| Age | Commit message (Collapse) | Author |
|---|
|
This commit adds some easy to use way to check whether the graph
was properly optimized. The idea is based on using mock glog sync
which keeps track on all messages and expects specific log message
to appear in the logging.
This means each optimization step should report that it optimized
something (specifying what exactly was optimized) and what was used
for optimization.
There's also some ease-ish way to define shader networks which we
might want to make a bit more global and available for all tests
in the future.
For now only RGBToBW node is covered as an example, really hope
to get help from active guys in the community to finish covering
all existing optimization cases. Ready to fix any possible issues
with builder when needed tho :)
|
|
Points behind camera will be flipped on projection, but that did not
happen.
This solves second issue reported at T48848.
|
|
|
|
|
|
|
|
Subdivision options can now be found in the subsurf modifier. The modifier must
be the last in the stack or the options will be unavailable. Catmull-Clark
subdivision is still unavailable and will fallback to linear subdivision instead
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D2109
|
|
This adds support for ngons and attributes on subdivision meshes. Ngons are
needed for proper attribute interpolation as well as correct Catmull-Clark
subdivision. Several changes are made to achieve this:
- new primitive `SubdFace` added to `Mesh`
- 3 more textures are used to store info on patches from subd meshes
- Blender export uses loop interface instead of tessface for subd meshes
- `Attribute` class is updated with a simplified way to pass primitive counts
around and to support ngons.
- extra points for ngons are generated for O(1) attribute interpolation
- curves are temporally disabled on subd meshes to avoid various bugs with
implementation
- old unneeded code is removed from `subd/`
- various fixes and improvements
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D2108
|
|
|
|
|
|
The issue was caused by some NaN appearing in calculations.
Visible with scifi_armor_concept.blend from the cloud.
|
|
This is a bit weak, but better than tagging whole mesh manager for update.
Maybe we'll solve such dual-look up in the future.
This commit finally solves T48963: Noise when changing Diffuse node to Emission node
|
|
Shader's has_surface_emission will only be known for sure after compilation.
|
|
Once again, should be no functional changes.
|
|
Should be no functional changes.
|
|
|
|
|
|
Seems we are using `void *` much much more often in Cycles.
|
|
- Nodification of shader nodes changed default strength from 10 to 1.
If that was intentional, IMO should happen as a separate commit.
- Fixed indentation.
|
|
|
|
While they prevent legit write past the array boundary error
those fixes introduced regression in behavior when having exact
max_hits transparent intersections and nothing else.
Previous code would have considered such case a totally opaque,
but it's not correct.
Fixes T48941: Some materials don't get transparent shadows anymore
|
|
code
In the triangle intersection refinement code, rays that are parallel to the triangle caused a divide by zero.
These rays might initially hit the triangle due to the watertight intersection test, but are very rare - therefore, just skipping the refinement for them works fine.
Also, a few remaining issues in the MultiGGX code are fixed that were caused by rays parallel to the surface (which happened more often there due to smooth shading).
|
|
The Multiscattering GGX closures didn't set the omega_i differentials, which could cause undefined behaviour.
|
|
|
|
|
|
|
|
Found by PVS-Studio T48917
|
|
|
|
|
|
It became rather annoying to have those functions duplicated.
Surely, it's not really nice it's actually a bad level call,
but similar thing is happening in OCIO and Cycles.
IMO, it's better than having functions re-implemented, and
have this solved with new OpenGL pipeline.
|
|
|
|
|
|
|
|
This allows us to store more than one UV layer in the UVs array.
|
|
|
|
|
|
Some of these check that dimensions match before running code that
assumes they do match.
For imb_stereo3d_write_anaglyph I *assume* this change reflects the
intended behavior. Before it was always grabbing alpha from buffer 0.
Found with PVS-Studio T48917
|
|
|
|
Previous patch was wrong apparently... :|
|
|
Patch from @fjuhec.
|
|
Mainly the changes are related on establishing API to feed UV islands
to OpenSubdiv, so it will know all the connectivity information and
will be able to do proper interpolation.
Island calculation is currently rather slow, not sure how to make it
fast and not use lots of allocations.
Shouldn't be THAT bad, since this code is only runs once when creating
OSD mesh, and it's probably still faster than our orientation code.
|
|
|
|
refactoring.
|
|
doesn't work then.
|
|
A bit work in progress, currently the following limitations:
- Texture shading only, Material shading will come later
- No UVs subdivision yet
- Always uses active UV and currently changing active UV will
not properly update the viewport.
Well, need to start somewhere :)
|
|
This is a way for us to store extra data, such as UVs which we can
collect now on topology refiner stage.
|
|
|
|
|
|
This probably makes code somewhat slower, but we can't easily know
whether we can use a shortcut and only use directional lighting from
the scene.
Need some better integration between GPU and OpenSubdiv for that.
|
|
|
|
Reviewed By: brecht, sergey
Differential Revision: https://developer.blender.org/D2089
|
