[FLASH-USERS] Seg Fault in Optimized UHD Solver

[Sean Couch]

Hi James,

I tried a 3D Sedov with the same setup line (minus the +parallelIO) in 4.2.2 and I didn’t get a segfault.  The transOrder option is obsolete and should always be =1 in the code.  The fact that it even appears in the code anymore is vestigial.  I would suggest removing it from your Config file, then resetup without the -noclobber option.  It can sometimes cause strange behavior.

Sean


-----------------------------------------------------------
Sean M. Couch
Theoretical Astrophysics Including Relativity
California Institute of Technology
Mail Code 350-17
Pasadena, CA  91125
(626) 395-4282
www.tapir.caltech.edu/~smc <http://www.tapir.caltech.edu/~smc>
> On Dec 18, 2014, at 3:43 PM, James Guillochon <jfg at ucolick.org> wrote:
> 
> Hi all, I'm experiencing a segmentation fault at line 117 in hy_uhd_upwindTransverseFlux.F90 in FLASH 4.2.2 when taking the very first hydro step. It appears that the transverse flux array is not defined for some reason. I've also noticed, oddly, that the "transOrder" option does not appear with the +uhd flag, but only when I include the unit in my Config file, however, adding this unit does not solve the seg fault. When I switch to the unoptimized solver (+uhdold), the code *does* work. Here is my setup line:
> 
> ./setup MultiTidal -noclobber -maxblocks=500 -nxb=8 -nyb=8 -nzb=8 -auto -3d -objdir=object_multitidal +newMpole +parallelIO +pm4dev +uhd
> 
> Included units:
> 
> REQUIRES Driver
> REQUIRES Simulation/SimulationComposition
> REQUIRES Multispecies/MultispeciesMain
> REQUIRES physics/sourceTerms/Cool/CoolMain
> REQUIRES IO
> REQUIRES IO/IOMain/hdf5/parallel/PM
> 
> And modified core files:
> 
> Simulation_initSpecies.F90
> gr_sanitizeDataAfterInterp.F90
> Simulation_sendOutputData.F90
> gr_markVarThreshold.F90
> Gravity_accelOneRow.F90
> gr_mpoleCen3Dcartesian.F90
> Simulation_initBlock.F90
> Simulation_finalize.F90
> Simulation_data.F90
> Simulation_init.F90
> Gravity_potentialListOfBlocks.F90
> IO_writePlotfile.F90
> Particles_sinkAdvanceParticles.F90
> Driver_sourceTerms.F90
> Grid_markRefineDerefine.F90
> Particles_sinkAccelSinksOnGas.F90
> pt_sinkAccelSinksOnSinks.F90
> 
> Any help would be much appreciated!

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20141220/8afeeaf7/attachment-0001.htm>