One of the many features added during the 2.5 development series was the
"futex" - a sort of fast, user-space mutual exclusion primitive.  In the
non-contended case, futexes can be obtained and released with no kernel
involvement at all, making them quite fast.  When contention does happen
(one process tries to obtain a futex currently owned by another), the
kernel is called in to queue any waiting processes and wake them up when
the futex becomes available.  When queueing is not needed, however, the
kernel maintains no knowledge of the futex, keeping its overhead low.