While this didn't cause any user visible bugs, this wouldn't
have behaved as intended since the timer would never run again once
wmTimer.ntime was set to NAN.
GPU_select originally used GL_SELECT which defined the format for
storing the selection result.
Now this is no longer the case, define our own struct - making the code
easier to follow:
- Avoid having to deal with arrays in both `uint*` and `uint(*)[4]`
multiplying offsets by 4 in some cases & not others.
- No magic numbers for the offsets of depth & selection-ID.
- No need to allocate unused members to match GL_SELECT
(halving the buffer size).
Previously, macros were ifdefed using the cmake option `WITH_INTERNATIONAL`
However, the is unnecessary as withen the functions themselves have checks for building without internationalization.
This also means that many `add_definitions(-DWITH_INTERNATIONAL)` are also unnecessary.
Reviewed By: mont29, LazyDodo
Differential Revision: https://developer.blender.org/D13929
Buttons can hold context and it's very useful to use this as a way to
let buttons provide context for drop operators.
For example, with this D13549 can make the material slot list set the
material-slot pointer for each row, and the drop operator can just query
that.
Does two main changes:
* Handle regions in the order as visible on screen. Practically this
just means handling overlapping regions before non-overlapping ones.
* Don't handle any other regions after having found one containing the
mouse pointer.
Fixes: T94016, T91538, T91579, T71899 (and a whole bunch of duplicates)
Addresses: T92364
Differential Revision: https://developer.blender.org/D13539
Reviewed by: Campbell Barton
Asserts that such events actually always lead to a handler return value
that actually keeps the event passing.
Reviewed by Campbell Barton as part of
https://developer.blender.org/D13539.
During sprite fright loading of complex scenes would spend a long time in remapping ID's
The remapping process is done on a per ID instance that resulted in a very time consuming
process that goes over every possible ID reference to find out if it needs to be updated.
If there are N of references to ID blocks and there are M ID blocks that needed to be remapped
it would take N*M checks. These checks are scattered around the place and memory.
Each reference would only be updated at most once, but most of the time no update is needed at all.
Idea: By grouping the changes together will reduce the number of checks resulting in improved performance.
This would only require N checks. Additional benefits is improved data locality as data is only loaded once
in the L2 cache.
It has be implemented for the resyncing process and UI editors.
On an Intel(R) Core(TM) i7-6700 CPU @ 3.40GHz 16Gig the resyncing process went
from 170 seconds to 145 seconds (during hotspot recording).
After this patch has been applied we could add similar approach
to references (references between data blocks) and functionality (tagged deletion).
In my understanding this could reduce the resyncing process to less than a second.
Opening the village production file between 10 and 20 seconds.
Flame graphs showing that UI remapping isn't visible anymore (`WM_main_remap_editor_id_reference`)
* Master {F12769210 size=full}
* This patch {F12769211 size=full}
Reviewed By: mont29
Maniphest Tasks: T94185
Differential Revision: https://developer.blender.org/D13615
Use more efficient logic for detecting when gizmos are under the cursor.
Even though this isn't a bottleneck, it runs on cursor motion in the
3D viewport, so avoiding any lag here is beneficial.
The common case for cursor motion without any gizmos was always
drawing two passes (one small, then again if nothing was found).
Now a single draw call at the larger size is used.
In isolation this gives around 1.2x-1.4x speedup.
When there are multiple gizmos a depth-buffer picking is used
(similar to object / bone selection) which is more involved but
still only performs 2x draw calls since the result is cached for reuse.
See note in gizmo_find_intersected_3d for a more detailed explanation.
Also restore the depth values in the selection result as they're
needed for gizmos to use selection bias.
Broken since support for GL_SELECT was removed.
Early on in 2.8x development gizmo-depth used GL_SELECT,
which has been removed. Bind the depth buffer so occlusion queries
use the front-most gizmo.
While this report only mentions face-maps, gizmo depth was ignored in
all cases. This wasn't noticeable in most cases though since the
transform gizmo for example was placed so gizmos didn't overlap.
This reverts commit 948211679f.
This commit introduced some regressions in the test suite.
As this change is a core part of blender Bastien and I decided to revert
it as the solution isn't clear and needs more investigation.
The following tests FAILED:
62 - blendfile_liblink (SEGFAULT)
63 - blendfile_library_overrides (SEGFAULT)
It fails in (id_us_ensure_real)
During sprite fright loading of complex scenes would spend a long time in remapping ID's
The remapping process is done on a per ID instance that resulted in a very time consuming
process that goes over every possible ID reference to find out if it needs to be updated.
If there are N of references to ID blocks and there are M ID blocks that needed to be remapped
it would take N*M checks. These checks are scattered around the place and memory.
Each reference would only be updated at most once, but most of the time no update is needed at all.
Idea: By grouping the changes together will reduce the number of checks resulting in improved performance.
This would only require N checks. Additional benefits is improved data locality as data is only loaded once
in the L2 cache.
It has be implemented for the resyncing process and UI editors.
On an Intel(R) Core(TM) i7-6700 CPU @ 3.40GHz 16Gig the resyncing process went
from 170 seconds to 145 seconds (during hotspot recording).
After this patch has been applied we could add similar approach
to references (references between data blocks) and functionality (tagged deletion).
In my understanding this could reduce the resyncing process to less than a second.
Opening the village production file between 10 and 20 seconds.
Flame graphs showing that UI remapping isn't visible anymore (`WM_main_remap_editor_id_reference`)
* Master {F12769210 size=full}
* This patch {F12769211 size=full}
Reviewed By: mont29
Maniphest Tasks: T94185
Differential Revision: https://developer.blender.org/D13615
- Use defines instead of magic numbers for F-Key & NDOF range checks.
- Use explicit values for NDOF event types.
- Minor clarification to doc-strings.
- Use doxy-sections.
Some of the message-bus macros are not safe to use in C++. This has come
up before, but no good solution was found. Now @LazyDodo, @HooglyBoogly
and I concluded this is the best duct tape "solution" for the moment.
The message-bus API should address this.
We want to refactor quite some of the Outliner code using C++, this is a
logical step to help the transition to a new architecture.
Includes plenty of fixes to make this compile without warnings, trying
not to change logic. The usual stuff (casts from `void *`, designated
initializers, compound literals, etc.).
When drawing windows on monitors that differ in DPI, we can sometimes
have UI elements draw at an incorrect scale. This patch just ensures
that `wm_window_make_drawable` always updates DPI.
See D10483 for more details.
Differential Revision: https://developer.blender.org/D10483
Reviewed by Brecht Van Lommel
This patch implements the vector types (i.e:`float2`) by making heavy
usage of templating. All vector functions are now outside of the vector
classes (inside the `blender::math` namespace) and are not vector size
dependent for the most part.
In the ongoing effort to make shaders less GL centric, we are aiming
to share more code between GLSL and C++ to avoid code duplication.
####Motivations:
- We are aiming to share UBO and SSBO structures between GLSL and C++.
This means we will use many of the existing vector types and others
we currently don't have (uintX, intX). All these variations were
asking for many more code duplication.
- Deduplicate existing code which is duplicated for each vector size.
- We also want to share small functions. Which means that vector
functions should be static and not in the class namespace.
- Reduce friction to use these types in new projects due to their
incompleteness.
- The current state of the `BLI_(float|double|mpq)(2|3|4).hh` is a
bit of a let down. Most clases are incomplete, out of sync with each
others with different codestyles, and some functions that should be
static are not (i.e: `float3::reflect()`).
####Upsides:
- Still support `.x, .y, .z, .w` for readability.
- Compact, readable and easilly extendable.
- All of the vector functions are available for all the vectors types
and can be restricted to certain types. Also template specialization
let us define exception for special class (like mpq).
- With optimization ON, the compiler unroll the loops and performance
is the same.
####Downsides:
- Might impact debugability. Though I would arge that the bugs are
rarelly caused by the vector class itself (since the operations are
quite trivial) but by the type conversions.
- Might impact compile time. I did not saw a significant impact since
the usage is not really widespread.
- Functions needs to be rewritten to support arbitrary vector length.
For instance, one can't call `len_squared_v3v3` in
`math::length_squared()` and call it a day.
- Type cast does not work with the template version of the `math::`
vector functions. Meaning you need to manually cast `float *` and
`(float *)[3]` to `float3` for the function calls.
i.e: `math::distance_squared(float3(nearest.co), positions[i]);`
- Some parts might loose in readability:
`float3::dot(v1.normalized(), v2.normalized())`
becoming
`math::dot(math::normalize(v1), math::normalize(v2))`
But I propose, when appropriate, to use
`using namespace blender::math;` on function local or file scope to
increase readability.
`dot(normalize(v1), normalize(v2))`
####Consideration:
- Include back `.length()` method. It is quite handy and is more C++
oriented.
- I considered the GLM library as a candidate for replacement. It felt
like too much for what we need and would be difficult to extend / modify
to our needs.
- I used Macros to reduce code in operators declaration and potential
copy paste bugs. This could reduce debugability and could be reverted.
- This touches `delaunay_2d.cc` and the intersection code. I would like
to know @howardt opinion on the matter.
- The `noexcept` on the copy constructor of `mpq(2|3)` is being removed.
But according to @JacquesLucke it is not a real problem for now.
I would like to give a huge thanks to @JacquesLucke who helped during this
and pushed me to reduce the duplication further.
Reviewed By: brecht, sergey, JacquesLucke
Differential Revision: https://developer.blender.org/D13791
This patch implements the vector types (i.e:float2) by making heavy
usage of templating. All vector functions are now outside of the vector
classes (inside the blender::math namespace) and are not vector size
dependent for the most part.
In the ongoing effort to make shaders less GL centric, we are aiming
to share more code between GLSL and C++ to avoid code duplication.
Motivations:
- We are aiming to share UBO and SSBO structures between GLSL and C++.
This means we will use many of the existing vector types and others we
currently don't have (uintX, intX). All these variations were asking
for many more code duplication.
- Deduplicate existing code which is duplicated for each vector size.
- We also want to share small functions. Which means that vector functions
should be static and not in the class namespace.
- Reduce friction to use these types in new projects due to their
incompleteness.
- The current state of the BLI_(float|double|mpq)(2|3|4).hh is a bit of a
let down. Most clases are incomplete, out of sync with each others with
different codestyles, and some functions that should be static are not
(i.e: float3::reflect()).
Upsides:
- Still support .x, .y, .z, .w for readability.
- Compact, readable and easilly extendable.
- All of the vector functions are available for all the vectors types and
can be restricted to certain types. Also template specialization let us
define exception for special class (like mpq).
- With optimization ON, the compiler unroll the loops and performance is
the same.
Downsides:
- Might impact debugability. Though I would arge that the bugs are rarelly
caused by the vector class itself (since the operations are quite trivial)
but by the type conversions.
- Might impact compile time. I did not saw a significant impact since the
usage is not really widespread.
- Functions needs to be rewritten to support arbitrary vector length. For
instance, one can't call len_squared_v3v3 in math::length_squared() and
call it a day.
- Type cast does not work with the template version of the math:: vector
functions. Meaning you need to manually cast float * and (float *)[3] to
float3 for the function calls.
i.e: math::distance_squared(float3(nearest.co), positions[i]);
- Some parts might loose in readability:
float3::dot(v1.normalized(), v2.normalized())
becoming
math::dot(math::normalize(v1), math::normalize(v2))
But I propose, when appropriate, to use
using namespace blender::math; on function local or file scope to
increase readability. dot(normalize(v1), normalize(v2))
Consideration:
- Include back .length() method. It is quite handy and is more C++
oriented.
- I considered the GLM library as a candidate for replacement.
It felt like too much for what we need and would be difficult to
extend / modify to our needs.
- I used Macros to reduce code in operators declaration and potential
copy paste bugs. This could reduce debugability and could be reverted.
- This touches delaunay_2d.cc and the intersection code. I would like to
know @Howard Trickey (howardt) opinion on the matter.
- The noexcept on the copy constructor of mpq(2|3) is being removed.
But according to @Jacques Lucke (JacquesLucke) it is not a real problem
for now.
I would like to give a huge thanks to @Jacques Lucke (JacquesLucke) who
helped during this and pushed me to reduce the duplication further.
Reviewed By: brecht, sergey, JacquesLucke
Differential Revision: http://developer.blender.org/D13791
The issue was caused by Cycles display driver not being able to restore
window's OpenGL context after disposing Cycles-side OpenGL context.
This is due to the window OpenGL re-activation needing to access window
manager which gets cleared out form global main during file reading.
Defer clearing window manager from the global main to until after all
screens are "exited". This allows Cycles to properly stop rendering,
dispose its OpenGL context, and restore window's drawable context.
It is unclear why it was required to clear window manager list early
on. Guess is that it comes from an original code in a1c8543f2a where
there was an early return which then got replaced with an actual logic
without changing the order of de-initialization and window manager list
clear.
Differential Revision: https://developer.blender.org/D13799
It it rather an old experiment now which didn't pay off.
The initial idea was to have main and jobs threads on fast
nodes of TR2 processors. This didn't really work reliably
because in Blender we need to be able to create nested
threads without their affinity set. This is not how some of
OS are creating nested threads, and we don't always have
access to child threads to reset their affinity.
So overall complexity of the initial idea implementation
became too much compared to the performance gain.
This was originally written by Ankit Meel as a GSoC 2020 project.
Howard Trickey added some tests and made some corrections/modifications.
See D13046 for more details.
This commit inserts a new menu item into the export menu called
"Wavefront OBJ (.obj) - New".
For now the old Python exporter remains in the menu, along with
the Python importer, but we plan to remove it soon (leaving the
old addon bundled with Blender but not enabled by default).
This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.
When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).
This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.
We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.
In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.
Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).
See patch description for benchmarks.
Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport
Differential Revision: https://developer.blender.org/D12406
Saving with only a filename (from Python) wasn't being prevented,
while it would successfully write the file to the working-directory,
path remapping and setting relative paths wouldn't work afterwards
as `Main.filepath` would have no directory component.
Disallow this since it's a corner case which only ever occurs
when path names without any directories are used from Python,
the overhead of expanding the working-directory for all data saving
operations isn't worthwhile.
The following changes have been made:
- bpy.ops.wm.save_mainfile() without a filepath argument
fails & reports and error when the file hasn't been saved.
Previously it would write to "untitled.blend" and set the
`G.main->filepath` to this as well.
- bpy.ops.wm.save_mainfile(filepath="untitled.blend")
fails & reports and error as the filename has no directory component.
- `BLI_path_is_abs_from_cwd` was added to check if the path would
attempt to expand to the CWD.
Revert part of the fix from 073669dd85
that initialized the file-path on first save as it's no longer needed.
Also remove relbase argument to BLI_path_normalize as the destination
file paths shouldn't use relative locations.
In almost all cases there is no difference between `G.relbase_valid`
and checking `G.main->filepath` isn't an empty string.
In many places a non-empty string is already being used instead of
`G.relbase_valid`.
The only situation where this was needed was when saving from
`wm_file_write` where they temporarily became out of sync.
This has been replaced by adding a new member to `BlendFileWriteParams`
to account for saving an unsaved file for the first time.
Reviewed By: brecht
Ref D13564
Fix an error if "File Preview Type" is "Auto" and there is no screen.
See D13574 for details.
Differential Revision: https://developer.blender.org/D13574
Reviewed by Julian Eisel
Whenever an exception happens in VR session code, we cancel the entire
session. Alongside that, we removed the "surface" item used to draw into
an offscreen context. This would mess up the iterator of the surface
draw loop.
Similar to 7afd84df40.
Whenever an exception happens in VR session code, we cancel the entire
session. Alongside that, we removed the "surface" item used to draw into
an offscreen context. But this may still be stored as active surface,
leading to a use-after-free when deactivating this active surface, for
example.
Dead-lock when VR viewport drawing and depsgraph updates would fight for
the draw-manager GL lock. This didn't usually cause issues because the
depsgraph would be evaluated at this point already, except in rare
exceptions like after file writing.
Fix this by ensuring the XR surface gets its depsgraph updated after
handling notifiers, which is where regular windows also do the depsgraph
updating.