Commit be63ebd961
This is causing issues with CUDA kernel compilation in some setups, even
though the builbot is ok. Since this isn't yet working for oneAPI anyway,
revert all the changes to Cycles kernel compilation for now.
Includes win32 specific extensions definitions when including
`vk_common.hh`. Inside `gpu_context.cc` vulkan needs to be
included before opengl, otherwise windows 10 builders will
report a warning.
```
[6421/7520] Building CXX object source\blender\gpu\CMakeFiles\bf_gpu.dir\intern\gpu_context.cc.obj
C:\Program Files (x86)\Windows Kits\10\include\10.0.19041.0\shared\minwindef.h(130): warning C4005: 'APIENTRY': macro redefinition
C:\Users\blender\git\blender-vexp\blender.git\lib\windows_x64\epoxy\include\epoxy/gl.h(59): note: see previous definition of 'APIENTRY'
```
Pull Request: https://projects.blender.org/blender/blender/pulls/137134
In the OpenXR/Vulkan specs it is mentioned that the swapchain layout
should be set to optimal color attachment. This wasn't the case and
could lead to validation errors.
Also fixes a memory size validation error. Not concerning as it was
abount already backed memory.
Pull Request: https://projects.blender.org/blender/blender/pulls/137143
This only worked for Ubuntu's discontinued Unity desktop.
Even though `libunity.so` can be used outside of Unity,
it's no longer actively developed.
Ref: !137128
When using SteamVR without starting SteamVR, the physical device cannot
be found. This isn't reported to the user nicely. This change will
tell the user why OpenXR couldn't be started.
Note that the physical device can be created when SteamVR has already
launched. OpenGL will start SteamVR automatically, but seems Vulkan
isn't able to do that.
Make the 'Available' keying set look at the F-Curves for the assigned
slot, instead of the backward-compatible API (which only sees the F-Curves
for the first slot).
Pull Request: https://projects.blender.org/blender/blender/pulls/137131
Remove the space-padding from the channel name in the Debug Channel List
operator.
This is a follow-up commit after !137105, to present things a little
cleaner.
Expand the "Debug Channel List" operator so that it also shows the channel
settings. The setting name is printed in ALL-CAPS if it is on, and in
lower case letters when it is off.
Note that this operator is only available in debug builds of Blender.
Pull Request: https://projects.blender.org/blender/blender/pulls/137105
Support rendering to BMP images with alpha since OIIO supports this.
This was likely left disabled from when Blender's own BMP I/O didn't
support alpha.
Ref !137123
Remove full-screen support from GHOST API's.
Note that this only had back-end implements for X11 and WIN32.
This was last used for the Game Engine to run games full-screen,
removing as it's unused and it doesn't seem likely to be used in the
future.
This doesn't impact making Blender full-screen from the window menu
which uses a window decoration setting.
Ref: !137050
The gaps between editors need to be selected and manipulated in order
to resize areas. How wide these need to be depends on the accuracy and
resolution of your pointing device. And also on the fine motor control
and visual acuity of the user. If you are using a tablet pen, touch
device, or are visually or physically challenged, then you need these
gaps to be wider. This PR allows a wide latitude in this.
Pull Request: https://projects.blender.org/blender/blender/pulls/126334
ProRes is a common intra-frame codec in post-production work, supported
by a wide range of post-production software.
This PR adds support for direct output from Blender using the ProRes
codec from FFmpeg. Alpha is supported, along with 8 and 10-bit channel
images.
Co-authored-by: mvji <33432858+mvji@users.noreply.github.com>
Pull Request: https://projects.blender.org/blender/blender/pulls/136405
The multi-frame falloff calculation had multiple issues.
It was also not the same as in 4.2 LTS.
This fixes the issues and should result in the exact same
falloffs as it did in 4.2 LTS.
Also resolves#137055.
Pull Request: https://projects.blender.org/blender/blender/pulls/137091
When multiple F-Curves have the `FCURVE_ACTIVE` flag set, only one of them
was drawn in the graph editor.
Only one F-Curve should be marked with this flag, but that's not something
the datamodel enforces. Because of that, it apparently is possible to get
into a situation where multiple F-Curves are marked as such. They are now
all drawn properly.
Pull Request: https://projects.blender.org/blender/blender/pulls/137102
This PR implements dynamic viewport state for the Vulkan gpu backend.
By doing so, it fixes#130914.
The following high-level changes were made:
1. The pipeline pool no longer uses the viewport and scissor
states to identify graphics pipelines, only the number of viewports
and the number of scissors. Graphics pipelines are configured with
dynamic viewport and scissor states upon construction.
2. The desired viewport and scissor configurations for drawing are set
in the data of the draw nodes in the render graph.
3. The draw nodes use these viewport and scissors settings in
`build_commands`. If the viewport and scissor settings have changed
between nodes, then vkCmdSetViewport and vkCmdSetScissor commands
are sent to the command buffer.
4. Tests are updated to verify that set_viewport and set_scissor commands
are executed the correct number of times. (Also note that I needed to
#136987 in order to avoid skipping some Vulkan tests).
See the attached screencast for verification. The number of graphics pipelines
no longer grow when resizing the viewport.
Pull Request: https://projects.blender.org/blender/blender/pulls/137002
Removes '/' prefix from gtest filter for parameterized tests. This prefix
was causing tests in fixtures like VKRenderGraphTestScheduler and
VKRenderGraphTestRender to be skipped in `make test`.
Only parameterized tests with an InstantiationName will have a '*/' in
their names as part of a prefix. But according to the spec, InstantiationName
is optional, and if not set it wont generate prefixes. In these cases, the
gtest_filter produced will not match the test names, and cause `make test`
to not run the tests.
Removing this '*/' prefix should still match any tests with a non-empty
InstantiationName.
Pull Request: https://projects.blender.org/blender/blender/pulls/136987
Use VERBATIM to ensure spaces inside command line arguments don't get
escaped automatically.
On Linux and Windows the oneAPI kernel compilation still has problems.
There is an apparent bug with single quote escaping in add_custom_command
which means it's not easy to use VERBATIM.
This showed: argument 1 range [18446744071562067968, 18446744073709551615]
exceeds maximum object size 9223372036854775807.
It seems to think signed integers might lead to this kind of overflow,
but it's unclear why these two cases are causing warnings specifically.
Refactor the code to sidestep it.
This was caused by the cross product changing
sign in view space during planar lightprobe capture.
This can be avoided by making sure the view matrix
doesn't change handedness.
Pull Request: https://projects.blender.org/blender/blender/pulls/137092
The "new/duplicate" button in the Action Slot selector did not actually
duplicate, and always acted as a "new" button.
This introduces the RNA function `ActionSlot.duplicate()`, which takes
care of duplicating all the animation data associated with the slot as
well. The semantics of this function should remain valid in the
future, when Actions support multiple layers & strips. Note that this
new function does not assign the slot, it just duplicates it and its
data. The assignment of this duplicated slot is done in Python,
through the already-existing API for this.
Pull Request: https://projects.blender.org/blender/blender/pulls/137087
Update the `ShaderCompilerGeneric` to support deferred compilation
using the batch compilation API, so we can get rid of
`drw_manager_shader`.
This approach also allows supporting non-blocking compilation
for static shaders.
This shouldn't cause any behavior changes at the moment, since batch
compilation is not yet used when parallel compilation is disabled.
This adds a `GPUWorker` and a `GPUSecondaryContext` as an easy to use
wrapper for managing secondary GPU contexts.
(Part of #133674)
Pull Request: https://projects.blender.org/blender/blender/pulls/136518
Blender crashes when using math interpolation functions to sample Float2
images. That's because 3 components is intentionally left out of various
switch case. To fix this, we handle 3 components and assert for expected
components count.
Pull Request: https://projects.blender.org/blender/blender/pulls/137094
