Sprinter/mame
2
0
Fork 0
mirror of https://github.com/holub/mame synced 2025-10-04 16:34:53 +03:00
Code Issues Actions 6 Packages Projects Releases Wiki Activity

more detailed comment (nw)

Browse Source
This commit is contained in:
Vas Crabb 2016-06-10 17:50:52 +10:00
parent 37d8de37fa
commit cfece7e1bc
1 changed files with 29 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
src/mame/drivers/mystwarr.cpp
View File

@ -95,25 +95,40 @@ Notes:
Notes on metamrpha. Research done by Vas Crabb
The one rom that is different than the parent set has 2 bytes which have switched values (0 to 1 and 1 to 0).
This appears to be a deliberate change in the main CPU progam function that begins at 03A86E, affecting a
fragment towards the end of the function where it sets the return value:
03A90E: B247 cmp.w D7, D1
03A910: 6604 bne $3a916
03A912: 08C0 0001 bset #$1, D0
03A916: 3E3C 6B6E move.w #$6b6e, D7
03A91A: 0647 7F92 addi.w #$7f92, D7
03A91E: B447 cmp.w D7, D2
03A920: 6604 bne $3a926
03A922: 08C0 0000 bset #$0, D0
03A8FC moveq #$0, D0
03A8FE move.w ($0,A5), D1
03A902 move.w ($2,A5), D2
03A906 move.w #$b61d, D7
03A90A addi.w #$3f1b, D7
03A90E cmp.w D7, D1
03A910 bne $3a916
03A912 bset #$1, D0
03A916 move.w #$6b6e, D7
03A91A addi.w #$7f92, D7
03A91E cmp.w D7, D2
03A920 bne $3a926
03A922 bset #$0, D0
In this snippet of code where the change is, from the original metamrph set
D0 is the integer return value of the function in C ABI, so it's
if (blah) result |= 2; ... if (blah) result |= 1; ...
The bset #$1,D0 and bset #$0,D0 instructions are reversed. This code looks like poorly optimised compiled
C code. Remember D0 is the integer return value register in M68k C ABI and int is a 16-bit integer.
The code as above amounts to:
In the new dump, the "bset #$1, D0" and "bset #$0, D0" are reversed.
if (blah) result |= 1; ... if (blah) result |= 2; ...
unsigned result = 0;
if (a == (46621U + 16155U)) result |= 2;
if (b == (27502U + 32658U)) result |= 1;
In the alternate set this is changed to:
unsigned result = 0;
if (a == (46621U + 16155U)) result |= 1;
if (b == (27502U + 32658U)) result |= 2;
So I would bet one is a bug fix to the other because someone forgot what each bit in the return value meant.
There's no way a bad dump could be that precise.
There's no way a bad dump could be that precise and logical. However without further investigation/tracing
we have no way of knowing which is the later/corrected version.
**************************************************************************/