Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S935611AbXHIMSk (ORCPT ); Thu, 9 Aug 2007 08:18:40 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1762911AbXHIMS3 (ORCPT ); Thu, 9 Aug 2007 08:18:29 -0400 Received: from ms-smtp-03.nyroc.rr.com ([24.24.2.57]:47088 "EHLO ms-smtp-03.nyroc.rr.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1763610AbXHIMSR (ORCPT ); Thu, 9 Aug 2007 08:18:17 -0400 Date: Thu, 9 Aug 2007 08:17:26 -0400 (EDT) From: Steven Rostedt X-X-Sender: rostedt@gandalf.stny.rr.com To: Jeremy Fitzhardinge cc: Andi Kleen , Glauber de Oliveira Costa , LKML , akpm@linux-foundation.org, rusty@rustcorp.com.au, Ingo Molnar , chrisw@sous-sol.org, avi@qumranet.com, anthony@codemonkey.ws, virtualization@lists.linux-foundation.org, lguest@ozlabs.org Subject: Re: [PATCH 18/25] [PATCH] turn priviled operations into macros in entry.S In-Reply-To: <46BAA447.9010504@goop.org> Message-ID: References: <11865467522495-git-send-email-gcosta@redhat.com> <200708081138.23018.ak@suse.de> <200708081424.49896.ak@suse.de> <20070808133001.GD14419@one.firstfloor.org> <46BAA447.9010504@goop.org> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1188 Lines: 29 -- On Wed, 8 Aug 2007, Jeremy Fitzhardinge wrote: > Steven Rostedt wrote: > > /* > > * x86 arch doesn't have an easy way to find out where > > * gs is located. So we need to read the MSR. But first > > * we need to save off the rcx, rax and rdx. > > > Why don't you store it in gs? movq %gs:my_gs_base, %rax? Because it can't be trusted. After the swapgs, we are pointing to the RW section of the HV. But by running the guest kernel in ring 1, we have no protection from the guest writing into that area too. So we can't put anything in that section and expect it to be safe after jumping to guest code. The only trusted values, is jumping in after getting there by an interrupt, where the hardware places the values onto the stack. But with Andi's comments, I realized I can point the gs pointer to the RO area. And make a constant offset that will point up into the RW area, so we could save the stack and replace it. -- Steve - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/