`BLI_task_parallel_range` counts the number of tasks depending on the
number of items.
In the case of `BLI_bvhtree_overlap` the number of items is always
between 2 and 16, which makes it always run in single thread.
So, set the maximum number of items per thread to 1.
In my tests the cloth collision system (which calls that function)
went from 0.80fps to 0.88fps.
Differential Revision: https://developer.blender.org/D6523
As you can see, here is testing the "children" of `node1`.
So bvhtree is `tree1`.
This problem has never been observed because usually `tree_type`
of `tree1` equals `tree_type` of `tree2`.
Although not using multithreading, the `thread` parameter in the
`BVHTree_OverlapCallback` callback always returned a value between the
"number of threads".
This parameter should always be 0 in such cases.
Also a `BLI_Stack` was created for each "thread" and used.
This small overhead is no longer seen.
Differential Revision: https://developer.blender.org/D6510
Currently the action channels are applied after the existing
transformation, as if the action controlled a child of the
bone. This is not very natural, but more importantly, the
transform tools are not designed to work conveniently with an
additional 'pseudo-child' transformation, resulting in effects
like an unexpected pivot location.
Implementing a Before mode that integrates the action channels
as if applied to a parent allows using the special transform
tool code intended for dealing with such constraints.
Note that in either mode, Action constraints should be added
in reverse order, putting a new constraint before the existing
ones that the Action was keyframed to work together.
In order to implement the option, extract a utility from
the Copy Transform constraint code for combining transforms
with special anti-shear scale handling that matches the
Aligned Inherit Scale mode.
The Before mode also requires switching the constraint to
the Local owner space, while the After mode can still use the
World space for efficiency as before. Since the constraint
doesn't have an Owner space option in the UI, this has to be
handled in an RNA setter.
For full backward compatibility, the original simple matrix
multiplication mode is preserved as the third option, but it
is not recommended due to creating shear.
Differential Revision: https://developer.blender.org/D6297
Previously this limit was rather high, but with UDIMs it's fairly easy
to reach this many images. Even though this exceeds the texture limit
on most hardware as far as I can tell, it should at least not crash.
The old code uses a fixed array which overflows eventually, this fix
replaces the array with a GSet.
Reviewed By: fclem
Differential Revision: https://developer.blender.org/D6416
This patch contains the work that I did during my week at the Code Quest - adding support for tiled images to Blender.
With this patch, images now contain a list of tiles. By default, this just contains one tile, but if the source type is set to Tiled, the user can add additional tiles. When acquiring an ImBuf, the tile to be loaded is specified in the ImageUser.
Therefore, code that is not yet aware of tiles will just access the default tile as usual.
The filenames of the additional tiles are derived from the original filename according to the UDIM naming scheme - the filename contains an index that is calculated as (1001 + 10*<y coordinate of the tile> + <x coordinate of the tile>), where the x coordinate never goes above 9.
Internally, the various tiles are stored in a cache just like sequences. When acquired for the first time, the code will try to load the corresponding file from disk. Alternatively, a new operator can be used to initialize the tile similar to the New Image operator.
The following features are supported so far:
- Automatic detection and loading of all tiles when opening the first tile (1001)
- Saving all tiles
- Adding and removing tiles
- Filling tiles with generated images
- Drawing all tiles in the Image Editor
- Viewing a tiled grid even if no image is selected
- Rendering tiled images in Eevee
- Rendering tiled images in Cycles (in SVM mode)
- Automatically skipping loading of unused tiles in Cycles
- 2D texture painting (also across tiles)
- 3D texture painting (also across tiles, only limitation: individual faces can not cross tile borders)
- Assigning custom labels to individual tiles (drawn in the Image Editor instead of the ID)
- Different resolutions between tiles
There still are some missing features that will be added later (see T72390):
- Workbench engine support
- Packing/Unpacking support
- Baking support
- Cycles OSL support
- many other Blender features that rely on images
Thanks to Brecht for the review and to all who tested the intermediate versions!
Differential Revision: https://developer.blender.org/D3509
When creating shaders and using maths functions it is expected that Blender should match functions in other DCC applications, game engines and shading languages such as GLSL and OSL.
This patch adds missing functions to the Blender maths node.
Ideally, it would be nice to have these functions available to vectors too but that is not part of this patch.
This patch adds the following functions trunc, snap, wrap, compare, pingpong, sign, radians, degrees, cosh, sinh, tanh, exp, smoothmin and inversesqrt.
Sign function is based on GLSL and OSL functions and returns zero when x == 0.
Differential Revision: https://developer.blender.org/D5957
After this commit it should be possible to share the same deformation
formulas that are used in the Elastic Deform brush with other areas of
Blender such as Grease Pencil or proportional editing.
This also removes a lot of code from sculpt.c that is not direclty
related to sculpting.
Reviewed By: jbakker
Differential Revision: https://developer.blender.org/D6281
This adds some basic simulation of internal air pressure inside of
closed cloth mesh objects.
Reviewed By: Jacques Lucke
Differential Revision: http://developer.blender.org/D5473
This code allows to push a set of different operations all based on
iterations over a range of indices, and then process them all at once
over multiple threads.
This commit also adds unit tests for both old un-pooled, and new pooled
task_parallel_range family of functions, as well as some basic
performances tests.
This is mainly interesting for relatively low amount of individual
tasks, as expected.
E.g. performance tests on a 32 threads machine, for a set of 10
different tasks, shows following improvements when using pooled version
instead of ten sequential calls to BLI_task_parallel_range():
| Num Items | Sequential | Pooled | Speed-up |
| --------- | ---------- | ------- | -------- |
| 10K | 365 us | 138 us | 2.5 x |
| 100K | 877 us | 530 us | 1.66 x |
| 1000K | 5521 us | 4625 us | 1.25 x |
Differential Revision: https://developer.blender.org/D6189
Note: Compared to previous commit yesterday, this reworks atomic handling in
parallel iter code, and fixes a dummy double-free bug.
Now we should only use the two critical values for synchronization from
atomic calls results, which is the proper way to do things.
Reading a value after an atomic operation does not guarantee you will
get the latest value in all cases (especially on Windows release builds
it seems).
This code allows to push a set of different operations all based on
iterations over a range of indices, and then process them all at once
over multiple threads.
This commit also adds unit tests for both old un-pooled, and new pooled
`task_parallel_range` family of functions, as well as some basic
performances tests.
This is mainly interesting for relatively low amount of individual
tasks, as expected.
E.g. performance tests on a 32 threads machine, for a set of 10
different tasks, shows following improvements when using pooled version
instead of ten sequential calls to `BLI_task_parallel_range()`:
| Num Items | Sequential | Pooled | Speed-up |
| --------- | ---------- | ------- | -------- |
| 10K | 365 us | 138 us | 2.5 x |
| 100K | 877 us | 530 us | 1.66 x |
| 1000K | 5521 us | 4625 us | 1.25 x |
Differential Revision: https://developer.blender.org/D6189
`BLI_strncpy_wchar_from_utf8` internally assumes `wchar_t` is 32 bits
which is not the case on windows.
The solution is to replace `wchar_t` with `char32_t`.
Thanks to @robbott for compatibility on macOS.
Differential Revision: https://developer.blender.org/D6198
As correctly pointed out by a comment in the code, adding
new springs wasn't thread safe, and caused crashes.
Fix by buffering new springs in intermediate thread-local
arrays, which are flushed on the main thread. This is valid
because the new springs are not used until the next sim step.
Differential Revision: https://developer.blender.org/D6133
Check was losing precision -- adjust by translating points
before calculating circumcircle.
Also, needed to check for flippability of edges before flipping.
Check was losing precision -- adjust by translating points
before calculating circumcircle.
Also, needed to check for flippability of edges before flipping.
This new function is part of the 'parallel for loops' functions. It
takes an iterator callback to generate items to be processed, in
addition to the usual 'process' func callback.
This allows to use common code from BLI_task for a wide range of custom
iteratiors, whithout having to re-invent the wheel of the whole tasks &
data chuncks handling.
This supports all settings features from `BLI_task_parallel_range()`,
including dynamic and static (if total number of items is knwon)
scheduling, TLS data and its finalize callback, etc.
One question here is whether we should provide usercode with a spinlock
by default, or enforce it to always handle its own sync mechanism.
I kept it, since imho it will be needed very often, and generating one
is pretty cheap even if unused...
----------
Additionaly, this commit converts (currently unused)
`BLI_task_parallel_listbase()` to use that generic code. This was done
mostly as proof of concept, but performance-wise it shows some
interesting data, roughly:
- Very light processing (that should not be threaded anyway) is several
times slower, which is expected due to more overhead in loop management
code.
- Heavier processing can be up to 10% quicker (probably thanks to the
switch from dynamic to static scheduling, which reduces a lot locking
to fill-in the per-tasks chunks of data). Similar speed-up in
non-threaded case comes as a surprise though, not sure what can
explain that.
While this conversion is not really needed, imho we should keep it
(instead of existing code for that function), it's easier to have
complex handling logic in as few places as possible, for maintaining and
for improving it.
Note: That work was initially done to allow for D5372 to be possible... Unfortunately that one proved to be not better than orig code on performances point of view.
Reviewed By: sergey
Differential Revision: https://developer.blender.org/D5371
* Add proper adjustment for scale in the solver epsilon computation.
* Run at least one full iteration of the solver, even if the initial
state meets the epsilon requirement.
* When applying offset, blend normal into the offset direction
as the initial point moves very close to the target mesh.
Also random improvements to debug trace output in the console.
