Re: [avr-libc-dev] [bugs #4101] setjmp/longjmp destroy changes in global registers

[Joerg Wunsch]

As Paul Schlie wrote:

> What were you checking for?

Dmitry's test suite, running against GCC 3.4.1.

> ..., I haven't been successful in breaking any code with the newest
> setjump patch, as GCC 4.0 seems to be only using registers r2-r17 as
> temporaries within statement expression evaluations, not across
> them, as such I haven't found a case where it matters yet?

Hmm, I have yet to try GCC pre-4 myself, by now I only relied on
Dmitry's investigation here.  For GCC 3.4.x, the test suite breaks on
Dmitry's "Preservation call-saved regs" test.

Anyway, as GCC 4.0 has not yet been released, we cannot release it in
a version of avr-libc before.  We could commit it to CVS already, but
then we'd have to require any CVS users to upgrade to pre-4.0
experimental versions of GCC, which I'd rather like to avoid.

I'll see to produce a pre-4 GCC version for my own tests though.

-- 
cheers, J"org               .-.-.   --... ...--   -.. .  DL8DTL

http://www.sax.de/~joerg/                        NIC: JW11-RIPE
Never trust an operating system you don't have sources for. ;-)