From mboxrd@z Thu Jan 1 00:00:00 1970 From: Evgeniy Polyakov Subject: Re: [PATCH 1/1] network memory allocator. Date: Tue, 15 Aug 2006 14:02:28 +0400 Message-ID: <20060815100228.GC1092@2ka.mipt.ru> References: <20060814110359.GA27704@2ka.mipt.ru> <20060815002724.a635d775.akpm@osdl.org> Mime-Version: 1.0 Content-Type: text/plain; charset=koi8-r Cc: Andrew Morton , David Miller , netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-mm@kvack.org Return-path: Received: from relay.2ka.mipt.ru ([194.85.82.65]:32468 "EHLO 2ka.mipt.ru") by vger.kernel.org with ESMTP id S965352AbWHOKDO (ORCPT ); Tue, 15 Aug 2006 06:03:14 -0400 To: Andi Kleen Content-Disposition: inline In-Reply-To: Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Tue, Aug 15, 2006 at 10:08:23AM +0200, Andi Kleen (ak@suse.de) wrote: > Andrew Morton writes: > > > > There will be heaps of cacheline pingpong accessing these arrays. I'd have > > though that > > > > static struct whatever { > > avl_t avl_node_id; > > struct avl_node **avl_node_array; > > struct list_head *avl_container_array; > > struct avl_node *avl_root; > > struct avl_free_list *avl_free_list_head; > > spinlock_t avl_free_lock; > > } __cacheline_aligned_in_smp whatevers[NR_CPUS]; > > > > would be better. > > Or even better per cpu data. New global/static NR_CPUS arrays should be really discouraged. I had a version with per-cpu data - it is not very convenient to use here with it's per_cpu_ptr dereferencings.... > -Andi -- Evgeniy Polyakov