[uint256] Force intermediate results of bitwise operations to avoid AVX stalls

[m-alvarez]

As noted in #1602, bitwise operations on uint256_t are vectorized by GCC/Clang and incur store forwarding stalls when mixed with non-vectorized operations. This can be fixed by forcing the intermediate results.

[Copilot]

Pull Request Overview

This PR addresses store forwarding stalls in uint256_t bitwise operations by forcing intermediate results to prevent AVX vectorization issues when mixed with non-vectorized operations.

Key Changes:

• Wraps bitwise operation results in force() calls within the BITWISE_BINOP macro to prevent compiler vectorization that causes store forwarding stalls

Comments suppressed due to low confidence (1)

category/vm/runtime/uint256.hpp:1

• [nitpick] The force() function is not defined or documented in the visible context. Consider adding a comment explaining what force() does (e.g., preventing compiler vectorization) to improve code readability for future maintainers.

// Copyright (C) 2025 Category Labs, Inc.

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[dhil]

What is the generated code diff before/after these changes?

[andreaslyn]

Did you forget NOT?