Source library

Memory region allocator — red team v3 (panel re-review of v4)

docs/ps1/memory-region-allocator-red-team-v3.md

This is the website shelf page for docs/ps1/memory-region-allocator-red-team-v3.md. It exists because the project has hundreds of Markdown files and the website should not pretend the interesting work only happened in the polished essays.

Why this file matters

Active PS1 reference. This is active engineering documentation for the PS1 port. It is close enough to the code that stale claims here become real bugs.

Companion to memory-region-allocator-plan.md (v4). Same five-reviewer panel from memory-region-allocator-red-team-v2.md reconvenes. Each reviewer re-evaluates whether their prior concerns are properly resolved and flags new issues introduced by the v3→v4 revisions (especially the JCBSOD integration, which is fresh).

File facts

Field Value
Source path docs/ps1/memory-region-allocator-red-team-v3.md
Lines 483
Approx. words 3197
Code fences 1

Outline

  • Memory region allocator — red team v3 (panel re-review of v4)
  • Reviewer 1 — Pat, Embedded Systems Veteran
  • Closed in v4
  • Reopened / new
  • 🔴 P1-bis. ISR enforcement is debug-only — re-classified as blocker
  • 🔴 P8. ps1Bsod reads stale heap diagnostics that no longer mean anything
  • 🟠 P9. ps1Bsod is noreturn and assumes graphics is up
  • 🟠 P10. PSX printf does not allocate (verify, document)
  • 🟡 P11. Watchdog / emulator timeout while in BSOD halt
  • 🟡 P12. CRC-32 vs SHA-256
  • Reviewer 2 — Sarah, SRE / Operations
  • Closed in v4
  • Reopened / new
  • 🔴 S7. Per-file PR sequencing across Phase 2 is unsafe as written

Read it in context