[FLASH-BUGS] FLASH 2.2 bugs

[Sean Matt]

Mike,

	Thanks for the quick response.  We are running on Tru64 (OSF1
V5.1) using a Compaq fortran compiler (V5.5-1877-48BBF).  Please let me
know if you require more information.

	-Sean

On Wed, 15 Jan 2003, Mike Zingale wrote:

> Thank you for the e-mail.  What compilers and OS are you using?  I'll look
> into the runtime problems you pointed out if we can reproduce them.  I
> think we've run into that runtime_parameters, but the fix may not have
> made it into the release.  I think that only affected Compaq machines.
> 
> Mike
> 
> ------------------------------------------------------------------------------
> Michael Zingale
> UCO/Lick Observatory
> UCSC
> Santa Cruz, CA 95064
> 
> phone:  (831) 459-5246 
> fax:    (831) 459-5265
> e-mail: zingale at ucolick.org
> web:    http://www.ucolick.org/~zingale
> 
> On Wed, 15 Jan 2003, Sean Matt wrote:
> 
> > Hello, my name is Sean Matt, and Robi Banerjee and I are part of the
> > McMaster University Computational Astrophysics group.  The group began
> > testing and use of the FLASH code since roughly 9 months ago.  Lately,
> > we have been working with the latest release, FLASH2.2, but have had
> > some problems.  Below, I list some bugs we have found.  In subsequent
> > emails, I'll list some of the less important issues we've noticed.
> > 
> > 	-Sean
> > 
> > 
> > 
> > We have identified numerous bugs in the new release of FLASH2.2.
> > These include one compile-time bug, and at least three run-time bugs.
> > They are as follows:
> > 
> > 
> > In the file "magnetic_resistivity.F90" in
> > "source/materials/magnetic_resistivity/" the relative order of the
> > declaration of some variables was changed from FLASH2.1 to FLASH2.2.
> > In particular, the declaration of
> > 
> >   real, DIMENSION(ns)    :: Zi
> > 
> > appears BEFORE the declaration of 
> > 
> >   integer                :: ns
> > 
> > in version 2.2.  This resulted in a compile-time error for us.  We fixed
> > it by switching the order of the variable declaration so that "ns" is
> > defined first.
> > 
> > 
> > After compiling, we get runtime errors.  We have found fixes to two of
> > them, but there is still at least one more.  The ones we have found
> > are as follows:
> > 
> > 
> > 1) runtime error in module
> > 
> >         runtime_parameters.F90
> > 
> >    error because the the newly introduced optional
> >    variable 'state' is explored while not set.
> > 
> >    we fixed by changing all appearances of
> > 
> >       if (present(state) .and. state == ...
> > 
> >    to
> > 
> >       if (present(state) ) then
> >          if (state == ...
> > 
> > 
> > 2) runtime error
> > 
> >    error because the the variable 'eos_mode' is set
> >    during runtime in 'init_global_parms.F90' but is
> >    of type 'parameter' in
> > 
> >       global_runtime_dec.F90
> > 
> >    we fixed by changing the line
> > 
> >      integer, parameter :: eos_mode = 2
> > 
> >    to
> > 
> >      integer, save      :: eos_mode
> > 
> > 
> > 3) we still get the following runtime error:
> > 
> > forrtl: error (75): floating point exception
> >    0: for__issue_diagnostic [0x1203918b8]
> >    1: for__signal_handler [0x120390720]
> >    2: __sigtramp [0x12045af60]
> >    3: $modulemonot$monot_ [monot.F90: 68, 0x1200dd240]
> >    4: $moduleintrfc$intrfc_ [intrfc.F90: 248, 0x1200c3670]
> >    5: $modulehydro_1d$hydro_1d_ [hydro_1d.F90: 275, 0x120093818]
> >    6: $modulehydrosweep$hydro_sweep_ [hydro_sweep.F90: 381, 0x1200abb60]
> >    7: $hydro$hydro_3d_ [hydro.F90: 109, 0x12008b620]
> >    8: evolve_ [evolve.F90: 69, 0x120086568]
> >    9: flash_ [flash.F90: 94, 0x120077480]
> >   10: main [for_main.c: 203, 0x1201b259c]
> >   11: __start [0x120058828]
> > 
> > 
> > 
> > 
> 

-- 

_______________________________________________________________________
Sean P. Matt				     CITA National Fellow
phone: (905) 525-9140 x 23189		     Physics & Astronomy Dept.
http://www.astro.washington.edu/matt	     McMaster University
-----------------------------------------------------------------------