Blender Git Commits

Prevent adding objects to be added to a group which is used as dupli-group for this object
when using "Add Selected to Active Group" operator,

Cleanup only:
Comments clarified, whitespaces adjusted for consistency, unneccesary code removed.

remove leftover files from last commit

revert 45924, not a very clean solution, especially for external libraries and looks like -fpermissive is used in linux too

svn merge ^/trunk/blender -r44562:HEAD

Merging r45877 through r45927 from trunk into soc-2011-tomato

BVHTree accellerated: ED_vgroup_copy_by_closest_single

Cycles:
* Fixing wrong windows define, causing compile errors.

MinGW-w64 enable full optimization for C++ too

Remove mingw-w64 errors from loss of precision by converting 64bit pointers to ints. All cases found were harmless and the error behaviour could be turned off by the -fpermissive flag but I'd rather keep that off to detect any real problems should they arise.

Add libMV and Scons support for MinGW-w64, patches by Caleb Joseph with slight modifications.

Thanks!

crash fix for brush select when no object is active

Cycles: cuda build fix.

Cycles: BVH build time optimizations:

* BVH building was multithreaded. Not all building is multithreaded, packing
and the initial bounding/splitting is still single threaded, but recursive
splitting is, which was the main bottleneck.

* Object splitting now uses binning rather than sorting of all elements, using
code from the Embree raytracer from Intel.
http://software.intel.com/en-us/articles/embree-photo-realistic-ray-tracing-kernels/

* Other small changes to avoid allocations, pack memory more tightly, avoid
some unnecessary operations, ...

These optimizations do not work yet when Spatial Splits are enabled, for that
more work is needed. There's also other optimizations still needed, in
particular for the case of many low poly objects, the packing step and node
memory allocation.

BVH raytracing time should remain about the same, but BVH build time should be
significantly reduced, test here show speedup of about 5x to 10x on a dual core
and 5x to 25x on an 8-core machine, depending on the scene.

Cycles: centralized task scheduler for multithreading, which is basically the
CPU device threading code wrapped into something reusable.

Basic idea is that there is a single TaskScheduler that keeps a pool of threads,
one for each core. Other places in the code can then create a TaskPool that they
can drop Tasks in to be executed by the scheduler, and wait for them to complete
or cancel them early.

fix [#31079] Revision 45893 crashes when I try to open a certain file created in it.

nasty bug going from 2.63 --> 2.62 --> 2.63, could have incorrect/corrupt data.
fix checks for this case and clears the customdata layer.

Add cycles support for MinGW-w64

fix for possible error building with debian

fix [#30997] Bevel angle option is broken (bevel vertex only)

Peer pressure :) Fix some very public (but probably harmless) errors in extern/bullet2, it will propagate to the Bullet soon from here:
https://www.assembla.com/code/bullet3/subversion/nodes
Thanks to Campbell for letting me know
Fixed described by Sean here:
http://stackoverflow.com/questions/818535/how-can-i-set-all-bits-to-1-in-a-binary-number-of-an-unknown-size