<div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">-on_error_attach_debugger does not start the debugger. I traced it upto the following call:
MatPartitioningApply(). It runs fine up until that function call and then gives the error.<div><br></div><div>Also, to make things even look more weird, I got the following once I was running it:</div><div><br></div><div>
<div>Key 10176 not found!</div><div>INTERNAL ERROR: Invalid error class (66) encountered while returning from</div><div>PMPI_Waitall. Please file a bug report.</div><div><br></div><div>Do you happen to know what that could mean? The funny thing is, it only happened once!</div>
</div></div></blockquote><div><br></div><div>This is a parmetis error from libparmetis/util.c:78 and not with petsc (well, at least, not directly an error from petsc). How did you build parmetis? Which parmetis version and which petsc version are you running? You'll probably have to send configure.log to <a href="mailto:petsc-maint@mcs.anl.gov">petsc-maint@mcs.anl.gov</a>.</div>
<div><br></div><div>By the way, the relevant function that is having the error is:</div><div><br></div><div><div>/*************************************************************************</div><div>* This function does a binary search on an array for a key and returns</div>
<div>* the index</div><div>**************************************************************************/</div><div>idx_t BSearch(idx_t n, idx_t *array, idx_t key)</div></div><div><br></div><div>... which doesn't help much unless you can step into the debugger and set a breakpoint here to give a traceback.</div>
</div>