From mboxrd@z Thu Jan 1 00:00:00 1970 From: Lukas Wunner Subject: Re: kmem_cache_alloc fail with unable to handle paging request after pci hotplug remove. Date: Tue, 5 Jul 2016 09:53:46 +0200 Message-ID: <20160705075346.GA2918@wunner.de> References: <577A7203.9010305@linux.intel.com> <577A7B0A.4090107@linux.intel.com> <20160704152131.GA2766@wunner.de> <577A84B4.8020505@linux.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mailout1.hostsharing.net ([83.223.95.204]:60107 "EHLO mailout1.hostsharing.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751052AbcGEHtX (ORCPT ); Tue, 5 Jul 2016 03:49:23 -0400 Content-Disposition: inline In-Reply-To: Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: AceLan Kao Cc: Mathias Nyman , "Rafael J. Wysocki" , Bjorn Helgaas , Andrew Morton , Linux Memory Management List , LKML , Linux PCI , ACPI Devel Maling List , "Kirill A. Shutemov" , USB On Tue, Jul 05, 2016 at 11:00:21AM +0800, AceLan Kao wrote: > These are logs from my machine. > > *** Before plug-in the USB key > > u@u-XPS-13-9xxx:~$ sudo lspci -vvvv -s 00:1c.0 > 00:1c.0 PCI bridge: Intel Corporation Device 9d10 (rev f1) (prog-if 00 > [Normal decode]) [...] > LnkSta: Speed 2.5GT/s, Width x0, TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt- > SltCap: AttnBtn- PwrCtrl- MRL- AttnInd- PwrInd- HotPlug+ Surprise+ The link is down (DLActive-), but the root port is a hotplug port, so apparently with Alpine Ridge the controller is powered down if nothing is plugged in and this results in the controller being "unplugged" from the root port. This looks less fishy than I originally thought, it's just very different from the power management of pre Alpine Ridge controllers on Macs (which is the only thing I'm really familiar with), where the root port is not a hotplug port and the controller does not disappear from the system when powered down. (It's config space just becomes inaccessible.) Best regards, Lukas