| Age | Commit message (Collapse) | Author |
|---|
|
This patch introduces a new `CacheMutex` which makes it easy to implement
lazily computed caches in e.g. `Curves`. For more details see `BLI_cache_mutex.hh`.
Differential Revision: https://developer.blender.org/D16419
|
|
Remove NURBS or Bezier specific attributes after removing points or
curves. In theory we could avoid copying those attributes in the
first place, but that doesn't seem worth the extra complexity here,
since we don't necessarily know the result curve type counts before
copying attributes.
|
|
Attribute copying often uses identical logic for copying selected
elements or copying with an index map. Instead of reimplementing
this in each file, use the common implementation in the array_utils
namespace. This makes the commonality more obvious, gives improved
performance (this implementation is multithreaded), reduces binary
size (I observed a 173KB reduction), and probably reduces compile time.
|
|
String attributes are intentionally not fully supported in geometry nodes
yet because more design work is necessary to decide how they should behave.
For now just disable handling string attributes to avoid crashes.
|
|
This patch contains an initial set of nodes to access basic
mesh topology information, as explored in T100020.
The nodes allow six direct topology mappings for meshes:
- **Corner -> Face** The face a corner is in, the index in the face
- **Vertex -> Edge** Choose an edge attached to the vertex
- **Vertex -> Corner** Choose a corner attached to the vertex
- **Corner -> Edge** The next and previous edge at each face corner
- **Corner -> Vertex** The vertex associated with a corner
- **Corner -> Corner** Offset a corner index within a face
And two new topology mappings for curves:
- **Curve -> Points** Choose a point within a curve
- **Point -> Curve** The curve a point is in, the index in the curve
The idea is that some of the 16 possible mesh mappings are more
important, and that this is a useful set of nodes to start exploring
this area. For mappings with an arbitrary number of connections, we
must sort them and use an index to choose a single element, because
geometry nodes does not support list fields. Note that the sort
index has repeating behavior as it goes over the "Total" number of
connections, and negative sort indices choose from the end.
Currently which of the "start" elements is used is determined by the
field context, so the "Field at Index" and "Interpolate Domain" nodes
will be quite important. Also, currently the "Sort Index" inputs are
clamped to the number of connections.
One important feature that isn't implemented here is using the winding
order for the output elements. This can be a separate mode for some
of these nodes. It will be optional because of the performance impact.
There are several todos for separate commits after this:
- Rename "Control Point Neighbors" to be consistent with this naming
- Version away the "Vertex Neighbors" node which is fully redundant now
- Implement a special case for when no weights are used for performance
- De-duplicating some of the sorting logic between the nodes
- Improve performance and memory use of topology mappings
- Look into caching some of the mappings on meshes
Differential Revision: https://developer.blender.org/D16029
|
|
When adapting the domain of a single value virtual array, skip
allocating an array for the result and just return another single
value. Among other cases, this can help when everything is selected
in sculpt mode, moving domain interpolation from 5% of perf samples
to 0% when sculpting.
|
|
|
|
|
|
When resizing mesh and curves attribute storage, avoid initializing the
new memory for basic types. Also, avoid skipping "no free" layers; all
layers should be reallocated to the new size since they may be accessed.
The semantics introduced in 25237d2625078c6d1 are essential for this
change, because otherwise we don't have a way to construct non-trivial
types in the new memory.
In a basic test of the extrude node, I observed a performance
improvement of about 30%, from 55ms to 42ms.
Differential Revision: https://developer.blender.org/D15818
|
|
|
|
Differential Revision: https://developer.blender.org/D15880
|
|
|
|
Reversing Bezier handle types and positions would skip the middle point
of curves with an odd number of segments, which is still necessary to
swap in order to avoid changing the curve's shape.
|
|
When creating a curves data-block, one is expected to set the new
position values. We can slightly improve performance by avoiding
doing that redundantly.
Similar to cccc6d6905be7ac32cb.
|
|
When allocating new `CustomData` layers, often we do redundant
initialization of arrays. For example, it's common that values are
allocated, set to their default value, and then set to some other
value. This is wasteful, and it negates the benefits of optimizations
to the allocator like D15082. There are two reasons for this. The
first is array-of-structs storage that makes it annoying to initialize
values manually, and the second is confusing options in the Custom Data
API. This patch addresses the latter.
The `CustomData` "alloc type" options are rearranged. Now, besides
the options that use existing layers, there are two remaining:
* `CD_SET_DEFAULT` sets the default value.
* Usually zeroes, but for colors this is white (how it was before).
* Should be used when you add the layer but don't set all values.
* `CD_CONSTRUCT` refers to the "default construct" C++ term.
* Only necessary or defined for non-trivial types like vertex groups.
* Doesn't do anything for trivial types like `int` or `float3`.
* Should be used every other time, when all values will be set.
The attribute API's `AttributeInit` types are updated as well.
To update code, replace `CD_CALLOC` with `CD_SET_DEFAULT` and
`CD_DEFAULT` with `CD_CONSTRUCT`. This doesn't cause any functional
changes yet. Follow-up commits will change to avoid initializing
new layers where the correctness is clear.
Differential Revision: https://developer.blender.org/D15617
|
|
When the curve type attribute doesn't exist, there is no reason to
create an array for it only to fill the default value, which will add
overhead to subsequent "add" operations. I added a "get_if_single"
method to virtual array to simplify this check. Also use the existing
functions for filling curve types.
Differential Revision: https://developer.blender.org/D15560
|
|
|
|
|
|
|
|
The `DefaultMixer` for mixing generic data types has some issues:
1. The full buffer is always zeroed, even if only some is used.
2. Finalizing also works on all values, even if only some are used.
3. "mixing" doesn't allow setting the first value, requiring that
everything is cleared beforehand.
This commit adds the following functionality:
1. Constructor with the specified `IndexMask` for preliminary zeroing.
2. `set` method to overwrite the value.
3. `finalize` with the specified mask to process a subset of values.
This is useful in situations where you want to use the
DefaultMixer without having to overwrite all the values many times.
A performance improvement was observed for NURBS curve evaluation and
attribute interpolation from the point to curve domain of about 15% and
35% respectively (100,000 curves).
Differential Revision: https://developer.blender.org/D15434
|
|
This was removed in cacdea7f4a5b49d to fix a bug, but copying point
and curve attributes should be fine as long as the attribute arrays are
retrieved before-hand.
Differential Revision: https://developer.blender.org/D15541
|
|
`parallel_invoke` allows executing functions on separate threads.
However, creating tasks in tbb has a measurable amount of overhead.
Therefore, it can be benefitial to disable parallelization when
the amount of work done per function is small.
See D15539 for some benchmark results.
Differential Revision: https://developer.blender.org/D15539
|
|
Cyclic curves don't need the tangent correction based on the first
and last handle position.
|
|
Calling two non-const methods on a `MutableAttributeAccessor`
at the same time in multiple threads is not safe.
While I don't know what caused the crash here exactly, I do know
that it happens while looking up the attribute for writing, which
may modify the unterlying geometry. I couldn't reproduce the
bug with a debug build or without threading.
|
|
Use the attribute API instead of the CustomData API, to correctly
handle anonymous attributes and simplify the code. One non-obvious
thing to note is that the type counts are recalculated by the "finish"
function of the `curve_type` attribute, so they don't need to be copied
explicitly. Also, the mutable attribute accessor cannot be an reference
if we want to give it an rvalue, which is convenient in this case.
|
|
These mutable pointers present problems with ownership in relation to
proper copy-on-write for attributes. The simplest solution is to just
remove them and retrieve the layers from `CustomData` when they are
needed. This also removes the complexity and redundancy of having to
update the pointers as the curves change. A similar change will apply
to meshes and point clouds.
One downside of this change is that it makes random access with RNA
slower. However, it's simple to just use the RNA attribute API instead,
which is unaffected. In this patch I updated Cycles to do that. With
the future attribute CoW changes, this generic approach makes sense
because Cycles can just request ownership of the existing arrays.
Differential Revision: https://developer.blender.org/D15486
|
|
* Make the class names more consistent.
* Implement missing move-constructors and assignment-operators.
|
|
The offsets array that encodes the sizes of each curve must be filled
anyway, or the curves will be in an invalid state. Calloc is unnecessary
here. To make that situation clearer, fill the offsets with -1 in debug
builds. Always set the first offset to zero though, since that can save
some boilerplate in other areas.
|
|
|
|
Add a method to remove points from the new curves type, just like
the existing curve removal function. No functional changes are expected.
The code is simpler because all data is just stored as attributes, but
also different because the point data for all curves is stored in the same
arrays.
Similar performance improvements as other commits in T95443 are
expected, expecially for cases where there are many small curves.
Differential Revision: https://developer.blender.org/D15130
|
|
This commit ports the "Set Spline Type" node to the new curves type.
Performance should be improved in similar ways to the other refactors
from the conversion task (T95443). Converting to and from Catmull Rom
curves is now supported. There are a few cases where a lot of work can
be skipped: when the number of points doesn't change, and when the
types already match the goal type.
The refactor has a few other explicit goals as well:
- Don't count on initialization of attribute arrays when they are
first allocated.
- Avoid copying the entire data-block when possible.
- Make decisions about which attributes to copy when changing curves
more obvious.
- Use higher-level methods to copy data between curve points.
- Optimize for the common cases of single types and full selections.
- Process selected curves of the same types in the same loop.
The Bezier to NURBS conversion is written by Piotr Makal (@pmakal).
Differential Revision: https://developer.blender.org/D14769
|
|
Curve attributes were allocated to the size of the point domain, which
wouldn't cause bad behavior, just potentially worse performance.
|
|
This implements the new way to attach curves to a mesh surface using
a uv map (based on the recent discussion in T95776).
The curves data block now not only stores a reference to the surface object
but also a name of a uv map on that object. Having a uv map is optional
for most operations, but it will be required later for animation (when the
curves are supposed to be deformed based on deformation of the surface).
The "Empty Hair" operator in the Add menu sets the uv map name automatically
if possible. It's possible to start working without a uv map and to attach the
curves to a uv map later on. It's also possible to reattach the curves to a new
uv map using the "Curves > Snap to Nearest Surface" operator in curves sculpt
mode.
Note, the implementation to do the reverse lookup from uv to a position on the
surface is trivial and inefficient now. A more efficient data structure will be
implemented separately soon.
Differential Revision: https://developer.blender.org/D15125
|
|
- CustomDataType -> eCustomDataType
- CustomDataMask -> eCustomDataMask
- AttributeDomain -> eAttrDomain
- NamedAttributeUsage -> eNamedAttrUsage
|
|
This commit adds a float selection to curve control points or curves,
a sculpt tool to paint the selection, and uses the selection influence
in the existing sculpt brushes.
The selection is the inverse of the "mask" from mesh sculpt mode
currently. That change is described in more detail here: T97903
Since some sculpt tools are really "per curve" tools, they use the
average point selection of all of their points. The delete brush
considers a curve selected if any of its points have a non-zero
selection.
There is a new option to choose the selection domain, which affects how
painting the selection works. You can also turn the selection off by
clicking on the active domain.
Sculpt brushes can be faster when the selection is small, because
finding selected curves or points is generally faster than the
existing brush intersection and distance checks.
The main limitation currently is that we can't see the selection in the
viewport by default. For now, to see the selection one has to add a
simple material to the curves object as shown in the differential
revision. And one has to switch to Material Preview in the 3d view.
Differential Revision: https://developer.blender.org/D14934
|
|
Replace tot/amount & size with num, in keeping with T85728.
|
|
Follow conventions from T85728.
|
|
When the curve types array isn't allocated, the default type
is Catmull Rom. Because the type counts are calculated eagerly,
they must be in a valid state.
|
|
For example, this can be used to find the indices of all Bezier curves
inside an existing selection. The important part is that it is optimized
for the case when all curves have the same type.
|
|
If all curves are selected for setting the new type,
skip counting the types and just set it directly.
|
|
NURBS curves can be invalid when the order is less than the number
of points, or in a few other situations. Currently the evaluated data of
an invalid NURBS curve is empty. This is inconvenient because it
requires checking for empty curves when it otherwise wouldn't be
necessary. This patch replaces that fallback with copying the original
data to the evaluated points. This makes conceptual sense too, as if
the curve couldn't be evaluated-- which wouldn't necessarily delete it.
Usually the UI protects against this happening, but it's currently
possible to create an invalid curve with some operations like the
delete geometry node.
Differential Revision: https://developer.blender.org/D14837
|
|
Remembering the number of curves of every type makes it fast to know
whether processing specific to a single curve type has to be done.
This information was accessed in quite a few places, so this should be
an overall reduction in overhead for the new curves type.
The cache is computed eagerly, in other words every time after changing
the curve types. In order to reduce verbosity I added helper functions
for some common ways to set the types.
Differential Revision: https://developer.blender.org/D14732
|
|
This commit changes the Curve to Mesh node to work with `Curves`
instead of `CurveEval`. The change ends up basically completely
rewriting the node, since the different attribute storage means that
the decisions made previously don't make much sense anymore.
The main loops are now "for each attribute: for each curve combination"
rather than the other way around, with the goal of taking advantage
of the locality of curve attributes. This improvement is quite
noticeable with many small curves; I measured a 4-5x improvement
(around 4-5s to <1s) when converting millions of curves to tens of
millions of faces. I didn't obverse any change in performance compared
to 3.1 with fewer curves though.
The changes also solve an algorithmic flaw where any interpolated
attributes would be evaluated for every curve combination instead
of just once per curve. This can be a large improvement when there
are many profile curves.
The code relies heavily on a function `foreach_curve_combination`
which calculates some basic information about each combination and
calls a templated function. I made assumptions about unnecessary reads
being removed by compiler optimizations. For further performance
improvements in the future that might be an area to investigate.
Another might be using a "for a group of curves: for each attribute:
for each curve" pattern to increase the locality of memory access.
Differential Revision: https://developer.blender.org/D14642
|
|
The ported normal calculation from ceed37fc5cbb466a0 neglected to
use the tilt attribute to rotate the normals around the tangents.
This commit adds that behavior back, adding a new math header file
to avoid duplicating the rotation function for normalized axes.
Differential Revision: https://developer.blender.org/D14655
|
|
If all of the curves are poly curves, the evaluated positions are the
same as the original positions. In this case just reuse the original
positions span as the evaluated positions.
|
|
Port the "Normal" and "Curve Tangent" nodes to the new curves data-block
to avoid the conversion to `CurveEval`. This should make them faster by
avoiding all that copying, but otherwise nothing else has changed.
This also includes a fix to move the normal mode as a built-in curve
attribute when converting to and from `CurveEval`. The attribute is
needed because the option is used implicitly in many nodes currently.
Differential Revision: https://developer.blender.org/D14609
|
|
|
|
Previously, the new attributes were zero-initialized. However, sometimes
the default has to be something else. The old behavior led to unexpected
behavior in the Snap Curves to Surface operator in Deform mode, when the
curves were not attached to the surface before.
Differential Revision: https://developer.blender.org/D14588
|
|
- Use "curve" instead of "spline" in comments
- Use non-plural variable names
- Tag topology dirty after resolution modified rather than positions
- Reorder enum values to change which value is zero (and the default)
- Remove a duplicate unused variable
|
|
Found in T96889.
|
