From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andi Kleen Subject: Re: [net-next 02/11] igb: Use node specific allocations for the q_vectors and rings Date: Sun, 09 Oct 2011 11:08:47 -0700 Message-ID: References: <1318056461-19562-1-git-send-email-jeffrey.t.kirsher@intel.com> <1318056461-19562-3-git-send-email-jeffrey.t.kirsher@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: davem@davemloft.net, Alexander Duyck , netdev@vger.kernel.org, gospo@redhat.com, sassmann@redhat.com To: Jeff Kirsher Return-path: Received: from mga14.intel.com ([143.182.124.37]:49376 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752017Ab1JISIt (ORCPT ); Sun, 9 Oct 2011 14:08:49 -0400 In-Reply-To: <1318056461-19562-3-git-send-email-jeffrey.t.kirsher@intel.com> (Jeff Kirsher's message of "Fri, 7 Oct 2011 23:47:32 -0700") Sender: netdev-owner@vger.kernel.org List-ID: Jeff Kirsher writes: > > for (i = 0; i < adapter->num_tx_queues; i++) { > - ring = kzalloc(sizeof(struct igb_ring), GFP_KERNEL); > + if (orig_node == -1) { > + int cur_node = next_online_node(adapter->node); > + if (cur_node == MAX_NUMNODES) > + cur_node = first_online_node; RR seems quite arbitrary. Who guarantees those nodes have any relationship with the CPUs submitting on those queues? Or the node the device is on. Anyways if it's a good idea probably need to add a dma_alloc_coherent_node() too -Andi -- ak@linux.intel.com -- Speaking for myself only