<div dir="ltr">that is, even if one use MPI_COMM to create an IS object. it is always a sequential object, right? thanks, Matt.<br><br><div class="gmail_quote">On Mon, Sep 29, 2008 at 11:40 AM, Matthew Knepley <span dir="ltr"><<a href="mailto:knepley@gmail.com">knepley@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div><div class="Wj3C7c">On Mon, Sep 29, 2008 at 1:33 PM, Yujie <<a href="mailto:recrusader@gmail.com">recrusader@gmail.com</a>> wrote:<br>
> I am wondering how to use ISCreateGeneral().<br>
><br>
> Assuming I use 2cpus when I run my codes. I use MPI_COMM to create an object<br>
> of IS using ISCreateGeneral() and an array on a single cpu. how to<br>
> distribute this IS object on these 2cpus? If I use MatGetSubMatrices() to<br>
> get different submatrices on different cpus. I am wondering how to create IS<br>
> object, using PETSC_COMM_SELF? thanks a lot.<br>
<br>
</div></div>You provide the indices you want for each process \emph{locally} on<br>
that process.<br>
There is no facility for communicating indices in IS.<br>
<br>
Matt<br>
<br>
> Regards,<br>
><br>
> Yujie<br>
<font color="#888888">--<br>
What most experimenters take for granted before they begin their<br>
experiments is infinitely more interesting than any results to which<br>
their experiments lead.<br>
-- Norbert Wiener<br>
<br>
</font></blockquote></div><br></div>