From: Ravikiran G Thirumalai <kiran@scalex86.org>

Following change places each element of the futex_queues hashtable on a
cacheline otherwise.

akpm: It's expected t0 consume 32k on bigsmp, 16k on smallsmp and 1MB on
CONFIG_X86_VSMP.  But the patch sped up a microbenchmark by 30% on
CONFIG_X86_VSMP, so the scalex guys feel it's a reasonable tradeoff.

Signed-off-by: Ravikiran Thirumalai <kiran@scalex86.org>
Signed-off-by: Shai Fultheim <shai@scalex86.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
---

 kernel/futex.c |    5 +++--
 1 files changed, 3 insertions(+), 2 deletions(-)

diff -puN kernel/futex.c~cache-align-futex-hash-buckets kernel/futex.c
--- devel/kernel/futex.c~cache-align-futex-hash-buckets	2006-02-27 20:58:29.000000000 -0800
+++ devel-akpm/kernel/futex.c	2006-02-27 20:58:29.000000000 -0800
@@ -100,9 +100,10 @@ struct futex_q {
 struct futex_hash_bucket {
        spinlock_t              lock;
        struct list_head       chain;
-};
+} ____cacheline_internodealigned_in_smp;
 
-static struct futex_hash_bucket futex_queues[1<<FUTEX_HASHBITS];
+static struct futex_hash_bucket futex_queues[1<<FUTEX_HASHBITS] 
+				__cacheline_aligned_in_smp;
 
 /* Futex-fs vfsmount entry: */
 static struct vfsmount *futex_mnt;
_