Re: [FLASH-USERS] orbit test problem

thanks. THis is not what I expected but it is a problem we found in other
parts of the code for TRU64. Thanks for tracking down. -andrew

On Wed, 23 Apr 2003, James Wadsley wrote:

>
> Hi,
>
> Thanks for your quick response. Unfortunately, I can't wait around
> for new releases as I am trying to get a simulation done for a
> deadline.
>
> I found the problem was with line 133 in MapParticlesToMesh.F90:
> if ((present(mode)) .and. (mode /= 0)) then
>
> It seems that the compaq compiler doesn't ignore the second
> test if present(mode) is false. Thus, if mode is not specified it
> causes a problem. I am using this (somewhat awkward) fix:
>
> if ((present(mode))) then
> if ((mode /= 0)) then
> ! addition mode, only zero boundary regions
>
> solnVec(ivar,il_bnd:il_bnd+nguard-1,:,:,:) = 0.
> solnVec(ivar,iu_bnd-nguard+1:iu_bnd,:,:,:) = 0.
> if (ndim >= 2) then
> solnVec(ivar,:,jl_bnd:jl_bnd+nguard-1,:,:)= 0.
> solnVec(ivar,:,ju_bnd-nguard+1:ju_bnd,:,:)= 0.
> endif
> if (ndim == 3) then
> solnVec(ivar,:,:,kl_bnd:kl_bnd+nguard-1,:) = 0.
> solnVec(ivar,:,:,ku_bnd-nguard+1:ku_bnd,:) = 0.
> endif
>
> else ! replacement mode, zero mesh variable before proceeding
>
> solnVec(ivar,:,:,:,:) = 0.
>
> endif
>
> else ! replacement mode, zero mesh variable before proceeding
>
> solnVec(ivar,:,:,:,:) = 0.
>
> endif
>
>
> Cheers,
>
> James.
>
> On Tue, 22 Apr 2003, Andrew Siegel wrote:
>
> >
> > Thanks for this message. We had some problems with 2.2 on TRU64 that
> > didn't manifest themselves on any of our testing platforms. We tracked
> > them down after getting a user bug report and fixed them for 2.3 (released
> > in ~2 weeks). The problem you mention isn't one that we found, but I
> > strongly suspect it is the same issue. I'll look into this and get back to
> > you asap ... -andrew siegel
> >
> >
> >
> > On Tue, 22 Apr 2003, James Wadsley wrote:
> >
> > >
> > > Hi,
> > >
> > > I am working on an AlphaServer SC running Tru64 unix with FLASH 2.2. We
> > > have run FLASH on several test problems so far including hydro, MHD and
> > > gravity.
> > >
> > > I was just trying the orbit test problem:
> > >
> > > ./setup orbit -auto -maxblocks=1000 -3d
> > >
> > > When I try to run it, it crashes with a segfault at line 149 of
> > > MapParticlesToMesh.F90 . I am not particularly sure where to start
> > > tracking this down. I have included the messages below.
> > >
> > > I am particularly interested in running using the particle modules with
> > > gravity and ultimately the cosmology modules as well. I'd be interested
> > > in the status of these. In particular, if the cosomology module works
> > > it would be really useful to have a simple test case for it in the
> > > standard setups.
> > >
> > > Thanks,
> > >
> > > James.
> > >
> > >
> > > prun -n 1 flash2
> > > read_parameters: ignoring unknown parameter "mgrid_nsmooth"...
> > > read_parameters: ignoring unknown parameter "mgrid_max_residual_norm"...
> > > read_parameters: ignoring unknown parameter "mgrid_print_norm"...
> > > read_parameters: ignoring unknown parameter "mpole_lmax"...
> > > read_parameters: ignoring unknown parameter "monotone"...
> > > read_parameters: ignoring unknown parameter "grav_boundary_type"...
> > > read_parameters: ignoring unknown parameter "eint_switch"...
> > > forrtl: severe (174): SIGSEGV, segmentation fault occurred
> > > 0: for__issue_diagnostic [0x120302c08]
> > > 1: for__signal_handler [0x120302024]
> > > 2: __sigtramp [0x1203cd5e0]
> > > 3: $modulemapparticlestomesh$mapparticlestomesh_
> > > [MapParticlesToMesh.F90: 149, 0x12010f030]
> > > 4: init_from_scratch_ [init_from_scratch.F90: 208, 0x120080f40]
> > > 5: init_flash_ [init_flash.F90: 290, 0x120074d84]
> > > 6: flash_ [flash.F90: 57, 0x1200736a0]
> > > 7: main [for_main.c: 203, 0x120122ecc]
> > > 8: __start [0x120056508]
> > > forrtl: error (76): IOT trap signal
> > > 0: for__issue_diagnostic [0x120302c08]
> > > 1: for__signal_handler [0x120301eb0]
> > > 2: __sigtramp [0x1203cd5e0]
> > > 3: __kill [0x1203b4424]
> > > 4: [0x1203b62a8]
> > >
> > > 5: __tis_raise [0x1203b3c84]
> > > 6: raise [0x1203cd4f0]
> > > 7: abort [0x1203a20a0]
> > > 8: for__issue_diagnostic [0x120302d44]
> > > 9: for__signal_handler [0x120302024]
> > > 10: __sigtramp [0x1203cd5e0]
> > > 11: $modulemapparticlestomesh$mapparticlestomesh_
> > > [MapParticlesToMesh.F90: 149, 0x12010f030]
> > > 12: init_from_scratch_ [init_from_scratch.F90: 208, 0x120080f40]
> > > 13: init_flash_ [init_flash.F90: 290, 0x120074d84]
> > > 14: flash_ [flash.F90: 57, 0x1200736a0]
> > > 15: main [for_main.c: 203, 0x120122ecc]
> > > 16: __start [0x120056508]
> > > prun: .//flash2 (pid 16820010) killed by signal 6
> > > prun: no core file for job 5803 in /local/core/rms/5803
> > >
> > >
> > >
> > >
> > >
> >
>
>
Received on Wed Apr 23 10:58:40 2003