2.1 Loading and running an application on an FVP

There are different ways to run an application on an FVP, for example from a command prompt, or from Model Debugger, or DS-5.

To run an FVP from the command prompt, change to the directory where your model is located. At the command prompt, enter the model name followed by the model options. To see all available options, use the --help option. The following options are commonly used when running an application on an FVP:

-a filename.axf
Specifies the ELF application to load.
--data filename.bin@address
Loads binary data into memory at the address you specify.
-C instance.parameter=value
Sets a single model parameter. For example, -C bp.flashloader0.fname=fip.bin loads a program into flash. To list all the available parameters, with their type, default value, and description, invoke the model with the --list-params option.
-f config_file.txt
Specifies the name of a plain text configuration file. Configuration files simplify managing multiple model parameters. You can set the same parameters using this option as with the -C option.
-S
Starts a CADI debug server. This option allows a CADI-enabled debugger, such as Model Debugger or DS-5 Debugger, to connect to the running model. The model waits for the debugger to connect before starting.

For example:

models_directory/FVP_Base_Cortex-A57x1 -a __image.axf -f params.txt

You can also launch and debug bare metal and Linux applications on an FVP from Model Debugger or DS-5 Debugger. These debuggers use CADI to communicate with the FVP, so you must use the -S option when launching the FVP.

Starting the model opens the FVP CLCD display, which shows the contents of the simulated color LCD framebuffer.

Non-ConfidentialPDF file icon PDF versionARM DUI0837I
Copyright © 2014-2016 ARM Limited or its affiliates. All rights reserved.