Handle extreme diagonal values in CPU Trilu

[Akshay Sonawane (apsonawane)]

This pull request improves the robustness of the Trilu operator implementation and its test coverage, particularly for edge cases involving extreme diagonal values. The main changes include handling out-of-bounds diagonal indices in the implementation and adding new tests to ensure correct behavior for these cases.

Implementation robustness:

• Updated TriluImpl in trilu.cc to handle extreme values of k_val (diagonal index), ensuring the output is correctly set or skipped when k_val is out of bounds for both upper and lower triangular cases.

Test coverage improvements:

• Added new tests in trilu_op_test.cc to verify behavior when k is set to the minimum and maximum possible int64_t values, ensuring the operator's correctness for these edge cases.
• Included <limits> header to facilitate the use of extreme integer values in tests.

[Copilot]

Pull request overview

This pull request hardens the CPU implementation of the ONNX Trilu operator against extreme diagonal (k) values (e.g., int64_t min/max) by short-circuiting cases where the requested diagonal is fully out of bounds, and adds targeted unit tests to validate those edge conditions.

Changes:

• Add early-exit handling in the CPU TriluImpl to correctly return identity/all-zero outputs when k is far outside the matrix bounds (avoiding problematic arithmetic with extreme values).
• Add new CPU-only tests covering k = std::numeric_limits<int64_t>::min() and k = std::numeric_limits<int64_t>::max() for both upper and lower triangular modes.
• Update test includes to support extreme int64_t constants.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
onnxruntime/test/providers/cpu/tensor/trilu_op_test.cc	Adds CPU-only unit tests for extreme k values using std::numeric_limits<int64_t>::min/max().
onnxruntime/core/providers/cpu/tensor/trilu.cc	Adds bounds checks/short-circuits in TriluImpl to handle out-of-range diagonals safely and deterministically.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.