ShuriZma
/
desmume
mirror of https://github.com/TASEmulators/desmume.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

fixed y-sorting that was messed up by the more convoluted multi-purpose indirect clipping approach that was added in early 2019 (e.g. 4cd19ce522 and e06d11f6df ) 

fixes SF bug 1724 (7th dragon HP renderer) but this was probably affecting many other games.
This commit is contained in:
zeromus 2020-12-15 00:58:19 -05:00
parent b138666bf6
commit 50a24fbbd7
1 changed files with 6 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
desmume/src/gfx3d.cpp
View File

@ -2293,8 +2293,8 @@ void gfx3d_glFlush(u32 v)
static bool gfx3d_ysort_compare(int num1, int num2)
{
const POLY &poly1 = gfx3d.polylist->list[num1];
const POLY &poly2 = gfx3d.polylist->list[num2];
const POLY &poly1 = *_clippedPolyUnsortedList[num1].poly;
const POLY &poly2 = *_clippedPolyUnsortedList[num2].poly;
//this may be verified by checking the game create menus in harvest moon island of happiness
//also the buttons in the knights in the nightmare frontend depend on this and the perspective division
@ -2401,8 +2401,10 @@ void gfx3d_GenerateRenderLists(const ClipperMode clippingMode)
//find the min and max y values for each poly.
//the w-division here is just an approximation to fix the shop in harvest moon island of happiness
//also the buttons in the knights in the nightmare frontend depend on this
size_t ysortCount = (gfx3d.state.sortmode) ? gfx3d.clippedPolyOpaqueCount : gfx3d.clippedPolyCount;
for (size_t i = 0; i < ysortCount; i++)
//we process all polys here instead of just the opaque ones (which have been sorted to the beginning of the index list earlier) because
//1. otherwise it is basically two passes through the list and will probably run slower than one pass through the list
//2. most geometry is opaque which is always sorted anyway
for (size_t i = 0; i < gfx3d.clippedPolyCount; i++)
{
// TODO: Possible divide by zero with the w-coordinate.
// Is the vertex being read correctly? Is 0 a valid value for w?