Use a big-lock for kernel sychronization instead of fine-grained locking

kernel/device/storage/ramdrv.c

@@ -116,24 +116,16 @@ DEFINE_DEVICE_OP(ramdrv_write) {
 }
 
 DEFINE_DEVICE_OP(ramdrv_partition_rescan) {
-  uint64_t fsd;
-
   struct list_node_link *subdevice_link, *tmp_subdevice_link;
   list_foreach(device->subdevices, subdevice_link, tmp_subdevice_link) {
     struct device* subdevice = list_entry(subdevice_link, struct device, subdevices_link);
 
-    spin_lock(&subdevice->lock, &fsd);
     list_remove(device->subdevices, &subdevice->subdevices_link);
-    spin_unlock(&subdevice->lock, fsd);
 
     device_delete(subdevice->key, proc, rctx);
   }
 
-  spin_unlock(&device->lock, *lockflags);
-
   int r = device_probe_partitions(proc, rctx, device);
 
-  spin_lock(&device->lock, lockflags);
-
   return r;
 }