Erroneous cache warming case where the generated material is
identical to default material and cached shader is re-used,
resulting in case where the parent shader is identical to the
source.
Authored by Apple: Michael Parkin-White
Ref #96261
Certain material node graphs can be very expensive to run. This feature aims to produce secondary GPUPass shaders within a GPUMaterial which provide optimal runtime performance. Such optimizations include baking constant data into the shader source directly, allowing the compiler to propogate constants and perform aggressive optimization upfront.
As optimizations can result in reduction of shader editor and animation interactivity, optimized pass generation and compilation is deferred until all outstanding compilations have completed. Optimization is also delayed util a material has remained unmodified for a set period of time, to reduce excessive compilation. The original variant of the material shader is kept to maintain interactivity.
Also adding a new concept to gpu::Shader allowing assignment of a parent shader from which a shader can pull PSO descriptors and any required metadata for asynchronous shader cache warming. This enables fully asynchronous shader optimization, without runtime hitching, while also reducing runtime hitching for standard materials, by using PSO descriptors from default materials, ahead of rendering.
Further shader graph optimizations are likely also possible with this architecture. Certain scenes, such as Wanderer benefit significantly. Viewport performance for this scene is 2-3x faster on Apple-silicon based GPUs.
Authored by Apple: Michael Parkin-White
Ref T96261
Pull Request #104536
Avoid computing the non-derivative height twice.
The height is now computed as part of the main function, while the height at x and y offsets are still computed on a separate function.
The differentials are now computed directly at node_bump.
Co-authored-by: Miguel Pozo <pragma37@gmail.com>
Pull Request #104595
Texture usage flags can now be provided during texture creation specifying
the ways in which a texture can be used. This allows the GPU backends to
perform contextual optimizations which were not previously possible. This
includes enablement of hardware lossless compression which can result in
a 15%+ performance uplift for bandwidth-limited scenes on hardware such
as Apple-Silicon using Metal.
GPU_TEXTURE_USAGE_GENERAL can be used by default if usage is not known
ahead of time. Patch will also be relevant for the Vulkan backend.
Authored by Apple: Michael Parkin-White
Ref T96261
Reviewed By: fclem
Differential Revision: https://developer.blender.org/D15967
The attribute node already allows accessing attributes associated
with objects and meshes, which allows changing the behavior of the
same material between different objects or instances. The same idea
can be extended to an even more global level of layers and scenes.
Currently view layers provide an option to replace all materials
with a different one. However, since the same material will be applied
to all objects in the layer, varying the behavior between layers while
preserving distinct materials requires duplicating objects.
Providing access to properties of layers and scenes via the attribute
node enables making materials with built-in switches or settings that
can be controlled globally at the view layer level. This is probably
most useful for complex NPR shading and compositing. Like with objects,
the node can also access built-in scene properties, like render resolution
or FOV of the active camera. Lookup is also attempted in World, similar
to how the Object mode checks the Mesh datablock.
In Cycles this mode is implemented by replacing the attribute node with
the attribute value during sync, allowing constant folding to take the
values into account. This means however that materials that use this
feature have to be re-synced upon any changes to scene, world or camera.
The Eevee version uses a new uniform buffer containing a sorted array
mapping name hashes to values, with binary search lookup. The array
is limited to 512 entries, which is effectively limitless even
considering it is shared by all materials in the scene; it is also
just 16KB of memory so no point trying to optimize further.
The buffer has to be rebuilt when new attributes are detected in a
material, so the draw engine keeps a table of recently seen attribute
names to minimize the chance of extra rebuilds mid-draw.
Differential Revision: https://developer.blender.org/D15941
This code slipped through the final review step surely caused by a faulty
merge.
Fixes T101372 Regression: World shader setup crashes Blender in rendered view
Regression introduced by rB697b447c2069bbbbaa9929aab0ea1f66ef8bf4d0
MTLContext provides functionality for command encoding, binding management and graphics device management. MTLImmediate provides simple draw enablement with dynamically encoded data. These draws utilise temporary scratch buffer memory to provide minimal bandwidth overhead during workload submission.
This patch also contains empty placeholders for MTLBatch and MTLDrawList to enable testing of first pixels on-screen without failure.
The Metal API also requires access to the GHOST_Context to ensure the same pre-initialized Metal GPU device is used by the viewport. Given the explicit nature of Metal, explicit control is also needed over presentation, to ensure correct work scheduling and rendering pipeline state.
Authored by Apple: Michael Parkin-White
Ref T96261
(The diff is based on 043f59cb3b)
Reviewed By: fclem
Differential Revision: https://developer.blender.org/D15953
Sun Disc is currently not supported because it'll need special handling - on the one hand, I'm not sure if Eevee would handle a 1e6 coming out of a background shader without issues, and on the other hand it won't actually cast sharp shadows anyways.
I guess we'd want to internally add a sun to the lamps if Sun Disc is enabled, but getting that right is tricky since the user could e.g. swap RGB channels in the node tree and the lamp wouldn't match that.
Anyways, that can be handled later, the sky itself is already a start.
Reviewed By: fclem
Differential Revision: https://developer.blender.org/D13522
This patch implements the necessary changes to the GPU module that are
needed by the realtime compositor.
A new function GPU_material_from_callbacks was added to construct a GPU
material from a number of callbacks. A callback to construct the
material graph by adding and linking the necessary GPU material nodes.
And the existing code generator callback. This essentially allows the
construction of GPU materials independent of node trees and without the
need to do any node tree localization.
A new composite source output to the code generator was added. This
output contains the serialization of nodes that are tagged with
GPU_NODE_TAG_COMPOSITOR, which are the nodes linked to the newly added
composite output links.
Two new GPU uniform setters were added for int2 and matrix3 types.
Shader create info now supports generated compute sources.
Shaders starting with gpu_shader_compositor are now considered part of
the shader library.
Additionally, two fixes were implemented. First, GPU setter node
de-duplication now appropriately increments the reference count of the
references resources. Second, unlinked sockets now get their value from
their associated GPU node stack instead of the socket itself.
Differential Revision: https://developer.blender.org/D14690
Reviewed By: Clement
This avoid leaving a `GPUMaterial` in a `GPU_MAT_QUEUED` state which would
block rendering.
Fix T98603: Hang when saving project in material preview mode
Maniphest Tasks: T98603
Differential Revision: https://developer.blender.org/D15135
On certain systems when eevee is used in a 3d viewport could crash. It
happened more often on slower systems or systems with slower glsl compilers.
For example an Intel Mac Mini. The cause was that even if a GPUMaterial
was in used it could be freed.
This uses refcounter instead of double thread mutexes. This should be
more robust and avoir use after free situation.
Also remove redundant structures and the use of scene as the job owner.
This is to make the codegen and shading nodes object type agnostic. This
is essential for flexibility of the engine to use the nodetree as it see
fits.
The essential volume attributes struct properties are moved to the
`GPUMaterialAttribute` which see its final input name set on creation.
The binding process is centralized into `draw_volume.cc` to avoid
duplicating the code between multiple engines. It mimics the hair attributes
process.
Volume object grid transforms and other per object uniforms are packed into
one UBO per object. The grid transform is now based on object which simplify
the matrix preparations.
This also gets rid of the double transforms and use object info orco factors
for volume objects.
Tagging @brecht because he did the initial implementation of Volume Grids.
This commit removes all EEVEE specific code from the `gpu_shader_material*.glsl`
files. It defines a clear interface to evaluate the closure nodes leaving
more flexibility to the render engine.
Some of the long standing workaround are fixed:
- bump mapping support is no longer duplicating a lot of node and is instead
compiled into a function call.
- bump rewiring to Normal socket is no longer needed as we now use a global
`g_data.N` for that.
Closure sampling with upstread weight eval is now supported if the engine needs
it.
This also makes all the material GLSL sources use `GPUSource` for better
debugging experience. The `GPUFunction` parsing now happens in `GPUSource`
creation.
The whole `GPUCodegen` now uses the `ShaderCreateInfo` and is object type
agnostic. Is has also been rewritten in C++.
This patch changes a view behavior for EEVEE:
- Mix shader node factor imput is now clamped.
- Tangent Vector displacement behavior is now matching cycles.
- The chosen BSDF used for SSR might change.
- Hair shading may have very small changes on very large hairs when using hair
polygon stripes.
- ShaderToRGB node will remove any SSR and SSS form a shader.
- SSS radius input now is no longer a scaling factor but defines an average
radius. The SSS kernel "shape" (radii) are still defined by the socket default
values.
Appart from the listed changes no other regressions are expected.
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
Currently there are many function declarations in `BKE_node.h` that
don't actually have implementations in blenkernel. This commit moves
the declarations to `NOD_composite.h`, `NOD_texture.h`, and
`NOD_shader.h` instead. This helps to clarify the purpose of the
different modules.
Differential Revision: https://developer.blender.org/D13869
This includes much improved GPU rendering performance, viewport interactivity,
new shadow catcher, revamped sampling settings, subsurface scattering anisotropy,
new GPU volume sampling, improved PMJ sampling pattern, and more.
Some features have also been removed or changed, breaking backwards compatibility.
Including the removal of the OpenCL backend, for which alternatives are under
development.
Release notes and code docs:
https://wiki.blender.org/wiki/Reference/Release_Notes/3.0/Cycleshttps://wiki.blender.org/wiki/Source/Render/Cycles
Credits:
* Sergey Sharybin
* Brecht Van Lommel
* Patrick Mours (OptiX backend)
* Christophe Hery (subsurface scattering anisotropy)
* William Leeson (PMJ sampling pattern)
* Alaska (various fixes and tweaks)
* Thomas Dinges (various fixes)
For the full commit history, see the cycles-x branch. This squashes together
all the changes since intermediate changes would often fail building or tests.
Ref T87839, T87837, T87836
Fixes T90734, T89353, T80267, T80267, T77185, T69800
ID data-blocks that could be accessed from Python and weren't freed
using BKE_id_free_ex did not release the Python reference count.
Add BKE_libblock_free_data_py function to clear the Python reference
in this case.
Add asserts to ensure no Python reference is held in situations
when ID's are copied for internal use (not exposed through the RNA API),
to ensure these kinds of leaks don't go by unnoticed again.
This patch adds support for AOVs in EEVEE. AOV Outputs can be defined in the
render pass tab and used in shader materials. Both Object and World based
shaders are supported. The AOV can be previewed in the viewport using the
renderpass selector in the shading popover.
AOV names that conflict with other AOVs are automatically corrected. AOV
conflicts with render passes get a warning icon. The reason behind this is that
changing render engines/passes can change the conflict, but you might not notice
it. Changing this automatically would also make the materials incorrect, so best
to leave this to the user.
**Implementation**
The patch adds a copies the AOV structures of Cycles into Blender. The goal is
that the Cycles will use Blenders AOV defintions. In the Blender kernel
(`layer.c`) the logic of these structures are implemented.
The GLSL shader of any GPUMaterial can hold multiple outputs (the main output
and the AOV outputs) based on the renderPassUBO the right output is selected.
This selection uses an hash that encodes the AOV structure. The full AOV needed
to be encoded when actually drawing the material pass as the AOV type changes
the behavior of the AOV. This isn't known yet when the GLSL is compiled.
**Future Developments**
* The AOV definitions in the render layer panel isn't shared with Cycles.
Cycles should be migrated to use the same viewlayer aovs. During a previous
attempt this failed as the AOV validation in cycles and in Blender have
implementation differences what made it crash when an aov name was invalid.
This could be fixed by extending the external render engine API.
* Add support to Cycles to render AOVs in the 3d viewport.
* Use a drop down list for selecting AOVs in the AOV Output node.
* Give user feedback when multiple AOV output nodes with the same AOV name
exists in the same shader.
* Fix viewing single channel images in the image editor [T83314]
* Reduce viewport render time by only render needed draw passes. [T83316]
Reviewed By: Brecht van Lommel, Clément Foucault
Differential Revision: https://developer.blender.org/D7010
This patch allows the user to type a property name into the
Attribute node, which will then output the value of the property
for each individual object, allowing to e.g. customize shaders
by object without duplicating the shader.
In order to make supporting this easier for Eevee, it is necessary
to explicitly choose whether the attribute is varying or uniform
via a dropdown option of the Attribute node. The dropdown also
allows choosing whether instancing should be taken into account.
The Cycles design treats all attributes as one common namespace,
so the Blender interface converts the enum to a name prefix that
can't be entered using keyboard.
In Eevee, the attributes are provided to the shader via a UBO indexed
with resource_id, similar to the existing Object Info data. Unlike it,
however, it is necessary to maintain a separate buffer for every
requested combination of attributes.
This is done using a hash table with the attribute set as the key,
as it is expected that technically different but similar materials
may use the same set of attributes. In addition, in order to minimize
wasted memory, a sparse UBO pool is implemented, so that chunks that
don't contain any data don't have to be allocated.
The back-end Cycles code is already refactored and committed by Brecht.
Differential Revision: https://developer.blender.org/D2057
This is to modernize the API:
- Add meaningful name to all textures (except DRW textures).
- Remove unused err_out argument: only used for offscreen python.
- Add mipmap count to creation functions for future changes.
- Clarify the data usage in creation functions.
This is a cleanup commit, there is no functional change.
# Conflicts:
# source/blender/gpu/GPU_texture.h
This is in preparation of vulkan backend. We move all opengl
functionnalities behind an abstract class.
This also cleansup the "dynamic" ubo create and rename it to
`GPU_uniformbuf_from_list()`
Contains, no functional change.
Part of T68990 Vulkan support.
This follows the GPU module naming of other buffers.
We pass name to distinguish each GPUUniformBuf in debug mode.
Also remove DRW_uniform_buffer interface.
- add the use of DRWShaderLibrary to EEVEE's glsl codebase to reduce code
complexity and duplication.
- split bsdf_common_lib.glsl into multiple sub library which are now shared
with other engines.
- the surface shader code is now more organised and have its own files.
- change default world to use a material nodetree and make lookdev shader
more clear.
Reviewed By: jbakker
Differential Revision: https://developer.blender.org/D8306
