<html xmlns:v="urn:schemas-microsoft-com:vml" 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 14 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0in;
margin-bottom:.0001pt;
font-size:11.0pt;
font-family:"Calibri","sans-serif";}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:blue;
text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
{mso-style-priority:99;
color:purple;
text-decoration:underline;}
pre
{mso-style-priority:99;
mso-style-link:"HTML Preformatted Char";
margin:0in;
margin-bottom:.0001pt;
font-size:10.0pt;
font-family:"Courier New";}
span.HTMLPreformattedChar
{mso-style-name:"HTML Preformatted Char";
mso-style-priority:99;
mso-style-link:"HTML Preformatted";
font-family:"Courier New";}
span.EmailStyle19
{mso-style-type:personal;
font-family:"Calibri","sans-serif";
color:windowtext;}
span.EmailStyle20
{mso-style-type:personal;
font-family:"Calibri","sans-serif";
color:#1F497D;}
span.EmailStyle21
{mso-style-type:personal-reply;
font-family:"Calibri","sans-serif";
color:#1F497D;}
.MsoChpDefault
{mso-style-type:export-only;
font-size:10.0pt;}
@page WordSection1
{size:8.5in 11.0in;
margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]-->
</head>
<body lang="EN-US" link="blue" vlink="purple">
<div class="WordSection1">
<p class="MsoNormal"><span style="font-size:10.0pt;color:black">Thanks Huiwei for trying this out.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;color:black">Looks like this is not allowed. Look at Pavan’s reply.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;color:black">-Hirak<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;color:black"><o:p> </o:p></span></p>
<div style="mso-element:para-border-div;border:none;border-bottom:solid windowtext 1.0pt;padding:0in 0in 1.0pt 0in">
<p class="MsoNormal" style="border:none;padding:0in"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
</div>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">Hi Hirak,<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">I can reproduce your error with the attached program using one process:<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"> mpicc -g -o mpi_comm_accept mpi_comm_accept.c -pthread<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"> mpiexec -n 1 ./mpi_comm_accept<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">I found MPI_Send and MPI_Recv were not using the same communicator to communicate, that’s why MPI_Recv will never receive the message. So either the communicator creation
was wrong or the application was wrong.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">The standard said MPI_Comm_accept and MPI_Comm_connect are used for “establishing contact between two groups of processes that do not share an existing communicator”.
But in this case, thread 1 and 2 do share an existing communicator and want to create a new communicator based on the common communicator. I don’t know if it is allowed. If it is allowed, then MPI_Comm_accept and MPI_Comm_connect should be fixed to support
multiple thread case; if it is not allowed, we may need to change the application to use another way to terminate MPI_Comm_accept.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">Thanks,<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">—<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">Huiwei<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">On Oct 8, 2014, at 12:14 AM, Roy, Hirak <<a href="https://lists.mpich.org/mailman/listinfo/discuss">Hirak_Roy at mentor.com</a>> wrote:<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Hi Pavan,<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Here is my code for thread2 :<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> do {<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> MPI_Comm newComm ;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> MPI_Comm_accept (m_serverPort.c_str(), MPI_INFO_NULL, 0, MPI_COMM_SELF, &newComm);<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Log ("Accepted a connection");<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> int buf = 0 ;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> MPI_Status status ;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> MPI_Recv(&buf, 1, MPI_INT, MPI_ANY_SOURCE, MPI_ANY_TAG, newComm, &status);<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> if (status.MPI_TAG == MPI_MSG_TAG_NEW_CONN) {<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> m_clientComs[m_clientCount] = newComm ;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> m_clientCount++;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> } else if (status.MPI_TAG == MPI_MSG_TAG_SHUTDOWN) {<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Log ("Shutdown");<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> //MPI_Comm_disconnect (&newComm);<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Log ("Disconnect");<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> break;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> } else {<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Log ("Unmatched Receive");<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> }<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> } while(1) ;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Here is my code for thread1 to terminate thread2 :<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> MPI_Comm newComm ;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> MPI_Comm_connect (m_serverPort.c_str(), MPI_INFO_NULL, 0, MPI_COMM_SELF, &newComm);<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Log ("Connect to Self");<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> int val = 0 ;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> MPI_Request req ;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> MPI_Send(&val, 1, MPI_INT, 0, MPI_MSG_TAG_SHUTDOWN, newComm);<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Log ("Successful");<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> //MPI_Status stat ;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> //MPI_Wait(&req, &stat);<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Log ("Complete");<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> //MPI_Comm_disconnect(&newComm);<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> The MPI_Send/Recv waits.<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> I am using sock channel.<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> For nemesis, I get the following crash :<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Assertion failed in file ./src/mpid/ch3/channels/nemesis/include/mpid_nem_inline.h at line 58: vc_ch->is_local<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> internal ABORT - process 0<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> I tried non-blocking send and receive followed by wait. However, that also does not solve the problem.<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Thanks,<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Hirak<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> -----<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Hirak,<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> Your approach should work fine. I’m not sure what issue you are facing. I assume thread 1 is doing this:<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> while (1) {<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> MPI_Comm_accept(..);<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> MPI_Recv(.., tag, ..);<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> if (tag == REGULAR_CONNECTION)<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> continue;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> else if (tag == TERMINATION) {<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> MPI_Send(..);<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> break;<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> }<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> }<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> In this case, all clients do an MPI_Comm_connect and then send a message with tag = REGULAR_CONNECTION. When thread 2 is done with its work, it’ll do an MPI_Comm_connect
and then send a message with tag = TERMINATION, wait for a response from thread 1, and call finalize.<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i><o:p> </o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><i> — Pavan<o:p></o:p></i></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:black">><o:p> </o:p></span></p>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
</body>
</html>