Eric Dumazet [off-list ref] writes:

1) Introduce hlist_nulls variant of hlist

  hlist uses NULL value to finish a chain.
  hlist_nulls variant use the low order bit set to 1 to signal an end marker.
  This allows to store many different end markers, so that some RCU lockless
  algos (used in TCP/UDP stack for example) can save some memory barriers in
  fast paths.

Do you have any numbers that demonstrate the read memory barriers being
a performance problem? At least on x86 they should be very cheap
because they're normally nops.

-Andi

-- 
ak@linux.intel.com