--- Ullrich von Bassewitz <uz@musoftware.de> wrote: > disadvantage of this > approach is that you must recompile all modules when > you make changes to the > main code, because the location of the runtime > functions may have changed. Following approach may solve this problem: One can create a run-time jump table in the main program with JMPs to the actual addresses and get the linker place this jump table to a specific address. Then modules can use these addresses instead of actual locations. Ofcourse it will have a performance impact but at least modules can run with different versions of the main program. Something like this should work: In the main program: << .segment "RUN_TIME_STUB" ; adjust the linker to place this segment to a known ; location. say $2000-$2300 jmp _printf jmp _malloc ... >> Note that this will ensure that all the run-time routines will get linked. So one may want to left out some of them that will not be used by the modules. Then one can write a stub run-time for the modules: << _printf = $2000 _malloc = $2003 ... >> But as I said, it will have a big performance impact given that cc65 calls some library routines quite often. Fatih. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ---------------------------------------------------------------------- To unsubscribe from the list send mail to majordomo@musoftware.de with the string "unsubscribe cc65" in the body(!) of the mail.Received on Sun Jul 30 02:36:55 2006
This archive was generated by hypermail 2.1.8 : 2006-07-30 02:36:58 CEST