perf: speed up job setup, action preparation and the cache server (#1327)

- Look up containers by name instead of listing all of them
- On Linux, cache the docker proxy discovery instead of starting a probe container per job
- Download a job's actions concurrently and share identical fetches between concurrent jobs
- Cache `runner.arch` per run and poll service health every 250ms
- Cache server: queue on an in-process lock instead of polling the database file lock, upload S3 objects up to 64MiB in one request and larger ones in 4 concurrent parts

| `exec`, median of 5 | Before | After |
|---|---:|---:|
| 1 step | 1.70s | 0.87s |
| 10 steps | 2.84s | 1.97s |
| Service + step | 3.78s | 2.18s |
| checkout + Node action | 3.64s | 2.52s |

Reviewed-on: https://gitea.com/gitea/runner/pulls/1327
Reviewed-by: Zettat123 <[email protected]>

act/runner/expression_test.go

@@ -6,6 +6,7 @@ package runner
 
 import (
 	"strings"
+	"sync"
 	"testing"
 
 	"gitea.dev/actionslib/pkg/exprparser"
@@ -500,3 +501,22 @@ jobs:
 	assert.NotContains(t, a, "s3cr3t") // it reaches the container name, which no log masker covers
 	assert.NotEqual(t, a, b)           // masking the name must not collapse two jobs onto one container
 }
+
+func TestNewExpressionEvaluatorReadsJobOutputsWhileCombinationsWriteThem(t *testing.T) {
+	rc := createRunContext(t)
+	rc.caller = &caller{runContext: createRunContext(t)}
+	rc.ExprEval = rc.NewExpressionEvaluator(t.Context())
+	job := rc.Run.Job()
+	var writes sync.WaitGroup
+	writes.Go(func() {
+		for range 100 {
+			unlock := lockJob(job)
+			job.Outputs = map[string]string{"out": "value"}
+			unlock()
+		}
+	})
+	for range 100 {
+		rc.NewExpressionEvaluator(t.Context())
+	}
+	writes.Wait()
+}