From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759138Ab2EDR50 (ORCPT ); Fri, 4 May 2012 13:57:26 -0400 Received: from merlin.infradead.org ([205.233.59.134]:55989 "EHLO merlin.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753164Ab2EDR5Z convert rfc822-to-8bit (ORCPT ); Fri, 4 May 2012 13:57:25 -0400 Message-ID: <1336154226.6509.49.camel@twins> Subject: Re: [PATCH 4/9] perf: Generic intel uncore support From: Peter Zijlstra To: "Yan, Zheng" Cc: mingo@elte.hu, andi@firstfloor.org, eranian@google.com, jolsa@redhat.com, ming.m.lin@intel.com, linux-kernel@vger.kernel.org Date: Fri, 04 May 2012 19:57:06 +0200 In-Reply-To: <4FA3862E.7030203@intel.com> References: <1335924440-11242-1-git-send-email-zheng.z.yan@intel.com> <1335924440-11242-5-git-send-email-zheng.z.yan@intel.com> <1336065156.22523.34.camel@twins> <4FA3862E.7030203@intel.com> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7BIT X-Mailer: Evolution 3.2.2- Mime-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 2012-05-04 at 15:33 +0800, Yan, Zheng wrote: > Thank you for your suggestion. One reason I choose hash table is that I'm uncertain > the system behavior when hot-plug a CPU, the PCI uncore device is probed first or > the CPU data is initialized first? If the PCI device is probed first, I think your > code won't work because topology_physical_package_id may return incorrect id. It had better first initialize the cpu data before it starts probing its devices. Doing it the other way around just doesn't make any sense.