Also applies to some other sculpt tools like filter and mask expand.
The full update happens after the paint stroke is finished, so it does not
happen on view navigation, which would cause a delay.
Ref T70295
Differential Revision: https://developer.blender.org/D5922
This improves performance of some sculpt tools, particularly those that modify
many vertices like filter and mask tools, or use brushes with large radius.
For mask expand it can make updates up to 2x faster on heavy meshes, but for
most tools it's more on the order of 1-1.1x. There are bigger bottlenecks to
solve, like normal updates.
Ref T70295
Differential Revision: https://developer.blender.org/D5926
Was caused by DRW_mesh_batch_cache_get_edituv_faces_stretch_area called
after DRW_mesh_batch_cache_create_requested. So it was created on the wrong
object/mesh.
Workaround that does not fix the real issue.
The bug is caused by glBufferData inside
DRW_instance_buffer_finish > GPU_vertbuf_use
but only after the selection code which resets the number of items in
idatalist->pool_buffers.
I don't understand why this is happening as the vbo ids are all valid and
no error is reported. What is even more strange, is that it affects another
vbo which has no connection with the ones in DRW_instance_buffer_finish.
Was caused by some drawcall not being done if the volumetric resolve pass
was drawn (using dual source blending seems to be the cause).
Not sure why this is needed since it is still reset before the next
drawcall.
These were only strictly valid for texture space calculation, don't store them
since they should not be used after that. Only store a flag to indicate if the
auto texture space has been evaluated.
In the future it might make sense to store bounding boxes at the mesh level to
speed up bounding box computation for multiple objects using the same mesh, but
then it will need to be implemented differently.
without a valid target
rB95b9680597f5 introduced code that would skip creation of GPUVertBuf
for ParticlePointCache if the keyed physics would not have a valid
target. Missing vertex buffer would lead to assert/crash.
This code is now removed (dont see a reason why this was done? afaict
2.79 also just displayed the particles without physics in this case and
this seems to be working just fine in 2.8 as well)
part of T69741
Reviewers: fclem
Maniphest Tasks: T69741
Differential Revision: https://developer.blender.org/D5781
Support per-viewport collection visibility options.
Note 1: There is no way to show a collection that was not visible before
due to depsgraph. Otherwise we would risk having all the collections in
the depsgraph and I believe this is not the idea.
An alternative would be to have a new depsgraph for viewports that are
not local. Something to keep in mind if we do per-viewport current frame
in the future.
So for now what we do is to only allow collections visibility to be
disabled/hidden in this mode.
Note 2: hide_viewport (the eye icon) doesn't really matter for
depsgraph. So after the merge we can still ignore it to show the
collections locally in a viewport with no problems for the depsgraph.
Reviewers: brecht, sergey
Subscribers: billreynish
Related task: T61327
Differential Revision: https://developer.blender.org/D5611
Currently unused, but will allow to keep of an owner of the depsgraph.
Could also simplify other APIs in the future by avoiding to pass bmain
explicitly to relation update functions and things like that.
PTCacheEditPoint flag PEP_HIDE was not respected at all...
Reviewers: brecht
Maniphest Tasks: T69432
Differential Revision: https://developer.blender.org/D5739
I believed the crash I experienced happened because:
1. The `extract_pos_nor_init` function is called.
2. Tasks are added to the task pool for `extract_pos_nor`.
3. The tasks begin to be executed while more tasks are added.
4. In some rare cases, all existing tasks are finished, but not all have been added yet.
5. This let the task-counter go down to zero.
6. This triggered a call to `extract_pos_nor_finish`.
7. Then more tasks are added and in the end `extract_pos_nor_finish` is called again.
A solution is to use a task pool that is suspended when created.
Unfortunately, there was an outdated comment, that was probably the root cause of the issue.
Reviewers: fclem, sergey
Differential Revision: https://developer.blender.org/D5680
Support for UV Stretching overlay during multi object editing. The
VBO now holds the ratios per fase. In the shader these ratios will
be compared against the global ratios. The global rations are created
from all selected objects.
The current implementation does not fit well with the draw module. The
plan is to move the drawing of other spaces towards the draw manager what
leads to a better fit. Currently the details on this solution is unclear
but this requirement will become an attentionpoint in the future design.
Reviewed By: fclem
Differential Revision: https://developer.blender.org/D5665
In weightpaint it is possible to enable the bone selection mode. During
drawing the overlay was rendered, but during selection this was ignored.
Users needed to double click in order to select bones even when the overlay
was enabled.
This patch makes bone selection possible during weight painting using the pose mode bone
selection overlay with a single click.
Reviewed By: fclem, campbellbarton
Differential Revision: https://developer.blender.org/D5629
Calculating UV Stretching on large meshes showed garbage. The reason
is that the calculation is not thread save. Temporarily disable
threading for UV Stretching
Before there were two options: Paste to original layer called "Paste" and Paste to active layer called "Paste & Merge"
Now, by default the paste is in active layer and the "Paste & Merge" has been renamed "Paste".
For old "Paste", now is called "Paste by Layer" and it's not the default value anymore.
Note: Minor edits to add icons not present in Differential revision.
Differential Revision: https://developer.blender.org/D5591
The correct fix would be to avoid all those hacks but this is needed if
we want to be able to parallelize object vbo extractions.
This is the fixed version of the hack. The issue was that the ibo ranges
were saved before the batch were reset and the IBO was discarded, leading
to a read after free error. All previous reported crash were tested and
are now not reproductible.
