Emulator-Archive/ppsspp
1
0
Fork 0
mirror of https://github.com/hrydgard/ppsspp.git synced 2024-11-23 21:39:52 +00:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

Small optimizations and fixes to vertex decoder

Browse Source
This commit is contained in:
Henrik Rydgard 2013-02-15 21:41:29 +01:00
parent 61e13e4475
commit 0eabc80c3e
2 changed files with 25 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
GPU/GLES/TransformPipeline.cpp
View File

@ -283,8 +283,13 @@ void Lighter::Light(float colorOut0[4], float colorOut1[4], const float colorIn[
bool doSpecular = (comp != GE_LIGHTCOMP_ONLYDIFFUSE);
bool poweredDiffuse = comp == GE_LIGHTCOMP_BOTHWITHPOWDIFFUSE;
float dot = toLight * norm;
float distanceToLight = toLight.Length();
float dot = 0.0f;
if (distanceToLight > 0.0f)
{
toLight /= distanceToLight;
dot = toLight * norm;
}
// Clamp dot to zero.
if (dot < 0.0f) dot = 0.0f;
@ -292,10 +297,9 @@ void Lighter::Light(float colorOut0[4], float colorOut1[4], const float colorIn[
dot = powf(dot, specCoef_);
float lightScale = 1.0f;
float distance = toLight.Normalize();
if (type != GE_LIGHTTYPE_DIRECTIONAL)
{
lightScale = 1.0f / (gstate_c.lightatt[l][0] + gstate_c.lightatt[l][1]*distance + gstate_c.lightatt[l][2]*distance*distance);
lightScale = 1.0f / (gstate_c.lightatt[l][0] + gstate_c.lightatt[l][1]*distanceToLight + gstate_c.lightatt[l][2]*distanceToLight*distanceToLight);
if (lightScale > 1.0f) lightScale = 1.0f;
}

33
GPU/GLES/VertexDecoder.cpp
View File

@ -144,18 +144,16 @@ void VertexDecoder::Step_WeightsFloat() const
void VertexDecoder::Step_TcU8() const
{
u8 *uv = (u8*)(decoded_ + decFmt.uvoff);
const u8 *uvdata = (const u8*)(ptr_ + tcoff);
uv[0] = uvdata[0];
uv[1] = uvdata[1];
u16 *uv = (u16*)(decoded_ + decFmt.uvoff);
const u16 *uvdata = (const u16*)(ptr_ + tcoff);
*uv = *uvdata;
}
void VertexDecoder::Step_TcU16() const
{
u16 *uv = (u16 *)(decoded_ + decFmt.uvoff);
const u16 *uvdata = (const u16*)(ptr_ + tcoff);
uv[0] = uvdata[0];
uv[1] = uvdata[1];
u32 *uv = (u32 *)(decoded_ + decFmt.uvoff);
const u32 *uvdata = (const u32*)(ptr_ + tcoff);
*uv = *uvdata;
}
void VertexDecoder::Step_TcU16Through() const
@ -305,7 +303,7 @@ void VertexDecoder::Step_NormalS16() const
xorval = 0xFFFF;
const s16 *sv = (const s16*)(ptr_ + nrmoff);
for (int j = 0; j < 3; j++)
normal[j] = sv[j] ^ xorval ;
normal[j] = sv[j] ^ xorval;
normal[3] = 0;
}
@ -330,9 +328,10 @@ void VertexDecoder::Step_NormalS8Morph() const
if (gstate.reversenormals & 1) {
multiplier = -multiplier;
}
const s8 *sv = (const s8*)(ptr_ + onesize_*n + nrmoff);
const s8 *bv = (const s8*)(ptr_ + onesize_*n + nrmoff);
multiplier *= (1.0f/127.0f);
for (int j = 0; j < 3; j++)
normal[j] += (sv[j]/32767.0f) * multiplier;
normal[j] += bv[j] * multiplier;
}
}
@ -346,9 +345,10 @@ void VertexDecoder::Step_NormalS16Morph() const
if (gstate.reversenormals & 1) {
multiplier = -multiplier;
}
const float *fv = (const float*)(ptr_ + onesize_*n + nrmoff);
const s16 *sv = (const s16 *)(ptr_ + onesize_*n + nrmoff);
multiplier *= (1.0f/32767.f);
for (int j = 0; j < 3; j++)
normal[j] += fv[j] * multiplier;
normal[j] += sv[j] * multiplier;
}
}
@ -425,9 +425,10 @@ void VertexDecoder::Step_PosS8Morph() const
float *v = (float *)(decoded_ + decFmt.posoff);
memset(v, 0, sizeof(float) * 3);
for (int n = 0; n < morphcount; n++) {
float multiplier = 1.0f / 127.0f;
const s8 *sv = (const s8*)(ptr_ + onesize_*n + posoff);
for (int j = 0; j < 3; j++)
v[j] += (sv[j] / 127.f) * gstate_c.morphWeights[n];
v[j] += (float)sv[j] * (multiplier * gstate_c.morphWeights[n]);
}
}
@ -437,9 +438,9 @@ void VertexDecoder::Step_PosS16Morph() const
memset(v, 0, sizeof(float) * 3);
for (int n = 0; n < morphcount; n++) {
float multiplier = 1.0f / 32767.0f;
const short *sv = (const short*)(ptr_ + onesize_*n + posoff);
const s16 *sv = (const s16*)(ptr_ + onesize_*n + posoff);
for (int j = 0; j < 3; j++)
v[j] += (sv[j] * multiplier) * gstate_c.morphWeights[n];
v[j] += (float)sv[j] * (multiplier * gstate_c.morphWeights[n]);
}
}