<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Darius Buntinas wrote:
<blockquote cite="mid:2B160B4A-18E4-4419-9243-D979286A0008@mcs.anl.gov"
type="cite">
<pre wrap="">On Sep 22, 2010, at 12:04 PM, Terry Dontje wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Darius Buntinas wrote:
Ok I need a clarification here because I feel that I might be misinterpreting something. So is the CANNOT_CONTINUE error class only returned by MPI after a previous error condition has been returned that has caused problems? For example let's say we did an MPI_Bcast that resulted in a return of MPI_ERR_OP and for whatever reason the MPI library is borked. So the next call to MPI would return the CANNOT_CONTINUE error class?
</pre>
</blockquote>
<pre wrap=""><!---->
Yes, I believe that's the behavior we talked about at the forum.
</pre>
<blockquote type="cite">
<pre wrap="">So is this an escape hatch for an implementation that does not support any type of fault tolerance to explicitly notify the user they shouldn't proceed any further? I really wonder how many implementations will do such.
</pre>
</blockquote>
<pre wrap=""><!---->
Well, I wouldn't say it's an escape hatch, since an implementation that doesn't support any fault tolerance needn't ever return CANNOT_CONTINUE. Because we still haven't defined what happens after an error, operation after an error is still undefined, the implementation is still free to do anything including returning MPI_SUCCESS.
However, a high quality implementation, would return CANNOT_CONTINUE on subsequent MPI calls when it knows that something is borked beyond repair.
Note that there's still a lot of middle ground between "no errors" and "totally borked". We're giving the implementation an error to return if it finds itself totally borked. This is why I agree with Josh's statement that this proposal is a minor change.
</pre>
</blockquote>
That's funny, because I was thinking a high quality implementation
would never return CANNOT_CONTINUE but more distinct error codes that
lets an application to recover. I would think very few errors would
actually completely obliterate an MPI library's internal structures.
At least the implementations I've seen that's seems to be the case.<br>
<br>
--td<br>
<blockquote cite="mid:2B160B4A-18E4-4419-9243-D979286A0008@mcs.anl.gov"
type="cite">
<pre wrap="">-d
_______________________________________________
mpi3-ft mailing list
<a class="moz-txt-link-abbreviated" href="mailto:mpi3-ft@lists.mpi-forum.org">mpi3-ft@lists.mpi-forum.org</a>
<a class="moz-txt-link-freetext" href="http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft">http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft</a>
</pre>
</blockquote>
<br>
<br>
<div class="moz-signature">-- <br>
<meta http-equiv="content-type" content="text/html; charset=ISO-8859-1">
<title></title>
<img moz-do-not-send="false"
src="cid:part1.07000309.04040200@oracle.com" alt="Oracle"><br>
<div class="moz-signature">
<div class="moz-signature">
<div class="moz-signature">
<div class="moz-signature">Terry D. Dontje | Principal Software Engineer<br>
<div class="moz-signature"><font color="#666666" face="Verdana" size="2">Developer
Tools
Engineering | +1.781.442.2631<br>
</font>
<font color="#ff0000" face="Verdana" size="2">Oracle
</font><font color="#666666" face="Verdana" size="2"><b> - Performance
Technologies</b></font><br>
<font color="#666666" face="Verdana" size="2">
95 Network Drive, Burlington, MA 01803<br>
Email <a href="mailto:terry.dontje@oracle.com">terry.dontje@oracle.com</a><br>
</font><br>
</div>
</div>
</div>
</div>
</div>
</div>
</body>
</html>