Improve the number of images that are requested in the swapchain. For
mailbox presentation mode we select tripple buffering. For V-Sync
presentation mode we select double buffering.
In future we might want to make the presentation mode and swapchain
images dynamic based on the actual action that the user is performing.
When doing action that benefit from low latency (paint cursor) we should
select mailbox, otherwise we can use fifo to reduce CPU/GPU usage and
safe some energy usage along the way. See #136923 for design task.
Pull Request: https://projects.blender.org/blender/blender/pulls/136922
It's better for performance to use a single thread pool for all areas of
Blender, and this gets us closer to that.
Bullet, Quadriflow, Mantaflow and Ceres still contain OpenMP code, but it
was already disabled.
On macOS, our OpenMP libraries are no longer compatible with the latest
Xcode 16.3. By removing OpenMP we no longer have to solve that problem.
OpenMP was disabled for bpy module builds on Windows ARM64, which also no
longer needs to be solved.
Pull Request: https://projects.blender.org/blender/blender/pulls/136865
Only the parallel sparse matrix code was updated. This is used by e.g.
LSCM and ABF unwrap, and performance seems about the same or better.
Parallel GEMM (dense matrix-matrix multiplication) is used by libmv,
for example in libmv_keyframe_selection_test for a 54 x 54 matrix.
However it appears to harm performance, removing parallelization makes
that test run 5x faster on a Apple M3 Max.
There has been no new Eigen release since 2021, however there is active
development in master and it includes support for a C++ thread pool for
GEMM. So we could upgrade, but the algorithm remains the same and
looking at the implementation it just does not seem designed for modern
many core CPUs. Unless the matrix is much larger, there's too much thread
synchronization overhead. So it does not seem useful to enable that
thread pool for us.
Pull Request: https://projects.blender.org/blender/blender/pulls/136865
To make adding a dependeny on TBB easier.
Additional changes:
* Using LIB for libmv tests, as it now brings in includes
* Removing Eigen header listing in iTaSC
Pull Request: https://projects.blender.org/blender/blender/pulls/136865
The goal is to be able to switch away from OpenMP by default.
In order to achieve this a new parallel_for() function is added,
which follows the same declaration as the one from TBB. For now
the simplest formulation is used where range is provided by start
and last indices (the last one is excluded from the range).
The side effect is that Libmv now expects the threading limits
to be imposed by the default thread area (or have an explicit
thread area with the limit in the caller). In a way this is
similar to other external libraries.
Pull Request: https://projects.blender.org/blender/blender/pulls/136834
MacOS is currently using one of our custom "hourglass" cursors during
busy times. This PR changes it to an OS-supplied cursor, a pointer
arrow with an animated blue spinner, meant for this purpose. Although
undocumented it has been used by many applications for many years.
Including Firefox for 11 years now.
Pull Request: https://projects.blender.org/blender/blender/pulls/136735
Reduce the register pressure and branching in the switch() by using
subclass and cast from void* to the base class.
This ensures intersection functions are not inlined multiple times,
bringing performance back.
Alternative could be to avoid functions (they are quite large) but
that only partially resolves the performance regression.
Pull Request: https://projects.blender.org/blender/blender/pulls/136823
This PR refactors the way how swapchains are used.
Allow scaling of the swapchain content to the actual resolution of the swapchain.
can reduce artefacts when resizing windows when supported.
When frame rate is to fast the previous implementation could use a semaphore
that were still in use, leading to unwanted stuttering on certain platforms. Waiting
when the rendering has finished (GHOST_Frame.submission_fence), before the
next image is acquired from the swap chain.
Mailbox has been disabled as it can calculate more frames then actually been
presented, leading to a lag and increased power usage on others.
Pull Request: https://projects.blender.org/blender/blender/pulls/136603
Instead of relying on the Intel extensions that may not be implemented,
we can use max_work_group_size until there is a better alternative.
Thanks to Codeplay for this proposal.
Co-authored-by: Georgi Mirazchiyski <georgi.mirazchiyski@codeplay.com>
Area light with a size of zero should not contribute to the scene, so
set the light as disabled.
This does not only fix the reported bug where such light is visible to
the camera, but also a regression in 4.2 where the light contributes to
the scene when light tree is off.
Pull Request: https://projects.blender.org/blender/blender/pulls/136763
This API is not properly implemented in other SYCL backends at the
moment and we don't want it to fail at runtime, so we conservatively
enable it only for Level-Zero.
Mailbox is lowest latency V-Sync enabled mode. Previously we selected
FIFO as that was always available and has more support when using debug
tools.
Improve the #136521 situation as well.
Pull Request: https://projects.blender.org/blender/blender/pulls/136753
The initial issues that led to the choice of forcing the use of
linker.exe seem gone and there is currently no strong reason to use
linker.exe explicitly, so let's simplify and use the default setting.
Match the behavior of other platforms by clearing the active window
when it's de-activated.
On Win32 Blender always considered the last-active window to be active
because de-activation wasn't handled.
Part of !136122, needed to fix#134733.
Co-authored-by: Kamil Galik <kgalik@3dconnexion.com>
The Blender config.ocio does not have any rules, but custom ones can.
The default file rule is ignored if default_byte or default_float roles
exist. These roles are Blender specific, so would not be found in a typical
OCIO config. But when they are set appropriately, they help provide better
default guesses than what is possible with standard OCIO rules.
Pull Request: https://projects.blender.org/blender/blender/pulls/136516
This makes it available in Cycles standalone, and the implementation
can be shared with Blender. This also makes it possible to compute
tangents after tessellation for adaptive subdivision.
There is a difference in UV map tangents when there are no UVs. They
are now generated from object space coordinates instead of auto
texture space coordinates. This is more efficient, and a corner case
that we don't have to keep compatible.
Co-authored-by: Brecht Van Lommel <brecht@blender.org>
Pull Request: https://projects.blender.org/blender/cycles/pulls/25
Add support for OSL parameter metadata named `defaultgeomprop`, whose
values are interpreted the same way as the property on MaterialX node
inputs. When set to `Tworld` the tangent is then automatically linked
to the shader and generated for the mesh.
Pull Request: https://projects.blender.org/blender/cycles/pulls/25
The initial limitation preventing from using -ffast-math, worked around
in 09df1f4caf, got fixed upstream in LLVM
and the fix is part of current DPC++ compiler:
63ecd2a725
We're now able to go back to using -ffast-math, which helps simplifying
the set of compiler flags.
No performance nor conformance change is expected from this change (most
of the gain is achieved already with the use of -cl-fast-relaxed-math
since 284b89a0a3) and this has been
verified on Arc B580 under Windows.
The Blender's VkInstance cannot be shared with OpenXR VkInstance. The
reason is a chicken and egg problem where OpenXR needs to be started
before Vulkan. OpenXR can add special vulkan specific requirements
(instance&device) that are only available when the user starts an OpenXR
session.
The goal implementation is to share memory between both instances using
[VK_KHR_external_memory](https://registry.khronos.org/vulkan/specs/latest/man/html/VK_KHR_external_memory.html) and related extensions. However this seems
to be a bridge to far as a initial step. Reason: There are not that many
samples/ guides and documentation to be found to handle the workflow that
we require. We want to do a smaller step by step approach to gain the needed
knowledge.
For that reason this PR does the most stupidest thing that can be done to
share memory between instances. Download the render result to CPU RAM share
the host pointer with the OpenXR instance which copies it to the swap chain.
Also the synchronization is done using wait idle commands.
<video src="attachments/32a0d69b-c3fa-4272-aea0-d207609afaaf" title="Screencast From 2025-03-18 11-16-17.webm" controls></video>
**Gaining knowledge**
- Experiment with `VK_KHR_external_memory_host` extension for uploading vertex buffers (not related to OpenXR).
- Import host pointer with `VK_KHR_external_memory_host`. This reduces the additional
memcpy on OpenXR side.
- Export host pointer from Blender side from a mappable buffer.
- Replace host pointers with fd/dmabuf/winhandle
- Remove mappable buffer.
Ref #133718
Pull Request: https://projects.blender.org/blender/blender/pulls/133824
Instead of returning 0 in case the Intel extension for getting the count
of Execution Units isn't available, we now use
sycl::info::device::max_compute_units.
We keep using the Intel extension in priority since it logically goes
with sycl::ext::intel::info::device::gpu_hw_threads_per_eu used in
get_max_num_threads_per_multiprocessor(), for which there is no
sycl::info::device::max_threads_per_compute_unit replacement yet.
The debug set of Embree prebuilt libraries currently lacks SYCL support
while the release ones have it.
This case was not gracefully handled for debug builds with Embree on GPU
enabled, leading to linking errors, trying to resolve rtcNewSYCLDevice
and rtcIsSYCLDeviceSupported.
We now test for this case to explicitly disable the use of Embree on GPU
for debug builds on Windows and print this status from CMake.
This fixes the following warning with MSVC:
device_impl.cpp(287): warning C4805: '|=': unsafe mix of type 'bool' and type 'ccl::uint' in operation
The similar fix is applied to Metal code as well.
There is no short-circuiting boolean operator ||=, so expand the expression.
Pull Request: https://projects.blender.org/blender/blender/pulls/136561
HIP-RT functions do have access to kg, and it was used inconsistently:
some functions were passed actual kg, other were passed nullptr.
This change makes it consistent and passes kg everywhere.
Pull Request: https://projects.blender.org/blender/blender/pulls/136503
The code before this change was relying on the ShadowPayload have
the same "header" as RayPayload for some of the primitive types
(curve, motion triangle, point): intersection functions were shared
between "regular" and shadow rays (shadow in this case is shadow_all),
but extra filter function was used for shadow rays.
This is fragile if someone changes one of these structures. What is
worse is that compiler might actually decide to shuffle things in
some structs, or remove unused fields.
This change also solves confusion about ShadowPayload::prim_type
seemingly only being assigned to PRIMITIVE_NONE. With time it is
not impossible that compiler will also see this, and constant-fold
some checks, or even remove the field. If that happens then the
render result will be wrong. Maybe it is already happening as there
are some GPU and driver and optimization flag specific bugs in the
area.
It is unclear whether it was causing any actual problem: W7800
seems to render all hair correctly on Linux.
