9. Communicate with your Pico over RPC#

Now, let’s revisit pw_console and pw_rpc. This time, we’ll send commands to and view logs from the real Pico device.

Interact with the Pico#

  1. Fire up a pw_console instance that’s connected to the Pico:

    In Bazel Build Targets right-click :rp2040_console (native_binary) (under //apps/blinky) and then select Run target.

    $ bazelisk run //apps/blinky:rp2040_console

  2. Toggle the Pico’s LED by typing the following into Python Repl and then pressing Enter:

    >>> device.rpcs.blinky.Blinky.ToggleLed()

    You should see your Raspberry Pi Pico’s LED turn either on or off. If you run the command again you should see the LED switch to its opposite state.

    (The next few commands should be executed the same way.)

  3. Blink the LED 10 times:

    >>> device.rpcs.blinky.Blinky.Blink(interval_ms=1000, blink_count=10)

  4. Write some custom automation in the Python REPL to achieve the same blinking effect:

    >>> def my_blinky(count, delay):
...     from time import sleep
...     toggle = device.rpcs.blinky.Blinky.ToggleLed
...     for _ in range(count):
...         toggle()
...         sleep(delay)
...
>>> my_blinky(20, 1)

    Note

    The REPL doesn’t currently support top-level execution of multiple statements. You can workaround this by wrapping your multi-statement logic in a function and then invoking the function, as seen in my_blinky().

  5. View the RP2040’s onboard temperature:

    >>> device.rpcs.board.Board.OnboardTemp()

    In Python Results you should see output like this:

    >>> device.rpcs.board.Board.OnboardTemp()
(Status.OK, board.rpc.OnboardTempResponse(temp=23.861492156982422))

    Put your finger on the RP2040 chip in the middle of your Raspberry Pi Pico for a few seconds and then run the temperature command again and you should see the temperature increase.

    Exercise

    Can you figure out the code to read the temperature 10 times with a 1-second interval between readings, and then output the average temperature? See Temperature averaging solution for one option.

  6. Leave the console open and proceed to the next section.

Search and filter logs#

You can search and filter your device’s logs. Try it now:

  1. Click any row in the Device Logs table to focus that part of the UI.

  2. Press / to search the logs.

  3. Type Stopped blinking and press Enter. A log that matches that string should be highlighted.

    Troubleshooting

    No logs are shown. There probably has just not been any logs that match the filter you entered. Try filtering by other values, such as 00:00 to only show logs that occurred during the first 60 seconds of logs.

  4. Press n to go to next match and N to go to previous match. (If there are 0 matches or only 1 match then this naturally won’t work.)

  5. Press Ctrl+Alt+F to filter out logs that don’t match your query.

  6. Press Ctrl+Alt+R or click Clear Filters to clear your filter and return to the original table view.

Handle crashes (full setup only)#

Warning

This workflow requires the full hardware setup.

pw_cpu_exception provides a consistent interface for entering a CPU exception handler and makes it easier to consistently collect CPU state that may otherwise get clobbered by application exception handlers. Try triggering a crash now and downloading a snapshot:

  1. In the Python Repl of pw_console send the following command:

    >>> device.rpcs.pw.system.proto.DeviceService.Crash()

    You should see error logs pop up in Device Logs.

    The traceback in Python Output is expected because you just manually triggered a crash:

    https://storage.googleapis.com/pigweed-media/sense/crash_python_results.png

  2. Download the crash snapshot to /tmp (or whatever directory you prefer) by running the following command in the Python Repl:

    >>> device.get_crash_snapshots('/tmp')

    In Host Logs you should see confirmation that a snapshot was downloaded:

    https://storage.googleapis.com/pigweed-media/sense/crash_snapshot.png

  3. You can leave pw_console running. We’ll keep exploring it on the next page.

Summary#

On this page we revisited our old friends pw_console and pw_rpc, except this time we used them to communicate with a real embedded device rather than a simulated device running on our development host. In other words, when it’s time to switch from simulated devices to real ones, you don’t need to learn new tools.

We also met a new friend, pw_cpu_exception, that can help us debug crashes.

Next, head over to 10. Automate common workflows to learn how to package up common development tasks into small scripts so that your whole team can benefit from them.

Appendix#

Temperature averaging solution#

Here’s one possible solution to the temperature averaging exercise at the bottom of Interact with the Pico.

def average(count, delay):
    from time import sleep
    total = 0
    sample = device.rpcs.board.Board.OnboardTemp
    for _ in range(count):
        status, data = sample()
        total += data.temp
        sleep(delay)
    return total / count

average(10, 1)
On this page