Support building on ARM by providing scalar histogram fallback

The metrics histogram update code unconditionally included <immintrin.h>
and used __attribute__((target("avx"))) SSE/AVX intrinsics, which only
exist on x86-64. This prevented the project from compiling on ARM64.

Guard the x86-64 SIMD implementation and the <immintrin.h> include with
an architecture check, and add a portable scalar fallback for non-x86-64
platforms (e.g., ARM64). A thin wrapper function keeps the call sites
unchanged and preserves the AVX fast path on x86-64.

Closes #3

src/metric.cpp

@@ -22,7 +22,9 @@
 #include <unordered_set>
 #include <vector>
 
+#if defined(__x86_64__) || defined(__amd64__) || defined(_M_X64)
 #include <immintrin.h>
+#endif
 #include <simdutf.h>
 
 #include "arena.hpp"
@@ -1398,8 +1400,10 @@ void Gauge::set(double x) {
 Histogram::Histogram() = default;
 
 // Vectorized histogram bucket updates with mutex protection for consistency
-// AVX-optimized implementation for high performance
+// AVX-optimized implementation for high performance on x86-64, scalar fallback
+// on other architectures (e.g., ARM64).
 
+#if defined(__x86_64__) || defined(__amd64__) || defined(_M_X64)
 __attribute__((target("avx"))) static void
 update_histogram_buckets_simd(std::span<const double> thresholds,
                               std::span<uint64_t> counts, double x,
@@ -1439,6 +1443,22 @@ update_histogram_buckets_simd(std::span<const double> thresholds,
     }
   }
 }
+#endif
+
+static void update_histogram_buckets(std::span<const double> thresholds,
+                                     std::span<uint64_t> counts, double x,
+                                     size_t start_idx) {
+#if defined(__x86_64__) || defined(__amd64__) || defined(_M_X64)
+  update_histogram_buckets_simd(thresholds, counts, x, start_idx);
+#else
+  const size_t size = thresholds.size();
+  for (size_t i = start_idx; i < size; ++i) {
+    if (x <= thresholds[i]) {
+      counts[i]++;
+    }
+  }
+#endif
+}
 
 void Histogram::observe(double x) {
   assert(p->thresholds.size() == p->shared.bucket_counts.size());
@@ -1459,14 +1479,14 @@ void Histogram::observe(double x) {
     }
 
     // Update shared directly
-    update_histogram_buckets_simd(p->thresholds, p->shared.bucket_counts, x, 0);
+    update_histogram_buckets(p->thresholds, p->shared.bucket_counts, x, 0);
     p->shared.sum += x;
     p->shared.observations++;
 
     p->mutex.unlock();
   } else {
     // Slow path: accumulate in pending (lock-free)
-    update_histogram_buckets_simd(p->thresholds, p->pending.bucket_counts, x,
+    update_histogram_buckets(p->thresholds, p->pending.bucket_counts, x,
                              0);
     p->pending.sum += x;
     p->pending.observations++;