Skip to content

Fix max nbor size related issues #3157

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
njzjz merged 6 commits into from
Jan 23, 2024
Merged

Fix max nbor size related issues #3157

Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
9c95657
fix max_nbor_size related issues
denghuilu Jan 20, 2024
ed213cb
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Jan 20, 2024
5f6b2a9
add UTs
denghuilu Jan 22, 2024
5a1466b
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Jan 22, 2024
1b7965c
Revert "[pre-commit.ci] auto fixes from pre-commit.com hooks"
njzjz Jan 23, 2024
8f66c8e
Revert "add UTs"
njzjz Jan 23, 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
29 changes: 29 additions & 0 deletions source/op/prod_env_mat_multi_device.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -577,6 +577,15 @@ class ProdEnvMatAOp : public OpKernel {
mesh_tensor.flat<int>().data(), mesh_tensor_size, nloc, nei_mode,
rcut_r, max_cpy_trial, max_nnei_trial);

// max_nbor_size may be changed after _prepare_coord_nlist_gpu
// So we need to update the uint64_temp tensor if necessary
if (uint64_temp.NumElements() < int_64(nloc) * max_nbor_size * 2) {
TensorShape uint64_shape;
uint64_shape.AddDim(int_64(nloc) * max_nbor_size * 2);
OP_REQUIRES_OK(context, context->allocate_temp(
DT_UINT64, uint64_shape, &uint64_temp));
array_longlong = uint64_temp.flat<unsigned long long>().data();
}
// launch the gpu(nv) compute function
deepmd::prod_env_mat_a_gpu(em, em_deriv, rij, nlist, coord, type,
gpu_inlist, array_int, array_longlong,
Expand Down Expand Up @@ -875,6 +884,16 @@ class ProdEnvMatROp : public OpKernel {
mesh_tensor.flat<int>().data(), mesh_tensor_size, nloc, nei_mode,
rcut, max_cpy_trial, max_nnei_trial);

// max_nbor_size may be changed after _prepare_coord_nlist_gpu
// So we need to update the uint64_temp tensor if necessary
if (uint64_temp.NumElements() < int_64(nloc) * max_nbor_size * 2) {
TensorShape uint64_shape;
uint64_shape.AddDim(int_64(nloc) * max_nbor_size * 2);
OP_REQUIRES_OK(context, context->allocate_temp(
DT_UINT64, uint64_shape, &uint64_temp));
array_longlong = uint64_temp.flat<unsigned long long>().data();
}

// launch the gpu(nv) compute function
deepmd::prod_env_mat_r_gpu(em, em_deriv, rij, nlist, coord, type,
gpu_inlist, array_int, array_longlong,
Expand Down Expand Up @@ -1221,6 +1240,16 @@ class ProdEnvMatAMixOp : public OpKernel {
mesh_tensor.flat<int>().data(), mesh_tensor_size, nloc, nei_mode,
rcut_r, max_cpy_trial, max_nnei_trial);

// max_nbor_size may be changed after _prepare_coord_nlist_gpu
// So we need to update the uint64_temp tensor if necessary
if (uint64_temp.NumElements() < int_64(nloc) * max_nbor_size * 2) {
TensorShape uint64_shape;
uint64_shape.AddDim(int_64(nloc) * max_nbor_size * 2);
OP_REQUIRES_OK(context, context->allocate_temp(
DT_UINT64, uint64_shape, &uint64_temp));
array_longlong = uint64_temp.flat<unsigned long long>().data();
}

// launch the gpu(nv) compute function
deepmd::prod_env_mat_a_gpu(em, em_deriv, rij, nlist, coord, type,
gpu_inlist, array_int, array_longlong,
Expand Down