test(e2e): speed up and cover exec and Kubernetes modes (#1363)

Follow-up to https://gitea.com/gitea/runner/pulls/1362, whose crash e2e missed because it only drove the daemon on Docker.

- `dispatch`, `failure_paths` and `reusable_workflow` run through the daemon and `exec`, each on Docker and one shared k3s cluster.
- Fixed sleeps become gates the test releases, the job timeout uses a 15s runner deadline, k3s boots alongside Gitea, Gitea's data is on tmpfs.
- Outside a container, `make test-e2e` runs `go test` in a Linux container on its own network like CI, so local runs take the same paths.

Local wall time:

| | Before | After |
|---|---|---|
| Suite | 84s | 27–35s |
| `job_timeout` | 64s | 17s |
| `reconnect` | 22s | 4.5s |
| `capacity/serial` | 20s | 2s |
| `cli_shutdown_drains` | 14s | 2.3s |
| `kubernetes` | 31s | 21–23s |

_Written by Claude (Opus 5.5)._

---------

Co-authored-by: bircni <[email protected]>
Reviewed-on: https://gitea.com/gitea/runner/pulls/1363
Reviewed-by: bircni <[email protected]>
Co-authored-by: silverwind <[email protected]>

e2e/reconnect_test.go

@@ -10,6 +10,8 @@ import (
 	"io"
 	"net"
 	"net/url"
+	"strconv"
+	"strings"
 	"sync"
 	"sync/atomic"
 	"testing"
@@ -101,6 +103,10 @@ func (p *flakyProxy) restore() { p.down.Store(false) }
 func testReconnect(t *testing.T) {
 	t.Parallel()
 	api, repo := newScenario(t)
+	gate, release := newGate(t)
+	if err := api.CreateVariable(t.Context(), repo, "E2E_GATE", gate); err != nil {
+		t.Fatalf("create gate variable: %v", err)
+	}
 	proxy := newFlakyProxy(t, fixture.baseURL)
 	startRunner(t, repo, "e2e-reconnect", runnerOptions{host: true, address: proxy.URL()})
 
@@ -112,14 +118,19 @@ func testReconnect(t *testing.T) {
 
 	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
+	waitForRunningJobLog(t, api, repo, slow.ID, "e2e-before-outage", 1)
+	// the job logs a line every 200ms until released, so the outage spans several that can only arrive after it
 	proxy.cut(t, 8)
 	proxy.restore()
+	release()
 	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 {
+	ticks, err := strconv.Atoi(strings.TrimPrefix(commandRow(logs, "ticks="), "ticks="))
+	if err != nil {
+		t.Fatalf("the job did not report its tick count: %v\n%s", err, logs)
+	}
+	for i := 1; i <= ticks; i++ {
+		if tick := fmt.Sprintf("tick-%d", i); commandRow(logs, tick) != tick {
 			t.Errorf("log line %s written during or after the outage was lost:\n%s", tick, logs)
 			break
 		}