linux-kernel - Re: [PATCH v9 0/9] x86: Show in sysfs if a memory node is able to do encryption

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] [day] [month] [year] [list]

Date:   Thu, 27 Oct 2022 17:33:59 +0200
From:   Borislav Petkov <bp@...en8.de>
To:     Dave Hansen <dave.hansen@...el.com>
Cc:     Martin Fernandez <martin.fernandez@...ypsium.com>,
        linux-kernel@...r.kernel.org, linux-efi@...r.kernel.org,
        platform-driver-x86@...r.kernel.org, linux-mm@...ck.org,
        kunit-dev@...glegroups.com, linux-kselftest@...r.kernel.org,
        tglx@...utronix.de, mingo@...hat.com, dave.hansen@...ux.intel.com,
        x86@...nel.org, hpa@...or.com, ardb@...nel.org,
        dvhart@...radead.org, andy@...radead.org,
        gregkh@...uxfoundation.org, rafael@...nel.org, rppt@...nel.org,
        akpm@...ux-foundation.org, daniel.gutson@...ypsium.com,
        hughsient@...il.com, alex.bazhaniuk@...ypsium.com,
        alison.schofield@...el.com, keescook@...omium.org
Subject: Re: [PATCH v9 0/9] x86: Show in sysfs if a memory node is able to do
 encryption

On Thu, Oct 27, 2022 at 08:21:02AM -0700, Dave Hansen wrote:
> On 10/27/22 01:57, Borislav Petkov wrote:
> > Well, I still think this is not going to work in all cases. SME/TME can
> > be enabled but the kernel can go - and for whatever reason - map a bunch
> > of memory unencrypted.
> 
> For TME on Intel systems, there's no way to make it unencrypted.  The
> memory controller is doing all the encryption behind the back of the OS
> and even devices that are doing DMA.  Nothing outside of the memory
> controller really knows or cares that encryption is happening.

Ok, Tom just confirmed that AMD's TSME thing also encrypts all memory.

So I guess the code should check for TME or TSME. If those are set, then
you can assume that all memory is encrypted.

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette