I have incorporated this patch in superlu_dist_2.5 tarball:<br><a href="http://crd.lbl.gov/~xiaoye/SuperLU/superlu_dist_2.5.tar.gz">http://crd.lbl.gov/~xiaoye/SuperLU/superlu_dist_2.5.tar.gz</a><br><br>Thanks!<br>Sherry<br>
<br><br><div class="gmail_quote">On Tue, Dec 21, 2010 at 12:23 PM, Satish Balay <span dir="ltr"><<a href="mailto:balay@mcs.anl.gov">balay@mcs.anl.gov</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
Sherry,<br>
<br>
One of the patches which we still have in our superlu_dist taball is:<br>
<br>
<a href="http://petsc.cs.iit.edu/petsc/externalpackages/SuperLU_DIST_2.4/raw-diff/ddb59df1a8a5/SRC/superlu_grid.c" target="_blank">http://petsc.cs.iit.edu/petsc/externalpackages/SuperLU_DIST_2.4/raw-diff/ddb59df1a8a5/SRC/superlu_grid.c</a><br>
<br>
could you add this to your tarball?<br>
<br>
thanks,<br>
<font color="#888888">Satish<br>
</font><div><div></div><div class="h5"><br>
On Fri, 3 Dec 2010, Barry Smith wrote:<br>
<br>
><br>
> Sherry,<br>
><br>
> Thanks, we'll upgrade to use those versions with petsc-dev.<br>
><br>
> Hong and Satish,<br>
><br>
> Could you please update petsc-dev with these newest versions of superlu[_dist]<br>
><br>
> Thanks<br>
><br>
> Barry<br>
><br>
> On Dec 3, 2010, at 6:31 PM, Xiaoye S. Li wrote:<br>
><br>
> > Sorry, but just got around to look at this.<br>
> ><br>
> > In SuperLU_4.0, we introduced ILU preconditioner, which works better with MC64 permutation (which is written in f77). But I forgot to include this in slu_Cnames.h, to handle the name mangling issue.<br>
> ><br>
> > In SuperLU_4.1, I converted MC64 from f77 to C using f2c. Therefore, there is no such name mangling problem. The C version is now also used in SuperLU_DIST_2.5.<br>
> ><br>
> > In summary, SuperLU_4.1 and SuperLU_DIST_2.5 upward versions should not need this name mangling for MC64.<br>
> ><br>
> > Sherry<br>
> ><br>
> > On Thu, Dec 2, 2010 at 1:40 PM, Satish Balay <<a href="mailto:balay@mcs.anl.gov">balay@mcs.anl.gov</a>> wrote:<br>
> > Looks like a bug in superlu. [this bug doesn't exist in superlu_dist].<br>
> ><br>
> > I suspect we've always used superlu with superlu_dist on bgp - so<br>
> > didn't see this before. [symbols from -lsuperlu_dist must have been<br>
> > picked up by the linker (before superlu symbols) - and these symbols<br>
> > resolved without errors]<br>
> ><br>
> > The following fixes the issue.<br>
> ><br>
> > login2.surveyor:SuperLU_4.0/SRC > diff -Nru slu_Cnames.h~ slu_Cnames.h<br>
> > --- slu_Cnames.h~ 2009-07-05 00:55:49.000000000 -0500<br>
> > +++ slu_Cnames.h 2010-12-02 15:19:30.005247114 -0600<br>
> > @@ -251,6 +251,9 @@<br>
> > #define izmax1_ IZMAX1<br>
> > #define zlacon_ ZLACON<br>
> ><br>
> > +#define mc64id_ MC64ID<br>
> > +#define mc64ad_ MC64AD<br>
> > +<br>
> > /* Fortran interface */<br>
> > #define c_bridge_dgssv_ C_BRIDGE_DGSSV<br>
> > #define c_fortran_sgssv_ C_FORTRAN_SGSSV<br>
> > @@ -350,6 +353,9 @@<br>
> > #define izmax1_ izmax1<br>
> > #define zlacon_ zlacon<br>
> ><br>
> > +#define mc64id_ mc64id<br>
> > +#define mc64ad_ mc64ad<br>
> > +<br>
> > /* Fortran interface */<br>
> > #define c_bridge_dgssv_ c_bridge_dgssv<br>
> > #define c_fortran_sgssv_ c_fortran_sgssv<br>
> > login2.surveyor:SuperLU_4.0/SRC ><br>
> ><br>
> ><br>
> > I've respun the superlu download tarball with this fix..<br>
> ><br>
> ><br>
> > Note: Looks like Superlu_4.1 tarball has this bug..<br>
> ><br>
> > Satish<br>
> ><br>
> > On Thu, 2 Dec 2010, Satish Balay wrote:<br>
> ><br>
> > > I'll attempt a build and see how it goes..<br>
> > ><br>
> > > satish<br>
> > ><br>
> > > On Thu, 2 Dec 2010, Matthew Knepley wrote:<br>
> > ><br>
> > > > We always always always always need configure.log.<br>
> > > ><br>
> > > > Matt<br>
> > > ><br>
> > > > On Thu, Dec 2, 2010 at 8:41 AM, John R. Cary <<a href="mailto:cary@txcorp.com">cary@txcorp.com</a>> wrote:<br>
> > > ><br>
> > > > > I am trying to link facets on a FEN of <a href="http://surveyor.alcf.anl.gov" target="_blank">surveyor.alcf.anl.gov</a>. It ends<br>
> > > > > with the error,<br>
> > > > ><br>
> > > > > /gpfs/home/projects/facets/surveyor/contrib-xlc-9.0/petsc-3.1-p4-ser/lib/libsuperlu_4.0.a(dldperm.o):<br>
> > > > > In function `dldperm':<br>
> > > > > /gpfs/home/cary/facetspkgs/builds/petsc-3.1-p4/ser/externalpackages/SuperLU_4.0/SRC/dldperm.c:127:<br>
> > > > > undefined reference to `mc64id_'<br>
> > > > > /gpfs/home/cary/facetspkgs/builds/petsc-3.1-p4/ser/externalpackages/SuperLU_4.0/SRC/dldperm.c:134:<br>
> > > > > undefined reference to `mc64ad_'<br>
> > > > ><br>
> > > > > which indicates that the SuperLU compiled with PETSc did<br>
> > > > > not get the fortran underscoring flag correct (which should<br>
> > > > > be no underscore with xlf). nm shows<br>
> > > > ><br>
> > > > > login1.surveyor$ nm<br>
> > > > > /gpfs/home/projects/facets/surveyor/contrib-xlc-9.0/petsc-3.1-p4-ser/lib/libsuperlu_4.0.a<br>
> > > > > | grep mc64<br>
> > > > > U mc64ad_<br>
> > > > > U mc64id_<br>
> > > > > mc64ad.o:<br>
> > > > > 0000000000000018 D mc64ad<br>
> > > > > 0000000000000030 D mc64bd<br>
> > > > > 0000000000000048 D mc64dd<br>
> > > > > 0000000000000060 D mc64ed<br>
> > > > > 0000000000000078 D mc64fd<br>
> > > > > 0000000000000000 D mc64id<br>
> > > > > 00000000000000c0 D mc64qd<br>
> > > > > 0000000000000090 D mc64rd<br>
> > > > > 00000000000000a8 D mc64sd<br>
> > > > > 00000000000000d8 D mc64ud<br>
> > > > > 00000000000000f0 D mc64wd<br>
> > > > > U mc64ad_<br>
> > > > > U mc64id_<br>
> > > > ><br>
> > > > > that the underscored symbol is being called, but the<br>
> > > > > underscore-free symbol is what was defined.<br>
> > > > ><br>
> > > > > My PETSc configure line was<br>
> > > > ><br>
> > > > > #!/bin/sh<br>
> > > > > /gpfs/home/cary/facetspkgs/builds-surveyor-xlc/facetspkgs/petsc-3.1-p4/ser/configure<br>
> > > > > \<br>
> > > > > --prefix=/gpfs/home/projects/facets/surveyor/contrib-xlc-9.0/petsc-3.1-p4-ser<br>
> > > > > \<br>
> > > > > --with-mpi=0 \<br>
> > > > > --with-debugging=0 \<br>
> > > > > --with-x=0 \<br>
> > > > > --with-cc='xlc_r' \<br>
> > > > > --with-cxx='xlC_r' \<br>
> > > > > --with-fc='xlf_r' \<br>
> > > > > --COPTFLAGS='-O2 -g' \<br>
> > > > > --download-superlu \<br>
> > > > > --with-lapack-lib=/home/projects/facets/intrepid/contrib/lapack-ser/lib/liblapack.a<br>
> > > > > \<br>
> > > > > --with-blas-lib=/home/projects/facets/intrepid/contrib/lapack-ser/lib/libblas.a<br>
> > > > > \<br>
> > > > > PETSC_DIR=/gpfs/home/cary/facetspkgs/builds-surveyor-xlc/facetspkgs/petsc-3.1-p4/ser<br>
> > > > > \<br>
> > > > > PETSC_ARCH=facets-ser \<br>
> > > > > --CFLAGS='-q64 -qlanglvl=redefmac' \<br>
> > > > > --CXXFLAGS='-q64 -qlanglvl=redefmac' \<br>
> > > > > --FFLAGS='-q64 -qextname=flush' \<br>
> > > > > --with-shared=1<br>
> > > > ><br>
> > > > > so I think that PETSc had enough info to figure out the underscoring.<br>
> > > > > Perhaps this is a bug.<br>
> > > > ><br>
> > > > > But regardless, is there a workaround?<br>
> > > > ><br>
> > > > > Thx....John<br>
> > > > ><br>
> > > ><br>
> > > ><br>
> > > ><br>
> > > ><br>
> > ><br>
> > ><br>
> ><br>
> ><br>
><br>
><br>
<br>
</div></div></blockquote></div><br>