Test With FVP

FVP

The firmware build can be executed on the Neoverse N2 Reference Design FVP that you can download from Arm Ecosystem FVPs .

Set up the FVP

Download the FVP from the previous page, or directly with:

    

        
        
            wget https://developer.arm.com/-/media/Arm%20Developer%20Community/Downloads/OSS/FVP/Neoverse-N2/Neoverse-N2-11-24-12/FVP_RD_N2_11.24_12_Linux64.tgz
        
    

Unpack the tarball and run the install script:

    

        
        
            tar -xf FVP_RD_N2_11.24_12_Linux64.tgz 
./FVP_RD_N2.sh --i-agree-to-the-contained-eula --no-interactive
        
    

Export the path to the FVP_RD_N2 model binary as the MODEL environment variable.

    

        
        
            export MODEL=/home/ubuntu/FVP_RD_N2/models/Linux64_GCC-9.3/FVP_RD_N2
        
    

Screen configuration for UARTs

The model will output UARTs to local ports 5000..5010. If you were running the model on a local machine, or had X11 forwarding set up, the model opens a number of xterm terminals with the UART output piped to them, one per port.

If you do not have X11 forwarding and you are executing on a remote server, you can use screen to spawn persistent terminals that listen on the ports and obtain the information that way.

Open a new terminal where you start a screen session and connect to it.

To install screen use:

    

        
        
            sudo apt-get install screen
        
    

Use a text editor to create the configuration file below, which will set up screen windows for each UART.

Create screen-uart.cfg

Create a config file so that when you start a session, there are ten windows and each periodically tries to connect to one of the local ports where a UART is running. You can change the titles of the windows so that you can identify each terminal. The resulting screen-uart.cfg file looks like this:

    

        
        
            # Split horizontally into two
split -v

# Start the SCP UART telnet
screen -t  "SCP UART" run-one-constantly telnet localhost 5000

# Split screen and start MCP UART
split
focus
screen -t "MCP term" run-one-constantly telnet localhost 5001

# Split screen and start AP-NS UART
split
focus
screen -t "AP-NS term" run-one-constantly telnet localhost 5002

# Split screen and start AP-S UART
split
focus
screen -t "AP-S term" run-one-constantly telnet localhost 5003

# Focus on the second vertical pane
focus

# Start a second set of terminals
screen -t "IO-1" run-one-constantly telnet localhost 5004

# Split 
split
focus
screen -t "IO-2" run-one-constantly telnet localhost 5005

# Split 
split
focus
screen -t "S-0" run-one-constantly telnet localhost 5006

# Split 
split
focus
screen -t "S-1" run-one-constantly telnet localhost 5007

# Split 
split
focus
screen -t "MCP-extern" run-one-constantly telnet localhost 5008

# Split 
split
focus
screen -t "term-0" run-one-constantly telnet localhost 5009

# Split 
split
focus
screen -t "term-1" run-one-constantly telnet localhost 5010

# Focus back on the AP-NS UART
focus
focus
focus
        
    

Start the screen session with this configuration file:

    

        
        
            screen -c screen-uart.cfg
        
    

The result should be similar to:

These errors are expected as there is nothing talking to the ports. You can quit from within the screen by the getting a prompt using Ctrl+A: key combination, followed by the quit command. Alternatively, Ctrl+A D will detach the screen session and send it to background.

Running the FVP

In your original terminal, launch the FVP using the supplied script:

    

        
        
            ./uefi.sh -p rdn2
        
    

Observe the platform is running successfully: Figure 2. FVP Terminals

To boot into busy-box, use:

    

        
        
            ./boot.sh -p rdn2
        
    

Figure 3. Docker Terminal

You have successfully booted the reference software stack you built in the previous step onto the FVP.