Note from Robert Osfield, I've temporarily re-enabled the old focing of of color and depth attachment to avoid regressions on some OpenGL driver. We'll revist this once
we have a mechanism for controlling this override at runtime.
#define FORCE_COLOR_ATTACHMENT 1
#define FORCE_DEPTH_ATTACHMENT 1
even if they have had DataVariance DYNAMIC explicitly specified. Then
when an application attempts to dynamically update the geometry in the
frame loop the primitive sets and data arrays are no longer as
expected, leading to display and/or memory corruption.
Attached is a simple fix. "
Note from Robert Osfield, tweaked Colin's changes to that it use != DYNAMIC as the the test rather than == STATIC.
Therefore I have changed all the occurances of atof by asciiToFloat or
asciiToDouble.
I believe that it is safe to do so at least for all the plugins.
Included here are also asciiToFloat conversion of environment variables. One
might argue that these should be locale dependent. But IMO these should be
set and interpreted by osg independent of the current locale.
"
of these methods in src/osgViewer/Renderer.cpp to make sure that the draw thread keeps references to all in scene graph Cameras
that are being used by the drawing threads, to keep the Camera's alive even when the main thread removes these Cameras from the scene graph.
The problem is in SpatializeGroupsVisitor::divide(osg::Geode*, unsigned int) where the code creates a new Group and divides up the input Geode into one Geode per Drawable.
I fixed the problem by assigning the Geode's stateset to the new parent group.
To replicate the bug, see attached osg/dds files:
osgviewer b.osg -- model renders correctly
set OSG_OPTIMIZER="SPATIALIZE_GROUPS"
osgviewer b.osg -- textures are missing."
Also, there was also a small bug in osgDB's CMakeLists.txt that was causing an error when I tested with CMake 2.4.4.
IF(${OSG_DEFAULT_IMAGE_PLUGIN_FOR_OSX} STREQUAL "quicktime")
was changed to
IF(OSG_DEFAULT_IMAGE_PLUGIN_FOR_OSX STREQUAL "quicktime")
"
Very simple cases of state configuration are supported (all the ones I really need):
- single per pixel not attenuated non spot light source ON/OFF
- exp2 fog ON/OFF
- diffuse texture in rgb + optional specular gloss in alpha (Texture unit 0) ON/OFF
- normal map texture (Texture unit 1 and Tangent in VertexAttribArray 6) ON/OFF
- blending and alpha testing (not in shader pipeline)
To view fixed function pipeline files and paged databases simply run >osgshadergen myfile.osg"
When 2 geometries are merged, the primitive sets of the second geometry
are copied to the first geometry.
The primitive sets were copied with a std::insert into the first geometry
primitive set vector. It doesn't work when the geometry is using VBOs (because
the element buffer object of the primitive set is not updated).
The correction replaces
lhs.getPrimitiveSetList().insert( lhs.getPrimitiveSetList().end(),
rhs.getPrimitiveSetList().begin(),
rhs.getPrimitiveSetList().end() );
by
for( primItr=rhs.getPrimitiveSetList().begin();
primItr!=rhs.getPrimitiveSetList().end();
++primItr )
{
lhs.addPrimitiveSet(primItr->get());
}"
Because there already exists the option to convert slow path geometry to the fast path by computing an internal fast path alternative, I added a new optimizer option that automatically does this. To check the results I also made some changes to the statistics gathering and rendering.
Somewhat unrelated, but also part of the optimizer I disabled removal of CameraView nodes during RemoveRedundantNodes optimization.
As discussed on the ML, CameraViews were removed from the scenegraph. This solves that issue.
Summary:
-Geometry::areFastPathsUsed now also looks at internalOptimizedGeometry
-Added Optimize option to make all slow path geometry compute their internal fast path alternative
-Added fast geometry counter to the statistics
-Disabled removel of CameraViews in optimizer
"
first email:
"in the current implementation of osgUtil::RenderStage::drawInner() method, there is some wrong assumptions made. The problem is, that whenever one does use multisampling functionality, the Blit operation (which suppose to copy the content of multisampled FBO into the usual one) doesn't perform well in some cases.
I've attached a corrected version of the RenderStage. It do just add one line, which enables the multisampled FBO as a readable just before the usual FBO is set as writable. With these corrections the Blit operation performs now correct and allows using of multisampled rendering results further."
second email:
"There was a problem when blitting the multisampled FBO with internal formats. The default internal format of color buffered multisample FBO was GL_RGBA. This has converted the color values whenever the FBO content was copied. I've added couple of lines, which do just enable the multisampled FBO internal format in respect to the attached color texture's internal format. This makes it possible to blit even float valued texture, so make HDR with multisampling possible ;)"
used but never restored to the decimal notation. That made OSG print messages
like the following after some notifications:
Warning: detected OpenGL error 'invalid value' after RenderBin::draw(,)
RenderStage::drawInner(,) FBO status= 0x8cd5
[...]
Scaling image 'brick_side.JPG' from (1b4,24f) to (200,200) <--- Values in hex
because of previous error.
[...]"
* When used PDS RenderStage::runCameraSetUp sets flag that FBO has already stencil,depth buffer attached. Prevents adding next depth buffer.
* Sets correct traits for p-buffer if used PDS and something goes wrong with FBO setup or p-buffer is used directly.
* Adds warning to camera if user add depth/stencil already attached through PDS.
* Sets blitMask when use blit to resolve buffer.
There is also new example with using multisampled FBO."
osg::ColorMask* leftColorMask = _renderStageLeft->getColorMask();
if (!leftColorMask)
{
leftColorMask = new osg::ColorMask();
_renderStageLeft->setColorMask(leftColorMask);
^^^^ here it said right, I think this should be Left.
}
// ensure that right eye color planes are active.
osg::ColorMask* rightColorMask = _renderStageRight->getColorMask();
^^^^ similar here, I think this should be right
if (!rightColorMask)
{
rightColorMask = new osg::ColorMask();
_renderStageRight->setColorMask(rightColorMask);
}
and i further removed an unnecessary setColorMask."
set.
The optimization is based on the observation that matrix matrix multiplication
with a dense matrix 4x4 is 4^3 Operations whereas multiplication with a
transform, or scale matrix is only 4^2 operations. Which is a gain of a
*FACTOR*4* for these special cases.
The change implements these special cases, provides a unit test for these
implementation and converts uses of the expensiver dense matrix matrix
routine with the specialized versions.
Depending on the transform nodes in the scenegraph this change gives a
noticable improovement.
For example the osgforest code using the MatrixTransform is about 20% slower
than the same codepath using the PositionAttitudeTransform instead of the
MatrixTransform with this patch applied.
If I remember right, the sse type optimizations did *not* provide a factor 4
improovement. Also these changes are totally independent of any cpu or
instruction set architecture. So I would prefer to have this current kind of
change instead of some hand coded and cpu dependent assembly stuff. If we
need that hand tuned stuff, these can go on top of this changes which must
provide than hand optimized additional variants for the specialized versions
to give a even better result in the end.
An other change included here is a change to rotation matrix from quaterion
code. There is a sqrt call which couold be optimized away. Since we divide in
effect by sqrt(length)*sqrt(length) which is just length ...
"
