[yt-users] Enzo compile under Debian 6.04 or Ubuntu 12.04 via YT script

Matthew Turk matthewturk at gmail.com
Thu Sep 13 05:34:21 PDT 2012


Hi Carlos,

This is more a question for enzo-users (
https://groups.google.com/forum/#!forum/enzo-users ) but I'll try to
answer your question here -- if it doesn't work, I would suggest you
search the enzo group and pose your question there.

On Thu, Sep 13, 2012 at 5:44 AM, CCAC South
<ccacsouthengredtech at gmail.com> wrote:
> Hi all,
>
> I am new to YT and Enzo.  Currently I am trying to get Enzo to compiled to
> run  the test simulations and analyze with YT.  I  was able to get YT
> installed and working and go thru some of the analysis examples.
>
> The Enzo package that downloaded with YT requires several dependencies.  Is
> there anyone that is currently running Debian 6.04 or Ubuntu 12.04 that can
> provide some help in listing the compiler dependencies?
>
> My current compiler output is below.
>
> gmake
>
> Updating DEPEND
> gmake: [dep] Error 127 (ignored)
> Compiling auto_show_config.C
> Compiling auto_show_flags.C
> Compiling auto_show_version.C
> Compiling s90_st1.F90
>
> gfortran  -c -o s90_st1.o -DLINUX -DH5_USE_16_API   -D__max_subgrids=100000
> -D__max_baryons=20 -D__max_cpu_per_node=8 -D__memory_pool_size=100000
> -DINITS64 -DLARGE_INTS -DCONFIG_PINT_8 -DIO_64     -DUSE_MPI
> -DCONFIG_PFLOAT_8 -DCONFIG_BFLOAT_8  -DUSE_HDF5_GROUPS    -DTRANSFER
> -DNEW_GRID_IO -DFAST_SIB -DBITWISE_IDENTICALITY -DFLUX_FIX
> -DSET_ACCELERATION_BOUNDARY -fno-second-underscore   -g s90_st1.F90
>
> s90_st1.F90:16.18:
>
>         call fftn(a, sn, inv=.false. )
>                   1
> Error: Type mismatch in argument 'array' at (1); passed COMPLEX(4) to
> COMPLEX(8)
> s90_st1.F90:21.18:
>
>         call fftn(a, sn, inv=.true. )
>                   1
> Error: Type mismatch in argument 'array' at (1); passed COMPLEX(4) to
> COMPLEX(8)
>
> gmake: *** [s90_st1.o] Error 1
>
>
> Thanks for any help or pointers.

I believe this is a result of incomplete settings for MACH_FFLAGS_* in
your Makefile.  For instance, here are mine:

MACH_FFLAGS_INTEGER_32 =
MACH_FFLAGS_INTEGER_64 = -fdefault-integer-8
MACH_FFLAGS_REAL_32    =
MACH_FFLAGS_REAL_64    = -fdefault-real-8 -fdefault-double-8

This works for me with gfortran.  You may need to "make ; make clean".
 -fdefault-double-8 is in particular a recent setting that may be
missing from your makefile.

-Matt

>
> Carlos
>
>
>
>
>
>
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>



More information about the yt-users mailing list