The idea is to make main thread and job threads to be scheduled
on CPU dies which has direct access to memory (those are NUMA
nodes 0 and 2).
We also do this for new EPYC CPUs since their NUMA nodes 1 and 3
do have access but only to a higher range DDR slots. By preferring
nodes 0 and 2 on EPYC we make it so users with partially filled
DDR slots has fast memory access.
One thing which is not really solved yet is localization of
memory allocation: we do not guarantee that memory is allocated
on the closest to the NUMA node DDR slot and hope that memory
manager of OS is acting in favor of us.
There is a new `bpy.app.timers` api.
For more details, look in the Python API documentation.
Reviewers: campbellbarton
Differential Revision: https://developer.blender.org/D3994
The goal is to address performance regression when going from
few threads to 10s of threads. On a systems with more than 32
CPU threads the benefit of threaded loop was actually harmful.
There are following tweaks now:
- The chunk size is adaptive for the number of threads, which
minimizes scheduling overhead.
- The number of tasks is adaptive to the list size and chunk
size.
Here comes performance comparison on the production shot:
Number of threads DEG time before DEG time after
44 0.09 0.02
32 0.055 0.025
16 0.025 0.025
8 0.035 0.033
The Nearest Surface Point shrink method, while fast, is neither
smooth nor continuous: as the source point moves, the projected
point can both stop and jump. This causes distortions in the
deformation of the shrinkwrap modifier, and the motion of an
animated object with a shrinkwrap constraint.
This patch implements a new mode, which, instead of using the simple
nearest point search, iteratively solves an equation for each triangle
to find a point which has its interpolated normal point to or from the
original vertex. Non-manifold boundary edges are treated as infinitely
thin cylinders that cast normals in all perpendicular directions.
Since this is useful for the constraint, and having multiple
objects with constraints targeting the same guide mesh is a quite
reasonable use case, rather than calculating the mesh boundary edge
data over and over again, it is precomputed and cached in the mesh.
Reviewers: mont29
Differential Revision: https://developer.blender.org/D3836
Simple find_nearest relies on a heuristic for efficient culling of
the BVH tree, which involves a fast callback that always updates the
result, and the caller reusing the result of the previous find_nearest
to prime the process for the next vertex.
If the callback is slow and/or applies significant restrictions on
what kind of nodes can qualify for the result, the heuristic can't
work. Thus for such tasks it is necessary to order and prune nodes
before the callback at BVH tree level using a priority queue.
Since, according to code history, for simple find_nearest the
heuristic approach is faster, this mode has to be an option.
The main use one can imagine for this is adding tweak controls to
parts of a model that are already deformed by multiple other major
bones. It is natural to expect such locations to deform as if the
tweaks aren't there by default; however currently there is no easy
way to make a bone follow multiple other bones.
This adds a new constraint that implements the math behind the Armature
modifier, with support for explicit weights, bone envelopes, and dual
quaternion blending. It can also access bones from multiple armatures
at the same time (mainly because it's easier to code it that way.)
This also fixes dquat_to_mat4, which wasn't used anywhere before.
Differential Revision: https://developer.blender.org/D3664
If the user only needs insertion and removal from top, there is
no need to allocate and manage separate HeapNode objects: the
data can be stored directly in the main tree array.
This measured a 24% FPS increase on a ~50% heap-heavy workload.
Reviewers: brecht
Differential Revision: https://developer.blender.org/D3898
The index field of nodes is supposed to be its actual index, so
there is no need to read it in swap. On a 64-bit processor i and
j are already in registers, so this removes two memory reads.
In addition, cache the tree pointer, use branch hints, and
put the most frequently accessed 'value' field at 0 offset.
Produced a 20% FPS improvement for a 50% heap-heavy workload.
The code this was taken from assumes a 'size_t' result,
which isn't the case here.
In practice the bucket distribution wasn't bad,
even so this was a nop so best fix.
Partially revert 41216d5ad4
Some of this code had comments to be left as is for readability,
or comment the code should be kept.
Other functions were only for debugging.
This commit includes several performance, stability, and reliability
improvements to cloth collisions.
Most notably:
* The implementation of a new self-collisions system.
* Multithreading of collision detection.
* Implementation of single sided collisions and normal overrides.
* Replacement of the `plNearestPoints` function from Bullet with a
dedicated solution.
Further, this also includes several bug fixes, and algorithmic
improvements.
Reviewed By: brecht
Differential Revision: http://developer.blender.org/D3712
Terms get/set don't make much sense when casting values.
Name macros so the conversion is obvious,
use common prefix for easier completion.
- GET_INT_FROM_POINTER -> POINTER_AS_INT
- SET_INT_IN_POINTER -> POINTER_FROM_INT
- GET_UINT_FROM_POINTER -> POINTER_AS_UINT
- SET_UINT_IN_POINTER -> POINTER_FROM_UINT
Also minor changes:
- Remove unused headers.
- rename SimpleExprParseState -> ExprParseState
- rename max_ops -> ops_count_alloc
Was misleading since maximums are often limits which can't be exceeded.
Simple isn't a good prefix for library names since
lots of unrelated modules could be called 'simple'.
Include 'py' in module name since this is a subset of Python,
one of the main motivations for this is to be Python like/compatible.
The goal here is to make app templates usable for default templates
that we can ship with Blender. These only have a custom startup.blend
currently and so are quite limited compared to app templates that fully
customize Blender.
But still it seems like the same kind of concept where we should be
sharing the code and UI. It is useful to be able to save a startup.blend
per template, and I can imagine some scripting being useful in the future
as well.
Changes made:
* File > New and Ctrl+N now list the templates, replacing a separate
Application Templates menu that was not as easy to discover.
* File menu now shows name of active template above Save Startup File
and Load Factory Settings to indicate these are saved/loaded per
template.
* The "Default" template was renamed to "General".
* Workspaces can now be added from any of the template startup.blend
files when clicking the (+) button in the topbar.
* User preferences are now fully shared between app templates, unless
the template includes a custom userpref.blend. I think this will be
useful in general, not all app templates need their own keymaps for
example.
* Previously Save User Preferences would save the current app template
and then Blender would start using that template by default. I've
disabled this, to me it seems it was unintentional, or at least not
clear at all that saving user preferences also makes the current
Differential Revision: https://developer.blender.org/D3690
Recently @sergey found that hard-coding evaluation of certain very
common driver expressions without calling the Python interpreter
produces a 30-40% performance improvement. Since hard-coding is
obviously not suitable for production, I implemented a proper
parser and interpreter for simple arithmetic expressions in C.
The evaluator supports +, -, *, /, (), ==, !=, <, <=, >, >=,
and, or, not, ternary if; driver variables, frame, pi, True, False,
and a subset of standard math functions that seem most useful.
Booleans are represented as numbers, since within the supported
operation set it seems to be impossible to distinguish True/False
from 1.0/0.0. Boolean operations properly implement lazy evaluation
with jumps, and comparisons support chaining like 'a < b < c...'.
Expressions are parsed into a very simple stack machine program
that can then be safely evaluated in multiple threads.
Reviewers: sergey, campbellbarton
Differential Revision: https://developer.blender.org/D3698
