Hi all,
I am new to the ASPECT. I have successfully installed the dockerized ASPECT in both my Macbook (x86, macOS14.5) and Raspberry Pi 5 (arm64, with Ubuntu Linux). My purpose in installing ASPECT in Raspberry Pi is not to run real simulations but to serve as a way for me to learn the Linux system. The successful running of ASPECT on Pi itself is pretty fun.
The general workflow I try to install fastscape is the following (after reading the post How to run cookbook "fastscape_eroding_box.prm"? and consulting with ChatGPT):
(1) docker pull the lastest ASPECT (3.1.0-pre) and docker run -it geodynamics/aspect
(2) Once I am in the docker (at /home/dealii/aspect), I clone the fastscape using git clone https://github.com/Djneu/fastscapelib-fortran
, and then create a “build” folder under this path.
(3) Under /home/dealii/aspect/fastscapelib-fortran/build, I run cmake -DBUILD_FASTSCAPELIB_SHARED=ON ..
and then make
. No errors and warnings.
(4) Then I goto /home/dealii/aspect and create a “build” folder, in which I run
cmake -DASPECT_WITH_FASTSCAPE=ON \
-DFASTSCAPE_DIR=/home/dealii/aspect/fastscapelib-fortran/build \
..
After running the above “cmake”, I see output such like:
"-- Using ASPECT_WITH_FASTSCAPE = ‘ON’ …
– FastScape library found at /home/dealii/aspect/fastscapelib-fortran/build … " and “-- Linking ASPECT against Fastscape …”
(5) Build ASPECT using make -j2
On Ubuntu Linux, no errors and no warnings. Complete at 100%.
(6) Then I do export GFORTRAN_CONVERT_UNIT='big_endian'
, source ~/.bashrc
, and confirm with echo $GFORTRAN_CONVERT_UNIT
. The output is big_endian without ’ ’ (see How to run cookbook "fastscape_eroding_box.prm"?)
(7) run “fastscape_eroding_box.prm”. Error shows “There is no such subsection to be entered: Mesh deformation.Fastscape”.
(8) I check the following things:
- libfastscapelib_fortran.so existed in the /home/dealii/aspect/fastscapelib-fortran/build
- run
ldd ./aspect
, I see all linked shared library but see no linked libfastscapelib_fortran.so.
I think I am getting close but somehow failed… that is so far what I can get on Raspberry Pi with Ubuntu. On my MacBook, step (5) was failed at 73% (sorry I forgot to record the error message).
Any suggestions are much appreciated! Thanks!