Re: c264 series (c16/c116/+4) complete ROM disassembly

From: Ullrich von Bassewitz <uz_at_musoftware.de>
Date: Tue, 3 Mar 2015 21:02:27 +0100
Message-ID: <20150303200226.GA14861@trixie.musoftware.de>
On Tue, Mar 03, 2015 at 08:22:48AM +0100, Gerrit Heitsch wrote:
> Most systems did have the -04 and -05 though. Commodore even
> replaced the kernal in some systems with a -05 afterwards (the
> kernal ROM had a '86 datecode while the rest of the chips were all
> '84). That makes me assume that some bugs were pretty bad.

The code in my last mail shows part of the problem. According to the book, the
descriptor for DS$ isn't initialized correctly. This leads to a crash when a
garbage collection for strings is executed. The following sequence of commands
can be used to check for the bug in the kernal (page 35):

  1. Enter "NEW"
  2. Enter "PRINT FRE(0)" and remember the output
  3. Enter "PRINT DS$", so DS$ is created (floppy must be powered on)
  4. Enter "PRINT FRE(0)" again

If the values from 2. and 4. differ by 42 bytes, the machine has the new ROM.
Any other value means that the machine has the old ROM.

> Here's the 'cmp -l' output between -03 and -05:

Yes, looks like the "old" version shown in the book is -03.

Regards


        Uz


-- 
Ullrich von Bassewitz                                  uz@musoftware.de
Encrypted email preferred                          PGP Key-Id: 29D93B10

       Message was sent through the cbm-hackers mailing list
Received on 2015-03-03 21:00:05

Archive generated by hypermail 2.2.0.