Blender Git Commits

Cycles: Attempt to workaround compilation error on new CUDA toolkit and sm_2x

Prevent crash when seeing the window new on keymaps

CTX_data_main(C) is NULL in those cases :/

Merge branch 'master' into blender2.8

Remove non-bmesh case from test

Fix skin mark operator

Accessed custom-data layer offset before creating.

Fix crash closing window in background mode

Fix missing NULL check in gpencil poll

Also de-duplicate poll functions

Use 'empty' option for clearing factory settings

Option to load startup file with empty-data

Useful for batch conversion and tests.

Buildbot: Revert previous change, older toolkit has same exact behavior

Buildbot: Use older NVCC on 32bit linux

Newer toolkit has some weird issue with cross0-compiling 32bit kernels
from 64bit environment.

Buildbot: Remove global hardcoded NVCC path

This was initially needed for heterogeneous setup of two toolkits which
we no longer need.

Cycles: Bring back preview AA samples when using BPT

This was removed in 93426cb. Please be more accurate when
changing interface.

PyAPI: minor path init simplification

Fix 'bl_app_override' wrapping multiple times.

Calling `SomeClass.draw(self, context)` instead of `self.draw()`
would try to wrap the argument `self` multiple times, causing an error.

Fix memory leak re-registering operators

Re-registering an operator used by the keymap would lead memory.
Reload scripts for eg leaked over ~1600 blocks.

PyAPI: check modules are registered before unregister

Needed since templates may unregister classes.

Also replace old modules on reloading.

Merge branch 'master' into blender2.8

Fix T51068: Place props in their own row

This allows the props to extend into the blank space that is to the right.

Cycles: Switch to reformulated Pluecker ray/triangle intersection

The intention of this commit it to address issues mentioned in the
reports T43865,T50164 and T50452.

The code is based on Embree code with some extra vectorization
to speed up single ray to single triangle intersection.

Unfortunately, such a fix is not coming for free. There is some
slowdown for AVX2 processors, mainly due to different vectorization
code, which caused different number of instructions to be executed
and different instructions-per-cycle counters. But on another hand
this commit makes pre-AVX2 platforms such as AVX and SSE4.1 a bit
faster. The prerformance goes as following:

2.78c AVX2 2.78c AVX Patch AVX2 Patch AVX
BMW 05:21.09 06:05.34 05:32.97 (+3.5%) 05:34.97 (-8.5%)
Classroom 16:55.36 18:24.51 17:10.41 (+1.4%) 17:15.87 (-6.3%)
Fishy Cat 08:08.49 08:36.26 08:09.19 (+0.2%) 08:12.25 (-4.7%
Koro 11:22.54 11:45.24 11:13.25 (-1.5%) 11:43.81 (-0.3%)
Barcelone 14:18.32 16:09.46 14:15.20 (-0.4%) 14:25.15 (-10.8%)

On GPU the performance is about 1.5-2% slower in my tests on GTX1080
but afraid we can't do much as a part of this chaneg here and
consider it a price to pay for more proper intersection check.

Made in collaboration with Maxym Dmytrychenko, big thanks to him!

Reviewers: brecht, juicyfruit, lukasstockner97, dingto

Differential Revision: https://developer.blender.org/D1574