[CSDb] - User Forums - Differences between KERNALs versions

You are not logged in - nap

CSDb User Forums

Forums > C64 Coding > Differences between KERNALs versions

2020-01-02 19:08

Golara
Account closed

Registered: Jan 2018
Posts: 212

Differences between KERNALs versions

I wanted to do something, but I thought I'll ask first, maybe someone already done it... We all know that there's several KERNAL versions, some minor changes here and there. I wonder, if someone ever made a disassembly of all the variants and then compare them all to highlights the parts that are in common, leaving the parts that do change between versions (marking them with ???? for example). What for ? Well, I think it would come handy in 4K intros for finding gadgets, i.e any portion of some code ending with an rts that you can use for yourself.

2020-01-02 19:49

Rastah Bar

Registered: Oct 2012
Posts: 336

These may be useful:
http://commodore64.se/wiki/index.php/Commodore_64_KERNAL_ROM_ve..
http://unusedino.de/ec64/technical/aay/c64/krnver.htm

2020-01-02 20:01

Golara
Account closed

Registered: Jan 2018
Posts: 212

Quote: These may be useful:
http://commodore64.se/wiki/index.php/Commodore_64_KERNAL_ROM_ve..
http://unusedino.de/ec64/technical/aay/c64/krnver.htm

Thanks, these list only the differences, since these are usually more interesting. What I wanted is just a plain kernel listing (doesn't have to have comments or anything) with just the common code so one can pick and chose gadgets. Maybe I'll just make it then, there's less difference than I expected (for some reason)

2020-01-02 20:34

Krill

Registered: Apr 2002
Posts: 2839

The differences are minimal indeed.

Might be worth another thread, but as for "gadgets", i've so far found a few routines quite helpful.

1. POLYX - to evaluate a polynomial, comes in handy to compress tables down to a few 5-byte CBMFLOAT coefficients, depending on degree. There are some caveats, such as the polynomials being required to model steady functions (obviously) with a certain degree of lossiness and rather longish "depack" times, but it can save quite a bit of disk space for small demos, even more so with many tables and zeroed LSBs of the coefficient mantissas.

2. Various subroutines to add 8- or 16-bit values to 16-bit zeropage variables. Handy for very small demos.

Refresh

Subscribe to this thread: