Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.57%. Comparing base (40048f0) to head (b0ba151).
Report is 2 commits behind head on master.

@@           Coverage Diff           @@
##           master   #15902   +/-   ##
=======================================
  Coverage   98.57%   98.57%           
=======================================
  Files         164      164           
  Lines       21336    21336           
=======================================
  Hits        21031    21031           
  Misses        305      305           

☔ 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:   +56 +0.015% TEENSY40
        rp2:   +40 +0.004% RPI_PICO_W
       samd:   +68 +0.025% ADAFRUIT_ITSYBITSY_M4_EXPRESS[incl +4(bss)]
  qemu rv32:    +0 +0.000% VIRT_RV32

Tested with

• MIMXRT 1062 (Teensy 4.0), MIMXRT1011, MIMXRT1015, MIMXRT1021, MIMXRT1052, MIMXRT1172
• SAMD21 with and without external flash ,
• SAMD51,
• RP2040, RP2350 ARM, RP2350 RISCV
• NRF52840

all build and work. NRF seems better than before. Code size increase for SAMD21 is 48bytes. Lacking a ARDUINO_PORTENTA_C33 board I could not test it with Renesas-RA. But at least the firmware builds.

Thanks @robert-hh I've got a renesas FS USB board I'll be able to test with.

@iabdalkader no pressure but if you get a chance to test on the portenta renesas or would be good to confirm it hasn't been broken by this change.

@robert-hh thanks for testing! That's very helpful.

I also tested with the following:

• rp2 / Pico
• samd / Itsy Bitsy M0 Express
• mimxrt / Teensy 4.0

I didn't see any regressions. USB CDC throughput is unchanged.

USB CDC throughput is unchanged.

Hi @dpgeorge do you have a script/pattern you could share that you use to test CDC throughout? I'd like to check that before/after on renesas and esp boards.

do you have a script/pattern you could share that you use to test CDC throughout?

Yes! See #15909.

Thanks for that @dpgeorge !
Renesas FS USB is slightly slower (right is with this PR applied), but not dramatically:

You probably need to do a few runs before you can compare code changes. Eg the load of your PC can change the throughput.

Yeah I did a few repeats and got similar results, though to be fair I didn't go back to original to re-check that.
I did also try applying [renesas-ra/usb: Use interrupt rather than polling for usb task.(]#15550 ) but didn't see any change to the speed - I guess the polling will be working just fine with this test code where it calls sleep in loops etc.

edit: Ok I did rerun both MR code then went back to master and got similar results on the large block sizes: 0.53 MBits/sec vs 0.56 MBits/sec. Not a deal breaker I don't think, but curious.

@andrewleech so you were able to test the renesas-ra port? If it worked then I think we are good to go with this PR.

@dpgeorge sorry I wasn't entirely clear - yes renesas-ra port is tested for FS connections and it does work. I don't have hardware to test HS myself though. FS is slightly slower but not enough to prevent moving forward I don't think.

@andrewleech OK, thanks. So the remaining test would be for renesas-ra port in HS mode.

@iabdalkader are you able to test this PR on an Arduino Portenta C33?

@iabdalkader are you able to test this PR on an Arduino Portenta C33?

@andrewleech @dpgeorge Yes I can test it today. I tested this too #15550 and it works fine, if you merge it first it will be easier for me to test both.

I tested this too #15550 and it works fine, if you merge it first it will be easier for me to test both.

OK, that's now merged.

I pulled this PR locally, and rebased on master and tested C33, and it seems to work fine.

Thanks @iabdalkader for testing.