Hi Uz, > This has been broken since ages (version 2.4 or similar) and no one has > complained so I almost forgot about it:-) That's mor or less what I guessed ;-) > The compiler can be fixed for the code shown above with not too much effort. > But the code you posted is a different thing, because it would be required to > move allocation of variables to the top of the function, and this doesn't > really fit into the compiler as it is now. I see. Again sort of what I guessed... Just to make sure: I am _not_ working on a project relying on that issue fixed ! > Maybe I can at least make this an > error .. Sounds reasonable to me - and/or maybe an entry in http://www.cc65.org/doc/cc65-4.html > Thanks for the report! You're welcome :-) In case you're interested: Today I noticed that the 6502-specific implementation of our 'Local Continuations' used in Contiki don't work with variables in local blocks: http://contiki.git.sourceforge.net/git/gitweb.cgi?p=contiki/contiki;a=blob_plain;f=cpu/6502/sys/lc-asm.S;hb=HEAD After understanding why it doesn't work (and then removing all variables in local blocks inside functions using Local Continuations) I was just curious to see if you built some "workaround" into the "official" goto statement... Regards, Oliver ---------------------------------------------------------------------- To unsubscribe from the list send mail to majordomo@musoftware.de with the string "unsubscribe cc65" in the body(!) of the mail.Received on Tue Apr 12 22:15:41 2011
This archive was generated by hypermail 2.1.8 : 2011-04-12 22:15:44 CEST