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/failure_test.go

@@ -5,55 +5,41 @@
 
 package e2e
 
-import "testing"
+import (
+	"testing"
+	"time"
+)
 
-func testFailurePaths(t *testing.T) {
+func testFailurePaths(t *testing.T, m mode) {
+	result := runWorkflow(t, m, workflowSpec{files: []string{"failure.yml"}})
+	if result.conclusion != "failure" {
+		t.Errorf("run concluded %q, want failure", result.conclusion)
+	}
+	for job, want := range map[string]string{"fail": "failure", "tolerated": "success", "dependent": "skipped"} {
+		if got := result.job(job); got != want {
+			t.Errorf("job %q concluded %q, want %q", job, got, want)
+		}
+	}
+	for _, marker := range []string{"e2e-before-fail", "e2e-failure-cleanup", "e2e-tolerated-next"} {
+		if !result.hasLine(marker) {
+			t.Errorf("logs are missing %q", marker)
+		}
+	}
+	if result.hasLine("e2e-skipped-step") {
+		t.Error("a success() step ran after a failed step")
+	}
+}
+
+func testJobTimeout(t *testing.T) {
 	t.Parallel()
-	ctx := t.Context()
-
-	api, repo := newScenario(t)
-	pushWorkflow(t, api, repo, "failure.yml")
-
+	api, repo, _ := startIsolatedScenario(t, "timeout.yml", "e2e-timeout", runnerOptions{timeout: 15 * time.Second})
 	wfRun := waitForRun(t, api, repo)
-	completed, err := api.WaitForRunConclusion(ctx, repo, wfRun.ID, runTimeout)
+	completed, err := api.WaitForRunConclusion(t.Context(), repo, wfRun.ID, runTimeout)
 	if err != nil {
 		dumpRunLogs(t, api, repo, wfRun.ID)
 		t.Fatalf("wait for run: %v", err)
 	}
-	if completed.Conclusion != "failure" {
-		t.Errorf("run concluded %q, want failure", completed.Conclusion)
-	}
-
-	jobs, err := api.Jobs(ctx, repo, wfRun.ID)
-	if err != nil {
-		t.Fatalf("list jobs: %v", err)
-	}
-	want := map[string]string{
-		"fail":      "failure",
-		"tolerated": "success",
-		"dependent": "skipped",
-		"timeout":   "failure",
-	}
-	for _, job := range jobs {
-		if w, ok := want[job.Name]; ok && job.Conclusion != w {
-			t.Errorf("job %q concluded %q, want %q", job.Name, job.Conclusion, w)
-		}
-		delete(want, job.Name)
-	}
-	for name := range want {
-		t.Errorf("job %q not found", name)
-	}
-
-	logs := runLogs(t, api, repo, wfRun.ID)
-	for _, marker := range []string{"e2e-before-fail", "e2e-failure-cleanup", "e2e-tolerated-next", "e2e-timeout-started"} {
-		if commandRow(logs, marker) != marker {
-			t.Errorf("logs are missing %q", marker)
-		}
-	}
-	if commandRow(logs, "e2e-skipped-step") != "" {
-		t.Error("a success() step ran after a failed step")
-	}
-	if t.Failed() {
-		t.Logf("full stored logs:\n%s", logs)
+	if logs := runLogs(t, api, repo, wfRun.ID); completed.Conclusion != "failure" || commandRow(logs, "e2e-timeout-started") == "" {
+		t.Errorf("run concluded %q, want failure after the job started, logs:\n%s", completed.Conclusion, logs)
 	}
 }