Blender Git Commits

Code cleanup

And checking whether mail indeed works now.

Speedup for guarded allocator

- Re-arrange locks, so no actual memory allocation
(which is relatively slow) happens from inside
the lock. operation system will take care of locks
which might be needed there on it's own.

- Use spin lock instead of mutex, since it's just
list operations happens from inside lock, no need
in mutex here.

- Use atomic operations for memory in use and total
used blocks counters.

This makes guarded allocator almost the same speed
as non-guarded one in files from Tube project.

There're still MemHead/MemTail overhead which might
be bad for CPU cache utilization.

TODO: We need smarter 32/64bit compile-time check,
currently i'm afraid only x86 CPU family is
detecting reliably.

Merging r59153 through r59161 from trunk into soc-2013-depsgraph_mt

Compilation error fix when using scons

OpenEXR libraries shall go after OpenImageIO,
otherwise it'll fail to find symbols needed
for OIIO.

== filebrowser ==
* FIX: selecting bookmark or recent file caused UNDO push, also caused issue with the 'Open' and 'Cancel' buttons being grayed out.

Reported by Sergey Sharybin in IRC, many thanks.

Added detailed timing information output

Used to visualize CPU threads activity during threaded
object update.

It prints whole bunch of timing information to STDERR
which then could be parsed and nicely visualized.

Detailed statistics is disabled by default.

File to parse and visualize statistics:
http://blender.nazg-gul.dyndns.org/depsgraph/threaded_update_stats.blend

And screenshot for fun:
http://blender.nazg-gul.dyndns.org/depsgraph/threaded_update_stats.blend.png

Skip objects which doesn't need update from update statistics

Merging r59136 through r59152 from trunk into soc-2013-depsgraph_mt

Fix #36384: Inconsistent curve fill behaviour

Creating new splines via Python API didn't take
curve dimension into account.

Now adding new splines will set 2D/3D flag for
spline according to curve's dimension.

Blender might be compiled without guardedalloc again

This is useful for benchmark tests, to make CPU cache
utilization as good as we could with current design.

fix [#36455] importing obj data after saving doesn't work

BGE: The Action Actuator can now make use of additive blending.

BGE: Finally adding support for additive layer blending.

Currently this is only for the Python API. The logic brick will be updated in a future commit.

Adding a new state for sound handles in audaspace: stopped.

Now sounds that stopped playing but are still kept in the device can be differentiated from paused sounds with this state.
This should also fix the performance issues mentioned in [#36466] End of SequencerEntrys not set correctly.
Please test if sound pausing, resuming and stopping works fine in the BGE and sequencer, my tests all worked fine, but there might be a use case that needs some fixing.

Fix #36467: ends of curves - 2D differs to 3D

This is getting confused, why bevel is handled different
for 2D and 3D curves?

Anyway, made bevel work for 2D case nice again, but it's
probably nice to unify 2D and 3D cases.

Cycles:
* Remove an unneeded return in shader_bsdf_eval(), function is void.
* if / else if tweak in shader_bsdf_ao().

fix [#36349] Separate mesh by material creates meshes with all the materials from the original

add materials.clear() method, matching other python list method.

quiet compiler warnings.

Merging r59130 through r59135 from trunk into soc-2013-depsgraph_mt