* Added a limited "fast draw" mode to pbvh drawing
that tries to limit data sent to the GPU.
- Facesets and mask data are disabled.
- Indexed mode is forced.
- Does not work (at all) for multires.
* Finally fixed some outstanding bmesh sculpt undo bugs:
- Forgot to mark pbvh nodes to update their bounds, which
produced a bug I had thought was caused by something else.
- Hopefully fixed a long-standing occasional memory corruption
bug.
This addresses reduced visibility of scenes (as displayed in the VR
headset) that can result from the 8-bit color depth format currently
used for XR swapchain images.
By switching to a swapchain format with higher color depth (RGB10_A2,
RGBA16, RGBA16F) for supported runtimes, visibility in VR should be
noticeably improved.
However, current limitations are lack of support for these higher
color depth formats by some XR runtimes, especially for OpenGL.
Also important to note that GPU_offscreen_create() now explicitly
takes in the texture format (eGPUTextureFormat) instead of a
"high_bitdepth" boolean.
Reviewed By: Julian Eisel, Clément Foucault
Differential Revision: http://developer.blender.org/D9842
ID data-blocks that could be accessed from Python and weren't freed
using BKE_id_free_ex did not release the Python reference count.
Add BKE_libblock_free_data_py function to clear the Python reference
in this case.
Add asserts to ensure no Python reference is held in situations
when ID's are copied for internal use (not exposed through the RNA API),
to ensure these kinds of leaks don't go by unnoticed again.
* The PBVH draw subsystem is now told whether any viewports
have drawtype >= OB_MATERIAL before anything in any windows
are drawn. There are no alternatives given the design
constraints of sculpting, where simply uploading data to the GPU
quickly becomes a bottleneck.
* Fixed flat vcol shading mode.
rBfb87d236edb7 made the values returned by `projmat_dimensions` more
standardized following the documentations. But the functions in Blender
that called `projmat_dimensions` followed a proposal that these values
corresponded to a distance of 1m of clip.
Adjust these functions to follow the new algorithm.
Also:
* added BMLog function to save mesh IDs.
- Used by SCULPT_UNDO_DYNTOPO_BEGIN/END instead of
saving the whole mesh, which was the previous behavior.
* SCULPT_UNDO_DYNTOPO_BEGIN no longer pushes a non-dyntopo
geomtry undo node, as this is no longer necassary.
This greatly speeds up going into/out of sculpt mode
with dyntopo enabled, as before it was copying
the mesh twice.
* PBVH_BMESH now supports faces other then triangles;
* Dyntopo triangulates faces as it finds them.
- I looked into methods of preserving quads and failed to
find anything that worked well in practice; it actually
worked better to use topology rake to align triangles
into quads and then mark diagonal edges for later dissolving
then to try to preserve quads explicitly (I've not
implementated that here, that was research code).
- To avoid excessive cache-destroying loops over vertex-faces,
DynTopo flags which verts have non-triangle faces.
* PBVHTriBuf now builds edge buffers so we can avoid drawing
tesselation phantom edges.
* BMLog also now supports arbitrary faces. It still does not
support edges though.
TODO:
* Fix vcol cell shading mode
* Make sure indexed drawing works
The crash happens because `GPU_offscreen_create` is called with `err_out` `NULL`.
This patch proposes a solution within the `GPU_offscreen_create` itself
and raises an error report in the interface if a menu is called with
dimensions beyond what is supported.
Ref T89782
Maniphest Tasks: T89782
Differential Revision: https://developer.blender.org/D11927
pbvh drawing.
* Dyntopo now stores a list of PBVHTriBufs in leaf nodes, one per material
used by the node.
* Actual drawing buffers live in a new mat_draw_buffers PBVHNode member.
- Added functions to check if the cursor is at a number.
- Added function to parse a number.
- Joined skip_separator functions.
- Added function to check if cursor is at any given set of characters.
Old implementation has a single parser of many different
formats. With the introduction of Vulkan this would lead
to another parser in the same function. This patch
separates the log parsing using a visitor pattern so the
log parsing can be configured per GPU backend or even
per driver.
With Vulkan we manage the compiler our self so the parsing
will become more straight forward. The OpenGL part depends
on many factors (OS, Driver) and perhaps even GPU.
One drawback to trying to predict the number of threads that will be
used in the `task_graph` is that we are only sure of the number when the
threads are running.
Using `BLI_task_parallel_range` allows the driver to
choose the best thread distribution through `parallel_reduce`.
The benefit is most evident on hardware with fewer cores.
This is the result on an 4-core laptop:
||before:|after:
|---|---|---|
|large_mesh_editing:|Average: 5.203638 FPS|Average: 5.398925 FPS
||rdata 15ms iter 43ms (frame 193ms)|rdata 14ms iter 36ms (frame 187ms)
Differential Revision: https://developer.blender.org/D11558
This is an adaptation of {D11488}.
A disadvantage of manually setting the iter ranges per thread is that
we don't know how many threads are running in the background and so we
don't know how to best distribute the ranges.
To solve this limitation we can use `parallel_reduce` and thus let the
driver choose the best distribution of ranges among the threads.
This proved to be especially beneficial for computers with few cores.
**Benchmarking:**
Here's the result on an 4-core laptop:
||master:|PATCH:
|---|---|---|
|large_mesh_editing:|Average: 5.203638 FPS|Average: 5.398925 FPS
||rdata 15ms iter 43ms (frame 193ms)|rdata 14ms iter 36ms (frame 187ms)
Here's the result on an 8-core PC:
||master:|PATCH:
|---|---|---|
|large_mesh_editing:|Average: 15.267482 FPS|Average: 15.906881 FPS
||rdata 9ms iter 28ms (frame 65ms)|rdata 9ms iter 25ms (frame 63ms)
|large_mesh_editing_ledge: |Average: 15.145966 FPS|Average: 15.520474 FPS
||rdata 9ms iter 29ms (frame 65ms)|rdata 9ms iter 25ms (frame 64ms)
|looptris_test:|Average: 4.001917 FPS|Average: 4.061105 FPS
||rdata 12ms iter 90ms (frame 236ms)|rdata 12ms iter 87ms (frame 230ms)
|subdiv_mesh_cage_and_final:|Average: 1.917769 FPS|Average: 1.971790 FPS
||rdata 7ms iter 37ms (frame 261ms)|rdata 7ms iter 31ms (frame 258ms)
||rdata 7ms iter 38ms (frame 252ms)|rdata 7ms iter 33ms (frame 249ms)
|subdiv_mesh_final_only:|Average: 6.387240 FPS|Average: 6.591251 FPS
||rdata 3ms iter 25ms (frame 151ms)|rdata 3ms iter 16ms (frame 145ms)
|subdiv_mesh_final_only_ledge:|Average: 6.247393 FPS|Average: 6.596024 FPS
||rdata 3ms iter 26ms (frame 158ms)|rdata 3ms iter 16ms (frame 148ms)
**Notes:**
- The improvement can only be noticed if all extracts are multithreaded.
- This patch touches different areas of the code, so it can be split into another patch if the idea is accepted.
These screenshots show how threads behave in a quadcore:
Master:
{F10164664}
Patch:
{F10164666}
Differential Revision: https://developer.blender.org/D11558
Current index builder is designed to be used in a single thread.
This makes all index buffer extractions single threaded.
This patch adds a thread safe solution enabling multithreaded
building of index buffers.
To reduce locking the solution would provide a task/thread local
index buffer builder (called sub builder).
When a thread is finished this thread local index buffer builder
can be joined with the initial index buffer builder.
`GPU_indexbuf_subbuilder_init`: Initialized a sub builder. The
index list is shared between the parent and sub buffer, but the
counters are localized. Ensuring that updating counters would
not need any locking.
`GPU_indexbuf_subbuilder_finish`: merge the information of the
sub builder back to the parent builder. Needs to be invoked outside
the worker thread, or when sure that all worker threads have been
finished. Internal the function is not thread safe.
For testing purposes the extract_points extractor has been migrated to
the new API. Herefore changes to the mesh extractor were needed.
* When creating tasks, the task number of current task is stored in
ExtractTaskData including the total number of tasks.
* Adding two functions in `MeshExtract`.
** `task_init` will initialize the task specific userdata.
** `task_finish` should merge back the task specific userdata back.
* adding task_id parameter to the iteration functions so they can
access the correct task data without any need for locking.
There is no noticeable change in end user performance.
Reviewed By: mano-wii
Differential Revision: https://developer.blender.org/D11499
Moving the bounds code to the builder can be useful
for future optimizations like building multithreaded.
Reviewed By: fclem, jbakker
Differential Revision: https://developer.blender.org/D11455
When projecting into screen space Z value isn't always needed.
Add 2D projection functions, renaming them to avoid accidents
happening again.
- Add GPU_matrix_project_2fv
- Add ED_view3d_project_v2
- Rename ED_view3d_project to ED_view3d_project_v3
- Use the 2D versions of these functions when the Z value isn't used.
