From mboxrd@z Thu Jan 1 00:00:00 1970 From: Chao Fan Subject: Re: [POC] Get acpi table to limit memory scope chosen by KASLR to home SB when "movable_node" is specified Date: Thu, 3 Aug 2017 17:03:59 +0800 Message-ID: <20170803090359.GF25014@localhost.localdomain> References: <20170803085155.GE25014@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Return-path: Received: from [183.91.158.132] ([183.91.158.132]:46662 "EHLO heian.cn.fujitsu.com" rhost-flags-FAIL-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S1751067AbdHCJES (ORCPT ); Thu, 3 Aug 2017 05:04:18 -0400 Content-Disposition: inline In-Reply-To: Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: Dou Liyang Cc: linux-kernel@vger.kernel.org, x86@kernel.org, hpa@zytor.com, tglx@linutronix.de, mingo@redhat.com, keescook@chromium.org, bhe@redhat.com, dyoung@redhat.com, arnd@arndb.de, dave.jiang@intel.com, indou.takao@jp.fujitsu.com, izumi.taku@jp.fujitsu.com, "Rafael J., Wysocki" , "Zheng, Lv" , ACPI Devel Maling List On Thu, Aug 03, 2017 at 04:59:26PM +0800, Dou Liyang wrote: >Cc ACPI maintainers and mail list So many thanks. > >At 08/03/2017 04:51 PM, Chao Fan wrote: >> Hi all, >> >> Here is an issue about movable_node and KASLR. >> Currently, KASLR works well with "mem=", by which we can limit the scope of >> memory chosen by KASLR. But if we use "movable_node" independently without >> "mem=", we don't know the scope of memory in the home SB, which will cause >> the memory of hotplug SB may be chosen by KASLR. >> To solve this problem, I want to do this job: >> If "movable_node" is specified, try to get the acpi table and find the scope >> of memory in home SB, then set "mem_limit". >> >> The order of getting table is: >> boot_params >> \-systab >> \-efi tables >> \-rsdp >> \-rsdt or xsdt >> \-acpi root table list >> \-acpi srat mem affinity >> I am not sure if it's good to do this job in this period, so I want to get the feedback of community. Thanks, Chao Fan >> I am not sure if it's OK for community. >> If there are any problems, please let me know. >> Any comments will be welcome. >> >> Thanks, >> Chao Fan >> >> >>