Blender Git Loki

Workspace: SETLOOPER should get render layer from workspace

Right now we are still passing NULL instead of the workspace in most of
those cases. That shouldn't matter much since this will simply bring
back the previous behaviour.

Also, most of those calls (gpu_material.c, view3d_draw_legacy.c,
render_internal.c) are from deprecated areas of Blender anyways.

The areas that will need to be addressed are:
* physics (collision.c)
* particles (pointcache.c)
* sound (sound.c)

And of course:
* game engine (BL_BlenderDataCoversion.cpp)

It's not clear to me, however, why some of those areas are using the
SET_LOOPER in the first place. SET_LOOPER acts on top of bases so we
want to iterate over the SceneLayer objects, not all of scene ones.

However in some cases it seems that the original design was expecting to
loop over all the objects (+ the set objects).

Pass workspace for space listeners

Space Image needs this to get active layer to get active object.

Workspace: "use_scene_settings" flag

This option determine whether we get the engine, layer and render settings
from the workspace or from the scene.

It is useful to allow for quick preview of the final render (F12) while in
a workspace,

Merge branch 'id_override_static' into asset-engine

Conflicts:
source/blender/blenloader/intern/versioning_270.c

Cycles: Safer fix for infinite recursion

Previous fix wasn't working correct for certain compiler and CPU intrinsics
mode, causing quite some crashes.

This should be a safer fix, which is closer in behavior to previous release
but which should still fix issues with robust curve intersection.

Merge branch 'master' into id_override_static

Fix T52650:Grease pencil selection its not automatically updating in Clip Editor

Math Lib: normalized vector project functions

Fix T52149: LoopTriArray computation was not correctly protected against concurrency.

Note: this commit seems to work as expected (also with transform
snapping etc.). However, it is rather unsafe - not enough for 2.79 at
least, unless we get much more testing on it. It also depends on three
previous ones.

Note that using a global lock here is far from ideal, we should rather
have a lock per DM, but that will do for now, whole DM thing is doomed
to oblivion anyway in 2.8.

Also, we may need a `DM_DIRTY_LOOPTRIS` dirty flag at some point. Looks
like we can survive without it for now though... Probably because cached
looptris are never copied accross DM's?

Cleanup: remove useless `DM_ensure_looptri()`.

That one was doing exactly same thing as `dm->getLoopTriArray()`, no
point in having twice the same code here...

Fix transform snap code using 'allocated' flags to get verts/edges/etc. arrays again from DM.

This was... horribly wrong, CDDM will often *not* need to allocate
anything to return arrays of mesh items! Just check whether array
pointer is NULL.

Also, remove `DM_get_looptri_array`, that one is useless currently,
`dm->getLoopTriArray` will always return cached array (computing it if
needed).

Cleanup: deduplicate DM's getLoopTriArray() callback.

All three functions were doing exactly the same thing, simpler to only
have one in that case!

Fix T52639: Weight paint smooth tool crash

UI: fix memory leak when copy-to-selected failed

Fix T52678: Crash editing gpencil w/ frame-lock

Screw Modifier: remove doubles option

Vertices on the axis can be optionally merged,
nice for creating objects which close at the end-points.

View3D Remap: skip defmaterial

This isn't library data.

Fix T52663: Remap used invalid local-view data

Only the camera from View3D.localvd is used,
other pointers may be invalid.

Longer term we should probably clear these to ensure no accidents.
For now just follow the rest of Blender's code and don't access.

Fix T52522: VSE renders with alpha transparent PNG image incorrectly

Need some extra checks and should be probably end up in 2.79 since that's a regression.

Merge branch 'blender2.8' into greasepencil-object