Skip to content

Fix the overwriting between xc kernel components #5654

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 7 commits into from
Dec 5, 2024
Merged

Fix the overwriting between xc kernel components #5654

Changes from 1 commit
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
377b9aa
fix the overwriting between xc kernel components
maki49 Nov 30, 2024
8fbdead
[pre-commit.ci lite] apply automatic fixes
pre-commit-ci-lite[bot] Nov 30, 2024
cf7d9de
use for-loop with omp
maki49 Dec 1, 2024
7855e11
add duration test
maki49 Dec 3, 2024
f4e5635
replace lambda by individual function
maki49 Dec 3, 2024
db170d4
expand the for-loop
maki49 Dec 3, 2024
ca1ab07
small change
maki49 Dec 3, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
45 changes: 30 additions & 15 deletions source/module_lr/potentials/xc_kernel.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -70,6 +70,29 @@ LR::KernelXC::KernelXC(const ModulePW::PW_Basis& rho_basis,
#endif
}

template <typename T>
inline void add_op(const T* const src1, const T* const src2, T* const dst, const int size)
{
#ifdef _OPENMP
#pragma omp parallel for schedule(static, 4096)
#endif
for (int i = 0;i < size;++i)
{
dst[i] = src1[i] + src2[i];
}
}
template <typename T>
inline void add_op(const std::vector<T>& src1, const std::vector<T>& src2, std::vector<T>& dst)
{
assert(dst.size() >= src1.size() && src2.size() >= src1.size());
add_op(src1.data(), src2.data(), dst.data(), src1.size());
}
template <typename T>
inline void add_assign_op(const std::vector<T>& src, std::vector<T>& dst)
{
add_op(src, dst, dst);
}

#ifdef USE_LIBXC
void LR::KernelXC::f_xc_libxc(const int& nspin, const double& omega, const double& tpiba, const double* const* const rho_gs, const double* const rho_core)
{
Expand Down Expand Up @@ -192,20 +215,12 @@ void LR::KernelXC::f_xc_libxc(const int& nspin, const double& omega, const doubl
break;
}
// add onto the total components
auto omp_add_vector = [](const std::vector<double>& src, std::vector<double>& dst)
{
assert(src.size() == dst.size());
#ifdef _OPENMP
#pragma omp parallel for schedule(static, 4096)
#endif
for (size_t i = 0; i < src.size(); ++i) { dst[i] += src[i]; }
};
// auto start = std::chrono::high_resolution_clock::now();
omp_add_vector(vrho_tmp, this->vrho_);
omp_add_vector(v2rho2_tmp, this->v2rho2_);
omp_add_vector(vsigma_tmp, this->vsigma_);
omp_add_vector(v2rhosigma_tmp, this->v2rhosigma_);
omp_add_vector(v2sigma2_tmp, this->v2sigma2_);
add_assign_op(vrho_tmp, this->vrho_);
add_assign_op(v2rho2_tmp, this->v2rho2_);
add_assign_op(vsigma_tmp, this->vsigma_);
add_assign_op(v2rhosigma_tmp, this->v2rhosigma_);
add_assign_op(v2sigma2_tmp, this->v2sigma2_);
// auto end = std::chrono::high_resolution_clock::now();
// auto duration = std::chrono::duration_cast<std::chrono::milliseconds>(end - start);
// std::cout << "Time elapsed adding XC components: " << duration.count() << " ms\n";
Expand All @@ -229,14 +244,14 @@ void LR::KernelXC::f_xc_libxc(const int& nspin, const double& omega, const doubl
// 1. $2f^{\rho\sigma}*\nabla\rho$
this->v2rhosigma_2drho_.resize(nrxx);
#ifdef _OPENMP
#pragma omp parallel for
#pragma omp parallel for schedule(static, 4096)
#endif
for (size_t i = 0; i < nrxx; ++i) { this->v2rhosigma_2drho_[i] = gradrho[0][i] * v2rs[i] * 2.; }

// 2. $4f^{\sigma\sigma}*\nabla\rho$
this->v2sigma2_4drho_.resize(nrxx);
#ifdef _OPENMP
#pragma omp parallel for
#pragma omp parallel for schedule(static, 4096)
#endif
for (size_t i = 0; i < nrxx; ++i) { this->v2sigma2_4drho_[i] = gradrho[0][i] * v2s2[i] * 4.; }
}
Expand Down
Loading