This introduces an internal glLoader library which allows
most of the implementation to be agnostic about the
implementation of the GL loading library. Specifically,
this removes references to the GLEW headers and libraries
from the rest of the source code and build system.
Split the examples/common objects into graphics API dependent object
libraries (following the pattern used for cemmon Ptex dependencies)
so that DX11 and Metal examples are no longer forced to link with GLFW.
- updated CMakeLists to account for the moved files
- updated references to argOptions.h in all examples
- added method to ArgOptions to populate vector<ShapeDesc>
- minor changes to ViewerArgUtils to use forward references
Updated the mtlViewer to make use of common/argOptions.h
like the other viewers and also implemented a "fit to frame"
hotkey. Both changes help improve manual testing.
Removed the use of the LOOP preprocessor symbol from
the glFVarViewer shaders code. The shader code is now
configured according to the types of the resulting patches
without depending on the subdivision scheme of the mesh
topology.
This allows the implementation of face-varying interpolation
to be improved and simplifies both the shader code and
application code.
Also, fixed a couple of incorrect buffer/texture bindings
that were raising GL errors.
- added checkboxes for Control Edges and Vertices as used elsewhere
- added checkbox for UV Backface Culling specific to this example
- relabeled the Boundaries pulldown to Linear Interpolation
- renamed variables in source to better reflect the affected enum
- added definition to regression/common/shape_utils.h
- removed examples/common/shapeDesc.h and all references to it
- removed local definitions of ShapeDesc from examples, regressions, etc.
- overloaded Shape::parseObj() with ShapeDesc
- updated examples to use Shape::parseObj(ShapeDesc const&);
- removed axis argument to Shape::ParseObj() and ObjAnim::Create()
- replaced -axis with -yup in examples/common/argOptions
- updated clients of ArgOptions: glViewer, glFVarViewer, dxViewer
- updated documentation for common arguments and usage of examples
Started refactoring how we parse arguments for sample viewer
applications into a common place. This will make it easier
to keep args consistent between applications.
- addition of new shapes -- mainly Loop and Bilinear
- removal of redundant shapes and those not suited to a particular example
- enforced consistent "Patch Type" coloring where provided
- enforced consistent command line usage for common options and Obj input
- enforced consistent default settings on initialization
- added errors and/or assertions where legitimate limitations remain
- improved error handling (OSD errors not being flushed before termination)
- Removed Catmark restrictions in the application code and shaders
- Added command line options to specifiy the Scheme for .obj input files
- Added support for Linear end-cap approximation
- Updated GUI labels for end-cap types, i.e. "Linear" instead of "Bilinear"
and "Regular" instead of "BSpline".
- Added new shading color to glViewer for "Patch Depth"
- Disabled screen-space and fractional tessellation by default
This updates the patch basis evaluation functions in Osd
to match recent changes to far/patchBasis.
This also exposes a common facility for dealing with PatchCoord,
PatchArray, and PatchParam. These are exposed as global functions
operating on struct data, since C++ style class methods are not
supported by all of the Osd shader and kernel execution envirionments.
Changes:
- Merged far/patchBasis.cpp to osd/patchBasisCommon{,Types,Eval}.h
- Exposed PatchCoord, PatchArray, and PatchParam to Osd kernels
- exposed OsdEvaluatePatchBasis and OsdEvaluatePatchBasisNormalized
- Updated CPU, TBB, Omp, CUDA, OpenCL, GLSL, HLSL, and Metal evaluators
- Updated glFVarViewer
- changed the main OPENSUBDIV_INCLUDE_DIR to exclude ./opensubdiv
- updated CMakeLists.txt files in non-examples to use only this path
- updated CMakeLists.txt files in examples to append ./opensubdiv to path
- updated source in regression/common to use #include <opensubdiv/...>
- updated source in examples/common to use #include <opensubdiv/...>
- deferred source in examples to be updated on a case-by-case basis
- added detection of shapes without UVs and report fatal error
- fixed command line parsing of shape file arguments and other options
- added missing UVs from shapes/catmark_fan
Read all comments and made corrections to files that aren't part of
OpenSubdiv itself but are packaged with it.
Commandline output of glPtexViewer is affected. Otherwise no functional
changes.
The methods which return arrays of FVarPatchParam have
been made plural, e.g. GetFVarPatchParams(), for consistency
with the other methods in PatchTable.
Also fixed a missing doxygen tag.
Now the channel specifier is the last parameter in a method's
parameter list with a default of 0. This is consistent with the
topological face-varying queries and also simplifies the common
case of just a single face-varying channel.
- the framebuffer class in examples/common is unstable in certain drivers.
removing offscreen rendering for now.
- move screenshot function to GLUtils.
- fix vertex attrib binding bug (not showing control mesh on osx)
- add GLControlMeshDisplay and D3D11ControlMeshDisplay into
examples/common
- delete all drawCageEdges/drawCageVertices from viewers and
use ControlMeshDisplay class
The GLFW context version hint is a minimum version, not maximum version so
requesting 4.4 and then falling back to lower versions doesn't make sense.
This change sets the minimum version to 3.2 and attempts to standardize this
across all example apps.
Also print the maximum supported GL version along with the context version
at startup.
All examples, regression tests and tutorials directly looked into
opensubdiv source directory to grab the header files. This is somewhat
convenient during development but they can mistakenly access private
header files.
With this change, when OPENSUBDIV_INCLUDE_DIR is given to cmake,
it will be used as an include search path to build examples etc.
Otherwise it follows the same behavior as before.
Also replaces include references to the files in regression dir
to be relative, and cleanups some copy-paste patterns.
Add EvalStencils and EvalPatches API for most of CPU and GPU evaluators.
with this change, Eval API in the osd layer consists of following parts:
- Evaluators (Cpu, Omp, Tbb, Cuda, CL, GLXFB, GLCompute, D3D11Compute)
implements EvalStencils and EvalPatches(*). Both supports derivatives
(not fully implemented though)
- Interop vertex buffer classes (optional, same as before)
Note that these classes are not necessary to use Evaluators.
All evaluators have EvalStencils/Patches which take device-specific
buffer objects. For example, GLXFBEvaluator can take GLuint directly
for both stencil tables and input primvars. Although using these
interop classes makes it easy to integrate osd into relatively
simple applications.
- device-dependent StencilTable and PatchTable (optional)
These are also optional, but can be used simply a substitute of
Far::StencilTable and Far::PatchTable for osd evaluators.
- PatchArray, PatchCoord, PatchParam
They are tiny structs used for GPU based patch evaluation.
(*) TODO and known issues:
- CLEvaluator and D3D11Evaluator's EvalPatches() have not been implemented.
- GPU Gregory patch evaluation has not been implemented in EvalPatches().
- CudaEvaluator::EvalPatches() is very unstable.
- All patch evaluation kernels have not been well optimized.
- Currently GLXFB kernel doesn't support derivative evaluation.
There's a technical difficulty for the multi-stream output.
