search

UniMainzGeo / LaMEM

LaMEM (Lithosphere and Mantle Evolution Model) - a 3D parallel code to simulate geodynamic and geomechanical processes
MIT License
50 stars 16 forks source link

Segmentation Violation error when using plume inflow boundaries #18

Open jaime-almeida opened 8 months ago

jaime-almeida commented 8 months ago

Hello!

I'm trying to use LaMEM to run a 3D model which requires the existence of a plume boundary. I've used these boundaries before with success, but after updating the code recently I've been met with this error:


[124]PETSC ERROR: ------------------------------------------------------------------------
[124]PETSC ERROR: Caught signal number 11 SEGV: Segmentation Violation, probably memory access out of range
[124]PETSC ERROR: Try option -start_in_debugger or -on_error_attach_debugger
[124]PETSC ERROR: or see https://petsc.org/release/faq/#valgrind and https://petsc.org/release/faq/
[124]PETSC ERROR: configure using --with-debugging=yes, recompile, link, and run
[124]PETSC ERROR: to get more information on the crash.
[124]PETSC ERROR: Run with -malloc_debug to check if memory corruption is causing the crash.
--------------------------------------------------------------------------

The plume inflow was defined like this:

# Set a plume in the plateau region:
        Plume_InflowBoundary    =   1   
        Plume_Type                     =      Inflow_Type
    Plume_Dimension           = 3D  
    Plume_Phase                 =   7   
    Plume_Depth             =   0
    Plume_Temperature   =   1700    
    Plume_Inflow_Velocity   =   15.0
    Plume_VelocityType        = Gaussian
    Plume_Center                =   -200.0 -100.0
    Plume_Radius                =   50

The only change from my previously used plumes are the coordinates and the number of the inflow phase. It was attempting to run on 128 cores.

boriskaus commented 8 months ago

Can you attach the input file so we can actually reproduce this? Does this also occur for the test that we have in place for this?

boriskaus commented 7 months ago

how was this solved?

jaime-almeida commented 7 months ago

Used the conditions on your test model and for now it seems to have fixed my problem. Changed the plume type from Gaussian. Also removed the plume depth parameter.

On 7 Mar 2024, 17:01, at 17:01, Boris Kaus @.***> wrote:

how was this solved?

-- Reply to this email directly or view it on GitHub: https://github.com/UniMainzGeo/LaMEM/issues/18#issuecomment-1984010050 You are receiving this because you modified the open/close state.

Message ID: @.***>

jaime-almeida commented 6 months ago

Sorry to reopen this. I did it earlier but then remembered to first update LaMEM to the latest version to make sure it hadn't been fixed already. I did previously "fix" my issue by not using a Gaussian plume but it really wasn't a fix.

I've recently tried to run older models with working plumes and they raise the same Segmentation Violation errors when attempted. The test models do work though, I've ran a simple gaussian plume there without issue. I I've attached the input file as well as the MATLAB script I used to build the markers.

inputs.zip