[FLASH-BUGS] FLASH 2.2 bugs (fwd)

[Robi Banerjee]

Hey, 

> ---------- Forwarded message ----------
> Date: Thu, 16 Jan 2003 17:23:03 -0600
> From: Alvaro Caceres <acaceres at uchicago.edu>
> To: Sean Matt <matt at physics.mcmaster.ca>
> Subject: Re: [FLASH-BUGS] FLASH 2.2 bugs
> 
> Hi Sean,
> 
> 
> I'm trying to reproduce this bug, could you tell me how you're calling
> setup and what your compiler flags are? Thanks,

  we used the following flags for the setup of our 'Evrard' problem

         ./setup evrard -auto -3d -maxblocks=1024

  and we used the 'f95' fortran compiler with the options

         -fast  -r8 -i4

  I attached the Makefile.h file to this e-mail for the
case you need more informations about our compiling options.

  thanks,
Robi


> 
> Alvaro Caceres
> 
> Sean Matt wrote:
> 
> I forgot to mention that the problem runs for a few seconds, writes the
> first checkpoint file, gives some multigrid non-converging warnings, then
> crashes.  Here's some of the screen output:
> 
>   INITIAL TIMESTEP =   1.000000000000000E-004
>  *** Wrote output to evrardhdf5_chk_0000
>  number of variables found for plotfile storage =            4
>  dens pres velx temp
>  *** Wrote output to evrardhdf5_plt_cnt_0000 ***
>  multigrid:  WARNING:  V-cycles not converging
>  multigrid:  WARNING:  V-cycles not converging
>  multigrid:  WARNING:  V-cycles not converging
>  multigrid:  WARNING:  V-cycles not converging
>  multigrid:  WARNING:  V-cycles not converging
>  multigrid:  WARNING:  V-cycles not converging
>  multigrid:  WARNING:  V-cycles not converging
>  multigrid:  WARNING:  V-cycles not converging
> forrtl: error (75): floating point exception
> forrtl: error (75): floating point exception
> forrtl: error (75): floating point exception
> forrtl: error (75): floating point exception
>    0: for__issue_diagnostic [0x12036f0a8]
>    1: for__signal_handler [0x12036df10]
>    2: __sigtramp [0x120438750]
>    3: $modulehydro_1d$hydro_1d_    0: for__issue_diagnostic [0x12036f0a8]
>    1: for__signal_handler [0x12036df10]
>    2: __sigtramp [0x120438750]
>    3: $modulemonot$monot_    0: for__issue_diagnostic [0x12036f0a8]
>    1: for__signal_handler [0x12036df10]
>    2: __sigtramp [0x120438750]
>    3: $modulemonot$monot_    0: for__issue_diagnostic [0x12036f0a8]
>    1: for__signal_handler [0x12036df10]
>    2: __sigtramp [0x120438750]
>    3: $modulemonot$monot_ [hydro_1d.F90: 341, 0x1200927f0]
>    4: $modulehydrosweep$hydro_sweep_ [monot.F90: 68, 0x1200d77d0]
>    4: $moduleintrfc$intrfc_ [monot.F90: 68, 0x1200d77d0]
>    4: $moduleintrfc$intrfc_ [monot.F90: 68, 0x1200d77d0]
>    4: $moduleintrfc$intrfc_ [hydro_sweep.F90: 381, 0x1200a7110]
>    5: $hydro$hydro_3d_ [intrfc.F90: 253, 0x1200bdd00]
> 
> ...
> 
>         -Sean
> 
> On Thu, 16 Jan 2003, Jonathan Dursi wrote:
> 
> > Hi, Sean:
> >
> > Every time we use a new compiler, we end up finding a couple things in
> > the code that the other compilers had let us get away with.  As Mike has
> > mentioned, about a month ago we started using a Compaq ourselves, and
> > found the first two runtime errors you had pointed out.  They're fixed,
> > and will be part of the new incremental release which should be out soon.
> > We hadn't seen the magnetic resistivity bug; that's fixed now too.
> >
> > I've been running a few problems on a Compaq machine, though, and
> > haven't seen that third runtime error you describe.  So that we can try
> > to reproduce it, can you tell me what problem you were running and how
> > long into the run it was before the error occured?  A flash.par and the
> > setup information would be helpful.
> >
> >       - Jonathan
> >
> 
> --
> 
> _______________________________________________________________________
> Sean P. Matt                                 CITA National Fellow
> phone: (905) 525-9140 x 23189                Physics & Astronomy Dept.
> http://www.astro.washington.edu/matt         McMaster University
> -----------------------------------------------------------------------

-------------- next part --------------
# FLASH Definitions for Idra.mcmaster.ca OSF5.1

#	Compiler and linker commands

FCOMP   = f95 -cpp 
CCOMP   = cc
CPPCOMP = CC
LINK    = f90  

#	Compiler flags

PP     = -D

FFLAGS_OPT   =  -c -fast  -r8 -i4 -I/usr/mpi/include
FFLAGS_DEBUG =  -c -g -r8   -i4 -I/usr/mpi/include
FFLAGS_TEST  =  -c -r8  -i4  -I/usr/mpi/include

CFLAGS  = -c -I/work/smatt/lib/5-1.4.4-osf1/include
#-I/usr/mpi/include

#	Linker flags

LFLAGS = -fast -r8 -i4 -non_shared -o

#	Libraries to include

# This doesn't work
#LIB =  -L -L/work/colinm/lib -L/home/colinm/lib -L/usr/lib -L/usr/local/lib -L/usr/shlib  -L/usr/ccs/lib -L/usr/lib/cmplrs/cc -L/var/shlib -lmpi -lelan -lelan3 
#LIB_HDF4= -lmfhdf -ldf -ljpeg -lz

#LIB_HDF4= -L/usr/local/lib -lmfhdf  /work/colinm/lib/libdf.a  -lz -ljpeg 
LIB_HDF5= -L/work/smatt/lib/5-1.4.4-osf1/lib -lhdf5

#-lhdf5_fortran -lhdf5_cpp

LIB = -L -L/work/colinm/lib -L/usr/lib -L/usr/shlib  -L/usr/ccs/lib -L/usr/lib/cmplrs/cc -L/var/shlib -lmpi -lelan -lelan3 -lrmscall -lmach -lz $(LIB_HDF5)

#       Additional commands                                                

#	Additional machine-dependent object files

MACHOBJ = 

#	Additional commands

MV = mv -f
AR = ar -r
RM = rm -f
CD = cd
RL = ranlib
ECHO = echo