Age | Commit message (Collapse) | Author |
|
|
|
Choosing a UV layer would actually affect the overlay in the viewport
and also painting with the mask brush was in that UV space, but the
resulting stencil mask was always applied with the active UV (not the
explicitly selected stencil UV -- the one one is looking at in the
viewport!) to painting.
This has been like that as far as I have checked back (at least 2.79b),
I am surprised this has not come up before, but it does not seem to make
sense at all...
Now use the UV specified for the stencil layer when applying the mask for
painting, so it corresponds to the stencil mask one is looking at in the
viewport.
Maniphest Tasks: T91557
Differential Revision: https://developer.blender.org/D12583
|
|
The resolution for surfaces was 12 for U, 4 for V,
where both should have been set to 4.
Regression in 9a076dd95a01135ea50f9ccc675668db9f2155f4
|
|
|
|
Follow up to fix for T89868.
|
|
Mistake in a30a8179331d689c9e599fb9a530c0b6b155f689.
|
|
This was working differently in 2.79, tried tracking this down and it
seems this was wrong since the 2.8 beginning in {rB7907dfc40018}.
This would not only crash without an active scene camera, but would also
result in different tracks from different camera's constraints could not
be selected.
So select id depends on corresponding camera, remove the dependency on
scene camera completely.
Maniphest Tasks: T90651
Differential Revision: https://developer.blender.org/D12230
|
|
Reading paths over 512 bytes would cause a buffer overrun.
|
|
|
|
|
|
It wasn't obvious this didn't free the memory of the mesh it's self
leading to memory leaks.
|
|
|
|
|
|
Those were caused by various tools used on degenerate geometry, see
T79775.
Note that fixes are as low-level as possible, to ensure they cover as
much as possible of unreported issues too.
We still probably have many more of those hidden in BLI_math though.
|
|
Code freeing the array would not properly reset its length value to
zero.
Note that this corrupted data could also be saved in .blend files, so
had to bump fileversion and add some doversion code too.
Fix T90166: crash when creating a liboverride.
|
|
cases.
This was apparently done in two places only, with a very cryptic comment
(`/* for raw_access, untested */`), and... I cannot see how returning a
non-zero length value for an array that does not exist or is not
accessible at least, would be anything but an obvious source of issues.
Note that both commits adding those lines are from stone ages (2009):
rBcbc2c1886dee and rB50e3bb7f5f34.
|
|
local.
Even though the ID itself remain the same after being made local, from
depsgraph point of view this is a different ID. Hence we need to tag all
of its users for COW update, as well as rebuild depsgraph relationships.
Should be also backported to LTS 2.93 (and 2.83 if possible).
|
|
Only the "changed" state from the last edit-object was used,
this meant the operator would not perform the necessary update
with multi-object edit-mode.
Use "changed" & "changed_multi" naming convention.
|
|
Also repair any errors in existing files.
Error from e0dd3fe5872ba37ff188e292b80b46fcf8df413c.
|
|
Slicing with indices greater than the length of the sequence would crash.
|
|
Negative indices that remained negative after adding the sequence length
caused incorrect slicing.
With the default scene for example:
bpy.context.scene.objects[-4:2]
Gave a different result to:
tuple(bpy.context.scene.objects)[-4:2]
Clamp indices above zero so loops that step forward works as intended.
|
|
Only vertex indices were ensured to be correct.
|
|
The crash occurred calling because mesh_get_eval_final in edit-mode
freed all derived mesh data without tagging the object for updating.
However meshes in edit-mode weren't meant to be used as knife-project
source-data, adding support for multi object edit-mode caused this.
|
|
Disable wrapping when "scale to fit" is used, assert the error is
small so an invalid scale-to-fit value wont go by unnoticed.
|
|
|
|
|
|
File is replaced by pipeline_config.yaml.
|
|
Failure to return a list of the expected size & type wasn't
decrementing the value, leaking a reference.
Caused by 127b5423d6203d521acb2b96b7de5534e8dbe79a a workaround for the
real error that was fixed f5e020a7a6ad6451fcaf075ae14f7014b8a4faea.
|
|
This fixes a performance regression on Ampere cards, on specific scenes like
classroom. For cycles-x there is little difference, but this is still helpful
for LTS releases, and we need to upgrade at some point anyway.
|
|
|
|
|
|
|
|
|
|
|
|
The old code was added in {rBbaa4a9c7d4dd}
While the hotfix committed in {rB2ddecfffc3d3a3a1db4ae45e8665ca} fixed
the id_management test, the correct reason was found at
https://bugs.llvm.org/show_bug.cgi?id=47984
This matches the behavior on Linux.
|
|
Regression in d25747ee751096de2e417a7da1316bf5bf81c25a
|
|
Also resolve a crash when when displaying thumbnails, see T89868.
|
|
Scaling down images could create images with a width or height of zero.
Clamp at 1 to prevent a crash, also add an assert to scaling functions.
Ref D11956
|
|
AMD Drivers didn't report an additional space in the renderer. This made
testing for the HQ workaround fail and the issue appeared back on
certain cards.
This fix will test with surrounding spaces or if the renderer name
endswith the given string. If any of these are the case the hq normals
workaround will be enabled.
Original patch {2262d6c45adf}.
|
|
`bm_face_len_cmp` incorrectly interpreted its arguments as `BMFace *`
instead of `BMFace **`, causing an out-of-bounds read.
Ref D11637
|
|
Pre computed normals index wasn't properly aligned.
Regression from 2ec00ea0c1be1ace7cd0c7b68e43cc8e87dd07c7.
|
|
|
|
|
|
Apparently `textureSize` doesn't work with
`sampler1DArray` on this OS.
Thanks to @dave1853 for finding the source of the
problem.
|
|
Due to the way we ship the CRT on windows TBB's
malloc proxy was unable to attach it self to
the memory management functions on windows 10.
This change moves ucrtbase.dll out of the blender.crt
folder and back into the main blender folder to side
step some undesirable behaviour on win10 making TBB
once more able to attach it self.
Having this work again, should give a speed
boost in memory allocation heavy workloads
such as mantaflow.
For details on how this only failed on Win10
see T88813
|
|
Error in 87cafe92ce2f99d8da620b80e1c26f8078554f93
|
|
|
|
Enabled HQ normals workaround for this specific configuration.
|
|
The id-property iterator referenced a PyObject pointer without
increasing it's user count - allowing for errors if the value
goes out of scope during iteration.
|
|
Looping over IDProperty keys or calling iteritems leaked memory.
Error in original Python idproperty support from
8768707610fbc1cea2bde069cdfd6d3f3e2fc522
|