mm
mm copied to clipboard
Bss script update
- Update the bss script to use mapfile parser
- Add flag to specify sections other than
.bss, allowing to check for reordering in other sections. - Hardcode the path to the mapfile
- This way it is more consistent with the other scripts.
- Some code cleanups
Misc changes:
- Update mapfile_parser version to 2.x.x
- The most remarkable change is the actual mapfile parsing was rewritten in Rust, so it should be faster
- Changed the
python-Levenshteinrequirement toLevenshtein- According to its PyPI page the project was renamed, so I think it is better to use the new name instead.
If it sees a symbol is shifted then it will complain. Example:
f32 sKaleidoSetupRightPageEyeX[] = {
PAUSE_EYE_DIST * -PAUSE_MAP_X, // PAUSE_ITEM
PAUSE_EYE_DIST * -PAUSE_QUEST_X, // PAUSE_MAP
PAUSE_EYE_DIST * -PAUSE_MASK_X, // PAUSE_QUEST
PAUSE_EYE_DIST * -PAUSE_ITEM_X, // PAUSE_MASK
0 // Added this
};
f32 sKaleidoSetupRightPageEyeZ[] = {
PAUSE_EYE_DIST * -PAUSE_MAP_Z, // PAUSE_ITEM
PAUSE_EYE_DIST * -PAUSE_QUEST_Z, // PAUSE_MAP
PAUSE_EYE_DIST * -PAUSE_MASK_Z, // PAUSE_QUEST
PAUSE_EYE_DIST * -PAUSE_ITEM_Z, // PAUSE_MASK
};
Build mapfile: build/mm.map
Expected mapfile: expected/build/mm.map
Symbol Name,Build Address,Build File,Expected Address,Expected File,Difference,GOOD/BAD/MISSING
sKaleidoSetupRightPageEyeZ,801BDB1C,src/code/z_kaleido_setup.c,801BDB1C,src/code/z_kaleido_setup.c,4,BAD
BAD
Symbol reordering in src/code/z_kaleido_setup.c
Symbols are REORDERED!!
Oh! MY GOD!!
is it intended both the build and expected addresses are 801BDB1C in this example?
No, that's a bug on the library