Skip to content

Matrix Transpose Tutorial Cleanup #1917

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

Open
MrBurmark wants to merge 19 commits into
base: develop
Choose a base branch
from
+219 −161
Open

Matrix Transpose Tutorial Cleanup #1917

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
19 commits
Select commit Hold shift + click to select a range
3ac295a
Fix indenting
MrBurmark Sep 21, 2025
8060a32
Use unchecked policies
MrBurmark Sep 21, 2025
d0c36ce
Fix team synchronization
MrBurmark Sep 21, 2025
1e1d347
Fix device synchronization
MrBurmark Sep 21, 2025
410f2bb
Remap launch thread mapping
MrBurmark Sep 21, 2025
e5d6004
clean up pass for launch
artv3 Sep 21, 2025
93b5a2a
use row/col_t
artv3 Sep 22, 2025
ec57de9
missed updating variable names
artv3 Sep 22, 2025
53346ca
Merge branch 'develop' into bugfix/burmark1/transpose
artv3 Sep 30, 2025
72cf1c9
Merge branch 'develop' into bugfix/burmark1/transpose
artv3 Oct 1, 2025
258fea0
Merge branch 'develop' into bugfix/burmark1/transpose
artv3 Oct 2, 2025
4ed8196
fix tranpose c example
artv3 Oct 2, 2025
a9e8dc4
Merge branch 'bugfix/burmark1/transpose' of github.com:LLNL/RAJA into…
artv3 Oct 2, 2025
586b594
Merge branch 'develop' into bugfix/burmark1/transpose
rhornung67 Nov 3, 2025
6f489d4
Merge branch 'develop' into bugfix/burmark1/transpose
rhornung67 Dec 1, 2025
627f418
Fix compilation errors
rhornung67 Dec 10, 2025
9f9e97d
Merge branch 'develop' into bugfix/burmark1/transpose
artv3 Apr 2, 2026
44eb1c7
clean up the code
artv3 Apr 2, 2026
59560b3
Merge branch 'develop' into bugfix/burmark1/transpose
artv3 Apr 2, 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
7 changes: 4 additions & 3 deletions exercises/kernel-matrix-transpose-local-array.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -156,11 +156,12 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
for (int tx = 0; tx < TILE_DIM; ++tx) {
for (int ty = 0; ty < TILE_DIM; ++ty) {

int col = bx * TILE_DIM + tx; // Matrix column index
int row = by * TILE_DIM + ty; // Matrix row index
// Tranpose tile offset
int col = by * TILE_DIM + tx; // Matrix column index
int row = bx * TILE_DIM + ty; // Matrix row index

// Bounds check
if (row < N_r && col < N_c) {
if (row < N_c && col < N_r) {
Atview(col, row) = Tile[ty][tx];
}
}
Expand Down
149 changes: 86 additions & 63 deletions exercises/kernel-matrix-transpose-local-array_solution.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -81,6 +81,7 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
constexpr int N_c = 251;

constexpr int TILE_DIM = 16;
constexpr int BLOCK_ROWS = 8;

constexpr int outer_Dimc = (N_c - 1) / TILE_DIM + 1;
constexpr int outer_Dimr = (N_r - 1) / TILE_DIM + 1;
Expand Down Expand Up @@ -153,15 +154,16 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
// Note: loop order is swapped from above so that output matrix
// data access is stride-1.
//
for (int tx = 0; tx < TILE_DIM; ++tx) {
for (int ty = 0; ty < TILE_DIM; ++ty) {
for (int ty = 0; ty < TILE_DIM; ++ty) {
for (int tx = 0; tx < TILE_DIM; ++tx) {

int col = bx * TILE_DIM + tx; // Matrix column index
int row = by * TILE_DIM + ty; // Matrix row index
// Tranpose tile offset
int col_t = by * TILE_DIM + tx; // Matrix column index
int row_t = bx * TILE_DIM + ty; // Matrix row index

// Bounds check
if (row < N_r && col < N_c) {
Atview(col, row) = Tile[ty][tx];
if (row_t < N_c && col_t < N_r) {
Atview(row_t, col_t) = Tile[tx][ty];
}
}
}
Expand Down Expand Up @@ -191,6 +193,9 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
using TILE_MEM =
RAJA::LocalArray<int, RAJA::Perm<0, 1>, RAJA::SizeList<TILE_DIM, TILE_DIM>>;
TILE_MEM Tile_Array;
using THREAD_ROWS = RAJA::TypedRangeSegment<int>;
using RAJA::Params;
using RAJA::Segs;
// _mattranspose_localarray_end

// **NOTE** Although the LocalArray is constructed
Expand All @@ -209,18 +214,17 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))

RAJA::statement::InitLocalMem<RAJA::cpu_tile_mem, RAJA::ParamList<2>,

RAJA::statement::ForICount<1, RAJA::statement::Param<0>, RAJA::seq_exec,
RAJA::statement::ForICount<0, RAJA::statement::Param<1>, RAJA::seq_exec,
RAJA::statement::ForICount<1, RAJA::statement::Param<1>, RAJA::seq_exec,
RAJA::statement::ForICount<0, RAJA::statement::Param<0>, RAJA::seq_exec,
RAJA::statement::Lambda<0>
>
>,

RAJA::statement::ForICount<0, RAJA::statement::Param<1>, RAJA::seq_exec,
RAJA::statement::ForICount<1, RAJA::statement::Param<0>, RAJA::seq_exec,
RAJA::statement::ForICount<0, RAJA::statement::Param<0>, RAJA::seq_exec,
RAJA::statement::ForICount<1, RAJA::statement::Param<1>, RAJA::seq_exec,
RAJA::statement::Lambda<1>
>
>

>
>
>
Expand Down Expand Up @@ -273,8 +277,8 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
// loops. These loops copy data from the global matrices
// to the local tile.
//
RAJA::statement::ForICount<1, RAJA::statement::Param<0>, RAJA::seq_exec,
RAJA::statement::ForICount<0, RAJA::statement::Param<1>, RAJA::seq_exec,
RAJA::statement::ForICount<1, RAJA::statement::Param<1>, RAJA::seq_exec,
RAJA::statement::ForICount<0, RAJA::statement::Param<0>, RAJA::seq_exec,
RAJA::statement::Lambda<0>
>
>,
Expand All @@ -286,8 +290,8 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
// swapped! This enables us to swap which
// index has unit stride.
//
RAJA::statement::ForICount<0, RAJA::statement::Param<1>, RAJA::seq_exec,
RAJA::statement::ForICount<1, RAJA::statement::Param<0>, RAJA::seq_exec,
RAJA::statement::ForICount<0, RAJA::statement::Param<0>, RAJA::seq_exec,
RAJA::statement::ForICount<1, RAJA::statement::Param<1>, RAJA::seq_exec,
RAJA::statement::Lambda<1>
>
>
Expand All @@ -302,13 +306,11 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
RAJA::make_tuple((int)0, (int)0, Tile_Array),

[=](int col, int row, int tx, int ty, TILE_MEM &_Tile_Array) {

_Tile_Array(ty, tx) = Aview(row, col);

},

[=](int col, int row, int tx, int ty, TILE_MEM &_Tile_Array) {

Atview(col, row) = _Tile_Array(ty, tx);

}
Expand Down Expand Up @@ -401,21 +403,23 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
// These loops iterate over the number of
// tiles needed to carry out the transpose
//
RAJA::statement::Tile<1, RAJA::tile_fixed<TILE_DIM>, RAJA::cuda_block_y_loop,
RAJA::statement::Tile<0, RAJA::tile_fixed<TILE_DIM>, RAJA::cuda_block_x_loop,
RAJA::statement::TileTCount<1, RAJA::statement::Param<1>,
RAJA::tile_fixed<TILE_DIM>, RAJA::cuda_block_y_direct_unchecked,
RAJA::statement::TileTCount<0, RAJA::statement::Param<0>,
RAJA::tile_fixed<TILE_DIM>, RAJA::cuda_block_x_direct_unchecked,
// This statement will initalize local array memory inside a
// kernel. The cpu_tile_mem policy specifies that memory should be
// allocated on the stack. The entries in the RAJA::ParamList
// identify RAJA local arrays to intialize in the parameter tuple.
RAJA::statement::InitLocalMem<RAJA::cuda_shared_mem, RAJA::ParamList<2>,
RAJA::statement::InitLocalMem<RAJA::cuda_shared_mem, RAJA::ParamList<3>,
//
// (1) Execution policies for the first set of inner
// loops. These loops copy data from the global matrices
// to the local tile.
//
RAJA::statement::ForICount<1, RAJA::statement::Param<0>, RAJA::cuda_thread_y_direct,
RAJA::statement::ForICount<0, RAJA::statement::Param<1>, RAJA::cuda_thread_x_direct,
RAJA::statement::Lambda<0>
RAJA::statement::For<2, RAJA::cuda_thread_y_direct,
RAJA::statement::ForICount<0, RAJA::statement::Param<2>, RAJA::cuda_thread_x_direct,
RAJA::statement::Lambda<0, Segs<0, 2>, Params<0, 1, 2, 3>>
>
>,
// Synchronize threads to ensure all loads
Expand All @@ -429,14 +433,11 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
// swapped! This enables us to swap which
// index has unit stride.
//
RAJA::statement::ForICount<0, RAJA::statement::Param<1>, RAJA::cuda_thread_y_direct,
RAJA::statement::ForICount<1, RAJA::statement::Param<0>, RAJA::cuda_thread_x_direct,
RAJA::statement::Lambda<1>
RAJA::statement::For<2, RAJA::cuda_thread_y_direct,
RAJA::statement::ForICount<0, RAJA::statement::Param<2>, RAJA::cuda_thread_x_direct,
RAJA::statement::Lambda<1, Segs<0, 2>, Params<0, 1, 2, 3>>
>
>,
// Synchronize threads to ensure all reads
// from the local array are complete
RAJA::statement::CudaSyncThreads
>
>
>
>
Expand All @@ -446,18 +447,29 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))

RAJA::kernel_param<CUDA_EXEC_POL>(
RAJA::make_tuple(RAJA::TypedRangeSegment<int>(0, N_c),
RAJA::TypedRangeSegment<int>(0, N_r)),
RAJA::make_tuple((int)0, (int)0, Tile_Array),

[=] RAJA_DEVICE (int col, int row, int tx, int ty, TILE_MEM &Tile_Array) {

Tile_Array(ty, tx) = Aview(row, col);
RAJA::TypedRangeSegment<int>(0, N_r),
THREAD_ROWS(0, BLOCK_ROWS)),
RAJA::make_tuple((int)0, (int)0, (int)0, Tile_Array),

[=] RAJA_DEVICE (int col, int ty, int bx, int by, int tx, TILE_MEM &Tile_Array) {
for (int j = 0; j < TILE_DIM; j += BLOCK_ROWS) {
int row = by * TILE_DIM + ty + j;
if (row < N_r) {
Tile_Array(ty + j, tx) = Aview(row, col);
}
}

},

[=] RAJA_DEVICE(int col, int row, int tx, int ty, TILE_MEM &Tile_Array) {

Atview(col, row) = Tile_Array(ty, tx);
[=] RAJA_DEVICE(int RAJA_UNUSED_ARG(col), int ty, int bx, int by, int tx,
TILE_MEM &Tile_Array) {
for (int j = 0; j < TILE_DIM; j += BLOCK_ROWS) {
int row_t = bx * TILE_DIM + ty + j;
int col_t = by * TILE_DIM + tx;
if (row_t < N_c && col_t < N_r) {
Atview(row_t, col_t) = Tile_Array(tx, ty + j);
}
}

}
);
Expand Down Expand Up @@ -496,21 +508,23 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
// These loops iterate over the number of
// tiles needed to carry out the transpose
//
RAJA::statement::Tile<1, RAJA::tile_fixed<TILE_DIM>, RAJA::hip_block_y_loop,
RAJA::statement::Tile<0, RAJA::tile_fixed<TILE_DIM>, RAJA::hip_block_x_loop,
RAJA::statement::TileTCount<1, RAJA::statement::Param<1>,
RAJA::tile_fixed<TILE_DIM>, RAJA::hip_block_y_direct_unchecked,
RAJA::statement::TileTCount<0, RAJA::statement::Param<0>,
RAJA::tile_fixed<TILE_DIM>, RAJA::hip_block_x_direct_unchecked,
// This statement will initalize local array memory inside a
// kernel. The cpu_tile_mem policy specifies that memory should be
// allocated on the stack. The entries in the RAJA::ParamList
// identify RAJA local arrays to intialize in the parameter tuple.
RAJA::statement::InitLocalMem<RAJA::hip_shared_mem, RAJA::ParamList<2>,
RAJA::statement::InitLocalMem<RAJA::hip_shared_mem, RAJA::ParamList<3>,
//
// (1) Execution policies for the first set of inner
// loops. These loops copy data from the global matrices
// to the local tile.
//
RAJA::statement::ForICount<1, RAJA::statement::Param<0>, RAJA::hip_thread_y_direct,
RAJA::statement::ForICount<0, RAJA::statement::Param<1>, RAJA::hip_thread_x_direct,
RAJA::statement::Lambda<0>
RAJA::statement::For<2, RAJA::hip_thread_y_direct,
RAJA::statement::ForICount<0, RAJA::statement::Param<2>, RAJA::hip_thread_x_direct,
RAJA::statement::Lambda<0, Segs<0, 2>, Params<0, 1, 2, 3>>
>
>,
// Synchronize threads to ensure all loads
Expand All @@ -524,14 +538,11 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
// swapped! This enables us to swap which
// index has unit stride.
//
RAJA::statement::ForICount<0, RAJA::statement::Param<1>, RAJA::hip_thread_y_direct,
RAJA::statement::ForICount<1, RAJA::statement::Param<0>, RAJA::hip_thread_x_direct,
RAJA::statement::Lambda<1>
RAJA::statement::For<2, RAJA::hip_thread_y_direct,
RAJA::statement::ForICount<0, RAJA::statement::Param<2>, RAJA::hip_thread_x_direct,
RAJA::statement::Lambda<1, Segs<0, 2>, Params<0, 1, 2, 3>>
>
>,
// Synchronize threads to ensure all reads
// from the local array are complete
RAJA::statement::HipSyncThreads
>
>
>
>
Expand All @@ -541,23 +552,35 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))

RAJA::kernel_param<HIP_EXEC_POL>(
RAJA::make_tuple(RAJA::TypedRangeSegment<int>(0, N_c),
RAJA::TypedRangeSegment<int>(0, N_r)),
RAJA::make_tuple((int)0, (int)0, Tile_Array),

[=] RAJA_DEVICE (int col, int row, int tx, int ty, TILE_MEM &Tile_Array) {

Tile_Array(ty, tx) = d_Aview(row, col);
RAJA::TypedRangeSegment<int>(0, N_r),
THREAD_ROWS(0, BLOCK_ROWS)),
RAJA::make_tuple((int)0, (int)0, (int)0, Tile_Array),

[=] RAJA_DEVICE (int col, int ty, int bx, int by, int tx, TILE_MEM &Tile_Array) {
for (int j = 0; j < TILE_DIM; j += BLOCK_ROWS) {
int row = by * TILE_DIM + ty + j;
if (row < N_r) {
Tile_Array(ty + j, tx) = d_Aview(row, col);
}
}

},

[=] RAJA_DEVICE(int col, int row, int tx, int ty, TILE_MEM &Tile_Array) {

d_Atview(col, row) = Tile_Array(ty, tx);
[=] RAJA_DEVICE(int RAJA_UNUSED_ARG(col), int ty, int bx, int by, int tx,
TILE_MEM &Tile_Array) {
for (int j = 0; j < TILE_DIM; j += BLOCK_ROWS) {
int row_t = bx * TILE_DIM + ty + j;
int col_t = by * TILE_DIM + tx;
if (row_t < N_c && col_t < N_r) {
d_Atview(row_t, col_t) = Tile_Array(tx, ty + j);
}
}

}
);

CAMP_HIP_API_INVOKE_AND_CHECK(hipMemcpy, At, d_At, N_r * N_c * sizeof(int), hipMemcpyDeviceToHost);
CAMP_HIP_API_INVOKE_AND_CHECK(hipDeviceSynchronize);
checkResult<int>(Atview, N_c, N_r);
// printResult<int>(Atview, N_c, N_r);
#endif
Expand All @@ -573,7 +596,7 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
using RAJA::Offsets;
using RAJA::Params;

// _raja_mattranspose_lambdaargs_start
// _mattranspose_localarray_raja_lambdaargs_start
using SEQ_EXEC_POL_II =
RAJA::KernelPolicy<
RAJA::statement::Tile<1, RAJA::tile_fixed<TILE_DIM>, RAJA::seq_exec,
Expand Down Expand Up @@ -612,7 +635,7 @@ int main(int RAJA_UNUSED_ARG(argc), char **RAJA_UNUSED_ARG(argv[]))
Atview(col, row) = _Tile_Array(ty, tx);
}
);
// _raja_mattranspose_lambdaargs_start
// _mattranspose_localarray_raja_lambdaargs_end

checkResult<int>(Atview, N_c, N_r);
// printResult<int>(Atview, N_c, N_r);
Expand Down
Loading
Loading