linux-kernel - Re: [PATCH v2 08/10] KVM: x86: Switch KVM guest to using interrupts for page ready APF delivery

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:   Wed, 10 Jun 2020 23:06:58 +0200
From:   Paolo Bonzini <pbonzini@...hat.com>
To:     Vivek Goyal <vgoyal@...hat.com>,
        Vitaly Kuznetsov <vkuznets@...hat.com>
Cc:     kvm@...r.kernel.org, x86@...nel.org,
        Andy Lutomirski <luto@...nel.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Borislav Petkov <bp@...en8.de>,
        "H. Peter Anvin" <hpa@...or.com>,
        Wanpeng Li <wanpengli@...cent.com>,
        Sean Christopherson <sean.j.christopherson@...el.com>,
        Jim Mattson <jmattson@...gle.com>,
        Gavin Shan <gshan@...hat.com>,
        Peter Zijlstra <peterz@...radead.org>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 08/10] KVM: x86: Switch KVM guest to using interrupts
 for page ready APF delivery

On 10/06/20 22:51, Vivek Goyal wrote:
>> KVM now supports using interrupt for 'page ready' APF event delivery and
>> legacy mechanism was deprecated. Switch KVM guests to the new one.
> Hi Vitaly,
> 
> I see we have all this code in guest which tries to take care of
> cases where PAGE_READY can be delivered before PAGE_NOT_PRESENT. In
> this new schedume of things, can it still happen. We are using
> an exception to deliver PAGE_NOT_PRESENT while using interrupt to
> deliver PAGE_READY.
> 
> If re-ordeing is not possible, then it will be good to get rid of
> that extra complexity in guest.

It is perhaps less likely but still possible, because the interrupt
might be delivered to another CPU and race against the exception.

Paolo