* cache totally. This means that the cache becomes inconsistent, and,
* since we use normal loads/stores as well, this is really bad.
* Typically, this causes oopsen in filp_close, but could have other,
- * more disasterous effects. There are two work-arounds:
+ * more disastrous effects. There are two work-arounds:
* 1. Disable interrupts and emulate the atomic swap
* 2. Clean the cache, perform atomic swap, flush the cache
*