enhance: make shared tool cache safe at any capacity (#1401)

Co-authored-by: bircni <[email protected]>

internal/pkg/config/config.example.yaml

@@ -132,8 +132,8 @@ runner:
   # What to mount at RUNNER_TOOL_CACHE (/opt/hostedtoolcache), where setup actions install tools:
   #   none:   nothing. A docker job sees what its image ships there, a host job an empty dir, and
   #           either way what it installs is gone when the job ends.
-  #   shared: one tool cache every job reuses. Two jobs writing the same tool version at once
-  #           corrupt it, so use it only with capacity 1.
+  #   shared: one tool cache per capacity slot, reused by later jobs. It hides the tools later job
+  #           images ship, so use images without them.
   #tool_cache_mode: none
   # Optional executable on the host, run once after each task's built-in cleanup
   # (post-steps, container teardown, bind-workdir removal). Additive only.

internal/pkg/config/config.go

@@ -193,7 +193,7 @@ type Container struct {
 	Sweep                bool                          `yaml:"sweep"`                  // Sweep removes the containers, networks and volumes a job creates through its Docker proxy when the job ends.
 }
 
-// Values of Runner.ToolCacheMode: the runner mounts no tool cache, or one that every job reuses.
+// Values of Runner.ToolCacheMode: the runner mounts no tool cache, or one per capacity slot that its jobs reuse.
 const (
 	ToolCacheModeNone   = "none"
 	ToolCacheModeShared = "shared"
@@ -454,10 +454,6 @@ func LoadDefault(file string) (*Config, error) {
 	}
 
 	// Validate and fix invalid config combinations to prevent confusing behavior.
-	if cfg.Runner.ToolCacheMode == ToolCacheModeShared && cfg.Runner.Capacity > 1 {
-		logging.Warnf("runner.tool_cache_mode %q with capacity %d: two jobs writing the same tool version at once corrupt it",
-			ToolCacheModeShared, cfg.Runner.Capacity)
-	}
 	if cfg.Runner.FetchTimeout > RequestTimeout {
 		logging.Warnf("fetch_timeout (%v) exceeds the RPC timeout (%v), capping it", cfg.Runner.FetchTimeout, RequestTimeout)
 		cfg.Runner.FetchTimeout = RequestTimeout