forked from bartvdbraak/blender
42557f90bd
This commit extend the technique of dynamic linked list to the mesh slots so as to eliminate dumb scan or map lookup. It provides massive performance improvement in the culling and in the rasterizer when the majority of objects are static. Other improvements: - Compute the opengl matrix only for objects that are visible. - Simplify hash function for GEN_HasedPtr - Scan light list instead of general object list to render shadows - Remove redundant opengl calls to set specularity, shinyness and diffuse between each mesh slots. - Cache GPU material to avoid frequent call to GPU_material_from_blender - Only set once the fixed elements of mesh slot - Use more inline function The following table shows the performance increase between 2.48, 1st round and this round of improvement. The test was done with a scene containing 40000 objects, of which 1000 are in the view frustrum approximately. The object are simple textured cube to make sure the GPU is not the bottleneck. As some of the rasterizer processing time has moved under culling, I present the sum of scenegraph(includes culling)+rasterizer time Scenegraph+rasterizer(ms) 2.48 1st round 3rd round All objects static, 323.0 86.0 7.2 all visible, 1000 in the view frustrum All objects static, 219.0 49.7 N/A(*) all invisible. All objects moving, 323.0 105.6 34.7 all visible, 1000 in the view frustrum Scene destruction 40min 40min 4s (*) : this time is not representative because the frame rate was at 60fps. In that case, the GPU holds down the GE by frame sync. By design, the overhead of the rasterizer is 0 when the the objects are invisible. This table shows a global speed up between 9x and 45x compared to 2.48a for scenegraph, culling and rasterizer overhead. The speed up goes much higher when objects are invisible. An additional 2-4x speed up is possible in the scenegraph by upgrading the Moto library to use Eigen2 BLAS library instead of C++ classes but the scenegraph is already so fast that it is not a priority right now. Next speed up in logic: many things to do there...
163 lines
4.1 KiB
C++
163 lines
4.1 KiB
C++
|
|
#include "DNA_customdata_types.h"
|
|
#include "DNA_material_types.h"
|
|
#include "DNA_scene_types.h"
|
|
|
|
#include "BKE_global.h"
|
|
#include "BKE_main.h"
|
|
|
|
#include "BL_BlenderShader.h"
|
|
#include "BL_Material.h"
|
|
|
|
#include "GPU_extensions.h"
|
|
#include "GPU_material.h"
|
|
|
|
#include "RAS_BucketManager.h"
|
|
#include "RAS_MeshObject.h"
|
|
#include "RAS_IRasterizer.h"
|
|
|
|
BL_BlenderShader::BL_BlenderShader(KX_Scene *scene, struct Material *ma, int lightlayer)
|
|
:
|
|
mScene(scene),
|
|
mMat(ma),
|
|
mLightLayer(lightlayer),
|
|
mGPUMat(NULL)
|
|
{
|
|
mBlenderScene = scene->GetBlenderScene();
|
|
mBlendMode = GPU_BLEND_SOLID;
|
|
|
|
ReloadMaterial();
|
|
}
|
|
|
|
BL_BlenderShader::~BL_BlenderShader()
|
|
{
|
|
if(mGPUMat)
|
|
GPU_material_unbind(mGPUMat);
|
|
}
|
|
|
|
void BL_BlenderShader::ReloadMaterial()
|
|
{
|
|
mGPUMat = (mMat) ? GPU_material_from_blender(mBlenderScene, mMat) : NULL;
|
|
}
|
|
|
|
void BL_BlenderShader::SetProg(bool enable, double time)
|
|
{
|
|
if(VerifyShader()) {
|
|
if(enable)
|
|
GPU_material_bind(mGPUMat, mLightLayer, mBlenderScene->lay, time);
|
|
else
|
|
GPU_material_unbind(mGPUMat);
|
|
}
|
|
}
|
|
|
|
int BL_BlenderShader::GetAttribNum()
|
|
{
|
|
GPUVertexAttribs attribs;
|
|
int i, enabled = 0;
|
|
|
|
if(!VerifyShader())
|
|
return enabled;
|
|
|
|
GPU_material_vertex_attributes(mGPUMat, &attribs);
|
|
|
|
for(i = 0; i < attribs.totlayer; i++)
|
|
if(attribs.layer[i].glindex+1 > enabled)
|
|
enabled= attribs.layer[i].glindex+1;
|
|
|
|
if(enabled > BL_MAX_ATTRIB)
|
|
enabled = BL_MAX_ATTRIB;
|
|
|
|
return enabled;
|
|
}
|
|
|
|
void BL_BlenderShader::SetAttribs(RAS_IRasterizer* ras, const BL_Material *mat)
|
|
{
|
|
GPUVertexAttribs attribs;
|
|
GPUMaterial *gpumat;
|
|
int i, attrib_num;
|
|
|
|
ras->SetAttribNum(0);
|
|
|
|
if(!VerifyShader())
|
|
return;
|
|
|
|
gpumat = mGPUMat;
|
|
|
|
if(ras->GetDrawingMode() == RAS_IRasterizer::KX_TEXTURED) {
|
|
GPU_material_vertex_attributes(gpumat, &attribs);
|
|
attrib_num = GetAttribNum();
|
|
|
|
ras->SetTexCoordNum(0);
|
|
ras->SetAttribNum(attrib_num);
|
|
for(i=0; i<attrib_num; i++)
|
|
ras->SetAttrib(RAS_IRasterizer::RAS_TEXCO_DISABLE, i);
|
|
|
|
for(i = 0; i < attribs.totlayer; i++) {
|
|
if(attribs.layer[i].glindex > attrib_num)
|
|
continue;
|
|
|
|
if(attribs.layer[i].type == CD_MTFACE) {
|
|
if(!mat->uvName.IsEmpty() && strcmp(mat->uvName.ReadPtr(), attribs.layer[i].name) == 0)
|
|
ras->SetAttrib(RAS_IRasterizer::RAS_TEXCO_UV1, attribs.layer[i].glindex);
|
|
else if(!mat->uv2Name.IsEmpty() && strcmp(mat->uv2Name.ReadPtr(), attribs.layer[i].name) == 0)
|
|
ras->SetAttrib(RAS_IRasterizer::RAS_TEXCO_UV2, attribs.layer[i].glindex);
|
|
else
|
|
ras->SetAttrib(RAS_IRasterizer::RAS_TEXCO_UV1, attribs.layer[i].glindex);
|
|
}
|
|
else if(attribs.layer[i].type == CD_TANGENT)
|
|
ras->SetAttrib(RAS_IRasterizer::RAS_TEXTANGENT, attribs.layer[i].glindex);
|
|
else if(attribs.layer[i].type == CD_ORCO)
|
|
ras->SetAttrib(RAS_IRasterizer::RAS_TEXCO_ORCO, attribs.layer[i].glindex);
|
|
else if(attribs.layer[i].type == CD_NORMAL)
|
|
ras->SetAttrib(RAS_IRasterizer::RAS_TEXCO_NORM, attribs.layer[i].glindex);
|
|
else if(attribs.layer[i].type == CD_MCOL)
|
|
ras->SetAttrib(RAS_IRasterizer::RAS_TEXCO_VCOL, attribs.layer[i].glindex);
|
|
else
|
|
ras->SetAttrib(RAS_IRasterizer::RAS_TEXCO_DISABLE, attribs.layer[i].glindex);
|
|
}
|
|
}
|
|
}
|
|
|
|
void BL_BlenderShader::Update(const RAS_MeshSlot & ms, RAS_IRasterizer* rasty )
|
|
{
|
|
float obmat[4][4], viewmat[4][4], viewinvmat[4][4], obcol[4];
|
|
GPUMaterial *gpumat;
|
|
|
|
gpumat = mGPUMat;
|
|
|
|
if(!gpumat || !GPU_material_bound(gpumat))
|
|
return;
|
|
|
|
MT_Matrix4x4 model;
|
|
model.setValue(ms.m_OpenGLMatrix);
|
|
const MT_Matrix4x4& view = rasty->GetViewMatrix();
|
|
const MT_Matrix4x4& viewinv = rasty->GetViewInvMatrix();
|
|
|
|
// note: getValue gives back column major as needed by OpenGL
|
|
model.getValue((float*)obmat);
|
|
view.getValue((float*)viewmat);
|
|
viewinv.getValue((float*)viewinvmat);
|
|
|
|
if(ms.m_bObjectColor)
|
|
ms.m_RGBAcolor.getValue((float*)obcol);
|
|
else
|
|
obcol[0]= obcol[1]= obcol[2]= obcol[3]= 1.0f;
|
|
|
|
GPU_material_bind_uniforms(gpumat, obmat, viewmat, viewinvmat, obcol);
|
|
|
|
mBlendMode = GPU_material_blend_mode(gpumat, obcol);
|
|
}
|
|
|
|
int BL_BlenderShader::GetBlendMode()
|
|
{
|
|
return mBlendMode;
|
|
}
|
|
|
|
bool BL_BlenderShader::Equals(BL_BlenderShader *blshader)
|
|
{
|
|
/* to avoid unneeded state switches */
|
|
return (blshader && mMat == blshader->mMat && mLightLayer == blshader->mLightLayer);
|
|
}
|
|
|
|
// eof
|