Fix S3 concurrent map writes causing pod crashes

PROBLEM:
- Pod crashes with "fatal error: concurrent map writes" during S3 publications
- Root cause: pathCache map in PublishedStorage accessed without synchronization
- Occurs during concurrent LinkFromPool operations in S3 publishing

SOLUTION:
- Add sync.RWMutex to PublishedStorage struct for thread-safe map access
- Implement double-check locking pattern for cache initialization
- Protect all map operations (read/write/delete) with appropriate locks

CHANGES:
- s3/public.go: Add pathCacheMutex field and protect all map operations
  * Cache initialization with double-check locking in LinkFromPool
  * Read operations protected with RLock/RUnlock
  * Write operations protected with Lock/Unlock
  * Delete operations in Remove() and RemoveDirs() protected

IMPACT:
- Eliminates concurrent map writes panic
- Prevents pod crashes during S3 publications
- Maintains performance with minimal synchronization overhead
- Uses read-write locks allowing concurrent reads while serializing writes

.github/workflows/test-act.yml

@@ -1,12 +0,0 @@
-name: Test Act
-on: [push]
-jobs:
-  test:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-      - name: Test
-        run: |
-          echo "Hello from act\!"
-          pwd
-          ls -la