Remove the indirection previously used for the topology refiner
to separate C and C++ code. Instead retrieve the base level in
calling code and call opensubdiv API functions directly. This
avoids copying arrays of mesh indices and should reduce
function call overhead since index retrieval can now be inlined.
It also lets us remove a lot of boilerplate shim code.
The downside is increased need for WITH_OPENSUBDIV defines
in various parts of blenkernel, but I think that is required to avoid
the previous indirection and have the kernel deal with OpenSubdiv
more directly.
Pull Request: https://projects.blender.org/blender/blender/pulls/120825
Memory allocated to vertex buffer but the total verts are zero. This
triggers assert in bind function when `vbo_size_` is zero. Exit out of
funtion early to prevent the assert.
Also wrap `DRW_shgroup_call_no_cull` with if check.
Pull Request: https://projects.blender.org/blender/blender/pulls/128248
The first commit ensures IsectBoxes are not set up unless
they are valid.
The second commit renames
`drw_bounds_are_valid` to `drw_bounds_corners_are_valid`,
and `drw_bounds_culling_enabled` to `drw_bounds_are_valid`
so it's harder to set up an invalid `IsectBox` by mistake.
(Continuation of #127807)
Pull Request: https://projects.blender.org/blender/blender/pulls/128125
Adds antialiasing to curve's handles and thickness to active ones.
Also handles now react to
`Preferences > Interface > Display > Resolution Scale` and
`Preferences > Themes > 3D Viewport > Edge Width` as they do in
legacy curves.
Pull Request: https://projects.blender.org/blender/blender/pulls/122910
Supporting selection operators and overlays in sculpt mode is necessary
to support masking operations.
This enables selection drawing in the overlay in sculpt mode,
and ensures all the necessary operators (pick, lasso, circle, box)
handle the differences in modes correctly.
The selection code so far was expecting object mode, so the `vc->obedit`
object was used without further checks. Now the selection code can be
called outside of edit mode, in which case the `vc->obact` object must
be used instead. Similarly edit mode and sculpt mode have separate
`selectmode` flags, so the correct mode property must be used from the
tool settings depending on the object mode. These changes require some
refactoring of unrelated selection code.
Pull Request: https://projects.blender.org/blender/blender/pulls/128040
Currently the `sum_group_sizes` call used to count the number
of elements in GPU vertex buffers for each PBVH node stands out
in profiles, taking a few percent of the total time when building
PBVH GPU data. Since the number of corners in a node doesn't
change, it's simpler to just store it in the node. We could
eventually cache it somewhere else too, if there was a benefit
to not storing it in the node itself.
The unlit packing could write to the (non-existing) 4th
closure bin inside the gbuffer header leading to no
combined pass shader being invoked for these specific
(complex) materials.
Making sure the unlit workaround writes to bin 0 fixes the
issue.
Pull Request: https://projects.blender.org/blender/blender/pulls/127811
Object with degenerate transform matrix can lead to flat
bounds on GPU. This in turn lead to NaN intersection planes
inside `IsectBox`.
Compute (pseudo) size of matrix and bypass culling is any
axis is too small.
The other part of the patch makes sure that there is a
distinction between disabled culling and invalid
bounding boxes.
Pull Request: https://projects.blender.org/blender/blender/pulls/127807
When the opaque layer was populated with only emissive
material, no raytracing was used for it and no feedback
buffer was needed. Thus, the refraction layer had nothing
to raytrace against.
This fixes it by forcing the use of feedback buffer in this
corner case.
Pull Request: https://projects.blender.org/blender/blender/pulls/127771
These buffers are bound as texture buffers and will have
their backend object created whether or not the subsequent
drawcall is emited.
Ensuring a minimum size fixes the issue.
This uses the path that metal was using.
This doesn't seems to create any difference in render
tests. This simplify the backend code and avoid
specific path for metal.
Idea suggested by Kevin Chuang
Pull Request: https://projects.blender.org/blender/blender/pulls/127687
Updating attributes is now done specifically for each attribute,
and topology and visibility updates tag the draw cache data
explicitly too. That makes checking for updates with these
tags unnecessary.
Similar to the previous commits, previously all GPU buffers
were recreated when only masks changed. Now only
that masks are re-uploaded, using the same mechanism.
Part of #118145.
Similar to the previous commit, previously all GPU buffers
were recreated when only a color attribute changed. Now only
that attribute is re-uploaded, using the same mechanism.
Part of #118145.
Similar to the previous commit, previously all GPU buffers
were recreated when only face sets changed. Now only face
sets are re-uploaded, using the same mechanism.
Part of #118145.
Previously tagging node positions dirty set the `PBVH_UpdateDrawBuffers`
flag which was then used by `node_draw_update_mask` and
`tag_all_attributes_dirty`, which, as hinted by the name, causes a refresh
of all the PBVH GPU vertex buffers, not just positions and normals.
Instead of using that flag, add a function to the draw cache to tag only
position-related data dirty. This should give a performance improvement
when there are also face sets, masks, and/or generic attributes being
extracted for drawing.
Part of $118145.
Add a new function to the draw manager to only
issue a unique resource ID per `ObjectRef`.
This avoids the memory overhead of duplicating
the handle for each overlay.
The handle is stored inside the `ObjectRef` on
first query.
Rel #102179
Fixes Grid2D tiles background and boundaries drawing.
To reproduce switch to mesh edit mode and in UV editing view overlay
options set `Tiles X` or `Y` to more than 1.
Pull Request: https://projects.blender.org/blender/blender/pulls/127516
