Thank you very much!<br><br><div class="gmail_quote">2012/5/2 Bogdan Costescu <span dir="ltr"><<a href="mailto:bcostescu@gmail.com" target="_blank">bcostescu@gmail.com</a>></span><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im">On Tue, May 1, 2012 at 5:34 PM, Yorquant Wang <<a href="mailto:wangykoo@gmail.com">wangykoo@gmail.com</a>> wrote:<br>
</div><div class="im">> I recently want to add some myself pull code in the gromacs4.5.4 kernel.<br>
<br>
</div>The pull code is only called from mdrun, but mdrun is also able to run<br>
on multiple nodes using MPI - so you should not rely on the code<br>
always running multithreaded. The thread_mpi internal library<br>
implements the subset of the MPI calls which are used in GROMACS, so<br>
the function calls are the same and which ones are used (an external<br>
MPI lib or thread_mpi) is decided at compile time. You don't need to<br>
concern yourself with sharing memory, but use the MPI data model. If<br>
you only have experience with OpenMP/threads programming, it's time to<br>
look up some MPI tutorials.<br>
<br>
Good luck!<br>
<span class="HOEnZb"><font color="#888888">Bogdan<br>
</font></span><div class="HOEnZb"><div class="h5">--<br>
gmx-developers mailing list<br>
<a href="mailto:gmx-developers@gromacs.org">gmx-developers@gromacs.org</a><br>
<a href="http://lists.gromacs.org/mailman/listinfo/gmx-developers" target="_blank">http://lists.gromacs.org/mailman/listinfo/gmx-developers</a><br>
Please don't post (un)subscribe requests to the list. Use the<br>
www interface or send it to <a href="mailto:gmx-developers-request@gromacs.org">gmx-developers-request@gromacs.org</a>.<br>
</div></div></blockquote></div><br>