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

sparse.mm.backward: fix for non-contiguous grad values on CPU #106127

Closed
wants to merge 3 commits into from
Closed

sparse.mm.backward: fix for non-contiguous grad values on CPU #106127

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
068c5be
sparse.mm.backward: fix for non-contiguous grad values on CPU
nikitaved Jul 27, 2023
6be99c3
Update on "sparse.mm.backward: fix for non-contiguous grad values on …
nikitaved Jul 27, 2023
596c378
Update on "sparse.mm.backward: fix for non-contiguous grad values on …
nikitaved Jul 27, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
92 changes: 58 additions & 34 deletions aten/src/ATen/native/sparse/SparseMatMul.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -50,17 +50,18 @@ void csr_to_coo(const int64_t n_row, const int64_t Ap[], int64_t Bi[]) {
}
}

template<typename index_t_ptr = int64_t*>
int64_t _csr_matmult_maxnnz(
const int64_t n_row,
const int64_t n_col,
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
const int64_t Ap[],
const index_t_ptr Ap,
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
const int64_t Aj[],
const index_t_ptr Aj,
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
const int64_t Bp[],
const index_t_ptr Bp,
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
const int64_t Bj[]) {
const index_t_ptr Bj) {
/*
Compute needed buffer size for matrix `C` in `C = A@B` operation.

Expand Down Expand Up @@ -88,28 +89,28 @@ int64_t _csr_matmult_maxnnz(
return nnz;
}

template<class scalar_t>
template<typename index_t_ptr, typename scalar_t_ptr>
void _csr_matmult(
const int64_t n_row,
const int64_t n_col,
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
const int64_t Ap[],
const index_t_ptr Ap,
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
const int64_t Aj[],
const index_t_ptr Aj,
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
const scalar_t Ax[],
const scalar_t_ptr Ax,
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
const int64_t Bp[],
const index_t_ptr Bp,
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
const int64_t Bj[],
const index_t_ptr Bj,
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
const scalar_t Bx[],
const scalar_t_ptr Bx,
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
int64_t Cp[],
typename index_t_ptr::value_type Cp[],
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
int64_t Cj[],
typename index_t_ptr::value_type Cj[],
// NOLINTNEXTLINE(modernize-avoid-c-arrays,cppcoreguidelines-avoid-c-arrays)
scalar_t Cx[]) {
typename scalar_t_ptr::value_type Cx[]) {
Copy link
Collaborator Author

@nikitaved nikitaved Jul 27, 2023 •
edited

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's probably not assume that C-arrays are always contiguous...
EDIT: not needed. This function is used in a single place, and only receives contiguous output.

/*
Compute CSR entries for matrix C = A@B.

Expand All @@ -133,27 +134,30 @@ void _csr_matmult(
Note:
Output arrays Cp, Cj, and Cx must be preallocated
*/
std::vector<int64_t> next(n_col, -1);
using index_t = typename index_t_ptr::value_type;
using scalar_t = typename scalar_t_ptr::value_type;

std::vector<index_t> next(n_col, -1);
std::vector<scalar_t> sums(n_col, 0);

int64_t nnz = 0;

Cp[0] = 0;

for (const auto i : c10::irange(n_row)) {
int64_t head = -2;
int64_t length = 0;
index_t head = -2;
index_t length = 0;

int64_t jj_start = Ap[i];
int64_t jj_end = Ap[i + 1];
index_t jj_start = Ap[i];
index_t jj_end = Ap[i + 1];
for (const auto jj : c10::irange(jj_start, jj_end)) {
int64_t j = Aj[jj];
index_t j = Aj[jj];
scalar_t v = Ax[jj];

int64_t kk_start = Bp[j];
int64_t kk_end = Bp[j + 1];
index_t kk_start = Bp[j];
index_t kk_end = Bp[j + 1];
for (const auto kk : c10::irange(kk_start, kk_end)) {
int64_t k = Bj[kk];
index_t k = Bj[kk];

sums[k] += v * Bx[kk];

Expand All @@ -174,7 +178,7 @@ void _csr_matmult(
Cx[nnz] = sums[head];
nnz++;

int64_t temp = head;
index_t temp = head;
head = next[head];

next[temp] = -1; // clear arrays
Expand All @@ -183,6 +187,7 @@ void _csr_matmult(

// Make sure that col indices are sorted.
// TODO: a better approach is to implement a CSR @ CSC kernel.
// NOTE: Cx arrays are expected to be contiguous!
auto col_indices_accessor = StridedRandomAccessor<int64_t>(Cj + nnz - length, 1);
auto val_accessor = StridedRandomAccessor<scalar_t>(Cx + nnz - length, 1);
auto kv_accessor = CompositeRandomAccessorCPU<
Expand Down Expand Up @@ -212,13 +217,32 @@ void sparse_matmul_kernel(
const auto mat1_csr = mat1.to_sparse_csr();
const auto mat2_csr = mat2.to_sparse_csr();

const auto nnz = _csr_matmult_maxnnz(
M,
N,
auto mat1_crow_indices_ptr = StridedRandomAccessor<int64_t>(
mat1_csr.crow_indices().data_ptr<int64_t>(),
mat1_csr.crow_indices().stride(-1));
auto mat1_col_indices_ptr = StridedRandomAccessor<int64_t>(
mat1_csr.col_indices().data_ptr<int64_t>(),
mat1_csr.col_indices().stride(-1));
auto mat1_values_ptr = StridedRandomAccessor<scalar_t>(
mat1_csr.values().data_ptr<scalar_t>(),
mat1_csr.values().stride(-1));
auto mat2_crow_indices_ptr = StridedRandomAccessor<int64_t>(
mat2_csr.crow_indices().data_ptr<int64_t>(),
mat2_csr.col_indices().data_ptr<int64_t>());
mat2_csr.crow_indices().stride(-1));
auto mat2_col_indices_ptr = StridedRandomAccessor<int64_t>(
mat2_csr.col_indices().data_ptr<int64_t>(),
mat2_csr.col_indices().stride(-1));
auto mat2_values_ptr = StridedRandomAccessor<scalar_t>(
mat2_csr.values().data_ptr<scalar_t>(),
mat2_csr.values().stride(-1));

const auto nnz = _csr_matmult_maxnnz(
M,
N,
mat1_crow_indices_ptr,
mat1_col_indices_ptr,
mat2_crow_indices_ptr,
mat2_col_indices_ptr);

auto output_indices = output._indices();
auto output_values = output._values();
Expand All @@ -234,12 +258,12 @@ void sparse_matmul_kernel(
_csr_matmult(
M,
N,
mat1_csr.crow_indices().data_ptr<int64_t>(),
mat1_csr.col_indices().data_ptr<int64_t>(),
mat1_csr.values().data_ptr<scalar_t>(),
mat2_csr.crow_indices().data_ptr<int64_t>(),
mat2_csr.col_indices().data_ptr<int64_t>(),
mat2_csr.values().data_ptr<scalar_t>(),
mat1_crow_indices_ptr,
mat1_col_indices_ptr,
mat1_values_ptr,
mat2_crow_indices_ptr,
mat2_col_indices_ptr,
mat2_values_ptr,
output_indptr.data_ptr<int64_t>(),
output_col_indices.data_ptr<int64_t>(),
output_values.data_ptr<scalar_t>());
Expand Down
12 changes: 12 additions & 0 deletions test/test_sparse.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3700,6 +3700,17 @@ def different_dtypes():

self.assertRaisesRegex(RuntimeError, 'mat1 dtype Double does not match mat2 dtype Float', different_dtypes)

def test_backward_noncontiguous():
# Sparse.mm backward used to wrong with non-contiguous grads,
# see https://github.com/pytorch/pytorch/issues/102493.
n_reps = 7
for _ in range(n_reps):
A = torch.eye(5).to_sparse().requires_grad_(True)
B = torch.eye(5).to_sparse()
out = torch.sparse.mm(A, B)
out.coalesce().values().sum().backward()
self.assertEqual(A.grad, A)

for n in range(2, 5):
for m in range(2, 8):
for p in range(2, 8):
Expand All @@ -3708,6 +3719,7 @@ def different_dtypes():
test_sparse_matmul(2, 0, [0, 0], [0, 0])
test_sparse_matmul(2, 0, [0, 10], [10, 0])
test_error_cases()
test_backward_noncontiguous()

@coalescedonoff
@dtypes(torch.double)
Expand Down