Summary

Refactored the extern_facade_return_tys function to return a slice instead of a vector, improving performance by avoiding unnecessary allocations. Also modified extern_facade_expr to accept an iterator instead of a vector, allowing for more flexible usage patterns. Updated all call sites to accommodate these changes, resulting in cleaner code that avoids unnecessary vector allocations and copies.

Type of change

Please check one:

• Bug fix (fixes incorrect behavior)
• New feature
• Performance improvement
• Documentation change with concrete technical impact
• Style, wording, formatting, or typo-only change

Why is this change needed?

The previous implementation was creating unnecessary vector allocations when processing extern function return types. By changing the return type to a slice and accepting iterators instead of vectors, we reduce memory allocations and improve performance in the lowering phase.

What was the behavior or documentation before?

Previously, extern_facade_return_tys would always allocate a new vector even when returning a single type, and extern_facade_expr required a fully constructed vector to be passed in, leading to additional allocations and copies.

What is the behavior or documentation after?

Now extern_facade_return_tys returns a slice (either a reference to the existing tuple types or a single-element slice), and extern_facade_expr accepts any exact-size iterator of variable IDs, allowing callers to avoid unnecessary allocations.

Additional context

This change is part of ongoing efforts to optimize the compiler's memory usage and performance. The modifications maintain the same functionality while reducing allocations in the hot path of the lowering phase.