Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.sys.apple2.programmer > #1477 > unrolled thread

Method to address both Legend and Saturn RAM cards

Started byD Finnigan <dog_cow@macgui.com>
First post2015-01-15 19:09 +0000
Last post2015-01-20 17:06 -0600
Articles 15 — 5 participants

Back to article view | Back to comp.sys.apple2.programmer

Contents

  Method to address both Legend and Saturn RAM cards D Finnigan <dog_cow@macgui.com> - 2015-01-15 19:09 +0000
    Re: Method to address both Legend and Saturn RAM cards "Anton Treuenfels" <teamtempest@yahoo.com> - 2015-01-15 22:21 -0600
      Re: Method to address both Legend and Saturn RAM cards gids.rs@sasktel.net - 2015-01-16 09:41 -0800
        Re: Method to address both Legend and Saturn RAM cards D Finnigan <dog_cow@macgui.com> - 2015-01-16 19:08 +0000
        Re: Method to address both Legend and Saturn RAM cards qkumba <peter.ferrie@gmail.com> - 2015-01-17 21:32 -0800
          Re: Method to address both Legend and Saturn RAM cards gids.rs@sasktel.net - 2015-01-17 22:33 -0800
            Re: Method to address both Legend and Saturn RAM cards gids.rs@sasktel.net - 2015-01-17 22:54 -0800
              Re: Method to address both Legend and Saturn RAM cards Leandro Polimeno <leandro.polimeno@gmail.com> - 2015-02-10 04:53 -0800
                Re: Method to address both Legend and Saturn RAM cards gids.rs@sasktel.net - 2015-02-10 21:19 -0800
                  Re: Method to address both Legend and Saturn RAM cards Leandro Polimeno <leandro.polimeno@gmail.com> - 2015-02-11 05:22 -0800
            Re: Method to address both Legend and Saturn RAM cards D Finnigan <dog_cow@macgui.com> - 2015-01-18 10:09 +0000
              Re: Method to address both Legend and Saturn RAM cards "Anton Treuenfels" <teamtempest@yahoo.com> - 2015-01-19 08:33 -0600
                Re: Method to address both Legend and Saturn RAM cards D Finnigan <dog_cow@macgui.com> - 2015-01-19 18:29 +0000
                Re: Method to address both Legend and Saturn RAM cards qkumba <peter.ferrie@gmail.com> - 2015-01-20 09:33 -0800
                  Re: Method to address both Legend and Saturn RAM cards "Anton Treuenfels" <teamtempest@yahoo.com> - 2015-01-20 17:06 -0600

#1477 — Method to address both Legend and Saturn RAM cards

FromD Finnigan <dog_cow@macgui.com>
Date2015-01-15 19:09 +0000
SubjectMethod to address both Legend and Saturn RAM cards
Message-ID<dog_cow-1421349003@macgui.com>
After looking at the technical documentation for the Legend 64KC and the
Saturn 32/64/128K RAM cards, it seems that one could write a common routine
using 6502 indexed addressing to work on both models of RAM card.

The Legend 64KC provides a bank select address at location $C084. This is a
write only address used to select bank 0 through 3.

The Saturn RAM cards use a series of addresses to select a bank, starting
with $C084 for bank 0, $C085 for bank 1, $C086 for bank 2, and so on.

So it seems to me that we could use indexed addressing with the X register
and $C084 as the base address. First we should have an earlier subroutine
detect which card is installed, or just ask the user, and store the setting
in ZP location. The subroutine to select a bank should accept the bank
number in the X-reg.

Here's my first naive attempt at writing the routine. There's definitely
room for improvement, but this is the general idea:



BANKBASE	EQU	$C084	; BASE ADDRESS FOR BANK SELECT ADDR
RAMTYPE		EQU	$06	; TYPE OF RAM CARD INSTALLED
*
*
* SELECT A 16K RAM BANK. DESIRED BANK NUMBER
* SHOULD BE PASSED IN THE A-REG.
*
BANKSEL
		LDA	RAMTYPE	; WHAT TYPE OF RAM CARD?
		BEQ	:LEGEND	; 0 = LEGEND, 1 = SATURN
		TAX
		STA	BANKBASE,X	; SELECT BANK
		RTS
:LEGEND		LDX	#0
		STA	BANKBASE,X	; SELECT BANK
		RTS



Documentation:

<http://lostclassics.apple2.info/wp-content/files/Legend/64KC.pdf>
<http://mirrors.apple2.org.za/Apple%20II%20Documentation%20Project/Interface%20Cards/Memory/Saturn%20Systems%20RAM%20Board/Manuals/Saturn%20Systems%2064K%20and%20128K%20RAM%20Board%20-%20Operations%20Manual.pdf>

-- 
]DF$
Apple II Book: http://macgui.com/newa2guide/
Usenet: http://macgui.com/usenet/  <-- get posts by email!
Apple II Web & Blog hosting: http://a2hq.com/

[toc] | [next] | [standalone]

#1478

From"Anton Treuenfels" <teamtempest@yahoo.com>
Date2015-01-15 22:21 -0600
Message-ID<crCdnUWW-uNNDCXJnZ2dnUU7-dednZ2d@earthlink.com>
In reply to#1477
"D Finnigan" <dog_cow@macgui.com> wrote in message 
news:dog_cow-1421349003@macgui.com...

> BANKBASE EQU $C084 ; BASE ADDRESS FOR BANK SELECT ADDR
> RAMTYPE EQU $06 ; TYPE OF RAM CARD INSTALLED
> *
> *
> * SELECT A 16K RAM BANK. DESIRED BANK NUMBER
> * SHOULD BE PASSED IN THE A-REG.
> *
> BANKSEL
> LDA RAMTYPE ; WHAT TYPE OF RAM CARD?
> BEQ :LEGEND ; 0 = LEGEND, 1 = SATURN
> TAX
> STA BANKBASE,X ; SELECT BANK
> RTS
> :LEGEND LDX #0
> STA BANKBASE,X ; SELECT BANK
> RTS

I don't know squat about either card, but:

- if the bank number is passed in the A-register, doesn't the instruction

LDA RAMTYPE

wipe it out first thing?

- if the value of the X-register is zero, there is no practical difference 
between

STA BANKBASE,X

and

STA BANKBASE

unless there's some weird hardware quirk that "notices" the address mode and 
does something special about it.

- so how about this:

BANKSEL
  LDX RAMTYPE
  BEQ LEGEND
  TAX
LEGEND
  STA BANKBASE,X
  RTS

- Anton Treuenfels

[toc] | [prev] | [next] | [standalone]

#1479

Fromgids.rs@sasktel.net
Date2015-01-16 09:41 -0800
Message-ID<5f4fce0f-eef3-4e51-aaef-f3929de36fbd@googlegroups.com>
In reply to#1478
On Thursday, January 15, 2015 at 10:21:37 PM UTC-6, Anton Treuenfels wrote:
> "D Finnigan" wrote in message 
> 
> > BANKBASE EQU $C084 ; BASE ADDRESS FOR BANK SELECT ADDR
> > RAMTYPE EQU $06 ; TYPE OF RAM CARD INSTALLED
> > *
> > *
> > * SELECT A 16K RAM BANK. DESIRED BANK NUMBER
> > * SHOULD BE PASSED IN THE A-REG.
> > *
> > BANKSEL
> > LDA RAMTYPE ; WHAT TYPE OF RAM CARD?
> > BEQ :LEGEND ; 0 = LEGEND, 1 = SATURN
> > TAX
> > STA BANKBASE,X ; SELECT BANK
> > RTS
> > :LEGEND LDX #0
> > STA BANKBASE,X ; SELECT BANK
> > RTS
> 
> I don't know squat about either card, but:
> 
> - if the bank number is passed in the A-register, doesn't the instruction
> 
> LDA RAMTYPE
> 
> wipe it out first thing?
> 
> - if the value of the X-register is zero, there is no practical difference 
> between
> 
> STA BANKBASE,X
> 
> and
> 
> STA BANKBASE
> 
> unless there's some weird hardware quirk that "notices" the address mode and 
> does something special about it.
> 
> - so how about this:
> 
> BANKSEL
>   LDX RAMTYPE
>   BEQ LEGEND
>   TAX
> LEGEND
>   STA BANKBASE,X
>   RTS
> 
> - Anton Treuenfels


No.  That doesn't quite work either.

You need to be able to store a 0-$F in the Saturns register.

So Maybe something like this

LDY #0-3 ; for Legend
LDY #0-$F; for Saturn

BANKSEL  LDX RAMTYPE        ; WHAT TYPE OF RAM CARD?
         BEQ :LEGEND        ; 0 = LEGEND, 1 = SATURN
         TYA
         STA BANKBASE,X     ; SELECT BANK
         RTS

:LEGEND  CPY #4
         BCS RETURN
         TYA
         TAX
         STA BANKBASE,X     ; SELECT BANK
RETURN   RTS

[toc] | [prev] | [next] | [standalone]

#1480

FromD Finnigan <dog_cow@macgui.com>
Date2015-01-16 19:08 +0000
Message-ID<dog_cow-1421435342@macgui.com>
In reply to#1479
I grimly admit that the 6502 code in my initial article wasn't a very good
expression of my idea, but I hope that people can still understand my
intent.

Thanks for the fixes. When I've got some more time to do this properly, I'll
come up with something better.

-- 
]DF$
Apple II Book: http://macgui.com/newa2guide/
Usenet: http://macgui.com/usenet/  <-- get posts by email!
Apple II Web & Blog hosting: http://a2hq.com/

[toc] | [prev] | [next] | [standalone]

#1481

Fromqkumba <peter.ferrie@gmail.com>
Date2015-01-17 21:32 -0800
Message-ID<2e88c3d7-749a-43c2-8ab6-48af7884af5f@googlegroups.com>
In reply to#1479
That's not quite right, either.

LDA #0-3 ; for Legend 
LDA #0-$F; for Saturn 

BANKSEL  LDX RAMTYPE        ; WHAT TYPE OF RAM CARD? 
         BNE :SATURN        ; 0 = LEGEND, 1 = SATURN 
         CMP #4             ; LEGEND
         BCS RETURN 

         db  #$24           ; BIT masks TAX
:SATURN
         TAX
         STA BANKBASE,X     ; SELECT BANK 
RETURN   RTS

[toc] | [prev] | [next] | [standalone]

#1482

Fromgids.rs@sasktel.net
Date2015-01-17 22:33 -0800
Message-ID<276de793-dd18-47b7-87b7-f9c5f5316798@googlegroups.com>
In reply to#1481
On Saturday, 17 January 2015 23:32:16 UTC-6, qkumba  wrote:
> That's not quite right, either.
> 
> LDA #0-3 ; for Legend 
> LDA #0-$F; for Saturn 
> 
> BANKSEL  LDX RAMTYPE        ; WHAT TYPE OF RAM CARD? 
>          BNE :SATURN        ; 0 = LEGEND, 1 = SATURN 
>          CMP #4             ; LEGEND
>          BCS RETURN 
> 
>          db  #$24           ; BIT masks TAX
> :SATURN
>          TAX
>          STA BANKBASE,X     ; SELECT BANK 
> RETURN   RTS



I think you had better re-read David's first post.

This next line is not possible.

> The Saturn RAM cards use a series of addresses to select a bank, starting
with $C084 for bank 0, $C085 for bank 1, $C086 for bank 2, and so on.


A Saturn card can have up to 128 kb which is 16 banks of 16 kb.  Your code would be accessing $C084 - $C093.  You have a bug in your code.

Ergo, how the cards are stated to be accessed, are more than likely turned around.

See what you started David?  You listed buggy code.

[toc] | [prev] | [next] | [standalone]

#1483

Fromgids.rs@sasktel.net
Date2015-01-17 22:54 -0800
Message-ID<ebb13e2c-0c6d-4bc8-9353-7476a4f1e1d1@googlegroups.com>
In reply to#1482
Here is some Ram drive code I have for the Saturn.

There looks to be 8 subbanks within 1 or 2 main banks depending if its 64 or 128 kb.

The subbanks are accessed from $C084.C087 and $C00C.C00F.

; ProDOS disk driver for 
; 64k/128k Saturn ramcard 
; 
; Matthew Lee Stier 
; Sun Microsystems 


	ORG $2000 
	LDY #7 
NXTBLK16	LDX LC.INDEX,Y	; TEST FOR RAMCARD 
	LDA $C080,X 	; SIZE 
	BIT $C083 	; ENABLE BANK 2 
	BIT $C083 
	TYA 
	STA $D000 
	DEY 
	BPL NXTBLK16 
	BIT $C08F 
	LDY $D000 	; GET RAMCARD SIZE 
	BIT $C084 	; RESTORE BLOCK 0 
	BIT $C089 	; BANK 1 
	BIT $C089 
	CPY #8	; ILLEGAL BANK SIZE? 
	BCS TESTRTN 	; YES... 
	TYA 
	ASL 
	TAY 
	LDX BLKSIZE+1,Y 	; GET RAMCARD SIZE 
	LDA BLKSIZE,Y 	; IN PRODOS BLOCKS 
	BNE INSTALL 	; IF NOT 0 THEN INSTALL 
	SEC 		; ELSE FLAG AN ERROR 
TESTRTN	RTS 		; AND QUIT 

LC.INDEX	DB $4,$5,$6,$7,$C,$D,$E,$F 
BLKSIZE	DW 0,0,0,96,0,0,0,224

[toc] | [prev] | [next] | [standalone]

#1504

FromLeandro Polimeno <leandro.polimeno@gmail.com>
Date2015-02-10 04:53 -0800
Message-ID<94496630-650b-4c0b-8d64-bea0c23bcfde@googlegroups.com>
In reply to#1483
Humm,

But that code, RAMDrive for 64k/128k Saturn by Matthew Lee Stier, works on ProDOS 1.9 ?! I have it and can´t play with.

Can you provide an image for test ?!

Regards,

Polimeno

Em domingo, 18 de janeiro de 2015 04:54:46 UTC-2, gid...@sasktel.net  escreveu:
> Here is some Ram drive code I have for the Saturn.
> 
> There looks to be 8 subbanks within 1 or 2 main banks depending if its 64 or 128 kb.
> 
> The subbanks are accessed from $C084.C087 and $C00C.C00F.
> 
> ; ProDOS disk driver for 
> ; 64k/128k Saturn ramcard 
> ; 
> ; Matthew Lee Stier 
> ; Sun Microsystems 
> 
> 
> 	ORG $2000 
> 	LDY #7 
> NXTBLK16	LDX LC.INDEX,Y	; TEST FOR RAMCARD 
> 	LDA $C080,X 	; SIZE 
> 	BIT $C083 	; ENABLE BANK 2 
> 	BIT $C083 
> 	TYA 
> 	STA $D000 
> 	DEY 
> 	BPL NXTBLK16 
> 	BIT $C08F 
> 	LDY $D000 	; GET RAMCARD SIZE 
> 	BIT $C084 	; RESTORE BLOCK 0 
> 	BIT $C089 	; BANK 1 
> 	BIT $C089 
> 	CPY #8	; ILLEGAL BANK SIZE? 
> 	BCS TESTRTN 	; YES... 
> 	TYA 
> 	ASL 
> 	TAY 
> 	LDX BLKSIZE+1,Y 	; GET RAMCARD SIZE 
> 	LDA BLKSIZE,Y 	; IN PRODOS BLOCKS 
> 	BNE INSTALL 	; IF NOT 0 THEN INSTALL 
> 	SEC 		; ELSE FLAG AN ERROR 
> TESTRTN	RTS 		; AND QUIT 
> 
> LC.INDEX	DB $4,$5,$6,$7,$C,$D,$E,$F 
> BLKSIZE	DW 0,0,0,96,0,0,0,224

[toc] | [prev] | [next] | [standalone]

#1505

Fromgids.rs@sasktel.net
Date2015-02-10 21:19 -0800
Message-ID<926aa9f1-7a48-4737-a735-b4844b459cb7@googlegroups.com>
In reply to#1504
On Tuesday, February 10, 2015 at 6:53:38 AM UTC-6, Leandro Polimeno wrote:
> Humm,
> 
> But that code, RAMDrive for 64k/128k Saturn by Matthew Lee Stier, works on ProDOS 1.9 ?! I have it and can´t play with.
> 
> Can you provide an image for test ?!
> 
> Regards,
> 
> Polimeno
> 
> Em domingo, 18 de janeiro de 2015 04:54:46 UTC-2, gid...@sasktel.net  escreveu:
> > Here is some Ram drive code I have for the Saturn.
> > 
> > There looks to be 8 subbanks within 1 or 2 main banks depending if its 64 or 128 kb.
> > 
> > The subbanks are accessed from $C084.C087 and $C00C.C00F.
> > 
> > ; ProDOS disk driver for 
> > ; 64k/128k Saturn ramcard 
> > ; 
> > ; Matthew Lee Stier 
> > ; Sun Microsystems 
> > 
> > 
> > 	ORG $2000 
> > 	LDY #7 
> > NXTBLK16	LDX LC.INDEX,Y	; TEST FOR RAMCARD 
> > 	LDA $C080,X 	; SIZE 
> > 	BIT $C083 	; ENABLE BANK 2 
> > 	BIT $C083 
> > 	TYA 
> > 	STA $D000 
> > 	DEY 
> > 	BPL NXTBLK16 
> > 	BIT $C08F 
> > 	LDY $D000 	; GET RAMCARD SIZE 
> > 	BIT $C084 	; RESTORE BLOCK 0 
> > 	BIT $C089 	; BANK 1 
> > 	BIT $C089 
> > 	CPY #8	; ILLEGAL BANK SIZE? 
> > 	BCS TESTRTN 	; YES... 
> > 	TYA 
> > 	ASL 
> > 	TAY 
> > 	LDX BLKSIZE+1,Y 	; GET RAMCARD SIZE 
> > 	LDA BLKSIZE,Y 	; IN PRODOS BLOCKS 
> > 	BNE INSTALL 	; IF NOT 0 THEN INSTALL 
> > 	SEC 		; ELSE FLAG AN ERROR 
> > TESTRTN	RTS 		; AND QUIT 
> > 
> > LC.INDEX	DB $4,$5,$6,$7,$C,$D,$E,$F 
> > BLKSIZE	DW 0,0,0,96,0,0,0,224



I don't have the assembled code for this driver.  Just the source code that was down loaded from the internet.  If you have it, then you have what I have.  I don't have a Saturn card so didn't see a need to assemble the code.

Were you able to assemble the source and check it for any discrepancies?

[toc] | [prev] | [next] | [standalone]

#1506

FromLeandro Polimeno <leandro.polimeno@gmail.com>
Date2015-02-11 05:22 -0800
Message-ID<f560888f-bed1-4d01-8a26-9ae7505ddbda@googlegroups.com>
In reply to#1505
I have it assembled but for my Saturn clone didn´t work on any version of ProDOS all.

Let me got at home and send an image.

Regards,

Polimeno

Em quarta-feira, 11 de fevereiro de 2015 03:19:20 UTC-2, gid...@sasktel.net  escreveu:
> On Tuesday, February 10, 2015 at 6:53:38 AM UTC-6, Leandro Polimeno wrote:
> > Humm,
> > 
> > But that code, RAMDrive for 64k/128k Saturn by Matthew Lee Stier, works on ProDOS 1.9 ?! I have it and can´t play with.
> > 
> > Can you provide an image for test ?!
> > 
> > Regards,
> > 
> > Polimeno
> > 
> > Em domingo, 18 de janeiro de 2015 04:54:46 UTC-2, gid...@sasktel.net  escreveu:
> > > Here is some Ram drive code I have for the Saturn.
> > > 
> > > There looks to be 8 subbanks within 1 or 2 main banks depending if its 64 or 128 kb.
> > > 
> > > The subbanks are accessed from $C084.C087 and $C00C.C00F.
> > > 
> > > ; ProDOS disk driver for 
> > > ; 64k/128k Saturn ramcard 
> > > ; 
> > > ; Matthew Lee Stier 
> > > ; Sun Microsystems 
> > > 
> > > 
> > > 	ORG $2000 
> > > 	LDY #7 
> > > NXTBLK16	LDX LC.INDEX,Y	; TEST FOR RAMCARD 
> > > 	LDA $C080,X 	; SIZE 
> > > 	BIT $C083 	; ENABLE BANK 2 
> > > 	BIT $C083 
> > > 	TYA 
> > > 	STA $D000 
> > > 	DEY 
> > > 	BPL NXTBLK16 
> > > 	BIT $C08F 
> > > 	LDY $D000 	; GET RAMCARD SIZE 
> > > 	BIT $C084 	; RESTORE BLOCK 0 
> > > 	BIT $C089 	; BANK 1 
> > > 	BIT $C089 
> > > 	CPY #8	; ILLEGAL BANK SIZE? 
> > > 	BCS TESTRTN 	; YES... 
> > > 	TYA 
> > > 	ASL 
> > > 	TAY 
> > > 	LDX BLKSIZE+1,Y 	; GET RAMCARD SIZE 
> > > 	LDA BLKSIZE,Y 	; IN PRODOS BLOCKS 
> > > 	BNE INSTALL 	; IF NOT 0 THEN INSTALL 
> > > 	SEC 		; ELSE FLAG AN ERROR 
> > > TESTRTN	RTS 		; AND QUIT 
> > > 
> > > LC.INDEX	DB $4,$5,$6,$7,$C,$D,$E,$F 
> > > BLKSIZE	DW 0,0,0,96,0,0,0,224
> 
> 
> 
> I don't have the assembled code for this driver.  Just the source code that was down loaded from the internet.  If you have it, then you have what I have.  I don't have a Saturn card so didn't see a need to assemble the code.
> 
> Were you able to assemble the source and check it for any discrepancies?

[toc] | [prev] | [next] | [standalone]

#1484

FromD Finnigan <dog_cow@macgui.com>
Date2015-01-18 10:09 +0000
Message-ID<dog_cow-1421575793@macgui.com>
In reply to#1482
gids.rs wrote:
> 
> I think you had better re-read David's first post.
> 
> This next line is not possible.
> 
>> The Saturn RAM cards use a series of addresses to select a bank, starting
> with $C084 for bank 0, $C085 for bank 1, $C086 for bank 2, and so on.
> 
> 
> A Saturn card can have up to 128 kb which is 16 banks of 16 kb.  Your code
> would be accessing $C084 - $C093.  You have a bug in your code.
> 
> Ergo, how the cards are stated to be accessed, are more than likely turned
> around.

Links to documentation in my first post.

The Legend card definitely uses only 1 address, and that's $C084, as a bank
select. It's write-only. Refer to PDF page 15.

The problem is that Saturn uses multiple addressses, but they're not
contiguous. Once you get above a certain bank number, there's a jump. Refer
to PDF page 73.

> 
> See what you started David?  You listed buggy code.

OK, here's the code that came to my head this morning. This is closer to
what I had in mind.

As with my first listing, this is completely untested.

So you start out with

RAMTYPE EQU $06 ; ZERO-PAGE LOCATION
BANKSEL EQU $C084

same as last time, but now your subroutine looks like this:

AND RAMTYPE
TAX
STA BANKSEL,X
RTS

The premise here being that an earlier subroutine has either stored a 0 in
RAMTYPE for the Legend, or it has stored the maxinum number of banks for the
Saturn card. The desired bank number is still sent in A. So there's a mask
to set up the X-register to use either the bank index for Saturn, or just
$C084,0 for the Legend.

Short and sweet.

Except that apparently it will still fail on the high bank numbers for
Saturn because it jumps from $C087 to $C08C.

-- 
]DF$
Apple II Book: http://macgui.com/newa2guide/
Usenet: http://macgui.com/usenet/  <-- get posts by email!
Apple II Web & Blog hosting: http://a2hq.com/

[toc] | [prev] | [next] | [standalone]

#1485

From"Anton Treuenfels" <teamtempest@yahoo.com>
Date2015-01-19 08:33 -0600
Message-ID<LuOdne0v5K89iCDJnZ2dnUU7-I-dnZ2d@earthlink.com>
In reply to#1484
"D Finnigan" <dog_cow@macgui.com> wrote in message 
news:dog_cow-1421575793@macgui.com...
> So you start out with
>
> RAMTYPE EQU $06 ; ZERO-PAGE LOCATION
> BANKSEL EQU $C084
>
> same as last time, but now your subroutine looks like this:
>
> AND RAMTYPE
> TAX
> STA BANKSEL,X
> RTS
>
> The premise here being that an earlier subroutine has either stored a 0 in
> RAMTYPE for the Legend, or it has stored the maxinum number of banks for 
> the
> Saturn card. The desired bank number is still sent in A. So there's a mask
> to set up the X-register to use either the bank index for Saturn, or just
> $C084,0 for the Legend.

Uh, not quite. If you store $00 in RAMTYPE, then the AND will set the 
A-register to zero as well. So this code will never select anything but bank 
zero for the Legend.

One way out is to create an offset table for the Saturn:

  LDX RAMTYPE
  BEQ setLegend
  TAY
  LDX offsetSaturn
setLegend
  STA BANKSELECT,X
  RTS

offsetSaturn
  DFB 0,1,2,3,4,5,6,7,12,13,14,....

Without a table:

  LDX RAMTYPE
  BEQ setLegend
  CMP #8
  BCC lowSaturn
  ADC #4-1 ; 8->12, 9->13, etc
lowSaturn
  TAX
setLegend
  STA BANKSELECT,X
  RTS

That's assuming that the actual value stored from the A-register matters for 
the Legend but only the address written to matters for the Saturn.

- Anton Treuenfels

[toc] | [prev] | [next] | [standalone]

#1486

FromD Finnigan <dog_cow@macgui.com>
Date2015-01-19 18:29 +0000
Message-ID<dog_cow-1421692212@macgui.com>
In reply to#1485
Anton Treuenfels wrote:
> 
> Uh, not quite. If you store $00 in RAMTYPE, then the AND will set the 
> A-register to zero as well. So this code will never select anything but

I would have gotten it right on the third try. :-)

> 
> That's assuming that the actual value stored from the A-register matters
> for the Legend but only the address written to matters for the Saturn.

Yes.

[toc] | [prev] | [next] | [standalone]

#1490

Fromqkumba <peter.ferrie@gmail.com>
Date2015-01-20 09:33 -0800
Message-ID<81a9de46-b503-44aa-825c-6251d4221b8f@googlegroups.com>
In reply to#1485
>   LDX offsetSaturn

LDX offsetSaturn, Y

However, that's still too many accesses.  According to the documentation, it should be 4 and 4 for bank select, not 8 and 8.  The other set of 4 and 4 are ROM/RAM access and write protection control, within the currently selected bank.

> Without a table:
> 
>   LDX RAMTYPE
>   BEQ setLegend
>   CMP #8
>   BCC lowSaturn
>   ADC #4-1 ; 8->12, 9->13, etc
> lowSaturn
>   TAX
> setLegend
>   STA BANKSELECT,X
>   RTS

So, limiting the access and removing one branch:

    AND RAMTYPE
    CMP #4
    BCC lowSaturn
    ADC #4-1 ; 4->8, 5->9, 6->10, 7->11
lowSaturn
    TAX
    STA BANKSELECT,X
    RTS

where RAMTYPE is 0 for Legend, and the maximum bank-1 for Saturn (3 or 7).

[toc] | [prev] | [next] | [standalone]

#1491

From"Anton Treuenfels" <teamtempest@yahoo.com>
Date2015-01-20 17:06 -0600
Message-ID<2Mydndkf3ejzQiPJnZ2dnUU7-dOdnZ2d@earthlink.com>
In reply to#1490
"qkumba" <peter.ferrie@gmail.com> wrote in message 
news:81a9de46-b503-44aa-825c-6251d4221b8f@googlegroups.com...
>>   LDX offsetSaturn
>
> LDX offsetSaturn, Y

yes...that's what it should have been, and what I meant to write. But as you 
point out, a table isn't the best solution anyway!

- Anton Treuenfels

[toc] | [prev] | [standalone]

Back to top | Article view | comp.sys.apple2.programmer

csiph-web