This PR adds a global mutex to `path_create_directories` to fix a thread-safety issue which can occur when concurrently creating multiple subdirectories with common stems.
Pull Request: https://projects.blender.org/blender/blender/pulls/139266
Previously, we were drawing the context path in the header of the spreadsheet.
However, that had some problems:
* When using a viewer that's somewhere deep in a node group, the viewer path
wouldn't fit.
* Standard editor menus didn't fit in. Also we wanted to add spreadsheet
specific operators that should be in a new menu.
* Couldn't fit more useful data for the context path (like the inspection index
for repeat zones).
This patch solves this by moving the entire context path to the left side bar.
This frees up the header for menus.
The new context panel information is added at the top of the side bar because
then there is a proper hierarchy: `Object > Evaluation State > Viewer Path
(optional) > Instance Selection > Domain Selection`
This patch also adds information about the current inspection index for for-each
and repeat zones. They are not editable yet, but that can be implemented
separately as it might require a few more changes for everything to update
correctly.
The new Viewer Path panel is only displayed when actually showing viewer data.
Pull Request: https://projects.blender.org/blender/blender/pulls/138477
The stencil value was also considered and could lead to out of range
depth values. These were ignored by operators and could lead to
printing errors, canceling operators, or inaccurate depth
selection.
This is a NVIDIA only issue as these GPU support DEPTH24S8 textures.
We should consider defaulting to DEPTH32FS8.
Pull Request: https://projects.blender.org/blender/blender/pulls/139272
In detail:
- Direct accesses of state attributes are replaced with the INTEGRATOR_STATE and INTEGRATOR_STATE_WRITE macros.
- Unified the checks for the __PATH_GUIDING define to use # if defined (__PATH_GUIDING__).
- Even if __PATH_GUIDING__ is defined, we now check if the feature is enabled using if ((kernel_data.kernel_features & KERNEL_FEATURE_PATH_GUIDING)) {. This is important for later GPU ports.
- The kernel usage of the guiding field, surface, and volume sampling distributions is wrapped behind macros for each specific device (atm only CPU). This will make it easier for a GPU port later.
If the `transdata_check_local_islands` check does not succeed, we end up
with `CTX_NO_PET` for scaling/rotation.
For `Grease Pencil`, this was lost in the transition from GPv2 > GPv3
(type was renamed to "legacy", then removed from
`transdata_check_local_islands` in a804320138, but the new type was
never added).
Curves never specified support for this, but both types are (in theory)
taking into account `V3D_AROUND_LOCAL_ORIGINS` in
`curve_populate_trans_data_structs` (logic from 9d86fada03 might be a
bit broken for proportional editing "islands" without anything selected
as reported in #139101, but that will be corrected in another PR).
So to resolve, add `Grease Pencil` and `Curves` as supported object
types in `transdata_check_local_islands`
NOTE: `transdata_check_local_center` already has these two types as
well.
Part of #139101
Pull Request: https://projects.blender.org/blender/blender/pulls/139209
Add back the deprecated format to avoid asserts.
GPU backends should still have the code to support them.
Also add deprecation warnings as these types will be
removed for 5.0.
Pull Request: https://projects.blender.org/blender/blender/pulls/139213
This promotes any INT_TO_FLOAT attribute to F32 and
to the float promotion inside python `attr_fill`.
We issue a deprecation warning when hitting this path.
Raise an error is the component type is not integer.
Pull Request: https://projects.blender.org/blender/blender/pulls/139206
It can happen that the previous context drew with a different
colorspace. In the case where the new context is drawing with
the same shader that was previously bound (shader binding
optimization), the uniform would not be set again because the
dirty flag would not have been set (since the color space of
this new context never changed). The shader would reuse the same
colorspace as the previous context framebuffer (see #137855).
Fix#137855
Pull Request: https://projects.blender.org/blender/blender/pulls/139226
Support parsing C style comments such as:
`char filepath[/*FILE_MAX*/ 1024]`
This is done by skipping white-space inside square brackets
so the value is read as:
`char filepath[1024]`.
Ref: !139196
This was caused by the removal of some compatibility code
in the metal backend.
This patch reintroduce this compatibility code in a cleaner
way than before.
This should be followed up by a documentation commit that
explains what is supported and what is not.
Pull Request: https://projects.blender.org/blender/blender/pulls/139185
This patch adds snapping options for the playhead to all animation editors.
The options can be modified through a new dropdown in the editor header.
All editors will show all those options, and they are shared,
so toggling the option in on editor will change it for all other editors too.
Some options are not working/relevant in some editors for example
Strips in the Dope Sheet. However for consistency the option is still shown.
This is a separate menu from the transform snapping menu because
you can toggle the snapping for transform and playhead separately.
Putting it in the existing snapping transform menu would imply that it can
be turned off with the magnet which is not the case.
Playhead snapping is explicitly disabled for the drivers editor
because there is no playhead to drag around.
Snapping to Frame/Second intervals takes the scene start as a starting point.
That means you can snap to the n-th second of the animation even though
it might not start at frame 1. The preview range is NOT taken into account
by design since the use case is working on a sub-section of the animation
in which case the snap target should not change.
Snapping is toggled by pressing CTRL as indicated by the status bar.
Snapping to Frames/Seconds is absolute, meaning no matter
how far away your cursor it will snap to the closest snap point.
All others only snap to things if they are close to the cursor in pixel values.
When mixing those two behaviors, it prefers relative snapping.
If no point is close enough to snap relative,
it will fall back to absolute snapping.
Based on the prototype #135913
Part of #135794
Pull Request: https://projects.blender.org/blender/blender/pulls/137278
When landing #134860 we didn't yet have documentation for the Path
Templates feature. Now that we do, we can refer users to it.
This PR adds a reference to the docs in the type info of properties that
are marked as supporting Path Templates.
Pull Request: https://projects.blender.org/blender/blender/pulls/139099
For IDProperties, this flag is mostly used only to controll whether
IDProperty code should also handle ID refcounting or not (through
`LIB_ID_CREATE_NO_USER_REFCOUNT`).
Without this fix, IDProp Group Merge could end up trying to modify ID
refcount of its data-block properties (via `IDP_FreeProperty`), which
can lead to data corruption or even crashes (e.g. when used in readfile
code, before lib-linking process).
Issue discovered while working on the system IDProperties split project
(!135807).
Pull Request: https://projects.blender.org/blender/blender/pulls/139230
* Avoid usage of `applied` in favor of helper methods to indicate
undo and redo requirements
* Rename `bmesh_restore_begin` and `bmesh_restore_end` to avoid
similarity with geometry methods
* Swap `geometry_original` and `geometry_modified` to match
expected meaning
Pull Request: https://projects.blender.org/blender/blender/pulls/139244
For Sculpt Undo, certain operators will modify the topology count of the
mesh. These operators are handled separately from normal brush strokes,
and so having tests for an operator that uses this functionality is
beneficial in detecting regressions.
Pull Request: https://projects.blender.org/blender/blender/pulls/139249
Previously, when a socket was detected to be unused, it was just grayed out.
This patch adds support for automatically hiding unused sockets based on this
convention: Menu inputs control visibility while other inputs only control
whether something is grayed out.
More specifically, an input is visible if any of these conditions is met:
* It affects the output currently.
* It never affects the output. In this case its usage does not depend on any
menu input.
* It is used if all non-menu inputs are considered to be unknown.
In the future, we could support customizing which inputs are allowed to control
visibility. For now it's good to use the convention that Blender generally
follows itself.
As before, panels are grayed out if they only contain grayed out sockets and
panels are hidden when they don't contain any visible sockets.
Hiding inputs works in group nodes, the Geometry Nodes modifier and node
operators. In theory it will work for all node tree types, but since only
Geometry Nodes supports the Menu Switch node currently, this patch currently
only makes a difference there.
The implementation reuses the existing `SocketUsageInferencer` with a different
sets of inputs. So no new core-inferencing logic was needed.
Design task: #132706.
Pull Request: https://projects.blender.org/blender/blender/pulls/138186
Regression in [0] caused the poll function to compare the result with
nullptr instead of std::nullopt.
The operators exec function then de-referenced the the std::nullopt.
[0]: 550094b018
The particle system generates some particles with NaN values. The
set_if_different mechanism skipped copying those due to a refactor
in the matrix equality test. Revert that part of 689633d802 for now.
A better solution would be to improve handling of NaNs in Cycles,
and to find and fix the cause of the NaN in the particle system.
Pull Request: https://projects.blender.org/blender/blender/pulls/139238
