Gareth Farrington
Gareth Farrington
> It's faster than `shutdown()` for multi-mcu homing (as it is propagated between mcus without entering the Python code), and it also allows for better error reporting. ahh good to...
I got the C part in, shutdowns are gone. When an error happens the sensor just turns off its pending flag and flushes the error. That will be the last...
I rebased and now the ROM for stm32f042 is too large. How are we deciding what to cut from these space limited chips?
> rn both revisions fail on imports :( > > should I start worrying and trying to finish this pr, or am I missing something obvious?) Sorry, is this comment...
> Yes, and I mean both commits in the branch I just need to get some time to work on it. Looks like I have to exclude the hx71x sensor...
> I don't want to rush you, sorry if that sounded like it — I just wanted to ask if any help is needed. > > Thank you for working...
stm32f401 - "Too many message ids" So this PR can't be merged until klipper gets 2 bytes message IDs (or these sensors are excluded from all tests, which seems wrong)...
> What Mac do you have? Intel
Last word I know about this from Kevin: https://klipper.discourse.group/t/scalability-of-too-many-message-ids/8023/4 "Using a VLQ for the command id was just never implemented in the mcu. Implementing it is likely the best long-term...
I made a change to the HX71x code to disable interrupts for the entire read cycle. This should last about ~15us. I made the change because of 2 errors that...