Blender Git Commit Log

USDVolumeReader improvements.

Fixed Volume reference counting error (incorrect
call to id_us_min()). Fixed error reading fieldName
attribute, which was incorrectly being read as a string
rather than a token. Avoiding use of auto and unneeded
declararions. Logically grouping related code. Now
invoking attribute Get() functions with the given time.

CMake: Improve python version mismatch error

When CMake detects and incompatible Python version
it errors out with an error saying at-least python 3.9
is required, but doesn't mention the version it detected.

This makes troubleshooting the problem harder than it
needs to be.

This diff changes the error message to include the python
version CMake detected.

Differential Revision: https://developer.blender.org/D11666

Reviewed By: Ray Molenkamp

Build/Windows: Preliminary VS 2022 support.

This adds preliminary VS 2022 support, since
there currently is no CMake version that
supports the VS2022 IDE only ninja support
was tested.

IDE support should work without any additional
changes as soon as an updated CMake becomes
available.

As VS2022 appears to keep binary compatibility
with earlier MSVC versions, the current SVN
libraries will work for this version.

USD Import: remove variant selection mask.

Removed logic to process variant selection in
Prim Path Mask, as this code needs to be
extended to handle variants specified in child prims.
This functionality can be added back once it's
fully developed. Also, now printing warning message
if the Prim Path Mask doesn't specify a valid prim.

Twist brush: use relax instead of smooth

Refactor of Wintab to use Wintab supplied mouse movement once verified against system input.

Reviewed By: brecht, LazyDodo

Maniphest Tasks: T88852

Differential Revision: https://developer.blender.org/D11508

Geometry Nodes: Use multithreading for the curve to mesh node

This commit optimizes the node for the case where it works on many
splines by allowing it to generate mesh data from their combinations
in parallel. By itself, this made the node around twice as fast in my
test file with a result of 20 million vertices, around 600ms instead of
1.2s before.

That isn't actually a very good result; it reveals another bottleneck,
a single threaded loop over all face corners in the mesh normal
calculation code. As a simple change that might improve performance
in some situations, this commit moves normal calculation out of this
node, so at least the work isn't wasted if the mesh is changed later
on in the node tree anyway.

Node socket type callback for display shape.

This can change the display shape per socket type instead of using
per instance variable.

Refactor: Do not keep a copy of depth buffer in RegionView3D

The depth cache (located in `RegionView3D::depths`) is used for quick
and simple occlusion testing in:
- particle selection,
- "Draw Curve" operator and
- "Interactive Light Track to Cursor" operator,

However, keeping a texture buffer in cache is not a recommended practice.

For displays with high resolution like 8k this represents something
around 132MB.

Also, currently, each call to `ED_view3d_depth_override` invalidates
the depth cache. So that depth is never reused in multiple calls from
an operator (this was not the case in blender 2.79).

This commit allows to create a depth cache and release it in the same
operator. Thus, the buffer is kept in cache for a short time, freeing
up space.

No functional changes.

Rename USDStageReader prune functions.

Renamed USDStageReader prune_by_visibility()
and prune_by_purpose() member functions to
include_by_visibility() and include_by_purpse(),
respectively, to reflect the current semantics.
I.e., both functions are now answering the question,
'should the prim be included in the traversal',
returning true to include and false to exclude.

Added direct attribute search for inputs.

This way an explicit "Get Attribute" is rarely needed.
The "Get" node can still be useful for connecting multiple inputs
to a single attribute lookup, or when a string is passed in to be
turned into an attribute reference.

USDStageReader::prune_by_purpose() refactor.

Per suggestion by Sybren in his review, flipped the samanitcs
for prune_by_purpose() to return true=include and false=exclude.
Also simplified the conditional logic to reduce nesting.

bli: generational_arena: add iterator support

Made as a bidirectional iterator since movement can be in both
directions. Random access iterator is not possible since there can be
gaps in between the elements.

Had a very annoying bug- using the iterator with the standard library
would throw an error "`iterator_category` not defined" even though it
was defined. Spent a lot of time to realize that `difference_type`
should also be defined but other types within the iterator class are
optional. For some reason the compiler does not create the
`iterator_trait` correctly if `difference_type` is missing but throws
error about `iterator_category` which is extremely strange and time consuming to debug :(

GPencil: Fix wrong enum list after merge

GPencil: Fix merge errors

Merge branch 'master' into temp-gpencil-camera-reproject

Conflicts:
source/blender/editors/gpencil/gpencil_utils.c

Fix T89291: Objects with rotation deltas don't rotate in correct axes

Quaternion correction was not implemented and Euler values were being
incorrectly combined.

Fix deadlocks in mesh modifier evaluation and particles

The recent task isolation changes missed two mutex locks that also need task
isolation.

Ref D11603, T89194

USDStageReader::prune_by_visibility() refactor.

Per suggestion by Sybren in his review, flipped the samanitcs
for prune_by_visibility() to return true=include and false=exclude.
Also simplified the conditional logic to reduce nesting.

Merge branch 'master' into greasepencil-object