Skip to content
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

Allow passing smart pointers to functions directly (interface changes only) #1279

Merged
merged 9 commits into from
Feb 22, 2023
Merged

Allow passing smart pointers to functions directly (interface changes only) #1279

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
f9b1ae1
add pointer_param
upsj Jan 30, 2023
1b38812
review updates and interface improvements
upsj Feb 3, 2023
57a5ee7
make MPI find_package call more consistent
upsj Jan 31, 2023
8158573
review updates interface
upsj Feb 16, 2023
4f76a58
add MPI to README
upsj Feb 16, 2023
2893d02
improve documentation
upsj Feb 16, 2023
1d95a95
rename pointer_param to ptr_param
upsj Feb 20, 2023
9334b77
rename local_matrix_type to local_matrix_template
upsj Feb 20, 2023
14b7f69
use pointee where possible
upsj Feb 20, 2023
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
2 changes: 1 addition & 1 deletion CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -234,7 +234,7 @@ endif()
set(GINKGO_HAVE_GPU_AWARE_MPI OFF)
set(GINKGO_FORCE_SPMV_BLOCKING_COMM OFF)
if(GINKGO_BUILD_MPI)
find_package(MPI REQUIRED)
find_package(MPI 3.1 COMPONENTS CXX REQUIRED)
if(GINKGO_FORCE_GPU_AWARE_MPI)
set(GINKGO_HAVE_GPU_AWARE_MPI ON)
else()
Expand Down
4 changes: 4 additions & 0 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -70,6 +70,10 @@ The Ginkgo DPC++ module has the following __additional__ requirements:
* oneMKL
* oneDPL

The Ginkgo MPI module has the following __additional__ requirements:

* MPI 3.1+, ideally with GPUDirect support for best performance

In addition, if you want to contribute code to Ginkgo, you will also need the
following:

Expand Down
2 changes: 1 addition & 1 deletion cmake/GinkgoConfig.cmake.in
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -161,7 +161,7 @@ endif()

# Check for MPI if it is enabled
if(GINKGO_BUILD_MPI)
find_package(MPI REQUIRED)
find_package(MPI 3.1 COMPONENTS CXX REQUIRED)
endif()

# HIP and OpenMP depend on Threads::Threads in some circumstances, but don't find it
Expand Down
2 changes: 1 addition & 1 deletion cmake/autodetect_executors.cmake
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@ if (NOT DEFINED GINKGO_BUILD_OMP)
endif()

if (NOT DEFINED GINKGO_BUILD_MPI)
find_package(MPI 3.1)
find_package(MPI 3.1 COMPONENTS CXX)
if(MPI_FOUND)
message(STATUS "Enabling MPI support")
set(GINKGO_HAS_MPI ON)
Expand Down
2 changes: 1 addition & 1 deletion cmake/create_test.cmake
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,7 +14,7 @@ function(ginkgo_create_gtest_mpi_main)
target_sources(gtest_mpi_main
PRIVATE
${PROJECT_SOURCE_DIR}/core/test/mpi/gtest/mpi_listener.cpp)
find_package(MPI REQUIRED)
find_package(MPI 3.1 COMPONENTS CXX REQUIRED)
target_link_libraries(gtest_mpi_main PRIVATE GTest::GTest MPI::MPI_CXX)
endfunction(ginkgo_create_gtest_mpi_main)

Expand Down
23 changes: 14 additions & 9 deletions core/distributed/matrix.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,15 +66,16 @@ Matrix<ValueType, LocalIndexType, GlobalIndexType>::Matrix(
template <typename ValueType, typename LocalIndexType, typename GlobalIndexType>
Matrix<ValueType, LocalIndexType, GlobalIndexType>::Matrix(
std::shared_ptr<const Executor> exec, mpi::communicator comm,
const LinOp* local_matrix_type)
: Matrix(exec, comm, local_matrix_type, local_matrix_type)
ptr_param<const LinOp> local_matrix_template)
: Matrix(exec, comm, local_matrix_template, local_matrix_template)
{}


template <typename ValueType, typename LocalIndexType, typename GlobalIndexType>
Matrix<ValueType, LocalIndexType, GlobalIndexType>::Matrix(
std::shared_ptr<const Executor> exec, mpi::communicator comm,
const LinOp* local_matrix_template, const LinOp* non_local_matrix_template)
ptr_param<const LinOp> local_matrix_template,
ptr_param<const LinOp> non_local_matrix_template)
: EnableDistributedLinOp<
Matrix<value_type, local_index_type, global_index_type>>{exec},
DistributedBase{comm},
Expand Down Expand Up @@ -141,8 +142,10 @@ void Matrix<ValueType, LocalIndexType, GlobalIndexType>::move_to(
template <typename ValueType, typename LocalIndexType, typename GlobalIndexType>
void Matrix<ValueType, LocalIndexType, GlobalIndexType>::read_distributed(
const device_matrix_data<value_type, global_index_type>& data,
const Partition<local_index_type, global_index_type>* row_partition,
const Partition<local_index_type, global_index_type>* col_partition)
ptr_param<const Partition<local_index_type, global_index_type>>
row_partition,
ptr_param<const Partition<local_index_type, global_index_type>>
col_partition)
{
const auto comm = this->get_communicator();
GKO_ASSERT_EQ(data.get_size()[0], row_partition->get_size());
Expand Down Expand Up @@ -228,8 +231,10 @@ void Matrix<ValueType, LocalIndexType, GlobalIndexType>::read_distributed(
template <typename ValueType, typename LocalIndexType, typename GlobalIndexType>
void Matrix<ValueType, LocalIndexType, GlobalIndexType>::read_distributed(
const matrix_data<value_type, global_index_type>& data,
const Partition<local_index_type, global_index_type>* row_partition,
const Partition<local_index_type, global_index_type>* col_partition)
ptr_param<const Partition<local_index_type, global_index_type>>
row_partition,
ptr_param<const Partition<local_index_type, global_index_type>>
col_partition)
{
this->read_distributed(
device_matrix_data<value_type, global_index_type>::create_from_host(
Expand All @@ -241,7 +246,7 @@ void Matrix<ValueType, LocalIndexType, GlobalIndexType>::read_distributed(
template <typename ValueType, typename LocalIndexType, typename GlobalIndexType>
void Matrix<ValueType, LocalIndexType, GlobalIndexType>::read_distributed(
const matrix_data<ValueType, global_index_type>& data,
const Partition<local_index_type, global_index_type>* partition)
ptr_param<const Partition<local_index_type, global_index_type>> partition)
{
this->read_distributed(
device_matrix_data<value_type, global_index_type>::create_from_host(
Expand All @@ -253,7 +258,7 @@ void Matrix<ValueType, LocalIndexType, GlobalIndexType>::read_distributed(
template <typename ValueType, typename LocalIndexType, typename GlobalIndexType>
void Matrix<ValueType, LocalIndexType, GlobalIndexType>::read_distributed(
const device_matrix_data<ValueType, GlobalIndexType>& data,
const Partition<local_index_type, global_index_type>* partition)
ptr_param<const Partition<local_index_type, global_index_type>> partition)
{
this->read_distributed(data, partition, partition);
}
Expand Down
Loading