Ethan Reesor

8 bytes of the change are due to `Default_Handler`: ```diff : bf20 wfe e7fd b.n d4d4 bmi.n +d4d4 bmi.n +d4d4 bmi.n +d4d4 bmi.n +d4d4 bmi.n ```

@deadprogram Rebased

On interrupt, the link register is filled with an `EXC_RETURN` (exception return) value. Bit 2 indicates whether the interrupted code was using the PSP (bit 2 == 1) or the...

The following works as expected, but fails to compile for the Feather M0: ```go //go:export HardFault_Handler func hardFaultEntry() { pspStacked := arm.AsmFull("ands.w {}, lr, #(1

`HasBits` returns whether any bit is set at the given positions, which is why it returns a boolean, whereas `GetBits` extracts the value of the specified bits, regardless of whether...

At one point, it looked like this version was working. Or at least I was seeing far more log messages from the USB ISR. But now it doesn't get past...

What about a normally false constant flag to print out full messages? In the usual case, the user will get a minimal error but binary size won’t increase much, but...

I expected less size differences with `verboseFaultMessage = false`. I'll work on that tomorrow. ``` main-old0.hex 25992 bytes # baseline for feather-m0 main-m0.hex 26080 bytes # feather-m0 with new fault...

As of the latest commit, there's no change for M0, and for M4: * Default: +184 bytes * With `showFaultAddresses`: +360 bytes * With `verboseFaultMessage`: +2472 bytes * With both...

Well, clearly I misunderstood the CPU field. That being said, I successfully implemented compiler-defined constants, though it's not pretty. I'd much rather manipulate the SSA or IR than inject a...