casey-bianco-davis/blender
0
0
Fork 0
forked from blender/blender
Code Pull Requests Activity

GPv3: Add Compound Shapes Rendering (i.e. Hole rendering) #12

Closed
casey-bianco-davis wants to merge 8 commits from GPv3-Compound-Shapes-2 into main
pull from: GPv3-Compound-Shapes-2
merge into: casey-bianco-davis:main

When changing the target branch, be careful to rebase the branch in your fork to match. See documentation.
Conversation 0 Commits 8 Files Changed 6 +38 -6
1 changed files with 38 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files
Showing only changes of commit 75002ef892 - Show all commits

44
source/blender/blenkernel/intern/grease_pencil.cc
View File

@ -510,6 +510,18 @@ static bool check_valid_curves(Span<float2> projverts, const OffsetIndices<int>
Span<Vector<uint3>> Drawing::triangles() const
{
struct LocalMemArena {
MemArena *pf_arena = nullptr;
LocalMemArena() : pf_arena(BLI_memarena_new(BLI_MEMARENA_STD_BUFSIZE, "Drawing::triangles")) {}
~LocalMemArena()
{
if (pf_arena != nullptr) {
BLI_memarena_free(pf_arena);
}
}
};
this->runtime->triangles_cache.ensure([&](Vector<Vector<uint3>> &r_data) {
const CurvesGeometry &curves = this->strokes();
const Span<float3> positions = curves.evaluated_positions();
@ -521,11 +533,11 @@ Span<Vector<uint3>> Drawing::triangles() const
const Array<IndexMask> groups = this->get_shapes_index_masks(memory);
r_data.resize(groups.size() + 1);
// r_data.resize(curves.curves_num());
MutableSpan<Vector<uint3>> strokes_triangles = r_data.as_mutable_span();
threading::EnumerableThreadSpecific<LocalMemArena> all_local_mem_arenas;
threading::parallel_for(groups.index_range(), 32, [&](const IndexRange group_range) {
MemArena *pf_arena = all_local_mem_arenas.local().pf_arena;
for (const int group_id : group_range) {
const IndexMask &group = groups[group_id];
@ -540,18 +552,38 @@ Span<Vector<uint3>> Drawing::triangles() const
const int num_points = points_by_group.total_size();
Array<float2> projverts2(num_points);
float(*projverts)[2] = static_cast<float(*)[2]>(
BLI_memarena_alloc(pf_arena, sizeof(*projverts) * size_t(num_points)));
group.foreach_index(GrainSize(256), [&](const int64_t curve_i, const int64_t pos) {
const IndexRange point_group = points_by_group[pos];
const IndexRange points = points_by_curve[curve_i];
threading::parallel_for(points.index_range(), 512, [&](const IndexRange range) {
for (const int p_id : range) {
mul_v2_m3v3(projverts2[point_group[p_id]], axis_mat.ptr(), positions[points[p_id]]);
mul_v2_m3v3(projverts[point_group[p_id]], axis_mat.ptr(), positions[points[p_id]]);
}
});
});
if (!check_valid_curves(projverts2, points_by_group)) {
/* If there is only on stroke or the geometry can not meshed then use simple poly fill
* using the first curve in the group. */
if (group.size() == 1 ||
!check_valid_curves({reinterpret_cast<float2 *>(projverts), num_points},
points_by_group))
{
const IndexRange points = points_by_curve[group.first()];
strokes_triangles[group_id].resize(points.size() - 2);
MutableSpan<uint3> r_tris = strokes_triangles[group_id];
BLI_polyfill_calc_arena(projverts,
points.size(),
0,
reinterpret_cast<uint32_t(*)[3]>(r_tris.data()),
pf_arena);
for (const int i : r_tris.index_range()) {
r_tris[i] += points.first();
}
BLI_memarena_clear(pf_arena);
continue;
}
@ -568,7 +600,7 @@ Span<Vector<uint3>> Drawing::triangles() const
threading::parallel_for(points.index_range(), 512, [&](const IndexRange range) {
for (const int p_id : range) {
vert_to_point_map[point_group[p_id]] = points[p_id];
verts[point_group[p_id]] = double2(projverts2[point_group[p_id]]);
verts[point_group[p_id]] = double2(projverts[point_group[p_id]]);
edges[point_group[p_id]] = std::pair<int, int>(
point_group[p_id], point_group[(p_id + 1) % points.size()]);
faces[pos][p_id] = point_group[p_id];