linux-kernel - Re: [PATCH v7] KVM: x86/tsc: Don't sync user-written TSC against startup values

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]

Message-ID: <D764860F-5DBC-49DD-8909-D08A3C24BC42@infradead.org>
Date:   Tue, 10 Oct 2023 11:03:50 +0100
From:   David Woodhouse <dwmw2@...radead.org>
To:     Sean Christopherson <seanjc@...gle.com>,
        Paolo Bonzini <pbonzini@...hat.com>,
        Like Xu <like.xu.linux@...il.com>
CC:     Oliver Upton <oliver.upton@...ux.dev>, kvm@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v7] KVM: x86/tsc: Don't sync user-written TSC against startup values



On 10 October 2023 01:44:32 BST, Sean Christopherson <seanjc@...gle.com> wrote:
>On Sun, 08 Oct 2023 10:53:35 +0800, Like Xu wrote:
>> The legacy API for setting the TSC is fundamentally broken, and only
>> allows userspace to set a TSC "now", without any way to account for
>> time lost to preemption between the calculation of the value, and the
>> kernel eventually handling the ioctl.
>> 
>> To work around this we have had a hack which, if a TSC is set with a
>> value which is within a second's worth of a previous vCPU, assumes that
>> userspace actually intended them to be in sync and adjusts the newly-
>> written TSC value accordingly.
>> 
>> [...]
>
>Applied to kvm-x86 misc, thanks!  I massaged away most of the pronouns in the
>changelog.  Yes, they bug me that much, and I genuinely had a hard time following
>some of the paragraphs even though I already knew what the patch is doing.
>
>Everyone, please take a look and make sure I didn't botch anything.  I tried my
>best to keep the existing "voice" and tone of the changelog (sans pronouns
>obviously).  I definitely don't want to bikeshed this thing any further.  If
>I've learned anything by this patch, it's that the only guaranteed outcome of
>changelog-by-committee is that no one will walk away 100% happy :-)

LGTM. I forgive you for not respecting my pronouns. :)

Thanks.