The length modifier was not handling 2 point curves correctly.
This change does two things:
1) Fix the crash by copying the original indices for 2 point curves into `dst_to_src_point`.
2) Fallback to `extend_curves_straight` for extending 2 point curves.
Pull Request: https://projects.blender.org/blender/blender/pulls/130122
Previously, some places used `curves.points_num() == 0` some other
places `curves.curves_num() == 0` to check if the geometry is empty.
Rather than having these two ways, add an `is_empty()` function
that replaces all these different checks.
Also update the curves geometry tests to use this function.
Pull Request: https://projects.blender.org/blender/blender/pulls/130168
The Double Edge Mask node uses too much memory when it is not connected
to anything in the Full-Frame compositor. That's because it tries to
allocate a buffer of size 131072x131072, which is due to the cumulation
of multiple issues.
1. The Full-Frame compositor evaluates nodes that are not connected to
the output.
2. The constant folder assigns a bounds of +-65536 to its output memory
buffers as a maximum possible value.
3. Then the MemoryBuffer::inflate method doesn't know about the constant
folder special value, so it tries to allocate a buffer of that size.
The safest fix for now is to clear the output and return early if all
inputs are single value.
The problem is that this issue might happen for other nodes that also
satisfy the 3 issues above, that is, uses inflate and can be constant
folded. I manually checked for such nodes and there are no nodes that
satisfy all three, so hopefully this is isolated to the Double Edge Mask
node.
Pull Request: https://projects.blender.org/blender/blender/pulls/130163
Mistake in e494a44024
Even though collection/object PointerRNA argument is same in most cases,
use `layer_or_collection_prop` in if() before assigning value to
collection property (`base_or_object_prop`argument value is null in function
call from `_isolate_flag()`).
Pull Request: https://projects.blender.org/blender/blender/pulls/129250
Instead of looping through all drawings in the vertex group kernel
function, loop through all/visible drawings in the operators and
pass a single drawing to `remove_from_vertex_group`.
Continuing !129789.
Co-authored-by: Falk David <falk@blender.org>
Pull Request: https://projects.blender.org/blender/blender/pulls/129890
- Modifications for marker selection to use the same behavior as
key selection in the dopesheet.
- Animation editors now mask out events for the timeline (when there are markers).
- Resolve logical conflicts with the Dope sheet keymap.
- Add Shift Drag & Ctrl Drag for box select & de-select.
Pull Request: https://projects.blender.org/blender/blender/pulls/129841
Add some missing `nullptr` and `ale->type` checks to ensure the
assmptions the rest of the code makes holds.
The direct cause of this issue is likely
822907a68d. The root cause is the hard to
predict behaviour of the animation filtering system combined with a lot
of pointer casts without testing whether the data is actually of the
expected type.
Pull Request: https://projects.blender.org/blender/blender/pulls/130148
This fixes an issue where the outliner would not be immediately
updated after changing the parent of a bone. The issue was simply
that the outliner was not being flagged to update in the callback
function.
Pull Request: https://projects.blender.org/blender/blender/pulls/129933
`set_active_node` call from `grease_pencil_copy_data` resets inactive
layer again as locked before executing the drawing code. Since property
is only operated on layers, it is possible to remove the code from
`set_active_node()` and only have it in `set_active_layer()`.
Pull Request: https://projects.blender.org/blender/blender/pulls/129465
Make use of the struct and patterns from 005e02d008 to make
the view places that deform shape keys more similar. While they each are
slightly different for valid reasons, with this change they follow the
same structure and use the same helper struct.
Pull Request: https://projects.blender.org/blender/blender/pulls/129901
The Merge-Down operator for layers was inserting existing layers first,
then appending the 2 merged layers at the end (top of stack).
Now merged layers are inserted at the position of the layer below the
active, so the layer order remains unchanged.
Note that merging a layer group has a similar issue, with the new
layers getting appended at the top. This is a bit more difficult because
looping only over layers drops the relative ordering of groups and
layers. A separate fix is needed for that.
Pull Request: https://projects.blender.org/blender/blender/pulls/130146
When a layer group was active, new groups would be created at
the top of the layer tree which is inconsistent with how layers are
added.
This fixes the behavior. When a group is active, newly added groups
are moved into that group.
The code that was updating the active node after removing a layer/group
was not handling all cases correctly.
The fix updates the logic for setting the active node:
1. If the node is a non-empty group and we're keeping the children, then set
the active node to the top node in the group.
2. Otherwise, if the node has a node below it (in the same group), set it to be the active one.
3. Otherwise, if the node has a node above it (in the same group), set it to be the active one.
4. Otherwise, if the parent is not the root group, set the parent to be the active node.
5. Otherwise, clear the active node.
This behaves the same as in Krita.
Pull Request: https://projects.blender.org/blender/blender/pulls/130126
The code was relying on the layer cache through `groups_for_write`/`layers_for_write`.
The `group->runtime->layer_group_cache_` stores all the child groups instead
of groups on first level. This leads to crash when recursively deleting
group i.e. group at third level has been deleted but `layer_group_cache_`
of first level group will still point to address of freed memory.
The fix removes the use of the cache and iterates over the direct
children in `group.children` and deletes them recursively.
Pull Request: https://projects.blender.org/blender/blender/pulls/130082
The code was relying on the layer cache through `groups_for_write`/`layers_for_write`.
The `group->runtime->layer_group_cache_` stores all the child groups instead
of groups on first level. This leads to crash when recursively deleting
group i.e. group at third level has been deleted but `layer_group_cache_`
of first level group will still point to address of freed memory.
The fix removes the use of the cache and iterates over the direct
children in `group.children` and deletes them recursively.
Pull Request: https://projects.blender.org/blender/blender/pulls/130082
The PBVH vertex positions accessor functions have to match the logic in
`sculpt_update_object`. When there were topology changing modifiers
before deform modifiers it didn't. `BKE_crazyspace_build_sculpt` just
skips the topology changing modifiers, and the resulting positions
are stored in `ss.deform_cos`, which we need to access here.
Pull Request: https://projects.blender.org/blender/blender/pulls/130139
The Join operator for Grease Pencil was making a shallow copy of the drawings
array, with the assumption that the source object would no longer be needed.
This is not the case, the source object is still in the blend data and still
shares the same drawings pointer as the target object. This is invalid, the
drawings should owned by one object.
To fix this the patch makes a deep copy of the drawings so the source and target
have their own sets of drawings. Data is still shared at the CustomData level,
so these copies are not very expensive.
Pull Request: https://projects.blender.org/blender/blender/pulls/130140
Check the error status from the function that updates the active vertex.
Also add `PROP_HIDDEN` to the operator's properties so the redo panel
doesn't show; that seemed extra weird to display on failure.
Pull Request: https://projects.blender.org/blender/blender/pulls/130142
When the duration of a frame is shorter than the build time specified in
the modifier, build modifier should shorten the build time to allow the
frame to be fully built within the frame duration. This fix made the
timing behaviour the same as GPv2.
Pull Request: https://projects.blender.org/blender/blender/pulls/129894
This moves from `GHash` to the type-safe and better-performing `Map`.
It uses the construct on first use idiom to simplify creation and freeing
of the map, removing the need to register an at-exit callback.
The change was initially motivated by seeing the hash lookups in a profile
of drawing a large node tree. Though my testing wasn't exactly scientific,
I did observe a few percent performance improvement with this change.
Pull Request: https://projects.blender.org/blender/blender/pulls/129954
