Fixes NaN in Vector Displacement node caused by the normalization of
0, 0, 0 vectors.
This fixes both visual rendering issues and an "illegal address" error
on the GPU. The "illegal address" error came from the Light Tree
Sampling code not handling the NaN normals well, leading to weird code
paths being taken, eventually leading to a kernel_assert and a
user facing illegal address error.
In Blender 3.5+, this fixes two errors (visual artifacts and errors on
the GPU when using the light tree). In Blender 3.3 it only fixes one
(visual artifacts) because 3.3 doesn't have the light tree. The commit
message should be adjusted to represent that
Pull Request: https://projects.blender.org/blender/blender/pulls/111294
This breaks backwards compatibility some in that 3 sides will be mapped
differently now, but difficult to avoid and can be considered a bugfix.
Similar to rBdd8016f7081f.
Maniphest Tasks: T103615
Differential Revision: https://developer.blender.org/D16910
Known Issues:
- Command buffer failures when using binary archives (binary archives is disabled for Intel GPUs as a workaround)
- Wrong texture sampler being applied (to be addressed in the future)
Ref T92212
Differential Revision: https://developer.blender.org/D16253
This patch optimises the Metal inlining policy. It gives a small speedup
(2-3% on M1 Max) with no notable compilation slowdown vs what is already
in master. Previously noted compilation slowdowns (as reported in T100102)
were caused by forcing inlining for `ccl_device`, but we get better
rendering perf by relying on compiler heuristics in these cases.
Backported to 3.3 because this also fixes a test failure.
Differential Revision: https://developer.blender.org/D16081
Windows drivers 101.3430 fix an important GUI-related crash and it's
best to prevent users from running into it.
Linux drivers weren't affected but still had relevant gpu binary
compatibility fixes, so it makes sense to keep the min-supported version
aligned across OSes.
Port over small part of ray differentials refactor in e949d6da5b that
accidentally fixed this bug. Ensure ray differentials are orthogonal to
the normal.
The recent revert of Apple silicon inlining changes to avoid long compile times
worked on macOS 12, but in macOS 13 Beta it results in render errors. This may
be a compiler bug and perhaps get fixed in time, but try to be on the safe side
and ensure Blender 3.3.0 works regardless.
This brings part of the inlining back, which brings improved performance but
also longer compiler times again. Compile time is around 2min now, where the
previous full inlining was about 5-7min.
Patch by Michael Jones.
Differential Revision: https://developer.blender.org/D15897
sycl/L0 runtime reports compute-runtime version since Intel graphics
driver 101.3268 on Windows, when querying driver version from sycl.
Prior to this driver, it was 0. Now we can bump minimum requirement to
this one and filter-out devices returning 0.
Maniphest Tasks: T100648
The calculation was revised to address two issues:
* Discontinuities occurring when detail was a non-integer greater than 2.
* Levels of detail in the interval [0,1) repeating the levels of detail in
the interval [1,2).
This fixes Cycles, Eevee and geometry nodes.
Differential Revision: https://developer.blender.org/D15785
Fix typo in blender_release.cmake, and ensure that "make release" still works
when ocloc is not available. While a fatal error is useful for debugging, the
current convention is to disable features, especially in cases like this where
there is no simple way to make the feature work.
Differential Revision: https://developer.blender.org/D15774
This gave a 1.1x speedup, however also leads to very long compile times
that make it seems like Blender has stopped working.
This can be brought back in the future behind an option that users can
explicitly enabled.
Fix T100102
Ref D14923, D14763, T92212
Detect cases where a ray-intersection would miss the current triangle, which if
the intersection is strictly watertight, implies that a neighboring triangle would
incorrectly be hit instead.
When that is detected, apply a ray-offset. The idea being that we only want to
introduce potential error from ray offsets if we really need to.
This work for BVH2 and Embree, as we are able to match the ray-interesction
bit-for-bit, though doing so for Embree requires ugly hacks. Tiny differences
like fused-multiply-add or dot product intrinstics in matrix inversion and ray
intersection needed to be matched exactly, so this is fragile.
Unfortunately we're not able to do the same for OptiX or MetalRT, since those
implementations are unknown (and possibly impossible to match as hardware
instructions). Still artifacts are much reduced, though not eliminated.
Ref T97259
Differential Revision: https://developer.blender.org/D15559
zebin format is critical for the compatibility of AoT graphics binaries
across driver versions. It was previously disabled on Linux due to
runtime issues that are now fixed in
https://github.com/intel/compute-runtime/releases/tag/22.31.23852.
The minimum supported driver version isn't bumped to this one yet as
current codebase with current IGC compiler does actually run fine on
earlier drivers and is not running into these issues anymore.
It should consistently use the Cycles pirmitive ID for self intersection detection,
not the one from the OptiX or Embree acceleration structure.
Differential Revision: https://developer.blender.org/D15632
Assume that all faces using the smae material form a closed mesh, so that
joining meshes gives the same result as separate meshes.
It does mean that using different materials on different sides of one
closed mesh do not work, but the meaning of that is poorly defined anyway
if there is a volume interior.
Recently, performance with oneAPI have regressed due some recent
changes in Blender itself. This commit's changes is resolving this
and also improve compilation time for oneAPI backend first
execution (or Blender compilation time in case of AoT).
Regression have appeared after 5152c7c152 and not related to the
changes itself, but increase of kernels complexity introduced with
it. Changes in this commit is marking some Blender functions as
noinlined for oneAPI backend, which helps GPU compiler to deal with
this complexity without any negative side-effects on performance.
Simplifies intersection code a little and slightly improves precision regarding
self intersection.
The parametric texture coordinate in shader nodes is still the same as before
for compatibility.
The number of Execution Units and resident "threads" (simd width * threads
per EUs) are now exposed and used to select the number of states using
a simplified heuristic.
float8 is a reserved type in Metal, but is not implemented. So rename to
float8_t for now.
Also move back intersection handlers to kernel.metal, they can't be in the
class that encapsulates the other Metal kernel functions.
This was tested in some places to check if code was being compiled for the
CPU, however this is only defined in the kernel. Checking __KERNEL_GPU__
always works.
This patch adds required math functions for float8 to make it possible
using float8 instead of float3 for color data.
Differential Revision: https://developer.blender.org/D15525
Having the OptiX/MetalRT/Embree/MetalRT implementations all in one file with
many #ifdefs became too confusing. Instead split it up per device, and also
move it together with device specific hit/filter/intersect functions and
associated data types.
All our intersections functions now work with unnormalized ray direction,
which means we no longer need to transform ray distance between world and
object space, they can all remain in world space.
There doesn't seem to be any real performance difference one way or the
other, but it does simplify the code.
