test(e2e): cover host mode, reusable workflows, scheduling, reconnect and CLI flags (#1347)

Extends the e2e suite with scenarios for host labels, local `workflow_call`, label routing, `runner.capacity`, a dropped and restored connection to Gitea, and the daemon's `--once`, `--labels` and `shutdown_timeout` behaviour.

Stacked on https://gitea.com/gitea/runner/pulls/1346, so its commit shows here until that merges. The new scenarios pass locally against gitea/gitea:latest and main-nightly.

Assisted-by: Codet:codet
Co-authored-by: Zettat123 <[email protected]>
Co-authored-by: bircni <[email protected]>
Reviewed-on: https://gitea.com/gitea/runner/pulls/1347
Reviewed-by: Zettat123 <[email protected]>

e2e/reconnect_test.go

@@ -0,0 +1,127 @@
+// Copyright 2026 The Gitea Authors. All rights reserved.
+// SPDX-License-Identifier: MIT
+
+//go:build e2e
+
+package e2e
+
+import (
+	"fmt"
+	"io"
+	"net"
+	"net/url"
+	"sync"
+	"sync/atomic"
+	"testing"
+	"time"
+)
+
+// flakyProxy forwards TCP to Gitea and can drop every connection to simulate an outage.
+type flakyProxy struct {
+	ln      net.Listener
+	target  string
+	down    atomic.Bool
+	refused atomic.Int64
+	mu      sync.Mutex
+	conns   map[net.Conn]struct{}
+}
+
+func newFlakyProxy(t *testing.T, targetURL string) *flakyProxy {
+	t.Helper()
+	u, err := url.Parse(targetURL)
+	if err != nil {
+		t.Fatalf("parse %q: %v", targetURL, err)
+	}
+	ln, err := net.Listen("tcp", "127.0.0.1:0")
+	if err != nil {
+		t.Fatalf("listen: %v", err)
+	}
+	p := &flakyProxy{ln: ln, target: u.Host, conns: map[net.Conn]struct{}{}}
+	go p.serve()
+	t.Cleanup(func() {
+		_ = ln.Close()
+		p.dropAll()
+	})
+	return p
+}
+
+func (p *flakyProxy) URL() string { return "http://" + p.ln.Addr().String() }
+
+func (p *flakyProxy) serve() {
+	for {
+		c, err := p.ln.Accept()
+		if err != nil {
+			return
+		}
+		if p.down.Load() {
+			p.refused.Add(1)
+			_ = c.Close()
+			continue
+		}
+		up, err := net.Dial("tcp", p.target)
+		if err != nil {
+			_ = c.Close()
+			continue
+		}
+		p.mu.Lock()
+		p.conns[c], p.conns[up] = struct{}{}, struct{}{}
+		p.mu.Unlock()
+		go func() { _, _ = io.Copy(up, c); _ = up.Close() }()
+		go func() { _, _ = io.Copy(c, up); _ = c.Close() }()
+	}
+}
+
+func (p *flakyProxy) dropAll() {
+	p.mu.Lock()
+	defer p.mu.Unlock()
+	for c := range p.conns {
+		_ = c.Close()
+	}
+	clear(p.conns)
+}
+
+// cut starts the outage and returns once the runner has been refused at least refusals more times.
+func (p *flakyProxy) cut(t *testing.T, refusals int64) {
+	t.Helper()
+	before := p.refused.Load()
+	p.down.Store(true)
+	p.dropAll()
+	deadline := time.Now().Add(30 * time.Second)
+	for p.refused.Load() < before+refusals {
+		if time.Now().After(deadline) {
+			t.Fatal("runner never retried during the outage")
+		}
+		time.Sleep(pollInterval)
+	}
+}
+
+func (p *flakyProxy) restore() { p.down.Store(false) }
+
+// testReconnect checks that a runner recovers from a lost connection to Gitea, idle or mid-job.
+func testReconnect(t *testing.T) {
+	t.Parallel()
+	api, repo := newScenario(t)
+	proxy := newFlakyProxy(t, fixture.baseURL)
+	startRunner(t, repo, "e2e-reconnect", runnerOptions{host: true, address: proxy.URL()})
+
+	proxy.cut(t, 2)
+	pushWorkflow(t, api, repo, "reconnect_idle.yml")
+	idle := waitForRun(t, api, repo)
+	proxy.restore()
+	requireSuccess(t, api, repo, idle.ID)
+
+	pushWorkflow(t, api, repo, "reconnect.yml")
+	slow := waitForRunAfter(t, api, repo, idle.ID)
+	waitForRunningJobLog(t, api, repo, slow.ID, "e2e-before-outage")
+	// the job logs a line per second, so a long outage spans several that can only arrive after it
+	proxy.cut(t, 8)
+	proxy.restore()
+	requireSuccess(t, api, repo, slow.ID)
+	logs := runLogs(t, api, repo, slow.ID)
+	for i := 1; i <= 15; i++ {
+		if tick := fmt.Sprintf("tick-%02d", i); commandRow(logs, tick) != tick {
+			t.Errorf("log line %s written during or after the outage was lost:\n%s", tick, logs)
+			break
+		}
+	}
+}