Crash in touch.py

[p8mty]

Occasionally (2 or 3 times a day) my program crashes with the following:

>>> %Run -c $EDITOR_CONTENT

MPY: soft reboot
malloc self
set fb pointers
m_new_class(ST7701...
launch core1
launched core1
core1 returned

[[[[ Program output ]]]]

Traceback (most recent call last):
  File "<stdin>", line 265, in <module>
  File "touch.py", line 66, in poll
  File "touch.py", line 77, in _handle_touch
OSError: [Errno 5] EIO

The call in line 265 in my program is merely a call to touch.poll().
The line 77 in _handle_touch is dealing with i2c - I have no additional i2c devices attached to the Presto

It's not reproducible in that I don't know what to do to cause it, but if I poke at the screen for long enough, it will happen!

[MichaelBell]

I saw a crash a couple of times when running for a long time, which I suspect was this (but I wasn't running through Thonny so didn't see the stack). We should probably just ignore the failure and wait for the next poll - I2C can be a bit unreliable at times.

[Gadgetoid]

Yup swallow the error and hope the next poll works 😬 with maybe a more permanent failure for multiple sequential errors?

[kmohrf]

It seems like this problem doesn’t (always?) trigger a python exception and therefore can’t be handled in micropython. Can you confirm that? I’ve monkey-patched the presto.touch.poll function that is called in presto.update and presto.partial_update to catch the base Exception type and log all errors and I don’t see any log calls in my output. Instead, a few seconds later my watchdog kicks in and resets the presto.

In case this is true, I’d appreciate a timely fix as this is the only thing that keeps my presto from being a reliable little companion.