(41CL) Advantage Pac (41AD) mapped to page 2?

[Monte Dalrymple]

(02-26-2019 07:57 PM)cdmackay Wrote:  [I hope this is an OK place to ask 41CL questions? If there's a more appropriate forum, or list, please tell me off and I'll take it there]

I've just (yesterday) got my 41CL, and am having some initial plays with it. What a great piece of kit

I was looking around at what was mapped on the MMU, and was surprised to find the Advantage Pac (ROM image 41AD) apparently mapped to page 2. That's really page 2, not port 2 (pages 10 & 11).

To verify that, I thought I would reset the MMU and start from scratch.

I did a backspace-ON, verified the MMU was disabled, then did MMUCLR.

I then loaded YFNX (41CL Extreme Functions) to page 6 with PLUGP (having read that this ROM is safe there). Then enabled the MMU.

I then used the YFNX function PLUG to search the mapped modules, with "?IMG ?", which produced:

Code:

41AD 2
YFNX 6

Just to check that, I read the value at RAM physical address 0x80402C, which is the MMU entry for page 2, set 0:

Code:

ALPHA 80402C-0000 ALPHA
YPEEK

which returned 0x800C. The 0x8000 is the MMU Enable bit, and 0x00C is indeed the page for the 41AD module.


Is this expected? I thought pages 0–2 were for the OS itself, and weren't mapped by the MMU? So why does there appear to be a ROM image mapped at page 2?

Most likely I have deeply confused myself, in which case apologies, but would someone please straighten me out? thanks!

Robert is correct. MMUCLR and MMUCLP do not affect the MMU locations for pages 0-3. But that's okay, because these pages are never enabled unless you enable the special MMU mapping for the OS pages. Being located in RAM, the MMU registers can power up holding anything, and that's why you must do MMUCLR before enabling the MMU. In a similar fashion, you must initialize the MMU registers for the OS pages before you issue the MAPEN command.

I deliberately did not initialize the MMU pages for the OS sector with MMUCLR, thinking that if someone was actually using the MAPEN feature to run their own OS code they wouldn't appreciate having to reinitialize the MMU registers for the custom OS all the time. But I don't think anyone besides me has used the CL with their own OS. I have only done it when I am playing around with experimental patches.

If the random stuff in those MMU locations bothers you (like when you do a PLUG ?IMG ? command to see the MMU contents) just do PLUG EMPT x, where x is 0-3 to clear those MMU registers.

Enjoy your CL!

Monte