Re: EOR 02D6,Y For Disk II Family Again

[Holger Picker <pickerh@uni-muenster.de>]

On Mon, 24 Feb 2003, Bryan Parkoff wrote:

> Date: Mon, 24 Feb 2003 17:53:50 -0600
> From: Bryan Parkoff <BParkoff@satx.rr.com>
> Newsgroups: comp.sys.apple2
> Subject: EOR 02D6,Y For Disk II Family Again
>
>     I have taken a look at the PROM code below.  It does not make sense to
> me when I look at EOR 02D6,Y at address: C6AF.  I changed from EOR 02D6,Y to
> TYA NOP NOP before I execute at address C6A6.  I do see that all 56 bytes
> are stored in $300-$356 using encoded translated characters (96-->FC).  I
> changed from TYA NOP NOP back to EOR 02D6,Y before I execute at address
> C6A6.  I do see that all 56 bytes are stored in $300-$356 using decoded
> translated characters, but 6-2 is NOT decoded YET.
>     I am trying to figure out how 96-->FC are translated back to the
> original byte.  For example, $DB is in $354, $B6 is in $355, and $00 is in
> $356.  The loop starts at $56 before it decreases to $00.  LDY C08C,X reads
> $B6 in $355.  $B6 is stored in Y Register before EOR executes.  It may mean
> EOR 02D6,$B6.  $02D6+$B6=$38C.  It is how EOR locates $38C to get $12 value,
> but it does not translate correctly.  I know that translatation table is in
> $36C-$3D5.  Please advise.

I think, I see your problem here. Indeed, normally you would expect that an
instruction like LDA $2d6,y is used to decode the byte, but instead you find
an EOR here. The reason for the EOR is that at the same time you decode the
disk nibbles, you also calculate the checksum for the whole data field. Take
a look at the very last EOR instruction at address $c6d0. Here the
additional checksum value is loaded from disk, and again an EOR is used to
"xor" it with the previously calculated checksum. If the result is zero,
then the checksum test was successful (i.e. the calculated checksum and the
checksum read from disk are identical), if not, the program will start again
looking for the sector, because the data was not read correctly. To sum it
up: EOR is used for the additional checksum test. Don't worry about that the
bytes are xored while decoding them, because a similar method (the other way
round) was used to code the disk bytes before writing them to the disk.

Holger

> C6A6: A0 56        LDY #56
> C6A8: 84 3C        STY WorkByte (3C)
> C6AA: BC 8C C0     LDY StrobeDataLatch,X (C08C)
> C6AD: 10 FB        BPL C6AA {-05}
> C6AF: 59 D6 02     EOR 02D6,Y
> C6B2: A4 3C        LDY WorkByte (3C)
> C6B4: 88           DEY
> C6B5: 99 00 03     STA VectorPage,Y (0300)
> C6B8: D0 EE        BNE C6A8 {-12}
> C6BA: 84 3C        STY WorkByte (3C)
> C6BC: BC 8C C0     LDY StrobeDataLatch,X (C08C)
> C6BF: 10 FB        BPL C6BC {-05}
> C6C1: 59 D6 02     EOR 02D6,Y
> C6C4: A4 3C        LDY WorkByte (3C)
> C6C6: 91 26        STA (Low_SectorRdBufAddr),Y (26)
> C6C8: C8           INY
> C6C9: D0 EF        BNE C6BA {-11}
> C6CB: BC 8C C0     LDY StrobeDataLatch,X (C08C)
> C6CE: 10 FB        BPL C6CB {-05}
> C6D0: 59 D6 02     EOR 02D6,Y
> C6D3: D0 87        BNE C65C {-79}
>
>
>
> --
> Yours Truly,
>
> Bryan Parkoff
> BParkoff@satx.rr.com
>