From: Borislav Petkov Subject: Re: [Part1 PATCH v5 02/17] x86/mm: Add Secure Encrypted Virtualization (SEV) support Date: Fri, 29 Sep 2017 16:41:46 +0200 Message-ID: <20170929144146.ctidwarilxntpu3i@pd.tnic> References: <20170927151329.70011-1-brijesh.singh@amd.com> <20170927151329.70011-3-brijesh.singh@amd.com> <20170928090242.ber7gynwaldinafa@pd.tnic> <20170928192347.gkpmpetmmp7523wu@pd.tnic> <0ed8db87-3f18-fd69-19e3-7f23a43bb8d2@amd.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Cc: kvm@vger.kernel.org, linux-kernel@vger.kernel.org, x86@kernel.org, Tom Lendacky , Thomas Gleixner , Ingo Molnar , "H. Peter Anvin" , Andy Lutomirski To: Brijesh Singh Return-path: Content-Disposition: inline In-Reply-To: <0ed8db87-3f18-fd69-19e3-7f23a43bb8d2@amd.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: kvm.vger.kernel.org On Fri, Sep 29, 2017 at 07:28:47AM -0500, Brijesh Singh wrote: > if we are adding a chicken bits then I think we should do it for both > "smeonly" and "sevonly". We can boot host OS with SME disabled and SEV > enabled, and still be able to create the SEV guest from the hypervisor. Sure, but is that a real use case? I mean, who would want to run encrypted guests on an unencrypted hypervisor? > How about this ? > > mem_encrypt=on     both SME and SEV enabled > mem_encrypt=sev    only SEV enabled > mem_encrypt=sme   only SME enabled > mem_encrypt=off     neither SME/SEV are enabled I like those short mnemonics, ACK. Less typing is always good. Thx. -- Regards/Gruss, Boris. SUSE Linux GmbH, GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) --