Gooseberry farm: Correction to previous commit

Initial patch was done against imbuf with ftype refactor applied.

Jpeg2000: Fixed for DCP pipeline

Profile and color space were not properly set.

Conflicts:
source/blender/imbuf/intern/jp2.c

Cycles: Report to the console when the samples offset feature is used

Cycles: Implement sample offset feature

This is real quick implementation of the feature which allows to render
different sample ranges on different machines and merge the result into
final image as if it was rendered on a single machine. Same is feature
makes it possible to refine animation with more samples after it was
rendered with lower number of samples.

Control is happening via the following environment variables:

- CYCLES_RENDER_SAMPLE_START, which defines which sample to start
rendering from.

- CYCLES_RENDER_NUM_SAMPLES, which defines how much samples are to be
sampled by the Blender instance.

Limitations:

- Not really intuitive, final implementation will require better
integration into command line arguments.

- All the render layers are to have the same exact number of samples,
meaning using samples override for layer will lead to undesired
results.

- No compositor is allowed.

Fix for broken particle stuff when using particle instance modifier with
children.

Cycles: Respect duplicator's object motion blur settings

The idea is to make it possible to control linked duplicated objects motion
blur from the scene file without need to do overrides on the linked object
settings. Currently only supported for dupligroup duplication and all now
if duplicator object has motion blur disabled then it'll be inherited into
all the duplicated objects.

There should be no regressions/changes in look of existing files because
objects do have motion blur enabled by default.

Gooseberry farm: Enable ALembic on windows platform

Also do some nnoying fixes for stupidness of MSVC which
was searchinh for symbols in the wrong namespaces.

Should be backported to Alembic or Gooseberry branches
but that'd happen from another operation system on this
laptop,

Cycles: Report currently sampling tile when CPU is working on the last tile

This is mainly useful for the render farms output when logging might stop at
the "Path Tracing Tile N/N" string, which makes it a bit difficult to follow
what exactly is happening (node going crazy, hardware issues or just last tile
is too much heavy).

This is more like an experiment, might be changed in the future.

Return non-zero exit code when running blender from the command line and reading file has failed

This way automated scripts can actually see if some issue happened.

Msgfmt: Fix for generating .mo files from .po without comments

Report proper frame and time to the console when doing compositing

Cleanup, use define instead of magic number

Print elapsed time when rendering from the command line

Flush stdout prior of calling render stats callback

Without this extra flush order of stat prints is undefined in the
output. which makes it rather tricky to write custom output in a
reliable way.

Cycles: Merge density and color volume attributes if they both requested

This way we store quite reasonable amount of memory (density is stored as RGBA
textures, which ends up in gigabytes of RAM for hires smoke sims).

It's not really pretty implementation and we'll really need to support textures
with 1 and 3 channels instead of doing such packing, but we need to be able to
render some files here.

Cycles: Free hires smoke data from blender side after scene was fully synchronized

This happens if all object's smokes are backed to files and rendering is either
happening from command line of with locked interface.

The idea behind this change is to make as much memory available for path tracing
as possible, so after heavy swapping during synchronization period there's loads
of free memory to put important data back to RAM.

Gooseberry: Use more proper flag for psys reconstruction after memory optimization

Gooseberry: Fix for flickering grass

The issue was caused by memory optimization marking particle system to recalc,
and because of the way how particle flags works it was possible that it'll
cause particle's re-distribution. Now this memory optimization will act the
same as loading the file.

Would need to review if this fix is to go to master/cycles_memory branches
but for now it'll be cool to figure out what to do with the farm.

Fix outdated stub factory class for pointcache when Alembic is disabled.

Gooseberry Farm: Solve compilation error with strict flags