logoChat01.ai

#include "pch.h" #define XOR_RAND(state, result_v...

تم الإنشاء في: ١٠ أكتوبر ٢٠٢٥

تمت الإجابة باستخدام GPT-5 Thinking بواسطة Chat01

سؤال

#include "pch.h"

#define XOR_RAND(state, result_var)
do {
int s = state;
s ^= s << 13;
s ^= s >> 17;
s ^= s << 5;
state = s;
result_var = state * 0x1.0p-32f;
} while (0)

#define XOR_RAND_GRSH(state, result_var)
do {
int s = state;
s ^= s << 13;
s ^= s >> 17;
s ^= s << 5;
state = s;
result_var = fmaf(state, 0x1.0p-31f, -1.0f);
} while (0)

#define FABE13_COS(x, result_var)
do {
float abs_val = fabsf(x);
float reduced = fmodf(abs_val, 6.28318530718f);
if (reduced > 3.14159265359f) {
reduced = 6.28318530718f - reduced;
}
if (reduced < 1.57079632679f) {
float val2 = reduced * reduced;
result_var = fmaf(val2, fmaf(val2, 0.0416666667f, -0.5f), 1.0f);
} else {
reduced = 3.14159265359f - reduced;
float val2 = reduced * reduced;
result_var = -fmaf(val2, fmaf(val2, 0.0416666667f, -0.5f), 1.0f);
}
} while (0)

#define FABE13_SINCOS(in, sin_out, cos_out, n)
do {
int i = 0;
int limit = n & ~7;
if (n >= 8) {
static __declspec(align(32)) const __m256 VEC_TWOPI = _mm256_set1_ps(6.28318530718f);
static __declspec(align(32)) const __m256 VEC_PI = _mm256_set1_ps(3.14159265359f);
static __declspec(align(32)) const __m256 VEC_PI_2 = _mm256_set1_ps(1.57079632679f);
static __declspec(align(32)) const __m256 INV_TWOPI = _mm256_set1_ps(0.15915494309189535f);
static __declspec(align(32)) const __m256 BIAS = _mm256_set1_ps(12582912.0f);
static __declspec(align(32)) const __m256 VEC_COS_P3 = _mm256_set1_ps(0.0416666667f);
static __declspec(align(32)) const __m256 VEC_COS_P1 = _mm256_set1_ps(-0.5f);
static __declspec(align(32)) const __m256 VEC_COS_P0 = _mm256_set1_ps(1.0f);
static __declspec(align(32)) const __m256 VEC_SIN_P1 = _mm256_set1_ps(-0.16666666f);
static __declspec(align(32)) const __m256 VEC_SIN_P0 = _mm256_set1_ps(1.0f);
static __declspec(align(32)) const __m256 VEC_ZERO = _mm256_setzero_ps();
while (i < limit) {
__m256 vx = _mm256_load_ps(&in[i]);
__m256 vax = _mm256_andnot_ps(_mm256_set1_ps(-0.0f), vx);
__m256 q = _mm256_fmadd_ps(vax, INV_TWOPI, BIAS);
q = _mm256_sub_ps(q, BIAS);
__m256 r = _mm256_fnmadd_ps(VEC_TWOPI, q, vax);
__m256 r1 = _mm256_min_ps(r, _mm256_sub_ps(VEC_TWOPI, r));
__m256 r2 = _mm256_min_ps(r1, _mm256_sub_ps(VEC_PI, r1));
__m256 t2 = _mm256_mul_ps(r2, r2);
__m256 cosv = _mm256_fmadd_ps(t2, _mm256_fmadd_ps(t2, VEC_COS_P3, VEC_COS_P1), VEC_COS_P0);
__m256 sinv = _mm256_mul_ps(_mm256_fmadd_ps(t2, VEC_SIN_P1, VEC_SIN_P0), r2);
__m256 cflip = _mm256_cmp_ps(r1, VEC_PI_2, _CMP_GT_OQ);
__m256 sflip = _mm256_xor_ps(_mm256_cmp_ps(vx, VEC_ZERO, _CMP_LT_OQ),
_mm256_cmp_ps(r, VEC_PI, _CMP_GT_OQ));
cosv = _mm256_blendv_ps(cosv, _mm256_sub_ps(VEC_ZERO, cosv), cflip);
sinv = _mm256_blendv_ps(sinv, _mm256_sub_ps(VEC_ZERO, sinv), sflip);
_mm256_store_ps(&cos_out[i], cosv);
_mm256_store_ps(&sin_out[i], sinv);
i += 8;
}
}
while (i < n) {
float x = in[i];
float ax = fabsf(x);
float q = fmaf(ax, 0.15915494309189535f, 12582912.0f);
q -= 12582912.0f;
float r = fmaf(-6.28318530718f, q, ax);
bool sflip = (r > 3.14159265359f);
if (sflip) r = 6.28318530718f - r;
bool cflip = (r > 1.57079632679f);
if (cflip) r = 3.14159265359f - r;
float t2 = r * r;
float c = fmaf(t2, fmaf(t2, 0.0416666667f, -0.5f), 1.0f);
float s = fmaf(t2, -0.16666666f, 1.0f) * r;
cos_out[i] = cflip ? -c : c;
sin_out[i] = ((x < 0.0f) ^ sflip) ? -s : s;
++i;
}
} while (0)

static __declspec(noalias) __forceinline float ShekelFunc(const float x, const float seed) noexcept
{
int i = 0;
float current_state = seed, current_res, current_res2, res = 0.0f;
while (i < 10) {
XOR_RAND(current_state, current_res);
const float x_part = fmaf(-current_res, 10.0f, x);
XOR_RAND(current_state, current_res);
XOR_RAND(current_state, current_res2);
float delimiter = fmaf(fmaf(current_res, 20.0f, 5.0f),
x_part * x_part,
fmaf(current_res2, 0.2f, 1.0f));
delimiter = copysignf(fmaxf(fabsf(delimiter), FLT_MIN), delimiter);
res -= 1.0f / delimiter;
++i;
}
return res;
}

static __declspec(noalias) __forceinline float RastriginFunc(const float x1, const float x2) noexcept
{
const float term1 = fmaf(x1, x1, x2 * x2);
float cos1, cos2;
FABE13_COS(6.28318530717958647692f * x1, cos1);
FABE13_COS(6.28318530717958647692f * x2, cos2);
return (term1 - fmaf(cos1 + cos2, 10.0f, -14.6f)) * fmaf(-term1, 0.25f, 18.42f);
}

static __declspec(noalias) __forceinline float HillFunc(const float x, const float seed) noexcept
{
int j = 0;
static __declspec(align(32)) float angles[14u];
const float start_angle = 6.28318530717958647692f * x;
#pragma loop(ivdep)
while (j < 14) {
angles[j] = start_angle * (j + 1);
++j;
}
static __declspec(align(32)) float sin_vals[14u];
static __declspec(align(32)) float cos_vals[14u];
FABE13_SINCOS(angles, sin_vals, cos_vals, 14u);
float current_state = seed, current_res, current_res2;
XOR_RAND(current_state, current_res);
float res = fmaf(current_res, 2.0f, -1.1f);
--j;
while (j >= 0) {
XOR_RAND(current_state, current_res);
XOR_RAND(current_state, current_res2);
res += fmaf(fmaf(current_res, 2.0f, -1.1f), sin_vals[j],
fmaf(current_res, 2.0f, -1.1f) * cos_vals[j]);
--j;
}
return res;
}

static __declspec(noalias) __forceinline float GrishaginFunc(const float x1, const float x2, const float seed) noexcept
{
int j = 0;
static __declspec(align(32)) float angles_j[8u];
static __declspec(align(32)) float angles_k[8u];
#pragma loop(ivdep)
while (j < 8) {
const float pj_mult = 3.14159265358979323846f * (j + 1);
angles_j[j] = pj_mult * x1;
angles_k[j] = pj_mult * x2;
++j;
}
static __declspec(align(32)) float sin_j[8u], cos_j[8u];
static __declspec(align(32)) float sin_k[8u], cos_k[8u];
FABE13_SINCOS(angles_j, sin_j, cos_j, 8u);
FABE13_SINCOS(angles_k, sin_k, cos_k, 8u);
--j;
float part1 = 0.0f;
float part2 = 0.0f;
float current_state = seed, current_res, current_res2;
while (j >= 0) {
size_t k = 0u;
while (k < 8u) {
const float sin_term = sin_j[j] * sin_j[j];
const float cos_term = cos_k[k] * cos_k[k];
XOR_RAND_GRSH(current_state, current_res);
XOR_RAND_GRSH(current_state, current_res2);
part1 = fmaf(current_res, sin_term,
fmaf(current_res2, cos_term, part1));
XOR_RAND_GRSH(current_state, current_res);
XOR_RAND_GRSH(current_state, current_res2);
part2 = fmaf(-current_res, cos_term,
fmaf(current_res2, sin_term, part2));
++k;
}
--j;
}
return -sqrtf(fmaf(part1, part1, part2 * part2));
}

static __declspec(noalias) __forceinline float Shag(const float _m, const float x1, const float x2, const float y1,
const float y2, const float _N, const float _r) noexcept
{
const float diff = y2 - y1;
const float sign_mult = _mm_cvtss_f32(_mm_castsi128_ps(_mm_set1_epi32(
0x3F800000 | ((((const int)&diff) & 0x80000000) ^ 0x80000000)
)));
return _N == 1.0f
? fmaf(-(1.0f / _m), diff, x1 + x2) * 0.5f
: _N == 2.0f
? fmaf(sign_mult / (_m * _m), diff * diff * _r, x1 + x2) * 0.5f
: fmaf(sign_mult / powf(_m, _N), powf(diff, _N) * _r, x1 + x2) * 0.5f;
}

__declspec(align(64)) struct PreallocatedPeanoMemoryManager final {
__declspec(align(64)) struct MemoryBlock final {
char* __restrict data;
char* __restrict current;
const char* __restrict end;

text
	__declspec(noalias) __forceinline MemoryBlock() noexcept
		: data(nullptr), current(nullptr), end(nullptr)
	{
	}

	const __declspec(noalias) __forceinline void initialize(const size_t bytes) noexcept
	{
		const size_t aligned_need = (bytes + 1048638) & ~(1048575);
		void* p = VirtualAlloc(nullptr, aligned_need,
			MEM_RESERVE | MEM_COMMIT,
			PAGE_READWRITE);

		data = reinterpret_cast<char* __restrict>(p);
		const uintptr_t base = reinterpret_cast<uintptr_t>(data);
		const uintptr_t aligned = (base + 63) & ~63;
		current = reinterpret_cast<char* __restrict>(aligned);
		end = reinterpret_cast<char*>(current + bytes);
	}

	__declspec(noalias) __forceinline void* allocate_from_block(const size_t alloc_size) noexcept
	{
		const size_t n = (alloc_size + 63) & ~63;
		char* __restrict result = current;
		current += n;
		return result;
	}
};

tbb::enumerable_thread_specific<MemoryBlock> tls;

const __declspec(noalias) __forceinline size_t calculate_tree_size_bytes(const int max_depth) const noexcept
{
	static const size_t bytes[] = {
			1398080u, 5592384u, 22369600u, 89478464u, 357913920u, 1431655744u, 5726623040u
	};
	return bytes[max_depth - 7u];
}

const __declspec(noalias) __forceinline void preallocate_for_depth(const int max_depth) noexcept
{
	const size_t required_size = calculate_tree_size_bytes(max_depth);
	tls.clear();
	tls = tbb::enumerable_thread_specific<MemoryBlock>(
		[required_size]() {
			MemoryBlock b;
			b.initialize(required_size);
			return b;
		}
	);
}

__declspec(noalias) __forceinline void* allocate(const size_t size) noexcept
{
	return tls.local().allocate_from_block(size);
}

};

static PreallocatedPeanoMemoryManager peano_memory_manager;

__declspec(align(16)) struct Interval final {
const float x1;
const float x2;
const float y1;
const float y2;
const float delta_y;
const float N_factor;
const float M;
float R;

text
__declspec(noalias) __forceinline void* operator new(size_t sz) noexcept { return peano_memory_manager.allocate(sz); }

__declspec(noalias) __forceinline Interval(const float _x1, const float _x2,
	const float _y1, const float _y2, const float _N) noexcept
	: x1(_x1), x2(_x2), y1(_y1), y2(_y2),
	delta_y(_y2 - _y1),
	N_factor(_N == 1.0f ? _x2 - _x1 :
		_N == 2.0f ? sqrtf(_x2 - _x1) :
		powf(_x2 - _x1, 1.0f / _N)),
	M(fabsf(delta_y)* (1.0f / N_factor))
{
}

__declspec(noalias) __forceinline void ChangeCharacteristic(const float _m) noexcept
{
	R = fmaf(
		-(y2 + y1),
		2.0f,
		fmaf(_m, N_factor, (delta_y * delta_y) * (1.0f / (_m * N_factor))));
}

};

static __declspec(noalias) __forceinline bool ComparePtr(const Interval* __restrict a, const Interval* __restrict b) noexcept
{
return a->R < b->R;
}

const enum List : uint8_t {
Top = 0b00,
Down = 0b01,
Left = 0b10,
Right = 0b11
};

__declspec(align(16)) struct PeanoCurve_2D final {
const List Type;
const int razvertka;
const float a;
const float b;
const float c;
const float d;
const float x1;
const float x2;
PeanoCurve_2D* __restrict DownLeft;
PeanoCurve_2D* __restrict TopLeft;
PeanoCurve_2D* __restrict TopRight;
PeanoCurve_2D* __restrict DownRight;

text
__declspec(noalias) __forceinline void* operator new(size_t sz) noexcept { return peano_memory_manager.allocate(sz); }

__declspec(noalias) __forceinline PeanoCurve_2D(
	const List _Type,
	int _razvertka,
	const float _a,
	const float _b,
	const float _c,
	const float _d) noexcept
	: Type(_Type),
	razvertka(_razvertka),
	a(_a),
	b(_b),
	c(_c),
	d(_d),
	x1(fmaf(0.5f, _a, 0.5f * _b)),
	x2(fmaf(0.5f, _c, 0.5f * _d))
{
	if (_razvertka-- != 0) {
		if (_razvertka > 8) {
			switch (_Type) {
			case Top:
				tbb::parallel_invoke(
					[&]() { DownLeft = new PeanoCurve_2D(Right, _razvertka, _a, x1, _c, x2); },
					[&]() { TopLeft = new PeanoCurve_2D(Top, _razvertka, _a, x1, x2, _d); },
					[&]() { TopRight = new PeanoCurve_2D(Top, _razvertka, x1, _b, x2, _d); },
					[&]() { DownRight = new PeanoCurve_2D(Left, _razvertka, x1, _b, _c, x2); }
				);
				break;
			case Down:
				tbb::parallel_invoke(
					[&]() { TopRight = new PeanoCurve_2D(Left, _razvertka, x1, _b, x2, _d); },
					[&]() { DownRight = new PeanoCurve_2D(Down, _razvertka, x1, _b, _c, x2); },
					[&]() { DownLeft = new PeanoCurve_2D(Down, _razvertka, _a, x1, _c, x2); },
					[&]() { TopLeft = new PeanoCurve_2D(Right, _razvertka, _a, x1, x2, _d); }
				);
				break;
			case Right:
				tbb::parallel_invoke(
					[&]() { DownLeft = new PeanoCurve_2D(Top, _razvertka, _a, x1, _c, x2); },
					[&]() { DownRight = new PeanoCurve_2D(Right, _razvertka, x1, _b, _c, x2); },
					[&]() { TopRight = new PeanoCurve_2D(Right, _razvertka, x1, _b, x2, _d); },
					[&]() { TopLeft = new PeanoCurve_2D(Down, _razvertka, _a, x1, x2, _d); }
				);
				break;
			default:
				tbb::parallel_invoke(
					[&]() { TopRight = new PeanoCurve_2D(Down, _razvertka, x1, _b, x2, _d); },
					[&]() { TopLeft = new PeanoCurve_2D(Left, _razvertka, _a, x1, x2, _d); },
					[&]() { DownLeft = new PeanoCurve_2D(Left, _razvertka, _a, x1, _c, x2); },
					[&]() { DownRight = new PeanoCurve_2D(Top, _razvertka, x1, _b, _c, x2); }
				);
			}
		}
		else {
			switch (_Type) {
			case Top:
				DownLeft = new PeanoCurve_2D(Right, _razvertka, _a, x1, _c, x2);
				TopLeft = new PeanoCurve_2D(Top, _razvertka, _a, x1, x2, _d);
				TopRight = new PeanoCurve_2D(Top, _razvertka, x1, _b, x2, _d);
				DownRight = new PeanoCurve_2D(Left, _razvertka, x1, _b, _c, x2);
				break;
			case Down:
				TopRight = new PeanoCurve_2D(Left, _razvertka, x1, _b, x2, _d);
				DownRight = new PeanoCurve_2D(Down, _razvertka, x1, _b, _c, x2);
				DownLeft = new PeanoCurve_2D(Down, _razvertka, _a, x1, _c, x2);
				TopLeft = new PeanoCurve_2D(Right, _razvertka, _a, x1, x2, _d);
				break;
			case Right:
				DownLeft = new PeanoCurve_2D(Top, _razvertka, _a, x1, _c, x2);
				DownRight = new PeanoCurve_2D(Right, _razvertka, x1, _b, _c, x2);
				TopRight = new PeanoCurve_2D(Right, _razvertka, x1, _b, x2, _d);
				TopLeft = new PeanoCurve_2D(Down, _razvertka, _a, x1, x2, _d);
				break;
			default:
				TopRight = new PeanoCurve_2D(Down, _razvertka, x1, _b, x2, _d);
				TopLeft = new PeanoCurve_2D(Left, _razvertka, _a, x1, x2, _d);
				DownLeft = new PeanoCurve_2D(Left, _razvertka, _a, x1, _c, x2);
				DownRight = new PeanoCurve_2D(Top, _razvertka, x1, _b, _c, x2);
			}
		}
	}
}

const __declspec(noalias) __forceinline PeanoCurve_2D* __restrict HitTest_2D(float x) const noexcept
{
	int i = 0;
	const int _razvertka = this->razvertka;
	int num;
	const float this_a = this->a;
	x -= this_a;
	const float b_minus_a = this->b - this_a;
	const float inv_b_minus_a = 1.0f / b_minus_a;
	const PeanoCurve_2D* __restrict Curr = this;
	while (i != _razvertka) {
		const int shift = 1 << ++i + i;
		num = shift * x * inv_b_minus_a;
		x = fmaf(-ldexp(1.0f, -(i << 1)) * num, b_minus_a, x);
		const List currType = Curr->Type;
		switch (num) {
		case 0:
			Curr = (currType == Top || currType == Right)
				? Curr->DownLeft
				: Curr->TopRight;
			break;
		case 1:
			Curr = (currType == Top || currType == Left)
				? Curr->TopLeft
				: Curr->DownRight;
			break;
		case 2:
			Curr = (currType == Top || currType == Right)
				? Curr->TopRight
				: Curr->DownLeft;
			break;
		default:
			Curr = (currType == Top || currType == Left)
				? Curr->DownRight
				: Curr->TopLeft;
		}
	}
	return const_cast<PeanoCurve_2D * __restrict>(Curr);
}

const __declspec(noalias) __forceinline float FindX_2D(const float target_x1, const float target_x2) const noexcept
{
	int _razvertka = this->razvertka;
	int _razvertka1 = _razvertka;
	float x1, x2, x = this->a;
	const float b_minus_a = this->b - x;
	const PeanoCurve_2D* __restrict Curr = this;
	while (_razvertka != 0) {
		const int exponent = _razvertka1 - _razvertka-- << 1;
		x1 = Curr->x1;
		x2 = Curr->x2;
		const List currType = Curr->Type;
		if (target_x1 > x1 && target_x2 > x2) {
			Curr = Curr->TopRight;
			if (currType == Top || currType == Right) {
				x = fmaf(ldexpf(1.0f, -(exponent + 1)), b_minus_a, x);
			}
		}
		else if (target_x1 < x1 && target_x2 > x2) {
			Curr = Curr->TopLeft;
			if (currType == Top || currType == Left) {
				x = fmaf(ldexpf(1.0f, -(exponent + 2)), b_minus_a, x);
			}
			else {
				x = fmaf(ldexpf(1.0f, -exponent) * 0.75f, b_minus_a, x);
			}
		}
		else if (target_x1 < x1 && target_x2 < x2) {
			Curr = Curr->DownLeft;
			if (currType == Down || currType == Left) {
				x = fmaf(ldexpf(1.0f, -(exponent + 1)), b_minus_a, x);
			}
		}
		else {
			Curr = Curr->DownRight;
			if (currType == Top || currType == Left) {
				x = fmaf(ldexpf(1.0f, -exponent) * 0.75f, b_minus_a, x);
			}
			else {
				x = fmaf(ldexpf(1.0f, -(exponent + 2)), b_minus_a, x);
			}
		}
	}
	return x;
}

};

static std::vector<float, tbb::scalable_allocator<float>> Extr;
static std::vector<Interval*, tbb::scalable_allocator<Interval*>> R;
static boost::mpi::environment* __restrict g_env;
static boost::mpi::communicator* __restrict g_world;
static boost::optionalboost::mpi::status probe_status;
static std::aligned_storage_t<sizeof(PeanoCurve_2D), alignof(PeanoCurve_2D)> curveStorage;
static PeanoCurve_2D& __restrict Curve = reinterpret_cast<PeanoCurve_2D & __restrict>(curveStorage);

extern "C" __declspec(dllexport) __declspec(noalias) __forceinline int AgpInit(const int peanoLevel, const float a, const float b, const float c, const float d) noexcept
{
peano_memory_manager.preallocate_for_depth(peanoLevel);
g_env = ::new boost::mpi::environment();
g_world = ::new boost::mpi::communicator();
int rank = g_world->rank();
if (rank) {
::new (&Curve) PeanoCurve_2D(List::Down, peanoLevel, a, b, c, d);
}
else {
::new (&Curve) PeanoCurve_2D(List::Top, peanoLevel, a, b, c, d);
}
return rank;
}

extern "C" __declspec(dllexport) __declspec(noalias) __forceinline void AgpFinalize() noexcept
{
delete g_world;
delete g_env;
}

__declspec(align(16)) struct CrossMsg final {
float s_x1, s_x2;
float e_x1, e_x2;
float Rtop;
template <typename Archive>
__declspec(noalias) __forceinline void serialize(Archive& __restrict const ar, const unsigned int) noexcept { ar& s_x1& s_x2& e_x1& e_x2& Rtop; }
};

extern "C" __declspec(dllexport) __declspec(noalias)
void Base_LNA_1_2_Mer_AGP(
const float N, const float global_iterations,
const float a, const float b, const float c, const float d, const float r,
const bool mode, const float epsilon, const float seed,
float** __restrict out_data, size_t* __restrict out_len) noexcept
{
const int rank = g_world->rank();
const int partner = rank ^ 1;
const float inv_divider = ldexpf(1.0f, -((Curve.razvertka << 1) + 1));
const float x_addition = (b - a) * inv_divider;
const float y_addition = (d - c) * inv_divider;
const float true_start = a + x_addition;
const float true_end = b - x_addition;
float x_Rmax_1, x_Rmax_2;
float initial_length, dmax, threshold_03, inv_threshold_03;
float start_val, best_f, y_Rmax_1, y_Rmax_2;
float Mmax, m;
int schetchick = 0, mcQueenSpeed = 1;
float new_point, new_value;
const PeanoCurve_2D* __restrict pc;
float new_x1, new_x2;
Interval* __restrict promejutochny_otrezok;
Interval* __restrict curr;
Interval* __restrict curr1;
float currM, len1, len2, len_item;
size_t r_size;
float progress, alpha, betta, MULTIPLIER, global_coeff, GLOBAL_FACTOR;
Interval* __restrict top_ptr;
float interval_len;
int dummy;
const PeanoCurve_2D* __restrict p1L;
const PeanoCurve_2D* __restrict p2L;
CrossMsg outbound, inbound;
Interval* __restrict injected;
float cooling;
int T;
float k;

text
if (N == 1.0f) {
	initial_length = b - a;
	dmax = initial_length;
	threshold_03 = 0.3f * initial_length;
	inv_threshold_03 = 1.0f / threshold_03;
	start_val = ShekelFunc(a, seed);
	best_f = ShekelFunc(b, seed);
	x_Rmax_1 = a;
	x_Rmax_2 = b;
	y_Rmax_1 = start_val;
	y_Rmax_2 = best_f;
	Extr.reserve(static_cast<size_t>(global_iterations) << 2u);
	Extr.clear();
	R.reserve(static_cast<size_t>(global_iterations) << 1u);
	R.clear();
	R.emplace_back(new Interval(a, b, start_val, best_f, N));
	std::push_heap(R.begin(), R.end(), ComparePtr);
	Mmax = R.front()->M;
	m = r * Mmax;
	while (true) {
		new_point = Shag(m, x_Rmax_1, x_Rmax_2, y_Rmax_1, y_Rmax_2, N, r);
		new_value = ShekelFunc(new_point, seed);
		if (new_value < best_f) {
			best_f = new_value;
			Extr.emplace_back(best_f);
			Extr.emplace_back(new_point);
		}
		std::pop_heap(R.begin(), R.end(), ComparePtr);
		promejutochny_otrezok = R.back();
		R.pop_back();
		new_x1 = promejutochny_otrezok->x1;
		new_x2 = promejutochny_otrezok->x2;
		len2 = new_x2 - new_point;
		len1 = new_point - new_x1;
		interval_len = (len1 < len2) ? len1 : len2;
		if (interval_len < epsilon || ++schetchick == static_cast<int>(global_iterations)) {
			if (partner) {
				Extr.emplace_back(static_cast<float>(schetchick));
				Extr.emplace_back(interval_len);
				*out_len = Extr.size();
				*out_data = reinterpret_cast<float* __restrict>(CoTaskMemAlloc(sizeof(float) * (*out_len)));
				memcpy(*out_data, Extr.data(), sizeof(float) * (*out_len));
			}
			return;
		}
		curr = new Interval(new_x1, new_point, promejutochny_otrezok->y1, new_value, N);
		curr1 = new Interval(new_point, new_x2, new_value, promejutochny_otrezok->y2, N);
		currM = (curr->M > curr1->M) ? curr->M : curr1->M;
		r_size = R.size();
		if ((len2 + len1) == dmax) {
			dmax = (len2 > len1) ? len2 : len1;
			size_t i = 0u;

#pragma loop(ivdep)
while (i < r_size) {
len_item = R[i]->x2 - R[i]->x1;
if (len_item > dmax) dmax = len_item;
++i;
}
}
if (mode) {
if ((threshold_03 > dmax && fmodf(static_cast<float>(schetchick), 3.0f) == 0.0f) || 10.0f * dmax < initial_length) {
if (currM > Mmax) { Mmax = currM; m = r * Mmax; }
progress = fmaf(-inv_threshold_03, dmax, 1.0f);
alpha = fmaf(progress, progress, 1.0f);
betta = 2.0f - alpha;
MULTIPLIER = (1.0f / dmax) * Mmax;
global_coeff = fmaf(MULTIPLIER, r, -MULTIPLIER);
GLOBAL_FACTOR = betta * global_coeff;
curr->ChangeCharacteristic(fmaf(GLOBAL_FACTOR, len1, curr->M * alpha));
curr1->ChangeCharacteristic(fmaf(GLOBAL_FACTOR, len2, curr1->M * alpha));
size_t i = 0u;
#pragma loop(ivdep)
while (i < r_size) {
R[i]->ChangeCharacteristic(fmaf(GLOBAL_FACTOR, R[i]->x2 - R[i]->x1, R[i]->M * alpha));
++i;
}
std::make_heap(R.begin(), R.end(), ComparePtr);
}
else {
if (currM > Mmax) {
Mmax = currM; m = r * Mmax;
curr->ChangeCharacteristic(m);
curr1->ChangeCharacteristic(m);
size_t i = 0u;
#pragma loop(ivdep)
while (i < r_size) { R[i]->ChangeCharacteristic(m); ++i; }
std::make_heap(R.begin(), R.end(), ComparePtr);
}
else {
curr->ChangeCharacteristic(m);
curr1->ChangeCharacteristic(m);
}
}
}
else {
if (currM > Mmax) {
Mmax = currM; m = r * Mmax;
curr->ChangeCharacteristic(m);
curr1->ChangeCharacteristic(m);
size_t i = 0u;
#pragma loop(ivdep)
while (i < r_size) { R[i]->ChangeCharacteristic(m); ++i; }
std::make_heap(R.begin(), R.end(), ComparePtr);
}
else {
curr->ChangeCharacteristic(m);
curr1->ChangeCharacteristic(m);
}
}
R.emplace_back(curr);
std::push_heap(R.begin(), R.end(), ComparePtr);
R.emplace_back(curr1);
std::push_heap(R.begin(), R.end(), ComparePtr);
top_ptr = R.front();
x_Rmax_1 = top_ptr->x1;
x_Rmax_2 = top_ptr->x2;
y_Rmax_1 = top_ptr->y1;
y_Rmax_2 = top_ptr->y2;
}
}
else {
x_Rmax_1 = true_start;
x_Rmax_2 = true_end;
initial_length = true_end - true_start;
dmax = initial_length;
threshold_03 = 0.3f * initial_length;
inv_threshold_03 = 1.0f / threshold_03;
start_val = rank ? RastriginFunc(true_end, d - y_addition)
: RastriginFunc(true_start, c + y_addition);
y_Rmax_1 = start_val;
best_f = rank ? RastriginFunc(true_start, d - y_addition)
: RastriginFunc(true_end, c + y_addition);
y_Rmax_2 = best_f;
Extr.reserve(static_cast<size_t>(global_iterations) << 2u);
Extr.clear();
R.reserve(static_cast<size_t>(global_iterations) << 1u);
R.clear();
R.emplace_back(new Interval(true_start, true_end, start_val, best_f, N));
Mmax = R.front()->M;
m = r * Mmax;
while (true) {
probe_status = g_world->iprobe(partner, 2);
if (probe_status) {
if (partner) {
Extr.emplace_back(static_cast<float>(schetchick));
Extr.emplace_back(interval_len);
*out_len = Extr.size();
out_data = reinterpret_cast<float __restrict>(CoTaskMemAlloc(sizeof(float) * (*out_len)));
memcpy(*out_data, Extr.data(), sizeof(float) * (*out_len));
}
g_world->recv(partner, 2, dummy);
return;
}
cooling = ldexpf(1.0f, -(1.0f / 138.63f) * ++schetchick);
T = static_cast<int>(fmaf(20.0f, cooling, 10.0f));
k = fmaf(0.2f, cooling, 0.7f);
if (schetchick % (T >> 1) == 0) {
probe_status = g_world->iprobe(partner, 0);
if (probe_status) {
g_world->recv(partner, 0, mcQueenSpeed);
}
}
if (!(schetchick % T) == mcQueenSpeed) {
p1L = Curve.HitTest_2D(top_ptr->x1);
p2L = Curve.HitTest_2D(top_ptr->x2);
outbound = CrossMsg{ p1L->x1, p1L->x2, p2L->x1, p2L->x2, top_ptr->R };
if (mcQueenSpeed) {
g_world->send(partner, 0, 0);
}
else {
mcQueenSpeed = 2;
}
g_world->sendrecv(partner, 1, outbound, partner, 1, inbound);
injected = new Interval(Curve.FindX_2D(inbound.s_x1, inbound.s_x2), Curve.FindX_2D(inbound.e_x1, inbound.e_x2), RastriginFunc(inbound.s_x1, inbound.s_x2), RastriginFunc(inbound.e_x1, inbound.e_x2), N);
injected->R = inbound.Rtop * k;
R.emplace_back(injected);
std::push_heap(R.begin(), R.end(), ComparePtr);
}
new_point = Shag(m, x_Rmax_1, x_Rmax_2, y_Rmax_1, y_Rmax_2, N, r);
pc = Curve.HitTest_2D(new_point);
new_x1 = pc->x1;
new_x2 = pc->x2;
new_value = RastriginFunc(new_x1, new_x2);
if (new_value < best_f) {
best_f = new_value;
Extr.emplace_back(best_f);
Extr.emplace_back(new_x1);
Extr.emplace_back(new_x2);
}
std::pop_heap(R.begin(), R.end(), ComparePtr);
promejutochny_otrezok = R.back();
R.pop_back();
curr = new Interval(promejutochny_otrezok->x1, new_point, promejutochny_otrezok->y1, new_value, N);
curr1 = new Interval(new_point, promejutochny_otrezok->x2, new_value, promejutochny_otrezok->y2, N);
currM = (curr->M > curr1->M) ? curr->M : curr1->M;
len2 = promejutochny_otrezok->x2 - new_point;
len1 = new_point - promejutochny_otrezok->x1;
r_size = R.size();
if ((len2 + len1) == dmax) {
dmax = (len2 > len1) ? len2 : len1;
size_t i = 0u;
#pragma loop(ivdep)
while (i < r_size) {
len_item = R[i]->x2 - R[i]->x1;
if (len_item > dmax) dmax = len_item;
++i;
}
}
if (mode) {
if ((threshold_03 > dmax && fmodf(static_cast<float>(schetchick), 3.0f) == 0.0f) || 10.0f * dmax < initial_length) {
if (currM > Mmax) { Mmax = currM; m = r * Mmax; }
progress = fmaf(-inv_threshold_03, dmax, 1.0f);
alpha = fmaf(progress, progress, 1.0f);
betta = 2.0f - alpha;
MULTIPLIER = (1.0f / dmax) * Mmax;
global_coeff = fmaf(MULTIPLIER, r, -MULTIPLIER);
GLOBAL_FACTOR = betta * global_coeff;
curr->ChangeCharacteristic(fmaf(GLOBAL_FACTOR, len1, curr->M * alpha));
curr1->ChangeCharacteristic(fmaf(GLOBAL_FACTOR, len2, curr1->M * alpha));
size_t i = 0u;
#pragma loop(ivdep)
while (i < r_size) {
R[i]->ChangeCharacteristic(fmaf(GLOBAL_FACTOR, R[i]->x2 - R[i]->x1, R[i]->M * alpha));
++i;
}
std::make_heap(R.begin(), R.end(), ComparePtr);
}
else {
if (currM > Mmax) {
Mmax = currM; m = r * Mmax;
curr->ChangeCharacteristic(m);
curr1->ChangeCharacteristic(m);
size_t i = 0u;
#pragma loop(ivdep)
while (i < r_size) { R[i]->ChangeCharacteristic(m); ++i; }
std::make_heap(R.begin(), R.end(), ComparePtr);
}
else {
curr->ChangeCharacteristic(m);
curr1->ChangeCharacteristic(m);
}
}
}
else {
if (currM > Mmax) {
Mmax = currM; m = r * Mmax;
curr->ChangeCharacteristic(m);
curr1->ChangeCharacteristic(m);
size_t i = 0u;
#pragma loop(ivdep)
while (i < r_size) { R[i]->ChangeCharacteristic(m); ++i; }
std::make_heap(R.begin(), R.end(), ComparePtr);
}
else {
curr->ChangeCharacteristic(m);
curr1->ChangeCharacteristic(m);
}
}
R.emplace_back(curr);
std::push_heap(R.begin(), R.end(), ComparePtr);
R.emplace_back(curr1);
std::push_heap(R.begin(), R.end(), ComparePtr);
top_ptr = R.front();
x_Rmax_1 = top_ptr->x1;
x_Rmax_2 = top_ptr->x2;
y_Rmax_1 = top_ptr->y1;
y_Rmax_2 = top_ptr->y2;
interval_len = x_Rmax_2 - x_Rmax_1;
if (interval_len < epsilon || schetchick == static_cast<int>(global_iterations)) {
if (partner) {
Extr.emplace_back(static_cast<float>(schetchick));
Extr.emplace_back(interval_len);
*out_len = Extr.size();
out_data = reinterpret_cast<float __restrict>(CoTaskMemAlloc(sizeof(float) * (*out_len)));
memcpy(*out_data, Extr.data(), sizeof(float) * (*out_len));
}
g_world->send(partner, 2, 0);
return;
}
}
}
}

extern "C" __declspec(dllexport) __declspec(noalias) __forceinline void Base_LNA_1_2_Mer_AGP_Free(float* __restrict p) noexcept
{
CoTaskMemFree(p);
}

extern "C" __declspec(dllexport) __declspec(noalias) __forceinline void AgpStartWorkers() noexcept
{
g_world->send(1, 3, 1);
}

extern "C" __declspec(dllexport) __declspec(noalias) __forceinline void AgpStopWorkers() noexcept
{
g_world->send(1, 3, 2);
}

extern "C" __declspec(dllexport) __declspec(noalias) __forceinline void AgpWaitStartAndRun() noexcept
{
int command;
float* __restrict buf;
size_t len;
while (true) {
command = 0;
probe_status = g_world->iprobe(0, 3);
if (probe_status) {
g_world->recv(0, 3, command);
if (command == 1) {
Base_LNA_1_2_Mer_AGP(2.0f, 1000.0f, -2.2f, 1.8f, -2.2f, 1.8f,
2.5f, false, 0.001f, GetTickCount(),
&buf, &len);
}
else {
return;
}
}
}
} это мой код для поиска минимума, рассмотри подробно двумерную ветку в частности логику MPI пересылок, в частности логику кросс-инъекций, сейчас есть такая проблема что если бы я убрал это условие что принятие сообщения о кросс-инъекции происходит только по определённым итерациям счётчика и разрешил бы принимать сообщения о кросс-инъекциях сразу как они будут доступны то возникла бы такая проблема что оба процесса примерно одновременно приходили бы к участку отправки сообщения на завершения счёта что приводило к deadlock, это происходило потому что один процесс по счёту сильно опережал другой и когда у более быстрого процесса уже было много поисковой информации и он отправлял сообщение процессу у которого её меньше то медленный процесс мгновенно цеплялся за лучший интервал который уже близок к оптимальному и найден более быстрым процессом - и завешал счёт - это неплохо такое быстрое завершение счёта - проблема лишь в том что это приводило к дэдлоку и я был вынужден написать условие if (schetchick % (T >> 1) == 0) {, также в целом проанализируй код - как его ещё можно ускорить, не пиши про if (!(schetchick % T) == mcQueenSpeed) { это условие полностью корректно это моя разработка работающая похожим на мьютекс образом здесь

مشاركة هذا السؤال والجواب