This repository has been archived on 2023-10-09. You can view files and clone it. You cannot open issues or pull requests or push a commit.
Files
blender-archive/source/blender/modifiers/intern/MOD_array.c
Sergey Sharybin 709041ed0b Threaded object update and EvaluationContext
Summary:
Made objects update happening from multiple threads. It is a task-based
scheduling system which uses current dependency graph for spawning new
tasks. This means threading happens on object level, but the system is
flexible enough for higher granularity.

Technical details:

- Uses task scheduler which was recently committed to trunk
  (that one which Brecht ported from Cycles).

- Added two utility functions to dependency graph:
  * DAG_threaded_update_begin, which is called to  initialize threaded
    objects update. It will also schedule root DAG node to the queue,
    hence starting evaluation process.

    Initialization will calculate how much parents are to be evaluation
    before current DAG node can be scheduled. This value is used by task
    threads for faster detecting which nodes might be scheduled.

  * DAG_threaded_update_handle_node_updated which is  called from task
    thread function when node was fully handled.

	This function decreases num_pending_parents of node children and
	schedules children with zero valency.

    As it might have become clear, task thread receives DAG nodes and
    decides which callback to call for it.

    Currently only BKE_object_handle_update is called for object nodes.

    In the future it'll call node->callback() from Ali's new DAG.

- This required adding some workarounds to the render pipeline.
  Mainly to stop using get_object_dm() from modifiers' apply callback.
  Such a call was only a workaround for dependency graph glitch when
  rendering scene with, say, boolean modifiers before displaying
  this scene.

  Such change moves workaround from one place to another, so overall
  hackentropy remains the same.

- Added paradigm of EvaluaitonContext. Currently it's more like just a
  more reliable replacement for G.is_rendering which fails in some
  circumstances.

  Future idea of this context is to also store all the local data needed
  for objects evaluation such as local time, Copy-on-Write data and so.

  There're two types of EvaluationContext:

  * Context used for viewport updated and owned by Main. In the future
    this context might be easily moved to Window or Screen to allo
    per-window/per-screen local time.

  * Context used by render engines to evaluate objects for render purposes.
    Render engine is an owner of this context.

  This context is passed to all object update routines.

Reviewers: brecht, campbellbarton

Reviewed By: brecht

CC: lukastoenne

Differential Revision: https://developer.blender.org/D94
2013-12-26 17:24:42 +06:00

616 lines
18 KiB
C

/*
* ***** BEGIN GPL LICENSE BLOCK *****
*
* 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
* of the License, or (at your option) any later version.
*
* 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,
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
*
* The Original Code is Copyright (C) 2005 by the Blender Foundation.
* All rights reserved.
*
* Contributor(s): Daniel Dunbar
* Ton Roosendaal,
* Ben Batt,
* Brecht Van Lommel,
* Campbell Barton
*
* ***** END GPL LICENSE BLOCK *****
*
*/
/** \file blender/modifiers/intern/MOD_array.c
* \ingroup modifiers
*/
/* Array modifier: duplicates the object multiple times along an axis */
#include "MEM_guardedalloc.h"
#include "BLI_math.h"
#include "BLI_utildefines.h"
#include "BLI_string.h"
#include "BLI_ghash.h"
#include "BLI_edgehash.h"
#include "DNA_curve_types.h"
#include "DNA_meshdata_types.h"
#include "DNA_object_types.h"
#include "DNA_scene_types.h"
#include "BKE_cdderivedmesh.h"
#include "BKE_displist.h"
#include "BKE_curve.h"
#include "BKE_mesh.h"
#include "BKE_modifier.h"
#include "BKE_object.h"
#include "MOD_util.h"
#include "bmesh.h"
#include "depsgraph_private.h"
#include <ctype.h>
#include <stdlib.h>
#include <string.h>
static void initData(ModifierData *md)
{
ArrayModifierData *amd = (ArrayModifierData *) md;
/* default to 2 duplicates distributed along the x-axis by an
* offset of 1 object-width
*/
amd->start_cap = amd->end_cap = amd->curve_ob = amd->offset_ob = NULL;
amd->count = 2;
zero_v3(amd->offset);
amd->scale[0] = 1;
amd->scale[1] = amd->scale[2] = 0;
amd->length = 0;
amd->merge_dist = 0.01;
amd->fit_type = MOD_ARR_FIXEDCOUNT;
amd->offset_type = MOD_ARR_OFF_RELATIVE;
amd->flags = 0;
}
static void copyData(ModifierData *md, ModifierData *target)
{
#if 0
ArrayModifierData *amd = (ArrayModifierData *) md;
ArrayModifierData *tamd = (ArrayModifierData *) target;
#endif
modifier_copyData_generic(md, target);
}
static void foreachObjectLink(
ModifierData *md, Object *ob,
void (*walk)(void *userData, Object *ob, Object **obpoin),
void *userData)
{
ArrayModifierData *amd = (ArrayModifierData *) md;
walk(userData, ob, &amd->start_cap);
walk(userData, ob, &amd->end_cap);
walk(userData, ob, &amd->curve_ob);
walk(userData, ob, &amd->offset_ob);
}
static void updateDepgraph(ModifierData *md, DagForest *forest,
struct Scene *UNUSED(scene), Object *UNUSED(ob), DagNode *obNode)
{
ArrayModifierData *amd = (ArrayModifierData *) md;
if (amd->start_cap) {
DagNode *curNode = dag_get_node(forest, amd->start_cap);
dag_add_relation(forest, curNode, obNode,
DAG_RL_DATA_DATA | DAG_RL_OB_DATA, "Array Modifier");
}
if (amd->end_cap) {
DagNode *curNode = dag_get_node(forest, amd->end_cap);
dag_add_relation(forest, curNode, obNode,
DAG_RL_DATA_DATA | DAG_RL_OB_DATA, "Array Modifier");
}
if (amd->curve_ob) {
DagNode *curNode = dag_get_node(forest, amd->curve_ob);
dag_add_relation(forest, curNode, obNode,
DAG_RL_DATA_DATA | DAG_RL_OB_DATA, "Array Modifier");
}
if (amd->offset_ob) {
DagNode *curNode = dag_get_node(forest, amd->offset_ob);
dag_add_relation(forest, curNode, obNode,
DAG_RL_DATA_DATA | DAG_RL_OB_DATA, "Array Modifier");
}
}
static float vertarray_size(MVert *mvert, int numVerts, int axis)
{
int i;
float min_co, max_co;
/* if there are no vertices, width is 0 */
if (numVerts == 0) return 0;
/* find the minimum and maximum coordinates on the desired axis */
min_co = max_co = mvert->co[axis];
mvert++;
for (i = 1; i < numVerts; ++i, ++mvert) {
if (mvert->co[axis] < min_co) min_co = mvert->co[axis];
if (mvert->co[axis] > max_co) max_co = mvert->co[axis];
}
return max_co - min_co;
}
static int *find_doubles_index_map(BMesh *bm, BMOperator *dupe_op,
const ArrayModifierData *amd,
int *index_map_length)
{
BMOperator find_op;
BMOIter oiter;
BMVert *v, *v2;
BMElem *ele;
int *index_map, i;
BMO_op_initf(bm, &find_op, (BMO_FLAG_DEFAULTS & ~BMO_FLAG_RESPECT_HIDE),
"find_doubles verts=%av dist=%f keep_verts=%s",
amd->merge_dist, dupe_op, "geom");
BMO_op_exec(bm, &find_op);
i = 0;
BMO_ITER (ele, &oiter, dupe_op->slots_in, "geom", BM_ALL) {
BM_elem_index_set(ele, i); /* set_dirty */
i++;
}
BMO_ITER (ele, &oiter, dupe_op->slots_out, "geom.out", BM_ALL) {
BM_elem_index_set(ele, i); /* set_dirty */
i++;
}
/* above loops over all, so set all to dirty, if this is somehow
* setting valid values, this line can be removed - campbell */
bm->elem_index_dirty |= BM_VERT | BM_EDGE | BM_FACE;
(*index_map_length) = i;
index_map = MEM_callocN(sizeof(int) * (*index_map_length), "index_map");
/*element type argument doesn't do anything here*/
BMO_ITER (v, &oiter, find_op.slots_out, "targetmap.out", 0) {
v2 = BMO_iter_map_value_ptr(&oiter);
index_map[BM_elem_index_get(v)] = BM_elem_index_get(v2) + 1;
}
BMO_op_finish(bm, &find_op);
return index_map;
}
/* Used for start/end cap.
*
* this function expects all existing vertices to be tagged,
* so we can know new verts are not tagged.
*
* All verts will be tagged on exit.
*/
static void bm_merge_dm_transform(BMesh *bm, DerivedMesh *dm, float mat[4][4],
const ArrayModifierData *amd,
BMOperator *dupe_op,
BMOpSlot dupe_op_slot_args[BMO_OP_MAX_SLOTS], const char *dupe_slot_name,
BMOperator *weld_op)
{
const int is_input = (dupe_op->slots_in == dupe_op_slot_args);
BMVert *v, *v2, *v3;
BMIter iter;
/* Add the DerivedMesh's elements to the BMesh. The pre-existing
* elements were already tagged, so the new elements can be
* identified by not having the BM_ELEM_TAG flag set. */
DM_to_bmesh_ex(dm, bm, false);
if (amd->flags & MOD_ARR_MERGE) {
/* if merging is enabled, find doubles */
BMOIter oiter;
BMOperator find_op;
BMOpSlot *slot_targetmap;
BMO_op_initf(bm, &find_op, (BMO_FLAG_DEFAULTS & ~BMO_FLAG_RESPECT_HIDE),
is_input ? /* ugh */
"find_doubles verts=%Hv dist=%f keep_verts=%s" :
"find_doubles verts=%Hv dist=%f keep_verts=%S",
BM_ELEM_TAG, amd->merge_dist,
dupe_op, dupe_slot_name);
/* append the dupe's geom to the findop input verts */
if (is_input) {
BMO_slot_buffer_append(&find_op, slots_in, "verts",
dupe_op, slots_in, dupe_slot_name);
}
else if (dupe_op->slots_out == dupe_op_slot_args) {
BMO_slot_buffer_append(&find_op, slots_in, "verts",
dupe_op, slots_out, dupe_slot_name);
}
else {
BLI_assert(0);
}
/* transform and tag verts */
BM_ITER_MESH (v, &iter, bm, BM_VERTS_OF_MESH) {
if (!BM_elem_flag_test(v, BM_ELEM_TAG)) {
mul_m4_v3(mat, v->co);
BM_elem_flag_enable(v, BM_ELEM_TAG);
}
}
BMO_op_exec(bm, &find_op);
slot_targetmap = BMO_slot_get(weld_op->slots_in, "targetmap");
/* add new merge targets to weld operator */
BMO_ITER (v, &oiter, find_op.slots_out, "targetmap.out", 0) {
v2 = BMO_iter_map_value_ptr(&oiter);
/* check in case the target vertex (v2) is already marked
* for merging */
while ((v3 = BMO_slot_map_elem_get(slot_targetmap, v2))) {
v2 = v3;
}
BMO_slot_map_elem_insert(weld_op, slot_targetmap, v, v2);
}
BMO_op_finish(bm, &find_op);
}
else {
/* transform and tag verts */
BM_ITER_MESH (v, &iter, bm, BM_VERTS_OF_MESH) {
if (!BM_elem_flag_test(v, BM_ELEM_TAG)) {
mul_m4_v3(mat, v->co);
BM_elem_flag_enable(v, BM_ELEM_TAG);
}
}
}
}
static void merge_first_last(BMesh *bm,
const ArrayModifierData *amd,
BMOperator *dupe_first,
BMOperator *dupe_last,
BMOperator *weld_op)
{
BMOperator find_op;
BMOIter oiter;
BMVert *v, *v2;
BMOpSlot *slot_targetmap;
BMO_op_initf(bm, &find_op, (BMO_FLAG_DEFAULTS & ~BMO_FLAG_RESPECT_HIDE),
"find_doubles verts=%s dist=%f keep_verts=%s",
dupe_first, "geom", amd->merge_dist,
dupe_first, "geom");
/* append the last dupe's geom to the findop input verts */
BMO_slot_buffer_append(&find_op, slots_in, "verts",
dupe_last, slots_out, "geom.out");
BMO_op_exec(bm, &find_op);
/* add new merge targets to weld operator */
slot_targetmap = BMO_slot_get(weld_op->slots_in, "targetmap");
BMO_ITER (v, &oiter, find_op.slots_out, "targetmap.out", 0) {
if (!BMO_slot_map_contains(slot_targetmap, v)) {
v2 = BMO_iter_map_value_ptr(&oiter);
BMO_slot_map_elem_insert(weld_op, slot_targetmap, v, v2);
}
}
BMO_op_finish(bm, &find_op);
}
static DerivedMesh *arrayModifier_doArray(ArrayModifierData *amd,
Scene *scene, Object *ob, DerivedMesh *dm,
ModifierApplyFlag flag)
{
DerivedMesh *result;
BMesh *bm = DM_to_bmesh(dm, false);
BMOperator first_dupe_op, dupe_op, old_dupe_op, weld_op;
BMVert **first_geom = NULL;
int i, j;
int index_len = -1; /* initialize to an invalid value */
/* offset matrix */
float offset[4][4];
float final_offset[4][4];
float length = amd->length;
int count = amd->count, maxVerts;
int *indexMap = NULL;
DerivedMesh *start_cap = NULL, *end_cap = NULL;
MVert *src_mvert;
BMOpSlot *slot_targetmap = NULL; /* for weld_op */
/* need to avoid infinite recursion here */
if (amd->start_cap && amd->start_cap != ob && amd->start_cap->type == OB_MESH)
start_cap = get_dm_for_modifier(amd->start_cap, flag);
if (amd->end_cap && amd->end_cap != ob && amd->end_cap->type == OB_MESH)
end_cap = get_dm_for_modifier(amd->end_cap, flag);
unit_m4(offset);
src_mvert = dm->getVertArray(dm);
maxVerts = dm->getNumVerts(dm);
if (amd->offset_type & MOD_ARR_OFF_CONST)
add_v3_v3v3(offset[3], offset[3], amd->offset);
if (amd->offset_type & MOD_ARR_OFF_RELATIVE) {
for (j = 0; j < 3; j++)
offset[3][j] += amd->scale[j] * vertarray_size(src_mvert, maxVerts, j);
}
if ((amd->offset_type & MOD_ARR_OFF_OBJ) && (amd->offset_ob)) {
float obinv[4][4];
float result_mat[4][4];
if (ob)
invert_m4_m4(obinv, ob->obmat);
else
unit_m4(obinv);
mul_serie_m4(result_mat, offset,
obinv, amd->offset_ob->obmat,
NULL, NULL, NULL, NULL, NULL);
copy_m4_m4(offset, result_mat);
}
if (amd->fit_type == MOD_ARR_FITCURVE && amd->curve_ob) {
Curve *cu = amd->curve_ob->data;
if (cu) {
if (!amd->curve_ob->curve_cache || !amd->curve_ob->curve_cache->path) {
cu->flag |= CU_PATH; // needed for path & bevlist
BKE_displist_make_curveTypes(scene, amd->curve_ob, 0);
}
if (amd->curve_ob->curve_cache->path) {
float scale = mat4_to_scale(amd->curve_ob->obmat);
length = scale * amd->curve_ob->curve_cache->path->totdist;
}
}
}
/* calculate the maximum number of copies which will fit within the
* prescribed length */
if (amd->fit_type == MOD_ARR_FITLENGTH || amd->fit_type == MOD_ARR_FITCURVE) {
float dist = len_v3(offset[3]);
if (dist > 1e-6f)
/* this gives length = first copy start to last copy end
* add a tiny offset for floating point rounding errors */
count = (length + 1e-6f) / dist;
else
/* if the offset has no translation, just make one copy */
count = 1;
}
if (count < 1)
count = 1;
/* calculate the offset matrix of the final copy (for merging) */
unit_m4(final_offset);
for (j = 0; j < count - 1; j++) {
float tmp_mat[4][4];
mul_m4_m4m4(tmp_mat, offset, final_offset);
copy_m4_m4(final_offset, tmp_mat);
}
/* BMESH_TODO: bumping up the stack level avoids computing the normals
* after every top-level operator execution (and this modifier has the
* potential to execute a *lot* of top-level BMOps. There should be a
* cleaner way to do this. One possibility: a "mirror" BMOp would
* certainly help by compressing it all into one top-level BMOp that
* executes a lot of second-level BMOps. */
BM_mesh_elem_toolflags_ensure(bm);
BMO_push(bm, NULL);
bmesh_edit_begin(bm, 0);
if (amd->flags & MOD_ARR_MERGE) {
BMO_op_init(bm, &weld_op, (BMO_FLAG_DEFAULTS & ~BMO_FLAG_RESPECT_HIDE),
"weld_verts");
slot_targetmap = BMO_slot_get(weld_op.slots_in, "targetmap");
}
BMO_op_initf(bm, &dupe_op, (BMO_FLAG_DEFAULTS & ~BMO_FLAG_RESPECT_HIDE),
"duplicate geom=%avef");
first_dupe_op = dupe_op;
for (j = 0; j < count - 1; j++) {
BMVert *v, *v2, *v3;
BMOpSlot *geom_slot;
BMOpSlot *geom_out_slot;
BMOIter oiter;
if (j != 0) {
BMO_op_initf(bm, &dupe_op,
(BMO_FLAG_DEFAULTS & ~BMO_FLAG_RESPECT_HIDE),
"duplicate geom=%S", &old_dupe_op, "geom.out");
}
BMO_op_exec(bm, &dupe_op);
geom_slot = BMO_slot_get(dupe_op.slots_in, "geom");
geom_out_slot = BMO_slot_get(dupe_op.slots_out, "geom.out");
if ((amd->flags & MOD_ARR_MERGEFINAL) && j == 0) {
int first_geom_bytes = sizeof(BMVert *) * geom_slot->len;
/* make a copy of the initial geometry ordering so the
* last duplicate can be merged into it */
first_geom = MEM_mallocN(first_geom_bytes, "first_geom");
memcpy(first_geom, geom_slot->data.buf, first_geom_bytes);
}
/* apply transformation matrix */
BMO_ITER (v, &oiter, dupe_op.slots_out, "geom.out", BM_VERT) {
mul_m4_v3(offset, v->co);
}
if (amd->flags & MOD_ARR_MERGE) {
/*calculate merge mapping*/
if (j == 0) {
indexMap = find_doubles_index_map(bm, &dupe_op,
amd, &index_len);
}
#define _E(s, i) ((BMVert **)(s)->data.buf)[i]
/* ensure this is set */
BLI_assert(index_len != -1);
for (i = 0; i < index_len; i++) {
if (!indexMap[i]) continue;
/* merge v (from 'geom.out') into v2 (from old 'geom') */
v = _E(geom_out_slot, i - geom_slot->len);
v2 = _E(geom_slot, indexMap[i] - 1);
/* check in case the target vertex (v2) is already marked
* for merging */
while ((v3 = BMO_slot_map_elem_get(slot_targetmap, v2))) {
v2 = v3;
}
BMO_slot_map_elem_insert(&weld_op, slot_targetmap, v, v2);
}
#undef _E
}
/* already copied earlier, but after executation more slot
* memory may be allocated */
if (j == 0)
first_dupe_op = dupe_op;
if (j >= 2)
BMO_op_finish(bm, &old_dupe_op);
old_dupe_op = dupe_op;
}
if ((amd->flags & MOD_ARR_MERGE) &&
(amd->flags & MOD_ARR_MERGEFINAL) &&
(count > 1))
{
/* Merge first and last copies. Note that we can't use the
* indexMap for this because (unless the array is forming a
* loop) the offset between first and last is different from
* dupe X to dupe X+1. */
merge_first_last(bm, amd, &first_dupe_op, &dupe_op, &weld_op);
}
/* start capping */
if (start_cap || end_cap) {
BM_mesh_elem_hflag_enable_all(bm, BM_VERT, BM_ELEM_TAG, FALSE);
if (start_cap) {
float startoffset[4][4];
invert_m4_m4(startoffset, offset);
bm_merge_dm_transform(bm, start_cap, startoffset, amd,
&first_dupe_op, first_dupe_op.slots_in, "geom", &weld_op);
}
if (end_cap) {
float endoffset[4][4];
mul_m4_m4m4(endoffset, offset, final_offset);
bm_merge_dm_transform(bm, end_cap, endoffset, amd,
&dupe_op, (count == 1) ? dupe_op.slots_in : dupe_op.slots_out,
(count == 1) ? "geom" : "geom.out", &weld_op);
}
}
/* done capping */
/* free remaining dupe operators */
BMO_op_finish(bm, &first_dupe_op);
if (count > 2)
BMO_op_finish(bm, &dupe_op);
/* run merge operator */
if (amd->flags & MOD_ARR_MERGE) {
BMO_op_exec(bm, &weld_op);
BMO_op_finish(bm, &weld_op);
}
/* Bump the stack level back down to match the adjustment up above */
BMO_pop(bm);
result = CDDM_from_bmesh(bm, FALSE);
if ((dm->dirty & DM_DIRTY_NORMALS) ||
((amd->offset_type & MOD_ARR_OFF_OBJ) && (amd->offset_ob)))
{
/* Update normals in case offset object has rotation. */
result->dirty |= DM_DIRTY_NORMALS;
}
BM_mesh_free(bm);
if (indexMap)
MEM_freeN(indexMap);
if (first_geom)
MEM_freeN(first_geom);
return result;
}
static DerivedMesh *applyModifier(ModifierData *md, Object *ob,
DerivedMesh *dm,
ModifierApplyFlag flag)
{
DerivedMesh *result;
ArrayModifierData *amd = (ArrayModifierData *) md;
result = arrayModifier_doArray(amd, md->scene, ob, dm, flag);
return result;
}
ModifierTypeInfo modifierType_Array = {
/* name */ "Array",
/* structName */ "ArrayModifierData",
/* structSize */ sizeof(ArrayModifierData),
/* type */ eModifierTypeType_Constructive,
/* flags */ eModifierTypeFlag_AcceptsMesh |
eModifierTypeFlag_SupportsMapping |
eModifierTypeFlag_SupportsEditmode |
eModifierTypeFlag_EnableInEditmode |
eModifierTypeFlag_AcceptsCVs,
/* copyData */ copyData,
/* deformVerts */ NULL,
/* deformMatrices */ NULL,
/* deformVertsEM */ NULL,
/* deformMatricesEM */ NULL,
/* applyModifier */ applyModifier,
/* applyModifierEM */ NULL,
/* initData */ initData,
/* requiredDataMask */ NULL,
/* freeData */ NULL,
/* isDisabled */ NULL,
/* updateDepgraph */ updateDepgraph,
/* dependsOnTime */ NULL,
/* dependsOnNormals */ NULL,
/* foreachObjectLink */ foreachObjectLink,
/* foreachIDLink */ NULL,
/* foreachTexLink */ NULL,
};