From: Greg King (gngking_at_erols.com)
Date: 2000-12-01 02:16:13
-----Original Message----- From: Ullrich von Bassewitz <uz_at_musoftware.de> Date: Wednesday, November 29, 2000 06:10 PM > On Wed, Nov 29, 2000 at 05:32:12PM -0500, Greg King wrote: > > The library's "enter" subroutine can compute that length. It gets the size > > of the entire argument-list from the caller. Cc65 can give, to "enter," > > the size of the fixxed part of that list. The subroutine can subtract one > > size from the other, and put the difference on the stack. After that, the > > compiler can include code to add it to offsets whenever cc65 wants a named > > argument. > > Yes, that's a good idea. It is not even needed to change the enter subroutine, > since it does already store the size of all arguments into a stack location, so > that the correct number of parameters can be dropped on function exit. When > generating code, the compiler can account for the known size of fixxed parameters. I had become so "locked" into the way my va_par() macro works that it did not occur to me that cc65 can do one of the arithmetic operations during compilation. > > I don't think this will go into the next release, since there is not enough > time left to implement and test it, but I will have a look at it once the new > release is out. That is why I included my "stdarg.h" -- it lets us go halfway toward portable source code, until you fix cc65. ---------------------------------------------------------------------- To unsubscribe from the list send mail to majordomo_at_musoftware.de with the string "unsubscribe cc65" in the body(!) of the mail.
This archive was generated by hypermail 2.1.3 : 2001-12-14 22:05:37 CET