2011-02-23 10:52:22 +00:00
|
|
|
/*
|
2002-10-12 11:37:38 +00:00
|
|
|
* $Id$
|
2008-04-16 22:40:48 +00:00
|
|
|
* ***** BEGIN GPL LICENSE BLOCK *****
|
2002-10-12 11:37:38 +00:00
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU General Public License
|
|
|
|
* as published by the Free Software Foundation; either version 2
|
2008-04-16 22:40:48 +00:00
|
|
|
* of the License, or (at your option) any later version.
|
2002-10-12 11:37:38 +00:00
|
|
|
*
|
|
|
|
* This program is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License
|
|
|
|
* along with this program; if not, write to the Free Software Foundation,
|
2010-02-12 13:34:04 +00:00
|
|
|
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
|
2002-10-12 11:37:38 +00:00
|
|
|
*
|
|
|
|
* The Original Code is Copyright (C) 2001-2002 by NaN Holding BV.
|
|
|
|
* All rights reserved.
|
|
|
|
*
|
|
|
|
* The Original Code is: all of this file.
|
|
|
|
*
|
|
|
|
* Contributor(s): none yet.
|
|
|
|
*
|
2008-04-16 22:40:48 +00:00
|
|
|
* ***** END GPL LICENSE BLOCK *****
|
2002-10-12 11:37:38 +00:00
|
|
|
*/
|
|
|
|
|
2011-02-25 13:38:24 +00:00
|
|
|
/** \file gameengine/Rasterizer/RAS_BucketManager.cpp
|
|
|
|
* \ingroup bgerast
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
2010-09-15 16:13:32 +00:00
|
|
|
#if defined(WIN32) && !defined(FREE_WINDOWS)
|
2002-10-12 11:37:38 +00:00
|
|
|
// don't show these anoying STL warnings
|
|
|
|
#pragma warning (disable:4786)
|
|
|
|
#endif
|
|
|
|
|
2011-05-06 20:18:42 +00:00
|
|
|
#include "CTR_Map.h"
|
2002-10-12 11:37:38 +00:00
|
|
|
#include "RAS_MaterialBucket.h"
|
|
|
|
#include "STR_HashedString.h"
|
|
|
|
#include "RAS_MeshObject.h"
|
|
|
|
#include "RAS_IRasterizer.h"
|
|
|
|
#include "RAS_IRenderTools.h"
|
|
|
|
|
|
|
|
#include "RAS_BucketManager.h"
|
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
#include <algorithm>
|
2004-05-21 08:55:12 +00:00
|
|
|
#include <set>
|
2004-03-22 22:02:18 +00:00
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
/* sorting */
|
|
|
|
|
|
|
|
struct RAS_BucketManager::sortedmeshslot
|
2002-10-12 11:37:38 +00:00
|
|
|
{
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
public:
|
|
|
|
MT_Scalar m_z; /* depth */
|
|
|
|
RAS_MeshSlot *m_ms; /* mesh slot */
|
|
|
|
RAS_MaterialBucket *m_bucket; /* buck mesh slot came from */
|
2002-10-12 11:37:38 +00:00
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
sortedmeshslot() {}
|
2002-10-12 11:37:38 +00:00
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
void set(RAS_MeshSlot *ms, RAS_MaterialBucket *bucket, const MT_Vector3& pnorm)
|
|
|
|
{
|
|
|
|
// would be good to use the actual bounding box center instead
|
|
|
|
MT_Point3 pos(ms->m_OpenGLMatrix[12], ms->m_OpenGLMatrix[13], ms->m_OpenGLMatrix[14]);
|
2002-10-12 11:37:38 +00:00
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
m_z = MT_dot(pnorm, pos);
|
|
|
|
m_ms = ms;
|
|
|
|
m_bucket = bucket;
|
|
|
|
}
|
2004-07-17 05:23:17 +00:00
|
|
|
};
|
2002-10-12 11:37:38 +00:00
|
|
|
|
2004-07-17 05:23:17 +00:00
|
|
|
struct RAS_BucketManager::backtofront
|
|
|
|
{
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
bool operator()(const sortedmeshslot &a, const sortedmeshslot &b)
|
2004-07-17 05:23:17 +00:00
|
|
|
{
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
return (a.m_z < b.m_z) || (a.m_z == b.m_z && a.m_ms < b.m_ms);
|
2004-07-17 05:23:17 +00:00
|
|
|
}
|
|
|
|
};
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
|
|
|
|
struct RAS_BucketManager::fronttoback
|
|
|
|
{
|
|
|
|
bool operator()(const sortedmeshslot &a, const sortedmeshslot &b)
|
|
|
|
{
|
|
|
|
return (a.m_z > b.m_z) || (a.m_z == b.m_z && a.m_ms > b.m_ms);
|
|
|
|
}
|
|
|
|
};
|
|
|
|
|
|
|
|
/* bucket manager */
|
|
|
|
|
|
|
|
RAS_BucketManager::RAS_BucketManager()
|
|
|
|
{
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
RAS_BucketManager::~RAS_BucketManager()
|
|
|
|
{
|
|
|
|
BucketList::iterator it;
|
|
|
|
|
|
|
|
for (it = m_SolidBuckets.begin(); it != m_SolidBuckets.end(); it++)
|
|
|
|
delete (*it);
|
|
|
|
|
|
|
|
for (it = m_AlphaBuckets.begin(); it != m_AlphaBuckets.end(); it++)
|
|
|
|
delete(*it);
|
2004-07-17 05:23:17 +00:00
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
m_SolidBuckets.clear();
|
|
|
|
m_AlphaBuckets.clear();
|
|
|
|
}
|
|
|
|
|
|
|
|
void RAS_BucketManager::OrderBuckets(const MT_Transform& cameratrans, BucketList& buckets, vector<sortedmeshslot>& slots, bool alpha)
|
|
|
|
{
|
|
|
|
BucketList::iterator bit;
|
|
|
|
list<RAS_MeshSlot>::iterator mit;
|
|
|
|
size_t size = 0, i = 0;
|
|
|
|
|
|
|
|
/* Camera's near plane equation: pnorm.dot(point) + pval,
|
|
|
|
* but we leave out pval since it's constant anyway */
|
|
|
|
const MT_Vector3 pnorm(cameratrans.getBasis()[2]);
|
|
|
|
|
|
|
|
for (bit = buckets.begin(); bit != buckets.end(); ++bit)
|
BGE performance, 3rd round: culling and rasterizer.
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...
2009-05-07 09:13:01 +00:00
|
|
|
{
|
|
|
|
SG_DList::iterator<RAS_MeshSlot> mit((*bit)->GetActiveMeshSlots());
|
|
|
|
for(mit.begin(); !mit.end(); ++mit)
|
|
|
|
size++;
|
|
|
|
}
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
|
|
|
|
slots.resize(size);
|
|
|
|
|
|
|
|
for (bit = buckets.begin(); bit != buckets.end(); ++bit)
|
BGE performance, 3rd round: culling and rasterizer.
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...
2009-05-07 09:13:01 +00:00
|
|
|
{
|
|
|
|
RAS_MaterialBucket* bucket = *bit;
|
|
|
|
RAS_MeshSlot* ms;
|
|
|
|
// remove the mesh slot form the list, it culls them automatically for next frame
|
2009-08-04 05:14:10 +00:00
|
|
|
while((ms = bucket->GetNextActiveMeshSlot())) {
|
BGE performance, 3rd round: culling and rasterizer.
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...
2009-05-07 09:13:01 +00:00
|
|
|
slots[i++].set(ms, bucket, pnorm);
|
|
|
|
}
|
|
|
|
}
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
|
|
|
|
if(alpha)
|
|
|
|
sort(slots.begin(), slots.end(), backtofront());
|
|
|
|
else
|
|
|
|
sort(slots.begin(), slots.end(), fronttoback());
|
|
|
|
}
|
|
|
|
|
2004-05-21 08:55:12 +00:00
|
|
|
void RAS_BucketManager::RenderAlphaBuckets(
|
|
|
|
const MT_Transform& cameratrans, RAS_IRasterizer* rasty, RAS_IRenderTools* rendertools)
|
|
|
|
{
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
vector<sortedmeshslot> slots;
|
|
|
|
vector<sortedmeshslot>::iterator sit;
|
|
|
|
|
|
|
|
// Having depth masks disabled/enabled gives different artifacts in
|
|
|
|
// case no sorting is done or is done inexact. For compatibility, we
|
|
|
|
// disable it.
|
|
|
|
rasty->SetDepthMask(RAS_IRasterizer::KX_DEPTHMASK_DISABLED);
|
|
|
|
|
|
|
|
OrderBuckets(cameratrans, m_AlphaBuckets, slots, true);
|
2004-05-21 08:55:12 +00:00
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
for(sit=slots.begin(); sit!=slots.end(); ++sit) {
|
|
|
|
rendertools->SetClientObject(rasty, sit->m_ms->m_clientObj);
|
|
|
|
|
2008-09-14 00:32:18 +00:00
|
|
|
while(sit->m_bucket->ActivateMaterial(cameratrans, rasty, rendertools))
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
sit->m_bucket->RenderMeshSlot(cameratrans, rasty, rendertools, *(sit->m_ms));
|
2009-04-07 22:14:06 +00:00
|
|
|
|
|
|
|
// make this mesh slot culled automatically for next frame
|
|
|
|
// it will be culled out by frustrum culling
|
|
|
|
sit->m_ms->SetCulled(true);
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
rasty->SetDepthMask(RAS_IRasterizer::KX_DEPTHMASK_ENABLED);
|
|
|
|
}
|
|
|
|
|
|
|
|
void RAS_BucketManager::RenderSolidBuckets(
|
|
|
|
const MT_Transform& cameratrans, RAS_IRasterizer* rasty, RAS_IRenderTools* rendertools)
|
|
|
|
{
|
|
|
|
BucketList::iterator bit;
|
|
|
|
|
|
|
|
rasty->SetDepthMask(RAS_IRasterizer::KX_DEPTHMASK_ENABLED);
|
|
|
|
|
|
|
|
for (bit = m_SolidBuckets.begin(); bit != m_SolidBuckets.end(); ++bit) {
|
BGE performance, 3rd round: culling and rasterizer.
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...
2009-05-07 09:13:01 +00:00
|
|
|
#if 1
|
|
|
|
RAS_MaterialBucket* bucket = *bit;
|
|
|
|
RAS_MeshSlot* ms;
|
|
|
|
// remove the mesh slot form the list, it culls them automatically for next frame
|
2009-08-04 05:14:10 +00:00
|
|
|
while((ms = bucket->GetNextActiveMeshSlot()))
|
BGE performance, 3rd round: culling and rasterizer.
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...
2009-05-07 09:13:01 +00:00
|
|
|
{
|
|
|
|
rendertools->SetClientObject(rasty, ms->m_clientObj);
|
|
|
|
while (bucket->ActivateMaterial(cameratrans, rasty, rendertools))
|
|
|
|
bucket->RenderMeshSlot(cameratrans, rasty, rendertools, *ms);
|
|
|
|
|
|
|
|
// make this mesh slot culled automatically for next frame
|
|
|
|
// it will be culled out by frustrum culling
|
|
|
|
ms->SetCulled(true);
|
|
|
|
}
|
|
|
|
#else
|
|
|
|
list<RAS_MeshSlot>::iterator mit;
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
for (mit = (*bit)->msBegin(); mit != (*bit)->msEnd(); ++mit) {
|
|
|
|
if (mit->IsCulled())
|
|
|
|
continue;
|
|
|
|
|
|
|
|
rendertools->SetClientObject(rasty, mit->m_clientObj);
|
|
|
|
|
2008-09-14 00:32:18 +00:00
|
|
|
while ((*bit)->ActivateMaterial(cameratrans, rasty, rendertools))
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
(*bit)->RenderMeshSlot(cameratrans, rasty, rendertools, *mit);
|
2009-04-07 22:14:06 +00:00
|
|
|
|
|
|
|
// make this mesh slot culled automatically for next frame
|
|
|
|
// it will be culled out by frustrum culling
|
|
|
|
mit->SetCulled(true);
|
2004-05-21 08:55:12 +00:00
|
|
|
}
|
BGE performance, 3rd round: culling and rasterizer.
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...
2009-05-07 09:13:01 +00:00
|
|
|
#endif
|
2004-05-21 08:55:12 +00:00
|
|
|
}
|
|
|
|
|
2008-09-14 00:32:18 +00:00
|
|
|
/* this code draws meshes order front-to-back instead to reduce overdraw.
|
|
|
|
* it turned out slower due to much material state switching, a more clever
|
|
|
|
* algorithm might do better. */
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
#if 0
|
|
|
|
vector<sortedmeshslot> slots;
|
|
|
|
vector<sortedmeshslot>::iterator sit;
|
2004-05-21 08:55:12 +00:00
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
OrderBuckets(cameratrans, m_SolidBuckets, slots, false);
|
|
|
|
|
|
|
|
for(sit=slots.begin(); sit!=slots.end(); ++sit) {
|
|
|
|
rendertools->SetClientObject(rasty, sit->m_ms->m_clientObj);
|
|
|
|
|
|
|
|
while(sit->m_bucket->ActivateMaterial(cameratrans, rasty, rendertools))
|
|
|
|
sit->m_bucket->RenderMeshSlot(cameratrans, rasty, rendertools, *(sit->m_ms));
|
2004-05-21 08:55:12 +00:00
|
|
|
}
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
#endif
|
2004-05-21 08:55:12 +00:00
|
|
|
}
|
|
|
|
|
2002-10-12 11:37:38 +00:00
|
|
|
void RAS_BucketManager::Renderbuckets(
|
|
|
|
const MT_Transform& cameratrans, RAS_IRasterizer* rasty, RAS_IRenderTools* rendertools)
|
|
|
|
{
|
2008-09-14 00:32:18 +00:00
|
|
|
/* beginning each frame, clear (texture/material) caching information */
|
2002-10-12 11:37:38 +00:00
|
|
|
rasty->ClearCachingInfo();
|
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
RenderSolidBuckets(cameratrans, rasty, rendertools);
|
2004-05-21 08:55:12 +00:00
|
|
|
RenderAlphaBuckets(cameratrans, rasty, rendertools);
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
|
|
|
|
rendertools->SetClientObject(rasty, NULL);
|
2002-10-12 11:37:38 +00:00
|
|
|
}
|
|
|
|
|
2008-07-29 15:48:31 +00:00
|
|
|
RAS_MaterialBucket* RAS_BucketManager::FindBucket(RAS_IPolyMaterial * material, bool &bucketCreated)
|
2002-10-12 11:37:38 +00:00
|
|
|
{
|
2004-05-21 08:55:12 +00:00
|
|
|
BucketList::iterator it;
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
|
|
|
|
bucketCreated = false;
|
|
|
|
|
|
|
|
for (it = m_SolidBuckets.begin(); it != m_SolidBuckets.end(); it++)
|
2004-03-22 22:02:18 +00:00
|
|
|
if (*(*it)->GetPolyMaterial() == *material)
|
|
|
|
return *it;
|
|
|
|
|
|
|
|
for (it = m_AlphaBuckets.begin(); it != m_AlphaBuckets.end(); it++)
|
|
|
|
if (*(*it)->GetPolyMaterial() == *material)
|
|
|
|
return *it;
|
|
|
|
|
|
|
|
RAS_MaterialBucket *bucket = new RAS_MaterialBucket(material);
|
2008-03-23 23:12:40 +00:00
|
|
|
bucketCreated = true;
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
|
2008-07-29 15:48:31 +00:00
|
|
|
if (bucket->IsAlpha())
|
2004-03-22 22:02:18 +00:00
|
|
|
m_AlphaBuckets.push_back(bucket);
|
|
|
|
else
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
m_SolidBuckets.push_back(bucket);
|
2004-03-22 22:02:18 +00:00
|
|
|
|
2002-10-12 11:37:38 +00:00
|
|
|
return bucket;
|
|
|
|
}
|
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
void RAS_BucketManager::OptimizeBuckets(MT_Scalar distance)
|
2002-10-12 11:37:38 +00:00
|
|
|
{
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
BucketList::iterator bit;
|
2002-10-12 11:37:38 +00:00
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
distance = 10.0;
|
|
|
|
|
|
|
|
for (bit = m_SolidBuckets.begin(); bit != m_SolidBuckets.end(); ++bit)
|
|
|
|
(*bit)->Optimize(distance);
|
|
|
|
for (bit = m_AlphaBuckets.begin(); bit != m_AlphaBuckets.end(); ++bit)
|
|
|
|
(*bit)->Optimize(distance);
|
2002-10-12 11:37:38 +00:00
|
|
|
}
|
2008-07-29 15:48:31 +00:00
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
void RAS_BucketManager::ReleaseDisplayLists(RAS_IPolyMaterial *mat)
|
2008-07-29 15:48:31 +00:00
|
|
|
{
|
|
|
|
BucketList::iterator bit;
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
list<RAS_MeshSlot>::iterator mit;
|
2008-07-29 15:48:31 +00:00
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
for (bit = m_SolidBuckets.begin(); bit != m_SolidBuckets.end(); ++bit) {
|
|
|
|
if (mat == NULL || (mat == (*bit)->GetPolyMaterial())) {
|
|
|
|
for (mit = (*bit)->msBegin(); mit != (*bit)->msEnd(); ++mit) {
|
|
|
|
if(mit->m_DisplayList) {
|
|
|
|
mit->m_DisplayList->Release();
|
|
|
|
mit->m_DisplayList = NULL;
|
|
|
|
}
|
2008-07-29 15:48:31 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
for (bit = m_AlphaBuckets.begin(); bit != m_AlphaBuckets.end(); ++bit) {
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 20:51:28 +00:00
|
|
|
if (mat == NULL || (mat == (*bit)->GetPolyMaterial())) {
|
|
|
|
for (mit = (*bit)->msBegin(); mit != (*bit)->msEnd(); ++mit) {
|
|
|
|
if(mit->m_DisplayList) {
|
|
|
|
mit->m_DisplayList->Release();
|
|
|
|
mit->m_DisplayList = NULL;
|
|
|
|
}
|
2008-07-29 15:48:31 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
BGE performance, 3rd round: culling and rasterizer.
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...
2009-05-07 09:13:01 +00:00
|
|
|
void RAS_BucketManager::ReleaseMaterials(RAS_IPolyMaterial * mat)
|
|
|
|
{
|
|
|
|
BucketList::iterator bit;
|
|
|
|
list<RAS_MeshSlot>::iterator mit;
|
|
|
|
|
|
|
|
for (bit = m_SolidBuckets.begin(); bit != m_SolidBuckets.end(); ++bit) {
|
|
|
|
if (mat == NULL || (mat == (*bit)->GetPolyMaterial())) {
|
|
|
|
(*bit)->GetPolyMaterial()->ReleaseMaterial();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
for (bit = m_AlphaBuckets.begin(); bit != m_AlphaBuckets.end(); ++bit) {
|
|
|
|
if (mat == NULL || (mat == (*bit)->GetPolyMaterial())) {
|
|
|
|
(*bit)->GetPolyMaterial()->ReleaseMaterial();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2009-11-15 23:58:56 +00:00
|
|
|
/* frees the bucket, only used when freeing scenes */
|
|
|
|
void RAS_BucketManager::RemoveMaterial(RAS_IPolyMaterial * mat)
|
|
|
|
{
|
|
|
|
BucketList::iterator bit, bitp;
|
|
|
|
list<RAS_MeshSlot>::iterator mit;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
|
|
|
|
for(i=0; i<m_SolidBuckets.size(); i++) {
|
|
|
|
RAS_MaterialBucket *bucket = m_SolidBuckets[i];
|
|
|
|
if (mat == bucket->GetPolyMaterial()) {
|
|
|
|
m_SolidBuckets.erase(m_SolidBuckets.begin()+i);
|
|
|
|
delete bucket;
|
|
|
|
i--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
for(int i=0; i<m_AlphaBuckets.size(); i++) {
|
|
|
|
RAS_MaterialBucket *bucket = m_AlphaBuckets[i];
|
|
|
|
if (mat == bucket->GetPolyMaterial()) {
|
|
|
|
m_AlphaBuckets.erase(m_AlphaBuckets.begin()+i);
|
|
|
|
delete bucket;
|
|
|
|
i--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
//#include <stdio.h>
|
|
|
|
|
2010-06-16 19:07:20 +00:00
|
|
|
void RAS_BucketManager::MergeBucketManager(RAS_BucketManager *other, SCA_IScene *scene)
|
2009-11-15 23:58:56 +00:00
|
|
|
{
|
|
|
|
/* concatinate lists */
|
|
|
|
// printf("BEFORE %d %d\n", GetSolidBuckets().size(), GetAlphaBuckets().size());
|
2010-06-16 19:07:20 +00:00
|
|
|
BucketList::iterator it;
|
|
|
|
|
|
|
|
for (it = other->GetSolidBuckets().begin(); it != other->GetSolidBuckets().end(); ++it)
|
|
|
|
(*it)->GetPolyMaterial()->Replace_IScene(scene);
|
|
|
|
|
2009-11-15 23:58:56 +00:00
|
|
|
GetSolidBuckets().insert( GetSolidBuckets().end(), other->GetSolidBuckets().begin(), other->GetSolidBuckets().end() );
|
|
|
|
other->GetSolidBuckets().clear();
|
|
|
|
|
2010-06-16 19:07:20 +00:00
|
|
|
for (it = other->GetAlphaBuckets().begin(); it != other->GetAlphaBuckets().end(); ++it)
|
|
|
|
(*it)->GetPolyMaterial()->Replace_IScene(scene);
|
|
|
|
|
2009-11-15 23:58:56 +00:00
|
|
|
GetAlphaBuckets().insert( GetAlphaBuckets().end(), other->GetAlphaBuckets().begin(), other->GetAlphaBuckets().end() );
|
|
|
|
other->GetAlphaBuckets().clear();
|
|
|
|
//printf("AFTER %d %d\n", GetSolidBuckets().size(), GetAlphaBuckets().size());
|
|
|
|
}
|
|
|
|
|