----- Original Message ----- From: Marc 'BlackJack Rintsch <marc@rintsch.de> > It doesn't, you have to supply the trampoline code. > Calling such > a "banked" function puts the bank and address within the bank > into > known locations and then calls the trampoline code > `__sdcc_banked_call`. This code saves the current bank > onto the stack > changes the bank and calls the function. On return > `__sdcc_banked_ret` > is called which switches back to the callers bank. The two > functions > are of course dependent on the target hardware and have to be > supplied > by the programmer. I could see something like that being useful. How does it deal with the library runtime code? Do you have to load it into a portion of memory that doesn't get banked out? Ullrich, Is there currently any way at compile time to output what segment a piece of code is in, that could be used in a function to automate the banking? ie. so I could write a function that takes this argument and then switches to the correct bank without having to rebuild the banking table all the time... ---------------------------------------------------------------------- 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 Nov 10 01:24:53 2009
This archive was generated by hypermail 2.1.8 : 2009-11-10 01:24:56 CET