<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<meta name="generator" content="HTML Tidy for Windows (vers 25 March 2009), see www.w3.org">
<meta name="Generator" content="MS Exchange Server version 14.03.0157.000">
<title>Re: [mpiwg-tools] Ticket #383 -- Final text?</title>
</head>
<body>
Looks mostly good to me, but length arguments are INOUT, so that may be confusing - perhaps just drop the IN?<br>
<br>
Martin<br>
<br>
<br>
<br>
Sent with Good (www.good.com)<br>
<br>
<br>
-----Original Message-----<br>
<b>From: </b>Kathryn Mohror [<a href="mailto:kathryn@llnl.gov">kathryn@llnl.gov</a>]<br>
<b>Sent: </b>Monday, May 19, 2014 08:31 AM Pacific Standard Time<br>
<b>To: </b>Bronis R. de Supinski; mpiwg-tools@lists.mpi-forum.org<br>
<b>Subject: </b>Re: [mpiwg-tools] Ticket #383 -- Final text?<br>
<br>
<!-- Converted from text/plain format -->
<p><font size="2">Jeff,<br>
<br>
Is this what you want for the text, attached?<br>
<br>
Kathryn<br>
_________________________________________________________________<br>
Kathryn Mohror, kathryn@llnl.gov, <a href="http://scalability.llnl.gov/">http://scalability.llnl.gov/</a><br>
Scalability Team @ Lawrence Livermore National Laboratory, Livermore, CA,<br>
USA<br>
<br>
<br>
<br>
<br>
<br>
<br>
On 5/19/14, 8:03 AM, "Bronis R. de Supinski" <bronis@llnl.gov> wrote:<br>
<br>
><br>
>"sufficiently long enough" is redundant; "sufficiently long"<br>
>is sufficient.<br>
><br>
>On Mon, 19 May 2014, Jeff Squyres (jsquyres) wrote:<br>
><br>
>> On May 18, 2014, at 2:35 AM, Martin Schulz <schulzm@llnl.gov> wrote:<br>
>><br>
>>> Hi Kathryn, all,<br>
>>><br>
>>> A few comments:<br>
>>> - "Adhere to the same definition from the MPI implementation.² - not<br>
>>>sure<br>
>>> what we meant with this anymore. Perhaps we should just drop this part.<br>
>><br>
>> What I meant by that is that each MPI implementation will create a<br>
>>definition for each cvar/pvar/etc. The implementation must adhere to<br>
>>that same definition in all connected processes. I.e., it can't change<br>
>>the definition of exactly what that variable is in different processes<br>
>>(e.g., bytes_in = bytes received from TCP in one process, but bytes<br>
>>received across shared memory in another process).<br>
>><br>
>>> - "OUT index values² is not correct, since index values are IN<br>
>>>parameters<br>
>>> to the get_info calls. Perhaps just drop OUT?<br>
>><br>
>> Good catch; missed that. Yes, this whole clause can go.<br>
>><br>
>>> - The mandate of equal OUT/INPUT values is also not quite right - for<br>
>>>the<br>
>>> strings, it¹s OK to return substrings of the size argument is too<br>
>>>small. I<br>
>>> know that¹s a technicality, but I am not sure if this causes problems<br>
>>> later down the road.<br>
>><br>
>> Mmm. Good point. So the current text is:<br>
>><br>
>> * Return the same INOUT and OUT values from the relevant<br>
>>MPI_T_*_GET_INFO function.<br>
>><br>
>> (I took the liberty of dropping the index value clause)<br>
>><br>
>> I notice that we actually return 3 kinds of things from GET_INFO<br>
>>functions: integers/enums, strings, and MPI handles (i.e.,<br>
>>MPI_Datatype). So how about this:<br>
>><br>
>> * Return the same INOUT and OUT integer and enum values from the<br>
>>relevant MPI_T_*_GET_INFO function.<br>
>> * If sufficiently long enough string length IN parameters are supplied,<br>
>>return the same string OUT parameters from the relevant MPI_T_*_GET_INFO<br>
>>function.<br>
>> * Return a handle to an equivalent MPI object from the relevant<br>
>>MPI_T_*_GET_INFO function.<br>
>><br>
>><br>
>><br>
>> --<br>
>> Jeff Squyres<br>
>> jsquyres@cisco.com<br>
>> For corporate legal information go to:<br>
>><a href="http://www.cisco.com/web/about/doing_business/legal/cri/">http://www.cisco.com/web/about/doing_business/legal/cri/</a><br>
>><br>
>> _______________________________________________<br>
>> mpiwg-tools mailing list<br>
>> mpiwg-tools@lists.mpi-forum.org<br>
>> <a href="http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-tools">http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-tools</a><br>
<br>
</font></p>
</body>
</html>