63a44e29ac
Since UVs are now stored as 2D vectors in meshes, they can be copied directly to the vertex buffers. Somewhat surprisingly, multithreading the copying into the vertex buffer provides a good speedup-- on a CPU with many cores at least. Here is a test uploading two UV maps created in geometry nodes with a 1 million quad mesh, with a Ryzen 7950x: | | Before | After | Speedup | | ------- | ------- | ------ | ------- | | Average | 24.3 ms | 7.5 ms | 3.2x | | Min | 17.6 ms | 7.0 ms | 2.5x | I added the copying utilities to the array utils header, since the need for them has come up in a few different places already, and the existing function with a selection argument didn't make sense here. Pull Request: https://projects.blender.org/blender/blender/pulls/105793
56 lines
1.7 KiB
C++
56 lines
1.7 KiB
C++
/* SPDX-License-Identifier: GPL-2.0-or-later */
|
|
|
|
#include "BLI_array_utils.hh"
|
|
|
|
namespace blender::array_utils {
|
|
|
|
void copy(const GVArray &src, GMutableSpan dst, const int64_t grain_size)
|
|
{
|
|
BLI_assert(src.type() == dst.type());
|
|
BLI_assert(src.size() == dst.size());
|
|
threading::parallel_for(src.index_range(), grain_size, [&](const IndexRange range) {
|
|
src.materialize_to_uninitialized(range, dst.data());
|
|
});
|
|
}
|
|
|
|
void copy(const GVArray &src,
|
|
const IndexMask selection,
|
|
GMutableSpan dst,
|
|
const int64_t grain_size)
|
|
{
|
|
BLI_assert(src.type() == dst.type());
|
|
BLI_assert(src.size() >= selection.min_array_size());
|
|
BLI_assert(dst.size() >= selection.min_array_size());
|
|
threading::parallel_for(selection.index_range(), grain_size, [&](const IndexRange range) {
|
|
src.materialize_to_uninitialized(selection.slice(range), dst.data());
|
|
});
|
|
}
|
|
|
|
void gather(const GVArray &src,
|
|
const IndexMask indices,
|
|
GMutableSpan dst,
|
|
const int64_t grain_size)
|
|
{
|
|
BLI_assert(src.type() == dst.type());
|
|
BLI_assert(indices.size() == dst.size());
|
|
threading::parallel_for(indices.index_range(), grain_size, [&](const IndexRange range) {
|
|
src.materialize_compressed_to_uninitialized(indices.slice(range), dst.slice(range).data());
|
|
});
|
|
}
|
|
|
|
void gather(const GSpan src, const IndexMask indices, GMutableSpan dst, const int64_t grain_size)
|
|
{
|
|
gather(GVArray::ForSpan(src), indices, dst, grain_size);
|
|
}
|
|
|
|
void invert_booleans(MutableSpan<bool> span)
|
|
{
|
|
threading::parallel_for(span.index_range(), 4096, [&](IndexRange range) {
|
|
for (const int i : range) {
|
|
span[i] = !span[i];
|
|
}
|
|
});
|
|
}
|
|
|
|
} // namespace blender::array_utils
|