Re: [PATCH 1/8] drivers/random: Cache align ip_random better
From: Matt Mackall
Date: Wed Mar 16 2011 - 15:55:19 EST
On Wed, 2011-03-16 at 20:26 +0100, Eric Dumazet wrote:
> > I think this alignment exists to minimize the number of cacheline
> > bounces on SMP as this can be a pretty hot structure in the network
> > stack. It could probably benefit from a per-cpu treatment.
> Well, this is a mostly read area of memory, dirtied every 5 minutes.
> Compare this to 'jiffies' for example ;)
> What could be done is to embed 'ip_cnt' inside ip_keydata for
> example, to avoid wasting a cache line for one bit ;)
> c1606c40 b ip_cnt
> c1606c80 b ip_keydata
Yeah. I actually think we're due for rethinking this entire process. It
dates back to when we introduced syncookies in the 90s and it shows. The
fact that we've started abusing it for other things doesn't help.
Mathematics is the supreme nostalgia of our time.
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/