<div dir="ltr">Anton,<div>I guess, when you reuse matrix and its symbolic factor with updated numerical values, superlu_dist requires this option. I'm cc'ing Sherry to confirm it.</div><div><br></div><div>I'll check petsc/superlu-dist interface to set this flag for this case.</div><div><br></div><div>Hong<br><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Oct 25, 2016 at 8:20 AM, Anton Popov <span dir="ltr"><<a href="mailto:popov@uni-mainz.de" target="_blank">popov@uni-mainz.de</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div bgcolor="#FFFFFF" text="#000000">
<p>Hong,</p>
<p>I get all the problems gone and valgrind-clean output if I
specify this:</p>
<p>-mat_superlu_dist_fact SamePattern_SameRowPerm<br>
</p>
What does SamePattern_SameRowPerm actually mean? <br>
Row permutations are for large diagonal, column permutations are for
sparsity, right?<br>
Will it skip subsequent matrix permutations for large diagonal even
if matrix values change significantly?<br>
<br>
Surprisingly everything works even with:<br>
<br>
-mat_superlu_dist_colperm PARMETIS<br>
-mat_superlu_dist_parsymbfact TRUE<br>
<br>
Thanks,<br>
Anton<br>
<br>
<div class="m_-1905162172435090326moz-cite-prefix">On 10/24/2016 09:06 PM, Hong wrote:<br>
</div>
<blockquote type="cite">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote">Anton:
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div bgcolor="#FFFFFF"><span class="m_-1905162172435090326gmail-">
<blockquote type="cite">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote">If replacing
superlu_dist with mumps, does your code work?</div>
</div>
</div>
</blockquote>
</span> yes</div>
</blockquote>
<div> </div>
<div>You may use mumps in your code, or tests different
options for superlu_dist:</div>
<div><br>
</div>
<div>
<div> -mat_superlu_dist_equil: <TRUE> Equilibrate
matrix (None)</div>
<div> -mat_superlu_dist_rowperm <LargeDiag> Row
permutation (choose one of) LargeDiag NATURAL (None)</div>
<div> -mat_superlu_dist_colperm <METIS_AT_PLUS_A>
Column permutation (choose one of) NATURAL MMD_AT_PLUS_A
MMD_ATA METIS_AT_PLUS_A PARMETIS (None)</div>
<div> -mat_superlu_dist_<wbr>replacetinypivot: <FALSE>
Replace tiny pivots (None)</div>
<div> -mat_superlu_dist_parsymbfact: <FALSE>
Parallel symbolic factorization (None)</div>
<div> -mat_superlu_dist_fact <SamePattern> Sparsity
pattern for repeated matrix factorization (choose one
of) SamePattern SamePattern_SameRowPerm (None)</div>
</div>
<div><br>
</div>
<div>The options inside <> are defaults. You may try
others. This might help narrow down the bug.</div>
<div><br>
</div>
<div>Hong</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div bgcolor="#FFFFFF"><span class="m_-1905162172435090326gmail-"><br>
<blockquote type="cite">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote">Hong
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div bgcolor="#FFFFFF"><span class="m_-1905162172435090326gmail-m_-8985211249997465386gmail-">
<div class="m_-1905162172435090326gmail-m_-8985211249997465386gmail-m_4450561754203495905moz-cite-prefix">On
10/24/2016 05:47 PM, Hong wrote:<br>
</div>
<blockquote type="cite">
<div dir="ltr">Barry,
<div>Your change indeed fixed the
error of his testing code.</div>
<div>As Satish tested, on your
branch, ex16 runs smooth.</div>
<div><br>
</div>
<div>I do not understand why on
maint or master branch, ex16
creases inside superlu_dist, but
not with mumps. </div>
<div><br>
</div>
</div>
</blockquote>
<br>
</span> I also confirm that ex16 runs fine
with latest fix, but unfortunately not my
code.<br>
<br>
This is something to be expected, since my
code preallocates once in the beginning.
So there is no way it can be affected by
multiple preallocations. Subsequently I
only do matrix assembly, that makes sure
structure doesn't change (set to get error
otherwise).<br>
<br>
Summary: we don't have a simple test code
to debug superlu issue anymore.<span class="m_-1905162172435090326gmail-m_-8985211249997465386gmail-HOEnZb"><font color="#888888"><br>
<br>
Anton</font></span><span class="m_-1905162172435090326gmail-m_-8985211249997465386gmail-"><br>
<br>
<blockquote type="cite">
<div dir="ltr">
<div>Hong<br>
<div class="gmail_extra"><br>
<div class="gmail_quote">On Mon,
Oct 24, 2016 at 9:34 AM,
Satish Balay <span dir="ltr"><<a href="mailto:balay@mcs.anl.gov" target="_blank">balay@mcs.anl.gov</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><span>On Mon, 24 Oct 2016, Barry
Smith wrote:<br>
<br>
><br>
</span><span>> > [Or
perhaps Hong is using a
different test code and is
observing bugs<br>
> > with
superlu_dist interface..]<br>
><br>
> She states that
her test does a NEW
MatCreate() for each
matrix load (I cut and
pasted it in the email I
just sent). The bug I
fixed was only related to
using the SAME matrix from
one MatLoad() in another
MatLoad().<br>
<br>
</span>Ah - ok.. Sorry -
wasn't thinking clearly :(<br>
<span class="m_-1905162172435090326gmail-m_-8985211249997465386gmail-m_4450561754203495905HOEnZb"><font color="#888888"><br>
Satish<br>
</font></span></blockquote>
</div>
<br>
</div>
</div>
</div>
</blockquote>
<br>
</span></div>
</blockquote>
</div>
<br>
</div>
</div>
</blockquote>
<br>
</span></div>
</blockquote>
</div>
<br>
</div>
</div>
</blockquote>
<br>
</div>
</blockquote></div><br></div></div></div>