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

Jump to bottom

Move sterf to CPU; Add experimental parallelism for sterf #462

Open
wants to merge 3 commits into
base: develop
Choose a base branch
from
Open

Move sterf to CPU; Add experimental parallelism for sterf #462

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
def0e2a
Move sterf to CPU; add experimental parallelism for sterf
Aug 22, 2022
20c9487
Add dsterf calling, minor fixes
mdvizov Sep 1, 2022
4c0203d
Off LAPACK flag; Minor fixes
mdvizov Sep 7, 2022
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: 2 additions & 0 deletions CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,6 +33,8 @@ set(CMAKE_CXX_STANDARD_REQUIRED ON)

option(ROCSOLVER_EMBED_FMT "Hide libfmt symbols" OFF)
option(OPTIMAL "Build specialized kernels for small matrix sizes" ON)
option(HYBRID_CPU "Build hybrid schema with CPU using" ON)
option(EXPERIMENTAL "Experimental parallelization" OFF)
option(ROCSOLVER_FIND_PACKAGE_LAPACK_CONFIG "Skip module mode search for LAPACK" ON)

# Add our CMake helper files to the lookup path
Expand Down
9 changes: 9 additions & 0 deletions library/src/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -290,6 +290,15 @@ if(OPTIMAL)
target_compile_definitions(rocsolver PRIVATE OPTIMAL)
endif()

if(HYBRID_CPU)
target_compile_definitions(rocsolver PRIVATE HYBRID_CPU)
endif()

if(EXPERIMENTAL)
target_compile_definitions(rocsolver PRIVATE EXPERIMENTAL)
endif()


target_compile_definitions(rocsolver PRIVATE
ROCM_USE_FLOAT16
ROCBLAS_INTERNAL_API
Expand Down
21 changes: 21 additions & 0 deletions library/src/auxiliary/rocauxiliary_sterf.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,26 @@ rocblas_status
rocblas_stride strideE = 0;
rocblas_int batch_count = 1;

#ifdef EXPERIMENTAL
// additional memory for internal kernels (parallel sterf)
size_t size_ranges;
rocsolver_sterf_parallel_getMemorySize<T>(n, &size_ranges);

if(rocblas_is_device_memory_size_query(handle))
return rocblas_set_optimal_device_memory_size(handle, size_ranges);

// memory workspace allocation
void* ranges;
rocblas_device_malloc mem_range(handle, size_ranges);
if(!mem_range)
return rocblas_status_memory_error;

ranges = mem_range[0];

// execution
return rocsolver_sterf_template<T>(handle, n, D, shiftD, strideD, E, shiftE, strideE, info,
batch_count, (rocblas_int*)ranges);
#else
// memory workspace sizes:
// size for lasrt stack
size_t size_stack;
Expand All @@ -46,6 +66,7 @@ rocblas_status
// execution
return rocsolver_sterf_template<T>(handle, n, D, shiftD, strideD, E, shiftE, strideE, info,
batch_count, (rocblas_int*)stack);
#endif
}

/*
Expand Down
Loading