From mboxrd@z Thu Jan 1 00:00:00 1970 From: Viresh Kumar Subject: Re: [PATCH 00/18] cpufreq: don't loose cpufreq history on CPU hotplug Date: Tue, 27 Jan 2015 20:29:30 +0530 Message-ID: References: <18969256.kkPtbfLplS@vostro.rjw.lan> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Return-path: Received: from mail-ob0-f169.google.com ([209.85.214.169]:56310 "EHLO mail-ob0-f169.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754290AbbA0O7a (ORCPT ); Tue, 27 Jan 2015 09:59:30 -0500 Received: by mail-ob0-f169.google.com with SMTP id va8so13840964obc.0 for ; Tue, 27 Jan 2015 06:59:30 -0800 (PST) In-Reply-To: <18969256.kkPtbfLplS@vostro.rjw.lan> Sender: linux-pm-owner@vger.kernel.org List-Id: linux-pm@vger.kernel.org To: "Rafael J. Wysocki" Cc: Linaro Kernel Mailman List , "linux-pm@vger.kernel.org" , Stephen Boyd , Prarit Bhargava , Saravana Kannan On 27 January 2015 at 20:36, Rafael J. Wysocki wrote: > OK, and this is the last one I'm going to consider for 3.20 (in case you have > more of them queued up somewhere). Okay.. >> To make things simple we stop playing with sysfs files unless the driver is >> getting removed. > > And uses will see CPUs that are not present any more? s/uses/users ? Actually I have never seen such a situation on the ARM world and so don't know what exactly happens here. You are probably talking about the NUMA server space, where a cluster of CPUs can be removed physically from the NUMA network of CPUs. What actually happens in that case? Similar to echo 0 > online ? And do we support it currently? FWIW, I haven't paid any attention to side of the problem yet, but as soon as I understand it, I can fix it. Any tips on how can I emulate that on platforms where we can't take out the CPUs physically ?