Re: [avr-gcc-list] Weird optimization issue with avr-gcc 4.5.3, re "naked."

[Bill Westfield]

> How does it compile if you change the C code to this?:
> [manually factoring out the common code]

Yes, that will fix this particularly obvious instance, but it looks
like gcc is failing to factor out common code like this in less
obvious places as well...

(Hmm.  It behaves differently (correctly!) with attribute OS_main
instead of naked.  That's also "interesting.")