* Mention why PrimDepth is used in PreRender_FetchFbufCoverage
* Clarify that this only affects the memory alpha blender input, not all alpha blender inputs
* Try to make the subtle difference between coverage and memory alpha more clear
* progress
* progress
* change
* named most
* name prerender things
* fix and format
* play prefix
* address most comments (still a few more things to do)
* fix error
* reword input test comments
* some more
* rework env draw flags
* add hilite
* merge and format
* init to proper values
* rename env draw flag names
* 7 to 6
* more small changes
* More documentation for PreRender.c
* Comment fix
* Suggested changes
* More on coverage reading functions
* Further comment changes
* Further suggestions
* Fix various small gfx issues
* Use 0xFFFF instead of -1 for max prim depth
* More fixed point numbers
* scale -> texCoordScale, better cycle 2 render mode in z_title
* GFXP_CHAR_* defines renamed to SPACING
* Fix some gDPLoadMultiBlock uses
* Cleanup `UNK_TYPE`, `UNK_PTR` usage
* Add some missing empty lines after declarations
* Remove some legacy comments from non-matching times
* Fix some grammar (mostly "it's"/"its")
* Use proper names for two symbols after ZAPD bugfix
* Cleanup `place_title_cards.xml`
* Use `NULL` to check against `D_8012D260` pointer
* Parentheses around some macro arguments
* wip proofread headers up to z64animation.h
* Less warnings in boot & code segments
* few more warnings gone
* Ran formatter
* z_view warning gone
* -> 1
* f31 -> 31
* Remove function casts
* Few more small improvements
* Separate declaration and assignment in func_80091738 and Item_Give
Co-authored-by: Thar0 <maximilianc64@gmail.com>
* Make func_800C213C match original ROM
The C code for this function was very close to done, and was
functionally equivalent to the ROM. This commit uses a few tricks
to convince the compiler to allocate the same registers as the original
ROM. The two tricks used are:
1. Introduced a new temporary variable "yinc".
2. Added a "do while(0) {}" line in two places.
Both of these ideas came from the decomp-permuter tool. I took the
do/while idea and guess as to what it might have been in the original
code. I imagined that the "do while(0) {}" was probably left over from
a standard multi-line macro, which for some reason had an empty body.
I further suggest that maybe the empty body is because it was a feature
compiled out during RELEASE builds - perhaps a debug logging macro.
* Use simpler code for func_800C213C
Replaced dummy LOG macro with "do {} while(0);". This version makes
less assumptions about what the original code looked like.
* cleanup libultra
* fixes
- use quotes instead of <> for includes
- add macros for zelda specific thread priorities
- fix Makefile
- properly format the remaining pfs structs
* fix button macros + add CHECK_BTN_ANY/CHECK_BTN_ALL
* remove ULTRA_ABS
* fix includes
* update z_player.c/z_lib.c + run format.sh
* merge upstream/master
* fix include in En_Goroiwa
* fix includes
* Decompile PreRender.c
* use NON_EQUIVALENT and run format.sh
* Get rid of some magic values
* rename dList/dListp to dListHead/dList
* make func_800C213C NON_MATCHING
* replace more values with SCREEN_WIDTH/SCREEN_HEIGHT
* dList/dListHead with gfxp/gfx
* Split files
* Format rodata
* Some more code rodata migrated
* Some more actor rodata migrated
* Migrate rodata for ovl_Boss_Ganon
* Migrate rodata for code_800EC960
* Remove unused rodata
* x1b occurences all lowercase b
