From f86711413fa475466e7bcb67b36f58691bb395f6 Mon Sep 17 00:00:00 2001
From: rcourtman <8825017+rcourtman@users.noreply.github.com>
Date: Thu, 24 Sep 2026 14:22:23 +0100
Subject: [PATCH] Stop timing the one-cycle NormalizeRoute root fast path

The advisory Benchmarks check kept failing unrelated pull requests on
NormalizeRoute/root alone (#2228, #2242, #2245, #2250, and PR 1973 on
2026-09-08), and flagged it beside other results on #2230 and #2244.
The "/" path is a length compare and a constant return.

In the last 19 paired PR runs, the 11 on AMD EPYC 7763 measured root
at exactly 2.18ns or 2.50ns on both sides, each within 0-1%. The two
values are about one CPU cycle apart. In five of those runs base and
candidate landed on different values, four times up (+14%) and once
down (-13%), though none touched the normaliser. The earlier
investigation found the benchmark loop identical and the normaliser
differing only in relocated addresses. More samples cannot separate a
stable one-cycle shift from a real change, and a looser threshold
would blunt the gate for everything else.

Batching root calls does not escape it. On an amd64 host a 64-call
batch spread as widely in relative terms as a single call, because
every call repeats the same few instructions.

What the path must guarantee is that "/" and "" return before
splitting. TestNormalizeRoute_RootFastPathDoesNotAllocate pins that
deterministically: with the fast path removed, TestNormalizeRoute
still passes, but the new test fails with one allocation per call.

Real regressions stay visible. A paired run using the CI runner's own
alternating recipe (10 samples, 100ms), with normalizeRoute made to do
its work twice, flagged every remaining NormalizeRoute sub-benchmark
at +89% to +107%, FullMiddlewarePath at +54%, and bytes and allocs at
+100%. check-bench-regression.sh exited 1. The same tree compared with
itself exited 0.
---
 internal/api/http_metrics_bench_test.go |  7 ++++++-
 internal/api/http_metrics_test.go       | 10 ++++++++++
 2 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/internal/api/http_metrics_bench_test.go b/internal/api/http_metrics_bench_test.go
index d11f00a32b..5d2e3fc813 100644
--- a/internal/api/http_metrics_bench_test.go
+++ b/internal/api/http_metrics_bench_test.go
@@ -13,6 +13,12 @@ import (
 
 // BenchmarkNormalizeRoute measures route normalization with representative
 // API paths of varying complexity.
+//
+// The "/" fast path is deliberately not timed here. Paired CI runs measured it
+// at either 2.18ns or 2.50ns, about one CPU cycle apart, and changes that never
+// touched this code flipped it both ways by 13-30%.
+// TestNormalizeRoute_RootFastPathDoesNotAllocate pins what matters about the
+// path instead.
 func BenchmarkNormalizeRoute(b *testing.B) {
 	paths := []struct {
 		name string
@@ -23,7 +29,6 @@ func BenchmarkNormalizeRoute(b *testing.B) {
 		{"with_uuid", "/api/nodes/550e8400-e29b-41d4-a716-446655440000"},
 		{"deep_path_truncated", "/api/v1/orgs/123/resources/456/metrics"},
 		{"with_query_params", "/api/metrics-store/history?resourceType=vm&resourceId=pve1:node1:100&metric=cpu&range=1h"},
-		{"root", "/"},
 		{"metrics_history", "/api/metrics-store/history"},
 	}
 
diff --git a/internal/api/http_metrics_test.go b/internal/api/http_metrics_test.go
index 6370cc2838..3e6e98ab8a 100644
--- a/internal/api/http_metrics_test.go
+++ b/internal/api/http_metrics_test.go
@@ -246,6 +246,16 @@ func TestNormalizeRoute(t *testing.T) {
 	}
 }
 
+// The root and empty paths must return before splitting. Without the fast
+// path, "/" still normalizes to "/" but allocates a split slice on every call.
+func TestNormalizeRoute_RootFastPathDoesNotAllocate(t *testing.T) {
+	for _, path := range []string{"/", ""} {
+		if allocs := testing.AllocsPerRun(100, func() { _ = normalizeRoute(path) }); allocs != 0 {
+			t.Errorf("normalizeRoute(%q) allocated %v times per call, want 0", path, allocs)
+		}
+	}
+}
+
 // ---------------------------------------------------------------------------
 // recordAPIRequest integration tests — verify that the core recording
 // function correctly populates all three Prometheus metric vectors.