As Michal Ludvig wrote:

> One thing is reseting BusPirate during unclean exit (that may not be
> cleanly doable) and another thing is detecting that BP is in a
> non-default state and resetting it upon next avrdude run. At least
> that should be implemented [...]

OK, that sounds like a plan, yes.

> What's your deadline for the next release by the way?

"Really soon now".  Within days, I'd say.  There are some open Xmega
issues which I'd like to reproduce/fix, but then, with the many
bugfixes that came in, I think a new release would be due.

> You can register a cleanup function with atexit() to ensure it's called
> after any exit().

That's already done right now, in order to call the teardown handler
of each programmer.

> However that doesn't kick in for exits triggered by
> signals.

Exactly, and performing a synchronous exit() from within a signal
handler wouldn't do very well, as it would just extend the requirement
to not use things like malloc/free into the atexit handlers (where the
currently existing teardown handlers are all doing some memory
management).

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

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


_______________________________________________
avrdude-dev mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/avrdude-dev

Reply via email to