub.
 
- As a preview feature, Charm++ can now be built with CMake (version 3.11 or higher).

- Support for a new Unified Communication X (UCX) networking backend in LRTS,
  thanks to Mellanox and Charmworks staff.

- The Zero Copy API now supports broadcast operations, and is used internally
  for transmission of large readonly objects during startup.

- Addition of a new Zero Copy Post API, for avoiding the receive-side message
  copy. This can be used in both point-to-point and broadcast operations.

- Defined a new API, CkWithinNodeBroadcast, for broadcasting a message from a Group element
  to all other Group elements in the same process or logical node. If the target entry
  method is [nokeep], this API avoids making any copies of the message.

- Callbacks to [inline] entry methods are now executed inline by default. Previously,
  this was only done when the callback was constructed with an optional parameter.

- Eliminated the need for mainchares in user-driven interop mode by adding a new
  split-phase initialization API, fixed a bug in the interop exit sequence, and new
  support for using CkCallback::ckExit when using interop.

- Fixes to the perfReport and memory tracemodes as well as record/replay in SMP mode,
  and improvements to PAPI-enabled builds.

- Various improvements to Charm4Py, such as a new sections implementation, are
  described in the charm4py repository on GitHub.
 
- The CmiAbort and CkAbort functions now support printf-style format strings.
  Please make sure to replace '%' with '%%' in the argument string to print a '%'.

- AMPI now uses Charm++'s zero copy API to transfer large messages efficiently using
  RDMA and CMA wherever possible and profitable.

- More efficient implementations of MPI_Bcast, all MPI_(I)(all)gather(v) routines,
  and reductions with non-commutative operations.