linux-kernel - Re: [PATCH] svm: Fix AVIC incomplete IPI emulation

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: <828b24bb-a55b-dd5f-a514-ef3e182db5cd@amd.com>
Date:   Wed, 13 Mar 2019 19:03:31 +0000
From:   "Suthikulpanit, Suravee" <Suravee.Suthikulpanit@....com>
To:     Oren Twaig <oren@...lemp.com>
CC:     "kvm@...r.kernel.org" <kvm@...r.kernel.org>,
        "x86@...nel.org" <x86@...nel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "Shai (Shai@...leMP.com)" <Shai@...lemp.com>,
        "pbonzini@...hat.com" <pbonzini@...hat.com>,
        "rkrcmar@...hat.com" <rkrcmar@...hat.com>,
        "Grimm, Jon" <Jon.Grimm@....com>
Subject: Re: [PATCH] svm: Fix AVIC incomplete IPI emulation

Oren

On 3/13/19 8:05 PM, Oren Twaig wrote:
> Hi Suravee,
> 
> Turns out, the _same_ bug was already discussed in the past by
> yourself, Paolo and Radim (both now 'cc'-ed)
> 
> Please read it here:
> https://patchwork.kernel.org/patch/8292231/
> 
> 
> After reading that thread, I have couple of questions:
> 
> First,
> 
> You wrote : "I have tried NOT setting the IRR, and only
> kick_vcpu(). And things seem to work fine. Therefore, I think your
> analysis is likely to be correct."
> 
> AFAIU, it means that the below patch is wrong just as Paolo
> suggested in his original answer and you did fixed it back than, but the
> code is now back ?

Thanks for the recap. I was chasing a bug and totally forgot about this
discussion. I just found the reason why irr_pending was not set to 1 in my test,
which then causes the target vcpu to never get wake up and scheduled to process
the IRR bit that was set by AVIC as part of sending IPI.

I will send out a patch to revert and clean up this mess. Thanks for catching this.
And sorry for confusion.

Thanks,
Suravee