Skip to content

Latest commit

 

History

History
158 lines (142 loc) · 7.51 KB

README.Linux.md

File metadata and controls

158 lines (142 loc) · 7.51 KB
Raw

Linux/etc README for OpenModelica

Debian/Ubuntu Compile Cheat Sheet (or read on for the full guide)

echo deb http://build.openmodelica.org/apt precise nightly | sudo tee -a /etc/apt/sources.list
echo deb-src http://build.openmodelica.org/apt precise nightly | sudo tee -a /etc/apt/sources.list
sudo apt-get update
sudo apt-get build-dep openmodelica
git clone --recursive https://openmodelica.org/git-readonly/OpenModelica.git OpenModelica
cd OpenModelica
autoconf
./configure --with-omniORB
make -j4

How to compile on Linux/BSD (all from source)

autoconf
# Skip some pieces of software to ease installation and only compile the base omc executable
# If you have a working and compatible omc that is not on the PATH, you can use --with-omc=path/to/omc to speed up compilation
./configure --prefix=/usr/local --disable-modelica3d
make
sudo make install

But first you need to install dependencies:

  • autoconf, automake, libtool, pkgconfig, g++, gfortran (pretty standard compilers)
  • boost (optional; used with configure --with-cppruntime)
  • clang, clang++ (optional, but highly recommended; if you use gcc instead, use gcc 4.4 or 4.9+, not 4.5-4.8 as they are very slow)
  • cmake
  • hwloc (optional; queries the number of hardware CPU cores instead of logical CPU cores)
  • Java JRE (JDK is option; compiles the Java CORBA interface)
  • Lapack/BLAS
  • libhdf5 (optional part of the MSL tables library supported by few other Modelica tools, so it does not do much)
  • libexpat (it's actually included in the FMIL sources which are included... but we do not compile those and it's better to use the OS-provided dynamically linked version)
  • lpsolve55
  • omniORB or mico (optional; CORBA is used by OMOptim, OMShell, and OMPython)
  • Sundials (optional; adds more numerical solvers to the simulation runtime)

Setting your environment for compiling OpenModelica

If you plan to use mico corba with OMC you need to:

  • set the PATH to path/to/mico/bin (for the idl compiler and mico-cpp)
  • set the LD_LIBRARY_PATH to path/to/installed/mico/lib (for mico libs)
  • set the PATH (for executables: idl, mico-cpp and mico-config):
export PATH=${PATH}:/path/to/installed/mico/bin

To Compile OpenModelica

Run:

autoconf
# One of the following configure lines
./configure --with-omniORB=/path/to/omniORB (if you want omc to use omniORB corba)
./configure --with-CORBA=/path/to/mico (if you want omc to use mico corba)
./configure --without-CORBA            (if you want omc to use sockets)

in the source directory. Make sure that all makefiles are created. Check carefully for error messages.

make

After the compilation the results are in the path/to/trunk/build. To run the testsuite, you need to use the superproject OpenModelica.git, or clone OpenModelica-testsuite.git into the root directory under the name testsuite.

make -C testsuite omc-diff ReferenceFiles
cd testsuite/runtests && ./runtests.pl

If you run into problems read the GENERAL NOTES below and if that does not help, subscribe to the OpenModelicaInterest list and then sent us an email at [email protected].

How to run

Here is a short example session. This example uses OMShell-terminal, but OMShell, mos-scripts, or OMNotebook work the same way.

$ cd trunk/build/bin
$ ./OMShell-terminal
OMShell Copyright 1997-2015, Open Source Modelica Consortium (OSMC)
Distributed under OMSC-PL and GPL, see www.openmodelica.org

To get help on using OMShell and OpenModelica, type "help()" and press enter
Started server using:omc -d=interactive > /tmp/omshell.log 2>&1 &
>>> loadModel(Modelica)
true
>>> getErrorString()
""
>> instantiateModel(Modelica.Electrical.Analog.Basic.Resistor)
"class Modelica.Electrical.Analog.Basic.Resistor \"Ideal linear electrical resistor\"
  Real v(quantity = \"ElectricPotential\", unit = \"V\") \"Voltage drop between the two pins (= p.v - n.v)\";
  Real i(quantity = \"ElectricCurrent\", unit = \"A\") \"Current flowing from pin p to pin n\";
  Real p.v(quantity = \"ElectricPotential\", unit = \"V\") \"Potential at the pin\";
  Real p.i(quantity = \"ElectricCurrent\", unit = \"A\") \"Current flowing into the pin\";
  Real n.v(quantity = \"ElectricPotential\", unit = \"V\") \"Potential at the pin\";
  Real n.i(quantity = \"ElectricCurrent\", unit = \"A\") \"Current flowing into the pin\";
  parameter Boolean useHeatPort = false \"=true, if HeatPort is enabled\";
  parameter Real T(quantity = \"ThermodynamicTemperature\", unit = \"K\", displayUnit = \"degC\", min = 0.0, start = 288.15, nominal = 300.0) = T_ref \"Fixed device temperature if useHeatPort = false\";
  Real LossPower(quantity = \"Power\", unit = \"W\") \"Loss power leaving component via HeatPort\";
  Real T_heatPort(quantity = \"ThermodynamicTemperature\", unit = \"K\", displayUnit = \"degC\", min = 0.0, start = 288.15, nominal = 300.0) \"Temperature of HeatPort\";
  parameter Real R(quantity = \"Resistance\", unit = \"Ohm\", start = 1.0) \"Resistance at temperature T_ref\";
  parameter Real T_ref(quantity = \"ThermodynamicTemperature\", unit = \"K\", displayUnit = \"degC\", min = 0.0, start = 288.15, nominal = 300.0) = 300.15 \"Reference temperature\";
  parameter Real alpha(quantity = \"LinearTemperatureCoefficient\", unit = \"1/K\") = 0.0 \"Temperature coefficient of resistance (R_actual = R*(1 + alpha*(T_heatPort - T_ref))\";
  Real R_actual(quantity = \"Resistance\", unit = \"Ohm\") \"Actual resistance = R*(1 + alpha*(T_heatPort - T_ref))\";
equation
  assert(1.0 + alpha * (T_heatPort - T_ref) >= 1e-15, \"Temperature outside scope of model!\");
  R_actual = R * (1.0 + alpha * (T_heatPort - T_ref));
  v = R_actual * i;
  LossPower = v * i;
  v = p.v - n.v;
  0.0 = p.i + n.i;
  i = p.i;
  T_heatPort = T;
  p.i = 0.0;
  n.i = 0.0;
end Modelica.Electrical.Analog.Basic.Resistor;
"
>> a:=1:5;
>> b:=3:8
{3,4,5,6,7,8}
>>> a*b

>>> getErrorString()
"[<interactive>:1:1-1:0:writable] Error: Incompatible argument types to operation scalar product in component <NO COMPONENT>, left type: Integer[5], right type: Integer[6]
[<interactive>:1:1-1:0:writable] Error: Incompatible argument types to operation scalar product in component <NO COMPONENT>, left type: Real[5], right type: Real[6]
[<interactive>:1:1-1:0:writable] Error: Cannot resolve type of expression a * b. The operands have types Integer[5], Integer[6] in component <NO COMPONENT>.
"
>> b:=3:7;
>> a*b
85
>>> listVariables()
{b, a}
>>

CentOS 6 Hints (RPM, command-line only; for clients, add CORBA, readline)

yum install tar gcc-c++ autoconf sqlite-devel java expat-devel lpsolve-devel lapack-devel make patch gettext

also needs cmake > 2.8; not in default repos; try to install an rpm manually if needed

autoconf
./configure
make -j8

GENERAL NOTES:

  • Fedora Core 4 has a missing symlink. To fix it, in /usr/lib do:
ln -s libg2c.so.0 libg2c.so

Otherwise the testsuite will fail when generating simulation code.

  • On some Linux systems when running simulate(Model, ...) the executable for the Model enters an infinite loop. To fix this, add -ffloat-store to CFLAGS

Last updated 2015-06-10. Much is still outdated.