Commit Graph

483 Commits

Author SHA1 Message Date
nmlgc 21b4a16512 th02/main: Identify some more data referenced when displaying the BGM title 2014-07-06 01:52:54 +02:00
nmlgc 550b037521 th05/maine: Identify some missing All Cast data references 2014-07-05 23:45:26 +02:00
nmlgc 2bd664c5e4 Compress unitialized memory by replacing bytes with dwords 2014-07-05 23:08:34 +02:00
nmlgc 174e08fc60 th05: Fix all remaining alignment errors
These stray NOPs and "db 0" declarations were exactly what made the recompiled
executables comparatively large.
2014-07-05 22:29:58 +02:00
nmlgc 31f6554d1c th05/main: Identify the score numeral bitmaps 2014-07-05 01:36:16 +02:00
nmlgc 0b0013beb0 th04/main: Identify the score numeral bitmaps 2014-07-05 01:30:20 +02:00
nmlgc 8586dbe98c th02/main: Identify the score numeral bitmaps 2014-07-04 21:41:03 +02:00
nmlgc 92e5e772be th05/maine: Declare some missing strings as such 2014-07-04 18:19:46 +02:00
nmlgc 0ff83ca74e th05/op: Identify menu description strings 2014-07-04 15:00:44 +02:00
nmlgc 4eceb05d9d th05/op: Identify sound setup strings
More relocation changes?
2014-07-04 14:55:07 +02:00
nmlgc 5a1b691f21 th05/op: Identify Music Room strings 2014-07-04 14:50:35 +02:00
nmlgc 53fa9b1cfb th05/op: Declare some missing strings as such
Well, nobody actually creates perfect code the first time around, except
Linus Torvalds. (http://youtu.be/4XpnKHJAok8?t=22m18s)
2014-07-04 14:42:38 +02:00
nmlgc 17f6ac7d1d th02/main: Identify stage and BGM title pointers 2014-07-03 16:10:24 +02:00
nmlgc fc5b599a63 th04/main: Identify stage and BGM title pointers 2014-07-03 08:29:18 +02:00
nmlgc 60195806f8 th03/mainl: Identify character name and title string pointers
For some reason, this ends up modifying the relocations in the .exe header.
Oh well, it's not code, so... *shrug*
2014-07-02 23:49:19 +02:00
nmlgc 8106ddf80f th05/maine: Identify All Cast string pointers 2014-07-02 23:15:59 +02:00
nmlgc 267f9eb90a th02/main: Initial state 2014-07-02 21:00:04 +02:00
nmlgc 4ae365a60b th02/maine: Initial state 2014-07-02 19:43:15 +02:00
nmlgc 3186da7cec th02/op: Initial state 2014-07-02 19:18:32 +02:00
nmlgc d569ad64f2 th03/main: Initial state 2014-07-01 23:25:51 +02:00
nmlgc cdf5141dee th03/mainl: Initial state 2014-07-01 20:30:39 +02:00
nmlgc bd5099f904 th03/op: Initial state 2014-06-30 00:52:27 +02:00
nmlgc 93360e2a13 th05/main: Split off segments 3 and 4 into a separate file
Getting rid of all the hash space hacks.
2014-06-30 00:00:24 +02:00
nmlgc 48ce5beaea th04/main: Initial state
To anyone else encountering TASM's "Near jump or call to different CS" issue:
Just split off the offending segment into a different source file and include
it from the main one.
2014-06-29 23:04:24 +02:00
nmlgc 6afe3e89a3 th04/maine: Initial state 2014-06-29 18:45:48 +02:00
nmlgc bf430e4231 th04/op: Initial state 2014-06-29 16:51:19 +02:00
nmlgc 82bc23ec58 th05/main: Identify Stage 4 boss theme strings 2014-06-29 03:21:02 +02:00
nmlgc 467a7db913 th05/main: Initial state 2014-06-29 02:43:13 +02:00
nmlgc 21b8ca62ab th05/maine: Initial state 2014-06-28 22:13:04 +02:00
nmlgc 91cb632353 th05/op: Additional alignment fixes
Now, we're fully identical to the original build - save for three instructions
that are getting assembled using different, but equivalent opcodes.
2014-06-28 20:38:31 +02:00
nmlgc 224f92c77f th05/op: Remove the duplicate segment alignment parameter for seg000 2014-06-28 20:38:30 +02:00
nmlgc adb4a66881 Fix a typo in the build parameters
And I've been wondering why this didn't work anymore...
2014-06-28 20:38:14 +02:00
nmlgc 9e07c54aee Start out with th05's OP.EXE
This is the raw output of IDA's initial analysis of the original OP.EXE,
touched up in order to compile to an executable identical to the original
build.
2014-06-26 22:47:15 +02:00