[petsc-dev] UMFPACK + Intel compiler

[Jed Brown]

On Sat, Jan 5, 2013 at 11:15 AM, Garth N. Wells <gnw20 at cam.ac.uk> wrote:

> Note quite sure how to delve into exactly what the compiler is doing,
> but doing some printing from the Makefile right before the compiler is
> first called:
>
>         @echo "Testing CFLAGS: $(CFLAGS)"
>         @echo "Testing C:      $(C)"
>         $(C) -DDINT -c ../Source/amd_aat.c -o amd_i_aat.o
>
> I get
>
>     Testing CFLAGS: -fPIC -wd1572 -mt_mpi -O3 -xHost -DUF_long=long
> long -DUF_long_max=ULONG_MAX -DUF_long_id='%lld' -DNCHOLMOD
>     Testing C:      mpicc -fPIC -wd1572 -mt_mpi -O3 -xHost
> -DUF_long=long long -DUF_long_max=ULONG_MAX -DUF_long_id='%lld'
> -DNCHOLMOD   -I../Include -I../../UFconfig
>     mpicc -fPIC -wd1572 -mt_mpi -O3 -xHost -DUF_long=long long
> -DUF_long_max=ULONG_MAX -DUF_long_id='"%lld"' -DNCHOLMOD
> -I../Include -I../../UFconfig -DDINT -c ../Source/amd_aat.c -o
> amd_i_aat.o
>    icc: error #10236: File not found:  'long'
>

Doesn't the same error occur with gcc? Do you have configure.log? What
"make" is this using?


> Double quotes don't seem to help. Changing 'long long' to 'long', it
> builds fine.
>

Unfortunately that doesn't work with LLP64 (like Windows).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20130105/d20d8ba0/attachment.html>