Arm Fast Models are accurate, flexible programmer’s view models of Arm IP. They are used to build a virtual platform, either standalone, or as part of Hybrid Simulation environment within EDA partner environments. Use the virtual platform for software development and verification throughout the development process, even long before any real hardware is available.
You can connect the Arm Development Studio debugger to your virtual platform and interact with it as if were real hardware.
It is assumed that Arm Fast Models and Arm Development Studio are installed, and the appropriate license(s) configured.
If necessary refer to the appropriate Install Guide:
If you are unfamiliar with the Development Studio IDE, first see the Get started with Arm Development Studio learning path.
If you do not have access to Arm Fast Models, you can still learn how to connect to a custom virtual platform, using an FVP as supplied with Development Studio instead.
A number of ready made example systems are provided with Arm Fast Models, including all of the Fixed Virtual Platform (FVP) examples.
Use the supplied FVP_MPS2_Cortex-M3 Fast Models example, which is installed in the
FastModelsPortfolio_<version>\examples\LISA\FVP_MPS2\Build_Cortex-M3
directory of your Fast Models installation.
Open System Canvas, and select File > Load project from the menu.
Navigate to the FVP_MPS2_Cortex-M3.sgproj project as above. Click Open to load.
Click Build to create the FVP. The FVP executable will be in the appropriate subfolder depending on platform and compiler version, for example:
Win64-Release-VC2019\isim_system_Win64-Release-VC2019.exe
In the Arm Development Studio IDE menu, select File > New > Other > Configuration Database > Configuration Database, and give it a meaningful name. This creates a project folder where the debugger stores all user-made configurations.
Then select File > New > Other > Configuration Database > Model Configuration, which will be the actual configuration to create. When prompted, select the above Configuration Database to store in. Click Next.
You will be prompted to ask which debug interface to use, Iris or CADI. Iris is the default and recommended interface. Click Next.
You are then prompted to either launch the model, or browse for an already running model. Select Launch and connect to specific model.
Browse for the Fast Model executable. The debugger will append necessary command options to enable debug. This is the most straight forward option, and recommended for first time users.
A .mdf file will be created. Specify an appropriate manufacturer and platform name for identification, and click Import.
If not previously imported, use File > Import... > Arm Development Studio > Examples and Programming Libraries, and browse for startup_Cortex-M3_AC6 (use the text filter box to easily locate this).
Although this project is pre-configured to be debugged with the FVP supplied with Arm Development Studio, a new debug session will be configured for the newly built FVP.
Select File > New > Model Connection, and give it a meaningful name. It is recommended to associate with the specific project (startup_Cortex-M3_AC6).
Select the model configuration you created above (the text filter can assist if many targets defined), and click Finish.
You can again select to launch a new instance of the model or connect to an already running model. Select Launch a new model.
Navigate to the Files tab, and browse (within Workspace) to the startup_Cortex-M3_AC6.axf pre-built image. Then, in the Debugger tab, select Debug from entry point.
If connecting to an already running model with a loaded image (-a <image>), select Load symbols from file from the Files tab. In this scenario, in the Debugger tab, select Connect Only.
Click Debug to connect to the virtual platform, and commence your debug session.
When creating the model configuration, you can select to browse for a model running on local or remote hosts. As these models are not being launched by the debugger, additional command options are needed.
If this is selected, you must previously have launched the model with -I option to start the Iris server in the model.
If no port number (--port-number) is specified, recommend adding -p to output the port number used.
If this is selected (where the virtual platform is running on a different machine on the network), you must previously have launched the model with -I -A options to start the Iris server in the model and allow remote access.
If no port number (--port-number) is specified, recommend adding -p to output the port number used. You will be prompted for the server address (the machine running the virtual platform) and port number.
In the Connections tab, select Connect to already running model. Specify the Connection Address as localhost:<port> or hostname:<port> as appropriate.
If the model has an image loaded (-a <image>), select Load symbols from file from the Files tab. In this scenario, in the Debugger tab, you must select Connect Only.