From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hanjun Guo Subject: Re: [PATCH v3 0/3] ACPI: container hot remove support. Date: Tue, 27 Nov 2012 19:24:05 +0800 Message-ID: <50B4A2D5.2050603@huawei.com> References: <1351668471-31436-1-git-send-email-tangchen@cn.fujitsu.com> <509106E2.70008@jp.fujitsu.com> <50B3014E.3040107@huawei.com> <50B306E0.3050207@cn.fujitsu.com> <50B4127F.3080400@huawei.com> <50B42798.4030402@cn.fujitsu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <50B42798.4030402@cn.fujitsu.com> Sender: linux-pci-owner@vger.kernel.org To: Tang Chen Cc: Yasuaki Ishimatsu , "Rafael J. Wysocki" , yinghai@kernel.org, bhelgaas@google.com, lenb@kernel.org, jiang.liu@huawei.com, izumi.taku@jp.fujitsu.com, linux-acpi@vger.kernel.org, linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org, Wen Congyang , Toshi Kani List-Id: linux-acpi@vger.kernel.org On 2012/11/27 10:38, Tang Chen wrote: > On 11/27/2012 09:08 AM, Hanjun Guo wrote: >> On 2012/11/26 14:06, Tang Chen wrote: >>> On 11/26/2012 01:42 PM, Hanjun Guo wrote: >>>> >>>> Hi all, >>>> I think Yasuaki mentioned the key point for the container device r= emove, >>>> that is dependency. >>>> >>>> Currently, container, processor, and memory hotpulg are managed by= different ACPI >>>> hotplug drivers, the driver works when handle device hotplug indiv= idually, but they >>>> have no idea for each other. >>>> >>>> This may introduce some issues, such as Yasuaki mentioned above, t= hat is to say, we >>>> should remove its child before remove the device itself, and hot a= dd its parent before >>>> the device itself. >>>> >>>> According to the ACPI namespace, we can resolve most of dependency= issues. On a typical >>>> two processor sockets system, the namespace is like this: >>>> >>>> /_SB ---container device, with HID ACPI0004 >>>> |_SCK0 ---container device, with HID ACPI0004 >>>> |_CPU0 ---processor device, with HID ACPI0009 or LNXC= PU >>>> |_... >>>> |_CPUx >>>> |_MEM0 ---Memory device, with HID PNP0C80 >>>> |_SCK1 >>>> |_CPU0 >>>> |_... >>>> |_CPUx >>>> |_MEM1 >>>> |_PCI0 ---Host bridge, with HID PNP0A03 or PNP0A0= 8 >>>> >>>> If hot remove the container device, such as SCK0, we can easily kn= ow the dependency list >>>> is CPU0~CPUx and MEM0, but I think the ACPI hotplug driver haven't= resolve this now. >>>> >>>> And there is another corner case for hotplug devices in the namesp= ace above, that is: >>>> 1) Remove SCK0. yes, we can remove it with no dependency to the ho= st bridge PCI0; >>>> >>>> 2) Remove SCK1 after SCK0. we should remove the host bridge PCI0 f= irst, >>>> or the system will crash down. yes, dynamic dependency analys= is is needed here. >>>> and the ACPI hotplug driver totally have no idea of this. >>>> >>>> so, should we do something to settle this down ? >>> >>> Hi Guo, >>> >>> I am trying to do this too. :) >> >> Great=EF=BC=8C what's your idea of this? >> > Hi Guo, >=20 > I noticed they had a lot of discussion on this topic. > https://lkml.org/lkml/2012/11/15/159 >=20 > And Vasilis's latest patches are here: > https://lkml.org/lkml/2012/11/23/335 >=20 > I think we can have a review of this patchset first. :) Hi Tang, Thanks for your remind, I will review Vasilis's latest patches, and reply Vasilis's patch if I have any comments. >=20 > And also, as you said, the new ACPI hotplug framework from Liu Jiang > will settle this problem more properly. > So I think any solution now could be temporary. :) >=20 > . >=20 From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from szxga02-in.huawei.com ([119.145.14.65]:26443 "EHLO szxga02-in.huawei.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753636Ab2K0LYe (ORCPT ); Tue, 27 Nov 2012 06:24:34 -0500 Message-ID: <50B4A2D5.2050603@huawei.com> Date: Tue, 27 Nov 2012 19:24:05 +0800 From: Hanjun Guo MIME-Version: 1.0 To: Tang Chen CC: Yasuaki Ishimatsu , "Rafael J. Wysocki" , , , , , , , , , Wen Congyang , Toshi Kani Subject: Re: [PATCH v3 0/3] ACPI: container hot remove support. References: <1351668471-31436-1-git-send-email-tangchen@cn.fujitsu.com> <509106E2.70008@jp.fujitsu.com> <50B3014E.3040107@huawei.com> <50B306E0.3050207@cn.fujitsu.com> <50B4127F.3080400@huawei.com> <50B42798.4030402@cn.fujitsu.com> In-Reply-To: <50B42798.4030402@cn.fujitsu.com> Content-Type: text/plain; charset="UTF-8" Sender: linux-pci-owner@vger.kernel.org List-ID: On 2012/11/27 10:38, Tang Chen wrote: > On 11/27/2012 09:08 AM, Hanjun Guo wrote: >> On 2012/11/26 14:06, Tang Chen wrote: >>> On 11/26/2012 01:42 PM, Hanjun Guo wrote: >>>> >>>> Hi all, >>>> I think Yasuaki mentioned the key point for the container device remove, >>>> that is dependency. >>>> >>>> Currently, container, processor, and memory hotpulg are managed by different ACPI >>>> hotplug drivers, the driver works when handle device hotplug individually, but they >>>> have no idea for each other. >>>> >>>> This may introduce some issues, such as Yasuaki mentioned above, that is to say, we >>>> should remove its child before remove the device itself, and hot add its parent before >>>> the device itself. >>>> >>>> According to the ACPI namespace, we can resolve most of dependency issues. On a typical >>>> two processor sockets system, the namespace is like this: >>>> >>>> /_SB ---container device, with HID ACPI0004 >>>> |_SCK0 ---container device, with HID ACPI0004 >>>> |_CPU0 ---processor device, with HID ACPI0009 or LNXCPU >>>> |_... >>>> |_CPUx >>>> |_MEM0 ---Memory device, with HID PNP0C80 >>>> |_SCK1 >>>> |_CPU0 >>>> |_... >>>> |_CPUx >>>> |_MEM1 >>>> |_PCI0 ---Host bridge, with HID PNP0A03 or PNP0A08 >>>> >>>> If hot remove the container device, such as SCK0, we can easily know the dependency list >>>> is CPU0~CPUx and MEM0, but I think the ACPI hotplug driver haven't resolve this now. >>>> >>>> And there is another corner case for hotplug devices in the namespace above, that is: >>>> 1) Remove SCK0. yes, we can remove it with no dependency to the host bridge PCI0; >>>> >>>> 2) Remove SCK1 after SCK0. we should remove the host bridge PCI0 first, >>>> or the system will crash down. yes, dynamic dependency analysis is needed here. >>>> and the ACPI hotplug driver totally have no idea of this. >>>> >>>> so, should we do something to settle this down ? >>> >>> Hi Guo, >>> >>> I am trying to do this too. :) >> >> Great， what's your idea of this? >> > Hi Guo, > > I noticed they had a lot of discussion on this topic. > https://lkml.org/lkml/2012/11/15/159 > > And Vasilis's latest patches are here: > https://lkml.org/lkml/2012/11/23/335 > > I think we can have a review of this patchset first. :) Hi Tang, Thanks for your remind, I will review Vasilis's latest patches, and reply Vasilis's patch if I have any comments. > > And also, as you said, the new ACPI hotplug framework from Liu Jiang > will settle this problem more properly. > So I think any solution now could be temporary. :) > > . >