fix(ci): scan all interfaces and full /24 to locate DinD daemon

The previous fallback only scanned .1-.50 of the first interface's
subnet, missing the DinD container when its IP is higher (.51+) or
when the forgejo-jobs network is on a different interface than
hostname -I returned first.

Now iterates all non-loopback IPs from hostname -I, scans each
subnet's full /24 (.1-254), and uses a 0.3 s bash /dev/tcp probe
instead of nc -zw1 to keep the total scan time under ~80 s.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

.forgejo/workflows/ci.yml

@@ -53,17 +53,20 @@ jobs:
             exit 0
           fi
           # Codeberg's act runner may not register the service container with the
-          # 'docker' DNS alias.  Fall back: scan the local /24 subnet for port 2375.
+          # 'docker' DNS alias.  Scan all local /24 subnets (one per runner
+          # interface) for any host accepting connections on port 2375.
           echo "Docker unreachable via hostname; scanning network for DinD..." >&2
-          MY_IP=$(hostname -I | awk '{print $1}')
-          PREFIX=$(echo "$MY_IP" | cut -d. -f1-3)
           FOUND=""
-          for i in $(seq 1 50); do
-            ip="${PREFIX}.${i}"
-            if nc -zw1 "$ip" 2375 2>/dev/null; then
-              FOUND="$ip"
-              break
-            fi
+          for MY_IP in $(hostname -I); do
+            case "$MY_IP" in 127.*) continue ;; esac
+            PREFIX=$(echo "$MY_IP" | cut -d. -f1-3)
+            for i in $(seq 1 254); do
+              ip="${PREFIX}.${i}"
+              if (timeout 0.3 bash -c "echo >/dev/tcp/${ip}/2375" 2>/dev/null); then
+                FOUND="$ip"
+                break 2
+              fi
+            done
           done
           if [ -z "$FOUND" ]; then
             echo "ERROR: Could not locate Docker daemon on the network" >&2