McSim-Cycle-accurate-NoC: Manycore platform Simulation tool for NoC-based platform at a Cycle-accurate level
This repository contains a simulator able to simulate embedded applications described using the AMALTHEA application model on top of NoC-based manycore architectures in a cycle accurate way. This simulator is based on NoCTweak and has been developed by the LIRMM laboratory of Montpellier.
To get the simulator you must clone this repository and its submodules. To clone the repository if you have a GitHub account with an SSH key registered use git clone git@github.com:DreamCloud-Project/McSim-Cycle-accurate-NoC.git. Else use git clone https://github.com/DreamCloud-Project/McSim-Cycle-accurate-NoC.git. Then do cd McSim-Cycle-accurate-NoC followed by git submodule init followed by git submodule update to clone submodules.
To ease the usage of the simulator, two python scripts are provided:
- compile.py for compilation
- simulate.py for launching the simulation
The requirements for using these scripts are the following ones:
- have a Linux system
- have a C++ compiler
- have CMake installed on your system (https://cmake.org)
- have SystemC 2.3.1
- download SystemC (Core SystemC Language and Examples) from here http://accellera.org/downloads/standards/systemc/systemc-license-agreement
- decompress the archive
- run
configureand thenmake install - you should see a new
includefolder and a newlib-linux64folder in your root folder of SystemC - define the
SYSTEMC_HOMEenvironment variable to the root folder of SystemC
- have the xerces-c-dev library installed in standard includes and libs folders (using
apt-get install libxerces-c-devfor example on Ubuntu) or have xerces-c-dev library in a custom folder and defineXERCES_HOMEenvironment variable
Compilation is done through the compile.py script which documentation is the following:
>> ./compile.py --help
usage: compile.py [-h] [-v] {build,clean} ...
Cycle accurate simulator compiler script
optional arguments:
-h, --help show this help message and exit
-v, --verbose enable verbose output
valid subcommands:
{build,clean}
To run a particular simulation, just run the simulate.py script. By
default this script runs one iteration of the Demo Car application on
a 4x4 NoC using ZigZag mapping, First Come First Serve (fcfs)
scheduling and without repeating periodic runnables. You can play
with all these parameters which documentation is the following:
>>./simulate.py --help
usage: simulate.py [-h] [-d] [-bs {1,2,4,8,16}] [-da {DC}]
[-ca CUSTOM_APPLICATION] [-f FREQ] [-mf MODES_FILE]
[-m MAPPING_STRATEGY [MAPPING_STRATEGY ...]] [-mw]
[-mww MICRO_WORKLOAD_WIDTH] [-mwh MICRO_WORKLOAD_HEIGHT]
[-np] [-nvc {1,2,4,8,16}] [-o OUTPUT_FOLDER] [-r]
[-s {prio}] [-v] [-x ROWS] [-y COLS]
Cycle accurate simulator runner script
optional arguments:
-h, --help show this help message and exit
-d, --syntax_dependency
consider successive runnables in tasks call graph as
dependent
-bs {1,2,4,8,16}, --buffer_size {1,2,4,8,16}
specify the number of slots in buffers (default is 16)
-da {DC}, --def_application {DC}
specify the application to be simulated among the
default ones
-ca CUSTOM_APPLICATION, --custom_application CUSTOM_APPLICATION
specify a custom application file to be simulated
-f FREQ, --freq FREQ specify the frequency of cores in the NoC. Supported
frequency units are Hz, KHz, MHz and GHz e.g 400MHz or
1GHz
-mf MODES_FILE, --modes_file MODES_FILE
specify a modes switching file to be simulated
-m MAPPING_STRATEGY [MAPPING_STRATEGY ...], --mapping_strategy MAPPING_STRATEGY [MAPPING_STRATEGY ...]
specify the mapping strategy used to map runnables on
cores and labels on memories. Valide strategies are
['MinComm', 'Static', 'ZigZag', 'Random']
-mw, --micro_workload
simulate a micro workload of the application instead
of the real application
-mww MICRO_WORKLOAD_WIDTH, --micro_workload_width MICRO_WORKLOAD_WIDTH
the width of the simulated micro workload. To be used
with -mw only
-mwh MICRO_WORKLOAD_HEIGHT, --micro_workload_height MICRO_WORKLOAD_HEIGHT
the height of the simulated micro workload. To be used
with -mw only
-np, --no_periodicity
run periodic runnables only once
-nvc {1,2,4,8,16}, --nb_virtual_channels {1,2,4,8,16}
specify the number of virtual channels (default is 8)
-o OUTPUT_FOLDER, --output_folder OUTPUT_FOLDER
specify the absolute path of the output folder where
simulation results will be generated
-r, --random replace constant seed used to generate distributions
by a random one based on current time
-s {prio}, --scheduling_strategy {prio}
specify the scheduling strategy used by cores to
choose the runnable to execute
-v, --verbose enable verbose output
-x ROWS, --rows ROWS specify the number of rows in the NoC
-y COLS, --cols COLS specify the number of columns in the NoC
This software is made available under the GNU Lesser General Public License v3.0
Report bugs at: mcsim-support@lirmm.fr
(C)2016 CNRS and Université de Montpellier