@iabdalkader FYI

Code size report:

   bare-arm:    +0 +0.000% 
minimal x86:    +0 +0.000% 
   unix x64:   +64 +0.008% standard
      stm32:   +72 +0.018% PYBV10
     mimxrt:   +72 +0.020% TEENSY40
        rp2:   +72 +0.022% RPI_PICO
       samd:   +68 +0.026% ADAFRUIT_ITSYBITSY_M4_EXPRESS

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (51fbec2) 98.36% compared to head (49e78d4) 98.36%.

❗ Current head 49e78d4 differs from pull request most recent head 2ecbad4. Consider uploading reports for the commit 2ecbad4 to get more accurate results

@@           Coverage Diff           @@
##           master   #13474   +/-   ##
=======================================
  Coverage   98.36%   98.36%           
=======================================
  Files         159      159           
  Lines       21088    21088           
=======================================
  Hits        20743    20743           
  Misses        345      345           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

I found and fixed an edge case where all sub-tasks finish early.

@dpgeorge Thank you very much for this much needed fix 🙏🏼 I had a chance to test this today, and even though it fixes the main issue for me (gather() raising "can't gather"), I still see the following output... Is this expected, should I just ignore it ?

Task exception wasn't retrieved
future: <Task> coro= <generator object 'conn_task' at 240437a0>
Traceback (most recent call last):
  File "asyncio/core.py", line 1, in run_until_complete
  File "ucloud.py", line 341, in conn_task
DoneException: 

Also it seems the task exception is now saved, and I can retrieve it after gather() raises with t.data, but could we also retrieve it with Task.exception() for more compatibility with CPython's Task ? https://docs.python.org/3/library/asyncio-task.html#asyncio.Task.exception

# Save exception raised by the coro for later use.
t.data = exc

I still see the following output... Is this expected, should I just ignore it ?

Yes that's expected. Because of the difference in GC vs reference counting, MicroPython and CPython necessarily differ in when exceptions like this are printed. You can see in the test in this PR that those messages are sent to a dummy custom_handler() which just ignores them. That way the test runs the same under CPy and uPy.

Also it seems the task exception is now saved, and I can retrieve it after gather() raises with t.data

You should also be able to retrieve the exception from the return value of the gather() call. Is that enough? If not, why not?

Also it seems the task exception is now saved, and I can retrieve it after gather() raises with t.data

If the return value of gather() is not enough we could consider adding Task.exception(). See #13000.

You should also be able to retrieve the exception from the return value of the gather() call. Is that enough? If not, why not?

This is how I catch the task exception:

            task_except = None
            try:
                await asyncio.gather(*self.tasks.values(), return_exceptions=False)
            except Exception as e:
                task_except = e
# Then later
            for name in list(self.tasks):
                task = self.tasks[name]
                try:
                    if task.done():
                         if isinstance(task_except, DoneException)
                break   # Break after the first task is removed.

This counts on tasks being processed by gather() in order, i.e tasks that raise exceptions are returned in order (which is always true I think?). The only problem with this, is that I have to process/cleanup one task at a time, then call gather() again, but it's not a big issue.