Crash seems to happen due to parallelized access of data in liboverride
diffing code (presumably when a same linked data is used as reference
for several local liboverrides?).
Since there is zero reason to actually create IDGroup properties there
(it's actually fairly bad, also adding useless overhead and trash data
in system IDProperties), add a new API to access a Pointer PropertyRNA,
`RNA_property_pointer_get_never_create`, which ensures that the call
never creates data, and simply returns `PointerRNA_NULL` instead.
Also reverts abd683fcb5 and re-enable liboverride unittests.
NOTE: This only addresses the case from the report, the current behavior
of `RNA_property_pointer_get` is simply wrong and needs to be rethought.
This is not a simple change though most likely. See also #147072.
Pull Request: https://projects.blender.org/blender/blender/pulls/146990
While adding more tests does add a tiny bit of overhead, having tests
within the 10s range is better to ensure a good distribution of them
over many threads. It also helps narrowing down the actual blendfile in
case of issues.
This commit adds a UI test that activates each of the default tools for
a mesh in Sculpt Mode and verifies that no Python errors are emitted
when the corresponding tool settings are displayed to the user.
Note that this does not test all of the default bundled brushes.
Pull Request: https://projects.blender.org/blender/blender/pulls/144397
This adds a function that can turn an existing `bNodeTree` into an inlined one.
The new node tree has all node groups, repeat zones, closures and bundles
inlined. So it's just a flat tree that ideally can be consumed easily by render
engines. As part of the process, it also does constant folding.
The goal is to support more advanced features from geometry nodes (repeat zones,
etc.) in shader nodes which the evaluator is more limited because it has to be
able to run on the GPU. Creating an inlined `bNodeTree` is likely the most
direct way to get but may also be limiting in the future. Since this is a fairly
local change, it's likely still worth it to support these features in all render
engines without having to make their evaluators significantly more complex.
Some limitations apply here that do not apply in Geometry Nodes. For example,
the iterations count in a repeat zone has to be a constant after constant
folding.
There is also a `Test Inlining Shader Nodes` operator that creates the inlined
tree and creates a group node for it. This is just for testing purposes.
#145811 will make this functionality available to the Python API as well so that
external renderers can use it too.
* Store scene linear to XYZ conversion matrix in each blend file, along
with the colorspace name. The matrix is the source of truth. The name
is currently only used for error logging about unknown color spaces.
* Add Working Space option in color management panel, to change the
working space for the entire blend file. Changing this will pop up
a dialog, with a default enabled option to convert all colors in
the blend file to the new working space. Note this is necessarily only
an approximation.
* Link and append automatically converts to the color space of the main
open blend file.
* There is builtin support for Rec.709, Rec.2020 and ACEScg working spaces,
in addition to the working space of custom OpenColorIO configs.
* Undo of working space for linked datablocks isn't quite correct when going
to a smaller gamut working space. This can be fixed by reloading the file
so the linked datablocks are reloaded.
Compatibility with blend files saved with a custom OpenColorIO config
is tricky, as we can not detect this.
* We assume that if the blend file has no information about the scene
linear color space, it is the default one from the active OCIO config.
And the same for any blend files linked or appended. This is effectively
the same behavior as before.
* Now that there is a warning when color spaces are missing, it is more
likely that a user will notice something is wrong and only save the
blend file with the correct config active.
* As no automatic working space conversion happens on file load, there is
an opportunity to correct things by changing the working space with
"Convert Colors" disabled. This can also be scripted for all blend files
in a project.
Ref #144911
Pull Request: https://projects.blender.org/blender/blender/pulls/145476
Add geometry file import nodes support to the for-each-path logic.
This will make `bpy.data.file_path_map()` report the input files for
OBJ, PLY, etc. import nodes (and any other node that has a string
property of subtype `PROP_FILEPATH`).
Currently this only supports static file paths, so where the file path
is set as the input socket's default value. When the path is
determined via any noodle, this is ignored and the default value is
reported anyway.
This is necessary for the new version of Blender Asset Tracer, which
in turn is needed to resolvestudio/flamenco#104423.
Pull Request: https://projects.blender.org/blender/blender/pulls/144874
This test can't easily be updated because it relies on the old socket shape
design before #144119. Parts of it could be done from scratch again, but
most of this stuff is also tested by the new structure type inferencing test
in `bl_node_structure_type_inference.py`.
Pull Request: https://projects.blender.org/blender/blender/pulls/144185
Temporary directory handling had a logical error, assuming the
"session" temporary directory was owned and created by Blender
and could be recursively removed on exit.
However, it's possible creating the session sub-directory fails,
in that case the temporary directory was used for the "session".
This meant setting `C:\` as the temporary directory in the preferences
would attempt to recursively remove `C:\` on exit.
Resolve with the following changes:
- Only perform a recursive removal on the temporary directory
if a session sub-directory was created.
- If the creating the user-preferences temporary "session" sub-directory
fails fall back to the systems temporary directory and try to
create the "session" directory there.
Previously this was only done if the preference path didn't exist.
The preferences path was still used if it existed but couldn't be
written to.
Include a test to ensure this is working as expected.
Ref !144042
Add a new package `scripts/modules/_bpy_internal/http`, containing
classes to download files via HTTP.
The code is intentionally put into the `_bpy_internal` package, as I
don't intend it to be the end-all-be-all of downloaders for general
use in add-ons. It's been written to support the Remote Asset Library
project (#134495), where it will be used to download JSON files (to
get the list of assets on the server) as well as the asset files
themselves.
The module consists of several parts. The main ones are:
`class ConditionalDownloader`
: File downloader, which downloads a URL to a file on disk.
It supports conditional requests via `ETag`/`If-None-Match` and
`Last-Modified`/`If-Modified-Since` HTTP headers (RFC 7273, section 3.
Precondition Header Fields). A `304 Not Modified` response is
treated as a succesful download.
Metadata of the request (the response length in bytes, and the above
headers) are stored on disk, in a location that is determined by the
user of the class. Probably in the future it would be nice to have a
single sqlite database for this (there's a TODO in the code about
this).
The downloader uses the Requests library, and manages its own HTTP
session object. This way it can handle TCP/IP connection reuse,
automatically retry failing connections, and in the future
HTTP-level authentication.
`class BackgroundDownloader`
: Wrapper for a `ConditionalDownloader` that manages a background
process for the actual downloading.
It runs the downloader in a background process, while ensuring that
its reporters (see below) get called on the main process. This way
it's possible to do background downloading, while still receiving
progress reports in a modal operator, which in turn can directly
call Blender's Python API. Care was taken to [not use Python
threads][1]
`class DownloadReporter`
: Protocol class. Objects adhering to the protocol can be given to a
`ConditionalDownloader` or `BackgroundDownloader`. The protocol has
functions like `download_starts(…)`, `download_progress(…)`,
`download_error(…)`, which will be called by the downloader to
report on what it's doing.
I chose to make this a protocol, rather than an abstract superclass,
because then it's possible to make an Operator a DownloadReporter
without requiring multi-classing.
[1]: https://docs.blender.org/api/main/info_gotchas_threading.html
Pull Request: https://projects.blender.org/blender/blender/pulls/138327
**How to reproduce:**
1. Assign a node tree to the scene compositing node group in Python
2. Notice how the node tree has the wrong user count
Example:
```python
C.scene.compositing_node_group = \
D.node_groups.new("ntree", "CompositorNodeTree")
print(C.scene.compositing_node_group.users) # returns 0
```
Pull Request: https://projects.blender.org/blender/blender/pulls/143577
Previously code that was reading Strip data assumed that seqbasep
and channels members would stay at fixed offsets within a struct,
forever into the future. Fix this by inferring their offsets from
the file SDNA data where needed.
Actual Strip DNA layout is not changed in this commit yet.
Co-authored-by: Sergey Sharybin <sergey@blender.org>
Pull Request: https://projects.blender.org/blender/blender/pulls/142940
This includes a new list structure type and socket shape, a node
to create lists, a node to retrieve values from lists, and a node to
retrieve the length of lists. It also implements multi-function support
so that function nodes work on lists.
There are three nodes included in this PR.
- **List** Creates a list of elements with a given size. The values
are computed with a field that can use the index as an input.
- **Get List Item** A field node that retrieves an element from a
a list at a given index. The index input is dynamic, so if the input
is a list, the output will be a list too.
- **List Length** Just gives the length of a list.
When a function node is used with multiple list inputs, the shorter
lists are repeated to extend it to the length of the longest.
The list nodes and structure type are hidden behind an experimental
feature until we can be sure they're useful for an actual use case.
Pull Request: https://projects.blender.org/blender/blender/pulls/140679
HDR video files are properly read into Blender, and can be rendered out
of Blender.
HDR video reading / decoding:
- Two flavors of HDR are recognized, based on color related video
metadata: "PQ" (Rec.2100 Perceptual Quantizer, aka SMPTE 2084) and
"HLG" (Rec.2100 Hybrid-Log-Gamma, aka ARIB STD B67). Both are read
effectively into floating point images, and their color space
transformations are done through OpenColorIO.
- The OCIO config shipped in Blender has been extended to contain
Rec.2100-PQ and Rec.2100-HLG color spaces.
- Note that if you already had a HDR video in sequencer or movie clip,
it would have looked "incorrect" previously, and it will continue to
look incorrect, since it already has "wrong" color space assigned to
it. Either re-add it (which should assign the correct color space),
or manually change the color space to PQ or HLG one as needed.
HDR video writing / encoding"
- For H.265 and AV1 the video encoding options now display the HDR mode.
Similar to reading, there are PQ and HLG HDR mode options.
- Reference white is assumed to be 100 nits.
- YUV uses "full" ("PC/jpeg") color range.
- No mastering display metadata is written into the video file, since
generally that information is not known inside Blender.
More details and screenshots in the PR.
Co-authored-by: Sergey Sharybin <sergey@blender.org>
Pull Request: https://projects.blender.org/blender/blender/pulls/120033
Adds a new file for testing the Multires Apply Base operator.
This is not included with the other modifier tests as it is a bit of an
exception - the modifier is not applied at the end for comparison
between the expected result and actual result.
Pull Request: https://projects.blender.org/blender/blender/pulls/141571
This commit moves Curves and Grease Pencil to use `AttributeStorage`
instead of `CustomData`, except for vertex groups. This PR mostly
involves extending the changes from the above commit for point clouds
to generalize to other geometry types.
This is mostly straightforward, though a couple non-trivial places of
note are the joining of Grease Pencil objects (`merge_attributes`), the
"default render fallback" UV for curves objects which was previously
unused at the UI level and just ended up being the first attribute, and
the `update_curve_types()` call in the curves versioning function.
Similar to:
- fa03c53d4a
- f74e304b00
Part of #122398.
Pull Request: https://projects.blender.org/blender/blender/pulls/140936
This addresses issue #120949 to move compositor tests to reflect the
grouping used when adding a new node.
This PR only moves the relevant single tests and their renders into
matching directories. Folders such as 'multi-node setups' and
'pixel nodes' were not changed.
Pull Request: https://projects.blender.org/blender/blender/pulls/139757
This new file can parse the file header (first few bytes) as well as the block
headers.
Right now, this is used by two places:
* `blendfile.py` which is used by `blend2json.py`
* `blend_render_info.py`
This new module is shipped with Blender because it's needed for
`blend_render_info.py` which is shipped with Blender too. This makes using it in
`blendfile.py` (which is not shipped with Blender) a bit more annoying. However,
this is already not ideal, because e.g. `blend2json` also has to add to
`sys.path` already to be able to import `blendfile.py`.
This new file could also be used by blender-asset-tracer (BAT).
The new `BlendFileHeader` and `BlockHeader` types may be subclassed by code
using it, because it wants to store additional derived data (`blendfile.py` and
BAT need this).
New tests have been added that check that the file and block header is parsed
correctly for different kinds of .blend files.
Pull Request: https://projects.blender.org/blender/blender/pulls/140341
The way these tests work is similar to the existing field inferencing tests.
There is a .blend file that is opened and then we check the inferred structure
types from Python. A new `NodeSocket.inferred_structure_type` property is added
to be able to access this information. Other then the field inferencing tests,
this patch does not directly check the socket shapes, which are not always
exactly determined by the inferred structure type.
This also fixes a few issues I found while adding the tests.
Pull Request: https://projects.blender.org/blender/blender/pulls/140520
This changes the engine identifier back to `BLENDER_EEVEE`.
We keep the `BLENDER_EEVEE_NEXT` identifier around for
versioning reasons (have to detect when it is the active
engine of a older file).
This also rename a bunch of pannels that were using `next`
in their name.
This is a breaking change for Addons compatibility.
Pull Request: https://projects.blender.org/blender/blender/pulls/140282
While individual modes have UI tests related to undo, this new set of
tests in this new file is intended to be a set of very broad sanity
tests that catch the most egregious errors that cause crashing on start
up, whether due to python errors, UI rendering issues, or otherwise.
Running these tests takes approximately 4 seconds currently as it adds
and verifies the loading of each of the workspaces available "out of
the box" to a blender user.
Pull Request: https://projects.blender.org/blender/blender/pulls/139318
This commit adds 9 tests that check each of the default brush curve
strength preset options to ensure that none of them cause NaN
propagation. In total this takes approximately 0.7s to run to run.
By design, these tests are very broad and are not a replacement for
other testing, but they should help in reducing the chance of potential
regressions.
Related to #140162
Pull Request: https://projects.blender.org/blender/blender/pulls/140242
In the render test suite there is an OSL folder that contains tests that
need OSL to function.
Previously due to some missed logic, the OptiX OSL test suite would not run
tests in that folder because part of the test code assumed that if you aren't
testing on a CPU, then your device doesn't support OSL.
This commit fixes this issue.
Along with this change, some logic was changed in preparation for allowing
OptiX OSL camera tests to run without OSL shading enabled.
Pull Request: https://projects.blender.org/blender/blender/pulls/139433
For Sculpt Undo, certain operators will modify the topology count of the
mesh. These operators are handled separately from normal brush strokes,
and so having tests for an operator that uses this functionality is
beneficial in detecting regressions.
Pull Request: https://projects.blender.org/blender/blender/pulls/139249
When building Blender WITH_MOD_FLUID=OFF, the OpenVDB render tests
would fail as some of them make use of the WITH_MOD_FLUID features.
This commit fixes this by disabling OpenVDB render tests unless
WITH_MOD_FLUID is active.
Pull Request: https://projects.blender.org/blender/blender/pulls/138728
This change moves the tests data files and publish folder of assets
repository to the main blender.git repository as LFS files.
The goal of this change is to eliminate toil of modifying tests,
cherry-picking changes to LFS branches, adding tests as part of a
PR which brings new features or fixes.
More detailed explanation and conversation can be found in the
design task.
Ref #137215
Pull Request: https://projects.blender.org/blender/blender/pulls/137219
Imports in the following animated data from UsdGeomCameras:
- Focal length
- DOF distance
- DOF fstop
- Clip start, Clip end
- Tilt shift x, Tilt shift y
- Aperture sizes (with caveats)
Implementation wise, it's more complicated than I'd like due to needing
to read in both non-animated and animated data for each property. And
because I've tried to reduce the duplication of various transforms we
have to do on each value. E.g. scaling values by "tenth of scene units"
or the extraction of the USD clipping range from a GfVec2f into 2
separate properties, and 2 separate fcurves, in Blender etc. The current
approach was the best I could come up with so far.
Aperture sizes remain problematic for import, with animation data and
without, due to how Blender selects the largest sensor dimension to base
downstream calculations on and for which there's no concept in USD to
strictly dictate which dimension to use. Additionally, changing the
sensor size will impact the Tilt values as well. This means that if the
Aperture sizes are animated, we must also animate the tilt values; leads
to more fcurves being created than perhaps expected.
The `projection` attribute (perspective and orthographic) remains
unchanged (non animated only) due to differences in how USD<>Blender
interoperate with the Orthographic projection method. Note: Blender only
exports perspective cameras due to the same reason.
Pull Request: https://projects.blender.org/blender/blender/pulls/137487
The `USE_EXPERIMENTAL_TESTS` variable was not exposed as an option, so
the two tests that use the option to check whether tests should be run
had to be manually enabled by changing `tests/python/CMakeLists.txt`.
This commit renames the variable to `WITH_TESTS_EXPERIMENTAL`, defaults
the option to `OFF`, and marks it as an advanced option.
Pull Request: https://projects.blender.org/blender/blender/pulls/133831
