<div dir="ltr">It does not appear to me that any change is necessary. <div><br></div><div>The only part about MPI_COMM_WORLD is in an 'as an example' clause.</div><div><br></div><div>That being said, substituting MPI_COMM_WORLD with some generic 'communicator comm' would also work. </div><div><br></div><div>Aurelien</div><div><br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Tue, Jun 25, 2019 at 9:49 AM Pritchard Jr., Howard via mpiwg-ft <<a href="mailto:mpiwg-ft@lists.mpi-forum.org">mpiwg-ft@lists.mpi-forum.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div lang="EN-US">
<div class="gmail-m_-6390681469657327504WordSection1">
<p class="MsoNormal"><span style="font-size:11pt">Hello MPI FTer’s,<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11pt"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11pt">The Sessions WG could use some help/suggestions about how to adjust the following advice to implementors that accompanies the definition of MPI_Abort:<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11pt"><u></u> <u></u></span></p>
<p class="gmail-m_-6390681469657327504p1"><span class="gmail-m_-6390681469657327504s1">\begin</span><span class="gmail-m_-6390681469657327504s2">{implementors}</span><u></u><u></u></p>
<p class="gmail-m_-6390681469657327504p2"><span class="gmail-m_-6390681469657327504apple-converted-space"> </span><span class="gmail-m_-6390681469657327504s2">After aborting a subset of processes, a high quality implementation should</span><u></u><u></u></p>
<p class="gmail-m_-6390681469657327504p2"><span class="gmail-m_-6390681469657327504apple-converted-space"> </span><span class="gmail-m_-6390681469657327504s2">be able to provide error handling for communicators, windows, and files</span><u></u><u></u></p>
<p class="gmail-m_-6390681469657327504p2"><span class="gmail-m_-6390681469657327504apple-converted-space"> </span><span class="gmail-m_-6390681469657327504s2">involving both aborted and non-aborted processes. As an example, if the</span><u></u><u></u></p>
<p class="gmail-m_-6390681469657327504p2"><span class="gmail-m_-6390681469657327504apple-converted-space"> </span><span class="gmail-m_-6390681469657327504s2">user changes the error handler for
</span><span class="gmail-m_-6390681469657327504s1">\const</span><span class="gmail-m_-6390681469657327504s3">{</span><span class="gmail-m_-6390681469657327504s2">MPI</span><span class="gmail-m_-6390681469657327504s3">\_</span><span class="gmail-m_-6390681469657327504s2">COMM</span><span class="gmail-m_-6390681469657327504s3">\_</span><span class="gmail-m_-6390681469657327504s2">WORLD</span><span class="gmail-m_-6390681469657327504s3">}</span><span class="gmail-m_-6390681469657327504s2"> to</span><u></u><u></u></p>
<p class="gmail-m_-6390681469657327504p2"><span class="gmail-m_-6390681469657327504apple-converted-space"> </span><span class="gmail-m_-6390681469657327504s1">\const</span><span class="gmail-m_-6390681469657327504s3">{</span><span class="gmail-m_-6390681469657327504s2">MPI</span><span class="gmail-m_-6390681469657327504s3">\_</span><span class="gmail-m_-6390681469657327504s2">ERRORS</span><span class="gmail-m_-6390681469657327504s3">\_</span><span class="gmail-m_-6390681469657327504s2">RETURN</span><span class="gmail-m_-6390681469657327504s3">}</span><span class="gmail-m_-6390681469657327504s2">
or a custom error handler, when a subset of</span><u></u><u></u></p>
<p class="gmail-m_-6390681469657327504p2"><span class="gmail-m_-6390681469657327504apple-converted-space"> </span><span class="gmail-m_-6390681469657327504s1">\const</span><span class="gmail-m_-6390681469657327504s3">{</span><span class="gmail-m_-6390681469657327504s2">MPI</span><span class="gmail-m_-6390681469657327504s3">\_</span><span class="gmail-m_-6390681469657327504s2">COMM</span><span class="gmail-m_-6390681469657327504s3">\_</span><span class="gmail-m_-6390681469657327504s2">WORLD</span><span class="gmail-m_-6390681469657327504s3">}</span><span class="gmail-m_-6390681469657327504s2">
is aborted, the remaining processes in</span><u></u><u></u></p>
<p class="gmail-m_-6390681469657327504p2"><span class="gmail-m_-6390681469657327504apple-converted-space"> </span><span class="gmail-m_-6390681469657327504s1">\const</span><span class="gmail-m_-6390681469657327504s3">{</span><span class="gmail-m_-6390681469657327504s2">MPI</span><span class="gmail-m_-6390681469657327504s3">\_</span><span class="gmail-m_-6390681469657327504s2">COMM</span><span class="gmail-m_-6390681469657327504s3">\_</span><span class="gmail-m_-6390681469657327504s2">WORLD</span><span class="gmail-m_-6390681469657327504s3">}</span><span class="gmail-m_-6390681469657327504s2">
should be able to continue communicating with each</span><u></u><u></u></p>
<p class="gmail-m_-6390681469657327504p2"><span class="gmail-m_-6390681469657327504apple-converted-space"> </span><span class="gmail-m_-6390681469657327504s2">other and receive appropriate error codes when attempting communication</span><u></u><u></u></p>
<p class="gmail-m_-6390681469657327504p2"><span class="gmail-m_-6390681469657327504apple-converted-space"> </span><span class="gmail-m_-6390681469657327504s2">with an aborted process.</span><u></u><u></u></p>
<p class="gmail-m_-6390681469657327504p1"><span class="gmail-m_-6390681469657327504s1">\end</span><span class="gmail-m_-6390681469657327504s2">{implementors}</span><u></u><u></u></p>
<p class="MsoNormal"><span style="font-size:11pt"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11pt">We would like to generalize this advice to implementors to the case where MPI_COMM_WORLD isn’t a valid communicator, i.e. when an application is using the Sessions model.
<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11pt">We think that there would need to be some reworking of the existing text to cover the sessions use case. Since the FT group has worked quite a bit on this text, we’d defer to your group for suggestions on
how to generalize this text to cover the sessions use case.<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11pt"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11pt">Thanks very much for any help,<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11pt"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11pt">Howard<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11pt"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11pt">-- <u></u><u></u></span></p>
<div>
<p class="MsoNormal"><span style="font-size:11pt"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:10.5pt;color:black">Howard Pritchard<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:10.5pt;color:black">HPC-ENV<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:10.5pt;color:black">Los Alamos National Laboratory<u></u><u></u></span></p>
</div>
<p class="MsoNormal"><u></u> <u></u></p>
</div>
</div>
_______________________________________________<br>
mpiwg-ft mailing list<br>
<a href="mailto:mpiwg-ft@lists.mpi-forum.org" target="_blank">mpiwg-ft@lists.mpi-forum.org</a><br>
<a href="https://lists.mpi-forum.org/mailman/listinfo/mpiwg-ft" rel="noreferrer" target="_blank">https://lists.mpi-forum.org/mailman/listinfo/mpiwg-ft</a><br>
</blockquote></div>