If multiple bones have a custom property with the same name,
depsgraph didn't distinguish between them, potentially leading
to spurious cycles.
This patch moves ID_PROPERTY operation nodes for bone custom
properties from the parameters component to individual bone
components, thus decoupling them.
Differential Revision: https://developer.blender.org/D13729
Continuation of the D13404 which finished the design of not having
geometry-level nodes dependent on object-level.
Differential Revision: https://developer.blender.org/D13405
In the past that worked because the `GPUMaterial` referenced the
`ImageUser` from the image node. However, that design was incompatible
with the recent node tree update refactor (rB7e712b2d6a0d257d272e).
Also, in general it is a bad idea to have references between data that is
owned by two different data blocks.
This incompatibility was resolved by copying the image user from the node
to the `GPUMaterial` (rB28df0107d4a8). Unfortunately, eevee depended
on this reference, because the image user on the node was update when the
frame changed. Because the image user was copied, the image user in the
`GPUMaterial` did not receive the frame update anymore.
This frame update is added back by this commit. The main change is that
the image user iterator now also iterates over image users in `GPUMaterial`s
on material and world data blocks. An issue is that these materials don't
exist on the original data blocks and that caused the check in
`build_animation_images` in the depsgraph to give the wrong answer.
Therefore the check is extended.
Right now the check is not optimal, because it results in more depsgraph
nodes than are necessary. This can be improved when it becomes cheaper
to check if a node tree contains any references to a video texture.
The node tree update refactor mentioned before makes it much easier
to construct this kind of run-time data from the bottom up, instead of
scanning the entire node tree recursively every time some information
is needed.
Goals of this refactor:
* More unified approach to updating everything that needs to be updated
after a change in a node tree.
* The updates should happen in the correct order and quadratic or worse
algorithms should be avoided.
* Improve detection of changes to the output to avoid tagging the depsgraph
when it's not necessary.
* Move towards a more declarative style of defining nodes by having a
more centralized update procedure.
The refactor consists of two main parts:
* Node tree tagging and update refactor.
* Generally, when changes are done to a node tree, it is tagged dirty
until a global update function is called that updates everything in
the correct order.
* The tagging is more fine-grained compared to before, to allow for more
precise depsgraph update tagging.
* Depsgraph changes.
* The shading specific depsgraph node for node trees as been removed.
* Instead, there is a new `NTREE_OUTPUT` depsgrap node, which is only
tagged when the output of the node tree changed (e.g. the Group Output
or Material Output node).
* The copy-on-write relation from node trees to the data block they are
embedded in is now non-flushing. This avoids e.g. triggering a material
update after the shader node tree changed in unrelated ways. Instead
the material has a flushing relation to the new `NTREE_OUTPUT` node now.
* The depsgraph no longer reports data block changes through to cycles
through `Depsgraph.updates` when only the node tree changed in ways
that do not affect the output.
Avoiding unnecessary updates seems to work well for geometry nodes and cycles.
The situation is a bit worse when there are drivers on the node tree, but that
could potentially be improved separately in the future.
Avoiding updates in eevee and the compositor is more tricky, but also less urgent.
* Eevee updates are triggered by calling `DRW_notify_view_update` in
`ED_render_view3d_update` indirectly from `DEG_editors_update`.
* Compositor updates are triggered by `ED_node_composite_job` in `node_area_refresh`.
This is triggered by calling `ED_area_tag_refresh` in `node_area_listener`.
Removing updates always has the risk of breaking some dependency that no
one was aware of. It's not unlikely that this will happen here as well. Adding
back missing updates should be quite a bit easier than getting rid of
unnecessary updates though.
Differential Revision: https://developer.blender.org/D13246
- Added space below non doc-string comments to make it clear
these aren't comments for the symbols directly below them.
- Use doxy sections for some headers.
Ref T92709
The are few things in the dependency graph which lead to the issue:
- IDs are only built once.
- Object-data level (Armature, i,e,) builder dependent on the object
visibility.
This caused issues when an armature is first built as not directly
visible (via driver, i.e.) and then was built as a directly visible.
This did not update visibility flag on the node for the custom shape
object.
The idea behind the fix is to go away form passing object visibility
flag to the geometry-level builders and instead rely on the common
visibility flush post-processing to make sure certain objects are
fully visible when needed.
This is the safest minimal part of the change for 3.0 release which
acts as an additional way to ensure visibility. This means that it
might not be a complete fix (if some configuration was overseen) but
it should not make currently working cases to not work.
The fix should also make modifiers used on rigify widgets to work.
The more complete fix will have `is_object_visible` argument removed
from the geometry-level builder functions.
Differential Revision: https://developer.blender.org/D13404
This commit adds a node that generates a text paragraph as curve
instances. The inputs on the node control the overall shape of the
paragraph, and other nodes can be used to move the individual instances
afterwards. To output more than one line, the "Special Characters" node
can be used.
The node outputs instances instead of real geometry so that it doesn't
have to duplicate work for every character afterwards. This is much
more efficient, because all of the curve evaluation and nodes like fill
curve don't have to repeat the same calculation for every instance of
the same character.
In the future, the instances component will support attributes, and the
node can output attribute fields like "Word Index" and "Line Index".
Differential Revision: https://developer.blender.org/D11522
When the same stroke was used as a driver variable, this could make this
stroke already tagged as built in the course of building driver
variables (via `build_gpencil`), but then important stuff from
`build_object_data_geometry_datablock` could be missed later on (because
both of these funtions use `checkIsBuiltAndTag`). Most importantly,
setting up operations such as GEOMETRY_EVAL would be skipped entirely.
`build_object_data_geometry_datablock` seems to cover greasepencil just
fine (does the same as `build_gpencil` and more). Proposed solution is to
remove `build_gpencil` entirely. In `build_id` it would then also call
`build_object_data_geometry_datablock` for `ID_GD` IDs. Now the covered
types that _call_ `build_object_data_geometry_datablock` match exactly
to what is covered _inside_ `build_object_data_geometry_datablock`.
Think this "duplication" of functionality was just overseen in
rB66da2f537ae8 [`build_gpencil` existed long before and said commit made
greasepencil a real object with geometry and such].
thx @JacquesLucke for additional input!
Maniphest Tasks: T88433
Differential Revision: https://developer.blender.org/D12324
We now use a for_each function with callback to iterate through all sequences in the scene.
This has the benefit that we now only loop over the sequences in the scene once.
Before we would loop over them twice and allocate memory to store temporary data.
The allocation of temporary data lead to unintentional memory leaks if the code used returns to exit out of the iteration loop.
The new for_each callback method doesn't allocate any temporary data and only iterates though all sequences once.
Reviewed By: Richard Antalik, Bastien Montagne
Differential Revision: http://developer.blender.org/D12278
This makes the internal naming consistent with the public API. And also gives
us a visibility_flag rather than restrictflag that can be extended with more
flags.
During a mesh transformation in edit mode (Move, Rotate...), only part of
the batch cache needs to be updated.
This commit allows only update only the drawn batches seen in
`BKE_object_data_eval_batch_cache_deform_tag` if the new
`ID_RECALC_GEOMETRY_DEFORM` flag is used.
This new flag is used in the transforms operation for edit-mesh and
results in 1.6x overall speedup in heavy subdiv cube.
Differential Revision: https://developer.blender.org/D11599
Avoid computationally expensive copying operations
when only some settings have been modified.
This is done by adding support for updating parameters
without tagging for copy-on-write.
Currently only mesh data blocks are supported,
other data-blocks can be added individually.
This prepares for changing values such as edit-mesh auto-smooth angle
in edit-mode without duplicating all mesh-data.
The benefit will only be seen when the user interface no longer tags
all ID's for copy on write updates.
ID_RECALC_GEOMETRY_ALL_MODES has been added to support situations
where non edit-mode geometry is modified in edit-mode.
While this isn't something user are likely to do,
Python scripts may change the underlying mesh.
Reviewed By: sergey
Ref D11377
Would crash when adding or removing a collection directly to the master
collection of a scene.
Consequence of change to handle depsgraph-controlled evaluation of some
IDs (like excluded collections which do not get evaluated and do not get
a COW anymore). See rBcf4258673755 and D10907.
Note that this mostly demonstrates once again how weak and flacky our
handling of embedded IDs still remains, with some part of the code
handling them as independent IDs, some as fully local/private data, some
as a mix of both... and lots and lots of custom handling code and corner
cases that are a bottomless pit of issues.
Also quiet incredible that this was not reported already, luckily this
original change did not make it to 2.93 release.
Root of the issue was actually hidden deep in depsgraph itself: it would
not properly update all of its COW IDs using a datablock when depsgraph
decides to evaluate or un-evaluate it.
This would lead to evaluated IDs pointing to either:
- orig IDs when there was an evaluated version of those (annoying bug,
but not a crashing one).
- old address of previously evaluated IDs that no longer exists in the
depsgraph (causing the crash from the report e.g.).
This commit adds an extra step at the end of nodes building, that goes
over all of already existing IDs in the depsgraph to check whether they
do one of the two things above, and tag them for COW update if so.
NOTE: This only affects depsgraph (re-)building, not its evaluation.
This remains consistent with the fact that operations that may change
the depsgraph content (like Collection exclusion etc.) need to trigger a
rebuild.
NOTE: Performances: Worst case scenarii, like (un-)excluding a whole
character collection in a production file, lead to 5% to 10% extra
processing time in depsgraph building. Most of it comming from extra COW
processing (in depsgraph's update in `build_step_finalize`), the detection
loop itself only accounts for 1% to 2% of the whole building time.
Maniphest Tasks: T85752
Differential Revision: https://developer.blender.org/D10907
The sockets are not exposed in any nodes yet.
They work similar to the Object/Collection sockets, which also
just reference a data block.
This is part of D11222.
This fixes T87666 and T83252.
The boolean modifier and geometry nodes can depend on the geometry
of an entire collection. Before, the modifiers had to manually create relations
to all the objects in the collection. This worked for the most part, but was
cumbersome and did not solve all issues. For example, the modifiers were not
properly updated when objects were added/removed from the referenced collection.
This commit introduces the concept of "collection geometry" in the depsgraph.
The geometry of a collection depends on the transforms and geometry of all
the objects in it. The boolean modifier and geometry nodes can now just depend
on the collection geometry instead of creating all the dependencies themselves.
Differential Revision: https://developer.blender.org/D11053
Make the SHADING_PARAMETERS operation on node tree a real no-op.
The function used as a callback was only doing a debug print.
The issue with this function was that an original ID was passed
as an argument. This is a violation of design, and if anything it
should be left up to the implementation to check whether depsgraph
is active.
There are no functional changes for users in the interface. The
only difference is that the debug print will no longer happen.
This change removes copy-on-write operations from ID nodes which do not
need copy-on-write.
Should be no functional changes, as before the copy-on-write operation
would do nothing for those nodes anyway.
Building IDs which are not covered by copy-on-write process was not
implemented, which was causing parameters block not present, and, hence
causing crashes in areas which expected parameters to present.
First part of this change is related on making it so Copy-on-Write is
optional for ID nodes in the dependency graph.
Second part is related on using a generic builder for all ID types
which were not covered by Copy-on-Write before.
The final part is related on making it so build_id() is properly
handling ParticleSettings and Grease Pencil Data. Before they were not
covered there at all, and they need special handling because they do
have own build functions.
Not sure it worth trying to split those parts, as they are related to
each other and are not really possible to be tested standalone. Open
for a second opinion though.
Possible nut-tightening is to re-organize build_id() function so
that every branch does return and have an assert at the end, so that
missing ID type in the switch statement is easier to spot even when
using compilers which do not report missing switch cases.
As for question "why not use default" the answer is: to make it more
explicit and clear what is a decision when adding new ID types. We do
not want to quietly fall-back to a non-copy-on-write case for a newly
added ID types.
Differential Revision: https://developer.blender.org/D10075
If a force field was of type "Texture", any changes of that texture (e.g.
its type - as reported in T85139 - or also its properties) were not
properly updating rigid bodies and particle systems.
Now ensure that texture is actually in the depsgraph and set up relation
accordingly.
Also fixes T75198.
Maniphest Tasks: T85139
Differential Revision: https://developer.blender.org/D10234
This reverts commit 76fd41e9db. This should have
been reverted along with 0f95f51361, since this
change by itself is causing crashes when the depsgraph accesses a non-existent
copy-on-write component.
Ref T84717
Revert "Fix T83411: Crash when using a workspace/layout data path in a driver"
The fix for the crash exposed design violation in the viewport shading updates,
which is for some reason relying on dependency graph tag of interface data.
The viewport module did not respond to the issue in 2 weeks, and the architect
considered missing update for multiple users a more serious issue than a crash
in a very specific case.
This reverts commit 0f95f51361.
This change removes copy-on-write operations from ID nodes which do not
need copy-on-write.
Should be no functional changes, as before the copy-on-write operation
would do nothing for those nodes anyway.
Building IDs which are not covered by copy-on-write process was not
implemented, which was causing parameters block not present, and, hence
causing crashes in areas which expected parameters to present.
First part of this change is related on making it so Copy-on-Write is
optional for ID nodes in the dependency graph.
Second part is related on using a generic builder for all ID types
which were not covered by Copy-on-Write before.
The final part is related on making it so build_id() is properly
handling ParticleSettings and Grease Pencil Data. Before they were not
covered there at all, and they need special handling because they do
have own build functions.
Not sure it worth trying to split those parts, as they are related to
each other and are not really possible to be tested standalone. Open
for a second opinion though.
Possible nut-tightening is to re-organize build_id() function so
that every branch does return and have an assert at the end, so that
missing ID type in the switch statement is easier to spot even when
using compilers which do not report missing switch cases.
As for question "why not use default" the answer is: to make it more
explicit and clear what is a decision when adding new ID types. We do
not want to quietly fall-back to a non-copy-on-write case for a newly
added ID types.
Differential Revision: https://developer.blender.org/D10075
The root of the issue was caused by the dependency graph using ID pointer
to map evaluated state from old depsgraph to new one upon relations update.
This was failing when IDs were re-allocated rapidly: was possible that
Object ID's evaluated state assigned to Mesh and vice versa.
Now depsgraph uses Session UUID to identify which IDs to restore evaluated
state to. The session UUID is stored in the IDNode, so that id_orig is not
dereferenced on depsgraph update since the ID might be freed.
The root of the issue is identified by Campbell, original patch was done
by Bastien, thanks! Also thanks to Oliver and Ray and everyone else for
testing!
The implementation is pretty much the same as for Object sockets.
The socket color is the one that is used for collections in the outliner.
Part of D9739.
Modernize loops by using the `for(type variable : container)` syntax.
Some loops were trivial to fix, whereas others required more attention
to avoid semantic changes. I couldn't address all old-style loops, so
this commit doesn't enable the `modernize-loop-convert` rule.
Although Clang-Tidy's auto-fixer prefers to use `auto` for the loop
variable declaration, I made as many declarations as possible explicit.
To me this increases local readability, as you don't need to fully
understand the container in order to understand the loop variable type.
No functional changes.
Because depsgraph isn't rebuild for animated properies, we have to
assume that active bodies will always want to have updates from the
rigidbody simulation.
We need to have transforms from passive objects if they are animated or
driven by parent relations. This is not immediately obvious as the
object transform matrix will still be available, it is just one frame
behind in some cases.
Fixed dependency cycles if there is a constraint between two rigid
bodies. Because bullet keeps track of its simulated bodies, we do not
need to supply objects transforms as bullet should already have them.
I need combine these two fixes because otherwise we will get depsgraph
warnings that nodes are missing that it expects to be there.
Reviewed By: Sergey, Jacques
Differential Revision: http://developer.blender.org/D8732
