Closed cshsgy closed 1 year ago
Thanks for writing @cshsgy. Are you building KPP on a Mac? I've noticed this issue on MacOS as well. I think it may have to do with the much lower amount of stack memory on MacOS than on Linux-based systems. You can try setting
export OMP_STACKSIZE=500m
which should set the stack memory to the maximum value on the system, and then rebuild again.
Also, if you don't need the stoichometric numbers, you can also disable that option with
#STOICHIOM off
Also tagging @RolfSander @jimmielin
Re: stack memory on MacOSX, we have this documentation: https://kpp.readthedocs.io/en/stable/getting_started/01_installation.html#request-maximum-stack-memory
Hi. I am using the latest KPP on a Windows Linux Subsystem (Ubuntu) and I also encountered the exact problem as well. I managed to set the stack limit to unlimit and recompiled the source code of KPP. But the problem persists.
When I check the small_strato.log
file and generated .f90
files, it seems that they are interrupted. For example, below are line 71 to the rear of the file small_strato.log
:
### Subroutines ---------------------------------------
SUBROUTINE Fun ( V, F, RCT, Vdot, Aout )
SUBROUTINE Fun_SPLIT ( V, F, RCT, Vdot, P_VAR, D_VAR, Aout )
SUBROUTINE CalcStoichNum ( StoichNum )
SUBROUTINE Jac_SP ( V, F, RCT, JVS )
SUBROU
Thanks for writing @cshsgy. Are you building KPP on a Mac? I've noticed this issue on MacOS as well. I think it may have to do with the much lower amount of stack memory on MacOS than on Linux-based systems. You can try setting
export OMP_STACKSIZE=500m
which should set the stack memory to the maximum value on the system, and then rebuild again.
Also, if you don't need the stoichometric numbers, you can also disable that option with
#STOICHIOM off
Also tagging @RolfSander @jimmielin
Hi @yantosca thanks for your reply! Yes I am using KPP in a UNIX-based system, but a x86-64 Ubuntu system. I tried your suggestion of changing the OMP stack size but also tried to set the ulimit, neither of them work and the error persists. What is strange to me is that, the log file shows nothing (an empty file created), which makes it quite hard to trace down the error. Thanks again for your kind help :D
hello @cshsgy ! i was wondering if you solved the problem already ! I am facing the similar issue on my PC. I tried much to solve this but i failed to do so. can you help me ?
@prithvi09, there's a 2 line change in this pull request: https://github.com/KineticPreProcessor/KPP/pull/67, which fixes the stack overflow and allows the STOICMAT option to run for me. I'm not sure if this has been validated or will get pulled into KPP at some point, but it has worked for my purposes
hello @cshsgy ! i was wondering if you solved the problem already ! I am facing the similar issue on my PC. I tried much to solve this but i failed to do so. can you help me ?
Hi! For me eventually ulimit -s unlimited solved the problem. I have to say the tutorial was poorly organized and we have to go to the very bottom of the page to find out the solution to this problem.
This should be now solved by PR #67, which has been merged into KPP 3.0.1.
Hi, I was trying out the examples of KPP. However, an error comes up when I came to the compiling stage:
I used the files in the example folder, and they all fail at the same place. Is there a way to trace where the problem happens or does anyone have any ideas on how I can fix this?