build: raise x86-64 baseline from SSE2 to SSE4.2

[bruno-dasilva]

PERFORMANCE TESTING

---

Default GCC/Clang builds targeted SSE2 via a wall of -mno-sse3/ssse3/sse4.* flags in the generic MARCH fallback. Replace with -msse4.2, which pulls in SSE3/SSSE3/SSE4.1. AVX/FMA stay banned — FMA contraction changes FP bit patterns and desyncs the deterministic simulation.

Minimum x86 CPU is now Nehalem (2008) / Bulldozer (2011). Requires a replay-level sync validation pass before shipping — autovectorization output will differ from prior builds even without FMA.

Co-Authored-By: Claude Opus 4.7 (1M context) noreply@anthropic.com

[github-actions]

bar-benchmark — PR #4

candidate 24808db vs baseline eb1c69f

sim trimmed mean (ms) with 95% CI on the relative delta

scenario	candidate	baseline	Δ (95% CI)	n cand	n base
fightertest-bots	23.91 ms ♻️	23.86 ms ♻️	$\color{red}{+0.08\%} \text{ to } \color{red}{+0.36\%}$	50	80
fightertest-aircraft	19.23 ms ♻️	19.17 ms ♻️	$\color{red}{+0.22\%} \text{ to } \color{red}{+0.36\%}$	50	70
fightertest-tanks	24.88 ms ♻️	24.82 ms ♻️	$\color{green}{-0.00\%} \text{ to } \color{red}{+0.44\%}$	50	70
fightertest-pathfinding	21.82 ms ♻️	21.77 ms ♻️	$\color{red}{+0.05\%} \text{ to } \color{red}{+0.41\%}$	50	70
lategame1	23.08 ms	23.39 ms ♻️	$\color{green}{-2.06\%} \text{ to } \color{green}{-0.59\%}$	50	110

_{💰 compute cost: $0.28 · 1 fresh leg · 9 cached at $0}

_{last updated: 2026-04-25T15:30:04.654Z · workflow run}