linux-kernel - Re: [PATCH v2 0/3] SEV-SNP: Add KVM support for attestation

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: Fri, 28 Jun 2024 19:48:03 -0500
From: Michael Roth <michael.roth@....com>
To: <kvm@...r.kernel.org>
CC: <linux-coco@...ts.linux.dev>, <linux-kernel@...r.kernel.org>,
	<x86@...nel.org>, <pbonzini@...hat.com>, <seanjc@...gle.com>,
	<jroedel@...e.de>, <thomas.lendacky@....com>, <pgonda@...gle.com>,
	<ashish.kalra@....com>, <bp@...en8.de>, <pankaj.gupta@....com>,
	<liam.merwick@...cle.com>
Subject: Re: [PATCH v2 0/3] SEV-SNP: Add KVM support for attestation

On Fri, Jun 28, 2024 at 01:52:41PM -0500, Michael Roth wrote:
> Changes since v1:
> 
>  * Fix cleanup path when handling firmware error (Liam, Sean)
>  * Use bounce-pages for interacting with firmware rather than passing in the
>    guest-provided pages directly. (Sean)
>  * Drop SNP_GUEST_VMM_ERR_GENERIC and rely solely on firmware-provided error
>    code to report any firmware error to the guest. (Sean)
>  * Use kvm_clear_guest() to handle writing empty certificate table instead 
>    of kvm_write_guest() (Sean)
>  * Add additional comments in commit messages and throughout code to better
>    explain the interactions with firmware/guest. (Sean)
>  * Drop 4K-alignment restrictions on the guest-provided req/resp buffers,
>    since the GHCB-spec only specifically requires they fit within 4K,

It turns out my eyeballs were not functional when I reached that
conclusion and it's clearly documented that the pages needed to be
4K-aligned in the GHCB spec.

With the current implementation, KVM can actually handle unaligned
req/resp GPAs thanks to the bounce buffers, but it should still be
enforced. So I will resend a v3 with this change, but leave a bit more
time in case there are other review comments for v2.

Thanks,

Mike

>    not necessarily that they be 4K-aligned. Additionally, the bounce
>    pages passed to firmware will be 4K-aligned regardless.
> 
> Changes since splitting this off from v15 SNP KVM patchset:
> 
>  * Address clang-reported warnings regarding uninitialized variables 
>  * Address a memory leak of the request/response buffer pages, and refactor
>    the code based on Sean's suggestions:
>    https://lore.kernel.org/kvm/ZktbBRLXeOp9X6aH@google.com/
>  * Fix SNP Extended Guest Request handling to only attempt to fetch
>    certificates if handling MSG_REQ_REPORT (attestation) message types
>  * Drop KVM_EXIT_VMGEXIT and introduce KVM_EXIT_COCO events instead
>  * Refactor patch layout for easier handling/review
> 
> ----------------------------------------------------------------
> Brijesh Singh (1):
>       KVM: SEV: Provide support for SNP_GUEST_REQUEST NAE event
> 
> Michael Roth (2):
>       x86/sev: Move sev_guest.h into common SEV header
>       KVM: SEV: Provide support for SNP_EXTENDED_GUEST_REQUEST NAE event
> 
>  arch/x86/include/asm/sev.h              |  48 ++++++++
>  arch/x86/kvm/svm/sev.c                  | 187 ++++++++++++++++++++++++++++++++
>  arch/x86/kvm/svm/svm.h                  |   3 +
>  drivers/virt/coco/sev-guest/sev-guest.c |   2 -
>  drivers/virt/coco/sev-guest/sev-guest.h |  63 -----------
>  include/uapi/linux/sev-guest.h          |   3 +
>  6 files changed, 241 insertions(+), 65 deletions(-)
>  delete mode 100644 drivers/virt/coco/sev-guest/sev-guest.h
> 
>