2011-02-23 11:52:22 +01:00
|
|
|
/*
|
2005-03-27 22:34:18 +02:00
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU General Public License
|
|
|
|
* as published by the Free Software Foundation; either version 2
|
2008-01-07 20:13:47 +01:00
|
|
|
* of the License, or (at your option) any later version.
|
2005-03-27 22:34:18 +02:00
|
|
|
*
|
|
|
|
* This program is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License
|
|
|
|
* along with this program; if not, write to the Free Software Foundation,
|
2010-02-12 14:34:04 +01:00
|
|
|
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
|
2005-03-27 22:34:18 +02:00
|
|
|
*
|
2005-03-29 18:43:39 +02:00
|
|
|
* The Original Code is Copyright (C) 2005 Blender Foundation.
|
2005-03-27 22:34:18 +02:00
|
|
|
* All rights reserved.
|
|
|
|
*/
|
|
|
|
|
2019-02-17 22:08:12 +01:00
|
|
|
/** \file
|
|
|
|
* \ingroup bke
|
2011-02-27 21:40:57 +01:00
|
|
|
*/
|
|
|
|
|
|
|
|
|
2005-08-23 22:04:10 +02:00
|
|
|
#include <string.h>
|
2012-04-06 07:53:01 +02:00
|
|
|
#include <limits.h>
|
2005-07-23 21:15:08 +02:00
|
|
|
|
2005-03-27 22:34:18 +02:00
|
|
|
#include "MEM_guardedalloc.h"
|
|
|
|
|
2010-03-30 13:49:07 +02:00
|
|
|
#include "DNA_cloth_types.h"
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
#include "DNA_customdata_types.h"
|
2006-08-28 03:12:36 +02:00
|
|
|
#include "DNA_key_types.h"
|
2014-07-21 12:02:05 +02:00
|
|
|
#include "DNA_material_types.h"
|
2012-02-19 23:17:30 +01:00
|
|
|
#include "DNA_mesh_types.h"
|
2005-03-27 22:34:18 +02:00
|
|
|
#include "DNA_meshdata_types.h"
|
|
|
|
#include "DNA_object_types.h"
|
2014-04-27 18:22:03 +02:00
|
|
|
#include "DNA_scene_types.h"
|
2005-03-27 22:34:18 +02:00
|
|
|
|
Fix T47038: Particles in Particle Edit Mode get added in completely wrong location.
It also fixes another issue (crash) related to symmetric editing.
Quite involved, we (try to!) fix complete broken logic of parts of particle code, which would use poly index
as tessface one (or vice-versa). Issue most probably goes back to BMesh integration time...
This patch mostly fixes particle editing mode:
- Adding/removing particles when using generative modifiers (like subsurf) should now work.
- Adding/removing particles with a non-tessellated mesh (i.e. one having ngons) should also mostly work.
- X-axis-mirror-editing particles over ngons does not really work, not sure why currently.
- All this in both 'modes' (with or without using modifier stack for particles).
Tech side:
- Store a deformed-only DM in particle modifier data.
- Rename existing DM to make it clear it's a final one.
- Use deformed-only DM's tessface2poly mapping to 'solve' poly/tessface mismatches.
- Make (part of) mirror-editing code able to use a DM instead of raw mesh, so that we can mirror based on final DM
when editing particles using modifier stack (mandatory, since there is no way currently to find orig tessface
from an final DM tessface index).
Note that this patch is not really nice and clean (current particles are beyond hope on this side anyway),
it's more like some urgency bandage. Whole crap needs complete rewrite anyway,
BMesh's polygons make it really hard to work with current system (and looptri would not help much here).
Also, did not test everything possibly affected by those changes, so it needs some users' testing & validation too.
Reviewers: psy-fi
Subscribers: dfelinto, eyecandy
Maniphest Tasks: T47038
Differential Revision: https://developer.blender.org/D1685
2016-01-04 12:19:45 +01:00
|
|
|
#include "BLI_array.h"
|
2005-03-27 23:27:12 +02:00
|
|
|
#include "BLI_blenlib.h"
|
2014-09-04 09:53:10 +02:00
|
|
|
#include "BLI_bitmap.h"
|
2010-03-22 12:59:36 +01:00
|
|
|
#include "BLI_math.h"
|
2011-01-07 19:36:47 +01:00
|
|
|
#include "BLI_utildefines.h"
|
2011-10-22 03:53:35 +02:00
|
|
|
#include "BLI_linklist.h"
|
2016-04-26 10:43:02 +02:00
|
|
|
#include "BLI_task.h"
|
2005-03-27 22:34:18 +02:00
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
#include "BKE_cdderivedmesh.h"
|
2017-12-07 05:36:26 +01:00
|
|
|
#include "BKE_colorband.h"
|
2013-04-18 19:09:56 +02:00
|
|
|
#include "BKE_editmesh.h"
|
2007-10-31 14:56:07 +01:00
|
|
|
#include "BKE_key.h"
|
2017-04-18 14:55:31 +02:00
|
|
|
#include "BKE_layer.h"
|
2015-11-09 19:47:10 +01:00
|
|
|
#include "BKE_library.h"
|
2014-04-09 03:03:25 +02:00
|
|
|
#include "BKE_material.h"
|
2007-10-31 14:56:07 +01:00
|
|
|
#include "BKE_modifier.h"
|
2005-03-27 22:34:18 +02:00
|
|
|
#include "BKE_mesh.h"
|
2018-10-09 07:52:46 +02:00
|
|
|
#include "BKE_mesh_iterators.h"
|
2014-01-21 16:32:36 +01:00
|
|
|
#include "BKE_mesh_mapping.h"
|
2018-06-05 15:59:53 +02:00
|
|
|
#include "BKE_mesh_runtime.h"
|
2017-05-24 14:33:21 +02:00
|
|
|
#include "BKE_mesh_tangent.h"
|
2005-03-27 22:34:18 +02:00
|
|
|
#include "BKE_object.h"
|
2012-09-05 05:45:32 +02:00
|
|
|
#include "BKE_object_deform.h"
|
2009-08-15 21:48:50 +02:00
|
|
|
#include "BKE_paint.h"
|
2011-01-31 21:02:51 +01:00
|
|
|
#include "BKE_multires.h"
|
2009-05-23 05:24:15 +02:00
|
|
|
#include "BKE_bvhutils.h"
|
2011-12-07 08:13:33 +01:00
|
|
|
#include "BKE_deform.h"
|
2005-03-27 22:34:18 +02:00
|
|
|
|
2013-05-28 21:35:26 +02:00
|
|
|
#include "BLI_sys_types.h" /* for intptr_t support */
|
2008-08-17 19:08:00 +02:00
|
|
|
|
2018-04-16 20:45:43 +02:00
|
|
|
#include "DEG_depsgraph.h"
|
|
|
|
#include "DEG_depsgraph_query.h"
|
Shrinkwrap: new mode that projects along the target normal.
The Nearest Surface Point shrink method, while fast, is neither
smooth nor continuous: as the source point moves, the projected
point can both stop and jump. This causes distortions in the
deformation of the shrinkwrap modifier, and the motion of an
animated object with a shrinkwrap constraint.
This patch implements a new mode, which, instead of using the simple
nearest point search, iteratively solves an equation for each triangle
to find a point which has its interpolated normal point to or from the
original vertex. Non-manifold boundary edges are treated as infinitely
thin cylinders that cast normals in all perpendicular directions.
Since this is useful for the constraint, and having multiple
objects with constraints targeting the same guide mesh is a quite
reasonable use case, rather than calculating the mesh boundary edge
data over and over again, it is precomputed and cached in the mesh.
Reviewers: mont29
Differential Revision: https://developer.blender.org/D3836
2018-11-06 19:04:53 +01:00
|
|
|
#include "BKE_shrinkwrap.h"
|
2018-04-16 20:45:43 +02:00
|
|
|
|
2019-02-01 02:44:19 +01:00
|
|
|
#include "CLG_log.h"
|
|
|
|
|
2015-08-03 20:35:43 +02:00
|
|
|
#ifdef WITH_OPENSUBDIV
|
|
|
|
# include "DNA_userdef_types.h"
|
|
|
|
#endif
|
|
|
|
|
2013-09-04 03:29:34 +02:00
|
|
|
/* very slow! enable for testing only! */
|
2014-12-02 13:49:40 +01:00
|
|
|
//#define USE_MODIFIER_VALIDATE
|
2013-09-04 03:29:34 +02:00
|
|
|
|
|
|
|
#ifdef USE_MODIFIER_VALIDATE
|
|
|
|
# define ASSERT_IS_VALID_DM(dm) (BLI_assert((dm == NULL) || (DM_is_valid(dm) == true)))
|
2018-05-11 09:53:52 +02:00
|
|
|
# define ASSERT_IS_VALID_MESH(mesh) (BLI_assert((mesh == NULL) || (BKE_mesh_is_valid(mesh) == true)))
|
2013-09-04 03:29:34 +02:00
|
|
|
#else
|
|
|
|
# define ASSERT_IS_VALID_DM(dm)
|
2018-05-11 09:53:52 +02:00
|
|
|
# define ASSERT_IS_VALID_MESH(mesh)
|
2013-09-04 03:29:34 +02:00
|
|
|
#endif
|
|
|
|
|
2019-02-01 02:44:19 +01:00
|
|
|
static CLG_LogRef LOG = {"bke.derivedmesh"};
|
2017-09-19 13:09:35 +02:00
|
|
|
static ThreadRWMutex loops_cache_lock = PTHREAD_RWLOCK_INITIALIZER;
|
2017-08-11 16:51:19 +02:00
|
|
|
|
|
|
|
|
2011-04-15 07:20:18 +02:00
|
|
|
static void shapekey_layers_to_keyblocks(DerivedMesh *dm, Mesh *me, int actshape_uid);
|
|
|
|
|
2018-05-11 09:53:52 +02:00
|
|
|
static void mesh_init_origspace(Mesh *mesh);
|
|
|
|
|
2013-09-04 03:29:34 +02:00
|
|
|
|
|
|
|
/* -------------------------------------------------------------------- */
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
|
2008-09-29 19:08:11 +02:00
|
|
|
static MVert *dm_getVertArray(DerivedMesh *dm)
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
{
|
|
|
|
MVert *mvert = CustomData_get_layer(&dm->vertData, CD_MVERT);
|
|
|
|
|
|
|
|
if (!mvert) {
|
2006-12-12 22:29:09 +01:00
|
|
|
mvert = CustomData_add_layer(&dm->vertData, CD_MVERT, CD_CALLOC, NULL,
|
2012-05-12 21:18:02 +02:00
|
|
|
dm->getNumVerts(dm));
|
2006-12-12 22:29:09 +01:00
|
|
|
CustomData_set_layer_flag(&dm->vertData, CD_MVERT, CD_FLAG_TEMPORARY);
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
dm->copyVertArray(dm, mvert);
|
|
|
|
}
|
|
|
|
|
|
|
|
return mvert;
|
|
|
|
}
|
|
|
|
|
2008-09-29 19:08:11 +02:00
|
|
|
static MEdge *dm_getEdgeArray(DerivedMesh *dm)
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
{
|
|
|
|
MEdge *medge = CustomData_get_layer(&dm->edgeData, CD_MEDGE);
|
|
|
|
|
|
|
|
if (!medge) {
|
2006-12-12 22:29:09 +01:00
|
|
|
medge = CustomData_add_layer(&dm->edgeData, CD_MEDGE, CD_CALLOC, NULL,
|
2012-05-12 21:18:02 +02:00
|
|
|
dm->getNumEdges(dm));
|
2006-12-12 22:29:09 +01:00
|
|
|
CustomData_set_layer_flag(&dm->edgeData, CD_MEDGE, CD_FLAG_TEMPORARY);
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
dm->copyEdgeArray(dm, medge);
|
|
|
|
}
|
|
|
|
|
|
|
|
return medge;
|
|
|
|
}
|
|
|
|
|
2011-11-30 19:03:56 +01:00
|
|
|
static MFace *dm_getTessFaceArray(DerivedMesh *dm)
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
{
|
|
|
|
MFace *mface = CustomData_get_layer(&dm->faceData, CD_MFACE);
|
|
|
|
|
|
|
|
if (!mface) {
|
2012-03-14 15:09:56 +01:00
|
|
|
int numTessFaces = dm->getNumTessFaces(dm);
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2012-03-14 15:09:56 +01:00
|
|
|
if (!numTessFaces) {
|
|
|
|
/* Do not add layer if there's no elements in it, this leads to issues later when
|
|
|
|
* this layer is needed with non-zero size, but currently CD stuff does not check
|
|
|
|
* for requested layer size on creation and just returns layer which was previously
|
|
|
|
* added (sergey) */
|
|
|
|
return NULL;
|
|
|
|
}
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2012-03-14 15:09:56 +01:00
|
|
|
mface = CustomData_add_layer(&dm->faceData, CD_MFACE, CD_CALLOC, NULL, numTessFaces);
|
2006-12-12 22:29:09 +01:00
|
|
|
CustomData_set_layer_flag(&dm->faceData, CD_MFACE, CD_FLAG_TEMPORARY);
|
(NOTE: DO NOT TEST)
Start of planned DerivedMesh refactoring. The mface
interfaces in DerivedMesh have been renamed to reflect
their new status as tesselated face interfaces (rather
then the primary ones, which are now stored in mpolys).
short review: mpolys store "primary" face data, while
mfaces store the tesselated form of the mesh (generally
as triangles). mpolys are defined by mloops, and each
mpoly defines a range of loops it "owns" in the main
mloop array.
I've also added basic read-only face iterators, which
are implemented for CDDM, ccgsubsurf, and the bmeditmesh
derivedmesh. Since faces are now variable-length things,
trying to implement the same interface as mfaces would not
have worked well (especially since faces are stored as
an mpoly + a range of mloops).
I figure first we can evaluate these simple read-only
face iterators, then decide if a) we like using iterators
in DerivedMesh, b) how much of it should use them, and c)
if we want write-capable iterators.
I plan to write official docs on this design after I get
it more stable; I'm committing now because there's a rather
lot of changes, and I might do a merge soon.
2009-06-10 12:06:25 +02:00
|
|
|
dm->copyTessFaceArray(dm, mface);
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
return mface;
|
|
|
|
}
|
2006-08-28 03:12:36 +02:00
|
|
|
|
2011-06-14 05:16:08 +02:00
|
|
|
static MLoop *dm_getLoopArray(DerivedMesh *dm)
|
|
|
|
{
|
|
|
|
MLoop *mloop = CustomData_get_layer(&dm->loopData, CD_MLOOP);
|
|
|
|
|
|
|
|
if (!mloop) {
|
|
|
|
mloop = CustomData_add_layer(&dm->loopData, CD_MLOOP, CD_CALLOC, NULL,
|
2012-05-12 21:18:02 +02:00
|
|
|
dm->getNumLoops(dm));
|
2011-06-14 05:16:08 +02:00
|
|
|
CustomData_set_layer_flag(&dm->loopData, CD_MLOOP, CD_FLAG_TEMPORARY);
|
|
|
|
dm->copyLoopArray(dm, mloop);
|
|
|
|
}
|
|
|
|
|
|
|
|
return mloop;
|
|
|
|
}
|
|
|
|
|
|
|
|
static MPoly *dm_getPolyArray(DerivedMesh *dm)
|
|
|
|
{
|
|
|
|
MPoly *mpoly = CustomData_get_layer(&dm->polyData, CD_MPOLY);
|
|
|
|
|
|
|
|
if (!mpoly) {
|
|
|
|
mpoly = CustomData_add_layer(&dm->polyData, CD_MPOLY, CD_CALLOC, NULL,
|
2012-05-12 21:18:02 +02:00
|
|
|
dm->getNumPolys(dm));
|
2011-06-14 05:16:08 +02:00
|
|
|
CustomData_set_layer_flag(&dm->polyData, CD_MPOLY, CD_FLAG_TEMPORARY);
|
|
|
|
dm->copyPolyArray(dm, mpoly);
|
|
|
|
}
|
|
|
|
|
|
|
|
return mpoly;
|
|
|
|
}
|
|
|
|
|
2008-09-29 19:08:11 +02:00
|
|
|
static MVert *dm_dupVertArray(DerivedMesh *dm)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
2018-01-14 22:14:20 +01:00
|
|
|
MVert *tmp = MEM_malloc_arrayN(dm->getNumVerts(dm), sizeof(*tmp),
|
2012-05-12 21:18:02 +02:00
|
|
|
"dm_dupVertArray tmp");
|
2006-08-28 03:12:36 +02:00
|
|
|
|
2012-03-24 07:18:31 +01:00
|
|
|
if (tmp) dm->copyVertArray(dm, tmp);
|
2006-08-28 03:12:36 +02:00
|
|
|
|
|
|
|
return tmp;
|
|
|
|
}
|
|
|
|
|
2008-09-29 19:08:11 +02:00
|
|
|
static MEdge *dm_dupEdgeArray(DerivedMesh *dm)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
2018-01-14 22:14:20 +01:00
|
|
|
MEdge *tmp = MEM_malloc_arrayN(dm->getNumEdges(dm), sizeof(*tmp),
|
2012-05-12 21:18:02 +02:00
|
|
|
"dm_dupEdgeArray tmp");
|
2006-08-28 03:12:36 +02:00
|
|
|
|
2012-03-24 07:18:31 +01:00
|
|
|
if (tmp) dm->copyEdgeArray(dm, tmp);
|
2006-08-28 03:12:36 +02:00
|
|
|
|
|
|
|
return tmp;
|
|
|
|
}
|
|
|
|
|
2008-09-29 19:08:11 +02:00
|
|
|
static MFace *dm_dupFaceArray(DerivedMesh *dm)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
2018-01-14 22:14:20 +01:00
|
|
|
MFace *tmp = MEM_malloc_arrayN(dm->getNumTessFaces(dm), sizeof(*tmp),
|
2012-05-12 21:18:02 +02:00
|
|
|
"dm_dupFaceArray tmp");
|
2006-08-28 03:12:36 +02:00
|
|
|
|
2012-03-24 07:18:31 +01:00
|
|
|
if (tmp) dm->copyTessFaceArray(dm, tmp);
|
2006-08-28 03:12:36 +02:00
|
|
|
|
|
|
|
return tmp;
|
|
|
|
}
|
|
|
|
|
2011-06-14 05:16:08 +02:00
|
|
|
static MLoop *dm_dupLoopArray(DerivedMesh *dm)
|
|
|
|
{
|
2018-01-14 22:14:20 +01:00
|
|
|
MLoop *tmp = MEM_malloc_arrayN(dm->getNumLoops(dm), sizeof(*tmp),
|
2012-05-12 21:18:02 +02:00
|
|
|
"dm_dupLoopArray tmp");
|
2011-06-14 05:16:08 +02:00
|
|
|
|
2012-03-24 07:18:31 +01:00
|
|
|
if (tmp) dm->copyLoopArray(dm, tmp);
|
2011-06-14 05:16:08 +02:00
|
|
|
|
|
|
|
return tmp;
|
|
|
|
}
|
|
|
|
|
|
|
|
static MPoly *dm_dupPolyArray(DerivedMesh *dm)
|
|
|
|
{
|
2018-01-14 22:14:20 +01:00
|
|
|
MPoly *tmp = MEM_malloc_arrayN(dm->getNumPolys(dm), sizeof(*tmp),
|
2012-05-12 21:18:02 +02:00
|
|
|
"dm_dupPolyArray tmp");
|
2011-06-14 05:16:08 +02:00
|
|
|
|
2012-03-24 07:18:31 +01:00
|
|
|
if (tmp) dm->copyPolyArray(dm, tmp);
|
2011-06-14 05:16:08 +02:00
|
|
|
|
|
|
|
return tmp;
|
|
|
|
}
|
|
|
|
|
2015-07-16 19:36:03 +02:00
|
|
|
static int dm_getNumLoopTri(DerivedMesh *dm)
|
|
|
|
{
|
2016-11-23 20:19:03 +01:00
|
|
|
const int numlooptris = poly_to_tri_count(dm->getNumPolys(dm), dm->getNumLoops(dm));
|
|
|
|
BLI_assert(ELEM(dm->looptris.num, 0, numlooptris));
|
|
|
|
return numlooptris;
|
2015-07-16 19:36:03 +02:00
|
|
|
}
|
|
|
|
|
2017-08-11 15:41:53 +02:00
|
|
|
static const MLoopTri *dm_getLoopTriArray(DerivedMesh *dm)
|
|
|
|
{
|
2017-09-19 13:09:35 +02:00
|
|
|
MLoopTri *looptri;
|
|
|
|
|
|
|
|
BLI_rw_mutex_lock(&loops_cache_lock, THREAD_LOCK_READ);
|
|
|
|
looptri = dm->looptris.array;
|
|
|
|
BLI_rw_mutex_unlock(&loops_cache_lock);
|
|
|
|
|
|
|
|
if (looptri != NULL) {
|
2017-08-11 16:18:01 +02:00
|
|
|
BLI_assert(dm->getNumLoopTri(dm) == dm->looptris.num);
|
2017-08-11 15:41:53 +02:00
|
|
|
}
|
|
|
|
else {
|
2017-09-19 13:09:35 +02:00
|
|
|
BLI_rw_mutex_lock(&loops_cache_lock, THREAD_LOCK_WRITE);
|
2017-08-11 16:51:19 +02:00
|
|
|
/* We need to ensure array is still NULL inside mutex-protected code, some other thread might have already
|
|
|
|
* recomputed those looptris. */
|
|
|
|
if (dm->looptris.array == NULL) {
|
|
|
|
dm->recalcLoopTri(dm);
|
|
|
|
}
|
2017-09-19 13:09:35 +02:00
|
|
|
looptri = dm->looptris.array;
|
|
|
|
BLI_rw_mutex_unlock(&loops_cache_lock);
|
2017-08-11 15:41:53 +02:00
|
|
|
}
|
2017-09-19 13:09:35 +02:00
|
|
|
return looptri;
|
2017-08-11 15:41:53 +02:00
|
|
|
}
|
|
|
|
|
2011-05-11 04:14:43 +02:00
|
|
|
static CustomData *dm_getVertCData(DerivedMesh *dm)
|
2009-06-23 07:35:49 +02:00
|
|
|
{
|
|
|
|
return &dm->vertData;
|
|
|
|
}
|
|
|
|
|
2011-05-11 04:14:43 +02:00
|
|
|
static CustomData *dm_getEdgeCData(DerivedMesh *dm)
|
2009-06-23 07:35:49 +02:00
|
|
|
{
|
|
|
|
return &dm->edgeData;
|
|
|
|
}
|
|
|
|
|
2011-11-30 19:03:56 +01:00
|
|
|
static CustomData *dm_getTessFaceCData(DerivedMesh *dm)
|
2009-06-23 07:35:49 +02:00
|
|
|
{
|
|
|
|
return &dm->faceData;
|
|
|
|
}
|
|
|
|
|
2011-05-11 04:14:43 +02:00
|
|
|
static CustomData *dm_getLoopCData(DerivedMesh *dm)
|
2009-06-23 07:35:49 +02:00
|
|
|
{
|
|
|
|
return &dm->loopData;
|
|
|
|
}
|
|
|
|
|
2011-05-11 04:14:43 +02:00
|
|
|
static CustomData *dm_getPolyCData(DerivedMesh *dm)
|
2009-06-23 07:35:49 +02:00
|
|
|
{
|
|
|
|
return &dm->polyData;
|
|
|
|
}
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
/**
|
|
|
|
* Utility function to initialize a DerivedMesh's function pointers to
|
|
|
|
* the default implementation (for those functions which have a default)
|
|
|
|
*/
|
2006-08-28 03:12:36 +02:00
|
|
|
void DM_init_funcs(DerivedMesh *dm)
|
|
|
|
{
|
|
|
|
/* default function implementations */
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
dm->getVertArray = dm_getVertArray;
|
|
|
|
dm->getEdgeArray = dm_getEdgeArray;
|
2011-11-30 19:03:56 +01:00
|
|
|
dm->getTessFaceArray = dm_getTessFaceArray;
|
2011-06-14 05:16:08 +02:00
|
|
|
dm->getLoopArray = dm_getLoopArray;
|
|
|
|
dm->getPolyArray = dm_getPolyArray;
|
2006-08-28 03:12:36 +02:00
|
|
|
dm->dupVertArray = dm_dupVertArray;
|
|
|
|
dm->dupEdgeArray = dm_dupEdgeArray;
|
(NOTE: DO NOT TEST)
Start of planned DerivedMesh refactoring. The mface
interfaces in DerivedMesh have been renamed to reflect
their new status as tesselated face interfaces (rather
then the primary ones, which are now stored in mpolys).
short review: mpolys store "primary" face data, while
mfaces store the tesselated form of the mesh (generally
as triangles). mpolys are defined by mloops, and each
mpoly defines a range of loops it "owns" in the main
mloop array.
I've also added basic read-only face iterators, which
are implemented for CDDM, ccgsubsurf, and the bmeditmesh
derivedmesh. Since faces are now variable-length things,
trying to implement the same interface as mfaces would not
have worked well (especially since faces are stored as
an mpoly + a range of mloops).
I figure first we can evaluate these simple read-only
face iterators, then decide if a) we like using iterators
in DerivedMesh, b) how much of it should use them, and c)
if we want write-capable iterators.
I plan to write official docs on this design after I get
it more stable; I'm committing now because there's a rather
lot of changes, and I might do a merge soon.
2009-06-10 12:06:25 +02:00
|
|
|
dm->dupTessFaceArray = dm_dupFaceArray;
|
2011-06-14 05:16:08 +02:00
|
|
|
dm->dupLoopArray = dm_dupLoopArray;
|
|
|
|
dm->dupPolyArray = dm_dupPolyArray;
|
2006-08-28 03:12:36 +02:00
|
|
|
|
2017-08-11 15:41:53 +02:00
|
|
|
dm->getLoopTriArray = dm_getLoopTriArray;
|
|
|
|
|
2015-07-16 19:36:03 +02:00
|
|
|
/* subtypes handle getting actual data */
|
|
|
|
dm->getNumLoopTri = dm_getNumLoopTri;
|
|
|
|
|
2009-06-23 07:35:49 +02:00
|
|
|
dm->getVertDataLayout = dm_getVertCData;
|
|
|
|
dm->getEdgeDataLayout = dm_getEdgeCData;
|
2011-11-30 19:03:56 +01:00
|
|
|
dm->getTessFaceDataLayout = dm_getTessFaceCData;
|
2009-06-23 07:35:49 +02:00
|
|
|
dm->getLoopDataLayout = dm_getLoopCData;
|
2011-11-29 14:01:51 +01:00
|
|
|
dm->getPolyDataLayout = dm_getPolyCData;
|
2009-06-23 07:35:49 +02:00
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
dm->getVertData = DM_get_vert_data;
|
|
|
|
dm->getEdgeData = DM_get_edge_data;
|
2011-11-29 06:09:54 +01:00
|
|
|
dm->getTessFaceData = DM_get_tessface_data;
|
2012-10-30 20:20:17 +01:00
|
|
|
dm->getPolyData = DM_get_poly_data;
|
2006-08-28 03:12:36 +02:00
|
|
|
dm->getVertDataArray = DM_get_vert_data_layer;
|
|
|
|
dm->getEdgeDataArray = DM_get_edge_data_layer;
|
2009-08-15 19:31:28 +02:00
|
|
|
dm->getTessFaceDataArray = DM_get_tessface_data_layer;
|
2012-10-30 20:20:17 +01:00
|
|
|
dm->getPolyDataArray = DM_get_poly_data_layer;
|
2013-04-03 22:10:08 +02:00
|
|
|
dm->getLoopDataArray = DM_get_loop_data_layer;
|
2009-05-23 05:24:15 +02:00
|
|
|
|
2018-05-09 01:00:51 +02:00
|
|
|
dm->bvhCache = NULL;
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
/**
|
|
|
|
* Utility function to initialize a DerivedMesh for the desired number
|
|
|
|
* of vertices, edges and faces (doesn't allocate memory for them, just
|
|
|
|
* sets up the custom data layers)
|
|
|
|
*/
|
|
|
|
void DM_init(
|
|
|
|
DerivedMesh *dm, DerivedMeshType type, int numVerts, int numEdges,
|
|
|
|
int numTessFaces, int numLoops, int numPolys)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
2010-01-06 13:05:46 +01:00
|
|
|
dm->type = type;
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
dm->numVertData = numVerts;
|
|
|
|
dm->numEdgeData = numEdges;
|
2011-11-30 19:03:56 +01:00
|
|
|
dm->numTessFaceData = numTessFaces;
|
(NOTE: DO NOT TEST)
Start of planned DerivedMesh refactoring. The mface
interfaces in DerivedMesh have been renamed to reflect
their new status as tesselated face interfaces (rather
then the primary ones, which are now stored in mpolys).
short review: mpolys store "primary" face data, while
mfaces store the tesselated form of the mesh (generally
as triangles). mpolys are defined by mloops, and each
mpoly defines a range of loops it "owns" in the main
mloop array.
I've also added basic read-only face iterators, which
are implemented for CDDM, ccgsubsurf, and the bmeditmesh
derivedmesh. Since faces are now variable-length things,
trying to implement the same interface as mfaces would not
have worked well (especially since faces are stored as
an mpoly + a range of mloops).
I figure first we can evaluate these simple read-only
face iterators, then decide if a) we like using iterators
in DerivedMesh, b) how much of it should use them, and c)
if we want write-capable iterators.
I plan to write official docs on this design after I get
it more stable; I'm committing now because there's a rather
lot of changes, and I might do a merge soon.
2009-06-10 12:06:25 +02:00
|
|
|
dm->numLoopData = numLoops;
|
2011-11-30 19:03:56 +01:00
|
|
|
dm->numPolyData = numPolys;
|
2006-08-28 03:12:36 +02:00
|
|
|
|
|
|
|
DM_init_funcs(dm);
|
2018-06-17 17:05:51 +02:00
|
|
|
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
dm->needsFree = 1;
|
Fix [#30234] Various problems with CD layers and tesselation, related to modifiers stack.
Should also fix [#30266], [#29451], and partly [#30316].
Here are the changes made by this commit:
* It adds a "dirty" flag to DerivedMesh struct (for now, only DM_DIRTY_TESS_CDLAYERS, but more might be added as needed).
* It adds a new func, DM_update_tessface_data, which assumes tessfaces themselves are valid, but updates tessellated customdata from their poly/loop counter parts.
* At end of modstack, when valid tessellated faces are present in finaldm , but the cdlayers dirty flag is set, call that function (instead of recomputing the whole tessellation).
* Edits to the codes concerned (UVProject, DynamicPaint, and Subsurf modifiers).
* Also add to subsurf dm generation code the creation of a CD_POLYINDEX layer (mandatory for DM_update_tessface_data to work well, and imho all tessellated dm should have one).
Note: some pieces of old code are just #if 0’ed, will clean them later.
2012-03-18 23:06:57 +01:00
|
|
|
dm->dirty = 0;
|
2012-10-31 10:50:24 +01:00
|
|
|
|
|
|
|
/* don't use CustomData_reset(...); because we dont want to touch customdata */
|
2015-05-05 09:08:29 +02:00
|
|
|
copy_vn_i(dm->vertData.typemap, CD_NUMTYPES, -1);
|
|
|
|
copy_vn_i(dm->edgeData.typemap, CD_NUMTYPES, -1);
|
|
|
|
copy_vn_i(dm->faceData.typemap, CD_NUMTYPES, -1);
|
|
|
|
copy_vn_i(dm->loopData.typemap, CD_NUMTYPES, -1);
|
|
|
|
copy_vn_i(dm->polyData.typemap, CD_NUMTYPES, -1);
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
/**
|
|
|
|
* Utility function to initialize a DerivedMesh for the desired number
|
|
|
|
* of vertices, edges and faces, with a layer setup copied from source
|
|
|
|
*/
|
2015-09-23 14:57:00 +02:00
|
|
|
void DM_from_template_ex(
|
2015-07-02 08:20:22 +02:00
|
|
|
DerivedMesh *dm, DerivedMesh *source, DerivedMeshType type,
|
|
|
|
int numVerts, int numEdges, int numTessFaces,
|
2015-09-23 14:57:00 +02:00
|
|
|
int numLoops, int numPolys,
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
const CustomData_MeshMasks *mask)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_copy(&source->vertData, &dm->vertData, mask->vmask, CD_CALLOC, numVerts);
|
|
|
|
CustomData_copy(&source->edgeData, &dm->edgeData, mask->emask, CD_CALLOC, numEdges);
|
|
|
|
CustomData_copy(&source->faceData, &dm->faceData, mask->fmask, CD_CALLOC, numTessFaces);
|
|
|
|
CustomData_copy(&source->loopData, &dm->loopData, mask->lmask, CD_CALLOC, numLoops);
|
|
|
|
CustomData_copy(&source->polyData, &dm->polyData, mask->pmask, CD_CALLOC, numPolys);
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
|
2013-01-10 05:43:31 +01:00
|
|
|
dm->cd_flag = source->cd_flag;
|
|
|
|
|
2010-01-06 13:05:46 +01:00
|
|
|
dm->type = type;
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
dm->numVertData = numVerts;
|
|
|
|
dm->numEdgeData = numEdges;
|
2011-11-30 19:03:56 +01:00
|
|
|
dm->numTessFaceData = numTessFaces;
|
(NOTE: DO NOT TEST)
Start of planned DerivedMesh refactoring. The mface
interfaces in DerivedMesh have been renamed to reflect
their new status as tesselated face interfaces (rather
then the primary ones, which are now stored in mpolys).
short review: mpolys store "primary" face data, while
mfaces store the tesselated form of the mesh (generally
as triangles). mpolys are defined by mloops, and each
mpoly defines a range of loops it "owns" in the main
mloop array.
I've also added basic read-only face iterators, which
are implemented for CDDM, ccgsubsurf, and the bmeditmesh
derivedmesh. Since faces are now variable-length things,
trying to implement the same interface as mfaces would not
have worked well (especially since faces are stored as
an mpoly + a range of mloops).
I figure first we can evaluate these simple read-only
face iterators, then decide if a) we like using iterators
in DerivedMesh, b) how much of it should use them, and c)
if we want write-capable iterators.
I plan to write official docs on this design after I get
it more stable; I'm committing now because there's a rather
lot of changes, and I might do a merge soon.
2009-06-10 12:06:25 +02:00
|
|
|
dm->numLoopData = numLoops;
|
|
|
|
dm->numPolyData = numPolys;
|
2006-08-28 03:12:36 +02:00
|
|
|
|
|
|
|
DM_init_funcs(dm);
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
|
|
|
|
dm->needsFree = 1;
|
Fix [#30234] Various problems with CD layers and tesselation, related to modifiers stack.
Should also fix [#30266], [#29451], and partly [#30316].
Here are the changes made by this commit:
* It adds a "dirty" flag to DerivedMesh struct (for now, only DM_DIRTY_TESS_CDLAYERS, but more might be added as needed).
* It adds a new func, DM_update_tessface_data, which assumes tessfaces themselves are valid, but updates tessellated customdata from their poly/loop counter parts.
* At end of modstack, when valid tessellated faces are present in finaldm , but the cdlayers dirty flag is set, call that function (instead of recomputing the whole tessellation).
* Edits to the codes concerned (UVProject, DynamicPaint, and Subsurf modifiers).
* Also add to subsurf dm generation code the creation of a CD_POLYINDEX layer (mandatory for DM_update_tessface_data to work well, and imho all tessellated dm should have one).
Note: some pieces of old code are just #if 0’ed, will clean them later.
2012-03-18 23:06:57 +01:00
|
|
|
dm->dirty = 0;
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
2015-09-23 14:57:00 +02:00
|
|
|
void DM_from_template(
|
|
|
|
DerivedMesh *dm, DerivedMesh *source, DerivedMeshType type,
|
|
|
|
int numVerts, int numEdges, int numTessFaces,
|
|
|
|
int numLoops, int numPolys)
|
|
|
|
{
|
|
|
|
DM_from_template_ex(
|
|
|
|
dm, source, type,
|
|
|
|
numVerts, numEdges, numTessFaces,
|
|
|
|
numLoops, numPolys,
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
&CD_MASK_DERIVEDMESH);
|
2015-09-23 14:57:00 +02:00
|
|
|
}
|
2006-08-28 03:12:36 +02:00
|
|
|
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
int DM_release(DerivedMesh *dm)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
if (dm->needsFree) {
|
2009-05-23 05:24:15 +02:00
|
|
|
bvhcache_free(&dm->bvhCache);
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
CustomData_free(&dm->vertData, dm->numVertData);
|
|
|
|
CustomData_free(&dm->edgeData, dm->numEdgeData);
|
2011-11-30 19:03:56 +01:00
|
|
|
CustomData_free(&dm->faceData, dm->numTessFaceData);
|
(NOTE: DO NOT TEST)
Start of planned DerivedMesh refactoring. The mface
interfaces in DerivedMesh have been renamed to reflect
their new status as tesselated face interfaces (rather
then the primary ones, which are now stored in mpolys).
short review: mpolys store "primary" face data, while
mfaces store the tesselated form of the mesh (generally
as triangles). mpolys are defined by mloops, and each
mpoly defines a range of loops it "owns" in the main
mloop array.
I've also added basic read-only face iterators, which
are implemented for CDDM, ccgsubsurf, and the bmeditmesh
derivedmesh. Since faces are now variable-length things,
trying to implement the same interface as mfaces would not
have worked well (especially since faces are stored as
an mpoly + a range of mloops).
I figure first we can evaluate these simple read-only
face iterators, then decide if a) we like using iterators
in DerivedMesh, b) how much of it should use them, and c)
if we want write-capable iterators.
I plan to write official docs on this design after I get
it more stable; I'm committing now because there's a rather
lot of changes, and I might do a merge soon.
2009-06-10 12:06:25 +02:00
|
|
|
CustomData_free(&dm->loopData, dm->numLoopData);
|
|
|
|
CustomData_free(&dm->polyData, dm->numPolyData);
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
|
2014-04-09 17:37:54 +02:00
|
|
|
if (dm->mat) {
|
|
|
|
MEM_freeN(dm->mat);
|
|
|
|
dm->mat = NULL;
|
|
|
|
dm->totmat = 0;
|
|
|
|
}
|
|
|
|
|
2015-07-16 19:36:03 +02:00
|
|
|
MEM_SAFE_FREE(dm->looptris.array);
|
|
|
|
dm->looptris.num = 0;
|
|
|
|
dm->looptris.num_alloc = 0;
|
|
|
|
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
CustomData_free_temporary(&dm->vertData, dm->numVertData);
|
|
|
|
CustomData_free_temporary(&dm->edgeData, dm->numEdgeData);
|
2011-11-30 19:03:56 +01:00
|
|
|
CustomData_free_temporary(&dm->faceData, dm->numTessFaceData);
|
2009-06-16 22:08:40 +02:00
|
|
|
CustomData_free_temporary(&dm->loopData, dm->numLoopData);
|
|
|
|
CustomData_free_temporary(&dm->polyData, dm->numPolyData);
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2011-06-14 05:16:08 +02:00
|
|
|
void DM_DupPolys(DerivedMesh *source, DerivedMesh *target)
|
(NOTE: DO NOT TEST)
Start of planned DerivedMesh refactoring. The mface
interfaces in DerivedMesh have been renamed to reflect
their new status as tesselated face interfaces (rather
then the primary ones, which are now stored in mpolys).
short review: mpolys store "primary" face data, while
mfaces store the tesselated form of the mesh (generally
as triangles). mpolys are defined by mloops, and each
mpoly defines a range of loops it "owns" in the main
mloop array.
I've also added basic read-only face iterators, which
are implemented for CDDM, ccgsubsurf, and the bmeditmesh
derivedmesh. Since faces are now variable-length things,
trying to implement the same interface as mfaces would not
have worked well (especially since faces are stored as
an mpoly + a range of mloops).
I figure first we can evaluate these simple read-only
face iterators, then decide if a) we like using iterators
in DerivedMesh, b) how much of it should use them, and c)
if we want write-capable iterators.
I plan to write official docs on this design after I get
it more stable; I'm committing now because there's a rather
lot of changes, and I might do a merge soon.
2009-06-10 12:06:25 +02:00
|
|
|
{
|
2011-10-27 14:17:02 +02:00
|
|
|
CustomData_free(&target->loopData, source->numLoopData);
|
|
|
|
CustomData_free(&target->polyData, source->numPolyData);
|
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_copy(&source->loopData, &target->loopData, CD_MASK_DERIVEDMESH.lmask, CD_DUPLICATE, source->numLoopData);
|
|
|
|
CustomData_copy(&source->polyData, &target->polyData, CD_MASK_DERIVEDMESH.pmask, CD_DUPLICATE, source->numPolyData);
|
2009-08-18 22:05:08 +02:00
|
|
|
|
2011-06-14 05:16:08 +02:00
|
|
|
target->numLoopData = source->numLoopData;
|
|
|
|
target->numPolyData = source->numPolyData;
|
2009-08-18 22:05:08 +02:00
|
|
|
|
2011-06-14 05:16:08 +02:00
|
|
|
if (!CustomData_has_layer(&target->polyData, CD_MPOLY)) {
|
|
|
|
MPoly *mpoly;
|
|
|
|
MLoop *mloop;
|
2009-08-18 22:05:08 +02:00
|
|
|
|
2011-06-14 05:16:08 +02:00
|
|
|
mloop = source->dupLoopArray(source);
|
|
|
|
mpoly = source->dupPolyArray(source);
|
|
|
|
CustomData_add_layer(&target->loopData, CD_MLOOP, CD_ASSIGN, mloop, source->numLoopData);
|
|
|
|
CustomData_add_layer(&target->polyData, CD_MPOLY, CD_ASSIGN, mpoly, source->numPolyData);
|
(NOTE: DO NOT TEST)
Start of planned DerivedMesh refactoring. The mface
interfaces in DerivedMesh have been renamed to reflect
their new status as tesselated face interfaces (rather
then the primary ones, which are now stored in mpolys).
short review: mpolys store "primary" face data, while
mfaces store the tesselated form of the mesh (generally
as triangles). mpolys are defined by mloops, and each
mpoly defines a range of loops it "owns" in the main
mloop array.
I've also added basic read-only face iterators, which
are implemented for CDDM, ccgsubsurf, and the bmeditmesh
derivedmesh. Since faces are now variable-length things,
trying to implement the same interface as mfaces would not
have worked well (especially since faces are stored as
an mpoly + a range of mloops).
I figure first we can evaluate these simple read-only
face iterators, then decide if a) we like using iterators
in DerivedMesh, b) how much of it should use them, and c)
if we want write-capable iterators.
I plan to write official docs on this design after I get
it more stable; I'm committing now because there's a rather
lot of changes, and I might do a merge soon.
2009-06-10 12:06:25 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-05-30 19:36:43 +02:00
|
|
|
void DM_ensure_normals(DerivedMesh *dm)
|
|
|
|
{
|
|
|
|
if (dm->dirty & DM_DIRTY_NORMALS) {
|
|
|
|
dm->calcNormals(dm);
|
|
|
|
}
|
|
|
|
BLI_assert((dm->dirty & DM_DIRTY_NORMALS) == 0);
|
|
|
|
}
|
|
|
|
|
2015-07-16 19:36:03 +02:00
|
|
|
/**
|
|
|
|
* Ensure the array is large enough
|
2017-09-19 13:57:46 +02:00
|
|
|
*
|
2019-02-11 15:21:09 +01:00
|
|
|
* \note This function must always be thread-protected by caller. It should only be used by internal code.
|
2015-07-16 19:36:03 +02:00
|
|
|
*/
|
|
|
|
void DM_ensure_looptri_data(DerivedMesh *dm)
|
|
|
|
{
|
|
|
|
const unsigned int totpoly = dm->numPolyData;
|
|
|
|
const unsigned int totloop = dm->numLoopData;
|
|
|
|
const int looptris_num = poly_to_tri_count(totpoly, totloop);
|
|
|
|
|
2017-09-19 13:57:46 +02:00
|
|
|
BLI_assert(dm->looptris.array_wip == NULL);
|
|
|
|
|
|
|
|
SWAP(MLoopTri *, dm->looptris.array, dm->looptris.array_wip);
|
|
|
|
|
2015-07-16 19:36:03 +02:00
|
|
|
if ((looptris_num > dm->looptris.num_alloc) ||
|
|
|
|
(looptris_num < dm->looptris.num_alloc * 2) ||
|
|
|
|
(totpoly == 0))
|
|
|
|
{
|
2017-09-19 13:57:46 +02:00
|
|
|
MEM_SAFE_FREE(dm->looptris.array_wip);
|
2015-07-16 19:36:03 +02:00
|
|
|
dm->looptris.num_alloc = 0;
|
|
|
|
dm->looptris.num = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (totpoly) {
|
2017-09-19 13:57:46 +02:00
|
|
|
if (dm->looptris.array_wip == NULL) {
|
2018-01-14 22:14:20 +01:00
|
|
|
dm->looptris.array_wip = MEM_malloc_arrayN(looptris_num, sizeof(*dm->looptris.array_wip), __func__);
|
2015-07-16 19:36:03 +02:00
|
|
|
dm->looptris.num_alloc = looptris_num;
|
|
|
|
}
|
|
|
|
|
|
|
|
dm->looptris.num = looptris_num;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
void DM_to_mesh(DerivedMesh *dm, Mesh *me, Object *ob, const CustomData_MeshMasks *mask, bool take_ownership)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
|
|
|
/* dm might depend on me, so we need to do everything with a local copy */
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
Mesh tmp = *me;
|
2011-11-16 18:09:41 +01:00
|
|
|
int totvert, totedge /*, totface */ /* UNUSED */, totloop, totpoly;
|
2012-05-12 21:18:02 +02:00
|
|
|
int did_shapekeys = 0;
|
2018-05-30 10:28:24 +02:00
|
|
|
eCDAllocType alloctype = CD_DUPLICATE;
|
2015-02-18 20:00:23 +01:00
|
|
|
|
|
|
|
if (take_ownership && dm->type == DM_TYPE_CDDM && dm->needsFree) {
|
|
|
|
bool has_any_referenced_layers =
|
|
|
|
CustomData_has_referenced(&dm->vertData) ||
|
|
|
|
CustomData_has_referenced(&dm->edgeData) ||
|
|
|
|
CustomData_has_referenced(&dm->loopData) ||
|
|
|
|
CustomData_has_referenced(&dm->faceData) ||
|
|
|
|
CustomData_has_referenced(&dm->polyData);
|
|
|
|
if (!has_any_referenced_layers) {
|
|
|
|
alloctype = CD_ASSIGN;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2012-10-31 10:50:24 +01:00
|
|
|
CustomData_reset(&tmp.vdata);
|
|
|
|
CustomData_reset(&tmp.edata);
|
|
|
|
CustomData_reset(&tmp.fdata);
|
|
|
|
CustomData_reset(&tmp.ldata);
|
|
|
|
CustomData_reset(&tmp.pdata);
|
2006-08-28 03:12:36 +02:00
|
|
|
|
2013-06-04 18:02:54 +02:00
|
|
|
DM_ensure_normals(dm);
|
|
|
|
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
totvert = tmp.totvert = dm->getNumVerts(dm);
|
|
|
|
totedge = tmp.totedge = dm->getNumEdges(dm);
|
2011-12-03 03:11:38 +01:00
|
|
|
totloop = tmp.totloop = dm->getNumLoops(dm);
|
2011-11-29 14:01:51 +01:00
|
|
|
totpoly = tmp.totpoly = dm->getNumPolys(dm);
|
2013-01-10 06:50:21 +01:00
|
|
|
tmp.totface = 0;
|
2006-08-28 03:12:36 +02:00
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_copy(&dm->vertData, &tmp.vdata, mask->vmask, alloctype, totvert);
|
|
|
|
CustomData_copy(&dm->edgeData, &tmp.edata, mask->emask, alloctype, totedge);
|
|
|
|
CustomData_copy(&dm->loopData, &tmp.ldata, mask->lmask, alloctype, totloop);
|
|
|
|
CustomData_copy(&dm->polyData, &tmp.pdata, mask->pmask, alloctype, totpoly);
|
2013-02-26 15:32:53 +01:00
|
|
|
tmp.cd_flag = dm->cd_flag;
|
2018-05-15 13:26:24 +02:00
|
|
|
tmp.runtime.deformed_only = dm->deformedOnly;
|
2006-08-28 03:12:36 +02:00
|
|
|
|
2011-04-15 07:20:18 +02:00
|
|
|
if (CustomData_has_layer(&dm->vertData, CD_SHAPEKEY)) {
|
|
|
|
KeyBlock *kb;
|
2012-01-30 04:54:24 +01:00
|
|
|
int uid;
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2011-04-15 07:20:18 +02:00
|
|
|
if (ob) {
|
2012-05-12 21:18:02 +02:00
|
|
|
kb = BLI_findlink(&me->key->block, ob->shapenr - 1);
|
2012-01-30 04:54:24 +01:00
|
|
|
if (kb) {
|
|
|
|
uid = kb->uid;
|
2011-04-15 07:20:18 +02:00
|
|
|
}
|
2012-01-30 04:54:24 +01:00
|
|
|
else {
|
2019-02-01 02:44:19 +01:00
|
|
|
CLOG_ERROR(&LOG, "could not find active shapekey %d!", ob->shapenr - 1);
|
2012-01-30 04:54:24 +01:00
|
|
|
uid = INT_MAX;
|
2011-04-15 07:20:18 +02:00
|
|
|
}
|
2012-01-30 04:54:24 +01:00
|
|
|
}
|
|
|
|
else {
|
2012-08-17 16:43:20 +02:00
|
|
|
/* if no object, set to INT_MAX so we don't mess up any shapekey layers */
|
2012-01-30 04:54:24 +01:00
|
|
|
uid = INT_MAX;
|
2011-04-15 07:20:18 +02:00
|
|
|
}
|
2012-01-30 04:54:24 +01:00
|
|
|
|
|
|
|
shapekey_layers_to_keyblocks(dm, me, uid);
|
2011-04-15 07:20:18 +02:00
|
|
|
did_shapekeys = 1;
|
|
|
|
}
|
2013-06-10 15:01:41 +02:00
|
|
|
|
|
|
|
/* copy texture space */
|
2013-08-04 01:58:17 +02:00
|
|
|
if (ob) {
|
|
|
|
BKE_mesh_texspace_copy_from_object(&tmp, ob);
|
|
|
|
}
|
2018-06-17 17:05:51 +02:00
|
|
|
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
/* not all DerivedMeshes store their verts/edges/faces in CustomData, so
|
2012-03-09 19:28:30 +01:00
|
|
|
* we set them here in case they are missing */
|
2015-02-18 20:00:23 +01:00
|
|
|
if (!CustomData_has_layer(&tmp.vdata, CD_MVERT)) {
|
|
|
|
CustomData_add_layer(&tmp.vdata, CD_MVERT, CD_ASSIGN,
|
|
|
|
(alloctype == CD_ASSIGN) ? dm->getVertArray(dm) : dm->dupVertArray(dm),
|
|
|
|
totvert);
|
|
|
|
}
|
|
|
|
if (!CustomData_has_layer(&tmp.edata, CD_MEDGE)) {
|
|
|
|
CustomData_add_layer(&tmp.edata, CD_MEDGE, CD_ASSIGN,
|
|
|
|
(alloctype == CD_ASSIGN) ? dm->getEdgeArray(dm) : dm->dupEdgeArray(dm),
|
|
|
|
totedge);
|
|
|
|
}
|
2012-03-24 07:18:31 +01:00
|
|
|
if (!CustomData_has_layer(&tmp.pdata, CD_MPOLY)) {
|
2015-02-18 20:00:23 +01:00
|
|
|
tmp.mloop = (alloctype == CD_ASSIGN) ? dm->getLoopArray(dm) : dm->dupLoopArray(dm);
|
|
|
|
tmp.mpoly = (alloctype == CD_ASSIGN) ? dm->getPolyArray(dm) : dm->dupPolyArray(dm);
|
2011-06-14 05:16:08 +02:00
|
|
|
|
|
|
|
CustomData_add_layer(&tmp.ldata, CD_MLOOP, CD_ASSIGN, tmp.mloop, tmp.totloop);
|
|
|
|
CustomData_add_layer(&tmp.pdata, CD_MPOLY, CD_ASSIGN, tmp.mpoly, tmp.totpoly);
|
|
|
|
}
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
|
2010-09-26 20:29:54 +02:00
|
|
|
/* object had got displacement layer, should copy this layer to save sculpted data */
|
|
|
|
/* NOTE: maybe some other layers should be copied? nazgul */
|
2012-03-24 07:18:31 +01:00
|
|
|
if (CustomData_has_layer(&me->ldata, CD_MDISPS)) {
|
2011-04-15 07:20:18 +02:00
|
|
|
if (totloop == me->totloop) {
|
|
|
|
MDisps *mdisps = CustomData_get_layer(&me->ldata, CD_MDISPS);
|
2015-02-18 20:00:23 +01:00
|
|
|
CustomData_add_layer(&tmp.ldata, CD_MDISPS, alloctype, mdisps, totloop);
|
2010-09-26 20:29:54 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2012-03-20 23:56:26 +01:00
|
|
|
/* yes, must be before _and_ after tessellate */
|
2013-03-17 20:55:10 +01:00
|
|
|
BKE_mesh_update_customdata_pointers(&tmp, false);
|
2011-11-25 23:32:58 +01:00
|
|
|
|
2013-01-10 06:50:21 +01:00
|
|
|
/* since 2.65 caller must do! */
|
|
|
|
// BKE_mesh_tessface_calc(&tmp);
|
2011-11-25 23:32:58 +01:00
|
|
|
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
CustomData_free(&me->vdata, me->totvert);
|
|
|
|
CustomData_free(&me->edata, me->totedge);
|
|
|
|
CustomData_free(&me->fdata, me->totface);
|
2009-06-16 22:08:40 +02:00
|
|
|
CustomData_free(&me->ldata, me->totloop);
|
|
|
|
CustomData_free(&me->pdata, me->totpoly);
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
|
2012-03-09 19:28:30 +01:00
|
|
|
/* ok, this should now use new CD shapekey data,
|
2012-08-25 01:22:34 +02:00
|
|
|
* which should be fed through the modifier
|
2016-01-28 06:30:50 +01:00
|
|
|
* stack */
|
2012-03-24 07:18:31 +01:00
|
|
|
if (tmp.totvert != me->totvert && !did_shapekeys && me->key) {
|
2019-02-01 02:44:19 +01:00
|
|
|
CLOG_WARN(&LOG, "YEEK! this should be recoded! Shape key loss!: ID '%s'", tmp.id.name);
|
2018-05-18 15:10:48 +02:00
|
|
|
if (tmp.key && !(tmp.id.tag & LIB_TAG_NO_MAIN)) {
|
2015-11-09 19:47:10 +01:00
|
|
|
id_us_min(&tmp.key->id);
|
2018-05-18 15:10:48 +02:00
|
|
|
}
|
2010-06-23 12:18:51 +02:00
|
|
|
tmp.key = NULL;
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2013-02-04 12:28:31 +01:00
|
|
|
/* Clear selection history */
|
2016-03-03 14:44:05 +01:00
|
|
|
MEM_SAFE_FREE(tmp.mselect);
|
2013-02-04 12:28:31 +01:00
|
|
|
tmp.totselect = 0;
|
2016-03-03 14:44:05 +01:00
|
|
|
BLI_assert(ELEM(tmp.bb, NULL, me->bb));
|
2016-02-22 13:20:18 +01:00
|
|
|
if (me->bb) {
|
|
|
|
MEM_freeN(me->bb);
|
2016-03-03 14:44:05 +01:00
|
|
|
tmp.bb = NULL;
|
2016-02-22 13:20:18 +01:00
|
|
|
}
|
2013-02-04 12:28:31 +01:00
|
|
|
|
2014-06-24 15:50:12 +02:00
|
|
|
/* skip the listbase */
|
2019-03-12 06:59:04 +01:00
|
|
|
MEMCPY_STRUCT_AFTER(me, &tmp, id.prev);
|
2015-02-18 20:00:23 +01:00
|
|
|
|
|
|
|
if (take_ownership) {
|
|
|
|
if (alloctype == CD_ASSIGN) {
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_free_typemask(&dm->vertData, dm->numVertData, ~mask->vmask);
|
|
|
|
CustomData_free_typemask(&dm->edgeData, dm->numEdgeData, ~mask->emask);
|
|
|
|
CustomData_free_typemask(&dm->loopData, dm->numLoopData, ~mask->lmask);
|
|
|
|
CustomData_free_typemask(&dm->polyData, dm->numPolyData, ~mask->pmask);
|
2015-02-18 20:00:23 +01:00
|
|
|
}
|
|
|
|
dm->release(dm);
|
|
|
|
}
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2018-06-21 18:24:32 +02:00
|
|
|
/** Utility function to convert an (evaluated) Mesh to a shape key block. */
|
|
|
|
/* Just a shallow wrapper around BKE_keyblock_convert_from_mesh,
|
|
|
|
* that ensures both evaluated mesh and original one has same number of vertices. */
|
|
|
|
void BKE_mesh_runtime_eval_to_meshkey(Mesh *me_deformed, Mesh *me, KeyBlock *kb)
|
2009-11-22 14:44:09 +01:00
|
|
|
{
|
2018-06-21 18:24:32 +02:00
|
|
|
const int totvert = me_deformed->totvert;
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
if (totvert == 0 || me->totvert == 0 || me->totvert != totvert) {
|
|
|
|
return;
|
|
|
|
}
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2018-06-21 18:24:32 +02:00
|
|
|
BKE_keyblock_convert_from_mesh(me_deformed, me->key, kb);
|
2009-11-22 14:44:09 +01:00
|
|
|
}
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
/**
|
|
|
|
* set the CD_FLAG_NOCOPY flag in custom data layers where the mask is
|
|
|
|
* zero for the layer type, so only layer types specified by the mask
|
|
|
|
* will be copied
|
|
|
|
*/
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
void DM_set_only_copy(DerivedMesh *dm, const CustomData_MeshMasks *mask)
|
2006-12-05 18:42:03 +01:00
|
|
|
{
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_set_only_copy(&dm->vertData, mask->vmask);
|
|
|
|
CustomData_set_only_copy(&dm->edgeData, mask->emask);
|
|
|
|
CustomData_set_only_copy(&dm->faceData, mask->fmask);
|
2012-10-01 16:23:57 +02:00
|
|
|
/* this wasn't in 2.63 and is disabled for 2.64 because it gives problems with
|
|
|
|
* weight paint mode when there are modifiers applied, needs further investigation,
|
|
|
|
* see replies to r50969, Campbell */
|
|
|
|
#if 0
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_set_only_copy(&dm->loopData, mask->lmask);
|
|
|
|
CustomData_set_only_copy(&dm->polyData, mask->pmask);
|
2012-10-01 16:23:57 +02:00
|
|
|
#endif
|
2006-12-05 18:42:03 +01:00
|
|
|
}
|
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
static void mesh_set_only_copy(Mesh *mesh, const CustomData_MeshMasks *mask)
|
2018-05-11 09:53:52 +02:00
|
|
|
{
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_set_only_copy(&mesh->vdata, mask->vmask);
|
|
|
|
CustomData_set_only_copy(&mesh->edata, mask->emask);
|
|
|
|
CustomData_set_only_copy(&mesh->fdata, mask->fmask);
|
2018-05-11 09:53:52 +02:00
|
|
|
/* this wasn't in 2.63 and is disabled for 2.64 because it gives problems with
|
|
|
|
* weight paint mode when there are modifiers applied, needs further investigation,
|
|
|
|
* see replies to r50969, Campbell */
|
|
|
|
#if 0
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_set_only_copy(&mesh->ldata, mask->lmask);
|
|
|
|
CustomData_set_only_copy(&mesh->pdata, mask->pmask);
|
2018-05-11 09:53:52 +02:00
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2018-05-30 10:28:24 +02:00
|
|
|
void DM_add_vert_layer(DerivedMesh *dm, int type, eCDAllocType alloctype, void *layer)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
2006-12-12 22:29:09 +01:00
|
|
|
CustomData_add_layer(&dm->vertData, type, alloctype, layer, dm->numVertData);
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2018-05-30 10:28:24 +02:00
|
|
|
void DM_add_edge_layer(DerivedMesh *dm, int type, eCDAllocType alloctype, void *layer)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
2006-12-12 22:29:09 +01:00
|
|
|
CustomData_add_layer(&dm->edgeData, type, alloctype, layer, dm->numEdgeData);
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2018-05-30 10:28:24 +02:00
|
|
|
void DM_add_tessface_layer(DerivedMesh *dm, int type, eCDAllocType alloctype, void *layer)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
2011-11-30 19:03:56 +01:00
|
|
|
CustomData_add_layer(&dm->faceData, type, alloctype, layer, dm->numTessFaceData);
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2018-05-30 10:28:24 +02:00
|
|
|
void DM_add_loop_layer(DerivedMesh *dm, int type, eCDAllocType alloctype, void *layer)
|
2009-08-15 19:31:28 +02:00
|
|
|
{
|
|
|
|
CustomData_add_layer(&dm->loopData, type, alloctype, layer, dm->numLoopData);
|
|
|
|
}
|
|
|
|
|
2018-05-30 10:28:24 +02:00
|
|
|
void DM_add_poly_layer(DerivedMesh *dm, int type, eCDAllocType alloctype, void *layer)
|
2009-08-15 19:31:28 +02:00
|
|
|
{
|
|
|
|
CustomData_add_layer(&dm->polyData, type, alloctype, layer, dm->numPolyData);
|
|
|
|
}
|
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
void *DM_get_vert_data(DerivedMesh *dm, int index, int type)
|
|
|
|
{
|
2013-07-10 07:38:36 +02:00
|
|
|
BLI_assert(index >= 0 && index < dm->getNumVerts(dm));
|
2006-08-28 03:12:36 +02:00
|
|
|
return CustomData_get(&dm->vertData, index, type);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *DM_get_edge_data(DerivedMesh *dm, int index, int type)
|
|
|
|
{
|
2013-07-10 07:38:36 +02:00
|
|
|
BLI_assert(index >= 0 && index < dm->getNumEdges(dm));
|
2006-08-28 03:12:36 +02:00
|
|
|
return CustomData_get(&dm->edgeData, index, type);
|
|
|
|
}
|
|
|
|
|
2011-11-29 06:09:54 +01:00
|
|
|
void *DM_get_tessface_data(DerivedMesh *dm, int index, int type)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
2013-07-10 07:38:36 +02:00
|
|
|
BLI_assert(index >= 0 && index < dm->getNumTessFaces(dm));
|
2006-08-28 03:12:36 +02:00
|
|
|
return CustomData_get(&dm->faceData, index, type);
|
|
|
|
}
|
|
|
|
|
2012-10-30 20:20:17 +01:00
|
|
|
void *DM_get_poly_data(DerivedMesh *dm, int index, int type)
|
|
|
|
{
|
2013-07-10 07:38:36 +02:00
|
|
|
BLI_assert(index >= 0 && index < dm->getNumPolys(dm));
|
2012-10-30 20:20:17 +01:00
|
|
|
return CustomData_get(&dm->polyData, index, type);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
void *DM_get_vert_data_layer(DerivedMesh *dm, int type)
|
|
|
|
{
|
2012-03-24 07:18:31 +01:00
|
|
|
if (type == CD_MVERT)
|
2010-01-06 13:05:46 +01:00
|
|
|
return dm->getVertArray(dm);
|
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
return CustomData_get_layer(&dm->vertData, type);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *DM_get_edge_data_layer(DerivedMesh *dm, int type)
|
|
|
|
{
|
2012-03-24 07:18:31 +01:00
|
|
|
if (type == CD_MEDGE)
|
2010-01-06 13:05:46 +01:00
|
|
|
return dm->getEdgeArray(dm);
|
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
return CustomData_get_layer(&dm->edgeData, type);
|
|
|
|
}
|
|
|
|
|
2009-08-15 19:31:28 +02:00
|
|
|
void *DM_get_tessface_data_layer(DerivedMesh *dm, int type)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
2011-11-13 16:13:59 +01:00
|
|
|
if (type == CD_MFACE)
|
2010-01-13 08:26:11 +01:00
|
|
|
return dm->getTessFaceArray(dm);
|
2010-01-06 13:05:46 +01:00
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
return CustomData_get_layer(&dm->faceData, type);
|
|
|
|
}
|
|
|
|
|
2011-11-29 06:09:54 +01:00
|
|
|
void *DM_get_poly_data_layer(DerivedMesh *dm, int type)
|
2009-08-15 19:31:28 +02:00
|
|
|
{
|
|
|
|
return CustomData_get_layer(&dm->polyData, type);
|
|
|
|
}
|
|
|
|
|
2012-02-05 12:30:26 +01:00
|
|
|
void *DM_get_loop_data_layer(DerivedMesh *dm, int type)
|
|
|
|
{
|
|
|
|
return CustomData_get_layer(&dm->loopData, type);
|
|
|
|
}
|
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
void DM_copy_vert_data(DerivedMesh *source, DerivedMesh *dest,
|
2012-05-12 21:18:02 +02:00
|
|
|
int source_index, int dest_index, int count)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
|
|
|
CustomData_copy_data(&source->vertData, &dest->vertData,
|
2012-05-12 21:18:02 +02:00
|
|
|
source_index, dest_index, count);
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
/**
|
|
|
|
* interpolates vertex data from the vertices indexed by src_indices in the
|
|
|
|
* source mesh using the given weights and stores the result in the vertex
|
|
|
|
* indexed by dest_index in the dest mesh
|
|
|
|
*/
|
|
|
|
void DM_interp_vert_data(
|
|
|
|
DerivedMesh *source, DerivedMesh *dest,
|
|
|
|
int *src_indices, float *weights,
|
|
|
|
int count, int dest_index)
|
2006-08-28 03:12:36 +02:00
|
|
|
{
|
|
|
|
CustomData_interp(&source->vertData, &dest->vertData, src_indices,
|
2012-05-12 21:18:02 +02:00
|
|
|
weights, NULL, count, dest_index);
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2013-12-25 22:26:41 +01:00
|
|
|
DerivedMesh *mesh_create_derived(Mesh *me, float (*vertCos)[3])
|
2005-03-27 22:34:18 +02:00
|
|
|
{
|
2013-12-25 22:26:41 +01:00
|
|
|
DerivedMesh *dm = CDDM_from_mesh(me);
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2012-03-24 07:18:31 +01:00
|
|
|
if (!dm)
|
2008-07-25 20:57:16 +02:00
|
|
|
return NULL;
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2013-05-30 19:36:43 +02:00
|
|
|
if (vertCos) {
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
CDDM_apply_vert_coords(dm, vertCos);
|
2013-05-30 19:36:43 +02:00
|
|
|
}
|
2005-03-27 23:27:12 +02:00
|
|
|
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
return dm;
|
2005-03-27 22:34:18 +02:00
|
|
|
}
|
|
|
|
|
2012-12-21 08:28:14 +01:00
|
|
|
static float (*get_editbmesh_orco_verts(BMEditMesh *em))[3]
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 22:51:28 +02:00
|
|
|
{
|
2009-05-16 18:18:08 +02:00
|
|
|
BMIter iter;
|
|
|
|
BMVert *eve;
|
2012-12-21 08:28:14 +01:00
|
|
|
float (*orco)[3];
|
|
|
|
int i;
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 22:51:28 +02:00
|
|
|
|
|
|
|
/* these may not really be the orco's, but it's only for preview.
|
|
|
|
* could be solver better once, but isn't simple */
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2018-01-14 22:14:20 +01:00
|
|
|
orco = MEM_malloc_arrayN(em->bm->totvert, sizeof(float) * 3, "BMEditMesh Orco");
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 22:51:28 +02:00
|
|
|
|
2012-12-21 08:28:14 +01:00
|
|
|
BM_ITER_MESH_INDEX (eve, &iter, em->bm, BM_VERTS_OF_MESH, i) {
|
|
|
|
copy_v3_v3(orco[i], eve->co);
|
2011-11-06 16:39:20 +01:00
|
|
|
}
|
2018-06-17 17:05:51 +02:00
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 22:51:28 +02:00
|
|
|
return orco;
|
|
|
|
}
|
|
|
|
|
2009-05-23 05:24:15 +02:00
|
|
|
/* orco custom data layer */
|
2018-10-09 06:04:51 +02:00
|
|
|
static float (*get_orco_coords(Object *ob, BMEditMesh *em, int layer, int *free))[3]
|
2010-03-30 13:49:07 +02:00
|
|
|
{
|
2012-05-12 21:18:02 +02:00
|
|
|
*free = 0;
|
2010-03-30 13:49:07 +02:00
|
|
|
|
2012-03-24 07:18:31 +01:00
|
|
|
if (layer == CD_ORCO) {
|
2010-03-30 13:49:07 +02:00
|
|
|
/* get original coordinates */
|
2012-05-12 21:18:02 +02:00
|
|
|
*free = 1;
|
2010-03-30 13:49:07 +02:00
|
|
|
|
2012-03-24 07:18:31 +01:00
|
|
|
if (em)
|
2012-12-21 08:28:14 +01:00
|
|
|
return get_editbmesh_orco_verts(em);
|
2010-03-30 13:49:07 +02:00
|
|
|
else
|
2012-12-21 08:28:14 +01:00
|
|
|
return BKE_mesh_orco_verts_get(ob);
|
2010-03-30 13:49:07 +02:00
|
|
|
}
|
2012-03-24 07:18:31 +01:00
|
|
|
else if (layer == CD_CLOTH_ORCO) {
|
2010-03-30 13:49:07 +02:00
|
|
|
/* apply shape key for cloth, this should really be solved
|
2012-03-09 19:28:30 +01:00
|
|
|
* by a more flexible customdata system, but not simple */
|
2012-03-24 07:18:31 +01:00
|
|
|
if (!em) {
|
2010-03-30 13:49:07 +02:00
|
|
|
ClothModifierData *clmd = (ClothModifierData *)modifiers_findByType(ob, eModifierType_Cloth);
|
2012-09-19 12:12:07 +02:00
|
|
|
KeyBlock *kb = BKE_keyblock_from_key(BKE_key_from_object(ob), clmd->sim_parms->shapekey_rest);
|
2011-12-05 00:13:28 +01:00
|
|
|
|
2015-04-13 07:54:37 +02:00
|
|
|
if (kb && kb->data) {
|
2011-12-05 00:13:28 +01:00
|
|
|
return kb->data;
|
2015-04-13 07:54:37 +02:00
|
|
|
}
|
2010-03-30 13:49:07 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2018-05-11 09:53:52 +02:00
|
|
|
static Mesh *create_orco_mesh(Object *ob, Mesh *me, BMEditMesh *em, int layer)
|
|
|
|
{
|
|
|
|
Mesh *mesh;
|
|
|
|
float (*orco)[3];
|
|
|
|
int free;
|
|
|
|
|
|
|
|
if (em) {
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
mesh = BKE_mesh_from_bmesh_for_eval_nomain(em->bm, NULL);
|
2018-05-11 09:53:52 +02:00
|
|
|
}
|
|
|
|
else {
|
2018-09-20 12:33:45 +02:00
|
|
|
mesh = BKE_mesh_copy_for_eval(me, true);
|
2018-05-11 09:53:52 +02:00
|
|
|
}
|
|
|
|
|
2018-10-09 06:04:51 +02:00
|
|
|
orco = get_orco_coords(ob, em, layer, &free);
|
2018-05-11 09:53:52 +02:00
|
|
|
|
|
|
|
if (orco) {
|
|
|
|
BKE_mesh_apply_vert_coords(mesh, orco);
|
|
|
|
if (free) MEM_freeN(orco);
|
|
|
|
}
|
|
|
|
|
|
|
|
return mesh;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void add_orco_mesh(
|
|
|
|
Object *ob, BMEditMesh *em, Mesh *mesh,
|
2019-03-28 11:42:53 +01:00
|
|
|
Mesh *mesh_orco, int layer)
|
2018-05-11 09:53:52 +02:00
|
|
|
{
|
|
|
|
float (*orco)[3], (*layerorco)[3];
|
|
|
|
int totvert, free;
|
|
|
|
|
|
|
|
totvert = mesh->totvert;
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_orco) {
|
2018-05-11 09:53:52 +02:00
|
|
|
free = 1;
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_orco->totvert == totvert) {
|
|
|
|
orco = BKE_mesh_vertexCos_get(mesh_orco, NULL);
|
2018-05-11 09:53:52 +02:00
|
|
|
}
|
|
|
|
else {
|
|
|
|
orco = BKE_mesh_vertexCos_get(mesh, NULL);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* TODO(sybren): totvert should potentially change here, as ob->data
|
|
|
|
* or em may have a different number of vertices than dm. */
|
2018-10-09 06:04:51 +02:00
|
|
|
orco = get_orco_coords(ob, em, layer, &free);
|
2018-05-11 09:53:52 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
if (orco) {
|
|
|
|
if (layer == CD_ORCO) {
|
|
|
|
BKE_mesh_orco_verts_transform(ob->data, orco, totvert, 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!(layerorco = CustomData_get_layer(&mesh->vdata, layer))) {
|
|
|
|
CustomData_add_layer(&mesh->vdata, layer, CD_CALLOC, NULL, mesh->totvert);
|
|
|
|
BKE_mesh_update_customdata_pointers(mesh, false);
|
|
|
|
|
|
|
|
layerorco = CustomData_get_layer(&mesh->vdata, layer);
|
|
|
|
}
|
|
|
|
|
|
|
|
memcpy(layerorco, orco, sizeof(float) * 3 * totvert);
|
|
|
|
if (free) MEM_freeN(orco);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-10-10 05:28:31 +02:00
|
|
|
static void editmesh_update_statvis_color(const Scene *scene, Object *ob)
|
2013-04-17 11:27:23 +02:00
|
|
|
{
|
|
|
|
BMEditMesh *em = BKE_editmesh_from_object(ob);
|
2018-10-09 06:57:38 +02:00
|
|
|
Mesh *me = ob->data;
|
|
|
|
BKE_mesh_runtime_ensure_edit_data(me);
|
|
|
|
BKE_editmesh_statvis_calc(em, me->runtime.edit_data, &scene->toolsettings->statvis);
|
2013-04-17 11:27:23 +02:00
|
|
|
}
|
2011-04-15 07:20:18 +02:00
|
|
|
|
|
|
|
static void shapekey_layers_to_keyblocks(DerivedMesh *dm, Mesh *me, int actshape_uid)
|
|
|
|
{
|
|
|
|
KeyBlock *kb;
|
|
|
|
int i, j, tot;
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2011-04-15 07:20:18 +02:00
|
|
|
if (!me->key)
|
2012-10-21 07:46:41 +02:00
|
|
|
return;
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2011-04-15 07:20:18 +02:00
|
|
|
tot = CustomData_number_of_layers(&dm->vertData, CD_SHAPEKEY);
|
2012-05-12 21:18:02 +02:00
|
|
|
for (i = 0; i < tot; i++) {
|
2011-04-15 07:20:18 +02:00
|
|
|
CustomDataLayer *layer = &dm->vertData.layers[CustomData_get_layer_index_n(&dm->vertData, CD_SHAPEKEY, i)];
|
|
|
|
float (*cos)[3], (*kbcos)[3];
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2012-05-12 21:18:02 +02:00
|
|
|
for (kb = me->key->block.first; kb; kb = kb->next) {
|
2011-04-15 07:20:18 +02:00
|
|
|
if (kb->uid == layer->uid)
|
|
|
|
break;
|
|
|
|
}
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2011-04-15 07:20:18 +02:00
|
|
|
if (!kb) {
|
2012-09-19 12:12:07 +02:00
|
|
|
kb = BKE_keyblock_add(me->key, layer->name);
|
2011-04-15 07:20:18 +02:00
|
|
|
kb->uid = layer->uid;
|
|
|
|
}
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2011-04-15 07:20:18 +02:00
|
|
|
if (kb->data)
|
|
|
|
MEM_freeN(kb->data);
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2011-04-15 07:20:18 +02:00
|
|
|
cos = CustomData_get_layer_n(&dm->vertData, CD_SHAPEKEY, i);
|
|
|
|
kb->totelem = dm->numVertData;
|
|
|
|
|
2018-02-01 15:53:49 +01:00
|
|
|
kb->data = kbcos = MEM_malloc_arrayN(kb->totelem, 3 * sizeof(float), "kbcos DerivedMesh.c");
|
2011-04-15 07:20:18 +02:00
|
|
|
if (kb->uid == actshape_uid) {
|
|
|
|
MVert *mvert = dm->getVertArray(dm);
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2012-05-12 21:18:02 +02:00
|
|
|
for (j = 0; j < dm->numVertData; j++, kbcos++, mvert++) {
|
2011-04-15 07:20:18 +02:00
|
|
|
copy_v3_v3(*kbcos, mvert->co);
|
|
|
|
}
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
|
|
|
else {
|
2012-05-12 21:18:02 +02:00
|
|
|
for (j = 0; j < kb->totelem; j++, cos++, kbcos++) {
|
2011-04-15 07:20:18 +02:00
|
|
|
copy_v3_v3(*kbcos, *cos);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2012-05-12 21:18:02 +02:00
|
|
|
for (kb = me->key->block.first; kb; kb = kb->next) {
|
2011-04-15 07:20:18 +02:00
|
|
|
if (kb->totelem != dm->numVertData) {
|
|
|
|
if (kb->data)
|
|
|
|
MEM_freeN(kb->data);
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2011-04-15 07:20:18 +02:00
|
|
|
kb->totelem = dm->numVertData;
|
2018-01-14 22:14:20 +01:00
|
|
|
kb->data = MEM_calloc_arrayN(kb->totelem, 3 * sizeof(float), "kb->data derivedmesh.c");
|
2019-02-01 02:44:19 +01:00
|
|
|
CLOG_ERROR(&LOG, "lost a shapekey layer: '%s'! (bmesh internal error)", kb->name);
|
2011-04-15 07:20:18 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-12-04 17:52:30 +01:00
|
|
|
static void mesh_copy_autosmooth(Mesh *me, Mesh *me_orig)
|
|
|
|
{
|
|
|
|
if (me_orig->flag & ME_AUTOSMOOTH) {
|
|
|
|
me->flag |= ME_AUTOSMOOTH;
|
|
|
|
me->smoothresh = me_orig->smoothresh;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
static void mesh_calc_modifier_final_normals(
|
|
|
|
const Mesh *mesh_input,
|
|
|
|
const CustomData_MeshMasks *dataMask,
|
|
|
|
const bool sculpt_dyntopo,
|
2019-03-29 21:13:55 +01:00
|
|
|
Mesh *mesh_final)
|
2019-03-28 11:42:53 +01:00
|
|
|
{
|
|
|
|
/* Compute normals. */
|
|
|
|
const bool do_loop_normals = ((mesh_input->flag & ME_AUTOSMOOTH) != 0 ||
|
|
|
|
(dataMask->lmask & CD_MASK_NORMAL) != 0);
|
|
|
|
/* Some modifiers may need this info from their target (other) object, simpler to generate it here as well.
|
|
|
|
* Note that they will always be generated when no loop normals are comptuted,
|
|
|
|
* since they are needed by drawing code. */
|
|
|
|
const bool do_poly_normals = ((dataMask->pmask & CD_MASK_NORMAL) != 0);
|
|
|
|
|
|
|
|
if (do_loop_normals) {
|
|
|
|
/* In case we also need poly normals, add the layer here, then BKE_mesh_calc_normals_split() will fill it. */
|
|
|
|
if (do_poly_normals) {
|
|
|
|
if (!CustomData_has_layer(&mesh_final->pdata, CD_NORMAL)) {
|
|
|
|
CustomData_add_layer(&mesh_final->pdata, CD_NORMAL, CD_CALLOC, NULL, mesh_final->totpoly);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/* Compute loop normals (note: will compute poly and vert normals as well, if needed!) */
|
|
|
|
BKE_mesh_calc_normals_split(mesh_final);
|
|
|
|
BKE_mesh_tessface_clear(mesh_final);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (sculpt_dyntopo == false) {
|
|
|
|
/* watch this! after 2.75a we move to from tessface to looptri (by default) */
|
|
|
|
if (dataMask->fmask & CD_MASK_MFACE) {
|
|
|
|
BKE_mesh_tessface_ensure(mesh_final);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* without this, drawing ngon tri's faces will show ugly tessellated face
|
|
|
|
* normals and will also have to calculate normals on the fly, try avoid
|
|
|
|
* this where possible since calculating polygon normals isn't fast,
|
|
|
|
* note that this isn't a problem for subsurf (only quads) or editmode
|
|
|
|
* which deals with drawing differently.
|
|
|
|
*
|
|
|
|
* Only calc vertex normals if they are flagged as dirty.
|
|
|
|
* If using loop normals, poly nors have already been computed.
|
|
|
|
*/
|
|
|
|
if (!do_loop_normals) {
|
|
|
|
BKE_mesh_ensure_normals_for_display(mesh_final);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Some modifiers, like datatransfer, may generate those data as temp layer, we do not want to keep them,
|
|
|
|
* as they are used by display code when available (i.e. even if autosmooth is disabled). */
|
|
|
|
if (!do_loop_normals && CustomData_has_layer(&mesh_final->ldata, CD_NORMAL)) {
|
|
|
|
CustomData_free_layers(&mesh_final->ldata, CD_NORMAL, mesh_final->totloop);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
static void mesh_calc_modifiers(
|
2019-03-27 19:07:16 +01:00
|
|
|
struct Depsgraph *depsgraph,
|
|
|
|
Scene *scene, Object *ob,
|
2018-05-22 19:42:41 +02:00
|
|
|
int useDeform,
|
2019-03-27 19:07:16 +01:00
|
|
|
const bool need_mapping,
|
|
|
|
const CustomData_MeshMasks *dataMask,
|
|
|
|
const int index,
|
2019-03-28 11:42:53 +01:00
|
|
|
const bool use_cache,
|
2015-07-02 08:20:22 +02:00
|
|
|
/* return args */
|
2019-03-27 19:07:16 +01:00
|
|
|
Mesh **r_deform,
|
|
|
|
Mesh **r_final)
|
2005-03-27 22:34:18 +02:00
|
|
|
{
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Input and final mesh. Final mesh is only created the moment the first
|
|
|
|
* constructive modifier is executed, or a deform modifier needs normals
|
|
|
|
* or certain data layers. */
|
|
|
|
Mesh *mesh_input = ob->data;
|
|
|
|
Mesh *mesh_final = NULL;
|
|
|
|
Mesh *mesh_deform = NULL;
|
|
|
|
BLI_assert((mesh_input->id.tag & LIB_TAG_COPIED_ON_WRITE_EVAL_RESULT) == 0);
|
|
|
|
|
|
|
|
/* Deformed vertex locations array. Deform only modifier need this type of
|
|
|
|
* float array rather than MVert*. Tracked along with mesh_final as an
|
|
|
|
* optimization to avoid copying coordinates back and forth if there are
|
|
|
|
* multiple sequential deform only modifiers. */
|
|
|
|
float (*deformed_verts)[3] = NULL;
|
|
|
|
int num_deformed_verts = mesh_input->totvert;
|
2014-01-27 17:52:21 +01:00
|
|
|
bool isPrevDeform = false;
|
2019-03-28 11:42:53 +01:00
|
|
|
|
|
|
|
/* Mesh with constructive modifiers but no deformation applied. Tracked
|
|
|
|
* along with final mesh if undeformed / orco coordinates are requested
|
|
|
|
* for texturing. */
|
|
|
|
Mesh *mesh_orco = NULL;
|
|
|
|
Mesh *mesh_orco_cloth = NULL;
|
|
|
|
|
|
|
|
/* Modifier evaluation modes. */
|
|
|
|
const bool use_render = (DEG_get_mode(depsgraph) == DAG_EVAL_RENDER);
|
|
|
|
const int required_mode = use_render ? eModifierMode_Render : eModifierMode_Realtime;
|
|
|
|
|
|
|
|
/* Sculpt can skip certain modifiers. */
|
2012-05-12 21:18:02 +02:00
|
|
|
MultiresModifierData *mmd = get_multires_modifier(scene, ob, 0);
|
2014-01-27 17:52:21 +01:00
|
|
|
const bool has_multires = (mmd && mmd->sculptlvl != 0);
|
|
|
|
bool multires_applied = false;
|
2019-03-28 11:42:53 +01:00
|
|
|
const bool sculpt_mode = ob->mode & OB_MODE_SCULPT && ob->sculpt && !use_render;
|
|
|
|
const bool sculpt_dyntopo = (sculpt_mode && ob->sculpt->bm) && !use_render;
|
2013-02-22 05:09:04 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Modifier evaluation contexts for different types of modifiers. */
|
|
|
|
ModifierApplyFlag app_render = use_render ? MOD_APPLY_RENDER : 0;
|
|
|
|
ModifierApplyFlag app_cache = use_cache ? MOD_APPLY_USECACHE : 0;
|
|
|
|
const ModifierEvalContext mectx = {depsgraph, ob, app_render | app_cache};
|
|
|
|
const ModifierEvalContext mectx_orco = {depsgraph, ob, app_render | MOD_APPLY_ORCO};
|
2014-04-13 12:18:51 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Get effective list of modifiers to execute. Some effects like shape keys
|
|
|
|
* are added as virtual modifiers before the user created modifiers. */
|
2013-08-19 11:05:34 +02:00
|
|
|
VirtualModifierData virtualModifierData;
|
2019-03-28 11:42:53 +01:00
|
|
|
ModifierData *firstmd = modifiers_getVirtualModifierList(ob, &virtualModifierData);
|
|
|
|
ModifierData *md = firstmd;
|
2013-08-19 11:05:34 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Preview colors by modifiers such as dynamic paint, to show the results
|
|
|
|
* even if the resulting data is not used in a material. Only in object mode.
|
|
|
|
* TODO: this is broken, not drawn by the drawn manager. */
|
|
|
|
const bool do_mod_mcol = (ob->mode == OB_MODE_OBJECT);
|
|
|
|
ModifierData *previewmd = NULL;
|
|
|
|
CustomData_MeshMasks previewmask = {0};
|
2018-09-30 17:07:38 +02:00
|
|
|
if (do_mod_mcol) {
|
Add weight preview to WeightVG modifiers, and first, simple/basic refactor of how modifiers can generate preview.
User side:
* Preview for DynamicPaint should keep the same behavior (for now). Weight preview should be somawhat quicker, though.
* Preview for WeightVG modifiers is only active in WeightPaint mode, and if the affected vgroup is the active one.
* Last active preview modifier in stack wins!
Note: that modifier preview topic is yet to be further refined, quite raw/incomplete for now.
Dev side:
* In draw code, renamed DRAW_DYNAMIC_PAINT_PREVIEW flag to DRAW_MODIFIERS_PREVIEW
* Removed use of MOD_DPAINT_PREVIEW_READY in DynamicPaint code (seems unecessary, and if it was, should be of more general scope).
* Added eModifierTypeFlag_UsesPreview to ModifierTypeFlag, for modifiers that can generate some preview data.
* Added three new modifier funcs, to handle preview modifiers in draw code / mod stack.
* For weights preview: added the generic DM_update_weight_mcol func, which can update WEIGHT_MCOL layer with either a given array of weights (currently used by DynamicPaint only), or from current active vgroup(s).
So now, draw code is fully generic (i.e. no more modifier-type checking in it). Mod stack code is generic to some extent, but will need more work.
2012-01-22 18:54:23 +01:00
|
|
|
/* Find the last active modifier generating a preview, or NULL if none. */
|
|
|
|
/* XXX Currently, DPaint modifier just ignores this.
|
|
|
|
* Needs a stupid hack...
|
|
|
|
* The whole "modifier preview" thing has to be (re?)designed, anyway! */
|
|
|
|
previewmd = modifiers_getLastPreview(scene, md, required_mode);
|
|
|
|
}
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Compute accumulated datamasks needed by each modifier. It helps to do
|
|
|
|
* this fine grained so that for example vertex groups are preserved up to
|
|
|
|
* an armature modifier, but not through a following subsurf modifier where
|
|
|
|
* subdividing them is expensive. */
|
|
|
|
CDMaskLink *datamasks = modifiers_calcDataMasks(scene, ob, md, dataMask, required_mode, previewmd, &previewmask);
|
|
|
|
CDMaskLink *md_datamask = datamasks;
|
|
|
|
/* XXX Always copying POLYINDEX, else tessellated data are no more valid! */
|
|
|
|
CustomData_MeshMasks append_mask = CD_MASK_BAREMESH_ORIGINDEX;
|
2005-07-19 02:21:01 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Clear errors before evaluation. */
|
|
|
|
modifiers_clearErrors(ob);
|
2019-03-18 15:37:46 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Apply all leading deform modifiers. */
|
2012-03-24 07:18:31 +01:00
|
|
|
if (useDeform) {
|
2019-03-28 11:42:53 +01:00
|
|
|
for (; md; md = md->next, md_datamask = md_datamask->next) {
|
2015-03-30 12:17:07 +02:00
|
|
|
const ModifierTypeInfo *mti = modifierType_getInfo(md->type);
|
2005-07-19 22:14:17 +02:00
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
if (!modifier_isEnabled(scene, md, required_mode)) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (useDeform < 0 && mti->dependsOnTime && mti->dependsOnTime(md)) {
|
|
|
|
continue;
|
|
|
|
}
|
2006-08-28 03:12:36 +02:00
|
|
|
|
2012-12-30 19:29:41 +01:00
|
|
|
if (mti->type == eModifierTypeType_OnlyDeform && !sculpt_dyntopo) {
|
2019-03-28 11:42:53 +01:00
|
|
|
if (!deformed_verts) {
|
|
|
|
deformed_verts = BKE_mesh_vertexCos_get(mesh_input, &num_deformed_verts);
|
2019-03-18 15:37:46 +01:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
else if (isPrevDeform && mti->dependsOnNormals && mti->dependsOnNormals(md)) {
|
|
|
|
if (mesh_final == NULL) {
|
|
|
|
mesh_final = BKE_mesh_copy_for_eval(mesh_input, true);
|
|
|
|
ASSERT_IS_VALID_MESH(mesh_final);
|
2019-03-18 15:37:46 +01:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
BKE_mesh_apply_vert_coords(mesh_final, deformed_verts);
|
2019-03-18 15:37:46 +01:00
|
|
|
}
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
modwrap_deformVerts(md, &mectx, mesh_final, deformed_verts, num_deformed_verts);
|
2005-07-19 22:14:17 +02:00
|
|
|
|
2019-03-18 15:37:46 +01:00
|
|
|
isPrevDeform = true;
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
|
|
|
else {
|
2005-07-19 22:14:17 +02:00
|
|
|
break;
|
|
|
|
}
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2008-07-28 13:01:34 +02:00
|
|
|
/* grab modifiers until index i */
|
2015-07-02 08:20:22 +02:00
|
|
|
if ((index != -1) && (BLI_findindex(&ob->modifiers, md) >= index))
|
2008-07-28 13:01:34 +02:00
|
|
|
break;
|
2005-07-19 22:14:17 +02:00
|
|
|
}
|
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
/* Result of all leading deforming modifiers is cached for
|
|
|
|
* places that wish to use the original mesh but with deformed
|
2019-03-28 11:42:53 +01:00
|
|
|
* coordinates (like vertex paint). */
|
2018-10-10 06:00:37 +02:00
|
|
|
if (r_deform) {
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_deform = BKE_mesh_copy_for_eval(mesh_input, true);
|
2018-05-11 09:53:52 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (deformed_verts) {
|
|
|
|
BKE_mesh_apply_vert_coords(mesh_deform, deformed_verts);
|
2006-09-07 08:44:25 +02:00
|
|
|
}
|
2006-08-30 05:05:30 +02:00
|
|
|
}
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
2006-08-30 12:55:46 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Apply all remaining constructive and deforming modifiers. */
|
|
|
|
for (; md; md = md->next, md_datamask = md_datamask->next) {
|
2015-03-30 12:17:07 +02:00
|
|
|
const ModifierTypeInfo *mti = modifierType_getInfo(md->type);
|
2005-07-19 22:14:17 +02:00
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
if (!modifier_isEnabled(scene, md, required_mode)) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (mti->type == eModifierTypeType_OnlyDeform && !useDeform) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if ((mti->flags & eModifierTypeFlag_RequiresOriginalData) && mesh_final) {
|
2012-10-27 13:12:09 +02:00
|
|
|
modifier_setError(md, "Modifier requires original data, bad stack position");
|
- added eModifierTypeFlag_RequiresOriginalData for modifiers that
can only follow deform (for example, they store mesh vertex
indices)
- added ModifierType.foreachObjectLink for iterating over Object
links inside modifier data (used for file load, relinking, etc)
- switched various modifiers_ functions to take object argument
instead of ListBase
- added user editable name field to modifiers
- bug fix, duplicate and make single user didn't relink object
pointers in modifier data
- added modifiers to outliner, needs icon
- added armature, hook, and softbody modifiers (softbody doesn't
do anything atm). added conversion of old hooks to modifiers.
NOTE-THE-FIRST: User name field is not initialized on loading 2.38 files
so if you have saved stuff with a cvs blender you will see blank names.
NOTE-THE-SECOND: Since modifiers aren't evaluated yet for non-Mesh
objects, hooks for lattices and curves are broken. Don't updated if
you actually, say, *use* Blender.
NOTE-THE-THIRD: Old hooks used a quirky weighting system during
deformation which can't be extended to modifiers. On the upside,
I doubt anyone relied on the old quirky system and the new system
makes much more sense. (Although the way falloff works is still
quite stupid I think).
2005-08-11 00:05:52 +02:00
|
|
|
continue;
|
|
|
|
}
|
2015-07-02 08:20:22 +02:00
|
|
|
|
2012-12-30 19:29:41 +01:00
|
|
|
if (sculpt_mode &&
|
2014-05-09 10:29:21 +02:00
|
|
|
(!has_multires || multires_applied || sculpt_dyntopo))
|
2012-12-30 19:29:41 +01:00
|
|
|
{
|
2014-01-27 17:52:21 +01:00
|
|
|
bool unsupported = false;
|
2011-05-04 15:15:42 +02:00
|
|
|
|
2013-06-03 10:26:12 +02:00
|
|
|
if (md->type == eModifierType_Multires && ((MultiresModifierData *)md)->sculptlvl == 0) {
|
|
|
|
/* If multires is on level 0 skip it silently without warning message. */
|
2015-07-02 08:20:22 +02:00
|
|
|
if (!sculpt_dyntopo) {
|
2014-05-08 18:25:46 +02:00
|
|
|
continue;
|
2015-07-02 08:20:22 +02:00
|
|
|
}
|
2013-06-03 10:26:12 +02:00
|
|
|
}
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (sculpt_dyntopo)
|
2014-01-27 17:52:21 +01:00
|
|
|
unsupported = true;
|
2012-12-30 19:29:41 +01:00
|
|
|
|
2012-03-24 07:18:31 +01:00
|
|
|
if (scene->toolsettings->sculpt->flags & SCULPT_ONLY_DEFORM)
|
2014-01-27 17:52:21 +01:00
|
|
|
unsupported |= (mti->type != eModifierTypeType_OnlyDeform);
|
2011-05-04 15:15:42 +02:00
|
|
|
|
2012-05-12 21:18:02 +02:00
|
|
|
unsupported |= multires_applied;
|
2011-05-04 15:15:42 +02:00
|
|
|
|
2012-03-24 07:18:31 +01:00
|
|
|
if (unsupported) {
|
2014-05-08 18:25:46 +02:00
|
|
|
if (sculpt_dyntopo)
|
|
|
|
modifier_setError(md, "Not supported in dyntopo");
|
|
|
|
else
|
|
|
|
modifier_setError(md, "Not supported in sculpt mode");
|
2011-01-31 21:02:51 +01:00
|
|
|
continue;
|
|
|
|
}
|
2014-05-08 18:15:36 +02:00
|
|
|
else {
|
|
|
|
modifier_setError(md, "Hide, Mask and optimized display disabled");
|
|
|
|
}
|
2011-05-04 15:15:42 +02:00
|
|
|
}
|
2015-07-02 08:20:22 +02:00
|
|
|
|
|
|
|
if (need_mapping && !modifier_supportsMapping(md)) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (useDeform < 0 && mti->dependsOnTime && mti->dependsOnTime(md)) {
|
|
|
|
continue;
|
|
|
|
}
|
2005-07-19 22:14:17 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Add an orco layer if needed by this modifier. */
|
|
|
|
if (mesh_final && mti->requiredDataMask) {
|
|
|
|
CustomData_MeshMasks mask = {0};
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
mti->requiredDataMask(ob, md, &mask);
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mask.vmask & CD_MASK_ORCO) {
|
|
|
|
add_orco_mesh(ob, NULL, mesh_final, mesh_orco, CD_ORCO);
|
|
|
|
}
|
2018-05-11 09:53:52 +02:00
|
|
|
}
|
2007-12-05 13:40:54 +01:00
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
/* How to apply modifier depends on (a) what we already have as
|
2018-05-11 09:53:52 +02:00
|
|
|
* a result of previous modifiers (could be a Mesh or just
|
2019-03-28 11:42:53 +01:00
|
|
|
* deformed vertices) and (b) what type the modifier is. */
|
2012-03-24 07:18:31 +01:00
|
|
|
if (mti->type == eModifierTypeType_OnlyDeform) {
|
2006-08-28 03:12:36 +02:00
|
|
|
/* No existing verts to deform, need to build them. */
|
2019-03-28 11:42:53 +01:00
|
|
|
if (!deformed_verts) {
|
|
|
|
if (mesh_final) {
|
2018-05-11 09:53:52 +02:00
|
|
|
/* Deforming a mesh, read the vertex locations
|
2006-08-28 03:12:36 +02:00
|
|
|
* out of the mesh and deform them. Once done with this
|
2019-03-28 11:42:53 +01:00
|
|
|
* run of deformers verts will be written back. */
|
|
|
|
deformed_verts = BKE_mesh_vertexCos_get(mesh_final, &num_deformed_verts);
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
|
|
|
else {
|
2019-03-28 11:42:53 +01:00
|
|
|
deformed_verts = BKE_mesh_vertexCos_get(mesh_input, &num_deformed_verts);
|
2005-07-19 22:14:17 +02:00
|
|
|
}
|
|
|
|
}
|
2010-09-30 12:51:36 +02:00
|
|
|
/* if this is not the last modifier in the stack then recalculate the normals
|
|
|
|
* to avoid giving bogus normals to the next modifier see: [#23673] */
|
2019-03-28 11:42:53 +01:00
|
|
|
else if (isPrevDeform && mti->dependsOnNormals && mti->dependsOnNormals(md)) {
|
2010-09-30 12:51:36 +02:00
|
|
|
/* XXX, this covers bug #23673, but we may need normal calc for other types */
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_final) {
|
|
|
|
BKE_mesh_apply_vert_coords(mesh_final, deformed_verts);
|
2010-09-30 12:51:36 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
modwrap_deformVerts(md, &mectx, mesh_final, deformed_verts, num_deformed_verts);
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
|
|
|
else {
|
2010-10-21 03:08:12 +02:00
|
|
|
/* determine which data layers are needed by following modifiers */
|
2019-03-28 11:42:53 +01:00
|
|
|
CustomData_MeshMasks nextmask;
|
|
|
|
if (md_datamask->next)
|
|
|
|
nextmask = md_datamask->next->mask;
|
2010-10-21 03:08:12 +02:00
|
|
|
else
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
nextmask = *dataMask;
|
2010-10-21 03:08:12 +02:00
|
|
|
|
2018-05-11 09:53:52 +02:00
|
|
|
/* apply vertex coordinates or build a Mesh as necessary */
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_final) {
|
|
|
|
if (deformed_verts) {
|
|
|
|
BKE_mesh_apply_vert_coords(mesh_final, deformed_verts);
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
|
|
|
else {
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_final = BKE_mesh_copy_for_eval(mesh_input, true);
|
|
|
|
ASSERT_IS_VALID_MESH(mesh_final);
|
2006-08-28 03:12:36 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (deformed_verts) {
|
|
|
|
BKE_mesh_apply_vert_coords(mesh_final, deformed_verts);
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
2009-05-23 05:24:15 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Initialize original indices the first time we evaluate a
|
|
|
|
* constructive modifier. Modifiers will then do mapping mostly
|
|
|
|
* automatic by copying them through CustomData_copy_data along
|
|
|
|
* with other data.
|
2010-10-21 03:08:12 +02:00
|
|
|
*
|
2019-03-28 11:42:53 +01:00
|
|
|
* These are created when either requested by evaluation, or if
|
|
|
|
* following modifiers requested them. */
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
if (need_mapping || ((nextmask.vmask | nextmask.emask | nextmask.pmask) & CD_MASK_ORIGINDEX)) {
|
2010-10-21 03:08:12 +02:00
|
|
|
/* calc */
|
2019-03-28 11:42:53 +01:00
|
|
|
CustomData_add_layer(&mesh_final->vdata, CD_ORIGINDEX, CD_CALLOC, NULL, mesh_final->totvert);
|
|
|
|
CustomData_add_layer(&mesh_final->edata, CD_ORIGINDEX, CD_CALLOC, NULL, mesh_final->totedge);
|
|
|
|
CustomData_add_layer(&mesh_final->pdata, CD_ORIGINDEX, CD_CALLOC, NULL, mesh_final->totpoly);
|
2010-02-09 19:06:57 +01:00
|
|
|
|
2016-05-29 20:14:42 +02:00
|
|
|
/* Not worth parallelizing this, gives less than 0.1% overall speedup in best of best cases... */
|
2019-03-28 11:42:53 +01:00
|
|
|
range_vn_i(CustomData_get_layer(&mesh_final->vdata, CD_ORIGINDEX), mesh_final->totvert, 0);
|
|
|
|
range_vn_i(CustomData_get_layer(&mesh_final->edata, CD_ORIGINDEX), mesh_final->totedge, 0);
|
|
|
|
range_vn_i(CustomData_get_layer(&mesh_final->pdata, CD_ORIGINDEX), mesh_final->totpoly, 0);
|
2010-02-09 19:06:57 +01:00
|
|
|
}
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2018-05-11 09:53:52 +02:00
|
|
|
/* set the Mesh to only copy needed data */
|
2019-03-28 11:42:53 +01:00
|
|
|
CustomData_MeshMasks mask = md_datamask->mask;
|
2016-01-28 06:30:50 +01:00
|
|
|
/* needMapping check here fixes bug [#28112], otherwise it's
|
|
|
|
* possible that it won't be copied */
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_MeshMasks_update(&mask, &append_mask);
|
|
|
|
if (need_mapping) {
|
|
|
|
mask.vmask |= CD_MASK_ORIGINDEX;
|
|
|
|
mask.emask |= CD_MASK_ORIGINDEX;
|
|
|
|
mask.pmask |= CD_MASK_ORIGINDEX;
|
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_set_only_copy(mesh_final, &mask);
|
2018-05-11 09:53:52 +02:00
|
|
|
|
2016-01-28 06:30:50 +01:00
|
|
|
/* add cloth rest shape key if needed */
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
if (mask.vmask & CD_MASK_CLOTH_ORCO) {
|
2019-03-28 11:42:53 +01:00
|
|
|
add_orco_mesh(ob, NULL, mesh_final, mesh_orco, CD_CLOTH_ORCO);
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
}
|
2010-03-30 13:49:07 +02:00
|
|
|
|
2007-12-05 13:40:54 +01:00
|
|
|
/* add an origspace layer if needed */
|
2019-03-28 11:42:53 +01:00
|
|
|
if ((md_datamask->mask.lmask) & CD_MASK_ORIGSPACE_MLOOP) {
|
|
|
|
if (!CustomData_has_layer(&mesh_final->ldata, CD_ORIGSPACE_MLOOP)) {
|
|
|
|
CustomData_add_layer(&mesh_final->ldata, CD_ORIGSPACE_MLOOP, CD_CALLOC, NULL, mesh_final->totloop);
|
|
|
|
mesh_init_origspace(mesh_final);
|
2012-02-05 12:30:26 +01:00
|
|
|
}
|
|
|
|
}
|
2007-12-05 13:40:54 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
Mesh *mesh_next = modwrap_applyModifier(md, &mectx, mesh_final);
|
|
|
|
ASSERT_IS_VALID_MESH(mesh_next);
|
2006-08-28 03:12:36 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_next) {
|
2018-05-11 09:53:52 +02:00
|
|
|
/* if the modifier returned a new mesh, release the old one */
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_final != mesh_next) {
|
|
|
|
BLI_assert(mesh_final != mesh_input);
|
|
|
|
BKE_id_free(NULL, mesh_final);
|
2018-05-11 09:53:52 +02:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_final = mesh_next;
|
2005-07-26 04:44:59 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (deformed_verts) {
|
|
|
|
MEM_freeN(deformed_verts);
|
|
|
|
deformed_verts = NULL;
|
2005-07-19 22:14:17 +02:00
|
|
|
}
|
2018-12-04 17:52:30 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_copy_autosmooth(mesh_final, mesh_input);
|
2012-04-01 17:02:19 +02:00
|
|
|
}
|
2010-03-30 13:49:07 +02:00
|
|
|
|
2018-05-11 09:53:52 +02:00
|
|
|
/* create an orco mesh in parallel */
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
if (nextmask.vmask & CD_MASK_ORCO) {
|
2019-03-28 11:42:53 +01:00
|
|
|
if (!mesh_orco) {
|
|
|
|
mesh_orco = create_orco_mesh(ob, mesh_input, NULL, CD_ORCO);
|
2018-05-11 09:53:52 +02:00
|
|
|
}
|
2010-03-30 13:49:07 +02:00
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
nextmask.vmask &= ~CD_MASK_ORCO;
|
2019-03-11 22:21:13 +01:00
|
|
|
CustomData_MeshMasks temp_cddata_masks = {
|
|
|
|
.vmask = CD_MASK_ORIGINDEX,
|
|
|
|
.emask = CD_MASK_ORIGINDEX,
|
|
|
|
.fmask = CD_MASK_ORIGINDEX,
|
|
|
|
.pmask = CD_MASK_ORIGINDEX,
|
|
|
|
};
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
if (mti->requiredDataMask != NULL) {
|
|
|
|
mti->requiredDataMask(ob, md, &temp_cddata_masks);
|
|
|
|
}
|
|
|
|
CustomData_MeshMasks_update(&temp_cddata_masks, &nextmask);
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_set_only_copy(mesh_orco, &temp_cddata_masks);
|
2013-05-30 19:36:43 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_next = modwrap_applyModifier(md, &mectx_orco, mesh_orco);
|
|
|
|
ASSERT_IS_VALID_MESH(mesh_next);
|
2010-03-30 13:49:07 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_next) {
|
2018-05-11 09:53:52 +02:00
|
|
|
/* if the modifier returned a new mesh, release the old one */
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_orco != mesh_next) {
|
|
|
|
BLI_assert(mesh_orco != mesh_input);
|
|
|
|
BKE_id_free(NULL, mesh_orco);
|
2018-05-11 09:53:52 +02:00
|
|
|
}
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_orco = mesh_next;
|
2010-03-30 13:49:07 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-05-11 09:53:52 +02:00
|
|
|
/* create cloth orco mesh in parallel */
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
if (nextmask.vmask & CD_MASK_CLOTH_ORCO) {
|
2019-03-28 11:42:53 +01:00
|
|
|
if (!mesh_orco_cloth) {
|
|
|
|
mesh_orco_cloth = create_orco_mesh(ob, mesh_input, NULL, CD_CLOTH_ORCO);
|
2018-05-16 17:35:54 +02:00
|
|
|
}
|
2010-03-30 13:49:07 +02:00
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
nextmask.vmask &= ~CD_MASK_CLOTH_ORCO;
|
|
|
|
nextmask.vmask |= CD_MASK_ORIGINDEX;
|
|
|
|
nextmask.emask |= CD_MASK_ORIGINDEX;
|
|
|
|
nextmask.pmask |= CD_MASK_ORIGINDEX;
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_set_only_copy(mesh_orco_cloth, &nextmask);
|
2013-05-30 19:36:43 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_next = modwrap_applyModifier(md, &mectx_orco, mesh_orco_cloth);
|
|
|
|
ASSERT_IS_VALID_MESH(mesh_next);
|
2010-03-30 13:49:07 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_next) {
|
2018-05-11 09:53:52 +02:00
|
|
|
/* if the modifier returned a new mesh, release the old one */
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_orco_cloth != mesh_next) {
|
|
|
|
BLI_assert(mesh_orco != mesh_input);
|
|
|
|
BKE_id_free(NULL, mesh_orco_cloth);
|
2015-07-02 08:20:22 +02:00
|
|
|
}
|
2018-05-11 09:53:52 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_orco_cloth = mesh_next;
|
2010-03-30 13:49:07 +02:00
|
|
|
}
|
|
|
|
}
|
2011-07-08 13:03:37 +02:00
|
|
|
|
|
|
|
/* in case of dynamic paint, make sure preview mask remains for following modifiers */
|
Add weight preview to WeightVG modifiers, and first, simple/basic refactor of how modifiers can generate preview.
User side:
* Preview for DynamicPaint should keep the same behavior (for now). Weight preview should be somawhat quicker, though.
* Preview for WeightVG modifiers is only active in WeightPaint mode, and if the affected vgroup is the active one.
* Last active preview modifier in stack wins!
Note: that modifier preview topic is yet to be further refined, quite raw/incomplete for now.
Dev side:
* In draw code, renamed DRAW_DYNAMIC_PAINT_PREVIEW flag to DRAW_MODIFIERS_PREVIEW
* Removed use of MOD_DPAINT_PREVIEW_READY in DynamicPaint code (seems unecessary, and if it was, should be of more general scope).
* Added eModifierTypeFlag_UsesPreview to ModifierTypeFlag, for modifiers that can generate some preview data.
* Added three new modifier funcs, to handle preview modifiers in draw code / mod stack.
* For weights preview: added the generic DM_update_weight_mcol func, which can update WEIGHT_MCOL layer with either a given array of weights (currently used by DynamicPaint only), or from current active vgroup(s).
So now, draw code is fully generic (i.e. no more modifier-type checking in it). Mod stack code is generic to some extent, but will need more work.
2012-01-22 18:54:23 +01:00
|
|
|
/* XXX Temp and hackish solution! */
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
if (md->type == eModifierType_DynamicPaint) {
|
|
|
|
append_mask.lmask |= CD_MASK_PREVIEW_MLOOPCOL;
|
|
|
|
}
|
2017-05-29 16:05:35 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_final->runtime.deformed_only = false;
|
2005-07-19 22:14:17 +02:00
|
|
|
}
|
2010-05-03 18:06:36 +02:00
|
|
|
|
2012-05-12 21:18:02 +02:00
|
|
|
isPrevDeform = (mti->type == eModifierTypeType_OnlyDeform);
|
2010-09-30 12:51:36 +02:00
|
|
|
|
2008-07-28 13:01:34 +02:00
|
|
|
/* grab modifiers until index i */
|
2015-07-02 08:20:22 +02:00
|
|
|
if ((index != -1) && (BLI_findindex(&ob->modifiers, md) >= index))
|
2008-07-28 13:01:34 +02:00
|
|
|
break;
|
2011-01-31 21:02:51 +01:00
|
|
|
|
2014-01-27 17:52:21 +01:00
|
|
|
if (sculpt_mode && md->type == eModifierType_Multires) {
|
|
|
|
multires_applied = true;
|
|
|
|
}
|
2005-07-19 22:14:17 +02:00
|
|
|
}
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
BLI_linklist_free((LinkNode *)datamasks, NULL);
|
|
|
|
|
2012-05-12 21:18:02 +02:00
|
|
|
for (md = firstmd; md; md = md->next)
|
2007-12-17 12:47:24 +01:00
|
|
|
modifier_freeTemporaryData(md);
|
|
|
|
|
2018-05-11 09:53:52 +02:00
|
|
|
/* Yay, we are done. If we have a Mesh and deformed vertices
|
|
|
|
* need to apply these back onto the Mesh. If we have no
|
2019-03-28 11:42:53 +01:00
|
|
|
* Mesh then we need to build one. */
|
|
|
|
if (mesh_final == NULL) {
|
|
|
|
mesh_final = BKE_mesh_copy_for_eval(mesh_input, true);
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
if (deformed_verts) {
|
|
|
|
BKE_mesh_apply_vert_coords(mesh_final, deformed_verts);
|
|
|
|
MEM_freeN(deformed_verts);
|
|
|
|
deformed_verts = NULL;
|
2005-03-27 22:34:18 +02:00
|
|
|
}
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Add orco coordinates to final and deformed mesh if requested. */
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
if (dataMask->vmask & CD_MASK_ORCO) {
|
2019-03-28 11:42:53 +01:00
|
|
|
add_orco_mesh(ob, NULL, mesh_final, mesh_orco, CD_ORCO);
|
2007-12-05 13:40:54 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_deform)
|
|
|
|
add_orco_mesh(ob, NULL, mesh_deform, NULL, CD_ORCO);
|
2008-01-09 15:40:25 +01:00
|
|
|
}
|
2007-12-05 13:40:54 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_orco) {
|
|
|
|
BKE_id_free(NULL, mesh_orco);
|
2014-04-13 12:18:51 +02:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_orco_cloth) {
|
|
|
|
BKE_id_free(NULL, mesh_orco_cloth);
|
2012-02-06 07:56:54 +01:00
|
|
|
}
|
2011-12-07 02:12:53 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Compute normals. */
|
|
|
|
mesh_calc_modifier_final_normals(mesh_input, dataMask, sculpt_dyntopo, mesh_final);
|
2017-01-27 18:47:11 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Return final mesh */
|
|
|
|
*r_final = mesh_final;
|
|
|
|
if (r_deform) {
|
|
|
|
*r_deform = mesh_deform;
|
2018-05-11 09:53:52 +02:00
|
|
|
}
|
2005-03-27 22:34:18 +02:00
|
|
|
}
|
|
|
|
|
2018-12-05 21:44:25 +01:00
|
|
|
|
2014-03-15 17:24:05 +01:00
|
|
|
float (*editbmesh_get_vertex_cos(BMEditMesh *em, int *r_numVerts))[3]
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
{
|
2009-05-16 18:18:08 +02:00
|
|
|
BMIter iter;
|
|
|
|
BMVert *eve;
|
2012-12-21 08:28:14 +01:00
|
|
|
float (*cos)[3];
|
|
|
|
int i;
|
|
|
|
|
2014-03-15 17:24:05 +01:00
|
|
|
*r_numVerts = em->bm->totvert;
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
|
2018-01-14 22:14:20 +01:00
|
|
|
cos = MEM_malloc_arrayN(em->bm->totvert, 3 * sizeof(float), "vertexcos");
|
2009-05-16 18:18:08 +02:00
|
|
|
|
2012-12-21 08:28:14 +01:00
|
|
|
BM_ITER_MESH_INDEX (eve, &iter, em->bm, BM_VERTS_OF_MESH, i) {
|
2011-11-06 16:39:20 +01:00
|
|
|
copy_v3_v3(cos[i], eve->co);
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
}
|
|
|
|
|
2005-10-09 18:59:10 +02:00
|
|
|
return cos;
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
}
|
|
|
|
|
2018-10-09 06:04:51 +02:00
|
|
|
bool editbmesh_modifier_is_enabled(Scene *scene, ModifierData *md, bool has_prev_mesh)
|
2007-07-28 23:04:30 +02:00
|
|
|
{
|
2015-03-30 12:17:07 +02:00
|
|
|
const ModifierTypeInfo *mti = modifierType_getInfo(md->type);
|
2015-07-02 08:20:22 +02:00
|
|
|
const int required_mode = eModifierMode_Realtime | eModifierMode_Editmode;
|
|
|
|
|
|
|
|
if (!modifier_isEnabled(scene, md, required_mode)) {
|
|
|
|
return false;
|
|
|
|
}
|
2007-07-28 23:04:30 +02:00
|
|
|
|
2018-10-09 06:04:51 +02:00
|
|
|
if ((mti->flags & eModifierTypeFlag_RequiresOriginalData) && has_prev_mesh) {
|
2012-10-27 13:12:09 +02:00
|
|
|
modifier_setError(md, "Modifier requires original data, bad stack position");
|
2014-12-01 17:11:18 +01:00
|
|
|
return false;
|
2007-07-28 23:04:30 +02:00
|
|
|
}
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2014-12-01 17:11:18 +01:00
|
|
|
return true;
|
2007-07-28 23:04:30 +02:00
|
|
|
}
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
static void editbmesh_calc_modifier_final_normals(
|
|
|
|
const Mesh *mesh_input,
|
|
|
|
const CustomData_MeshMasks *dataMask,
|
2019-03-29 21:13:55 +01:00
|
|
|
Mesh *mesh_final)
|
2019-03-28 11:42:53 +01:00
|
|
|
{
|
|
|
|
const bool do_loop_normals = ((mesh_input->flag & ME_AUTOSMOOTH) != 0 ||
|
|
|
|
(dataMask->lmask & CD_MASK_NORMAL) != 0);
|
|
|
|
/* Some modifiers may need this info from their target (other) object, simpler to generate it here as well. */
|
|
|
|
const bool do_poly_normals = ((dataMask->pmask & CD_MASK_NORMAL) != 0);
|
|
|
|
|
|
|
|
if (do_loop_normals) {
|
|
|
|
/* In case we also need poly normals, add the layer here, then BKE_mesh_calc_normals_split() will fill it. */
|
|
|
|
if (do_poly_normals) {
|
|
|
|
if (!CustomData_has_layer(&mesh_final->pdata, CD_NORMAL)) {
|
|
|
|
CustomData_add_layer(&mesh_final->pdata, CD_NORMAL, CD_CALLOC, NULL, mesh_final->totpoly);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/* Compute loop normals */
|
|
|
|
BKE_mesh_calc_normals_split(mesh_final);
|
|
|
|
BKE_mesh_tessface_clear(mesh_final);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* BMESH_ONLY, ensure tessface's used for drawing,
|
|
|
|
* but don't recalculate if the last modifier in the stack gives us tessfaces
|
|
|
|
* check if the derived meshes are DM_TYPE_EDITBMESH before calling, this isn't essential
|
|
|
|
* but quiets annoying error messages since tessfaces wont be created. */
|
|
|
|
if (dataMask->fmask & CD_MASK_MFACE) {
|
|
|
|
if (mesh_final->edit_mesh == NULL) {
|
|
|
|
BKE_mesh_tessface_ensure(mesh_final);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* same as mesh_calc_modifiers (if using loop normals, poly nors have already been computed). */
|
|
|
|
if (!do_loop_normals) {
|
|
|
|
BKE_mesh_ensure_normals_for_display(mesh_final);
|
|
|
|
|
|
|
|
/* Some modifiers, like datatransfer, may generate those data, we do not want to keep them,
|
|
|
|
* as they are used by display code when available (i.e. even if autosmooth is disabled). */
|
|
|
|
if (CustomData_has_layer(&mesh_final->ldata, CD_NORMAL)) {
|
|
|
|
CustomData_free_layers(&mesh_final->ldata, CD_NORMAL, mesh_final->totloop);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
static void editbmesh_calc_modifiers(
|
2019-03-27 19:07:16 +01:00
|
|
|
struct Depsgraph *depsgraph,
|
|
|
|
Scene *scene,
|
|
|
|
Object *ob,
|
2019-03-28 11:42:53 +01:00
|
|
|
BMEditMesh *em_input,
|
2019-03-27 19:07:16 +01:00
|
|
|
const CustomData_MeshMasks *dataMask,
|
2015-07-02 08:20:22 +02:00
|
|
|
/* return args */
|
2019-03-27 19:07:16 +01:00
|
|
|
Mesh **r_cage,
|
|
|
|
Mesh **r_final)
|
2013-07-05 02:13:14 +02:00
|
|
|
{
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Input and final mesh. Final mesh is only created the moment the first
|
|
|
|
* constructive modifier is executed, or a deform modifier needs normals
|
|
|
|
* or certain data layers. */
|
|
|
|
Mesh *mesh_input = ob->data;
|
|
|
|
Mesh *mesh_final = NULL;
|
|
|
|
Mesh *mesh_cage = NULL;
|
|
|
|
|
|
|
|
/* Deformed vertex locations array. Deform only modifier need this type of
|
|
|
|
* float array rather than MVert*. Tracked along with mesh_final as an
|
|
|
|
* optimization to avoid copying coordinates back and forth if there are
|
|
|
|
* multiple sequential deform only modifiers. */
|
|
|
|
float (*deformed_verts)[3] = NULL;
|
|
|
|
int num_deformed_verts = 0;
|
2019-03-18 15:37:46 +01:00
|
|
|
bool isPrevDeform = false;
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Mesh with constructive modifiers but no deformation applied. Tracked
|
|
|
|
* along with final mesh if undeformed / orco coordinates are requested
|
|
|
|
* for texturing. */
|
|
|
|
Mesh *mesh_orco = NULL;
|
2018-05-01 17:33:04 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Modifier evaluation modes. */
|
|
|
|
const int required_mode = eModifierMode_Realtime | eModifierMode_Editmode;
|
|
|
|
const bool do_init_statvis = false; /* FIXME: use V3D_OVERLAY_EDIT_STATVIS. */
|
2014-04-13 12:18:51 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Modifier evaluation contexts for different types of modifiers. */
|
|
|
|
const ModifierEvalContext mectx = {depsgraph, ob, MOD_APPLY_USECACHE};
|
|
|
|
const ModifierEvalContext mectx_orco = {depsgraph, ob, MOD_APPLY_ORCO};
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Evaluate modifiers up to certain index to get the mesh cage. */
|
|
|
|
int cageIndex = modifiers_getCageIndex(scene, ob, NULL, 1);
|
2015-07-02 08:20:22 +02:00
|
|
|
if (r_cage && cageIndex == -1) {
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_cage = BKE_mesh_from_editmesh_with_coords_thin_wrap(em_input, dataMask, NULL);
|
|
|
|
mesh_copy_autosmooth(mesh_cage, mesh_input);
|
2005-08-03 06:04:05 +02:00
|
|
|
}
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Get effective list of modifiers to execute. Some effects like shape keys
|
|
|
|
* are added as virtual modifiers before the user created modifiers. */
|
|
|
|
VirtualModifierData virtualModifierData;
|
|
|
|
ModifierData *md = modifiers_getVirtualModifierList(ob, &virtualModifierData);
|
2006-12-05 18:42:03 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Compute accumulated datamasks needed by each modifier. It helps to do
|
|
|
|
* this fine grained so that for example vertex groups are preserved up to
|
|
|
|
* an armature modifier, but not through a following subsurf modifier where
|
|
|
|
* subdividing them is expensive. */
|
|
|
|
CDMaskLink *datamasks = modifiers_calcDataMasks(scene, ob, md, dataMask, required_mode, NULL, NULL);
|
|
|
|
CDMaskLink *md_datamask = datamasks;
|
|
|
|
CustomData_MeshMasks append_mask = CD_MASK_BAREMESH;
|
2018-10-10 06:00:37 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Clear errors before evaluation. */
|
|
|
|
modifiers_clearErrors(ob);
|
2018-10-10 06:00:37 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
for (int i = 0; md; i++, md = md->next, md_datamask = md_datamask->next) {
|
2015-03-30 12:17:07 +02:00
|
|
|
const ModifierTypeInfo *mti = modifierType_getInfo(md->type);
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (!editbmesh_modifier_is_enabled(scene, md, mesh_final != NULL)) {
|
- added eModifierTypeFlag_RequiresOriginalData for modifiers that
can only follow deform (for example, they store mesh vertex
indices)
- added ModifierType.foreachObjectLink for iterating over Object
links inside modifier data (used for file load, relinking, etc)
- switched various modifiers_ functions to take object argument
instead of ListBase
- added user editable name field to modifiers
- bug fix, duplicate and make single user didn't relink object
pointers in modifier data
- added modifiers to outliner, needs icon
- added armature, hook, and softbody modifiers (softbody doesn't
do anything atm). added conversion of old hooks to modifiers.
NOTE-THE-FIRST: User name field is not initialized on loading 2.38 files
so if you have saved stuff with a cvs blender you will see blank names.
NOTE-THE-SECOND: Since modifiers aren't evaluated yet for non-Mesh
objects, hooks for lattices and curves are broken. Don't updated if
you actually, say, *use* Blender.
NOTE-THE-THIRD: Old hooks used a quirky weighting system during
deformation which can't be extended to modifiers. On the upside,
I doubt anyone relied on the old quirky system and the new system
makes much more sense. (Although the way falloff works is still
quite stupid I think).
2005-08-11 00:05:52 +02:00
|
|
|
continue;
|
2015-07-02 08:20:22 +02:00
|
|
|
}
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Add an orco layer if needed by this modifier. */
|
|
|
|
if (mesh_final && mti->requiredDataMask) {
|
|
|
|
CustomData_MeshMasks mask = {0};
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
mti->requiredDataMask(ob, md, &mask);
|
|
|
|
if (mask.vmask & CD_MASK_ORCO) {
|
2019-03-28 11:42:53 +01:00
|
|
|
add_orco_mesh(ob, em_input, mesh_final, mesh_orco, CD_ORCO);
|
2018-10-09 06:04:51 +02:00
|
|
|
}
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 22:51:28 +02:00
|
|
|
}
|
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
/* How to apply modifier depends on (a) what we already have as
|
2019-03-28 11:42:53 +01:00
|
|
|
* a result of previous modifiers (could be a mesh or just
|
|
|
|
* deformed vertices) and (b) what type the modifier is. */
|
2012-03-24 07:18:31 +01:00
|
|
|
if (mti->type == eModifierTypeType_OnlyDeform) {
|
2006-08-28 03:12:36 +02:00
|
|
|
/* No existing verts to deform, need to build them. */
|
2019-03-28 11:42:53 +01:00
|
|
|
if (!deformed_verts) {
|
|
|
|
if (mesh_final) {
|
2006-08-28 03:12:36 +02:00
|
|
|
/* Deforming a derived mesh, read the vertex locations
|
|
|
|
* out of the mesh and deform them. Once done with this
|
2019-03-28 11:42:53 +01:00
|
|
|
* run of deformers verts will be written back. */
|
|
|
|
deformed_verts = BKE_mesh_vertexCos_get(mesh_final, &num_deformed_verts);
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
|
|
|
else {
|
2019-03-28 11:42:53 +01:00
|
|
|
deformed_verts = editbmesh_get_vertex_cos(em_input, &num_deformed_verts);
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
}
|
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
else if (isPrevDeform && mti->dependsOnNormals && mti->dependsOnNormals(md)) {
|
|
|
|
if (mesh_final == NULL) {
|
|
|
|
mesh_final = BKE_mesh_from_bmesh_for_eval_nomain(em_input->bm, NULL);
|
|
|
|
ASSERT_IS_VALID_MESH(mesh_final);
|
|
|
|
mesh_copy_autosmooth(mesh_final, mesh_input);
|
2019-03-18 15:37:46 +01:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
BLI_assert(deformed_verts != NULL);
|
|
|
|
BKE_mesh_apply_vert_coords(mesh_final, deformed_verts);
|
2019-03-18 15:37:46 +01:00
|
|
|
}
|
|
|
|
|
2018-09-19 16:59:05 +02:00
|
|
|
if (mti->deformVertsEM)
|
2019-03-28 11:42:53 +01:00
|
|
|
modwrap_deformVertsEM(md, &mectx, em_input, mesh_final, deformed_verts, num_deformed_verts);
|
2012-05-09 17:00:26 +02:00
|
|
|
else
|
2019-03-28 11:42:53 +01:00
|
|
|
modwrap_deformVerts(md, &mectx, mesh_final, deformed_verts, num_deformed_verts);
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
|
|
|
else {
|
2006-08-28 03:12:36 +02:00
|
|
|
/* apply vertex coordinates or build a DerivedMesh as necessary */
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_final) {
|
|
|
|
if (deformed_verts) {
|
|
|
|
Mesh *mesh_tmp = BKE_mesh_copy_for_eval(mesh_final, false);
|
|
|
|
if (mesh_final != mesh_cage) {
|
|
|
|
BKE_id_free(NULL, mesh_final);
|
2018-10-09 06:04:51 +02:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_final = mesh_tmp;
|
|
|
|
BKE_mesh_apply_vert_coords(mesh_final, deformed_verts);
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
else if (mesh_final == mesh_cage) {
|
2018-10-09 06:04:51 +02:00
|
|
|
/* 'me' may be changed by this modifier, so we need to copy it. */
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_final = BKE_mesh_copy_for_eval(mesh_final, false);
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
|
|
|
else {
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_final = BKE_mesh_from_bmesh_for_eval_nomain(em_input->bm, NULL);
|
|
|
|
ASSERT_IS_VALID_MESH(mesh_final);
|
2006-08-28 03:12:36 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_copy_autosmooth(mesh_final, mesh_input);
|
2018-12-04 17:52:30 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (deformed_verts) {
|
|
|
|
BKE_mesh_apply_vert_coords(mesh_final, deformed_verts);
|
2006-08-28 03:12:36 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 22:51:28 +02:00
|
|
|
/* create an orco derivedmesh in parallel */
|
2019-03-28 11:42:53 +01:00
|
|
|
CustomData_MeshMasks mask = md_datamask->mask;
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
if (mask.vmask & CD_MASK_ORCO) {
|
2019-03-28 11:42:53 +01:00
|
|
|
if (!mesh_orco) {
|
|
|
|
mesh_orco = create_orco_mesh(ob, mesh_input, em_input, CD_ORCO);
|
2018-10-09 06:04:51 +02:00
|
|
|
}
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 22:51:28 +02:00
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
mask.vmask &= ~CD_MASK_ORCO;
|
|
|
|
mask.vmask |= CD_MASK_ORIGINDEX;
|
|
|
|
mask.emask |= CD_MASK_ORIGINDEX;
|
|
|
|
mask.pmask |= CD_MASK_ORIGINDEX;
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_set_only_copy(mesh_orco, &mask);
|
2010-05-03 12:09:26 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
Mesh *mesh_next = modwrap_applyModifier(md, &mectx_orco, mesh_orco);
|
|
|
|
ASSERT_IS_VALID_MESH(mesh_next);
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 22:51:28 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_next) {
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 22:51:28 +02:00
|
|
|
/* if the modifier returned a new dm, release the old one */
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_orco && mesh_orco != mesh_next) {
|
|
|
|
BKE_id_free(NULL, mesh_orco);
|
2018-10-09 06:04:51 +02:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_orco = mesh_next;
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 22:51:28 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2006-12-05 18:42:03 +01:00
|
|
|
/* set the DerivedMesh to only copy needed data */
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_MeshMasks_update(&mask, &append_mask);
|
2019-04-16 17:55:11 +02:00
|
|
|
/* XXX WHAT? ovewrites mask ??? */
|
|
|
|
/* CD_MASK_ORCO may have been cleared above */
|
|
|
|
mask = md_datamask->mask;
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
mask.vmask |= CD_MASK_ORIGINDEX;
|
|
|
|
mask.emask |= CD_MASK_ORIGINDEX;
|
|
|
|
mask.pmask |= CD_MASK_ORIGINDEX;
|
2006-12-05 18:42:03 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_set_only_copy(mesh_final, &mask);
|
2010-10-21 03:08:12 +02:00
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
if (mask.lmask & CD_MASK_ORIGSPACE_MLOOP) {
|
2019-03-28 11:42:53 +01:00
|
|
|
if (!CustomData_has_layer(&mesh_final->ldata, CD_ORIGSPACE_MLOOP)) {
|
|
|
|
CustomData_add_layer(&mesh_final->ldata, CD_ORIGSPACE_MLOOP, CD_CALLOC, NULL, mesh_final->totloop);
|
|
|
|
mesh_init_origspace(mesh_final);
|
2012-02-05 12:30:26 +01:00
|
|
|
}
|
|
|
|
}
|
2013-05-30 19:36:43 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
Mesh *mesh_next = modwrap_applyModifier(md, &mectx, mesh_final);
|
|
|
|
ASSERT_IS_VALID_MESH(mesh_next);
|
2005-08-03 06:04:05 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_next) {
|
|
|
|
if (mesh_final && mesh_final != mesh_next) {
|
|
|
|
BKE_id_free(NULL, mesh_final);
|
2018-10-09 06:04:51 +02:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_final = mesh_next;
|
2005-08-03 06:04:05 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (deformed_verts) {
|
|
|
|
MEM_freeN(deformed_verts);
|
|
|
|
deformed_verts = NULL;
|
2005-08-03 06:04:05 +02:00
|
|
|
}
|
2018-12-04 17:52:30 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_copy_autosmooth(mesh_final, mesh_input);
|
2005-08-03 06:04:05 +02:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_final->runtime.deformed_only = false;
|
2005-08-03 06:04:05 +02:00
|
|
|
}
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
if (r_cage && i == cageIndex) {
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_final && deformed_verts) {
|
|
|
|
mesh_cage = BKE_mesh_copy_for_eval(mesh_final, false);
|
|
|
|
BKE_mesh_apply_vert_coords(mesh_cage, deformed_verts);
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
else if (mesh_final) {
|
|
|
|
mesh_cage = mesh_final;
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
|
|
|
else {
|
2019-03-28 11:42:53 +01:00
|
|
|
Mesh *me_orig = mesh_input;
|
2018-10-09 06:04:51 +02:00
|
|
|
if (me_orig->id.tag & LIB_TAG_COPIED_ON_WRITE) {
|
|
|
|
BKE_mesh_runtime_ensure_edit_data(me_orig);
|
2019-03-28 11:42:53 +01:00
|
|
|
me_orig->runtime.edit_data->vertexCos = MEM_dupallocN(deformed_verts);
|
2018-04-19 11:03:58 +02:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_cage = BKE_mesh_from_editmesh_with_coords_thin_wrap(
|
|
|
|
em_input, dataMask,
|
|
|
|
deformed_verts ? MEM_dupallocN(deformed_verts) : NULL);
|
|
|
|
mesh_copy_autosmooth(mesh_cage, mesh_input);
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
}
|
|
|
|
}
|
2019-03-18 15:37:46 +01:00
|
|
|
|
|
|
|
isPrevDeform = (mti->type == eModifierTypeType_OnlyDeform);
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
}
|
|
|
|
|
2012-05-12 21:18:02 +02:00
|
|
|
BLI_linklist_free((LinkNode *)datamasks, NULL);
|
2006-12-05 18:42:03 +01:00
|
|
|
|
2006-08-28 03:12:36 +02:00
|
|
|
/* Yay, we are done. If we have a DerivedMesh and deformed vertices need
|
|
|
|
* to apply these back onto the DerivedMesh. If we have no DerivedMesh
|
2019-03-28 11:42:53 +01:00
|
|
|
* then we need to build one. */
|
|
|
|
if (mesh_final) {
|
|
|
|
if (deformed_verts) {
|
|
|
|
Mesh *mesh_tmp = BKE_mesh_copy_for_eval(mesh_final, false);
|
|
|
|
if (mesh_final != mesh_cage) {
|
|
|
|
BKE_id_free(NULL, mesh_final);
|
|
|
|
}
|
|
|
|
mesh_final = mesh_tmp;
|
2019-03-29 14:53:56 +01:00
|
|
|
BKE_mesh_apply_vert_coords(mesh_final, deformed_verts);
|
2015-07-02 08:20:22 +02:00
|
|
|
}
|
2012-02-28 11:22:21 +01:00
|
|
|
}
|
2019-03-28 11:42:53 +01:00
|
|
|
else if (!deformed_verts && mesh_cage) {
|
2012-02-28 11:22:21 +01:00
|
|
|
/* cage should already have up to date normals */
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_final = mesh_cage;
|
2013-04-13 22:20:21 +02:00
|
|
|
|
|
|
|
/* In this case, we should never have weight-modifying modifiers in stack... */
|
2018-10-09 06:04:51 +02:00
|
|
|
if (do_init_statvis) {
|
2018-10-10 05:28:31 +02:00
|
|
|
editmesh_update_statvis_color(scene, ob);
|
2018-10-09 06:04:51 +02:00
|
|
|
}
|
2012-02-28 11:22:21 +01:00
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* this is just a copy of the editmesh, no need to calc normals */
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_final = BKE_mesh_from_editmesh_with_coords_thin_wrap(em_input, dataMask, deformed_verts);
|
|
|
|
deformed_verts = NULL;
|
2013-04-13 22:20:21 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
mesh_copy_autosmooth(mesh_final, mesh_input);
|
2018-12-04 17:52:30 +01:00
|
|
|
|
2013-04-13 22:20:21 +02:00
|
|
|
/* In this case, we should never have weight-modifying modifiers in stack... */
|
2018-10-09 06:04:51 +02:00
|
|
|
if (do_init_statvis) {
|
2018-10-10 05:28:31 +02:00
|
|
|
editmesh_update_statvis_color(scene, ob);
|
2018-10-09 06:04:51 +02:00
|
|
|
}
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
}
|
2007-02-03 15:46:30 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (deformed_verts) {
|
|
|
|
MEM_freeN(deformed_verts);
|
2014-04-13 12:18:51 +02:00
|
|
|
}
|
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Add orco coordinates to final and deformed mesh if requested. */
|
|
|
|
if (dataMask->vmask & CD_MASK_ORCO) {
|
|
|
|
add_orco_mesh(ob, em_input, mesh_final, mesh_orco, CD_ORCO);
|
2012-02-28 11:22:21 +01:00
|
|
|
}
|
2018-12-04 17:52:30 +01:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
if (mesh_orco) {
|
|
|
|
BKE_id_free(NULL, mesh_orco);
|
2014-04-13 12:18:51 +02:00
|
|
|
}
|
2013-05-30 19:36:43 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Compute normals. */
|
|
|
|
editbmesh_calc_modifier_final_normals(mesh_input, dataMask, mesh_final);
|
|
|
|
if (mesh_cage && (mesh_cage != mesh_final)) {
|
|
|
|
editbmesh_calc_modifier_final_normals(mesh_input, dataMask, mesh_cage);
|
2018-10-09 06:04:51 +02:00
|
|
|
}
|
Merge of first part of changes from the apricot branch, especially
the features that are needed to run the game. Compile tested with
scons, make, but not cmake, that seems to have an issue not related
to these changes. The changes include:
* GLSL support in the viewport and game engine, enable in the game
menu in textured draw mode.
* Synced and merged part of the duplicated blender and gameengine/
gameplayer drawing code.
* Further refactoring of game engine drawing code, especially mesh
storage changed a lot.
* Optimizations in game engine armatures to avoid recomputations.
* A python function to get the framerate estimate in game.
* An option take object color into account in materials.
* An option to restrict shadow casters to a lamp's layers.
* Increase from 10 to 18 texture slots for materials, lamps, word.
An extra texture slot shows up once the last slot is used.
* Memory limit for undo, not enabled by default yet because it
needs the .B.blend to be changed.
* Multiple undo for image painting.
* An offset for dupligroups, so not all objects in a group have to
be at the origin.
2008-09-04 22:51:28 +02:00
|
|
|
|
2019-03-28 11:42:53 +01:00
|
|
|
/* Return final mesh. */
|
|
|
|
*r_final = mesh_final;
|
|
|
|
if (r_cage) {
|
|
|
|
*r_cage = mesh_cage;
|
2018-10-09 06:04:51 +02:00
|
|
|
}
|
- added data arguments to deformer modifiers, in case someone wants
to write one that is based on geometry (and not just vertex position)
- added editmode versions of modifier deform/apply calls and flag
to tag modifiers that support editmode
- added isFinalCalc param to applyModifier, basically a switch to let
subsurf know if it is calc'ng orco or not (so it can deal with cache
appropriately). This is kinda hacky and perhaps I can come up with
a better solution (its also a waste to do a complete subdivide just
to get vertex locations).
- changed ccgsubsurf to not preallocate hash's to be approximately correct
size... this was probably not a big performance savings but means that
the order of faces returned by the iterator can vary after the first
call, this messes up orco calculation so dropped for time being.
- minor bug fix, meshes with only key didn't get vertex normals correctly
calc'd
- updated editmesh derivedmesh to support auxiliary locations
- changed mesh_calc_modifiers to alloc deformVerts on demand
- added editmesh_calc_modifiers for calculating editmesh cage and final
derivedmesh's
- bug fix, update shadedisplist to always calc colors (even if totvert==0)
- changed load_editMesh and make_edge to build me->medge even if totedge==0
(incremental subsurf checks this)
todo: add drawFacesTex for ccgderivedmesh
So, modifiers in editmode are back (which means auto-mirror
in edit mode works now) although still not finished. Currently
no cage is computed, the cage is always the base mesh (in
other words, Optimal edge style editing is off), and the final
mesh currently includes all modifiers that work in edit mode
(including lattice and curve). At some point there will be toggles
for which modifiers affect the final/cage editmode derivedmesh's.
Also, very nice new feature is that incremental subsurf in object
mode returns a ccgderivedmesh object instead of copying to a new
displistmesh. This can make a *huge* speed difference, and is very
nice for working with deformed armatures (esp. with only small
per frame changes).
2005-07-22 09:37:15 +02:00
|
|
|
}
|
|
|
|
|
2018-05-30 12:56:20 +02:00
|
|
|
static void mesh_finalize_eval(Object *object)
|
|
|
|
{
|
|
|
|
Mesh *mesh = (Mesh *)object->data;
|
|
|
|
Mesh *mesh_eval = object->runtime.mesh_eval;
|
|
|
|
/* Special Tweaks for cases when evaluated mesh came from
|
|
|
|
* BKE_mesh_new_nomain_from_template().
|
|
|
|
*/
|
|
|
|
BLI_strncpy(mesh_eval->id.name, mesh->id.name, sizeof(mesh_eval->id.name));
|
|
|
|
if (mesh_eval->mat != NULL) {
|
|
|
|
MEM_freeN(mesh_eval->mat);
|
|
|
|
}
|
2018-06-04 15:11:09 +02:00
|
|
|
/* Set flag which makes it easier to see what's going on in a debugger. */
|
|
|
|
mesh_eval->id.tag |= LIB_TAG_COPIED_ON_WRITE_EVAL_RESULT;
|
2018-05-30 12:56:20 +02:00
|
|
|
mesh_eval->mat = MEM_dupallocN(mesh->mat);
|
|
|
|
mesh_eval->totcol = mesh->totcol;
|
|
|
|
/* Make evaluated mesh to share same edit mesh pointer as original
|
|
|
|
* and copied meshes.
|
|
|
|
*/
|
2019-02-17 08:05:18 +01:00
|
|
|
mesh_eval->edit_mesh = mesh->edit_mesh;
|
2018-05-30 12:56:20 +02:00
|
|
|
/* Copy autosmooth settings from original mesh.
|
|
|
|
* This is not done by BKE_mesh_new_nomain_from_template(), so need to take
|
|
|
|
* extra care here.
|
|
|
|
*/
|
|
|
|
mesh_eval->flag |= (mesh->flag & ME_AUTOSMOOTH);
|
|
|
|
mesh_eval->smoothresh = mesh->smoothresh;
|
|
|
|
/* Replace evaluated object's data with fully evaluated mesh. */
|
|
|
|
/* TODO(sergey): There was statement done by Sybren and Mai that this
|
|
|
|
* caused modifiers to be applied twice. which is weirtd and shouldn't
|
|
|
|
* really happen. But since there is no reference to the report, can not
|
|
|
|
* do much about this.
|
|
|
|
*/
|
2018-06-01 11:44:50 +02:00
|
|
|
|
|
|
|
/* Object is sometimes not evaluated!
|
|
|
|
* TODO(sergey): BAD TEMPORARY HACK FOR UNTIL WE ARE SMARTER */
|
2018-06-04 15:11:09 +02:00
|
|
|
if (object->id.tag & LIB_TAG_COPIED_ON_WRITE) {
|
2018-06-01 11:44:50 +02:00
|
|
|
object->data = mesh_eval;
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* evaluated will be available via: 'object->runtime.mesh_eval' */
|
|
|
|
}
|
2018-05-30 12:56:20 +02:00
|
|
|
}
|
|
|
|
|
Shrinkwrap: new mode that projects along the target normal.
The Nearest Surface Point shrink method, while fast, is neither
smooth nor continuous: as the source point moves, the projected
point can both stop and jump. This causes distortions in the
deformation of the shrinkwrap modifier, and the motion of an
animated object with a shrinkwrap constraint.
This patch implements a new mode, which, instead of using the simple
nearest point search, iteratively solves an equation for each triangle
to find a point which has its interpolated normal point to or from the
original vertex. Non-manifold boundary edges are treated as infinitely
thin cylinders that cast normals in all perpendicular directions.
Since this is useful for the constraint, and having multiple
objects with constraints targeting the same guide mesh is a quite
reasonable use case, rather than calculating the mesh boundary edge
data over and over again, it is precomputed and cached in the mesh.
Reviewers: mont29
Differential Revision: https://developer.blender.org/D3836
2018-11-06 19:04:53 +01:00
|
|
|
static void mesh_build_extra_data(struct Depsgraph *depsgraph, Object *ob)
|
|
|
|
{
|
|
|
|
uint32_t eval_flags = DEG_get_eval_flags_for_id(depsgraph, &ob->id);
|
|
|
|
|
|
|
|
if (eval_flags & DAG_EVAL_NEED_SHRINKWRAP_BOUNDARY) {
|
|
|
|
BKE_shrinkwrap_compute_boundary_data(ob->runtime.mesh_eval);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-11-20 19:25:42 +01:00
|
|
|
static void mesh_runtime_check_normals_valid(const Mesh *mesh)
|
2018-11-20 11:37:51 +01:00
|
|
|
{
|
|
|
|
UNUSED_VARS_NDEBUG(mesh);
|
|
|
|
BLI_assert(!(mesh->runtime.cd_dirty_vert & CD_MASK_NORMAL));
|
|
|
|
BLI_assert(!(mesh->runtime.cd_dirty_loop & CD_MASK_NORMAL));
|
|
|
|
BLI_assert(!(mesh->runtime.cd_dirty_poly & CD_MASK_NORMAL));
|
|
|
|
}
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
static void mesh_build_data(
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
struct Depsgraph *depsgraph, Scene *scene, Object *ob, const CustomData_MeshMasks *dataMask,
|
2019-03-27 19:07:16 +01:00
|
|
|
const bool need_mapping)
|
- shuffled editmesh derived function name/function
- added ModifierTypeInfo.freeData function
- added modifier_{new,free] utility function
- added ccgSubSurf_getUseAgeCounts to query info
- removed subsurf modifier faking (ME_SUBSURF flag is no
longer valid). subsurf modifier gets converted on file load
although there is obscure linked mesh situation where this
can go wrong, will fix shortly. this also means that some
places in the code that test/copy subsurf settings are broken
for the time being.
- shuffled modifier calculation to be simpler. note that
all modifiers are currently disabled in editmode (including
subsurf). don't worry, will return shortly.
- bug fix, build modifier didn't randomize meshes with only verts
- cleaned up subsurf_ccg and adapted for future editmode modifier
work
- added editmesh.derived{Cage,Final}, not used yet
- added SubsurfModifierData.{mCache,emCache}, will be used to cache
subsurf instead of caching in derivedmesh itself
- removed old subsurf buttons
- added do_modifiers_buttons to handle modifier events
- removed count_object counting of modifier (subsurfed) objects...
this would be nice to add back at some point but requires care.
probably requires rewrite of counting system.
New feature: Incremental Subsurf in Object Mode
The previous release introduce incremental subsurf calculation during
editmode but it was not turned on during object mode. In general it
does not make sense to have it always enabled during object mode because
it requires caching a fair amount of information about the mesh which
is a waste of memory unless the mesh is often recalculated.
However, for mesh's that have subsurfed armatures for example, or that
have other modifiers so that the mesh is essentially changing on every
frame, it makes a lot of sense to keep the subsurf'd object around and
that is what the new incremental subsurf modifier toggle is for. The
intent is that the user will enable this option for (a) a mesh that is
currently under active editing or (b) a mesh that is heavily updated
in the scene, such as a character.
I will try to write more about this feature for release, because it
has advantages and disadvantages that are not immediately obvious (the
first user reaction will be to turn it on for ever object, which is
probably not correct).
2005-07-21 22:30:33 +02:00
|
|
|
{
|
2012-11-21 01:31:47 +01:00
|
|
|
BLI_assert(ob->type == OB_MESH);
|
|
|
|
|
2018-12-01 17:06:44 +01:00
|
|
|
/* Evaluated meshes aren't supposed to be created on original instances. If you do,
|
|
|
|
* they aren't cleaned up properly on mode switch, causing crashes, e.g T58150. */
|
|
|
|
BLI_assert(ob->id.tag & LIB_TAG_COPIED_ON_WRITE);
|
|
|
|
|
2013-05-18 12:24:34 +02:00
|
|
|
BKE_object_free_derived_caches(ob);
|
|
|
|
BKE_object_sculpt_modifiers_changed(ob);
|
- shuffled editmesh derived function name/function
- added ModifierTypeInfo.freeData function
- added modifier_{new,free] utility function
- added ccgSubSurf_getUseAgeCounts to query info
- removed subsurf modifier faking (ME_SUBSURF flag is no
longer valid). subsurf modifier gets converted on file load
although there is obscure linked mesh situation where this
can go wrong, will fix shortly. this also means that some
places in the code that test/copy subsurf settings are broken
for the time being.
- shuffled modifier calculation to be simpler. note that
all modifiers are currently disabled in editmode (including
subsurf). don't worry, will return shortly.
- bug fix, build modifier didn't randomize meshes with only verts
- cleaned up subsurf_ccg and adapted for future editmode modifier
work
- added editmesh.derived{Cage,Final}, not used yet
- added SubsurfModifierData.{mCache,emCache}, will be used to cache
subsurf instead of caching in derivedmesh itself
- removed old subsurf buttons
- added do_modifiers_buttons to handle modifier events
- removed count_object counting of modifier (subsurfed) objects...
this would be nice to add back at some point but requires care.
probably requires rewrite of counting system.
New feature: Incremental Subsurf in Object Mode
The previous release introduce incremental subsurf calculation during
editmode but it was not turned on during object mode. In general it
does not make sense to have it always enabled during object mode because
it requires caching a fair amount of information about the mesh which
is a waste of memory unless the mesh is often recalculated.
However, for mesh's that have subsurfed armatures for example, or that
have other modifiers so that the mesh is essentially changing on every
frame, it makes a lot of sense to keep the subsurf'd object around and
that is what the new incremental subsurf modifier toggle is for. The
intent is that the user will enable this option for (a) a mesh that is
currently under active editing or (b) a mesh that is heavily updated
in the scene, such as a character.
I will try to write more about this feature for release, because it
has advantages and disadvantages that are not immediately obvious (the
first user reaction will be to turn it on for ever object, which is
probably not correct).
2005-07-21 22:30:33 +02:00
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
#if 0 /* XXX This is already taken care of in mesh_calc_modifiers()... */
|
2018-12-01 10:12:18 +01:00
|
|
|
if (need_mapping) {
|
|
|
|
/* Also add the flag so that it is recorded in lastDataMask. */
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
dataMask->vmask |= CD_MASK_ORIGINDEX;
|
|
|
|
dataMask->emask |= CD_MASK_ORIGINDEX;
|
|
|
|
dataMask->pmask |= CD_MASK_ORIGINDEX;
|
2018-12-01 10:12:18 +01:00
|
|
|
}
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
#endif
|
2018-12-01 10:12:18 +01:00
|
|
|
|
2018-05-30 12:56:20 +02:00
|
|
|
mesh_calc_modifiers(
|
2019-03-27 19:07:16 +01:00
|
|
|
depsgraph, scene, ob, 1, need_mapping, dataMask, -1, true,
|
2018-05-30 12:56:20 +02:00
|
|
|
&ob->runtime.mesh_deform_eval, &ob->runtime.mesh_eval);
|
|
|
|
|
2018-10-15 08:14:05 +02:00
|
|
|
BKE_object_boundbox_calc_from_mesh(ob, ob->runtime.mesh_eval);
|
2019-01-22 21:49:38 +01:00
|
|
|
/* Only copy texspace from orig mesh if some modifier (hint: smoke sim, see T58492)
|
|
|
|
* did not re-enable that flag (which always get disabled for eval mesh as a start). */
|
|
|
|
if (!(ob->runtime.mesh_eval->texflag & ME_AUTOSPACE)) {
|
|
|
|
BKE_mesh_texspace_copy_from_object(ob->runtime.mesh_eval, ob);
|
|
|
|
}
|
2005-07-19 04:36:21 +02:00
|
|
|
|
2018-11-25 18:39:57 +01:00
|
|
|
mesh_finalize_eval(ob);
|
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
ob->runtime.last_data_mask = *dataMask;
|
2018-12-03 16:09:45 +01:00
|
|
|
ob->runtime.last_need_mapping = need_mapping;
|
2012-03-14 07:30:55 +01:00
|
|
|
|
2018-04-05 18:20:27 +02:00
|
|
|
if ((ob->mode & OB_MODE_ALL_SCULPT) && ob->sculpt) {
|
2012-03-14 07:30:55 +01:00
|
|
|
/* create PBVH immediately (would be created on the fly too,
|
2012-04-22 13:54:53 +02:00
|
|
|
* but this avoids waiting on first stroke) */
|
2018-06-19 19:00:25 +02:00
|
|
|
/* XXX Disabled for now.
|
|
|
|
* This can create horrible nasty bugs by generating re-entrant call of mesh_get_eval_final! */
|
|
|
|
// BKE_sculpt_update_mesh_elements(depsgraph, scene, scene->toolsettings->sculpt, ob, false, false);
|
2012-03-14 07:30:55 +01:00
|
|
|
}
|
2013-05-30 19:36:43 +02:00
|
|
|
|
2018-11-20 11:37:51 +01:00
|
|
|
mesh_runtime_check_normals_valid(ob->runtime.mesh_eval);
|
Shrinkwrap: new mode that projects along the target normal.
The Nearest Surface Point shrink method, while fast, is neither
smooth nor continuous: as the source point moves, the projected
point can both stop and jump. This causes distortions in the
deformation of the shrinkwrap modifier, and the motion of an
animated object with a shrinkwrap constraint.
This patch implements a new mode, which, instead of using the simple
nearest point search, iteratively solves an equation for each triangle
to find a point which has its interpolated normal point to or from the
original vertex. Non-manifold boundary edges are treated as infinitely
thin cylinders that cast normals in all perpendicular directions.
Since this is useful for the constraint, and having multiple
objects with constraints targeting the same guide mesh is a quite
reasonable use case, rather than calculating the mesh boundary edge
data over and over again, it is precomputed and cached in the mesh.
Reviewers: mont29
Differential Revision: https://developer.blender.org/D3836
2018-11-06 19:04:53 +01:00
|
|
|
mesh_build_extra_data(depsgraph, ob);
|
- shuffled editmesh derived function name/function
- added ModifierTypeInfo.freeData function
- added modifier_{new,free] utility function
- added ccgSubSurf_getUseAgeCounts to query info
- removed subsurf modifier faking (ME_SUBSURF flag is no
longer valid). subsurf modifier gets converted on file load
although there is obscure linked mesh situation where this
can go wrong, will fix shortly. this also means that some
places in the code that test/copy subsurf settings are broken
for the time being.
- shuffled modifier calculation to be simpler. note that
all modifiers are currently disabled in editmode (including
subsurf). don't worry, will return shortly.
- bug fix, build modifier didn't randomize meshes with only verts
- cleaned up subsurf_ccg and adapted for future editmode modifier
work
- added editmesh.derived{Cage,Final}, not used yet
- added SubsurfModifierData.{mCache,emCache}, will be used to cache
subsurf instead of caching in derivedmesh itself
- removed old subsurf buttons
- added do_modifiers_buttons to handle modifier events
- removed count_object counting of modifier (subsurfed) objects...
this would be nice to add back at some point but requires care.
probably requires rewrite of counting system.
New feature: Incremental Subsurf in Object Mode
The previous release introduce incremental subsurf calculation during
editmode but it was not turned on during object mode. In general it
does not make sense to have it always enabled during object mode because
it requires caching a fair amount of information about the mesh which
is a waste of memory unless the mesh is often recalculated.
However, for mesh's that have subsurfed armatures for example, or that
have other modifiers so that the mesh is essentially changing on every
frame, it makes a lot of sense to keep the subsurf'd object around and
that is what the new incremental subsurf modifier toggle is for. The
intent is that the user will enable this option for (a) a mesh that is
currently under active editing or (b) a mesh that is heavily updated
in the scene, such as a character.
I will try to write more about this feature for release, because it
has advantages and disadvantages that are not immediately obvious (the
first user reaction will be to turn it on for ever object, which is
probably not correct).
2005-07-21 22:30:33 +02:00
|
|
|
}
|
|
|
|
|
2017-08-16 04:45:11 +02:00
|
|
|
static void editbmesh_build_data(
|
2018-04-06 12:07:27 +02:00
|
|
|
struct Depsgraph *depsgraph, Scene *scene,
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
Object *obedit, BMEditMesh *em, CustomData_MeshMasks *dataMask)
|
- shuffled editmesh derived function name/function
- added ModifierTypeInfo.freeData function
- added modifier_{new,free] utility function
- added ccgSubSurf_getUseAgeCounts to query info
- removed subsurf modifier faking (ME_SUBSURF flag is no
longer valid). subsurf modifier gets converted on file load
although there is obscure linked mesh situation where this
can go wrong, will fix shortly. this also means that some
places in the code that test/copy subsurf settings are broken
for the time being.
- shuffled modifier calculation to be simpler. note that
all modifiers are currently disabled in editmode (including
subsurf). don't worry, will return shortly.
- bug fix, build modifier didn't randomize meshes with only verts
- cleaned up subsurf_ccg and adapted for future editmode modifier
work
- added editmesh.derived{Cage,Final}, not used yet
- added SubsurfModifierData.{mCache,emCache}, will be used to cache
subsurf instead of caching in derivedmesh itself
- removed old subsurf buttons
- added do_modifiers_buttons to handle modifier events
- removed count_object counting of modifier (subsurfed) objects...
this would be nice to add back at some point but requires care.
probably requires rewrite of counting system.
New feature: Incremental Subsurf in Object Mode
The previous release introduce incremental subsurf calculation during
editmode but it was not turned on during object mode. In general it
does not make sense to have it always enabled during object mode because
it requires caching a fair amount of information about the mesh which
is a waste of memory unless the mesh is often recalculated.
However, for mesh's that have subsurfed armatures for example, or that
have other modifiers so that the mesh is essentially changing on every
frame, it makes a lot of sense to keep the subsurf'd object around and
that is what the new incremental subsurf modifier toggle is for. The
intent is that the user will enable this option for (a) a mesh that is
currently under active editing or (b) a mesh that is heavily updated
in the scene, such as a character.
I will try to write more about this feature for release, because it
has advantages and disadvantages that are not immediately obvious (the
first user reaction will be to turn it on for ever object, which is
probably not correct).
2005-07-21 22:30:33 +02:00
|
|
|
{
|
2018-12-05 23:38:08 +01:00
|
|
|
BLI_assert(em->ob->id.tag & LIB_TAG_COPIED_ON_WRITE);
|
|
|
|
|
2013-05-18 12:24:34 +02:00
|
|
|
BKE_object_free_derived_caches(obedit);
|
|
|
|
BKE_object_sculpt_modifiers_changed(obedit);
|
- shuffled editmesh derived function name/function
- added ModifierTypeInfo.freeData function
- added modifier_{new,free] utility function
- added ccgSubSurf_getUseAgeCounts to query info
- removed subsurf modifier faking (ME_SUBSURF flag is no
longer valid). subsurf modifier gets converted on file load
although there is obscure linked mesh situation where this
can go wrong, will fix shortly. this also means that some
places in the code that test/copy subsurf settings are broken
for the time being.
- shuffled modifier calculation to be simpler. note that
all modifiers are currently disabled in editmode (including
subsurf). don't worry, will return shortly.
- bug fix, build modifier didn't randomize meshes with only verts
- cleaned up subsurf_ccg and adapted for future editmode modifier
work
- added editmesh.derived{Cage,Final}, not used yet
- added SubsurfModifierData.{mCache,emCache}, will be used to cache
subsurf instead of caching in derivedmesh itself
- removed old subsurf buttons
- added do_modifiers_buttons to handle modifier events
- removed count_object counting of modifier (subsurfed) objects...
this would be nice to add back at some point but requires care.
probably requires rewrite of counting system.
New feature: Incremental Subsurf in Object Mode
The previous release introduce incremental subsurf calculation during
editmode but it was not turned on during object mode. In general it
does not make sense to have it always enabled during object mode because
it requires caching a fair amount of information about the mesh which
is a waste of memory unless the mesh is often recalculated.
However, for mesh's that have subsurfed armatures for example, or that
have other modifiers so that the mesh is essentially changing on every
frame, it makes a lot of sense to keep the subsurf'd object around and
that is what the new incremental subsurf modifier toggle is for. The
intent is that the user will enable this option for (a) a mesh that is
currently under active editing or (b) a mesh that is heavily updated
in the scene, such as a character.
I will try to write more about this feature for release, because it
has advantages and disadvantages that are not immediately obvious (the
first user reaction will be to turn it on for ever object, which is
probably not correct).
2005-07-21 22:30:33 +02:00
|
|
|
|
2014-02-28 10:02:30 +01:00
|
|
|
BKE_editmesh_free_derivedmesh(em);
|
- shuffled editmesh derived function name/function
- added ModifierTypeInfo.freeData function
- added modifier_{new,free] utility function
- added ccgSubSurf_getUseAgeCounts to query info
- removed subsurf modifier faking (ME_SUBSURF flag is no
longer valid). subsurf modifier gets converted on file load
although there is obscure linked mesh situation where this
can go wrong, will fix shortly. this also means that some
places in the code that test/copy subsurf settings are broken
for the time being.
- shuffled modifier calculation to be simpler. note that
all modifiers are currently disabled in editmode (including
subsurf). don't worry, will return shortly.
- bug fix, build modifier didn't randomize meshes with only verts
- cleaned up subsurf_ccg and adapted for future editmode modifier
work
- added editmesh.derived{Cage,Final}, not used yet
- added SubsurfModifierData.{mCache,emCache}, will be used to cache
subsurf instead of caching in derivedmesh itself
- removed old subsurf buttons
- added do_modifiers_buttons to handle modifier events
- removed count_object counting of modifier (subsurfed) objects...
this would be nice to add back at some point but requires care.
probably requires rewrite of counting system.
New feature: Incremental Subsurf in Object Mode
The previous release introduce incremental subsurf calculation during
editmode but it was not turned on during object mode. In general it
does not make sense to have it always enabled during object mode because
it requires caching a fair amount of information about the mesh which
is a waste of memory unless the mesh is often recalculated.
However, for mesh's that have subsurfed armatures for example, or that
have other modifiers so that the mesh is essentially changing on every
frame, it makes a lot of sense to keep the subsurf'd object around and
that is what the new incremental subsurf modifier toggle is for. The
intent is that the user will enable this option for (a) a mesh that is
currently under active editing or (b) a mesh that is heavily updated
in the scene, such as a character.
I will try to write more about this feature for release, because it
has advantages and disadvantages that are not immediately obvious (the
first user reaction will be to turn it on for ever object, which is
probably not correct).
2005-07-21 22:30:33 +02:00
|
|
|
|
2018-10-09 06:04:51 +02:00
|
|
|
Mesh *me_cage;
|
|
|
|
Mesh *me_final;
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
editbmesh_calc_modifiers(
|
2018-04-06 12:07:27 +02:00
|
|
|
depsgraph, scene, obedit, em, dataMask,
|
2018-10-09 06:04:51 +02:00
|
|
|
&me_cage, &me_final);
|
|
|
|
|
2018-10-09 07:52:46 +02:00
|
|
|
em->mesh_eval_final = me_final;
|
|
|
|
em->mesh_eval_cage = me_cage;
|
2015-07-02 08:20:22 +02:00
|
|
|
|
2018-10-15 08:14:05 +02:00
|
|
|
BKE_object_boundbox_calc_from_mesh(obedit, em->mesh_eval_final);
|
Added custom vertex/edge/face data for meshes:
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
2006-11-20 05:28:02 +01:00
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
em->lastDataMask = *dataMask;
|
2013-05-30 19:36:43 +02:00
|
|
|
|
2018-11-20 11:37:51 +01:00
|
|
|
mesh_runtime_check_normals_valid(em->mesh_eval_final);
|
2005-07-19 04:36:21 +02:00
|
|
|
}
|
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
static void object_get_datamask(const Depsgraph *depsgraph, Object *ob, CustomData_MeshMasks *r_mask, bool *r_need_mapping)
|
2013-05-01 16:34:12 +02:00
|
|
|
{
|
2018-04-06 12:07:27 +02:00
|
|
|
ViewLayer *view_layer = DEG_get_evaluated_view_layer(depsgraph);
|
2018-06-19 18:23:26 +02:00
|
|
|
Object *actob = view_layer->basact ? DEG_get_original_object(view_layer->basact->object) : NULL;
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
|
|
|
|
DEG_get_customdata_mask_for_object(depsgraph, ob, r_mask);
|
2015-06-04 17:39:43 +02:00
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
if (r_need_mapping) {
|
|
|
|
*r_need_mapping = false;
|
|
|
|
}
|
2013-05-01 16:34:12 +02:00
|
|
|
|
2018-06-19 18:29:40 +02:00
|
|
|
if (DEG_get_original_object(ob) == actob) {
|
2018-12-01 10:12:18 +01:00
|
|
|
bool editing = BKE_paint_select_face_test(actob);
|
2015-06-04 17:39:43 +02:00
|
|
|
|
|
|
|
/* weight paint and face select need original indices because of selection buffer drawing */
|
2015-07-02 08:20:22 +02:00
|
|
|
if (r_need_mapping) {
|
2018-04-05 18:20:27 +02:00
|
|
|
*r_need_mapping = (editing || (ob->mode & (OB_MODE_WEIGHT_PAINT | OB_MODE_VERTEX_PAINT)));
|
2015-07-02 08:20:22 +02:00
|
|
|
}
|
2015-06-04 17:39:43 +02:00
|
|
|
|
2013-05-01 16:34:12 +02:00
|
|
|
/* check if we need tfaces & mcols due to face select or texture paint */
|
2018-04-05 18:20:27 +02:00
|
|
|
if ((ob->mode & OB_MODE_TEXTURE_PAINT) || editing) {
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
r_mask->lmask |= CD_MASK_MLOOPUV | CD_MASK_MLOOPCOL;
|
|
|
|
r_mask->fmask |= CD_MASK_MTFACE;
|
2013-05-01 16:34:12 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/* check if we need mcols due to vertex paint or weightpaint */
|
2018-04-05 18:20:27 +02:00
|
|
|
if (ob->mode & OB_MODE_VERTEX_PAINT) {
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
r_mask->lmask |= CD_MASK_MLOOPCOL;
|
2013-05-01 16:34:12 +02:00
|
|
|
}
|
|
|
|
|
2018-10-09 13:21:23 +02:00
|
|
|
if (ob->mode & OB_MODE_WEIGHT_PAINT) {
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
r_mask->vmask |= CD_MASK_MDEFORMVERT;
|
2018-10-09 13:21:23 +02:00
|
|
|
}
|
|
|
|
|
2019-01-11 20:09:11 +01:00
|
|
|
if (ob->mode & OB_MODE_EDIT)
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
r_mask->vmask |= CD_MASK_MVERT_SKIN;
|
2013-05-01 16:34:12 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
void makeDerivedMesh(
|
2018-04-06 12:07:27 +02:00
|
|
|
struct Depsgraph *depsgraph, Scene *scene, Object *ob, BMEditMesh *em,
|
2019-03-27 19:07:16 +01:00
|
|
|
const CustomData_MeshMasks *dataMask)
|
2005-07-19 04:36:21 +02:00
|
|
|
{
|
2015-07-02 08:20:22 +02:00
|
|
|
bool need_mapping;
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_MeshMasks cddata_masks = *dataMask;
|
|
|
|
object_get_datamask(depsgraph, ob, &cddata_masks, &need_mapping);
|
2013-05-01 16:34:12 +02:00
|
|
|
|
2008-12-31 18:11:42 +01:00
|
|
|
if (em) {
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
editbmesh_build_data(depsgraph, scene, ob, em, &cddata_masks);
|
2012-03-24 07:18:31 +01:00
|
|
|
}
|
|
|
|
else {
|
2019-03-27 19:07:16 +01:00
|
|
|
mesh_build_data(depsgraph, scene, ob, &cddata_masks, need_mapping);
|
- shuffled editmesh derived function name/function
- added ModifierTypeInfo.freeData function
- added modifier_{new,free] utility function
- added ccgSubSurf_getUseAgeCounts to query info
- removed subsurf modifier faking (ME_SUBSURF flag is no
longer valid). subsurf modifier gets converted on file load
although there is obscure linked mesh situation where this
can go wrong, will fix shortly. this also means that some
places in the code that test/copy subsurf settings are broken
for the time being.
- shuffled modifier calculation to be simpler. note that
all modifiers are currently disabled in editmode (including
subsurf). don't worry, will return shortly.
- bug fix, build modifier didn't randomize meshes with only verts
- cleaned up subsurf_ccg and adapted for future editmode modifier
work
- added editmesh.derived{Cage,Final}, not used yet
- added SubsurfModifierData.{mCache,emCache}, will be used to cache
subsurf instead of caching in derivedmesh itself
- removed old subsurf buttons
- added do_modifiers_buttons to handle modifier events
- removed count_object counting of modifier (subsurfed) objects...
this would be nice to add back at some point but requires care.
probably requires rewrite of counting system.
New feature: Incremental Subsurf in Object Mode
The previous release introduce incremental subsurf calculation during
editmode but it was not turned on during object mode. In general it
does not make sense to have it always enabled during object mode because
it requires caching a fair amount of information about the mesh which
is a waste of memory unless the mesh is often recalculated.
However, for mesh's that have subsurfed armatures for example, or that
have other modifiers so that the mesh is essentially changing on every
frame, it makes a lot of sense to keep the subsurf'd object around and
that is what the new incremental subsurf modifier toggle is for. The
intent is that the user will enable this option for (a) a mesh that is
currently under active editing or (b) a mesh that is heavily updated
in the scene, such as a character.
I will try to write more about this feature for release, because it
has advantages and disadvantages that are not immediately obvious (the
first user reaction will be to turn it on for ever object, which is
probably not correct).
2005-07-21 22:30:33 +02:00
|
|
|
}
|
2005-07-19 04:36:21 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/***/
|
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
Mesh *mesh_get_eval_final(struct Depsgraph *depsgraph, Scene *scene, Object *ob, const CustomData_MeshMasks *dataMask)
|
2018-06-01 12:34:50 +02:00
|
|
|
{
|
2018-12-03 13:22:19 +01:00
|
|
|
/* This function isn't thread-safe and can't be used during evaluation. */
|
|
|
|
BLI_assert(DEG_debug_is_evaluating(depsgraph) == false);
|
|
|
|
|
|
|
|
/* Evaluated meshes aren't supposed to be created on original instances. If you do,
|
|
|
|
* they aren't cleaned up properly on mode switch, causing crashes, e.g T58150. */
|
|
|
|
BLI_assert(ob->id.tag & LIB_TAG_COPIED_ON_WRITE);
|
|
|
|
|
2018-06-01 12:34:50 +02:00
|
|
|
/* if there's no evaluated mesh or the last data mask used doesn't include
|
|
|
|
* the data we need, rebuild the derived mesh
|
|
|
|
*/
|
|
|
|
bool need_mapping;
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_MeshMasks cddata_masks = *dataMask;
|
|
|
|
object_get_datamask(depsgraph, ob, &cddata_masks, &need_mapping);
|
2018-06-01 12:34:50 +02:00
|
|
|
|
2018-06-08 15:42:27 +02:00
|
|
|
if (!ob->runtime.mesh_eval ||
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
!CustomData_MeshMasks_are_matching(&(ob->runtime.last_data_mask), &cddata_masks) ||
|
2018-12-03 16:09:45 +01:00
|
|
|
(need_mapping && !ob->runtime.last_need_mapping))
|
2018-06-01 12:34:50 +02:00
|
|
|
{
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_MeshMasks_update(&cddata_masks, &ob->runtime.last_data_mask);
|
|
|
|
mesh_build_data(depsgraph, scene, ob, &cddata_masks,
|
2019-03-27 19:07:16 +01:00
|
|
|
need_mapping || ob->runtime.last_need_mapping);
|
2018-06-01 12:34:50 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
if (ob->runtime.mesh_eval) { BLI_assert(!(ob->runtime.mesh_eval->runtime.cd_dirty_vert & CD_MASK_NORMAL)); }
|
|
|
|
return ob->runtime.mesh_eval;
|
|
|
|
}
|
2005-07-19 02:21:01 +02:00
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
Mesh *mesh_get_eval_deform(struct Depsgraph *depsgraph, Scene *scene, Object *ob, const CustomData_MeshMasks *dataMask)
|
2018-06-01 12:34:50 +02:00
|
|
|
{
|
2018-12-03 13:22:19 +01:00
|
|
|
/* This function isn't thread-safe and can't be used during evaluation. */
|
|
|
|
BLI_assert(DEG_debug_is_evaluating(depsgraph) == false);
|
|
|
|
|
|
|
|
/* Evaluated meshes aren't supposed to be created on original instances. If you do,
|
|
|
|
* they aren't cleaned up properly on mode switch, causing crashes, e.g T58150. */
|
|
|
|
BLI_assert(ob->id.tag & LIB_TAG_COPIED_ON_WRITE);
|
|
|
|
|
2018-06-01 12:34:50 +02:00
|
|
|
/* if there's no derived mesh or the last data mask used doesn't include
|
|
|
|
* the data we need, rebuild the derived mesh
|
|
|
|
*/
|
|
|
|
bool need_mapping;
|
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_MeshMasks cddata_masks = *dataMask;
|
|
|
|
object_get_datamask(depsgraph, ob, &cddata_masks, &need_mapping);
|
2018-06-01 12:34:50 +02:00
|
|
|
|
|
|
|
if (!ob->runtime.mesh_deform_eval ||
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
!CustomData_MeshMasks_are_matching(&(ob->runtime.last_data_mask), &cddata_masks) ||
|
2018-12-03 16:09:45 +01:00
|
|
|
(need_mapping && !ob->runtime.last_need_mapping))
|
2018-06-01 12:34:50 +02:00
|
|
|
{
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_MeshMasks_update(&cddata_masks, &ob->runtime.last_data_mask);
|
|
|
|
mesh_build_data(depsgraph, scene, ob, &cddata_masks,
|
2019-03-27 19:07:16 +01:00
|
|
|
need_mapping || ob->runtime.last_need_mapping);
|
2018-06-01 12:34:50 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
return ob->runtime.mesh_deform_eval;
|
|
|
|
}
|
|
|
|
|
2005-03-28 23:49:49 +02:00
|
|
|
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
Mesh *mesh_create_eval_final_render(Depsgraph *depsgraph, Scene *scene, Object *ob, const CustomData_MeshMasks *dataMask)
|
2018-10-12 09:05:26 +02:00
|
|
|
{
|
|
|
|
Mesh *final;
|
|
|
|
|
|
|
|
mesh_calc_modifiers(
|
2019-03-27 19:07:16 +01:00
|
|
|
depsgraph, scene, ob, 1, false, dataMask, -1, false,
|
2018-10-12 09:05:26 +02:00
|
|
|
NULL, &final);
|
|
|
|
|
|
|
|
return final;
|
|
|
|
}
|
2008-07-28 13:01:34 +02:00
|
|
|
|
2018-10-10 06:18:34 +02:00
|
|
|
Mesh *mesh_create_eval_final_index_render(
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
Depsgraph *depsgraph, Scene *scene,
|
|
|
|
Object *ob, const CustomData_MeshMasks *dataMask, int index)
|
2018-07-05 11:08:05 +02:00
|
|
|
{
|
|
|
|
Mesh *final;
|
|
|
|
|
|
|
|
mesh_calc_modifiers(
|
2019-03-27 19:07:16 +01:00
|
|
|
depsgraph, scene, ob, 1, false, dataMask, index, false,
|
2018-07-05 11:08:05 +02:00
|
|
|
NULL, &final);
|
|
|
|
|
|
|
|
return final;
|
|
|
|
}
|
2005-07-16 23:03:28 +02:00
|
|
|
|
2018-07-05 15:41:58 +02:00
|
|
|
Mesh *mesh_create_eval_final_view(
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
Depsgraph *depsgraph, Scene *scene,
|
|
|
|
Object *ob, const CustomData_MeshMasks *dataMask)
|
2018-07-05 15:41:58 +02:00
|
|
|
{
|
|
|
|
Mesh *final;
|
|
|
|
|
|
|
|
/* XXX hack
|
|
|
|
* psys modifier updates particle state when called during dupli-list generation,
|
|
|
|
* which can lead to wrong transforms. This disables particle system modifier execution.
|
|
|
|
*/
|
|
|
|
ob->transflag |= OB_NO_PSYS_UPDATE;
|
|
|
|
|
|
|
|
mesh_calc_modifiers(
|
2019-03-27 19:07:16 +01:00
|
|
|
depsgraph, scene, ob, 1, false, dataMask, -1, false,
|
2018-07-05 15:41:58 +02:00
|
|
|
NULL, &final);
|
|
|
|
|
|
|
|
ob->transflag &= ~OB_NO_PSYS_UPDATE;
|
|
|
|
|
|
|
|
return final;
|
|
|
|
}
|
2006-09-17 07:15:56 +02:00
|
|
|
|
2018-12-04 23:47:13 +01:00
|
|
|
Mesh *mesh_create_eval_no_deform(
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
Depsgraph *depsgraph, Scene *scene, Object *ob,
|
2019-03-27 19:07:16 +01:00
|
|
|
const CustomData_MeshMasks *dataMask)
|
2005-07-17 03:18:59 +02:00
|
|
|
{
|
2018-12-04 23:47:13 +01:00
|
|
|
Mesh *final;
|
2018-06-17 17:05:51 +02:00
|
|
|
|
2018-12-04 23:47:13 +01:00
|
|
|
mesh_calc_modifiers(
|
2019-03-27 19:07:16 +01:00
|
|
|
depsgraph, scene, ob, 0, false, dataMask, -1, false,
|
2018-12-04 23:47:13 +01:00
|
|
|
NULL, &final);
|
|
|
|
|
|
|
|
return final;
|
|
|
|
}
|
|
|
|
|
|
|
|
Mesh *mesh_create_eval_no_deform_render(
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
Depsgraph *depsgraph, Scene *scene, Object *ob,
|
2019-03-27 19:07:16 +01:00
|
|
|
const CustomData_MeshMasks *dataMask)
|
2018-12-04 23:47:13 +01:00
|
|
|
{
|
|
|
|
Mesh *final;
|
|
|
|
|
|
|
|
mesh_calc_modifiers(
|
2019-03-27 19:07:16 +01:00
|
|
|
depsgraph, scene, ob, 0, false, dataMask, -1, false,
|
2015-07-02 08:20:22 +02:00
|
|
|
NULL, &final);
|
2005-04-04 14:22:33 +02:00
|
|
|
|
2005-07-19 02:21:01 +02:00
|
|
|
return final;
|
|
|
|
}
|
2005-04-04 05:38:21 +02:00
|
|
|
|
2005-07-19 02:21:01 +02:00
|
|
|
/***/
|
|
|
|
|
2018-10-09 07:52:46 +02:00
|
|
|
Mesh *editbmesh_get_eval_cage_and_final(
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
Depsgraph *depsgraph, Scene *scene, Object *obedit, BMEditMesh *em,
|
|
|
|
const CustomData_MeshMasks *dataMask,
|
2015-07-02 08:20:22 +02:00
|
|
|
/* return args */
|
2018-10-09 07:52:46 +02:00
|
|
|
Mesh **r_final)
|
2005-07-19 02:21:01 +02:00
|
|
|
{
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_MeshMasks cddata_masks = *dataMask;
|
|
|
|
|
2006-12-05 18:42:03 +01:00
|
|
|
/* if there's no derived mesh or the last data mask used doesn't include
|
|
|
|
* the data we need, rebuild the derived mesh
|
|
|
|
*/
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
object_get_datamask(depsgraph, obedit, &cddata_masks, NULL);
|
2013-05-01 16:34:12 +02:00
|
|
|
|
2018-10-09 07:52:46 +02:00
|
|
|
if (!em->mesh_eval_cage ||
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
!CustomData_MeshMasks_are_matching(&(em->lastDataMask), &cddata_masks))
|
2012-04-28 08:31:57 +02:00
|
|
|
{
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
editbmesh_build_data(depsgraph, scene, obedit, em, &cddata_masks);
|
2012-04-28 08:31:57 +02:00
|
|
|
}
|
2005-03-27 22:34:18 +02:00
|
|
|
|
2018-10-09 07:52:46 +02:00
|
|
|
*r_final = em->mesh_eval_final;
|
|
|
|
if (em->mesh_eval_final) { BLI_assert(!(em->mesh_eval_final->runtime.cd_dirty_vert & DM_DIRTY_NORMALS)); }
|
|
|
|
return em->mesh_eval_cage;
|
2005-03-27 22:34:18 +02:00
|
|
|
}
|
2005-03-28 07:58:43 +02:00
|
|
|
|
2018-10-09 07:52:46 +02:00
|
|
|
Mesh *editbmesh_get_eval_cage(
|
2018-04-06 12:07:27 +02:00
|
|
|
struct Depsgraph *depsgraph, Scene *scene, Object *obedit, BMEditMesh *em,
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
const CustomData_MeshMasks *dataMask)
|
2005-03-28 07:58:43 +02:00
|
|
|
{
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
CustomData_MeshMasks cddata_masks = *dataMask;
|
|
|
|
|
2006-12-05 18:42:03 +01:00
|
|
|
/* if there's no derived mesh or the last data mask used doesn't include
|
|
|
|
* the data we need, rebuild the derived mesh
|
|
|
|
*/
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
object_get_datamask(depsgraph, obedit, &cddata_masks, NULL);
|
2013-05-01 16:34:12 +02:00
|
|
|
|
2018-10-09 07:52:46 +02:00
|
|
|
if (!em->mesh_eval_cage ||
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
!CustomData_MeshMasks_are_matching(&(em->lastDataMask), &cddata_masks))
|
2012-04-28 08:31:57 +02:00
|
|
|
{
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
editbmesh_build_data(depsgraph, scene, obedit, em, &cddata_masks);
|
2012-04-28 08:31:57 +02:00
|
|
|
}
|
2005-03-28 07:58:43 +02:00
|
|
|
|
2018-10-09 07:52:46 +02:00
|
|
|
return em->mesh_eval_cage;
|
2005-03-28 07:58:43 +02:00
|
|
|
}
|
2005-08-06 22:44:59 +02:00
|
|
|
|
2018-12-05 23:38:08 +01:00
|
|
|
Mesh *editbmesh_get_eval_cage_from_orig(
|
|
|
|
struct Depsgraph *depsgraph, Scene *scene, Object *obedit, BMEditMesh *UNUSED(em),
|
Refactor CDData masks, to have one mask per mesh elem type.
We already have different storages for cddata of verts, edges etc.,
'simply' do the same for the mask flags we use all around Blender code
to request some data, or limit some operation to some layers, etc.
Reason we need this is that some cddata types (like Normals) are
actually shared between verts/polys/loops, and we don’t want to generate
clnors everytime we request vnors!
As a side note, this also does final fix to T59338, which was the
trigger for this patch (need to request computed loop normals for
another mesh than evaluated one).
Reviewers: brecht, campbellbarton, sergey
Differential Revision: https://developer.blender.org/D4407
2019-03-07 11:13:40 +01:00
|
|
|
const CustomData_MeshMasks *dataMask)
|
2018-12-05 23:38:08 +01:00
|
|
|
{
|
|
|
|
BLI_assert((obedit->id.tag & LIB_TAG_COPIED_ON_WRITE) == 0);
|
|
|
|
Scene *scene_eval = (Scene *)DEG_get_evaluated_id(depsgraph, &scene->id);
|
|
|
|
Object *obedit_eval = (Object *)DEG_get_evaluated_id(depsgraph, &obedit->id);
|
|
|
|
BMEditMesh *em_eval = BKE_editmesh_from_object(obedit_eval);
|
|
|
|
return editbmesh_get_eval_cage(depsgraph, scene_eval, obedit_eval, em_eval, dataMask);
|
|
|
|
}
|
|
|
|
|
2014-08-01 15:42:17 +02:00
|
|
|
/***/
|
|
|
|
|
2014-09-04 09:53:10 +02:00
|
|
|
/* same as above but for vert coords */
|
|
|
|
typedef struct {
|
|
|
|
float (*vertexcos)[3];
|
|
|
|
BLI_bitmap *vertex_visit;
|
|
|
|
} MappedUserData;
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
static void make_vertexcos__mapFunc(
|
|
|
|
void *userData, int index, const float co[3],
|
|
|
|
const float UNUSED(no_f[3]), const short UNUSED(no_s[3]))
|
2014-09-04 09:53:10 +02:00
|
|
|
{
|
|
|
|
MappedUserData *mappedData = (MappedUserData *)userData;
|
|
|
|
|
|
|
|
if (BLI_BITMAP_TEST(mappedData->vertex_visit, index) == 0) {
|
|
|
|
/* we need coord from prototype vertex, not from copies,
|
|
|
|
* assume they stored in the beginning of vertex array stored in DM
|
|
|
|
* (mirror modifier for eg does this) */
|
|
|
|
copy_v3_v3(mappedData->vertexcos[index], co);
|
|
|
|
BLI_BITMAP_ENABLE(mappedData->vertex_visit, index);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-10-09 07:52:46 +02:00
|
|
|
void mesh_get_mapped_verts_coords(Mesh *me_eval, float (*r_cos)[3], const int totcos)
|
2014-09-04 09:53:10 +02:00
|
|
|
{
|
2018-10-09 07:52:46 +02:00
|
|
|
if (me_eval->runtime.deformed_only == false) {
|
2014-09-04 09:53:10 +02:00
|
|
|
MappedUserData userData;
|
2014-09-14 10:50:59 +02:00
|
|
|
memset(r_cos, 0, sizeof(*r_cos) * totcos);
|
2014-09-04 09:53:10 +02:00
|
|
|
userData.vertexcos = r_cos;
|
|
|
|
userData.vertex_visit = BLI_BITMAP_NEW(totcos, "vertexcos flags");
|
2018-10-09 07:52:46 +02:00
|
|
|
BKE_mesh_foreach_mapped_vert(me_eval, make_vertexcos__mapFunc, &userData, MESH_FOREACH_NOP);
|
2014-09-04 09:53:10 +02:00
|
|
|
MEM_freeN(userData.vertex_visit);
|
|
|
|
}
|
|
|
|
else {
|
2018-10-09 07:52:46 +02:00
|
|
|
MVert *mv = me_eval->mvert;
|
|
|
|
for (int i = 0; i < totcos; i++, mv++) {
|
|
|
|
copy_v3_v3(r_cos[i], mv->co);
|
2014-09-04 09:53:10 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2016-04-26 10:43:02 +02:00
|
|
|
void DM_calc_loop_tangents(
|
|
|
|
DerivedMesh *dm, bool calc_active_tangent,
|
2017-05-24 14:33:21 +02:00
|
|
|
const char (*tangent_names)[MAX_NAME], int tangent_names_len)
|
2016-04-26 10:43:02 +02:00
|
|
|
{
|
2017-05-24 14:33:21 +02:00
|
|
|
BKE_mesh_calc_loop_tangent_ex(
|
|
|
|
dm->getVertArray(dm),
|
|
|
|
dm->getPolyArray(dm), dm->getNumPolys(dm),
|
|
|
|
dm->getLoopArray(dm),
|
|
|
|
dm->getLoopTriArray(dm), dm->getNumLoopTri(dm),
|
|
|
|
&dm->loopData,
|
|
|
|
calc_active_tangent,
|
|
|
|
tangent_names, tangent_names_len,
|
|
|
|
CustomData_get_layer(&dm->polyData, CD_NORMAL),
|
|
|
|
dm->getLoopDataArray(dm, CD_NORMAL),
|
|
|
|
dm->getVertDataArray(dm, CD_ORCO), /* may be NULL */
|
|
|
|
/* result */
|
|
|
|
&dm->loopData, dm->getNumLoops(dm),
|
|
|
|
&dm->tangent_mask);
|
2010-03-05 17:47:52 +01:00
|
|
|
}
|
2011-10-09 23:11:51 +02:00
|
|
|
|
2018-05-11 09:53:52 +02:00
|
|
|
static void mesh_init_origspace(Mesh *mesh)
|
|
|
|
{
|
|
|
|
const float default_osf[4][2] = {{0, 0}, {1, 0}, {1, 1}, {0, 1}};
|
|
|
|
|
|
|
|
OrigSpaceLoop *lof_array = CustomData_get_layer(&mesh->ldata, CD_ORIGSPACE_MLOOP);
|
|
|
|
const int numpoly = mesh->totpoly;
|
|
|
|
// const int numloop = mesh->totloop;
|
|
|
|
MVert *mv = mesh->mvert;
|
|
|
|
MLoop *ml = mesh->mloop;
|
|
|
|
MPoly *mp = mesh->mpoly;
|
|
|
|
int i, j, k;
|
|
|
|
|
|
|
|
float (*vcos_2d)[2] = NULL;
|
|
|
|
BLI_array_staticdeclare(vcos_2d, 64);
|
|
|
|
|
|
|
|
for (i = 0; i < numpoly; i++, mp++) {
|
|
|
|
OrigSpaceLoop *lof = lof_array + mp->loopstart;
|
|
|
|
|
|
|
|
if (mp->totloop == 3 || mp->totloop == 4) {
|
|
|
|
for (j = 0; j < mp->totloop; j++, lof++) {
|
|
|
|
copy_v2_v2(lof->uv, default_osf[j]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
MLoop *l = &ml[mp->loopstart];
|
|
|
|
float p_nor[3], co[3];
|
|
|
|
float mat[3][3];
|
|
|
|
|
|
|
|
float min[2] = {FLT_MAX, FLT_MAX}, max[2] = {-FLT_MAX, -FLT_MAX};
|
|
|
|
float translate[2], scale[2];
|
|
|
|
|
|
|
|
BKE_mesh_calc_poly_normal(mp, l, mv, p_nor);
|
|
|
|
axis_dominant_v3_to_m3(mat, p_nor);
|
|
|
|
|
|
|
|
BLI_array_clear(vcos_2d);
|
|
|
|
BLI_array_reserve(vcos_2d, mp->totloop);
|
|
|
|
for (j = 0; j < mp->totloop; j++, l++) {
|
|
|
|
mul_v3_m3v3(co, mat, mv[l->v].co);
|
|
|
|
copy_v2_v2(vcos_2d[j], co);
|
|
|
|
|
|
|
|
for (k = 0; k < 2; k++) {
|
|
|
|
if (co[k] > max[k])
|
|
|
|
max[k] = co[k];
|
|
|
|
else if (co[k] < min[k])
|
|
|
|
min[k] = co[k];
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Brings min to (0, 0). */
|
|
|
|
negate_v2_v2(translate, min);
|
|
|
|
|
|
|
|
/* Scale will bring max to (1, 1). */
|
|
|
|
sub_v2_v2v2(scale, max, min);
|
|
|
|
if (scale[0] == 0.0f)
|
|
|
|
scale[0] = 1e-9f;
|
|
|
|
if (scale[1] == 0.0f)
|
|
|
|
scale[1] = 1e-9f;
|
|
|
|
invert_v2(scale);
|
|
|
|
|
2018-09-24 17:27:41 +02:00
|
|
|
/* Finally, transform all vcos_2d into ((0, 0), (1, 1)) square and assign them as origspace. */
|
2018-05-11 09:53:52 +02:00
|
|
|
for (j = 0; j < mp->totloop; j++, lof++) {
|
|
|
|
add_v2_v2v2(lof->uv, vcos_2d[j], translate);
|
|
|
|
mul_v2_v2(lof->uv, scale);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-05-22 17:53:18 +02:00
|
|
|
BKE_mesh_tessface_clear(mesh);
|
2018-05-11 09:53:52 +02:00
|
|
|
BLI_array_free(vcos_2d);
|
|
|
|
}
|
2012-02-05 12:30:26 +01:00
|
|
|
|
|
|
|
|
2012-01-19 01:18:25 +01:00
|
|
|
/* derivedmesh info printing function,
|
|
|
|
* to help track down differences DM output */
|
|
|
|
|
|
|
|
#ifndef NDEBUG
|
|
|
|
#include "BLI_dynstr.h"
|
|
|
|
|
2015-07-02 08:20:22 +02:00
|
|
|
static void dm_debug_info_layers(
|
|
|
|
DynStr *dynstr, DerivedMesh *dm, CustomData *cd,
|
|
|
|
void *(*getElemDataArray)(DerivedMesh *, int))
|
2012-01-19 01:18:25 +01:00
|
|
|
{
|
|
|
|
int type;
|
|
|
|
|
|
|
|
for (type = 0; type < CD_NUMTYPES; type++) {
|
2013-05-30 15:13:43 +02:00
|
|
|
if (CustomData_has_layer(cd, type)) {
|
2018-09-24 17:27:41 +02:00
|
|
|
/* note: doesn't account for multiple layers */
|
2012-01-19 01:18:25 +01:00
|
|
|
const char *name = CustomData_layertype_name(type);
|
|
|
|
const int size = CustomData_sizeof(type);
|
2013-05-30 15:13:43 +02:00
|
|
|
const void *pt = getElemDataArray(dm, type);
|
|
|
|
const int pt_size = pt ? (int)(MEM_allocN_len(pt) / size) : 0;
|
2012-01-19 01:18:25 +01:00
|
|
|
const char *structname;
|
|
|
|
int structnum;
|
|
|
|
CustomData_file_write_info(type, &structname, &structnum);
|
|
|
|
BLI_dynstr_appendf(dynstr,
|
|
|
|
" dict(name='%s', struct='%s', type=%d, ptr='%p', elem=%d, length=%d),\n",
|
2015-01-13 19:10:18 +01:00
|
|
|
name, structname, type, (const void *)pt, size, pt_size);
|
2012-01-19 01:18:25 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
char *DM_debug_info(DerivedMesh *dm)
|
|
|
|
{
|
2012-05-12 21:18:02 +02:00
|
|
|
DynStr *dynstr = BLI_dynstr_new();
|
2012-01-19 01:18:25 +01:00
|
|
|
char *ret;
|
|
|
|
const char *tstr;
|
|
|
|
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, "{\n");
|
2012-01-19 01:18:25 +01:00
|
|
|
BLI_dynstr_appendf(dynstr, " 'ptr': '%p',\n", (void *)dm);
|
|
|
|
switch (dm->type) {
|
|
|
|
case DM_TYPE_CDDM: tstr = "DM_TYPE_CDDM"; break;
|
|
|
|
case DM_TYPE_CCGDM: tstr = "DM_TYPE_CCGDM"; break;
|
|
|
|
default: tstr = "UNKNOWN"; break;
|
|
|
|
}
|
|
|
|
BLI_dynstr_appendf(dynstr, " 'type': '%s',\n", tstr);
|
|
|
|
BLI_dynstr_appendf(dynstr, " 'numVertData': %d,\n", dm->numVertData);
|
|
|
|
BLI_dynstr_appendf(dynstr, " 'numEdgeData': %d,\n", dm->numEdgeData);
|
2012-01-19 12:31:31 +01:00
|
|
|
BLI_dynstr_appendf(dynstr, " 'numTessFaceData': %d,\n", dm->numTessFaceData);
|
|
|
|
BLI_dynstr_appendf(dynstr, " 'numPolyData': %d,\n", dm->numPolyData);
|
2012-01-19 01:18:25 +01:00
|
|
|
BLI_dynstr_appendf(dynstr, " 'deformedOnly': %d,\n", dm->deformedOnly);
|
|
|
|
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, " 'vertexLayers': (\n");
|
2013-05-30 15:13:43 +02:00
|
|
|
dm_debug_info_layers(dynstr, dm, &dm->vertData, dm->getVertDataArray);
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, " ),\n");
|
2012-02-05 12:30:26 +01:00
|
|
|
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, " 'edgeLayers': (\n");
|
2013-05-30 15:13:43 +02:00
|
|
|
dm_debug_info_layers(dynstr, dm, &dm->edgeData, dm->getEdgeDataArray);
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, " ),\n");
|
2012-01-19 01:18:25 +01:00
|
|
|
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, " 'loopLayers': (\n");
|
2013-05-30 15:13:43 +02:00
|
|
|
dm_debug_info_layers(dynstr, dm, &dm->loopData, dm->getLoopDataArray);
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, " ),\n");
|
2012-01-19 12:31:31 +01:00
|
|
|
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, " 'polyLayers': (\n");
|
2013-05-30 15:13:43 +02:00
|
|
|
dm_debug_info_layers(dynstr, dm, &dm->polyData, dm->getPolyDataArray);
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, " ),\n");
|
2013-05-30 15:13:43 +02:00
|
|
|
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, " 'tessFaceLayers': (\n");
|
2013-05-30 15:13:43 +02:00
|
|
|
dm_debug_info_layers(dynstr, dm, &dm->faceData, dm->getTessFaceDataArray);
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, " ),\n");
|
2012-01-19 01:18:25 +01:00
|
|
|
|
2019-03-24 06:09:46 +01:00
|
|
|
BLI_dynstr_append(dynstr, "}\n");
|
2012-01-19 01:18:25 +01:00
|
|
|
|
|
|
|
ret = BLI_dynstr_get_cstring(dynstr);
|
|
|
|
BLI_dynstr_free(dynstr);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
void DM_debug_print(DerivedMesh *dm)
|
|
|
|
{
|
|
|
|
char *str = DM_debug_info(dm);
|
2012-03-18 07:01:33 +01:00
|
|
|
puts(str);
|
2012-01-19 01:18:25 +01:00
|
|
|
fflush(stdout);
|
2012-01-19 05:00:27 +01:00
|
|
|
MEM_freeN(str);
|
2012-01-19 01:18:25 +01:00
|
|
|
}
|
|
|
|
|
2012-09-03 04:41:12 +02:00
|
|
|
void DM_debug_print_cdlayers(CustomData *data)
|
|
|
|
{
|
|
|
|
int i;
|
2015-01-13 19:10:18 +01:00
|
|
|
const CustomDataLayer *layer;
|
2012-09-03 04:41:12 +02:00
|
|
|
|
|
|
|
printf("{\n");
|
|
|
|
|
|
|
|
for (i = 0, layer = data->layers; i < data->totlayer; i++, layer++) {
|
|
|
|
|
|
|
|
const char *name = CustomData_layertype_name(layer->type);
|
|
|
|
const int size = CustomData_sizeof(layer->type);
|
|
|
|
const char *structname;
|
|
|
|
int structnum;
|
|
|
|
CustomData_file_write_info(layer->type, &structname, &structnum);
|
|
|
|
printf(" dict(name='%s', struct='%s', type=%d, ptr='%p', elem=%d, length=%d),\n",
|
2015-01-13 19:10:18 +01:00
|
|
|
name, structname, layer->type, (const void *)layer->data, size, (int)(MEM_allocN_len(layer->data) / size));
|
2012-09-03 04:41:12 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
printf("}\n");
|
|
|
|
}
|
|
|
|
|
2013-09-04 03:29:34 +02:00
|
|
|
bool DM_is_valid(DerivedMesh *dm)
|
|
|
|
{
|
|
|
|
const bool do_verbose = true;
|
|
|
|
const bool do_fixes = false;
|
|
|
|
|
|
|
|
bool is_valid = true;
|
2013-11-25 20:39:14 +01:00
|
|
|
bool changed = true;
|
2013-09-04 03:29:34 +02:00
|
|
|
|
|
|
|
is_valid &= BKE_mesh_validate_all_customdata(
|
2018-12-03 16:19:08 +01:00
|
|
|
dm->getVertDataLayout(dm), dm->getNumVerts(dm),
|
|
|
|
dm->getEdgeDataLayout(dm), dm->getNumEdges(dm),
|
|
|
|
dm->getLoopDataLayout(dm), dm->getNumLoops(dm),
|
|
|
|
dm->getPolyDataLayout(dm), dm->getNumPolys(dm),
|
2015-02-05 14:03:01 +01:00
|
|
|
false, /* setting mask here isn't useful, gives false positives */
|
2013-11-25 20:39:14 +01:00
|
|
|
do_verbose, do_fixes, &changed);
|
2013-09-04 03:29:34 +02:00
|
|
|
|
|
|
|
is_valid &= BKE_mesh_validate_arrays(
|
|
|
|
NULL,
|
|
|
|
dm->getVertArray(dm), dm->getNumVerts(dm),
|
|
|
|
dm->getEdgeArray(dm), dm->getNumEdges(dm),
|
|
|
|
dm->getTessFaceArray(dm), dm->getNumTessFaces(dm),
|
|
|
|
dm->getLoopArray(dm), dm->getNumLoops(dm),
|
|
|
|
dm->getPolyArray(dm), dm->getNumPolys(dm),
|
|
|
|
dm->getVertDataArray(dm, CD_MDEFORMVERT),
|
2013-11-25 20:39:14 +01:00
|
|
|
do_verbose, do_fixes, &changed);
|
2013-09-04 03:29:34 +02:00
|
|
|
|
2013-11-25 20:39:14 +01:00
|
|
|
BLI_assert(changed == false);
|
2013-09-04 03:29:34 +02:00
|
|
|
|
|
|
|
return is_valid;
|
|
|
|
}
|
|
|
|
|
2012-01-19 01:18:25 +01:00
|
|
|
#endif /* NDEBUG */
|