From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S934378AbeE2N1q (ORCPT ); Tue, 29 May 2018 09:27:46 -0400 Received: from mx3-rdu2.redhat.com ([66.187.233.73]:50268 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S934320AbeE2N1m (ORCPT ); Tue, 29 May 2018 09:27:42 -0400 Date: Tue, 29 May 2018 09:27:37 -0400 From: Luiz Capitulino To: Baoquan He Cc: Ingo Molnar , linux-kernel@vger.kernel.org, keescook@chromium.org, tglx@linutronix.de, x86@kernel.org, hpa@zytor.com, fanc.fnst@cn.fujitsu.com, yasu.isimatu@gmail.com, indou.takao@jp.fujitsu.com, douly.fnst@cn.fujitsu.com Subject: Re: [PATCH 0/2] x86/boot/KASLR: Skip specified number of 1GB huge pages when do physical randomization Message-ID: <20180529092737.10a8138b@doriath> In-Reply-To: <20180528095418.GD31261@MiWiFi-R3L-srv> References: <20180516100532.14083-1-bhe@redhat.com> <20180518070046.GA18660@gmail.com> <20180518074359.GR24627@MiWiFi-R3L-srv> <20180518081919.GB11379@gmail.com> <20180518112836.GS24627@MiWiFi-R3L-srv> <20180523151022.102e3565@doriath> <20180528095418.GD31261@MiWiFi-R3L-srv> Organization: Red Hat MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 28 May 2018 17:54:18 +0800 Baoquan He wrote: > On 05/23/18 at 03:10pm, Luiz Capitulino wrote: > > On Fri, 18 May 2018 19:28:36 +0800 > > Baoquan He wrote: > > > > > > Note that it's not KASLR specific: if we had some other kernel feature that tried > > > > to allocate a piece of memory from what appears to be perfectly usable generic RAM > > > > we'd have the same problems! > > > > > > Hmm, this may not be the situation for 1GB huge pages. For 1GB huge > > > pages, the bug is that on KVM guest with 4GB ram, when user adds > > > 'default_hugepagesz=1G hugepagesz=1G hugepages=1' to kernel > > > command-line, if 'nokaslr' is specified, they can get 1GB huge page > > > allocated successfully. If remove 'nokaslr', namely KASLR is enabled, > > > the 1GB huge page allocation failed. > > > > Let me clarify that this issue is not specific to KVM in any way. The same > > issue happens on bare-metal, but if you have lots of memory you'll hardly > > notice it. On the other hand, it's common to create KVM guests with a few > > GBs of memory. In those guests, you may not be able to get a 1GB hugepage > > at all if kaslr is enabled. > > > > This series is a simple fix for this bug. It hooks up into already existing > > KASLR code that scans memory regions to be avoided. The memory hotplug > > issue is left for another day. > > Exactly. > > This issue is about kernel being randomized into good 1GB huge pages to > break later huge page allocation, and we can only scan memory to know > where 1GB huge page is located and avoid them. > > The memory hotplug issue is about kernel being randomized into movable > memory regions, and we need read ACPI SRAT table to retrieve the > attribute of memory regions to know if it's movable, then avoid it if > yes. Makes sense. Since the KASLR code already scans memory regions looking for regions to skip and since this series just uses that, I think this is a good solution to the problem: Reviewed-and-Tested-by: Luiz Capitulino > > > > > Now, if I understand what Ingo is saying is that he wants to see all problems > > solved with a generic solution vs. a specific solution for each problem. > > Hmm, if we understand Ingo's words correctly, for these two issues, > seems there isn't a generic solution to solve both of them. We can only > fix them separately. > > Hi Ingo, > > Ping! > > Not sure if my above understanding is correct. Could you confirm if I > have understood your comments and if the solution of this patchset is > right? > > Thanks > Baoquan >