This shared state between original data and depsgraphs was added in
98a0bcd425. Other physics systems also share
the pointcache, but not the simulation state to this extent, which leads
to this kind of crash.
The mutex lock is not a great solution, you don't really want both render and viewport to be filling the same cache in parallel. However
this kind of problem also exists in other physics systems, and solving
that is certainly beyond the scope of 4.4, and probably needs to wait
for a bigger physics rewrite. In general the recommendation is to bake
everything before rendering.
Pull Request: https://projects.blender.org/blender/blender/pulls/134779
Move `Library.runtime` to be a pointer, move the related
`LibraryRuntime` struct to `BKE_library.hh`. Similar to e.g.
Mesh.runtime, that pointer is expected to always be valid, and is
allocated at readtime or when creating a new Library ID.
Related smaller changes:
* Write code now uses standard ID writing codepath for Library IDs too.
* Runtime pointer is reset to nullptr before writing.
* Looking up a library by its absolute path is now handled through a
dedicated utils, `search_filepath_abs`, instead of using
`BLI_findstring`.
Pull Request: https://projects.blender.org/blender/blender/pulls/134188
This caused build errors on the docs builder, I can't seem to reproduce
locally, so revert for now and have another look at some point in the
future.
Sadly as these changes usually go, this took 5c515e26bb and
2f0fc7fc9f with it as well.
Pull Request: https://projects.blender.org/blender/blender/pulls/132559
Avoid rebuilding BVH trees when meshes are copied.
Similar to the other uses of the shared cache system,
this can arbitrarily improve performance when meshes
are copied but not deformed and BVH building is the
main bottleneck. In a simple test file I got a 6x speedup.
The amount of code is also reduced and the system is
much simpler overall-- built out of common threading
patterns like `SharedCache` with its double-checked lock.
RAII is used in a few places to simplify memory management
too.
The downside is storing more `SharedCache` items in the
mesh runtime struct. That has a slight cost when copying
a small mesh many times, but we have ideas to improve that
in the future anyway (#104327).
Pull Request: https://projects.blender.org/blender/blender/pulls/130865
Restore overlay depth drawing which was removed in [0].
For viewport navigation this is useful. Disable overlays
for other operations such as painting & object placement.
Also disable color drawing in the overlay_next engine drawing
when drawing depth, since it caused the viewport to show a
checkered backdrop.
Details:
A boolean argument has been added to ED_view3d_depth_override instead
of new enums in eV3DDepthOverrideMode since mixing object-filtering
and draw-type in the one enum gets verbose & awkward, where most
existing enums would have needed to include NO_OVERLAY in their name.
V3D_DEPTH_NO_OVERLAYS has been renamed to V3D_DEPTH_ALL,
the caller must pass in use_overlay as false.
[0]: 5fea1eda36
In lieu of fixing the root cause, this commit requires the object to be
in object mode to duplicate particle systems, which works around the
issue because the problem is with particle edit data. This is a pragmatic
choice to focus development efforts on replacing the particle system.
Operations on collections could tag the same collection multiple times.
Track the changed state instead. Also rename (ok/updated to "changed"),
for clarity & consistency.
Remove duplicate depsgraph tagging from ED_rigidbody_object_remove too.
This commit moves generated `RNA_blender.h`, `RNA_prototype.h` and
`RNA_blender_cpp.h` headers to become C++ header files.
It also removes the now useless `RNA_EXTERN_C` defines, and just
directly use the `extern` keyword. We do not need anymore `extern "C"`
declarations here.
Pull Request: https://projects.blender.org/blender/blender/pulls/124469
The attribute API defined in `attribute.cc` was dependent on
the assumption that `ID`s are always the "direct" owners of attributes.
For Grease Pencil drawings, this is not the case. The Grease Pencil ID
stores the attributes for layers, and the attributes for drawings are stored
in `CurvesGeometry` on the drawings themselves.
In order to make use of `rna_attribute.cc`, we need that API to handle
other types of attribute owners.
This adds an `AttributeOwner` which is basically just a type and a
pointer. We replace the `ID` pointers and pass `AttributeOwner`s instead.
For cases where we have to do a switch based on the type, all the
types are handled and the `default` statment is left out. This ensures
that we get a compiler warning when a new `AttributeOwnerType`
is added.
No functional changes expected.
Pull Request: https://projects.blender.org/blender/blender/pulls/122765
Add new ID_IS_EDITABLE macro that checks if the ID can be edited in the
user interface. Replace usage of ID_IS_LINKED where it is used with this
meaning.
Also add a corresponding ID.is_editable property for Python.
This prepares for the ability to edit some linked datablocks for brush
assets.
Pull Request: https://projects.blender.org/blender/blender/pulls/121838
Move the public functions from the editors/object (`ED_object.hh`)
header to the `blender::ed::object` namespace, and move all of the
implementation files to the namespace too. This provides better code
completion, makes it easier to use other C++ code, removes unnecessary
redundancy and verbosity from local uses of public functions, and more
cleanly separates different modules.
See the diff in `ED_object.hh` for the main renaming changes.
Pull Request: https://projects.blender.org/blender/blender/pulls/119947
Keep each function's declaration in the header file associated with its
module. Arguably mode switching should be more organized, but for now
it's better to just declare functions in more predictable places.
"Own" (the adjective) cannot be used on its own. It should be combined
with something like "its own", "our own", "her own", or "the object's own".
It also isn't used separately to mean something like "separate".
Also, "its own" is correct instead of "it's own" which is a misues of the verb.
The depsgraph CoW mechanism is a bit of a misnomer. It creates an
evaluated copy for data-blocks regardless of whether the copy will
actually be written to. The point is to have physical separation between
original and evaluated data. This is in contrast to the commonly used
performance improvement of keeping a user count and copying data
implicitly when it needs to be changed. In Blender code we call this
"implicit sharing" instead. Importantly, the dependency graph has no
idea about the _actual_ CoW behavior in Blender.
Renaming this functionality in the despgraph removes some of the
confusion that comes up when talking about this, and will hopefully
make the depsgraph less confusing to understand initially too. Wording
like "the evaluated copy" (as opposed to the original data-block) has
also become common anyway.
Pull Request: https://projects.blender.org/blender/blender/pulls/118338
The `object_to_world` and `world_to_object` matrices are set during
depsgraph evaluation, calculated from the object's animated location,
rotation, scale, parenting, and constraints. It's confusing and
unnecessary to store them with the original data in DNA.
This commit moves them to `ObjectRuntime` and moves the matrices to
use the C++ `float4x4` type, giving the potential for simplified code
using the C++ abstractions. The matrices are accessible with functions
on `Object` directly since they are used so commonly. Though for write
access, directly using the runtime struct is necessary.
The inverse `world_to_object` matrix is often calculated before it's
used, even though it's calculated as part of depsgraph evaluation.
Long term we might not want to store this in `ObjectRuntime` at all,
and just calculate it on demand. Or at least we should remove the
redundant calculations. That should be done separately though.
Pull Request: https://projects.blender.org/blender/blender/pulls/118210
This adds a new `ModifierData.persistent_uid` integer property with the following properties:
* It's unique within the object.
* Match between the original and evaluated object.
* Stable across Blender sessions.
* Stable across renames and reorderings of modifiers.
Potential use-cases:
* Everywhere where we currently use the name as identifier. For example,
`ModifierComputeContext` and `ModifierViewerPathElem`.
* Can be used as part of a key in `IDCacheKey` to support caches that stay
in-tact across undo steps.
* Can be stored in the `SpaceNode` to identify the modifier whose geometry node
tree is currently pinned (this could use the name currently, but that hasn't been
implemented yet).
This new identifier has some overlap with `ModifierData.session_uid`, but there
are some differences:
* `session_uid` is unique within the entire Blender session (except for duplicates
between the original and evaluated data blocks).
* `session_uid` is not stable across Blender sessions.
Especially due to the first difference, it's not immediately obvious that the new
`persistent_uid` can fulfill all use-cases of the existing `session_uid`. Nevertheless,
this seems likely and will be cleaned up separately.
Unfortunately, there is not a single place where modifiers are added to objects currently.
Therefore, there are quite a few places that need to ensure valid identifiers. I tried to catch
all the places, but it's hard to be sure. Therefore, I added an assert in `object_copy_data`
that checks if all identifiers are valid. This way, we should be notified relatively quickly if
issues are caused by invalid identifiers.
Pull Request: https://projects.blender.org/blender/blender/pulls/117347
