Skip to content

Fix Inverse kernel rank underflow before indexing trailing dims #28400

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
tianleiwu merged 2 commits into from
May 15, 2026
Merged

Fix Inverse kernel rank underflow before indexing trailing dims #28400

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
78c57f2
Guard Inverse rank before indexing trailing dims; add rank validation…
Copilot May 7, 2026
6b3bcff
Merge remote-tracking branch 'origin/main' into copilot/fix-inverse-r…
tianleiwu May 7, 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
2 changes: 2 additions & 0 deletions onnxruntime/contrib_ops/cpu/inverse.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -70,6 +70,8 @@ Status Inverse::Compute(OpKernelContext* ctx) const {
const auto num_dim = input_shape.NumDimensions();
auto* output = ctx->Output(0, input_shape);

ORT_RETURN_IF_NOT(num_dim >= 2, "Input tensor rank must be >= 2, got: ", num_dim);

int64_t num_batches = 1;
const int64_t rows = input_shape.GetDims()[num_dim - 2];
const int64_t cols = input_shape.GetDims()[num_dim - 1];
Expand Down
2 changes: 2 additions & 0 deletions onnxruntime/contrib_ops/cuda/inverse.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -142,6 +142,8 @@ Status Inverse::ComputeInternal(OpKernelContext* ctx) const {
const auto num_dim = input_shape.NumDimensions();
auto* output = ctx->Output(0, input_shape);

ORT_RETURN_IF_NOT(num_dim >= 2, "Input tensor rank must be >= 2, got: ", num_dim);

size_t num_batches = 1;
const size_t rows = static_cast<size_t>(input_shape.GetDims()[num_dim - 2]);
const size_t cols = static_cast<size_t>(input_shape.GetDims()[num_dim - 1]);
Expand Down
15 changes: 15 additions & 0 deletions onnxruntime/test/contrib_ops/inverse_test.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -86,5 +86,20 @@ TEST(InverseContribOpTest, four_by_four_batches_float) {
test.AddOutput<float>("Output", {3, 4, 4, 4}, output);
test.Run();
}

TEST(InverseContribOpTest, scalar_input_fails) {
OpTester test("Inverse", 1, kMSDomain);
test.AddInput<float>("X", {}, {4.f});
test.AddOutput<float>("Y", {}, {0.25f});
test.Run(OpTester::ExpectResult::kExpectFailure, "rank must be >= 2");
}

TEST(InverseContribOpTest, one_dim_input_fails) {
OpTester test("Inverse", 1, kMSDomain);
test.AddInput<float>("X", {4}, {4.f, 7.f, 2.f, 6.f});
test.AddOutput<float>("Y", {4}, {0.f, 0.f, 0.f, 0.f});
test.Run(OpTester::ExpectResult::kExpectFailure, "rank must be >= 2");
}

} // namespace test
} // namespace onnxruntime
Loading