Issue was caused by do_versions being used pdata as reference for active/render/
stencil/clone layer indices instead of fdata.
Added some utility functions used only by do_versions to be sure this indices
are set from fdata for pre-bmesh files.
When in material display mode, mesh_calc_modifiers() calculates the
orco DerivedMesh, which uses a different CustomDataMask. In
particular, it does not necessarily include the current modifier's
requiredDataMask, so those layers might get set to NO_COPY. For the
skin modifier, this resulted in a crash when the modifier internally
copies the DerivedMesh and the output does not contain the expected
MVertSkin layer.
Fixed by adding the requiredDataMask to the orco DM's CustomDataMask.
Also added a debugging function to customdata.c:
customData_mask_layers__print(CustomDataMask mask);
This will print out the names of all the CD layer types in the mask.
* The operator creates bones for each input edge (does not subdivide
them like the skin operator does), adds a fake root bone for skin
roots with multiple children.
* The operator adds vertex weight groups to the original mesh.
* Make copy_object_transform() public, used to match the armature
object to the mesh object.
Skin modifier documentation:
http://wiki.blender.org/index.php/User:Nicholasbishop/SkinModifier
CD_PAINT_MASK is a layer of per-vertex floats for non-multires
meshes. Multires meshes use CD_GRID_PAINT_MASK, which is a layer of
per-loop GridPaintMask structures. GridPaintMask is similar to MDisp,
but contains an array of scalar floats.
Note: the GridPaintMask could be folded into MDisp, but this way
should be easier to add mask layers in the future (if we do fold
GridPaintMask into MDisp, the mask array should probably be an array
of arrays with a 'totmask' field so that mask layers can be easily
supported.)
Includes blenload read/write support for CD_PAINT_MASK and
CD_GRID_PAINT_MASK.
In fact there were several issues fixed (all of them regressions since bmesh merge):
- Creating navmesh crashed because creating new faces for mesh was trying to set
default values for all customdata layers in this face. This requires memory
pool created for this datablock.
Usually this pool is creating on creating datablock if there're some elements
to be stored in this block. In cases of regular primitive creating it wasn't
an issue because they doesn't create customdata layers, they only creates
geometry.
Navigation mesh creates geometry and customdata layers (CD_RECAST layer)
which used to confuse a bit custom data functions. Solved by ensuring there's
memory pool created for polygons datablock after adding new custom data layer.
Most probably it's better to be resolved on CD level (like smarter track on
changed amount of stored data and so) but prefer not to make such global changes
so close to the release.
- Toggling edit mode lead to loosing recast datalayer. Solved by adding recast
layer to bmesh mask so it'll be copied to/from edit mesh.
- Some part of code assumed raycast layer is in face datablock, some that it's in
polygon datablock. Made it to be in polygons datablock.
Kind of temporary solution to make navmesh working, probably it'll fail if one
will want to edit navmesh by hand after it was generated.
Proper way would be to ensure the whole navmesh things are using ngons.
Another crash with array caps, was caused by not making a deep enough
copy of CD field.
Also fixed the type of the 'mask' parameter, was int where it should
be 64-bit.
When merging bmesh customdata, memset the new blocks to
zero. Otherwise can contain junk values (some layer types with
pointers like MDeformVert can then crash.)
* Skip calculation of merge indices if merging isn't enabled
* Clean up usage of BMesh operators to fix small memory leak
* Fix harmless BLI_assert in CustomData_bmesh_merge
* Another null-initialization fix in CustomData_bmesh_merge
old mesh MCol 'r' was blue, 'b' was red, but theres no reason to keep this for bmesh with MLoopCol.
Loading old files works, saving legacy format works too.
What wont work is loading a file after this revision and loading it in an older revision since the bmesh merge.
(it wont crash but the blue and red will be swapped on vertex color layers).
* double default edge allocation size (double the number of verts/faces).
* CustomData_bmesh_init_pool was using allocsize & chunksize as the same variable. Now use type specific chunk size.
* bmesh copy and editmode conversion now allocate the BMesh mempool size needed for the entire vert/edge/loop/face arrays since its known already.
