yggdrasil75/stupidsimcpp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fixed some issues. updated a bunch. pushing now.

Browse Source
This commit is contained in:
Yggdrasil75
2026-01-12 14:38:42 -05:00
parent b1cffb9a54
commit 5d614e6737
4 changed files with 581 additions and 42 deletions
Download Patch File Download Diff File Expand all files Collapse all files

491
util/grid/grid33.hpp Normal file
View File

@@ -0,0 +1,491 @@
#include <array>
#include <cstdint>
#include <memory>
#include <unordered_map>
#include <functional>
#include <cmath>
#include <iostream>
#include "../vectorlogic/vec3.hpp"
static inline uint32_t FindLowestOn(uint64_t v)
{
#if defined(_MSC_VER) && defined(TREEXY_USE_INTRINSICS)
unsigned long index;
_BitScanForward64(&index, v);
return static_cast<uint32_t>(index);
#elif (defined(__GNUC__) || defined(__clang__)) && defined(TREEXY_USE_INTRINSICS)
return static_cast<uint32_t>(__builtin_ctzll(v));
#else
static const unsigned char DeBruijn[64] = {
0, 1, 2, 53, 3, 7, 54, 27, 4, 38, 41, 8, 34, 55, 48, 28,
62, 5, 39, 46, 44, 42, 22, 9, 24, 35, 59, 56, 49, 18, 29, 11,
63, 52, 6, 26, 37, 40, 33, 47, 61, 45, 43, 21, 23, 58, 17, 10,
51, 25, 36, 32, 60, 20, 57, 16, 50, 31, 19, 15, 30, 14, 13, 12,
};
// disable unary minus on unsigned warning
#if defined(_MSC_VER) && !defined(__NVCC__)
#pragma warning(push)
#pragma warning(disable : 4146)
#endif
return DeBruijn[uint64_t((v & -v) * UINT64_C(0x022FDD63CC95386D)) >> 58];
#if defined(_MSC_VER) && !defined(__NVCC__)
#pragma warning(pop)
#endif
#endif
}
inline uint32_t CountOn(uint64_t v)
{
#if defined(_MSC_VER) && defined(_M_X64)
v = __popcnt64(v);
#elif (defined(__GNUC__) || defined(__clang__))
v = __builtin_popcountll(v);
#else
// Software Implementation
v = v - ((v >> 1) & uint64_t(0x5555555555555555));
v = (v & uint64_t(0x3333333333333333)) + ((v >> 2) & uint64_t(0x3333333333333333));
v = (((v + (v >> 4)) & uint64_t(0xF0F0F0F0F0F0F0F)) *
uint64_t(0x101010101010101)) >>
56;
#endif
return static_cast<uint32_t>(v);
}
template<uint32_t LOG2DIM>
class Mask {
private:
static constexpr uint32_t SIZE = std::pow(2, 3 * LOG2DIM);
static constexpr uint32_t WORD_COUNT = SIZE / 64;
uint64_t mWords[WORD_COUNT];
uint32_t findFirstOn() const {
const uint64_t *w = mWords;
uint32_t n = 0;
while (n < WORD_COUNT && !*w) {
++w;
++n;
}
return n == WORD_COUNT ? SIZE : (n << 6) + FindLowestOn(*w);
}
uint32_t findNextOn(uint32_t start) const {
uint32_t n start >> 6;
if (n >= WORD_COUNT) {
return SIZE;
}
uint32_t m = start & 63;
uint64_t b = mWords[n];
if (b & (uint64_t(1) << m)) {
return start;
}
b &= ~uint64_t(0) << m;
while (!b && ++n < WORD_COUNT) {
b = mWords[n];
}
return (!b ? SIZE : (n << 6) + FindLowestOn(b));
}
public:
static size_t memUsage() {
return sizeof(Mask);
}
static uint32_t bitCount() {
return SIZE;
}
static uint32_t wordCount() {
return WORD_COUNT;
}
uint64_t getWord(size_t n) const {
return mWords[n];
}
void setWord(size_t n, uint64_t v) {
mWords[n] = v;
}
uint32_t countOn() const {
uint32_t sum = 0;
uint32_t n = WORD_COUNT;
for (const uint64_t* w = mWords; n--; ++w) {
sum += CountOn(*w);
}
return sum;
}
class Iterator {
private:
uint32_t mPos;
const Mask* mParent;
public:
Iterator() : mPos(Mask::SIZE), mParent(nullptr) {}
Iterator(uint32_t pos, const Mask* parent) : mPos(pos), mParent(parent) {}
Iterator& operator=(const Iterator&) = default;
uint32_t opeator*() const {
return mPos;
}
operator bool() const {
return mPos != Mask::SIZE;
}
Iterator& operator++() {
mPos = mParent -> findNextOn(mPos + 1);
return *this;
}
}
Mask() {
for (uint32_t i = 0; i < WORD_COUNT; ++i) {
mWords[i] = 0;
}
}
Mask(bool on) {
const uint64_t v = on ? ~uint64_t(0) : uint64_t(0);
for (uint32_t i = 0; i < WORD_COUNT; ++i) {
mWords[i] = v;
}
}
Mask(const Mask &other) {
for (uint32_t i = 0; i < WORD_COUNT; ++i) {
mWords[i] = other.mWords[i];
}
}
template<typename WordT>
WordT getWord(int n) const {
return reinterpret_cast<const WordT *>(mWords)[n];
}
Mask &operator=(const Mask &other) {
// static_assert(sizeof(Mask) == sizeof(Mask), "Mismatching sizeof");
// static_assert(WORD_COUNT == Mask::WORD_COUNT, "Mismatching word count");
// static_assert(LOG2DIM == Mask::LOG2DIM, "Mismatching LOG2DIM");
uint64_t *src = reinterpret_cast<const uint64_t* >(&other);
uint64_t *dst = mWords;
for (uint32_t i = 0; i < WORD_COUNT; ++i) {
*dst++ = *src++;
}
return *this;
}
bool operator==(const Mask &other) const {
for (uint32_t i = 0; i < WORD_COUNT; ++i) {
if (mWords[i] != other.mWords[i]) return false;
}
return true;
}
bool operator!=(const Mask &other) const {
return !((*this) == other);
}
Iterator beginOn() const {
return Iterator(this->findFirstOn(), this);
}
/// @brief return true if bit n is set.
/// @param n the bit to check
/// @return true otherwise return false
bool isOn(uint32_t n) const {
return 0 != (mWords[n >> 6] & (uint64_t(1) << (n&63)));
}
bool isOn() const {
for (uint32_t i = 0; i < WORD_COUNT; ++i) {
if (mWords[i] != ~uint64_t(0)) return false;
}
return true;
}
bool isOff() const {
for (uint32_t i = 0; i < WORD_COUNT; ++i) {
if (mWords[i] != ~uint64_t(0)) return true;
}
return false;
}
bool setOn(uint32_t n) {
uint64_t &word = mWords[n >> 6];
const uint64_t bit = (uint64_t(1) << (n & 63));
bool wasOn = word & bit;
word |= bit;
return wasOn;
}
void setOff(uint32_t n) {
mWords[n >> 6] &= ~(uint64_t(1) << (n & 63));
}
template<bool UseBranchless = true>
void set(uint32_t n, bool On) {
if constexpr (UseBranchless) {
auto &word = mWords[n >> 6];
n &= 63;
word &= ~(uint64_t(1) << n);
word |= uint64_t(On) << n;
} else {
On ? this->setOn(n) : this->setOff(n);
}
}
void setOn() {
for (uint32_t i = 0; i < WORD_COUNT; ++i) {
mWords[i] = ~uint64_t(0);
}
}
void setOff() {
for (uint32_t i = 0; i < WORD_COUNT; ++i) {
mWords[i] = uint64_t(0);
}
}
void set(bool on) {
const uint64_t v = on ? ~uint64_t(0) : uint64_t(0);
for (uint32_t i = 0; i < WORD_COUNT; ++i) {
mWords[i] = v;
}
}
void toggle() {
uint32_t n = WORD_COUNT;
for (auto* w = mWords; n--; ++w) {
*w = ~*w;
}
}
void toggle(uint32_t n) {
mWords[n >> 6] ^= uint64_t(1) << (n & 63);
}
};
template <typename DataT, int Log2DIM>
class Grid {
constexpr static int DIM = 1 << Log2DIM;
constexpr static int SIZE = DIM * DIM * DIM;
std::array<DataT, SIZE> data;
Mask<Log2DIM> mask;
};
template <typename DataT, int INNER_BITS = 2, int LEAF_BITS = 3>
class VoxelGrid {
private:
public:
constexpr static int32_t Log2N = INNER_BITS + LEAF_BITS;
using LeafGrid = Grid<DataT, LEAF_BITS>;
using InnerGrid = Grid<std::shared_ptr<LeafGrid>, INNER_BITS>;
using RootMap = std::unordered_map<Vec3i, InnerGrid>;
RootMap root_map;
const double resolution;
const double inv_resolution;
const double half_resolution;
VoxelGrid(double voxel_size) : resolution(voxel_size), inv_resolution(1.0 / voxel_size), half_resolution(0.5 * voxel_size) {}
size_t getMemoryUsage() const {
size_t total_size = 0;
for (unsigned i = 0; i < root_map.bucket_count(); ++i) {
size_t bucket_size = root_map.bucket_size(i);
if (bucket_size == 0) {
total_size++;
} else {
total_size += bucket_size;
}
}
size_t entry_size = sizeof(Vec3i) + sizeof(InnerGrid) + sizeof(void *);
total_size += root_map.size() * entry_size;
for (const auto& [key, inner_grid] : root_map) {
total_size += inner_grid.mask.countOn() * sizeof(LeafGrid);
}
return total_size;
}
static inline Veci PosToCoord(float x, float y, float z) {
// union VI {
// __m128i m;
// int32_t i[4];
// };
// static __m128 RES = _mm_set1_ps(inv_resolution);
// __m128 vect = _mm_set_ps(x, y, z, 0.0);
// __m128 res = _mm_mul_ps(vect, RES);
// VI out;
// out.m = _mm_cvttps_epi32(_mm_floor_ps(res));
// return {out.i[3], out.i[2], out.i[1]};
return Vec3f(x,y,z).floorToI();
}
static inline Vec3i posToCoord(double x, double y, double z) {
return Vec3f(x,y,z).floorToI();
}
static inline Vec3i posToCoord(const Vec3d &pos) {
return pos.floorToI();
}
Vec3d Vec3ioPos(const Vec3i &coord) {
return (coord.toDouble() * resolution) + half_resolution;
}
template <class VisitorFunction>
void forEachCell(VisitorFunction func) {
constexpr static int32_t MASK_LEAF = ((1 << LEAF_BITS) - 1);
constexpr static int32_t MASK_INNER = ((1 << INNER_BITS) - 1);
for (auto& map_it : root_map) {
const auto& [xA, yA, zA] = (map_it.first);
InnerGrid& inner_grid = map_it.second;
auto& mask2 = inner_grid.mask;
for (auto inner_it = mask2.beginOn(); inner_it; ++inner_it) {
const int32_t inner_index = *inner_it;
int32_t xB = xA | ((inner_index & MASK_INNER) << LEAF_BITS);
int32_t yB = yA | (((inner_index >> INNER_BITS) & MASK_INNER) << LEAF_BITS);
int32_t zB = zA | (((inner_index >> (INNER_BITS* 2)) & MASK_INNER) << LEAF_BITS);
auto& leaf_grid = inner_grid.data[inner_index];
auto& mask1 = leaf_grid->mask;
for (auto leaf_it = mask1.beginOn(); leaf_it; ++leaf_it){
const int32_t leaf_index = *leaf_it;
Vec3i pos = Vec3i(xB | (leaf_index & MASK_LEAF),
yB | ((leaf_index >> LEAF_BITS) & MASK_LEAF),
zB | ((leaf_index >> (LEAF_BITS * 2)) & MASK_LEAF));
func(leaf_grid->data[leaf_index], pos);
}
}
}
}
class Accessor {
private:
RootMap &root_;
Vec3i prev_root_coord_;
Vec3i prev_inner_coord_;
InnerGrid* prev_inner_ptr_ = nullptr;
LeafGrid* prev_leaf_ptr_ = nullptr;
public:
Accessor(RootMap& root) : root_(root) {}
bool setValue(const Vec3i& coord, const DataT& value) {
LeafGrid* leaf_ptr = prev_leaf_ptr_;
const Vec3i inner_key = getInnerKey(coord);
if (inner_key != prev_inner_coord_ || !prev_leaf_ptr_) {
InnerGrid* inner_ptr = prev_inner_ptr_;
const Vec3i root_key = getRootKey(coord);
if (root_key != prev_root_coord_ || !prev_inner_ptr_) {
auto root_it = root_.find(root_key);
if (root_it == root_.end()) {
root_it = root_insert({root_key, InnerGrid()}).first;
}
inner_ptr = &(root_it->second);
prev_root_coord_ = root_key;
prev_inner_ptr_ = inner_ptr;
}
const uint32_t inner_index = getInnerIndex(coord);
auto& inner_data = inner_ptr->data[inner_index];
if (inner_ptr->mask.setOn(inner_index)) {
inner_data = std::make_shared<LeafGrid>();
}
leaf_ptr = inner_data.get();
prev_inner_coord_ = inner_key;
prev_leaf_ptr_ = leaf_ptr;
}
const uint32_t leaf_index = getLeafIndex(coord);
bool was_on = leaf_ptr->mask.setOn(leaf_index);
leaf_ptr->data[leaf_index] = value;
return was_on;
}
DataT* value(const Vec3i& coord) {
LeafGrid* leaf_ptr = prev_leaf_ptr_;
const Vec3i inner_key = getInnerKey(coord);
if (inner_key != prev_inner_coord_ || !prev_inner_ptr_) {
InnerGrid* inner_ptr = prev_inner_ptr_;
const Vec3i root_key = getRootKey(coord);
if (root_key != prev_root_coord_ || !prev_inner_ptr_) {
auto it = root_.find(root_key);
if (it == root_.end()) {
return nullptr;
}
inner_ptr = &(it->second);
prev_inner_coord_ = root_key;
prev_inner_ptr_ = inner_ptr;
}
const uint32_t inner_index = getInnerIndex(coord);
auto& inner_data - inner_ptr->data[inner_index];
if (!inner_ptr->mask.isOn(inner_index)) {
return nullptr;
}
leaf_ptr = inner_ptr->data[inner_index].get();
prev_inner_coord_ = inner_key;
prev_leaf_ptr_ = leaf_ptr;
}
const uint32_t leaf_index = getLeafIndex(coord);
if (!leaf_ptr->mask.isOn(leaf_index)) {
return nullptr;
}
return &(leaf_ptr->data[leaf_index]);
}
const InnerGrid* lastInnerGrid() const {
return prev_inner_ptr_;
}
const LeafGrid* lastLeafGrid() const {
return prev_leaf_ptr_;
}
}
Accessor createAccessor() {
return Accessor(root_map);
}
static inline Vec3i getRootKey(const Vec3i& coord) {
constexpr static int32_t MASK = ~((1 << Log2N) - 1);
return {coord.x & MASK, coord.y & MASK, coord.z & MASK};
}
static inline Vec3i getInnerKey(const Vec3i &coord)
{
constexpr static int32_t MASK = ~((1 << LEAF_BITS) - 1);
return {coord.x & MASK, coord.y & MASK, coord.z & MASK};
}
static inline uint32_t getInnerIndex(const Vec3i &coord)
{
constexpr static int32_t MASK = ((1 << INNER_BITS) - 1);
// clang-format off
return ((coord.x >> LEAF_BITS) & MASK) +
(((coord.y >> LEAF_BITS) & MASK) << INNER_BITS) +
(((coord.z >> LEAF_BITS) & MASK) << (INNER_BITS * 2));
// clang-format on
}
static inline uint32_t getLeafIndex(const Vec3i &coord)
{
constexpr static int32_t MASK = ((1 << LEAF_BITS) - 1);
// clang-format off
return (coord.x & MASK) +
((coord.y & MASK) << LEAF_BITS) +
((coord.z & MASK) << (LEAF_BITS * 2));
// clang-format on
}
};