Cleanup: Reduce use and scope of templates in vertex paint #2

Closed
Hans Goudey wants to merge 10 commits from paint-vertex-fewer-templates into main

When changing the target branch, be careful to rebase the branch in your fork to match. See documentation.
Owner

Many high-level functions with unrelated code were templated based on
the color attribute domain and type. In the end, those were just a few
branches though, similar to other branches. So to reduce binary bloat
and clarify code, move tempates to the point where separate types are
actually needed. Also move constant code out of loops, and use generic
arrays and spans to store some caches. The binary ended up 53 KB
smaller for me, which isn't much but it's in the right direction.

Performance wise nothing really changes. The vast majority of time in
vertex paint is spent on unrelated things like calculating normals or
uploading GPU buffers anyway.

The one ugly part I didn't account for when I started is the casting
between the "ColorGeometry" and "ColorPaint" types. I'm not sure that
this is correct, but it's just a more explicit version of what was
there already.

Many high-level functions with unrelated code were templated based on the color attribute domain and type. In the end, those were just a few branches though, similar to other branches. So to reduce binary bloat and clarify code, move tempates to the point where separate types are actually needed. Also move constant code out of loops, and use generic arrays and spans to store some caches. The binary ended up 53 KB smaller for me, which isn't much but it's in the right direction. Performance wise nothing really changes. The vast majority of time in vertex paint is spent on unrelated things like calculating normals or uploading GPU buffers anyway. The one ugly part I didn't account for when I started is the casting between the "ColorGeometry" and "ColorPaint" types. I'm not sure that this is correct, but it's just a more explicit version of what was there already.
Hans Goudey added 119 commits 2023-07-10 22:43:58 +02:00
In edit mode the uv map data length gets set to zero. The specialized
MLoopUV code used to have a check to detect this when trying to access
the UVs using foreach_get/set . Add this check for the Attribute code
path as well.

Pull Request: blender/blender#109179
Ref #109168
Add an offset indices utility to do fill constant size new offsets in
parallel, which was already done in the duplicate elements node.

For example, filling poly offsets for a new part of a mesh that is only
quads. In the extrude node this was single-threaded before, so the
new poly offsets is about 10x faster, saving about 10 out of 157 ms
when extruding 2 million faces.
a8a454287a which moved edge creases out of `MEdge` only
retrieved the result data array if the mesh had vertex creases. Before
the processing always happened though. So process the result creases
if the input has edge creases too, and try to retrieve an existing array
instead of always adding a new one.
Also remove redundant parenthesis.
To ensure consistency between the UV unwrapping operators, the
`UV > Follow Active Quads` operator will now use 3D selection when
called from the 3D editor, and UV selection when called from the
UV editor.
Improve 3D text selection & feedback when moving the mouse outside of
the text. Previously mouse selection only worked if the mouse was over
the text, now there is more margin to select characters or lines.

Ref !109229

Co-authored-by: Campbell Barton <campbell@blender.org>
Co-authored-by: Harley Acheson <harley.acheson@gmail.com>
The "Align Rotation" operator in the UV editor was producing
unexpected results if an image was displayed with non-square aspect
ratio. This change adds the "Correct Aspect" option to the operator.
The FPS in the viewport is displayed in a string of the form "fps:
%.2f". The string that stores it is of length 16, which is plenty
enough in English. But the translated message in Hindi uses
Devanagari, which takes up 4 bytes per char. Because of this, it is
too long and there is not enough room for the numbers to actually
appear.

To solve this, the containing string length was doubled to 32 chars.

Pull Request: blender/blender#109408
Currently start and end property of action don't have hardmin/hardmax. This results in an
error when exporting action having extreme range values in FBX and gltf formats (#107965).

Pull Request: blender/blender#107985
Only do `next++` if `next` is a non-`NULL` pointer. This doesn't change
the functionality of the code, but it does prevent a sanitizer warning.

No functional changes.
Remove the trailing period from the 'Disable constraint without moving'
operator.

No functional changes.
GHOST_ContextCGL used an incorrect compiler directive. When reading
the comment we could also remove it. In order to do the right thing
(fixing directive vs removing code) we opened this pull request to
ask feedback.

Pull Request: blender/blender#109686
The merging behavior of the Array, Mirror, Screw modifiers has changed
since 4369627e71.

Before, only the customdata of the destination vertex was kept. Source
vertices were completely removed.

With 4369627e71, all vertices in the merge group (whether source or
destination) now have their customdata interpolated.

But this can cause problems for example with the root vertex customdata
of the skin modifier, where if only one of the vertices of the group of
vertices has this flag, the resulting one will also have it.

This commit restores the behavior for the vertices customdata and does
not interpolate it if `do_mix_vert_data` is false.

Pull Request: blender/blender#109627
Using sycl::bitcast instead of union hack
Hierarchy lines (like we also have in the Outliner) make it easier to
visually parse the hierarchy, and avoid confusion about nesting level.
Especially when some items have icons and/or collapse chevrons and some
not (thus different levels of visual indentation).
They were planned for #93582 and #107881, also see
https://code.blender.org/2023/05/the-next-big-step-grease-pencil-3-0/#layer-groups.

The drawing is implemented as a general tree-view feature, so all
tree-views with collapsable items (which excludes the spreadsheet
data-set tree view) will get them without further setup.
Pull Request: blender/blender#109328
No functional changes.

Previously, each navigation operation had its own method for restoring
the 3D View's state upon cancellation.

However, this approach did not provide significant advantages.

The number of affected variables was too small to justify considering
it for performance optimization, and the cancellation code was
duplicated across different operations.

To address this, the present commit unifies all restore code into a
single function:

- `void viewops_data_state_restore(ViewOpsData *vod);`

As a result, the `viewmove_apply_reset` function has been removed.

Moreover, since the state now has a dedicated restore function, it
seemed appropriate to create a function for capturing the state as
well.

Therefore, the code responsible for capturing the state (previously
scattered within `viewops_data_init_navigation`) has been extracted to
create the function:

- `void viewops_data_state_capture(ViewOpsData *vod);`

During this separation process, a potential error in the code was
identified, where the state may not be fully restored.

Although this issue persists, a comment has been added to clarify this
situation.

Pull Request: blender/blender#109760
No user visible changes expected.

Part of #96713, continuation of work started in 249e4df110 and 2e221de4ce.
Refer to these for a motivation and design overview.

Adds a new class for bone elements.

Pull Request: blender/blender#109684
Pull Request: blender/blender#109777
Turns out uint wasn't defined this early in our kernels on Windows.
Using unsigned int instead should fix this.
fixes one uint missed in a0846a60c9
ebe8f8ce71 and 5669c5a61b added multithreading but
a typo meant it wasn't used for half of the process. Fixing that gives a
30% improvement in my tests, from about 80 to 60 ms converting a grid.
Instead of keeping track of a local array of positions in the modifier
stack itself, use the existing edit mode SoA "edit cache" which already
contains a contiguous array of positions. Combined with positions as a
generic attribute, this means the state is contained just in the mesh
(and the geometry set) making the code much easier to follow.

To do this we make more use of the mesh wrapper system, where we can
pass a `Mesh` that's actually stored with a `BMesh` and the extra
cached array of positions. This also resolves some confusion-- it was
weird to have the mesh wrapper system for this purpose but not use it.

Since we always created a wrapped mesh in edit mode, there's no need
for `MOD_deform_mesh_eval_get` at all anymore. That function was quite
confusing with "eval" in its name when it really retrieved the original
mesh.

Many deform modifiers had placeholder edit mode evaluation functions.
Since these didn't do anything and since the priority is node-based
deformation now, I removed these. The case is documented more in the
modifier type struct callbacks.

Pull Request: blender/blender#108637
See #103343
Locked `FCurves` had 2 visual issues

* the dashing is so short it just creates visual noise
* keyframes are drawn in white looking like they are selected

This PR changes this by

* Increasing the dash width
* Keyframes are drawn in black on locked curves
* To indicate that they can't be selected, draw them in a X shape
* To further reduce visual noise, locked curves no longer draw thicker when selected

This is part of the changes discussed in this design task #104867

Pull Request: blender/blender#106052
Split off from [#106952: Animation: Butterworth Smoothing filter](blender/blender#106952)

This patch allows the slider to take a different string than just "%",
the use case is on #106952 where we want to display "Hz" instead.

Additionally the slider got the ability to set modes, which determine
how the factor number is displayed.
`SLIDER_MODE_PERCENT` means it will multiply by 100 and display as whole numbers
`SLIDER_MODE_FLOAT` means it will display just floats with one digit after the comma

Additionally to that, because the slider range is now arbitrary and potentially deals with
large numbers, the mouse distance needed to travel from min to max has been normalized
to the range.

Pull Request: blender/blender#109768
This introduces an alias target `bf::intern::atomic` for
`bf_intern_atomic`. This has the following benefits:

- Any target name with `::` in it will be recognized as an actual
target by cmake, rather than a library name it may not know about.
and will be validated by cmake to exist. Which means if you make
a typo in the LIB section, CMake will error out telling you it
doesn't know about this specific target rather than passing it on
to the build system, where you'll either get build or linker errors
because of said typo.

- Given there is quite a cleanup still to do in the build system,
it won't always be obvious which targets have been updated to
modern targets and which still need to be done. Having a namespaced
target name is a good indicator there.

Pull Request: blender/blender#109784
All probes (including the world background probe) are stored in a single texture. Each probe
can be of any resolution as long as it is a power of 2 and not larger than 2048. So valid options
are (2048x2048, 1024x1024, 512x512, etc).

Each probe can be stored in their own resolution and can be set by the user.
> NOTE: Eventually we would like to add automatic resolution selection.

The probes are packed in an 2d texture array with the dimension of 2048*2048. The number of
layers depends on the actual needed layers. If more layers are needed the texture will be recreated.
This can happen when a new reflection probe is added, or an existing reflection probe is made visible
to the scene or its resolution is changed.

### Octahedral mapping

Probes are rendered into a cubemap. To reduce memory needs and improve sampling performance the cubemap
is stored in octahedral mapping space. This is done in `eevee_reflection_probe_remap_comp.glsl`.

The regular octahedral mapping has been extended to fix leakages at the edges of the texture
and to be able to be used on an atlas texture and by sampling the texture once.

To reduce sampling cost and improve the quality we add an border around the
octahedral map and extend the octahedral coordinates. This also allows us to
generate lower resolution mipmaps of the atlas texture using 2x2 box filtering
from a higher resolution.

### Subdivisions and areas

Probes data are stored besides the texture. The data is used to find out where the probe is stored
in the texture. It is also used to find free space to store new probes.

This approach ensures that we can be flexible at storing probes with different
resolutions on the same layer. Lets see an example how that works

Code-wise this is implemented by `ProbeLocationFinder`. ProbeLocationFinder can view a texture in a
given subdivision level and mark areas that are covered by probes. When finding a free spot it returns
the first empty area.

**Notes**

* Currently the cubemap is rendered with a fixed resolution and mipmaps are generated in order to
  increase the quality of the atlas. Eventually we should use dynamic resolution and no mipmaps.
  This will be done as part of the light probe baking change.

Pull Request: blender/blender#109688
Using " instead of ' caused a compiler warning
The correct conversion factor should be 1 / (pi * area), see #108505

Co-authored-by: Brecht Van Lommel <brecht@blender.org>
Pull Request: blender/blender#109153
The spotlight is now treated as a sphere instead of a view-aligned disk.
The implementation remains almost identical to that of a point light,
except for the spotlight attenuation and spot blend. There is no
attenuation inside the sphere. Ref #108505

Other changes include:
## Sampling
Instead of sampling the disk area, the new implementation samples either
the cone of the visible portion on the sphere or the spread cone, based
on which cone has a smaller solid angle. This reduces noise when the
spotlight has a large radius and a small spread angle.
| Before | After  |
|   --   |   --   |
|![spot_size_before.png](/attachments/04ea864a-6bf9-40fe-b11b-61c838ae70cf)|![spot_size_after.png](/attachments/7077eaf9-b7a8-41b1-a8b6-aabf1eadb4f4)
## Texture
Spot light can now project texture using UV coordinates.
<video src="/attachments/6db989d2-7a3c-4b41-9340-f5690d48c4fb"
title="spot_light_texture.mp4" controls></video>
## Normalization
Previously, the normalization factor for the spotlight was \(\pi r^2\),
the area of a disk. This factor has been adjusted to \(4\pi r^2\) to
account for the surface area of a sphere. This change also affects point
light since they share the same kernel type.
## Versioning
Some pipeline uses the `Normal` socket of the Texture Coordinate node for
projection, because `ls->Ng` was set to the incoming direction at the
current shading point. Now that `ls->Ng` corresponds to the normal
direction of a point on the sphere (except when the radius is zero),
we replace these nodes with a combination of the Geometry shader node
and the Vector Transform node, which gives the same result as before.
![versioning.png](/attachments/5bbfcacc-26c5-4f7f-8360-c42bcd851f68)
Example file see https://archive.blender.org/developer/T93676

Pull Request: blender/blender#109329
This fixes the issue described in blender/blender#108957.

Instead of modeling distant lights like a disk light at infinity, it models them as cones. This way, the radiance is constant across the entire range of directions that it covers.

For smaller angles, the difference is very subtle, but for very large angles it becomes obvious (here's the file from #108957, the angle is 179°):
| Old | New |
| - | - |
| ![old_bigsun.png](/attachments/4ef8e7a7-1a29-4bdf-a74c-3cfa103bf1e7) | ![new_bigsun.png](/attachments/d53c7749-2672-40b6-9048-ccf2fffceeb7) |

One notable detail is the sampling method: Using `sample_uniform_cone` can increase noise, since the sampling method no longer preserves the stratification of the samples. This is visible in the "light tree multi distant" test scene.
Turns out we can do better, and after a bit of testing I found a way to adapt the concentric Shirley mapping to uniform cone sampling. I hope the comment explains the logic behind it reasonably well.

Here's the result, note that even the noise distribution is the same when using the new sampling:
| Method | Old | New, basic sampling | New, concentric sampling |
| - | - |- | - |
| Image | ![old.png](/attachments/b3258a70-f015-4065-a774-193974cce439) | ![new_basic.png](/attachments/a9008576-0af6-4152-a687-c800fd958bbd) | ![new_concentric.png](/attachments/769b6c43-34bc-434e-a4fd-ce69addd1ba5) |
| Render time (at higher spp)| 9.03sec | 8.79sec | 8.96sec |

I'm not sure if I got the `light->normalized` handling right, since I don't really know what the expectation from Hydra is here.

Co-authored-by: Weizhen Huang <weizhen@blender.org>
Pull Request: blender/blender#108996
Minor tweaks to the API to support drag & drop management in the layer tree UI.

These changes were needed for #109826:
- Add `TreeNode.parent_group()`
- Add `LayerGroup.as_node()`
- Use `TreeNode` instead of `Layer` whenever a link is needed to link or unlink a layer.

Pull Request: blender/blender#109824
* Use pi factor to convert between radiant flux and intensity
* Mark lights as normalized on export
* Add spot light export support
* Add treatAsPoint support for import and export
* Empirically match normalized distant light
* Fix wrong unnormalized point/sphere/disk light unit in Cycles

Overall it should be much closer now for all light types. Point and distant
light units are inconsistent between renderers, so not possible to match
everything there.

Ref #109404

Pull Request: blender/blender#109795
So that there is some lighting when there are no lights in the scene,
and black when there are lights. This matches the behavior of other
Hydra renderers.

Ref #96731
`list_assert_duplicates` validates a list that should contain no
duplicates, contains no duplicates, with scope keywords now being
allowed in `INC` sections, there is the situation where multiple
include paths can have the same scope set on them causing
`list_assert_duplicates` to error out.

To remedy this we remove the scope keywords from the list first,
before running the test.
Ref #108505

Pull Request: blender/blender#109831
Change storage to consistently put xyz + radius in the motion blur attribute.

Pull Request: blender/blender#109830
Navigation operations, like those of the trackpad, are not modal and
therefore are confirmed on each call.

To prevent the transform operations from being updated in this case,
add a fake navigation flag.

The removal of this flag is postponed to the next call.
* Renamed BKE_pbvh_raycast_project_ray_root to
          BKE_pbvh_clip_ray_ortho for greater
	  clarity.
* BKE_pbvh_clip_ray_ortho no longer strictly clips
  within the input ray interval.  This is not necassary
  for orthographic views and was too prone to floating
  point error.  The function is only called to clip
  brush rays for orthographic views so this is acceptable.
The Alembic importer can optionnaly interpolate vertex and matrix data.
To detect if two samples can be interpolated `get_weight_and_index`
needs to be called which output the interpolation weight and the indices
for the floor and ceil samples separately. Either the weight or the
indices could be used to determine if interpolation was needed.

This adds a `SampleInterpolationSettings` structure to hold the weight
and indices together so we don't need to manage multiple local
variables, and replaces `get_weight_and_index` with
`get_sample_interpolation_settings` which returns either a
`SampleInterpolationSettings` or nothing if no interpolation is
necessary.

This also modifies `AbcMeshData` to have an optional
`SampleInterpolationSettings` and removes members used for interpolation
from `CDStreamConfig`, which simplifies the latter structure as well as
the check in `read_mverts` to verify if interpolation is needed.

`get_config` also no longer needs a parameter for setting the now removed
`use_vertex_interpolation` member from `CDStreamConfig`. This was only
used for Mesh vertex interpolation despite also being set in the points
reader (which does not yet support any interpolation).

No functional changes.

Pull Request: #109155
When blender is not focused and a selection is executed
with the mouse, since there is no conversion from `wl_fixed` to `int`,
the bounds of the selection can cause the selection box to be too large,
causing `draw_select_framebuffer_depth_only_setup` to fail when create
`g_select_buffer.texture_depth`.

Ref !109834
- BKE_action_frame_range_get  (was BKE_action_get_frame_range)
- BKE_action_frame_range_calc (was calc_action_range)
- BKE_action_has_motion       (was action_has_motion)
Use eAction_TransformFlags typed return value.
This is a bit more flexible and future proof than CustomData.
Avoid showing many obscure CMake options by default.
The file is easier to read and change when less context is needed at
each step. Also extract some complex-looking flag testing into a
separate function, and move some constant checks out of loops.
There is only one error value, prefer a specific check.
Once the channel was found the for-loop kept searching.
While in practice this didn't cause problems,
it's misleading & unnecessary.
This PR fixes a compiler issue when using NVIDIA. It was introduced
when adding the reflection probes. NVIDIA compiler doesn't like
to access elements in a bound resource and pass it directly
as a parameter. It should first be added as a local variable.

Pull Request: blender/blender#109891
Originally was noticed when adding drivers to a rigid body., but
it could potentially happen with any configuration.

The reason for the crash was that the ID which was modified was
not tagged as such.

Modifying drivers from the interface are likely tagging for updates
from the operator. This change makes it so the python function also
does tagging.

It is not really how one would design the system nowadays, but it
is how the Blender historically handles such cases. A bigger refactor
is possible to move tags to the places where modification actually
happens, but it seems to be a better idea to tackle it as a separate
project which will be considered no-functional-changes.

Pull Request: blender/blender#109895
At a glance, it can be hard to see if a Shape Key has been muted or not.
This change greys out all the UI elements, instead of just the current value number.
All the functionality still works the same, purely a visual change.

Pull Request: blender/blender#109857
When using many reflection probes with different subdivisions
it can happen that the reflection probe uses texture space that
is already been allocated for another probe.

The cause for that is that negative shift operations result into
undefined behavior, this was fixed previously by clamping the
number of bit shifts to perform, but ignoring to shift into the
other direction

This change also includes `ProbeLocationFinder.print_debug`
for future debugging.

Pull Request: blender/blender#109823
The WITH_VULKAN_BACKEND/WITH_METAL_BACKEND should be included per
library. However when doing this we used an none existing compile
directive for WITH_METAL_BACKEND. This wasn't detected as
WITH_METAL_BACKEND is also defined globally.

Pull Request: blender/blender#109896
`use_ambient_occlusion` is a legacy property and not
connected to anything. The clip editor still write
to it for enabling the ambient occlusion renderpass.

This change will remove the legaccy option and
won't enable ambient occlusion pass anymore due
to visual and performance impact.

Pull Request: blender/blender#109783
Using area-preserving mapping from cone to disk. Has somewhat distortion
near 90°.
The texture rotates with the transformation of the light object, can
have negative and non-uniform scaling.

Pull Request: blender/blender#109842
No user visible changes expected.

Part of #96713, continuation of work started in 249e4df110 and 2e221de4ce.
Refer to these for a motivation and design overview.

Adds a new class for edit bone elements.

Pull Request: blender/blender#109793
No reason to skip these, and this will be useful for faster shader node
preview rendering based on AOVs.

Pull Request: blender/blender#109709
Also fix some incorrect usages of `N_` macro instead of `TIP_` one
(these error messages typically need to get translated explicitely, not
only marked for extraction).
The "Single file with all the views" message was not extracted because
it is an enum item that was not exposed to the RNA, but only
selectable in some cases depending on the
`RNA_def_property_enum_funcs()`.

This commit changes the `RNA_def_property_enum_items()` so that this
enum item is exposed to the RNA.

Pull Request: blender/blender#109876
For overlapping regions, we currently offset the edge for resizing
regions a bit towards the contents. This makes sense for example in the
sidebar, where the region background is fully transparent by default,
and users expect the edge to be where the visible contents begin. This
was an issue in #104831 for two reasons:
- Although the region overlaps, we decided to make it fully opaque by
  default. The offset is still applied, making the edge detection feel
  broken.
- The offset would also be applied to header-like regions, where there
  is not much padding between the edge and the content. So the edge to
  resize the region would overlap buttons.

2 changes to address this:
- Only apply the offset if the region background is transparent-ish,
  with a rather arbitrary alpha threshold.
- Do not apply the offset for header-like regions.

Pull Request: blender/blender#109753
All files that use it are in C++ now.
When switching shading between material preview and solid modes, the
world reflective light could become uninitialized. The reason was that
the world was only updated when they actually changed. When switching
it might not be the case that the world changed and the world probe
wasn't uploaded to the texture.

This fix will reduce the locations where this information was stored
(removing the `do_world_update` attribute in `ReflectionProbeModule`).
It also doesn't reset the `do_render` flag during syncing, but post-
pone it to the actual drawing.

Pull Request: blender/blender#109901
Slight speed up of face topology building for blur node.
Avoid using increment and replacing by adding size.
Multi-threading for accumulating offsets. Early returns
in one place places. In a test with a large grid, the node
became 36% faster, with 13 out of 40 ms saved from
the topology building.

Pull Request: blender/blender#109764
Clay strips was using it's own brush local matrix,
which wasn't quite compatible with texture matrices.
This could lead to brush textures not lining up with
the stroke dabs.

There was also a bug where the stroke was starting
20 pixels into the stroke, which is much higher than
necassary to derive the initial rake angle.

Notes:
* The clay strips brush now uses SCULPT_cube_tip_init
  to calculate the local brush matrix.
* SCULPT_cube_tip_init now accepts custom brush location
  and radius arguments.
* The mouse sample preroll used to calculate initial brush
  rotation angle is now smaller than the update interval.
* Clay strips now supports tip_scale_x, which has also
  been added to DNA defaults.
- Deduplicate code;
- Support no customdata interpolation for edges;
- Reduce number of required elements in context arrays.

With 74772c6920, the merge by distance code for meshes now supports
optional interpolation of custom data for vertices.

As this can be advantageous for performance and memory, it seems
convenient to support the same for edges (and in the future polygons
and corners).

So this commit reworks the code that finds the edge groups to match the
same used for vertices and thus deduplicating and simplifying the code.

Conveniently the `edge_ctx_map` array is no longer needed and can be
removed to save memory.

However it was necessary to create another array (which is usually
smaller) called `double_edges` (to match `double_verts`).

---

**Results:**

The profiling result depends on some factors such as:
- interpolate or not interpolate customdata;
- affect a large or small portion of the mesh;

The best case is when the customdata **is not** interpolated and the operation affects **large** portions of the mesh:

| Before:    | After:     | Factor: |
| ---------- | ---------- | ------- |
| 1256.61 ms | 1073.218 ms| -14.44% |

The worst case is when the customdata **is not** interpolated and the operation affects **small** portions of the mesh:

| Before:   | After:    | Factor: |
| --------- | --------- | ------- |
| 1077.50 ms| 1086.7 ms| +0.85%  |

Pull Request: blender/blender#109836
This commit implements the `VIEW3D_OT_ndof_orbit`,
`VIEW3D_OT_ndof_orbit_zoom`, `VIEW3D_OT_ndof_pan` and
`VIEW3D_OT_ndof_all` operators to the `ED_view3d_navigation_`
utility so that these operations can be used in the transform
operators.

Pull Request: blender/blender#109792
There's quite a few libraries that depend on dna_type_offsets.h
but had gotten to it by just adding the folder that contains it to
their includes INC section without declaring a dependency to
bf_dna in the LIB section.

which occasionally lead to the lib building before bf_dna and the
header being missing, while this generally gets fixed in CMake by
adding bf_dna to the LIB section of the lib, however until last
week all libraries in the LIB section were linked as INTERFACE so
adding it in there did not resolve the build issue.

To make things still build, we sprinkled add_dependencies wherever
we needed it to force a build order.

This diff :

Declares public include folders for the bf_dna target so there's
no more fudging the INC section required to get to them.

Removes all dna related paths from the INC section for all
libraries.

Adds an alias target bf:dna to signify it has been updated to
modern cmake

Declares a dependency on bf::dna for all libraries that require it

Removes (almost) all calls to add_dependencies for bf_dna

Future work:

Because of the manual dependency management that was done, there is
now some "clutter" with libs depending on bf_dna that realistically
don't. Example bf_intern_opencolorio itself has no dependency on
bf_dna at all, doesn't need it, doesn't use it. However the
dna include folder had been added to it in the past since bf_blenlib
uses dna headers in some of its public headers and
bf_intern_opencolorio does use those blenlib headers.

Given bf_blenlib now correctly declares the dependency on bf_dna
as public bf_intern_opencolorio will get the dna header directory
automatically from CMake, hence some cleanup could be done for
bf_intern_opencolorio

Because 99% of the changes in this diff have been automated, this diff
does not seek to address these issues as there is no easy way to
determine why a certain dependency is in place. A developer will have
to make a pass a this at some later point in time. As I'd rather not
mix automated and manual labour.

There are a few libraries that could not be automatically processed
(ie bf_blendthumb) that also will need this manual look-over.

Pull Request: blender/blender#109835
Logic in FileBrowser and/or `BKE_blendfile_library_path_explode`
probably changed at some point, and the generic 'invalid filetype' error
message was reached before the path is compared to current blendfile path.
The "No Cache" simulation nodes option effectively changes the cache
to work in a "realtime mode" where there are only two states, the
current and previous frame. Whenever the current frame doesn't
increase, the previous state should reset. This didn't happen
properly, and it was hard to verify because the code was shared
with the regular "cache on" mode.

Instead, separate the caching more in the code, using a different
struct to store the two "realtime" states. Also clarify that we
don't support animation of the "No Cache" option by disabling
support for that in RNA.

Pull Request: blender/blender#109741
This isn't really helpful at this point, and is ugly/distracting when
starting Blender. It's not that bad if more menus show up with a
bit of a delay anyway.
Doing so avoids having duplicated logic for working with pixel
data which is being passed throughout the render pipeline.

Notable changes:

- ImBug can now store GPU texture.
  This is not very finished part of the API, which will be
  worked further to support tiling for very-high-res images.

- Implicit sharing is removed from the image buffer, as it is
  no longer needed.

There should be no functional changes on user level with this
change.

Ref #108618

Pull Request: blender/blender#109788
Although we had only 190 developers, the script
was accusing a total of 365 developers (with 188 inactive).

There were two problems here:
* The gitea API was only using the `limit` paramater when an explicit
  page was passed.
* There was a bug with the get_next_page code, that would skip the last
  page.

I now made the page always an explicit argument passed to fetch_single.

With this fix we now get: 190 developers with 98 inactive.
The Voronoi distance output is clamped at 8, which is apparent for distance
metrics like Minkowski with low exponents.

This patch fixes that by setting the initial distance of the search loop to
FLT_MAX instead of 8. And for the Smooth variant of F1, the "h" parameter is set
to 1 for the first iteration using a signal value, effectively ignoring the
initial distance and using the computed distance at the first iteration instead.

Pull Request: blender/blender#109286
Also remove unnecessary namespace specification.
The `BKE_bvhtree_from_pointcloud_get` function have requirements for
input point cloud argument and initialization of `BVHTreeFromPointCloud`
can be skipped. Due to `BVHTreeFromPointCloud` is not initialized by
default constructor, it can contains garbage data. To check if tree is
initialized field of `BVHTreeFromPointCloud`, return argument shouldn't
be ignored. `[[nodiscard]]` attributes is added.

Pull Request: blender/blender#109892
Socket type drawing of custom type in view template.

Pull Request: blender/blender#109871
Hans Goudey closed this pull request 2023-07-10 22:44:29 +02:00
Hans Goudey referenced this issue from a commit 2024-01-21 20:00:11 +01:00

Pull request closed

Sign in to join this conversation.
No reviewers
No Label
No Milestone
No Assignees
1 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: HooglyBoogly/blender#2
No description provided.