A mesh wrapper might be being accessed for read-only from one thread
while another thread converts the wrapper type to something else.
The proposes solution is to defer assignment of the mesh wrapper
type until the wrapper is fully converted. The good thing about this
approach is that it does not introduce extra synchronization (and,
potentially, evaluation pipeline stalls). The downside is that it
might not work with all possible wrapper types in the future. If a
wrapper type which does not clear data separation is ever added in
the future we will re-consider the threading safety then.
Unfortunately, some changes outside of the mesh wrapper file are
to be made to allow "incremental" construction of the mesh prior
changing its wrapper type.
Unfortunately, there is no simplified file which demonstrates the
issue. It was investigated using Heist production file checked at
the revision 1228: `pro/lib/char/einar/einar.shading.blend`. The
repro case is simple: tab into edit mode, possibly few times.
The gist is that there several surface deform and shrinkwrap
modifiers which uses the same target. While one of them is building
BVH tree (which changes wrapper type) the other one accesses it for
read-only via `BKE_mesh_wrapper_vert_coords_copy_with_mat4()`.
Differential Revision: https://developer.blender.org/D15424
Curves that are attached to a surface can now follow the surface when
it is modified using shape keys or modifiers (but not when the original
surface is deformed in edit or sculpt mode).
The surface is allowed to be changed in any way that keeps uv maps
intact. So deformation is allowed, but also some topology changes like
subdivision.
The following features are added:
* A new `Deform Curves on Surface` node, which deforms curves with
attachment information based on the surface object and uv map set
in the properties panel.
* A new `Add Rest Position` checkbox in the shape keys panel. When checked,
a new `rest_position` vector attribute is added to the mesh before shape
keys and modifiers are applied. This is necessary to support proper
deformation of the curves, but can also be used for other purposes.
* The `Add > Curve > Empty Hair` operator now sets up a simple geometry
nodes setup that deforms the hair. It also makes sure that the rest
position attribute is added to the surface.
* A new `Object (Attach Curves to Surface)` operator in the `Set Parent To`
(ctrl+P) menu, which attaches existing curves to the surface and sets the
surface object as parent.
Limitations:
* Sculpting the procedurally deformed curves will be implemented separately.
* The `Deform Curves on Surface` node is not generic and can only be used
for one specific purpose currently. We plan to generalize this more in the
future by adding support by exposing more inputs and/or by turning it into
a node group.
Differential Revision: https://developer.blender.org/D14864
It's helpful to make the separation of legacy data formats explicit,
because it declutters actively changed code and makes it clear which
areas do not follow Blender's current design. In this case I separated
the `MFace`/"tessface" conversion code into a separate blenkernel
.cc file and header. This also makes refactoring to remove these
functions simpler because they're easier to find.
In the future, conversions to the `MLoopUV` type and `MVert`
can be implemented here for the same reasons (see T95965).
Differential Revision: https://developer.blender.org/D15396
We store various lazily calculated caches on meshes, some of which
depend on the vertex positions staying the same. The current API to
invalidate these caches is a bit confusing. With an explicit set of
functions modeled after the functions in `BKE_node_tree_update.h`,
it becomes clear which function to call. This may become more
important if more lazy caches are added in the future.
Differential Revision: https://developer.blender.org/D14760
Add a property to the **Shade Smooth** operator to quickly enable the Mesh `use_auto_smooth` option.
The `Angle` property is exposed in the **Adjust Last Operation** panel to make it easy to tweak on multiple objects without having to go to the Properties editor.
The operator is exposed in the `Object` menu and `Object Context Menu`.
=== Demo ===
{F13066173, size=full}
Regarding the implementation, there are multiple ways to go about this (like making a whole new operator altogether), but I think a property is the cleanest/simplest.
I imagine there are simpler ways to achieve this without duplicating the `use_auto_smooth` property in the operator itself (getting it from the Mesh props?), but I couldn't find other operators doing something similar.
Reviewed By: #modeling, mont29
Differential Revision: https://developer.blender.org/D14894
Support merging UV's that share the same vertex and are very close when
applying modifiers.
This is needed to prevent UV's becoming "detached" which can happen when
applying the subdivision surface modifier.
This regression was caused by [0] which removed selection threshold for
nearby coordinates. While restoring the UV selection threshold could be
done - some selection operations that walk around connected UV fans
wouldn't behave in a deterministic way (such as select shortest path).
There are also other cases where UV's may be compared without a
threshold such as tangent calculation and exporters which have their own
logic to handling UV's.
Also resolves T86896, T89903.
[0]: b88dd3b8e7
Reviewed By: sergey
Ref D14841
A mistake in the mesh normal refactor caused the wrong mesh to
be used when calculating normals with a shape key's deformation.
This commit fixes the normal calculation by using the correct mesh,
with just adjusted vertex positions, and calculating the results
directly into the result arrays when possible. This completely avoids
the need to make a local copy of the mesh, which makes sense,
since the only thing that changes is the vertex positions.
Differential Revision: https://developer.blender.org/D14317
In some cases, the normal edit modifier calculated the normals on one
mesh with the "ensure" functions, then copied the mesh and retrieved
the layers "for write" on the copy. Since 59343ee162, normal
layers are never copied, and normals are allocated with malloc instead
of calloc, so the mutable memory was uninitialized.
Fix by calculating normals on the correct mesh, and also add a warning
to the "for write" functions in the header.
Currently, when normals are calculated for a const mesh, a custom data
layer might be added if it doesn't already exist. Adding a custom data
layer to a mesh is not thread-safe, so this can be a problem in some
situations.
This commit moves derived mesh normals for polygons and
vertices out of `CustomData` to `Mesh_Runtime`. Most of the
hard work for this was already done by rBcfa53e0fbeed7178.
Some changes to logic elsewhere are necessary/helpful:
- No need to call both `BKE_mesh_runtime_clear_cache` and
`BKE_mesh_normals_tag_dirty`, since the former also does the latter.
- Cleanup/simplify mesh conversion and copying since normals are
handled with other runtime data.
Storing these normals like other runtime data clarifies their status
as derived data, meaning custom data moves more towards storing
original/editable data. This means normals won't automatically benefit
from the planned copy-on-write refactor (T95845), so it will have to be
added manually like for the other runtime data.
Differential Revision: https://developer.blender.org/D14154
This removes manual handling of normals that was hard-coded
to false in the one place the function was called. This change
will help to make a fix to T95839 simpler.
Currently, whenever any BMesh is converted to a Mesh (except for edit
mode switching), original index (`CD_ORIGINDEX`) layers are added.
This is incorrect, because many operations just convert some Mesh into
a BMesh and then back, but they shouldn't make any assumption about
where their input mesh came from. It might even come from a primitive
in geometry nodes, where there are no original indices at all.
Conceptually, mesh original indices should be filled by the modifier
stack when first creating the evaluated mesh. So that's where they're
moved in this patch. A separate function now fills the indices with their
default (0,1,2,3...) values. The way the mesh wrapper system defers
the BMesh to Mesh conversion makes this a bit less obvious though.
The old behavior is incorrect, but it's also slower, because three
arrays the size of the mesh's vertices, edges, and faces had to be
allocated and filled during the BMesh to Mesh conversion, which just
ends up putting more pressure on the cache. In the many cases where
original indices aren't used, I measured an **8% speedup** for the
conversion (from 76.5ms to 70.7ms).
Generally there is an assumption that BMesh is "original" and Mesh is
"evaluated". After this patch, that assumption isn't quite as strong,
but it still exists for two reasons. First, original indices are added
whenever converting a BMesh "wrapper" to a Mesh. Second, original
indices are not added to the BMesh at the beginning of evaluation,
which assumes that every BMesh in the viewport is original and doesn't
need the mapping.
Differential Revision: https://developer.blender.org/D14018
Use a shorter/simpler license convention, stops the header taking so
much space.
Follow the SPDX license specification: https://spdx.org/licenses
- C/C++/objc/objc++
- Python
- Shell Scripts
- CMake, GNUmakefile
While most of the source tree has been included
- `./extern/` was left out.
- `./intern/cycles` & `./intern/atomic` are also excluded because they
use different header conventions.
doc/license/SPDX-license-identifiers.txt has been added to list SPDX all
used identifiers.
See P2788 for the script that automated these edits.
Reviewed By: brecht, mont29, sergey
Ref D14069
The evaluated mesh is a result of evaluated modifiers, and referencing
other evaluated IDs such as materials.
It can not be stored in the EditMesh structure which is intended to be
re-used by many areas. Such sharing was causing ownership errors causing
bugs like
T93855: Cycles crash with edit mode and simultaneous viewport and final render
The proposed solution is to store the evaluated edit mesh and its cage in
the object's runtime field. The motivation goes as following:
- It allows to avoid ownership problems like the ones in the linked report.
- Object level is chosen over mesh level is because the evaluated mesh
is affected by modifiers, which are on the object level.
This patch allows to have modifier stack of an object which shares mesh with
an object which is in edit mode to be properly taken into account (before
the change the modifier stack from the active object will be used for all
objects which share the mesh).
There is a change in the way how copy-on-write is handled in the edit mode to
allow proper state update when changing active scene (or having two windows
with different scenes). Previously, the copt-on-write would have been ignored
by skipping tagging CoW component. Now it is ignored from within the CoW
operation callback. This allows to update edit pointers for objects which are
not from the current depsgraph and where the edit_mesh was never assigned in
the case when the depsgraph was evaluated prior the active depsgraph.
There is no user level changes changes expected with the CoW handling changes:
should not affect on neither performance, nor memory consumption.
Tested scenarios:
- Various modifiers configurations of objects sharing mesh and be part of the
same scene.
- Steps from the reports: T93855, T82952, T77359
This also fixes T76609, T72733 and perhaps other reports.
Differential Revision: https://developer.blender.org/D13824
- Add BM_mesh_debug_print & BM_mesh_debug_info.
- Report flags in Mesh.cd_flag in BKE_mesh_debug_print
- Move custom data printing into customdata.cc (noted as a TODO).
Note that the term "runtime" has been removed from
`BKE_mesh_runtime_debug_print` since these are useful for debugging any
kind of mesh data.
As described in T91186, this commit moves mesh vertex normals into a
contiguous array of float vectors in a custom data layer, how face
normals are currently stored.
The main interface is documented in `BKE_mesh.h`. Vertex and face
normals are now calculated on-demand and cached, retrieved with an
"ensure" function. Since the logical state of a mesh is now "has
normals when necessary", they can be retrieved from a `const` mesh.
The goal is to use on-demand calculation for all derived data, but
leave room for eager calculation for performance purposes (modifier
evaluation is threaded, but viewport data generation is not).
**Benefits**
This moves us closer to a SoA approach rather than the current AoS
paradigm. Accessing a contiguous `float3` is much more efficient than
retrieving data from a larger struct. The memory requirements for
accessing only normals or vertex locations are smaller, and at the
cost of more memory usage for just normals, they now don't have to
be converted between float and short, which also simplifies code
In the future, the remaining items can be removed from `MVert`,
leaving only `float3`, which has similar benefits (see T93602).
Removing the combination of derived and original data makes it
conceptually simpler to only calculate normals when necessary.
This is especially important now that we have more opportunities
for temporary meshes in geometry nodes.
**Performance**
In addition to the theoretical future performance improvements by
making `MVert == float3`, I've done some basic performance testing
on this patch directly. The data is fairly rough, but it gives an idea
about where things stand generally.
- Mesh line primitive 4m Verts: 1.16x faster (36 -> 31 ms),
showing that accessing just `MVert` is now more efficient.
- Spring Splash Screen: 1.03-1.06 -> 1.06-1.11 FPS, a very slight
change that at least shows there is no regression.
- Sprite Fright Snail Smoosh: 3.30-3.40 -> 3.42-3.50 FPS, a small
but observable speedup.
- Set Position Node with Scaled Normal: 1.36x faster (53 -> 39 ms),
shows that using normals in geometry nodes is faster.
- Normal Calculation 1.6m Vert Cube: 1.19x faster (25 -> 21 ms),
shows that calculating normals is slightly faster now.
- File Size of 1.6m Vert Cube: 1.03x smaller (214.7 -> 208.4 MB),
Normals are not saved in files, which can help with large meshes.
As for memory usage, it may be slightly more in some cases, but
I didn't observe any difference in the production files I tested.
**Tests**
Some modifiers and cycles test results need to be updated with this
commit, for two reasons:
- The subdivision surface modifier is not responsible for calculating
normals anymore. In master, the modifier creates different normals
than the result of the `Mesh` normal calculation, so this is a bug
fix.
- There are small differences in the results of some modifiers that
use normals because they are not converted to and from `short`
anymore.
**Future improvements**
- Remove `ModifierTypeInfo::dependsOnNormals`. Code in each modifier
already retrieves normals if they are needed anyway.
- Copy normals as part of a better CoW system for attributes.
- Make more areas use lazy instead of eager normal calculation.
- Remove `BKE_mesh_normals_tag_dirty` in more places since that is
now the default state of a new mesh.
- Possibly apply a similar change to derived face corner normals.
Differential Revision: https://developer.blender.org/D12770
Most of the fields in Mesh had no comments, or outdated misleading
comments. For example, "BMESH ONLY" referred to the BMesh project,
not the data structure. Given how much these structs are used, it should
save a lot of time to have proper comments.
I also rearranged the fields in mesh to have a more logical order. Now
the most important fields come first. In the process I was able to
remove 19 bytes of unnecessary padding (31->12). I just had to
change a `short` flag to `char`.
Differential Revision: https://developer.blender.org/D13454
- 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.
- Minor improvements to doc-strings.
Ref T92709
When enabling or disabling a Mesh Sequence Cache modifier of an Object
with a hair particle system, the hair would switch positions. This is
caused because original coordinates in Blender are expected to be
normalized, and toggling the modifier would cause the usage of different
orco layers: one that is normalized, and the other which isn't.
This bug exposes a few related issues:
- if the Alembic file did not have orco data,
`MOD_deform_mesh_eval_get`, used by the particle system modifier, would
add an orco layer without normalization
- `MOD_deform_mesh_eval_get` would also ignore the presence of an orco
layer (e.g. one that could have been read from Alembic)
- if the Alembic file did have orco data, the data would be read
unnormalized
To fix those various issues, original coordinates are normalized when
read from Alembic and unnormalized when written to Alembic; and a new
utility function `BKE_mesh_orco_ensure` is added to add a normalized
orco layer if none exists on the mesh already, this function derives
from the code used in the particle system.
Reviewed By: brecht
Maniphest Tasks: T92561
Differential Revision: https://developer.blender.org/D13306
This supported calculating normals for MPoly array which was copied to
an MFace aligned array.
Remove the functions entirely since MFace use is being phased out and
these function isn't used anywhere.
Remove the 'only_face_normals' argument.
- BKE_mesh_calc_normals_poly for polygon normals.
- BKE_mesh_calc_normals_poly_and_vertex for poly and vertex normals.
Order arguments logically:
- Pair array and length arguments.
- Position normal array arguments (to be filled) last.
Because mesh vertex and face normals are just derived data, they can
be calculated lazily instead of eagerly. Often normal calculation is
a relatively expensive task, and the calculation is often redundant
if the mesh is deformed afterwards anyway.
Instead, normals should be calculated only when they are needed. This
commit moves in that direction by adding a new function to tag a mesh's
normals dirty and replacing normal calculation with it in some places.
Differential Revision: https://developer.blender.org/D12107
Supported multi-threading for bm_mesh_loops_calc_normals.
This is done by operating on vertex-loops instead of face-loops.
Single threaded operation still loops over faces since iterating
over vertices adds some overhead in the case of custom-normals
as the order used for accessing loops must be the same as iterating
of a faces loops.
From isolated timing tests of bm_mesh_loops_calc_normals on high
poly models, this gives between 3.5x to 10x speedup,
with larger gains for meshes with custom-normals.
NOTE: this is part one of two patches for multi-threaded auto-smooth,
tagging edges as sharp is still single threaded.
Reviewed By: mont29
Ref D11928
Also use `const Curve *` instead of `const Object *`, since the
function works at a lower level than objects anyway.
And also remove another unused function. Since this section of code
for converting curves to meshes will likely be replaced, it's nicer to
see which parts actually remain used at this point.
Normals now includes many functions including normal splitting &
custom normal manipulation split this into it's own file
to centralize related functions.
- Multi-thread BKE_mesh_recalc_looptri.
- Add BKE_mesh_recalc_looptri_with_normals,
this skips having to calculate normals for ngons.
Exact performance depends on number of faces, size of ngons and
available CPU cores.
For high poly meshes the isolated improvement to BKE_mesh_recalc_looptri
in my tests was between 6.7x .. 25.0x, with the largest gains seen in
meshes containing ngons with many sides.
The overall speedup for high poly meshes containing quads and triangles
is only ~20% although ngon heavy meshes can be much faster.
Sergey Sharybin
Campbell Barton
Jacques Lucke
Hans Goudey
Pablo Vazquez
Campbell Barton
Kévin Dietrich