Process message stacks without intermediate traversal allocations

[OskarEichler]

Summary

Traverse incoming sessions in reverse order without copying the array and carry the message through plain loops in both directions. Preserve error wrapping and pipeline order.

Reproduction and verification

# With active sessions a then b:
exts.process_outgoing_message(message) # a then b
exts.process_incoming_message(message) # b then a

Existing processing-order/error examples pass, plus 100 randomized selection/order scenarios. For 100,000 no-op messages through three sessions, incoming allocations fell from 300,001 to 1 and outgoing from 200,001 to 1; local times were 0.026354 to 0.015430s and 0.024441 to 0.014949s. These are synthetic local measurements, not production claims.

• Ruby 4.0.6 through rbenv; existing current-upstream suite: 64 examples, zero failures on this isolated change.
• The cumulative release-based candidate passes 283 focused checks (272 baseline failures → zero), 2,784 model checks, the current upstream's 64-example suite, and gem build/extraction.
• The historical 0.1.5 suite has three Ruby keyword-versus-options-hash mock failures on both baseline and candidate. Current upstream already corrected those expectations; they were run against the candidate through an external preload without changing repository tests.
• No new or modified tests/specs, following the consumer repository's explicit policy. Reproductions/models were run from external scratch scripts.

Breaking changes and limitations

No intended breaking change. The same message transformations, final return value and ExtensionError behavior remain; no extension API changes.

Only local Ruby 4.0.6/macOS execution is claimed; the repository's older Ruby/JRuby matrix needs upstream CI. No production access or unrelated release upgrades.