Most of our field inputs are currently specific to geometry. This patch introduces
a new `GeometryFieldInput` that reduces the overhead of adding new geometry
field input.
Differential Revision: https://developer.blender.org/D13489
For some underlying data (e.g. spans) we had two virtual array
implementations. One for the mutable and one for the immutable
case. Now that most code does not deal with the virtual array
implementations directly anymore (since rBrBd4c868da9f97a),
we can get away with sharing one implementation for both cases.
This means that we have to do a `const_cast` in a few places, but
this is an implementation detail that does not leak into "user code"
(only when explicitly casting a `VArrayImpl` to a `VMutableArrayImpl`,
which should happen nowhere).
Previously, there was a fixed grain size for all multi-functions. That was
not sufficient because some functions could benefit a lot from smaller
grain sizes.
This refactors adds a new `MultiFunction::call_auto` method which has the
same effect as just calling `MultiFunction::call` but additionally figures
out how to execute the specific multi-function efficiently. It determines
a good grain size and decides whether the mask indices should be shifted
or not.
Most multi-function evaluations benefit from this, but medium sized work
loads (1000 - 50000 elements) benefit from it the most. Especially when
expensive multi-functions (e.g. noise) is involved. This is because for
smaller work loads, threading is rarely used and for larger work loads
threading worked fine before already.
With this patch, multi-functions can specify execution hints, that allow
the caller to execute it most efficiently. These execution hints still
have to be added to more functions.
Some performance measurements of a field evaluation involving noise and
math nodes, ordered by the number of elements being evaluated:
```
1,000,000: 133 ms -> 120 ms
100,000: 30 ms -> 18 ms
10,000: 20 ms -> 2.7 ms
1,000: 4 ms -> 0.5 ms
100: 0.5 ms -> 0.4 ms
```
Under some circumstances that can lead to more than a 2x
performance increase, because math nodes can better optimize
for the case when the slice is a single value or span.
Currently the geometry nodes evaluator always stores a field for every
type that supports it, even if it is just a single value. This results in a lot
of overhead when there are many sockets that just contain a single
value, which is often the case.
This introduces a new `ValueOrField<T>` type that is used by the geometry
nodes evaluator. Now a field will only be created when it is actually
necessary. See D13307 for more details. In extrem cases this can speed
up the evaluation 2-3x (those cases are probably never hit in practice
though, but it's good to get rid of unnecessary overhead nevertheless).
Differential Revision: https://developer.blender.org/D13307
The idea behind this change is the same as in
rB6ee2abde82ef121cd6e927995053ac33afdbb438.
A `MultiFunction::debug_parameter_name` method could be
added separately when necessary.
Previously, the function names were stored in `std::string` and were often
created dynamically (especially when the function just output a constant).
This resulted in a lot of overhead.
Now the function name is just a `const char *` that should be statically
allocated. This is good enough for the majority of cases. If a multi-function
needs a more dynamic name, it can override the `MultiFunction::debug_name`
method.
In my test file with >400,000 simple math nodes, the execution time improves from
3s to 1s.
Goals of this refactor:
* Simplify creating virtual arrays.
* Simplify passing virtual arrays around.
* Simplify converting between typed and generic virtual arrays.
* Reduce memory allocations.
As a quick reminder, a virtual arrays is a data structure that behaves like an
array (i.e. it can be accessed using an index). However, it may not actually
be stored as array internally. The two most important implementations
of virtual arrays are those that correspond to an actual plain array and those
that have the same value for every index. However, many more
implementations exist for various reasons (interfacing with legacy attributes,
unified iterator over all points in multiple splines, ...).
With this refactor the core types (`VArray`, `GVArray`, `VMutableArray` and
`GVMutableArray`) can be used like "normal values". They typically live
on the stack. Before, they were usually inside a `std::unique_ptr`. This makes
passing them around much easier. Creation of new virtual arrays is also
much simpler now due to some constructors. Memory allocations are
reduced by making use of small object optimization inside the core types.
Previously, `VArray` was a class with virtual methods that had to be overridden
to change the behavior of a the virtual array. Now,`VArray` has a fixed size
and has no virtual methods. Instead it contains a `VArrayImpl` that is
similar to the old `VArray`. `VArrayImpl` should rarely ever be used directly,
unless a new virtual array implementation is added.
To support the small object optimization for many `VArrayImpl` classes,
a new `blender::Any` type is added. It is similar to `std::any` with two
additional features. It has an adjustable inline buffer size and alignment.
The inline buffer size of `std::any` can't be relied on and is usually too
small for our use case here. Furthermore, `blender::Any` can store
additional user-defined type information without increasing the
stack size.
Differential Revision: https://developer.blender.org/D12986
Previously, the computed value passed into the data socket could depend
on the actual field a bit. However, given that the link is marked as invalid
in the ui, the user should not depend on this behavior.
Using a default value is consistent with other cases when there are
invalid links.
In order to address feedback that the "Stable ID" was not easy enough
to use, remove the "Stable ID" output from the distribution node and
the input from the instance on points node. Instead, the nodes write
or read a builtin named attribute called `id`. In the future we may
add more attributes like `edge_id` and `face_id`.
The downside is that more behavior is invisible, which is les
expected now that most attributes are passed around with node links.
This behavior will have to be explained in the manual.
The random value node's "ID" input that had an implicit index input
is converted to a special implicit input that uses the `id` attribute
if possible, but otherwise defaults to the index. There is no way to
tell in the UI which it uses, except by knowing that rule and checking
in the spreadsheet for the id attribute.
Because it isn't always possible to create stable randomness, this
attribute does not always exist, and it will be possible to remove it
when we have the attribute remove node back, to improve performance.
Differential Revision: https://developer.blender.org/D12903
We do this in other nodes to reduce overhead of using the same node more
than once. I don't think it will make a difference with index nodes
currently, but at least it's consistent.
Previously, some multi-functions were allocated in a resource scope.
This was fine as long as the multi-functions were only needed during
the current evaluation of the node tree. However, now cases arise
that require the multi-functions to be alive after the modifier is finished.
For example, we want to evaluate fields created with geometry nodes
outside of geometry nodes.
To make this work, `std::shared_ptr` has to be used in a few more places.
Realistically, this shouldn't have a noticable impact on performance.
If this does become a bottleneck in the future, we can think about ways
to make this work without using `shared_ptr` for multi-functions that
are only used once.
Multi-functions are not allowed to throw exceptions that are not
caught in the same multi-function. Previously, it was difficult to
backtrack a crash to a previously thrown exception.
This adds a new `ParallelMultiFunction` which wraps another multi-function
and evaluates it with multiple threads. The speeds up field evaluation
quite a bit (the effect is most noticeable when the number of evaluations
and the field is large).
There are still other single-threaded performance bottlenecks in field
evaluation that will need to be solved separately. Most notably here
is the process of copying the computed data into the position attribute
in the Set Position node.
Differential Revision: https://developer.blender.org/D12457
Previously, a debug name had to be passed to all methods
that added a resource to the `ResourceScope`. The idea was
that this would make it easier to find certain bugs. In reality
I never found this to be useful, and it was mostly annoying.
The thing is, something that is in a resource scope never leaks
(unless the resource scope is not destructed of course).
Removing the name parameter makes the structure easier to use.
Sometimes not all outputs of a multi-function are required by the
caller. In those cases it would be a waste of compute resources
to calculate the unused values anyway. Now, the caller of a
multi-function can specify when a specific output is not used.
The called function can check if an output is unused and may
ignore it. Multi-functions can still computed unused outputs as
before if they don't want to check if a specific output is unused.
The multi-function procedure system has been updated to support
ignored outputs in call instructions. An ignored output just has no
variable assigned to it.
The field system has been updated to generate a multi-function
procedure where unused outputs are ignored.
Since fields were committed to master, socket inspection did
not work correctly for all socket types anymore. Now the same
functionality as before is back. Furthermore, fields that depend
on some input will now show the inputs in the socket inspection.
I added support for evaluating constant fields more immediately.
This has the benefit that the same constant field is not evaluated
more than once. It also helps with making the field independent
of the multi-functions that it uses. We might still want to change
the ownership handling for the multi-functions of nodes a bit,
but that can be done separately.
Differential Revision: https://developer.blender.org/D12444
Now an instruction knows the cursors where it is inserted instead
of just the instruction that references it. This has two benefits:
* An instruction knows when it is the entry instruction.
* The cursor can contain more information, e.g. if it is linked to the
true or false branch of a branch instruction.
This also simplifies updating the procedure in future optimization
passes.
This implements the initial core framework for fields and anonymous
attributes (also see T91274).
The new functionality is hidden behind the "Geometry Nodes Fields"
feature flag. When enabled in the user preferences, the following
new nodes become available: `Position`, `Index`, `Normal`,
`Set Position` and `Attribute Capture`.
Socket inspection has not been updated to work with fields yet.
Besides these changes at the user level, this patch contains the
ground work for:
* building and evaluating fields at run-time (`FN_fields.hh`) and
* creating and accessing anonymous attributes on geometry
(`BKE_anonymous_attribute.h`).
For evaluating fields we use a new so called multi-function procedure
(`FN_multi_function_procedure.hh`). It allows composing multi-functions
in arbitrary ways and supports efficient evaluation as is required by
fields. See `FN_multi_function_procedure.hh` for more details on how
this evaluation mechanism can be used.
A new `AttributeIDRef` has been added which allows handling named
and anonymous attributes in the same way in many places.
Hans and I worked on this patch together.
Differential Revision: https://developer.blender.org/D12414
The multi-function network system was able to compose multiple
multi-functions into a new one and to evaluate that efficiently.
This functionality was heavily used by the particle nodes prototype
a year ago. However, since then we only used multi-functions
without the need to compose them in geometry nodes.
The upcoming "fields" in geometry nodes will need a way to
compose multi-functions again. Unfortunately, the code removed
in this commit was not ideal for this different kind of function
composition. I've been working on an alternative that will be added
separately when it becomes needed.
I've had to update all the function nodes, because their interface
depended on the multi-function network data structure a bit.
The actual multi-function implementations are still the same though.
`CPPType` can wrap any C++ type so that code can work
with the wrapped type in a generic way. The goal of subclassing
`CPPType` is to provide additional methods for some types.
For example, the `CPPType` for `Array<int>` could have a `.element_type()`
method that returns the `CPPType` for `int`.
* Reduce code duplication.
* Give methods more standardized names (e.g. `move_to_initialized` -> `move_assign`).
* Support wrapping arbitrary C++ types, even those that e.g. are not copyable.
Colors are often thought of as being 4 values that make up that can make any color.
But that is of course too limited. In C we didn’t spend time to annotate what we meant
when using colors.
Recently `BLI_color.hh` was made to facilitate color structures in CPP. CPP has possibilities to
enforce annotating structures during compilation and can adds conversions between them using
function overloading and explicit constructors.
The storage structs can hold 4 channels (r, g, b and a).
Usage:
Convert a theme byte color to a linearrgb premultiplied.
```
ColorTheme4b theme_color;
ColorSceneLinear4f<eAlpha::Premultiplied> linearrgb_color =
BLI_color_convert_to_scene_linear(theme_color).premultiply_alpha();
```
The API is structured to make most use of inlining. Most notable are space
conversions done via `BLI_color_convert_to*` functions.
- Conversions between spaces (theme <=> scene linear) should always be done by
invoking the `BLI_color_convert_to*` methods.
- Encoding colors (compressing to store colors inside a less precision storage)
should be done by invoking the `encode` and `decode` methods.
- Changing alpha association should be done by invoking `premultiply_alpha` or
`unpremultiply_alpha` methods.
# Encoding.
Color encoding is used to store colors with less precision as in using `uint8_t` in
stead of `float`. This encoding is supported for `eSpace::SceneLinear`.
To make this clear to the developer the `eSpace::SceneLinearByteEncoded`
space is added.
# Precision
Colors can be stored using `uint8_t` or `float` colors. The conversion
between the two precisions are available as methods. (`to_4b` and
`to_4f`).
# Alpha conversion
Alpha conversion is only supported in SceneLinear space.
Extending:
- This file can be extended with `ColorHex/Hsl/Hsv` for different representations
of rgb based colors. `ColorHsl4f<eSpace::SceneLinear, eAlpha::Premultiplied>`
- Add non RGB spaces/storages ColorXyz.
Reviewed By: JacquesLucke, brecht
Differential Revision: https://developer.blender.org/D10978
Colors are often thought of as being 4 values that make up that can make any color.
But that is of course too limited. In C we didn’t spend time to annotate what we meant
when using colors.
Recently `BLI_color.hh` was made to facilitate color structures in CPP. CPP has possibilities to
enforce annotating structures during compilation and can adds conversions between them using
function overloading and explicit constructors.
The storage structs can hold 4 channels (r, g, b and a).
Usage:
Convert a theme byte color to a linearrgb premultiplied.
```
ColorTheme4b theme_color;
ColorSceneLinear4f<eAlpha::Premultiplied> linearrgb_color =
BLI_color_convert_to_scene_linear(theme_color).premultiply_alpha();
```
The API is structured to make most use of inlining. Most notable are space
conversions done via `BLI_color_convert_to*` functions.
- Conversions between spaces (theme <=> scene linear) should always be done by
invoking the `BLI_color_convert_to*` methods.
- Encoding colors (compressing to store colors inside a less precision storage)
should be done by invoking the `encode` and `decode` methods.
- Changing alpha association should be done by invoking `premultiply_alpha` or
`unpremultiply_alpha` methods.
# Encoding.
Color encoding is used to store colors with less precision as in using `uint8_t` in
stead of `float`. This encoding is supported for `eSpace::SceneLinear`.
To make this clear to the developer the `eSpace::SceneLinearByteEncoded`
space is added.
# Precision
Colors can be stored using `uint8_t` or `float` colors. The conversion
between the two precisions are available as methods. (`to_4b` and
`to_4f`).
# Alpha conversion
Alpha conversion is only supported in SceneLinear space.
Extending:
- This file can be extended with `ColorHex/Hsl/Hsv` for different representations
of rgb based colors. `ColorHsl4f<eSpace::SceneLinear, eAlpha::Premultiplied>`
- Add non RGB spaces/storages ColorXyz.
Reviewed By: JacquesLucke, brecht
Differential Revision: https://developer.blender.org/D10978
This is very similar to rB5613c61275fe6 and rB0061150e4c90d, basically
just exposing a `VMutableArray` method to its generic counterpart. This
is quite important for curve point attributes to avoid a lookup for
every point when there are multiple splines.
Creating a shallow copy is sometimes useful to get a unique ptr
for a virtual array when one only has a reference. It shouldn't
be used usually, but sometimes its the fastest way to do correct
ownership handling.
Sometimes functions expect a span instead of a virtual array.
If the virtual array is a span internally already, great. But if it is
not (e.g. the position attribute on a mesh), the elements have
to be copied over to a span.
This patch makes the copying process more efficient by giving
the compiler more opportunity for optimization.
This adds support for mutable virtual arrays and provides many utilities
for creating virtual arrays for various kinds of data. This commit is
preparation for D10994.
Previously, the signature of a `MultiFunction` was always embedded into the function.
There are two issues with that. First, `MFSignature` is relatively large, because it contains
multiple strings and vectors. Secondly, constructing it can add overhead that should not
be necessary, because often the same signature can be reused.
The solution is to only keep a pointer to a signature in `MultiFunction` that is set during
construction. Child classes are responsible for making sure that the signature lives
long enough. In most cases, the signature is either embedded into the child class or
it is allocated statically (and is only created once).
When a function is executed for many elements (e.g. per point) it is often the case
that some parameters are different for every element and other parameters are
the same (there are some more less common cases). To simplify writing such
functions one can use a "virtual array". This is a data structure that has a value
for every index, but might not be stored as an actual array internally. Instead, it
might be just a single value or is computed on the fly. There are various tradeoffs
involved when using this data structure which are mentioned in `BLI_virtual_array.hh`.
It is called "virtual", because it uses inheritance and virtual methods.
Furthermore, there is a new virtual vector array data structure, which is an array
of vectors. Both these types have corresponding generic variants, which can be used
when the data type is not known at compile time. This is typically the case when
building a somewhat generic execution system. The function system used these virtual
data structures before, but now they are more versatile.
I've done this refactor in preparation for the attribute processor and other features of
geometry nodes. I moved the typed virtual arrays to blenlib, so that they can be used
independent of the function system.
One open question for me is whether all the generic data structures (and `CPPType`)
should be moved to blenlib as well. They are well isolated and don't really contain
any business logic. That can be done later if necessary.
