Blender Git Commits

Fix T53686: VSE Render crash when zooming timeline.

Epic fail from recent 'security' fixes (rBe04d7c49dca9). ;)

To be backported to 2.79a!

Cycles: Added static asserts to make sure new kernel structures are properly aligned

Merge branch 'master' of git.blender.org:blender into cycles_refactor

Move int <-> float 2d vector funcs from GP code to Math Lib

We now have the following two new functions:
* copy_v2fl_v2i(float r[2], const int a[2]) <-- Copies 2D int vector to float vector
* round_v2i_v2fl(int r[2], const float a[2]) <-- Copies 2D float vector to int vector, by rounding the values

Merge branch '28' into temp-workspace-addons

Correct merge

Merge branch 'blender2.8' into temp-workspace-addons

Merge branch 'blender2.8' into temp-workspace-addons

Workspace addon reading

Merge branch 'blender2.8' into temp-workspace-addons

Fix T53962: Cycles OpenCL compile error in some scenes.

This part of 212a8d9e needed to be ported over for 2ca933f to work.

Workspace addons, initial commit

Basic working, no way for the UI to enable yet

Fix possible concurency issue in mesh normals computation.

Failure in own code from last December, thanks @sergey for finding it.

To be backported to 2.79a.

Merge branch 'blender2.8' into greasepencil-object

Cleanup: Use standard mathlib functions here, instead of inlined pointer magic/etc.

More self-contained fix for previous commit

Cleanup: Various fixes

* Remove redundant operations (e.g. extra copies of data)

* Remove duplicate functions (e.g. vector copying, for non-floats)
TODO: We still have copy_v2int_v2float() and copy_v2float_v2int()
to fix (i.e. rename + move to BLI_math.h)

* Cleanup redundant struct forward defs in headers (including some duplicates)

* Hacky fix for compiler warning in gpencil_colorpick.c (about const vs non-const)
UI_FSTYLE_WIDGET is const, but UI_text_clip_middle_ex() needs non-const

Cleanup: More simple fixes

Cleanup: Style fixes

Cleanup: Move ED_gpencil_stroke_minmax() to blenkernel

The rest of the boundbox/minmax code is already there. We might as well
have the function for grabbing this info from a single stroke there too.