Document gfxalloc.c #2153
Document gfxalloc.c #2153
Conversation
include/functions.h
Outdated
| Gfx* Gfx_Open(Gfx* gfx); | ||
| Gfx* Gfx_Close(Gfx* gfx, Gfx* dst); | ||
| Gfx* Gfx_Open(Gfx* gfxDisp); | ||
| Gfx* Gfx_Close(Gfx* gfxDisp, Gfx* gfx); |
There was a problem hiding this comment.
why "disp" ? afaik that stands for "display"
There was a problem hiding this comment.
opted for disp to make the distinction that the first argument should be a display buffer pointer (i.e. POLY_OPA_DISP) rather than any old Gfx*.
src/code/z_play.c
Outdated
| Gfx* tempGfxDisp; | ||
| Gfx* lockedGfxDisp; | ||
|
|
||
| gfxP = Gfx_Open(sp1CC); | ||
| gSPDisplayList(OVERLAY_DISP++, gfxP); | ||
| tempGfxDisp = Gfx_Open(lockedGfxDisp = POLY_OPA_DISP); |
There was a problem hiding this comment.
I'm not convinced by the temp/locked naming scheme (nor "disp")
There was a problem hiding this comment.
The Gfx_Open(lockedGfxDisp = POLY_OPA_DISP); is awful too
There was a problem hiding this comment.
Gfx_Open(lockedGfxDisp = POLY_OPA_DISP); is unusual yes, but I'd argue that it makes sense to write unusual code to force POLY_OPA_DISP to appear inside the call to Gfx_Open in this case. If you think of the high level picture, that call to Gfx_Open is creating a side effect on POLY_OPA_DISP, it's taking all of the memory it has and giving it to something else.
When you have the assignment above Gfx_Open I feel that you lose a little bit of that connection.
|
I opted to create GFX_ALLOC_OPEN/GFX_ALLOC_CLOSE macros, since I think it makes the open/close process much cleaner |
Changes made in this pr are licensed under CC0.
This covers the documentation pass on gfxalloc.c I made in #1972. It'd be ideal if
Gfx* tempGfxDisp; Gfx* lockedGfxDisp;could be combined into one struct, but there's one function where the regalloc simply doesn't work.