For a real-world distortion the payload is quite uniformly
distributed across scanlines. Surely, in the corners more
iterations of minimizer is needed, but that happens in threads
without scheduling overhead.
Need to communicate available number of threads to the camera
intrinsics implementation, otherwise default value of 1 is used.
Must have been single-threaded for a very long time.
All DirectX management happens on Ghost level now, higher level code can
just assume everything is OpenGL (except of the upside-down drawing that
still needs to be done for DirectX). This is similar to how the
metal-layer is hidden outside of Ghost.
The Ghost-XR graphics binding for DirectX is responsible for managing
the DirectX compatibility now.
This patch adds an AVX implementation of Perlin noise in Cycles.
An avxi type was also added as a utility based on the respective
type in Intel Embree.
Only 3D and 4D noise were implemented, there is no benefit for
utilizing AVX in 1D and 2D noise. The SSE trilinear interpolation
function was used in the AVX implementation because there is no
benefit from using AVX in interpolating the last three dimensions.
Differential Revision: https://developer.blender.org/D6680
Removing the GHOST_kKeyUnknown check from processKeyEvent() produces
epeated unknown key events whenever a modifier key is held down, due
to the way ghost uses GHOST_kKeyUnknown as a filter value for modifier
key events.
Differential Revision: https://developer.blender.org/D7257
This works similarly to the CYCLES_OPENCL_TEST
environment variable to allow testing on unsupported
hardware.
Note: like the OPENCL test override, this is
for *testing* only and bug reports on unsupported
hardware will *not* be accepted at this point in
time.
Events for keys specific to certain keyboard layouts unknown to Blender
were ignored. Now pass them along as unknown key events for which we
can still handle text input, like we already do for Linux and macOS.
Differential Revision: https://developer.blender.org/D7229
Random Walk subsurface scattering did look different with OptiX because transmittance is
calculated based on the hit distance, but the OptiX implementation of `scene_intersect_local`
would return the distance in world space, while the Cycles BVH version returns it in object
space. This fixes the problem by simply skipping the object->world transforms in all the
places using the result of `scene_intersect_local` with OptiX.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D7232
There is no need to have another font embedded in the Blender executable, we
can assume the bundled font exists. In the future we may provide a fallback
if the font specified by the user in the preferences is missing a character,
but that can use our bundled international font.
Differential Revision: https://developer.blender.org/D6854
The numbers here can probably be tweaked to be better, but it's hard to
predict and this should at least avoid excessive memory swapping.
Fixes T57064.
Tested with AMD Radeon Pro WX 9100, where it brings performance back to 2.80
level, and combined with recent changes is about 2-15% faster than 2.80 in
our benchmark scenes.
This somehow appears to specifically address the issue where adding more shader
nodes leads to slower runtime. I found no additional speedup by applying this
to change to 2.80 or removing the new shader node code.
Ref T71479
Patch by Jeroen Bakker.
Differential Revision: https://developer.blender.org/D6252
We appear to be hitting some limit where adding any amount of code causes a
significant performance regression, no matter what it does. To work around
that a new node level was added.
Ref T71479
Fixes an issue with secondary particles being out of sync with the main simulation. Cleaned up the secondary particle code in general too (making sure that all solver attributes - timestep, framelength, etc. - are set correctly).
Back in 2.79 you could either use the debug panel or an
environment variable to override using OpenCL for unsupported
hardware. Which was rather useful for developers when testing
on NVidia just to be sure the CL kernels at-least build properly.
This broke in rB949ab753bb2
This diff restores testing though the CYCLES_OPENCL_TEST
environment variable.
Differential Revision: https://developer.blender.org/D7202
Reviewers: brecht