Debug Counters Implementation - Clean History

Major Features:
- Debug counter infrastructure for Refill Stage tracking
- Free Pipeline counters (ss_local, ss_remote, tls_sll)
- Diagnostic counters for early return analysis
- Unified larson.sh benchmark runner with profiles
- Phase 6-3 regression analysis documentation

Bug Fixes:
- Fix SuperSlab disabled by default (HAKMEM_TINY_USE_SUPERSLAB)
- Fix profile variable naming consistency
- Add .gitignore patterns for large files

Performance:
- Phase 6-3: 4.79 M ops/s (has OOM risk)
- With SuperSlab: 3.13 M ops/s (+19% improvement)

This is a clean repository without large log files.

🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>

benchmarks/scripts/comprehensive/run_perf_hot_triad.sh

@@ -0,0 +1,56 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+# perf-stat triad (HAKMEM/System/mimalloc) for bench_tiny_hot at a single size/batch/cycles
+# Collects cycles/instructions/branches/branch-misses/L1-dcache-load-misses (CSV via perf -x ,)
+# Usage: scripts/run_perf_hot_triad.sh [size] [batch] [cycles] [reps]
+#   size  : bytes (default 32)
+#   batch : batch size (default 100)
+#   cycles: loop cycles per op (default 50000)
+#   reps  : perf repetitions (default 3)
+
+ROOT_DIR=$(cd "$(dirname "$0")/.." && pwd)
+cd "$ROOT_DIR"
+
+size=${1:-32}
+batch=${2:-100}
+cycles=${3:-50000}
+reps=${4:-3}
+
+echo "[build] benches (fast + mi)"
+make -s bench_fast bench_tiny_hot_mi >/dev/null
+
+# Ensure LD_LIBRARY_PATH is defined (set -u safety)
+export LD_LIBRARY_PATH="${LD_LIBRARY_PATH:-}"
+MI_LIBDIR="$ROOT_DIR/mimalloc-bench/extern/mi/out/release"
+
+TS=$(date +%Y%m%d_%H%M%S)
+OUTDIR="bench_results/perf_hot_triad_${TS}"
+mkdir -p "$OUTDIR"
+
+echo "[info] size=$size batch=$batch cycles=$cycles reps=$reps"
+echo "[info] results → $OUTDIR"
+
+run_perf() {
+  local alloc="$1"; shift
+  local bin="$1"; shift
+  local tag="$alloc"_s${size}_b${batch}_c${cycles}
+  local log="$OUTDIR/${tag}.perf.csv"
+  echo "[perf] $tag"
+  if [[ "$alloc" == "mimalloc" ]]; then
+    LD_LIBRARY_PATH="$LD_LIBRARY_PATH:$MI_LIBDIR" \
+      perf stat -x , -r "$reps" -e cycles,instructions,branches,branch-misses,L1-dcache-load-misses \
+        "$bin" "$size" "$batch" "$cycles" 1>/dev/null 2>"$log" || true
+  else
+    perf stat -x , -r "$reps" -e cycles,instructions,branches,branch-misses,L1-dcache-load-misses \
+      "$bin" "$size" "$batch" "$cycles" 1>/dev/null 2>"$log" || true
+  fi
+}
+
+run_perf hakmem ./bench_tiny_hot_hakmem
+run_perf system ./bench_tiny_hot_system
+run_perf mimalloc ./bench_tiny_hot_mi
+
+echo "[done] perf CSVs under: $OUTDIR"
+ls -1 "$OUTDIR" | sed -n '1,20p'
+