If one of the devices already used host happed memory but another not,
it would previously realloc both.
Thanks to Jorn Visser for investigating and finding this problem.
Pull Request: https://projects.blender.org/blender/blender/pulls/132912
Should be a bit more efficient, and it fixes host memory fallback bugs,
where host memory was incorrectly freed during re-copy. For the case
where memory should get reallocated on the host, a new mem_move_to_host
was added.
Thanks to Jorn Visser for investigating and finding this problem.
Pull Request: https://projects.blender.org/blender/blender/pulls/132912
Check was misc-const-correctness, combined with readability-isolate-declaration
as suggested by the docs.
Temporarily clang-format "QualifierAlignment: Left" was used to get consistency
with the prevailing order of keywords.
Pull Request: https://projects.blender.org/blender/blender/pulls/132361
* Use .empty() and .data()
* Use nullptr instead of 0
* No else after return
* Simple class member initialization
* Add override for virtual methods
* Include C++ instead of C headers
* Remove some unused includes
* Use default constructors
* Always use braces
* Consistent names in definition and declaration
* Change typedef to using
Pull Request: https://projects.blender.org/blender/blender/pulls/132361
Precompiled Cycles kernels make up a considerable fraction of the total size of
Blender builds nowadays. As we add more features and support for more
architectures, this will only continue to increase.
However, since these kernels tend to be quite compressible, we can save a lot
of storage by storing them in compressed form and decompressing the required
kernel(s) during loading.
By using Zstandard compression with a high level, we can get decent compression
ratios (~5x for the current kernels) while keeping decompression time low
(about 30ms in the worse case in my tests). And since we already require zstd
for Blender, this doesn't introduce a new dependency.
While the main improvement is to the size of the extracted Blender installation
(which is reduced by ~400-500MB currently), this also shrinks the download on
Windows, since .zip's deflate compression is less effective. It doesn't help on
Linux since we're already using .tar.xz there, but the smaller installed size
is still a good thing.
See #123522 for initial discussion.
Pull Request: https://projects.blender.org/blender/blender/pulls/123557
Since #118841 there are more cases where Cycles would check for the
graphics interop support. This could lead to a crash when graphics
interop functions are called without having active graphics context.
This change makes it so there is no graphics interop calls when doing
headless render. In order to achieve this the device creation is now
aware of the headless mode.
Pull Request: https://projects.blender.org/blender/blender/pulls/122844
This enables the new lazy module loading behavior introduced in OIDN 2.3,
without breaking compatibility with older versions of OIDN (using separate
code paths).
Also, the detection of OIDN support for devices is now much cleaner, and
devices do not need to be matched by PCI address or device name anymore.
Pull Request: https://projects.blender.org/blender/blender/pulls/121362
With the switch to using the primary CUDA context it became possible
for peer access between CUDA devices to already have been enabled for
that context, either by a previous Cycles session or third-party library,
thus causing the call to `cuCtxEnablePeerAccess` to return
`CUDA_ERROR_PEER_ACCESS_ALREADY_ENABLED`. This is not a failure
state however, so just needs to be handled like a success return value.
Pull Request: https://projects.blender.org/blender/blender/pulls/120255
Previously the CUDA context was always destroyed and the module along
with it. Now that this no longer happens, the missing module free became
a memory leak.
Also fix the same issue for HIP, though this is destroying the context
so it's not a problem yet.
Fix part of #119035
Co-authored-by: Brecht Van Lommel <brecht@blender.org>
Host memory fallback in CUDA and HIP devices is almost identical.
We remove duplicated code and create a shared generic version that
other devices (oneAPI) will be able to use.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D17173
This adds a new mirror image extension type for shaders and
geometry nodes (next to the existing repeat, extend and clip
options).
See D16432 for a more detailed explanation of `wrap_mirror`.
This also adds a new sampler flag `GPU_SAMPLER_MIRROR_REPEAT`.
It acts as a modifier to `GPU_SAMPLER_REPEAT`, so any `REPEAT`
flag must be set for the `MIRROR` flag to have an effect.
Differential Revision: https://developer.blender.org/D16432
If no OPTIX_ROOT is set, nvcc fails to compile because there is a stray "-I"
in the arguments. Detect if the include path is empty and act accordingly.
Differential Revision: https://developer.blender.org/D16308
This patch tunes the integrator state sizing for Metal (`num_concurrent_states` and `num_concurrent_busy_states`).
On all GPUs architecture, we adjust the busy:total states ratio to be 1:4 which gives better rendering performance than the previous 1:16 ratio (independent of total state count). This gives a small performance uplift (e.g. 2-3% on M1 Ultra).
Additionally for M2 architectures, we double the overall state size if there is available headroom. Inclusive of the first change, we can expect uplift of close to 10% in future, as this results in larger dispatch sizes and minimises work submission overheads. In order to make an accurate determination of available headroom, we defer the calculation of `num_concurrent_states` and `num_concurrent_busy_states` until the time of integrator state allocation (i.e. after all of the scene data has been allocated). We also refactor `alloc_integrator_soa` to calculate an *exact* single-state-size in a first pass, right before allocating the integrator SoA buffers in a second pass.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D16313
When verbose level 4 is enabled, Blender prints kernel performance
data for Cycles on GPU backends (except Metal that doesn't use
debug_enqueue_* methods) for groups of kernels.
These changes introduce a new CYCLES_DEBUG_PER_KERNEL_PERFORMANCE
environment variable to allow getting timings for each kernels
separately and not grouped with others. This is done by adding
explicit synchronization after each kernel execution.
Differential Revision: https://developer.blender.org/D15971
This patch causes the render buffers to be copied to the denoiser
device only once before denoising and output/display is then fed
from that single buffer on the denoiser device. That way usually all
but one copy (from all the render devices to the denoiser device)
can be eliminated, provided that the denoiser device is also the
display device (in which case interop is used to update the display).
As such this patch also adds some logic that tries to ensure the
chosen denoiser device is the same as the display device.
Differential Revision: https://developer.blender.org/D15657
* Rename "texture" to "data array". This has not used textures for a long time,
there are just global memory arrays now. (On old CUDA GPUs there was a cache
for textures but not global memory, so we used to put all data in textures.)
* For CUDA and HIP, put globals in KernelParams struct like other devices.
* Drop __ prefix for data array names, no possibility for naming conflict now that
these are in a struct.
Move MNEE to own kernel, separate from shader ray-tracing. This does introduce
the limitation that a shader can't use both MNEE and AO/bevel, but that seems
like the better trade-off for now.
We can experiment with bigger kernel organization changes later.
Differential Revision: https://developer.blender.org/D15070
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
Noticed while looking into oneAPI patch.
Seems to be unused, without clear indication why/when it might be
needed. Removing the function simplifies adding the new backend.
Differential Revision: https://developer.blender.org/D14652
* 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 root of the issue is caused by Cycles ignoring OpenGL limitation on
the maximum resolution of textures: Cycles was allocating texture of the
final render resolution. It was exceeding limitation on certain GPUs and
driver.
The idea is simple: use multiple textures for the display, each of which
will fit into OpenGL limitations.
There is some code which allows the display driver to know when to start
the new tile. Also added some code to allow force graphics interop to be
re-created. The latter one ended up not used in the final version of the
patch, but it might be helpful for other drivers implementation.
The tile size is limited to 8K now as it is the safest size for textures
on many GPUs and OpenGL drivers.
This is an updated fix with a workaround for freezing with the NVIDIA
driver on Linux.
Differential Revision: https://developer.blender.org/D13385
This reverts commit 5e37f70307.
It is leading to freezing of the entire desktop for a few seconds when stopping
3D viewport rendering on my Linux / NVIDIA system.
The root of the issue is caused by Cycles ignoring OpenGL limitation on
the maximum resolution of textures: Cycles was allocating texture of the
final render resolution. It was exceeding limitation on certain GPUs and
driver.
The idea is simple: use multiple textures for the display, each of which
will fit into OpenGL limitations.
There is some code which allows the display driver to know when to start
the new tile. Also added some code to allow force graphics interop to be
re-created. The latter one ended up not used in the final version of the
patch, but it might be helpful for other drivers implementation.
The tile size is limited to 8K now as it is the safest size for textures
on many GPUs and OpenGL drivers.
Differential Revision: https://developer.blender.org/D13385
