From mboxrd@z Thu Jan 1 00:00:00 1970 From: Viresh Kumar Subject: Re: [PATCH v4 1/1] PM / OPP: Fix get sharing cpus when hotplug is used Date: Thu, 27 Jul 2017 09:13:51 +0530 Message-ID: <20170727034351.GF352@vireshk-i7> References: <20170726121151.7576-1-waldemarx.rymarkiewicz@intel.com> <276e68a1-edfe-961b-e387-7855f409557b@arm.com> <592bc7b8-615d-5d80-f328-eabdfe6c3828@arm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mail-pf0-f171.google.com ([209.85.192.171]:35781 "EHLO mail-pf0-f171.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751502AbdG0Dn4 (ORCPT ); Wed, 26 Jul 2017 23:43:56 -0400 Received: by mail-pf0-f171.google.com with SMTP id e75so181239pfj.2 for ; Wed, 26 Jul 2017 20:43:55 -0700 (PDT) Content-Disposition: inline In-Reply-To: <592bc7b8-615d-5d80-f328-eabdfe6c3828@arm.com> Sender: linux-pm-owner@vger.kernel.org List-Id: linux-pm@vger.kernel.org To: Sudeep Holla Cc: Waldemar Rymarkiewicz , Waldemar Rymarkiewicz , linux-pm@vger.kernel.org, Viresh Kumar , Nishanth Menon , Stephen Boyd , "Rafael J. Wysocki" On 26-07-17, 15:08, Sudeep Holla wrote: > CPUs won't be unregistered on hotplug path You sure? I don't you will get confused with this as you already know all this, but I will still write it down for the sake of clarity. We aren't talking about offlining a CPU here, but a real physical hotplug. Why should the kernel keep a device structure for a device if it isn't physically present on the system? acpi_processor_remove() removes the CPU device and I thought it was all related to that hotplug. Am I missing something ? > and hence I can't figure out > why get_cpu_device would fail if all the CPUs were registered on boot. > Just trying to understand the scenario and how it differs from normal > hotplug case. -- viresh