linux-kernel - Re: [PATCH] x86/sev-es: Do not use copy_from_kernel

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 8 Sep 2023 07:48:58 -0700
From:   Dave Hansen <dave.hansen@...el.com>
To:     Tom Lendacky <thomas.lendacky@....com>,
        Adam Dunlap <acdunlap@...gle.com>
Cc:     linux-kernel@...r.kernel.org, x86@...nel.org,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        Dave Hansen <dave.hansen@...ux.intel.com>,
        "H. Peter Anvin" <hpa@...or.com>,
        David Hildenbrand <david@...hat.com>,
        Mike Rapoport <rppt@...nel.org>,
        "Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>,
        Nikunj A Dadhania <nikunj@....com>,
        Dionna Glaze <dionnaglaze@...gle.com>,
        "Peter Zijlstra (Intel)" <peterz@...radead.org>,
        Joerg Roedel <jroedel@...e.de>, Jacob Xu <jacobhxu@...gle.com>
Subject: Re: [PATCH] x86/sev-es: Do not use copy_from_kernel_nofault in early
 #VC handler

On 9/8/23 06:13, Tom Lendacky wrote:
> On 9/7/23 14:12, Dave Hansen wrote:
>> But seriously, is it even *possible* to spin up a SEV-SNP VM what
>> doesn't have NX?
> 
> It is a common path, so while an SEV guest would have NX support, you
> would first have to determine that it is an SEV guest. That would take
> issuing a CPUID instruction in order to determine if a particular MSR
> can be read...

I was thinking more along the lines of telling folks that if they want
to turn SEV-SNP support on, they also have to give up on running on a
!NX system.  That would be a _bit_ nicer than just refusing to boot on
all !NX systems.

> Ultimately, we could probably pass the encryption mask from the
> decompressor to the kernel and avoid some of the checks during early
> boot of the kernel proper. Is it possible to boot an x86 kernel without
> going through the decompressor?

I think it's possible, but it's very unusual.