Blender Git Loki

Merge branch 'blender2.8' into transform-manipulators

Conflicts:
intern/gawain/gawain/immediate.h
intern/gawain/src/immediate.c
source/blender/editors/physics/physics_ops.c
source/blender/editors/screen/glutil.c
source/blender/editors/space_view3d/space_view3d.c
source/blender/editors/space_view3d/view3d_draw.c
source/blender/editors/space_view3d/view3d_edit.c
source/blender/editors/space_view3d/view3d_ops.c
source/blender/editors/transform/transform_manipulator.c

Collada export cleanup

fix: Collada export selected worked only for the very first export (needs further testing and cleanup, see comments)

Immediate Mode: handle other cases of glPop/glPushClientAttrib

Those cases were not using ClientAttrib but they should :)
Rather use a OpenGL3.3 alternative anyways.

Fix collada exporting for Blender 2.8

Instead of exporting all the scene objects, I'm exporting the scene
layer objects instead.

FIXUP stupid rotation mode y=up bug

Cleanup: redundant casts

Alembic import: fixed crash on more complex model

There was a problem with parent-child relations not getting set up
correctly when an Alembic object was both the transform for a mesh object
and the parent of other mesh objects.

Alembic: comment typo/error fixes

install_deps: removed leftover compile_HDF5 command

It was a leftover from when Alembic with HDF5 was still officially
supported.

Immediate Mode: removing unecessary comment on gpencil (false positive)

Image Draw: remove unused code since 2009

(removing false positive deprecated OpenGL calls)

Some more cleanup, get rid of hack to protect against NULL RNA pointerprops.

Merge branch 'master' into datablock_idprops

Fix missing protection of `RNA_pointer_as_string()` against NULL pointers.

Odd that issue was never reached before? Looks like it hit in
datablock_idprops branch though...

Minor changes

- Remove utildefines from BKE_workspace header
- doxy example are typically for code examples, just use heading
- wrap args, avoid right shift

Merge branch 'master' into blender2.8

Cycles: Tag shaders for update after the threading part is over

This avoids write access happening in non-atomic manner in
Shader::tag_update which modifies the global managers. Even
for 1 byte data types it's quite dangerous.

Depsgraph: Use atomic operation to tag the changed ID

Cycles: Fix uninitialized memory access when comparing curve mapping nodes

The issue is coming from the fact that float3 is actually 16 bytes aligned
data type and the "padding" was not initialized. This caused memcmp() to
access non-initialized memory.