Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.44%. Comparing base (17fbc5a) to head (66a33e4).

@@           Coverage Diff           @@
##           master   #17688   +/-   ##
=======================================
  Coverage   98.44%   98.44%           
=======================================
  Files         171      171           
  Lines       22208    22215    +7     
=======================================
+ Hits        21863    21870    +7     
  Misses        345      345           

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

Code size report:

   bare-arm:    +0 +0.000% 
minimal x86:    +0 +0.000% 
   unix x64:    +0 +0.000% standard
      stm32:    +0 +0.000% PYBV10
     mimxrt:    +0 +0.000% TEENSY40
        rp2:    +0 +0.000% RPI_PICO_W
       samd:    +0 +0.000% ADAFRUIT_ITSYBITSY_M4_EXPRESS
  qemu rv32:    +0 +0.000% VIRT_RV32

Native & viper tests fail if --via-mpy even after accounting for small int size. I did some digging and I'm guessing it's because mpy-cross uses a fixed representation for integers in the code emitter, e.g.,

        } else if (si->vtype == VTYPE_INT || si->vtype == VTYPE_UINT) {
            ASM_MOV_REG_IMM(emit->as, reg_dest, (uintptr_t)MP_OBJ_NEW_SMALL_INT(si->data.u_imm));

I chose to simply skip running any of the "native" tests, even though locally they succeed when not going via mpy. Because this was the easiest of the choices available.

Just for the benefit of one test in the testsuite, I added micropython.small_int_max. Instead, a few tests could be changed so that they just use numbers sized to work on "any repr that gets tested".

I'd prefer not to add this constant, if we can help it. Could the tests automatically detect the max int? Note there is already sys.maxsize which may be useful. Could also add a feature check in tests/feature_check/ to detect the representation.

If we do need to add a constant, it might be better as sys.implementation._small_int.

I reworked it to not require the added property.

is this a known-to-be-flaky test? My changes shouldn't have touched it. 1 tests failed: extmod/select_poll_eintr.py

is this a known-to-be-flaky test? My changes shouldn't have touched it. 1 tests failed: extmod/select_poll_eintr.py

Yes, it's known flaky. PR #17655 works around the issue.

The one remaining failure appears to be 1 tests failed: extmod/select_poll_eintr.py