Eric Dumazet 6c9ae009b2 irq: use per_cpu kstat_irqs
Use modern per_cpu API to increment {soft|hard}irq counters, and use
per_cpu allocation for (struct irq_desc)->kstats_irq instead of an array.

This gives better SMP/NUMA locality and saves few instructions per irq.

With small nr_cpuids values (8 for example), kstats_irq was a small array
(less than L1_CACHE_BYTES), potentially source of false sharing.

In the !CONFIG_SPARSE_IRQ case, remove the huge, NUMA/cache unfriendly
kstat_irqs_all[NR_IRQS][NR_CPUS] array.

Note: we still populate kstats_irq for all possible irqs in
early_irq_init().  We probably could use on-demand allocations.  (Code
included in alloc_descs()).  Problem is not all IRQS are used with a prior
alloc_descs() call.

kstat_irqs_this_cpu() is not used anymore, remove it.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-01-13 17:32:31 -08:00
..
2010-12-22 18:57:02 +01:00
2011-01-13 17:32:31 -08:00
2010-08-09 16:48:42 -04:00
2010-07-14 11:29:46 +02:00
2010-10-30 01:42:19 -04:00
2010-11-03 13:49:58 -04:00
2010-10-30 01:42:19 -04:00
2010-10-30 08:45:43 -04:00
2011-01-12 20:02:47 -05:00
2010-10-29 04:17:06 -04:00
2009-09-18 09:48:52 -07:00
2010-11-10 15:01:34 +01:00
2010-12-16 11:36:43 +01:00
2010-11-01 15:38:34 -04:00
2010-10-01 10:50:58 -07:00
2010-10-18 18:44:26 +02:00
2010-08-11 23:04:20 +09:30
2010-08-19 17:18:02 -07:00
2010-11-18 13:27:46 +01:00
2011-01-13 08:03:08 -08:00
2011-01-13 08:03:18 -08:00
2009-09-23 18:13:10 -07:00
2010-12-29 11:31:38 -08:00
2009-06-18 13:03:55 -07:00