nick/WickedEngine
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

modified light culling compute shader

Browse Source
This commit is contained in:
turanszkij
2017-01-16 22:28:39 +01:00
parent 1bd2a1ddf5
commit 731acbb9be
1 changed files with 37 additions and 44 deletions
Download Patch File Download Diff File Expand all files Collapse all files
WickedEngine/lightCullingCS.hlsl
+37 -44
View File
@@ -227,6 +227,43 @@ void main(ComputeShaderInput IN)
t_LightGrid[IN.groupID.xy] = uint2(t_LightIndexStartOffset, t_LightCount);
}
// Sort the decals:
// TODO: Optimize for parallel sorting! (Merge sort?)
GroupMemoryBarrierWithGroupSync();
// first thread sorts opaques
if (IN.groupIndex == 0 && o_LightCount > 0)
{
for (i = 0; i < o_LightCount - 1; ++i)
{
for (uint j = i + 1; j < o_LightCount; ++j)
{
if (o_LightList[i] > o_LightList[j])
{
uint swap = o_LightList[i];
o_LightList[i] = o_LightList[j];
o_LightList[j] = swap;
}
}
}
}
// second thread sorts transparents
if (IN.groupIndex == 0 && t_LightCount > 0)
{
for (i = 0; i < t_LightCount - 1; ++i)
{
for (uint j = i + 1; j < t_LightCount; ++j)
{
if (t_LightList[i] > t_LightList[j])
{
uint swap = t_LightList[i];
t_LightList[i] = t_LightList[j];
t_LightList[j] = swap;
}
}
}
}
GroupMemoryBarrierWithGroupSync();
// Now update the light index list (all threads).
@@ -241,50 +278,6 @@ void main(ComputeShaderInput IN)
t_LightIndexList[t_LightIndexStartOffset + i] = t_LightList[i];
}
// Sort the decals:
// The sorting *could be optimized* when it is needed, now this should work
// TODO: Try some other algorithms and choose fastest
GroupMemoryBarrierWithGroupSync();
// first thread sorts opaques
if (IN.groupIndex == 0 && o_LightCount > 0)
{
for (i = 0; i < o_LightCount - 1; ++i)
{
if (Lights[o_LightIndexList[o_LightIndexStartOffset + i]].type == 100) /*DECAL*/
{
for (uint j = i + 1; j < o_LightCount; ++j)
{
if (o_LightIndexList[o_LightIndexStartOffset + i] > o_LightIndexList[o_LightIndexStartOffset + j])
{
uint swap = o_LightIndexList[o_LightIndexStartOffset + i];
o_LightIndexList[o_LightIndexStartOffset + i] = o_LightIndexList[o_LightIndexStartOffset + j];
o_LightIndexList[o_LightIndexStartOffset + j] = swap;
}
}
}
}
}
// second thread sorts transparents
if (IN.groupIndex == 1 && t_LightCount > 0)
{
for (i = 0; i < t_LightCount - 1; ++i)
{
if (Lights[t_LightIndexList[t_LightIndexStartOffset + i]].type == 100) /*DECAL*/
{
for (uint j = i + 1; j < t_LightCount; ++j)
{
if (t_LightIndexList[t_LightIndexStartOffset + i] > t_LightIndexList[t_LightIndexStartOffset + j])
{
uint swap = t_LightIndexList[t_LightIndexStartOffset + i];
t_LightIndexList[t_LightIndexStartOffset + i] = t_LightIndexList[t_LightIndexStartOffset + j];
t_LightIndexList[t_LightIndexStartOffset + j] = swap;
}
}
}
}
}
#ifdef DEBUG_TILEDLIGHTCULLING
DebugTexture[texCoord] = float4((float3)_counter / (float)lightCount,1);
#endif