Skip to content

Fix: add runtime check for cudaHostRegister in EXX PW #7157

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
mohanchen merged 5 commits into from
Mar 28, 2026
+22 −10
Merged

Fix: add runtime check for cudaHostRegister in EXX PW #7157

Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
c4f751d
fix: add runtime check for cudaHostRegister in EXX PW
Flying-dragon-boxing Mar 28, 2026
103cb69
Merge branch 'develop' into fix/exx_pw_cuda_host_register
Flying-dragon-boxing Mar 28, 2026
665e5a4
Update source/source_pw/module_pwdft/op_pw_exx_pot.cpp
Flying-dragon-boxing Mar 28, 2026
cd5b5c8
Update source/source_pw/module_pwdft/op_pw_exx_pot.cpp
Flying-dragon-boxing Mar 28, 2026
e13617b
Merge branch 'develop' into fix/exx_pw_cuda_host_register
Flying-dragon-boxing Mar 28, 2026
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
32 changes: 22 additions & 10 deletions source/source_pw/module_pwdft/op_pw_exx_pot.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -192,17 +192,23 @@ void get_exx_potential(const K_Vectors* kv,

// copy the potential to the device memory
#ifdef __CUDA
cudaError_t err = cudaHostRegister(pot_cpu, sizeof(Real) * npw, cudaHostRegisterPortable);
if (err != cudaSuccess) {
throw std::runtime_error("failed to register potential CPU memory operations");
if (PARAM.inp.device == "gpu")
{
cudaError_t err = cudaHostRegister(pot_cpu, sizeof(Real) * npw, cudaHostRegisterPortable);
if (err != cudaSuccess) {
throw std::runtime_error("failed to register potential CPU memory operations");
}
}
#endif
syncmem_real_c2d_op()(pot, pot_cpu, rhopw_dev->npw);
#ifdef __CUDA
cudaHostUnregister(pot_cpu);
if (PARAM.inp.device == "gpu")
{
cudaHostUnregister(pot_cpu);
}
#endif

delete pot_cpu;
delete[] pot_cpu;
}

template <typename Real, typename Device>
Expand Down Expand Up @@ -369,17 +375,23 @@ void get_exx_stress_potential(const K_Vectors* kv,

// copy the potential to the device memory
#ifdef __CUDA
cudaError_t err = cudaHostRegister(pot_cpu, sizeof(Real) * npw, cudaHostRegisterPortable);
if (err != cudaSuccess) {
throw std::runtime_error("failed to register potential CPU memory operations");
if (PARAM.inp.device == "gpu")
{
cudaError_t err = cudaHostRegister(pot_cpu, sizeof(Real) * npw, cudaHostRegisterPortable);
if (err != cudaSuccess) {
throw std::runtime_error("failed to register potential CPU memory operations");
}
}
#endif
syncmem_real_c2d_op()(pot, pot_cpu, rhopw_dev->npw);
#ifdef __CUDA
cudaHostUnregister(pot_cpu);
if (PARAM.inp.device == "gpu")
{
cudaHostUnregister(pot_cpu);
}
#endif

delete pot_cpu;
delete[] pot_cpu;
}

double exx_divergence(Conv_Coulomb_Pot_K::Coulomb_Type coulomb_type,
Expand Down
Loading