Commit Graph

76 Commits

Author SHA1 Message Date
manuelk
000cb400ca Revisit singular vertex fix for FarSubdivisionTables
The previous fix pointed far indexing tables to the origin vertex
of duped singular verts.

This fix goes one step further and actually shifts all vertex indexing
to start at the end of the coarse mesh vertices, using the space for
data that was previously occupied by duplicated singular verts.

The consequence is that client code no longer needs to duplicate vertex
data in vertex buffers (huzzah !).

- fix FarSubdivisionTablesFactory to shift factory vertex table offsets using Hbr's
  singular verts map
- fix schema table factories (Catmark, Loop...) to correctly use these offsets
- remove vertex data duplication code from osdPolySmooth example
- remove some (unrelated) cruft from glViewer example
- shape_utils unfortunately still needs to dubplicate the singular verts to
  allow the coarse edge drawing in our example viewers to work correctly
  (although it could be fixed to avoid data duplication too...)

fixes #241
2013-11-22 11:27:01 -08:00
manuelk
575d931f2d Fixing build break from hbr_regression 2013-11-21 16:27:13 -08:00
manuelk
c3cb17fa99 Modify Far remapping of singular vertices to point to their source vertex.
- Add a vector of index pairs to HbrMesh to track the index of a split
  vertex and its origin vertex
- Correct the Far remap tables in FarSubdivisionTablesFactory to point split
  vertices to their origin instead of themselves
- Fix regression/common/shape_utils.h to use the new HbrMesh::GetSplitVertices()
  method.
- Fix the osdPolySmooth example to use the new HbrMesh::GetSplitVertices()
  method.
- Add a paragraph to the documentation

fixes #241
2013-11-21 16:05:31 -08:00
manuelk
2dc8520938 Fix Chaikin rule
The Chaikin crease interpolation mode seems to be broken:
	   - Catmark / Loop / Bilinear are passing the wrong halfedge vertex to the
	     SubdivideCreaseWeight function which results in sub-edge crease weights
	     being swapped
	   - the loop that iterates over adjacent edges needs to check against both
	     the original edge and its opposite, otherwise it may be incorrectly
	     accumulated into summation of these adjacent edges (with a 0.25 weight)

	   The proposed fix:
	   - Swaps the Dest/Org vertex passed to the SubdivideCreaseWeight (and
	     we probably want Julian to confirm that this the correct fix)
	   - Checks against both the original edge and its opposite in the iteration
	     over adjacent edges
	   - Replaces the std::vector based query with an HbrHalfedgeOperator for
	     better performance (hopefully)

	   The similar fix to OpenSubdiv been reviewed by Tony DeRose.

    Also in the fix:
        - fix "obj" tag parsing of the smooth triangle tag that was incorrectly
          associated with the crease method (and reporting the wrong errors)
        - add regression shapes for both Loop & Catmark schemes to hbr_regression
        - add same shapes to the glViewer
        - improve hbr_regression output to be more readable
        - add command-line argument parsing to hbr_regression
        - add functionality to dump an obj file when regression fails for comparison

fixes #235
2013-11-07 17:06:55 -08:00
manuelk
83d9315e64 Fix Chaikin tag parsing
Also:
- add 2 shape examples with Chaikin rule tag
- add shapes to the glViewer
- add a stub in the documentation

Note: the Chaikin rule currently applied by Hbr appears to be somewhat off...

fixes #236
2013-11-01 11:38:20 -07:00
manuelk
9e9fd5c021 Fix Far to handle disconnected vertices in an Hbr mesh.
Also:
    - change our shape parser to skip non-manifold faces (print a warning)
    - add a 'fan' non-manifold test shape to the glViewer

fixes #233
2013-10-28 10:40:24 -07:00
manuelk
23a518b42f Add face-varying boundary interpolation parsing to shape_utils
Also change face-varying data declarations: 2 channels of 1 float instead of 1 channel of 2 floats for uv's
2013-10-10 10:37:39 -07:00
manuelk
3ae50d1c50 Amending Apache license language & file headers.
New text:

     Copyright 2013 Pixar

     Licensed under the Apache License, Version 2.0 (the "Apache License")
     with the following modification; you may not use this file except in
     compliance with the Apache License and the following modification to it:
     Section 6. Trademarks. is deleted and replaced with:

     6. Trademarks. This License does not grant permission to use the trade
        names, trademarks, service marks, or product names of the Licensor
        and its affiliates, except as required to comply with Section 4(c) of
        the License and to reproduce the content of the NOTICE file.

     You may obtain a copy of the Apache License at

         http://www.apache.org/licenses/LICENSE-2.0

     Unless required by applicable law or agreed to in writing, software
     distributed under the Apache License with the above modification is
     distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
     KIND, either express or implied. See the Apache License for the specific
     language governing permissions and limitations under the Apache License.
2013-09-26 12:04:57 -07:00
manuelk
924d994fdd Fix CMake executable macro
- Added a macro for non-cuda / non-glfw executables to add proper dependencies for ICC
- Switched various examples & regressions
2013-09-16 18:16:28 -07:00
Ian Stephenson
ff86ee087c Fix bhr baseline regression, so reference files are real OBJ's 2013-09-16 15:44:56 -07:00
manuelk
563ac5b55f Fix osd_regression build
osd_regression needs to link against all the appropriate OSD dependencies, including possibly CUDA, TBB, ICC...
2013-09-11 18:25:25 -07:00
takahito-tejima
dd91b2de8c Fix some portability issues on OSX. (GLEW will be used if exists) 2013-07-18 19:49:00 -07:00
Takahito Tejima
51a45b598d Updating EULA 2013-07-18 14:19:50 -07:00
manuelk
500defee4c - adding new build flag options (NO_LIB, NO_CUDA...)
- move python build section into the python directory (cleaning up)

- fix some broken dependencies

- remove the public_headers targets if doxygen was not found

TODO :
- fix MSVC targets for public headers (wbn if MSVC didn't require the pro version
  in order to support solution folders)
- fix osd_regression to not build if -DNO_LIB is present (ie. fix the broken dependency)
2013-07-11 18:55:17 -07:00
manuelk
7b36b4feed fix incorrect patch generation for patches with 2 non-consecutive boundary edges
add catmark_flap2 shape & baseline to regression suite

fixes #188
2013-07-08 18:13:44 -07:00
Manuel Kraemer
523ccee896 typo fix 2013-07-07 18:18:06 -07:00
Takahito Tejima
a9f028cdc0 Add face varying data to regression shapes. 2013-07-02 10:50:27 -07:00
David G Yu
7d16bffccb Continued cleanup of OpenGL dependencies 2013-06-26 00:52:57 -07:00
manuelk
30ec963112 fix example code to work with GLFW 3.0
- fix FindGLFW.cmake to match the new locations

- fix all viewer & examples to build with 3.0, 3.0.1 and 3.0.2

Note : ptxViewer and paintTest do not work with 3.0.1 and 3.0.2 (yet)
2013-06-20 11:53:33 -07:00
manuelk
cf13a8d755 Add check in shape_utils that detects dicsonnected vertices in a shape mesh and exits with an error message.
A more permanent fix will be to handle fully non-manifold topology at the Hbr level, but that is beyond the
scope of fixing this problem.

fixes #175
2013-06-13 16:22:59 -07:00
manuelk
419b6b9716 - add a message to limitEval when trying to display face-varying UVs on shapes
that do not have that type of data

- fix some regression shapes that had "empty" uv values
  (including some minor topo surgery that luckily appears to not break
  regression baselines)
2013-06-07 17:53:23 -07:00
manuelk
cf1b2f1334 - fix a face-varying bug in shape_utils : now face-varying UVs are indexed correctly
coming out of our shape format

- set blue component of point-cloud to 0 in face-varying mode (save a few micro-secs)
2013-06-07 14:47:41 -07:00
manuelk
307b353b13 shape_utils : make sure we don't create a mesh with face-varying data declared
if the shape does not actually contain usable face-varying data.
2013-06-07 10:40:11 -07:00
manuelk
c18cf5bff2 first pass at face-varying interpolation for CpuEvalLimit - still
a couple of kinks to be worked out...
2013-06-06 18:07:46 -07:00
manuelk
ad3bacbbbb remove topology data from FarMesh and refactor uniform / adaptive
code paths using FarPatchTables for all serialized topological data.
2013-05-15 17:53:40 -07:00
manuelk
fbcb02d39e Fixed dyu's fix (we do need the window title in Osd Regression with GLFW 3.0) 2013-05-07 14:25:37 -07:00
David G Yu
2a12b21ab2 Removed mutex implementation.
We added this to support earlier versions
of hbr, but it is no longer needed.
2013-05-07 11:27:14 -07:00
David G Yu
93476a368d Fixed a few OS X build errors. 2013-05-06 23:59:58 -07:00
manuelk
20d61cd93c A quad face with 2 non-consecutive boundary edges causes the 4 vertices
of this quad to be tagged as boundary. These vertices will cause the
feature adaptive pass to generate sub-faces where appropriate, however
the face itself will not be identified as "non-patch", which causes the
FarPatchTables factory to mis-identify it and fall-back on an assert.

This fix flags these particular quad faces in the first adaptive pass.

The particular shape that caused the crash has also been addded to our
regression suite.

fixes #159
2013-05-02 19:37:34 -07:00
manuelk
b815aff333 First pass at fixing the installation part of the Cmake build : now CMAKE_INSTALL_PREFIX can
be used as intended to specify an installation directory, which can be located anywhere on the
file system.

Also improved the doxygen target and made the doxy build "quiet".

fixes #154
2013-04-22 19:35:41 -07:00
Takahito Tejima
d026aca006 use std::string instead of char* for regression shapes definition,
to avoid MSVC's literal length limitation on some big shapes.
2013-03-19 14:43:21 -07:00
Manuel Kraemer
8e9e832f94 When OpenCL is detected, make sure all the includes / libs are tacked on the build lines 2013-03-16 18:44:07 -07:00
Manuel Kraemer
d703ba0205 fix some unused variable warnings 2013-03-16 18:43:20 -07:00
Takahito Tejima
fbc139d39c fix compile errors on gcd kernel 2013-03-07 22:43:47 -08:00
Takahito Tejima
8efecb0fca Batching stuffs: generalized kernel batches, table/dispatcher refactoring, multiMeshFactory, drawContext, etc.
2 client APIs are changed.
- VertexBuffer::UpdateData() takes start vertex offset
- ComputeController::Refine() takes FarKernelBatchVector

Also, ComputeContext no longer holds farmesh.
Client can free farmesh after OsdComputeContext is created.
(but still need FarKernelBatchVector to apply subdivision kernels)
2013-03-07 17:50:15 -08:00
manuelk
ea1a87441f Add support for hierarchical hole edits.
- add tag parsing for h-hole in shape_utils
- re-create Renderman's test shape from the documentation (catmark_square_hedit.h)
- fix Hbr to correctly pass the hole tag from parent to child face
- fix FarSubdivisionTables to handle disconnected face-vertices without crashing

fixes #75
2013-03-01 18:27:19 -08:00
manuelk
25f79e7ff2 - change adaptive refinement in FarMeshFactory to not refine inside holes,
while retaining a necessary 1-ring on the inside of a hole edge

- add IsInsideHole() function to HbrHalfEdge

- add HasChild() function to HbrVertex and HbrHalfedge

- add a regression shape with adjacent holes and creases (tests dart, crease & boundaries)

Note : this does not address hierarchical edits inside holes or hole tags in hierarchical edits

fixes #78
2013-02-20 14:12:09 -08:00
manuelk
a52c70ab8b First pass implementation of holes :
- make sure HBR passes down the hole tag to children when subdividing faces
- minor API modification : allow to unset the hole flag on a face
- modify uniform / adaptive FarMeshFactory to be aware of the flag
- make the FarSubdivisionTableFactory assert when finding unconnected HBR vertices (as it should)

* Uniform subdivision :
    The refinement scheme only creates faces & vertices necessary
    to maintain the one-ring around the edges of a hole, so this solution
    is quite efficient.

* Adaptive subdivision :
    At the moment we are still performing full topological analysis on holes and
    only skipping patches associated to holes. This is sub-optimal in 2 ways :
        1. the topological analysis can potentially be cranking on a lot of unnecessary
           geometry
        2. even though we may not be drawing the patches, the compute stage is still
           applying kernels on all the control vertices of these patches.
    We will have to revisit feature adaptive subdivision & holes, so keep the issue active.

fixes #78
2013-02-13 14:34:33 -08:00
Andrew Wong
b25226134f fix warnings for having assignments inside conditionals
warning C4706: assignment within conditional expression
2013-02-09 23:34:58 -05:00
manuelk
ef408fc67c - remove the now obsolete glutViewer (replaced by glViewer)
- remove some glut-era callbacks in glViewer / ptexViewer
- fix OSX includes in osd_regression

fixes #117
2013-02-08 14:06:44 -08:00
manuelk
c53f973651 - fixed GL context crashes for Linux / GLFW 3.0
- refactored CLI arguments parsing
- minor code style cleanups
2013-02-06 14:07:47 -08:00
Manuel Kraemer
758a788b32 Merge pull request #110 from aras-p/osd_regression_work
osd_regression: add "CL" backend that tests OpenCL kernel
2013-02-06 08:00:09 -08:00
David G Yu
88679ce187 Fixed WIN32 build breakage.
GLFW tries to avoid including windows.h, but by hijacking
macro definitions provided by windows.h GLFW actually
further pollutes the global namespace.
2013-02-05 17:05:59 -08:00
Aras Pranckevicius
663a62cb08 osd_regression: add "CL" backend that tests OpenCL kernel with GL buffer backing 2013-02-04 22:01:59 +02:00
Aras Pranckevicius
660fd15e3b osd_regression: implement raw CPU and CPU with GL VBO backing tests. Actually count test failures; they were not counted before ;) 2013-02-03 19:25:22 +02:00
Aras Pranckevicius
e0b231424f osd_regression: towards running all test cases per backend (checkBackend function) 2013-02-03 19:07:04 +02:00
Aras Pranckevicius
1fbdf77bb4 osd_regression: towards testing multiple backends; adding cmdline arg parsing 2013-02-03 19:00:28 +02:00
Aras Pranckevicius
2d00699c8d osd_regression does not need GLEW on OSX 2013-02-03 18:34:41 +02:00
manuelk
055ab17357 migrate osd_regression from GLUT to GLFW (this regression still needs work)
fixes #98
2013-01-31 18:19:56 -08:00
manuelk
4b199f7587 Make sure that shape strings are post-fixed with a null-char when read from file.
fixes #85
2012-12-11 11:40:57 -08:00