linux-kernel - Re: [PATCH] KVM: x86: Fix lapic timer interrupt lost after loading a snapshot.

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

Date:   Tue, 12 Sep 2023 18:14:25 +0000
From:   Sean Christopherson <seanjc@...gle.com>
To:     Haitao Shan <hshan@...gle.com>
Cc:     Paolo Bonzini <pbonzini@...hat.com>, kvm@...r.kernel.org,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        Dave Hansen <dave.hansen@...ux.intel.com>, x86@...nel.org,
        "H. Peter Anvin" <hpa@...or.com>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] KVM: x86: Fix lapic timer interrupt lost after loading a snapshot.

Please send mail as plain text, most of the mailing lists reject HTML, and in
general HTML makes everyone grumpy.

On Tue, Sep 12, 2023, Haitao Shan wrote:
> On Tue, Sep 12, 2023 at 10:07 AM Sean Christopherson <seanjc@...gle.com>
> wrote:
> 
> > Aha!  And I think the new apicv_pre_state_restore() needs to be invoked
> > even if APICv is not active, because I don't see anything that purges the
> > PIR when APICv is enabled.  VMX's APICv doesn't have many inhibits that can
> > go away, and I highly doubt userspace will restore into a vCPU with pending
> > posted interrupts, so in practice this is _extremely_ unlikely to be
> > problematic.  But it's still wrong.
> >
> Thanks for sharing what you would like to fix the bug. I will write a v2 for
> that.  Actually, I am sorry that I forgot to add RFC to the title, as I
> personally did not think the proposed fix looks clean.

No worries, thanks for taking this on!