From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-oi1-f169.google.com (mail-oi1-f169.google.com [209.85.167.169]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 10B3668 for ; Fri, 12 Nov 2021 21:43:25 +0000 (UTC) Received: by mail-oi1-f169.google.com with SMTP id o83so20331279oif.4 for ; Fri, 12 Nov 2021 13:43:24 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20210112; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc:content-transfer-encoding; bh=p6BJgAR2dKpdI5PQSFC6gOgUXqAUQZJtadmObr7tjTY=; b=AnRQXnEvNzvr8lmbdoVBi472Kzz/I2xvy53cquSxtFszpAe3zph3LuOPEc5/48deXh xzvEzM60QjlA+Vpeqcr5k2Xs9hIpwxJHjlaBy+UJ+tVrUBkcD59RCdQ+/DT38Ot+4p9h v0AIhXq1nE9873XX18Cuk0CjaQh/m87fMJI7jgMcSqhzslXXtaw6E3pz1r6NeJVqvqhN K7/P/WRPlJwR11gNsOk5JE/VBsulqTP5MhsyHwJIlWm8Cg31Kz6C4pLrDMcWcLhAv+2h 1h3Rk1x09vCadHQBu9BU9p2uxYWPHVizVC0N06lkIufqe7Tq8CJVeVZLg5AIwOIU7Q0y 7plA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc:content-transfer-encoding; bh=p6BJgAR2dKpdI5PQSFC6gOgUXqAUQZJtadmObr7tjTY=; b=s8hNij1WruGD6M6JUd6/0GXlYoXFhLZgKMfRRJgBC2liBhJjHu1CnF2dKZe7jJ0I0S lCi8AIhRY//vP5/PJaF+TUUafDtSG9WVCrn0IBHaWoMEvN5gyQ3axNVz2HfhefNHFdB4 5GsGvtARNtY8uSmcklVLLIPI5XWx9ZY28QZENVqHqwPZV5afsdaCDZ1IinH1HNDk98rB MMOQh5XhMQHFar7oW3x8jiYQTv/IRovilCO+n/rfrLuFInE3P1ZnfSEAD/P2Je1mREGE AM+3y2ypHozYUXJjFu3LJI6e7gK4dg7L7ow0tsawzH/FrCbLhnI/XV4lIGVXyhL4fNLH 3cxQ== X-Gm-Message-State: AOAM532s1+uxsg4fCf9MRpO/m1fdEdAP87XyR54OvP988cZZiJvtlUWP HpYduA1QXrHBCNhXP1cKN+m+IWIVtBFopQAKz0m3ew== X-Google-Smtp-Source: ABdhPJwN1DMFj8lwZlramd1I0OGWe2kSl7QSbL+hhwnl5gfxV3O28d9EPTR8alNht5b3hCsOpj1PmUPMkqmyQaJMInE= X-Received: by 2002:aca:2319:: with SMTP id e25mr28956699oie.164.1636753404063; Fri, 12 Nov 2021 13:43:24 -0800 (PST) Precedence: bulk X-Mailing-List: linux-coco@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 References: <20210820155918.7518-1-brijesh.singh@amd.com> <061ccd49-3b9f-d603-bafd-61a067c3f6fa@intel.com> <2cb3217b-8af5-4349-b59f-ca4a3703a01a@www.fastmail.com> In-Reply-To: <2cb3217b-8af5-4349-b59f-ca4a3703a01a@www.fastmail.com> From: Marc Orr Date: Fri, 12 Nov 2021 13:43:13 -0800 Message-ID: Subject: Re: [PATCH Part2 v5 00/45] Add AMD Secure Nested Paging (SEV-SNP) Hypervisor Support To: Andy Lutomirski Cc: Sean Christopherson , Borislav Petkov , Dave Hansen , Peter Gonda , Brijesh Singh , "the arch/x86 maintainers" , Linux Kernel Mailing List , kvm list , linux-coco@lists.linux.dev, linux-mm@kvack.org, Linux Crypto Mailing List , Thomas Gleixner , Ingo Molnar , Joerg Roedel , Tom Lendacky , "H. Peter Anvin" , Ard Biesheuvel , Paolo Bonzini , Vitaly Kuznetsov , Wanpeng Li , Jim Mattson , Dave Hansen , Sergio Lopez , "Peter Zijlstra (Intel)" , Srinivas Pandruvada , David Rientjes , Dov Murik , Tobin Feldman-Fitzthum , Michael Roth , Vlastimil Babka , "Kirill A . Shutemov" , Andi Kleen , Tony Luck , Sathyanarayanan Kuppuswamy Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable On Fri, Nov 12, 2021 at 1:39 PM Andy Lutomirski wrote: > > > > On Fri, Nov 12, 2021, at 1:30 PM, Marc Orr wrote: > > On Fri, Nov 12, 2021 at 12:38 PM Sean Christopherson wrote: > >> > >> On Fri, Nov 12, 2021, Borislav Petkov wrote: > >> > On Fri, Nov 12, 2021 at 07:48:17PM +0000, Sean Christopherson wrote: > >> > > Yes, but IMO inducing a fault in the guest because of _host_ bug i= s wrong. > >> > > >> > What do you suggest instead? > >> > >> Let userspace decide what is mapped shared and what is mapped private.= The kernel > >> and KVM provide the APIs/infrastructure to do the actual conversions i= n a thread-safe > >> fashion and also to enforce the current state, but userspace is the co= ntrol plane. > >> > >> It would require non-trivial changes in userspace if there are multipl= e processes > >> accessing guest memory, e.g. Peter's networking daemon example, but it= _is_ fully > >> solvable. The exit to userspace means all three components (guest, ke= rnel, > >> and userspace) have full knowledge of what is shared and what is priva= te. There > >> is zero ambiguity: > >> > >> - if userspace accesses guest private memory, it gets SIGSEGV or wha= tever. > >> - if kernel accesses guest private memory, it does BUG/panic/oops[*] > >> - if guest accesses memory with the incorrect C/SHARED-bit, it gets = killed. > >> > >> This is the direction KVM TDX support is headed, though it's obviously= still a WIP. > >> > >> And ideally, to avoid implicit conversions at any level, hardware vend= ors' ABIs > >> define that: > >> > >> a) All convertible memory, i.e. RAM, starts as private. > >> b) Conversions between private and shared must be done via explicit = hypercall. > >> > >> Without (b), userspace and thus KVM have to treat guest accesses to th= e incorrect > >> type as implicit conversions. > >> > >> [*] Sadly, fully preventing kernel access to guest private is not poss= ible with > >> TDX, especially if the direct map is left intact. But maybe in th= e future > >> TDX will signal a fault instead of poisoning memory and leaving a = #MC mine. > > > > In this proposal, consider a guest driver instructing a device to DMA > > write a 1 GB memory buffer. A well-behaved guest driver will ensure > > that the entire 1 GB is marked shared. But what about a malicious or > > buggy guest? Let's assume a bad guest driver instructs the device to > > write guest private memory. > > > > So now, the virtual device, which might be implemented as some host > > side process, needs to (1) check and lock all 4k constituent RMP > > entries (so they're not converted to private while the DMA write is > > taking palce), (2) write the 1 GB buffer, and (3) unlock all 4 k > > constituent RMP entries? If I'm understanding this correctly, then the > > synchronization will be prohibitively expensive. > > Let's consider a very very similar scenario: consider a guest driver sett= ing up a 1 GB DMA buffer. The virtual device, implemented as host process,= needs to (1) map (and thus lock *or* be prepared for faults) in 1GB / 4k p= ages of guest memory (so they're not *freed* while the DMA write is taking = place), (2) write the buffer, and (3) unlock all the pages. Or it can lock= them at setup time and keep them locked for a long time if that's appropri= ate. > > Sure, the locking is expensive, but it's nonnegotiable. The RMP issue is= just a special case of the more general issue that the host MUST NOT ACCES= S GUEST MEMORY AFTER IT'S FREED. Good point.