Blender Git Commits

Fix Cycles assert after recent half changes.

Fix return type in case of single channel half.

Cycles: Support half and half4 textures.

This is an initial commit for half texture support in Cycles.
It adds the basic infrastructure inside of the ImageManager and support for these textures on CPU.

Supported:
* Half Float OpenEXR images (can be used for e.g HDRs or Normalmaps) now use 1/2 the memory, when loaded via disk (OIIO).

ToDo:
Various things like support for inbuilt half textures, GPU... will come later, step by step.

Part of my GSoC 2016.

Fix broken combine XYZ node after recent refactoring.

Cleanup: style, whitespace, doxy filepaths

Fix T48671: Cycles crash with OSL nodes and bump after recent refactoring.

Fix T46492: GGX distribution produces black pixels

The issue was caused by some numerical instability.

Fix T48667: Bisect-fill crash

BMO iterator would loop over removed faces.

Recent changes to mempool FREEWORD size exposed this bug.

Fix T48662: Blender crash using node groups connected with displacement output in some situations

Fix T48657: Incorrect render with Cycles (CPU) using texture

The sockets of the RGB to BW node were set to the wrong type after the recent node refactor.

Fix/cleanup BKE libquery's ID looper.

Some area were still not in sync with readfile.c, now should be better.

Note that readfile.c has been used as r�f�rence here re us refcounting,
not sure how accurate it is, time will say :|

readfile.c: fix some wrong usages of newlibadr_us.

There are most likely some more still, but think this should now be inline with
libquery looper...

Readfile cleanup: add new newlibadr_real_us helper.

Vertex paint color operations

D2050 by @metaraptor with edits

Adds levels, brightness/contrast, hsv & invert operations.

Correct error in recent commit

Fix some rare asserts with new simple/random particle distribution code.

Optimization in binary search could lead to equality instead of expected strictly greater than value.
Harmless but noisy, and better be strict here.

reported by sergey on irc (with koro cycles benchmark file), thanks.

BMesh Decimate: use doubles to calculate optimized position

This allows the error threshold for calculating the optimized location to be much lower.

Resolves visible artifacts w/ 1m-tri happy-buddha example.

BLI_math: Add double versions of functions

- mul_v3_m3v3_db
- mul_m3_v3_db
- negate_v3_db

Fix ruler crash using stale snap-context

Since ruler allows other operators to run (such as mode-switching).
Only cache snap-context while dragging.

BMesh Decimate: support ngons