feat(metrics): instrument upstream latency

Add time-to-first-byte and total-duration histograms for every upstream
call, observation only. No throttling, priorities, or control logic
reads this data. The operator wants strain-based throttling later and
needs real traffic data first.

Extend watchdogBody (timeout.go) rather than add a new interception
layer: it already wraps every upstream response body for provider
adapters, native passthrough, and the MCP/A2A target proxy, and already
observes the first successful read. armLatency arms a watchdogBody
after construction, so newWatchdogBody keeps its existing signature and
behavior; its own pinned tests still pass unchanged. A timestamp plus a
few atomics are the only added cost; Read and Close each pay one nil
check when latency collection is disabled.

Expose llmgateway_upstream_ttfb_seconds and
llmgateway_upstream_duration_seconds on the existing Prometheus
endpoint (metrics.go). Both are in-process, per-replica histograms,
never store-backed, so sum() is always the correct cross-replica
aggregation. Buckets run 0.1s to 600s, chosen to resolve fast TTFB and
short completions finely, cover the bulk of real chat completions, and
bracket the five-minute default request timeout. Labels are provider
and stream always, model only when Config.Metrics.ModelLabel is set,
matching the existing per-model opt-in convention. Add a compact
per-provider, per-stream average summary to GET /admin/api/overview,
sourced from the same in-process store, without a full histogram or a
per-model breakdown.

Every write path is gated on metricsEnabled(cfg), so a deployment with
metrics disabled collects nothing and the histogram families never
render. No behavior change to request handling: same responses, status
codes, headers, and accounting.

Fix a Yaegi-only bug this feature's own context-threading change
exposed: two distinct empty struct types used as context.WithValue keys
compare equal to each other under the interpreter's interface equality,
though never under compiled Go. Threading a second recorder
(latencyRecorder) alongside the existing attemptRecorder made this
observable for the first time, since compiled tests never see it and
the plugin previously carried only one such key. Give
latencyRecorderCtxKey a field so its size differs from
attemptRecorderCtxKey's, confirmed against a minimal Yaegi reproduction
before and after the fix.

Fix a second bug, reproducible under compiled Go too: io.Reader may
legally return its final data together with io.EOF in one call, and
Go's real net/http body reader does this for small, fully-buffered
responses. The original TTFB capture lived inside the err == nil
branch, so it never fired for a typical small non-streaming completion.
Move the check ahead of the error branch so it fires on any n > 0,
leaving the pinned timer-reset and Close semantics untouched.

Tests mutate the real source per case and confirm each one fails first,
recorded in each test's own doc comment.

routes_unified.go

@@ -494,6 +494,15 @@ func (g *Gateway) runMeteredCall(sw *statusTrackingWriter, r *http.Request, scop
 		ctx := withAttemptRecorder(r.Context(), func(resp *http.Response, attemptErr error) {
 			g.limiter.recordProviderAttempt(providerName, upstreamModel, resp, attemptErr)
 		})
+		// feat: instrument upstream latency — gated on metricsEnabled so a
+		// deployment with metrics off wires nothing here at all (not even
+		// the closure allocation), one level up from watchdogBody's own
+		// per-Read nil check (timeout.go).
+		if metricsEnabled(g.cfg) {
+			ctx = withLatencyRecorder(ctx, func(sample latencySample) {
+				g.recordLatency(providerName, upstreamModel, sample)
+			})
+		}
 		result, callErr := call(cand.adapter, ctx, respWriter, attemptReq)
 
 		// Request-path health, recorded ONCE per candidate per logical