Age | Commit message (Collapse) | Author |
|
* OneAPI: remove separate float3 definition
* OneAPI: disable operator[] to match other GPUs
* OneAPI: make int3 compact to match other GPUs
* Use #pragma once
* Add __KERNEL_NATIVE_VECTOR_TYPES__ to simplify checks
* Remove unused vector3
|
|
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.
|
|
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
|
|
This helps with debugging, and gives a slightly closer match between CPU
and CUDA/HIP/Metal renders when it comes to ray tracing precision.
|
|
For transparency, volume and light intersection rays, adjust these distances
rather than the ray start position. This way we increment the start distance
by the smallest possible float increment to avoid self intersections, and be
sure it works as the distance compared to be will be exactly the same as
before, due to the ray start position and direction remaining the same.
Fix T98764, T96537, hair ray tracing precision issues.
Differential Revision: https://developer.blender.org/D15455
|
|
The Metal backend now compiles and caches a second set of kernels which are
optimized for scene contents, enabled for Apple Silicon.
The implementation supports doing this both for intersection and shading
kernels. However this is currently only enabled for intersection kernels that
are quick to compile, and already give a good speedup. Enabling this for
shading kernels would be faster still, however this also causes a long wait
times and would need a good user interface to control this.
M1 Max samples per minute (macOS 13.0):
PSO_GENERIC PSO_SPECIALIZED_INTERSECT PSO_SPECIALIZED_SHADE
barbershop_interior 83.4 89.5 93.7
bmw27 1486.1 1671.0 1825.8
classroom 175.2 196.8 206.3
fishy_cat 674.2 704.3 719.3
junkshop 205.4 212.0 257.7
koro 310.1 336.1 342.8
monster 376.7 418.6 424.1
pabellon 273.5 325.4 339.8
sponza 830.6 929.6 1142.4
victor 86.7 96.4 96.3
wdas_cloud 111.8 112.7 183.1
Code contributed by Jason Fielder, Morteza Mostajabodaveh and Michael Jones
Differential Revision: https://developer.blender.org/D14645
|
|
Add more math functions for float4 to make them on par with float3 ones. It
makes it possible to change the types of float3 variables to float4 without
additional work.
Differential Revision: https://developer.blender.org/D15318
|
|
This patch adds a new Cycles device with similar functionality to the
existing GPU devices. Kernel compilation and runtime interaction happen
via oneAPI DPC++ compiler and SYCL API.
This implementation is primarly focusing on Intel® Arc™ GPUs and other
future Intel GPUs. The first supported drivers are 101.1660 on Windows
and 22.10.22597 on Linux.
The necessary tools for compilation are:
- A SYCL compiler such as oneAPI DPC++ compiler or
https://github.com/intel/llvm
- Intel® oneAPI Level Zero which is used for low level device queries:
https://github.com/oneapi-src/level-zero
- To optionally generate prebuilt graphics binaries: Intel® Graphics
Compiler All are included in Linux precompiled libraries on svn:
https://svn.blender.org/svnroot/bf-blender/trunk/lib The same goes for
Windows precompiled binaries but for the graphics compiler, available
as "Intel® Graphics Offline Compiler for OpenCL™ Code" from
https://www.intel.com/content/www/us/en/developer/articles/tool/oneapi-standalone-components.html,
for which path can be set as OCLOC_INSTALL_DIR.
Being based on the open SYCL standard, this implementation could also be
extended to run on other compatible non-Intel hardware in the future.
Reviewed By: sergey, brecht
Differential Revision: https://developer.blender.org/D15254
Co-authored-by: Nikita Sirgienko <nikita.sirgienko@intel.com>
Co-authored-by: Stefan Werner <stefan.werner@intel.com>
|
|
Enables Vega and Vega II GPUs as well as Vega APU, using changes in HIP code
to support 64-bit waves and a new HIP SDK version.
Tested with Radeon WX9100, Radeon VII GPUs and Ryzen 7 PRO 5850U with Radeon
Graphics APU.
Ref T96740, T91571
Differential Revision: https://developer.blender.org/D15242
|
|
This patch unifies the names of math functions for different data types and uses
overloading instead. The goal is to make it possible to swap out all the float3
variables containing RGB data with something else, with as few as possible
changes to the code. It's a requirement for future spectral rendering patches.
Differential Revision: https://developer.blender.org/D15276
|
|
And use them more consistently than before.
|
|
For non-raw, non-sRGB color spaces, always use half float even if that uses
more memory. Otherwise the precision loss from conversion to scene linear or
sRGB (as natively understood by the texture sampling) can be too much.
This also required a change to do alpha association ourselves instead of OIIO,
because in OIIO alpha multiplication happens before conversion to half float
and that gives too much precision loss.
Ref T68926
|
|
|
|
After recent changes to Nishita sky to clamp negative colors, the pixels ended
up a bit brighter which lead to them exceeding the half float max value. The
CUDA float to half function seems to need clamping.
|
|
This patch makes it possible to change the precision with which to
store volume data in the NanoVDB data structure (as float, half, or
using variable bit quantization) via the previously unused precision
field in the volume data block.
It makes it possible to further reduce memory usage during
rendering, at a slight cost to the visual detail of a volume.
Differential Revision: https://developer.blender.org/D10023
|
|
|
|
Inspired by D12936 and D12929, this patch adds general purpose
"Combine Color" and "Separate Color" nodes to Geometry, Compositor,
Shader and Texture nodes.
- Within Geometry Nodes, it replaces the existing "Combine RGB" and
"Separate RGB" nodes.
- Within Compositor Nodes, it replaces the existing
"Combine RGBA/HSVA/YCbCrA/YUVA" and "Separate RGBA/HSVA/YCbCrA/YUVA"
nodes.
- Within Texture Nodes, it replaces the existing "Combine RGBA" and
"Separate RGBA" nodes.
- Within Shader Nodes, it replaces the existing "Combine RGB/HSV" and
"Separate RGB/HSV" nodes.
Python addons have not been updated to the new nodes yet.
**New shader code**
In node_color.h, color.h and gpu_shader_material_color_util.glsl,
missing methods hsl_to_rgb and rgb_to_hsl are added by directly
converting existing C code. They always produce the same result.
**Old code**
As requested by T96219, old nodes still exist but are not displayed in
the add menu. This means Python scripts can still create them as usual.
Otherwise, versioning replaces the old nodes with the new nodes when
opening .blend files.
Differential Revision: https://developer.blender.org/D14034
|
|
|
|
Keep the existing Rec.709 fit and convert to other colorspace if needed, it
seems accurate enough in practice, and keeps the same performance for the
default case.
|
|
From changes in 869a46df2980, ref D14454
|
|
|
|
Propagate the fp settings from the main thread to all the worker threads (the fp settings includes the FZ settings among other things) - this guarantees consistency in execution of floating point math regardless if its executed in tbb thread arena or on main thread
Add FZ mode to arm64/aarch64 in parallel to the way its been done on intel processors, currently compiling for arm target does not set this mode at all, hence potentially runs slower and with possible results mismatch with intel x86.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D14454
|
|
* Add missing GLEW and hgiGL libraries for Hydra
* Fix wrong case sensitive include
* Fix link errors by adding external libs to static Hydra lib
* Work around weird Hydra link error with MAX_SAMPLES
* Use Embree by default for Hydra
* Sync external libs code with standalone
* Update version number to match Blender
* Remove unneeded CLEW/GLEW from test executable
None of this should affect Cycles in Blender.
Ref T96731
|
|
This adds support for selective rendering of caustics in shadows of refractive
objects. Example uses are rendering of underwater caustics and eye caustics.
This is based on "Manifold Next Event Estimation", a method developed for
production rendering. The idea is to selectively enable shadow caustics on a
few objects in the scene where they have a big visual impact, without impacting
render performance for the rest of the scene.
The Shadow Caustic option must be manually enabled on light, caustic receiver
and caster objects. For such light paths, the Filter Glossy option will be
ignored and replaced by sharp caustics.
Currently this method has a various limitations:
* Only caustics in shadows of refractive objects work, which means no caustics
from reflection or caustics that outside shadows. Only up to 4 refractive
caustic bounces are supported.
* Caustic caster objects should have smooth normals.
* Not currently support for Metal GPU rendering.
In the future this method may be extended for more general caustics.
TECHNICAL DETAILS
This code adds manifold next event estimation through refractive surface(s) as a
new sampling technique for direct lighting, i.e. finding the point on the
refractive surface(s) along the path to a light sample, which satisfies Fermat's
principle for a given microfacet normal and the path's end points. This
technique involves walking on the "specular manifold" using a pseudo newton
solver. Such a manifold is defined by the specular constraint matrix from the
manifold exploration framework [2]. For each refractive interface, this
constraint is defined by enforcing that the generalized half-vector projection
onto the interface local tangent plane is null. The newton solver guides the
walk by linearizing the manifold locally before reprojecting the linear solution
onto the refractive surface. See paper [1] for more details about the technique
itself and [3] for the half-vector light transport formulation, from which it is
derived.
[1] Manifold Next Event Estimation
Johannes Hanika, Marc Droske, and Luca Fascione. 2015.
Comput. Graph. Forum 34, 4 (July 2015), 87–97.
https://jo.dreggn.org/home/2015_mnee.pdf
[2] Manifold exploration: a Markov Chain Monte Carlo technique for rendering
scenes with difficult specular transport Wenzel Jakob and Steve Marschner.
2012. ACM Trans. Graph. 31, 4, Article 58 (July 2012), 13 pages.
https://www.cs.cornell.edu/projects/manifolds-sg12/
[3] The Natural-Constraint Representation of the Path Space for Efficient
Light Transport Simulation. Anton S. Kaplanyan, Johannes Hanika, and Carsten
Dachsbacher. 2014. ACM Trans. Graph. 33, 4, Article 102 (July 2014), 13 pages.
https://cg.ivd.kit.edu/english/HSLT.php
The code for this samping technique was inserted at the light sampling stage
(direct lighting). If the walk is successful, it turns off path regularization
using a specialized flag in the path state (PATH_MNEE_SUCCESS). This flag tells
the integrator not to blur the brdf roughness further down the path (in a child
ray created from BSDF sampling). In addition, using a cascading mechanism of
flag values, we cull connections to caustic lights for this and children rays,
which should be resolved through MNEE.
This mechanism also cancels the MIS bsdf counter part at the casutic receiver
depth, in essence leaving MNEE as the only sampling technique from receivers
through refractive casters to caustic lights. This choice might not be optimal
when the light gets large wrt to the receiver, though this is usually not when
you want to use MNEE.
This connection culling strategy removes a fair amount of fireflies, at the cost
of introducing a slight bias. Because of the selective nature of the culling
mechanism, reflective caustics still benefit from the native path
regularization, which further removes fireflies on other surfaces (bouncing
light off casters).
Differential Revision: https://developer.blender.org/D13533
|
|
|
|
|
|
Function overloading of make_float4() doesn't work since it's a macro, just
don't do this minor cleanup then.
|
|
Differential Revision: https://developer.blender.org/D14426
|
|
Use templates to optimize the CPU texture sampler to interpolate using
float for single component datatypes instead of using float4 for all types.
Differential Revision: https://developer.blender.org/D14424
|
|
This patch adds a Hydra render delegate to Cycles, allowing Cycles to be used for rendering
in applications that provide a Hydra viewport. The implementation was written from scratch
against Cycles X, for integration into the Blender repository to make it possible to continue
developing it in step with the rest of Cycles. For this purpose it follows the style of the rest of
the Cycles code and can be built with a CMake option
(`WITH_CYCLES_HYDRA_RENDER_DELEGATE=1`) similar to the existing standalone version
of Cycles.
Since Hydra render delegates need to be built against the exact USD version and other
dependencies as the target application is using, this is intended to be built separate from
Blender (`WITH_BLENDER=0` CMake option) and with support for library versions different
from what Blender is using. As such the CMake build scripts for Windows had to be modified
slightly, so that the Cycles Hydra render delegate can e.g. be built with MSVC 2017 again
even though Blender requires MSVC 2019 now, and it's possible to specify custom paths to
the USD SDK etc. The codebase supports building against the latest USD release 22.03 and all
the way back to USD 20.08 (with some limitations).
Reviewed By: brecht, LazyDodo
Differential Revision: https://developer.blender.org/D14398
|
|
|
|
|
|
|
|
GLUT does not support offscreen contexts, which is required for the new
display driver. So we use SDL instead. Note that this requires using a
system SDL package, the Blender precompiled SDL does not include the video
subsystem.
There is currently no text display support, instead info is printed to
the terminal. This would require adding an embedded font and GLSL shaders,
or using GUI library.
Another improvement to be made is supporting OpenColorIO display transforms,
right now we assume Rec.709 scene linear and display.
All OpenGL, GLEW and SDL code was move out of core cycles and into
app/opengl. This serves as a template for apps that want to integrate
Cycles interactive rendering, with a simple OpenGLDisplayDriver example.
In general this would be adapted to the graphics API and color management
used by the app.
Ref T91846
|
|
* Replace license text in headers with SPDX identifiers.
* Remove specific license info from outdated readme.txt, instead leave details
to the source files.
* Add list of SPDX license identifiers used, and corresponding license texts.
* Update copyright dates while we're at it.
Ref D14069, T95597
|
|
|
|
The rbit instruction is only available starting with ARMv6T2 and
the register prefix is different from what AARCH64 uses.
Separate the 32 and 64 bit ARM branches, add missing ISA checks.
Made sure the code works as intended on macMini with Apple silicon,
and on Raspberry Pi 4 B running 32bit Raspbian OS.
Differential Revision: https://developer.blender.org/D14056
|
|
Use a shorter/simpler license convention, stops the header taking so
much space.
Follow the SPDX license specification: https://spdx.org/licenses
- C/C++/objc/objc++
- Python
- Shell Scripts
- CMake, GNUmakefile
While most of the source tree has been included
- `./extern/` was left out.
- `./intern/cycles` & `./intern/atomic` are also excluded because they
use different header conventions.
doc/license/SPDX-license-identifiers.txt has been added to list SPDX all
used identifiers.
See P2788 for the script that automated these edits.
Reviewed By: brecht, mont29, sergey
Ref D14069
|
|
There are two things achieved by this change:
- No possible downcast of size_t to int when calculating motion steps.
- Disambiguate call to `min()` which was for some reason considered
ambiguous on 32bit platforms `min(int, unsigned int)`.
- Do the same for the `max()` call to keep them symmetrical.
On an implementation side the `min()` is defined for a fixed width
integer type to disambiguate uint from size_t on 32bit platforms,
and yet be able to use it for 32bit operands on 64bit platforms without
upcast.
This ended up in a bit bigger change as the conditional compile-in of
functions is easiest if the functions is templated. Making the functions
templated required to remove the other source of ambiguity which is
`algorithm.h` which was pulling min/max from std.
Now it is the `math.h` which is the source of truth for min/max.
It was only one place which was relying on `algorithm.h` for these
functions, hence the choice of `math.h` as the safest and least
intrusive.
Fixes 32bit platforms (such as i386) in Debian package build system.
Differential Revision: https://developer.blender.org/D14062
|
|
This reverts commit d74bb7be1916744ae56347b49333eac22ebb7339.
Need to re-iterate to have a proper support of all platforms.
|
|
There are two things achieved by this change:
- No possible downcast of size_t to int when calculating motion steps.
- Disambiguate call to min() which was for some reason considered
ambiguous on 32bit platforms `min(int, unsigned int)`.
On an implementation side the `min()` is defined for a fixed width
integer type to disambiguate uint from size_t on 32bit platforms,
and yet be able to use it for 32bit operands on 64bit platforms without
upcast.
Fixes 32bit platforms (such as i386) in Debian package build system.
Differential Revision: https://developer.blender.org/D13992
|
|
|
|
This patch fixes a correctness issue discovered in the `int4 select(...)` function on Apple Silicon machines, which causes bad bvh2 builds. Although the generated bvh2s give correct renders, the resulting runtime performance is terrible. This fix allows us to switch over to bvh2 on Apple Silicon giving a significant performance uplift for many of the standard benchmarking assets. It also fixes some unit test failures stemming from the use of MetalRT, and trivially enables the new pointcloud primitive.
Ref T92212
Reviewed By: brecht
Maniphest Tasks: T92212
Differential Revision: https://developer.blender.org/D13877
|
|
The ustring is not a trivially copyable object from the C++ standard
point of view, so using memcpy on it is strictly wrong. In practice,
however, this is OK since it is just a thin wrapper around char*.
For now use explicit cast to void* same as it was done in other places
of ccl::array implementation. But also localize the place where memory
copy happens to make it easier to support proper non-trivial C++
objects in the future.
|
|
In some places the task scheduler was not initialized in time.
|
|
Consider temporary directory to be variant part of session configuration
which gets communicated to the tile manager on render reset.
This allows to be able to render with one temp directory, change the
directory, render again and have proper render result even with enabled
persistent data.
For the ease of access to the temp directory expose it via the render
engine API (engine.temp_directory).
Differential Revision: https://developer.blender.org/D13790
|
|
No need for it now since all the threading queries and
scheduling is done via TBB.
Should be no functional changes as all the removed code
is supposed to be unused.
|
|
Query TBB for the maximum allowed concurrency, which is free from a bug
in own concurrency detection code. One thing to keep in mind is that now
Cycles is limited by the number of threads in the TBB areana from which
Session is created. This isn't a problem for Blender since we do not limit
arena on Blender side. Could be something to watch out for in other Cycles
integrations.
Differential Revision: https://developer.blender.org/D13658
|