<html xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=us-ascii">
<meta name="Generator" content="Microsoft Word 15 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
{font-family:"Cambria Math";
panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
{font-family:Aptos;
panose-1:2 11 0 4 2 2 2 2 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0in;
font-size:12.0pt;
font-family:"Aptos",sans-serif;}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:blue;
text-decoration:underline;}
.MsoChpDefault
{mso-style-type:export-only;
font-size:10.0pt;
mso-ligatures:none;}
@page WordSection1
{size:8.5in 11.0in;
margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
{page:WordSection1;}
--></style>
</head>
<body lang="EN-US" link="blue" vlink="purple" style="word-wrap:break-word">
<div class="WordSection1">
<div>
<p class="MsoNormal"><span style="font-size:11.0pt">A new release candidate of MPICH, 4.3.0rc1, is now available for download. See the CHANGES below for a detailed list of features.<br>
<br>
At this stage, only critical bugs will be fixed before the 4.3.0 GA release. Please give it a try!<br>
<br>
You can find the release on our downloads page (<a href="https://urldefense.us/v3/__https://www.mpich.org/downloads/__;!!G_uCfscf7eWS!YH94mnV2Xl_axlZG1RNsLZw2AKXk3M5CgL9xY2hBsg4b1zNW73QDKHStC-6LkuzKWY_7lstH1nigr_j1$">https://www.mpich.org/downloads/</a>)<br>
<br>
Regards,<br>
The MPICH team<br>
<br>
===============================================================================<o:p></o:p></span></p>
<div style="mso-element:para-border-div;border:none;border-bottom:double windowtext 2.25pt;padding:0in 0in 1.0pt 0in">
<p class="MsoNormal" style="border:none;padding:0in"><span style="font-size:11.0pt"> Changes in 4.3<o:p></o:p></span></p>
</div>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Support MPI memory allocation kinds side document.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Support MPI ABI Proposal. Configure with --enable-mpi-abi and build with<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> mpicc_abi. By default, mpicc still builds and links with MPICH ABI.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Experimental API MPIX_Op_create_x. It supports user callback function with<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> extra_state context and op destructor callback. It supports language bindings<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> to use proxy function for language-specific user callbacks.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Experimental API MPIX_{Comm,File,Session,Win}_create_errhandler_x. They allow<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> user error handlers to have extra_state context and corresponding destructor.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> This allows language bindings to implement user error handlers via proxy.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Experimental API MPIX_Request_is_complete. This is a pure request state query<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> function that will not invoke progress, nor will free the request. This should<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> help applications that want separate task dependency checking from progress<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> engine to avoid progress contentions, especially in multi-threaded context.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> It is also useful for tools to profile non-deterministic calls such as<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> MPI_Test.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Experimental API MPIX_Async_start. This function let applications to inject<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> progress hooks to MPI progress. It allows application to implement custom<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> asynchronous operations that will be progressed by MPI. It avoids having to<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> implement separate progress mechanisms that may either take additional<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> resource or contend with MPI progress and negatively impact performance. It<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> also allows applications to create custom MPI operations, such as MPI<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> nonblocking collectives, and achieve near native performance.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Added benchmark tests test/mpi/bench/p2p_{latency,bw}.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Added CMA support in CH4 IPC.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Added IPC read algorithm for intranode Allgather and Allgatherv.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Added CVAR MPIR_CVAR_CH4_SHM_POSIX_TOPO_ENABLE to enable non-temporal memcpy<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> for inter-numa shm communication.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Added CVAR MPIR_CVAR_DEBUG_PROGRESS_TIMEOUT for debugging MPI deadlock issues.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># ch4:ucx now supports dynamic processes. MPI_Comm_spawn{_multiple} will work.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> MPI_Open_port will fail due to ucx port name exceeds current MPI_MAX_PORT_NAME<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> of 256. One can work around by use an info hint "port_name_size" and use a<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> larger port name buffer.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># PMI-1 defines PMI_MAX_PORT_NAME, which may be different from MPI_MAX_PORT_NAME.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> This is used by "PMI_Lookup_name". Consequently, MPI_Lookup_name accepts info<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> hint "port_name_size" that may be larger than MPI_MAX_PORT_NAME. If the port<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"> name does not fit in "port_name_size", it will return a truncation error.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Autogen default to use -yaksa-depth=2.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Default MPIR_CVAR_CH4_ROOTS_ONLY_PMI to on.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Added ch4 netmod API am_tag_send and am_tag_recv.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Added MPIR_CVAR_CH4_OFI_EAGER_THRESHOLD to force RNDV send mode.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt"># Make check target will run ROMIO tests.<o:p></o:p></span></p>
</div>
</div>
</body>
</html>