Avoid copying positions and normals from their source arrays. This is
simplified by using separate loops for the original data and accumulate
cases. I observed a performance improvement in the typical benchmark file
of about 13%, from 0.54s to 0.48s for a brush stroke affecting most of a
6 million vertex grid.
Avoid the need to acess mesh data for every node which will become more
important if we decide to make nodes smaller. Also makes some possible
performance improvements simpler.
Part of #118145.
Similar to 851505752f.
This doesn't affect performance much at all, the vast majority of time
in typical scenarios is spent modifying topology anyway. It does significantly
ease development though, because now we can just do slight modifications
on the "new" version of brush code instead also creating a refactored "old"
version.
This fixes a kernel crash on NVidia GP100.
This splits the tilemap finalize shader into another shader
and split the valid tile gather into its own loop.
Simplifying the shader seems to avoid the issue. But the
cause of the issue remains unknown.
Pull Request: https://projects.blender.org/blender/blender/pulls/123850
When preview is downscaled and transformation origin is not the center
of the image, this causes unexpected offset. This happened, because one
matrix combined image downscaling, so it fits into preview and user
defined scale. When origin was not center of the image, this results in
incorrect offset.
Solved by splitting 1 matrix in `sequencer_image_crop_transform_matrix`
into 2 matrices. First matrix just centers and scales image to expected
size. Second matrix performs rest of transform operations. This code is
bit easier to read as well.
Pull Request: https://projects.blender.org/blender/blender/pulls/123776
Delete operator was shared for strips and retiming keys, which was not
ideal. Operator logic relied on invoke function setting proper context
which was also hidden from users.
Split these operators and make dedicated operators for removing retiming
keys - `SEQUENCER_OT_retiming_key_delete`.
This operator returns `OPERATOR_CANCELLED | OPERATOR_PASS_THROUGH` if
executed in non-retiming context, or no key is selected.
Pull Request: https://projects.blender.org/blender/blender/pulls/123767
No need for an additional view item button lookup when the context menu
is built for that button itself. Only do a lookup if the context menu is
built for a button drawn on top of the view item button.
Caused by floating point error in `BLI_rctf_transform_pt_v()`
The error is not fixed, but intead of recalculating each strip offset
with this function, only offset for {0, 0} is calculated and added to
easch strip.
Pull Request: https://projects.blender.org/blender/blender/pulls/123825
So far the brush refactors haven't affected the multires implementations
much. Besides replacing the vertex iteration macro they are fundamentally
the same. This commit refactors the multires implementations to use the
same structure as the base mesh code. The motivations are improving CPU
cache use, removing constant checks from hot loops, and allowing use of
SIMD. This generally works by gathering the AoS multires format into
local arrays.
Quite a few brush evaluation functions have been duplicated to work on
non-indexed data. This will be necessary as we start to handle brushes
that use original data anyway. I don't think this duplication is very
bad though; since the functions are right next to each other it's easy
to change them at the same time. In the future we might investigate
gathering data into local arrays for the mesh implementations too.
In the brush benchmark file from #118145, this improved brush evaluation
performance by roughly 15%, mostly by removing the separate step of
accumulating changes with the proxies system.
Pull Request: https://projects.blender.org/blender/blender/pulls/123703
The code which is responsible for applying data from coordinate grids to
displacement grids could be run as part of modifier evaluation, hence
using the safe version of accessing mesh data was returning a nullptr.
Pull Request: https://projects.blender.org/blender/blender/pulls/123907
The GPU compositor frees some of the cached resources when it gets
canceled during interactive editing, making the experience less smooth.
This is because when the compositor gets canceled mid-evaluation, some
of the operations won't get the chance to mark their used resources as
still in use.
To fix this, we skip the cache manager reset after canceling,
effectively only resetting when a full evaluation happens, giving all
operations the chance to keep their cached resources.
Pull Request: https://projects.blender.org/blender/blender/pulls/123886
Linked images that share the same name as an image in the file will fail
to load in the GPU compositor. That's because cached resources are keyed
using only their ID name, while they should also be keyed by the library
name.
Pull Request: https://projects.blender.org/blender/blender/pulls/123898
Remove the `pbvh_colors.cc` file, since dealing with color attributes should
generally not be the responsibility of the PBVH. Move the color API functions
to the `color` namespace, replacing the two existing `SCULPT_color`
functions. Also replace `SCULPT_has_colors` and `SCULPT_has_loop_colors`
which are redundant with checking the easily accessible attributes directly.
Remove `PBVH::color_layer`, `PBVH::color_domain`, and remove `vcol`, `mcol`,
`vcol_domain`, and `vcol_type` from `SculptSession`. Instead, retrieve attributes
and other arrays from the mesh as necessary and pass them around separately.
This adds more arguments to some functions but it makes their responsibilities
clearer.
Part of #118145.
Use vertex indices instead of `PBVHVertRef`. This doesn't lose functionality
because color attribute painting is only supported for base meshes anyway.
The next steps are refactoring the code to have the same structure as the
new brushes and using the same methods to add code paths for the other
data structures.
This patch adds two theme entries for data before/after the current frame.
Those entries control what color to use when displaying
reference data from a different point in time (motion path, onion skins).
They are under the `3D Viewport` section of the theme in the user preferences.
## User Facing Changes
### Grease Pencil
When `Custom Colors` is disabled it will now use the theme instead of the layer default color.
Currently this setting is enabled by default. Disable it to get Blender to use the theme.
### Motion Path
If Custom Color is disabled, read from the Theme.
### Annotations
If onion skin is enabled, and custom onion skin colors is disabled, read from the theme.
**Grease Pencil**
* previously the color used for onion skinning (without custom colors) was `U.gpencil_new_layer_col`. This is used in other places as well so it cannot be removed with this PR
* By default, custom colors are enabled with Grease Pencil objects. I specifically did **not** change that with this PR. I will instead let the GP module handle that on their terms.
This PR has been discussed in the Grease Pencil module meeting
https://devtalk.blender.org/t/2024-05-21-grease-pencil-module-meeting/34755#meeting-notes-2
and in the A&R module meeting
https://devtalk.blender.org/t/2024-05-24-animation-rigging-module-meeting/34813#patches-review-decision-time-4
Pull Request: https://projects.blender.org/blender/blender/pulls/120558
Keep track of which IDs are animated by which Action Binding. This will
be necessary for display in the Action editor, where animation data that
is unrelated to the active object can be shown (when "show all bindings"
is on).
Note: animation evaluation will not be using this cache, at least not in
the near future. Potentially when we introduce animation-level
constraints this will change, but that's for the future.
The user cache isn't actually used in this commit. It will be used soon
in !122672.
Pull Request: https://projects.blender.org/blender/blender/pulls/123187
Pull Request: https://projects.blender.org/blender/blender/pulls/123187
Add a function to get the `Main *bmain` pointer from a
LibraryForeachIDData struct. Since the struct is opaque to the
`..._foreach_id` functions, a separate function like this is necessary
to get that pointer.
Not used yet, but will be in a followup commit.
Pull Request: https://projects.blender.org/blender/blender/pulls/123187
Pass a non-`nullptr` `bmain` to `BKE_library_foreach_ID_link()` wherever
that `bmain` is known at the call site. This will be used in a followup
commit.
This should not introduce any functional changes.
Pull Request: https://projects.blender.org/blender/blender/pulls/123187
This was the case when the slot index in the active object is greater
than the available slots of other selected objects.
Code was trying to avoid searching all materials by using
`BKE_object_material_get` to get an appropriate index (an addition to
D4441 added in 6b39dc7672). That function has the behavior of clamping
if a target index is greater than the available slots (for good
reasons), so we cant rely on the slot index being the same if
`BKE_object_material_get` finds a material. So in essence, this is not
what we want to use in this case.
Now use the much simpler `BKE_object_material_index_get` to get the
index [this might be less performant in certain scenarios, dont think
this is critical though and correctness should beat performance here]
Pull Request: https://projects.blender.org/blender/blender/pulls/123718
When copying the depth aspect of a depth+stencil image the incorrect
barriers where created. The barrier needs to have the full aspect, even
when we only copy a single aspect from the image.
Pull Request: https://projects.blender.org/blender/blender/pulls/123890
* Better separation between drawing backdrop and main line.
* Pass u and v coordinates of line to fragment shader for further processing.
* Remove `colorGradient` which can also be computed from `lineUV`.
* Simplify drawing potentially more than one parallel line as is done in #112677.
Adding object line art does not work because the wrong source
object (lineart itself is set). `add_type/BKE_object_add_for_data` function
deselects all bases and sets new line art object as active. Get active object
before calling this function to fix this.
Pull Request: https://projects.blender.org/blender/blender/pulls/123836
The goal here is to remove the assumption that multi-inputs will always
have a corresponding node that packs all individual values into a `Vector`
(which is what `LazyFunctionForMultiInput` is doing).
No user visible change expected.
There's no good way to identify items from their base class which is
annoying for development/debugging. I ended up adding a helper like this
a few times, so makes sense to just add this to the API.
