linux-kernel - Re: [PATCH 10/35] irqchip/gic-v4.1: VPE table (aka GICR

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] [thread-next>] [day] [month] [year] [list]

Message-ID: <56ddfedd-5d8c-0507-c164-b04d63f51a1e@huawei.com>
Date:   Fri, 27 Sep 2019 09:59:16 +0800
From:   Zenghui Yu <yuzenghui@...wei.com>
To:     Marc Zyngier <maz@...nel.org>, <kvmarm@...ts.cs.columbia.edu>,
        <linux-kernel@...r.kernel.org>
CC:     Lorenzo Pieralisi <lorenzo.pieralisi@....com>,
        Jason Cooper <jason@...edaemon.net>,
        Thomas Gleixner <tglx@...utronix.de>
Subject: Re: [PATCH 10/35] irqchip/gic-v4.1: VPE table (aka GICR_VPROPBASER)
 allocation

On 2019/9/26 23:57, Marc Zyngier wrote:
> On 26/09/2019 16:19, Zenghui Yu wrote:
>> On 2019/9/25 22:41, Marc Zyngier wrote:
>>>
>>> Indeed. The whole idea is that ITSs and RDs can share a vPE table if
>>> they are in the same affinity group (such as a socket, for example).
>>> This is what is missing from v4.0 where the ITS knows about vPEs, but
>>> doesn't know about residency. With that in place, the HW is able to do a
>>> lot more for us.
>>
>> Thanks for your education!
>>
>> I really want to know *how* does GICv4.1 ITS know about the vPE
>> residency (in hardware level)?
> 
> Hey, I'm a SW guy, I don't design the hardware! ;-)
> 
>> I can understand that Redistributor can easily achieve it by
>> VPENDBASER's Valid and vPEID field.  And it's necessary for ITS to
>> know the residency so that it can determine whether it's appropriate
>> to trigger default doorbell for vPE.  But I have no knowledge about
>> how can it be achieved in hardware details.
> 
> My take is that the RD and the ITS can communicate over the shared
> table, hence my remark above about SVPET==0. But as I said, I'm not a HW
> guy.

;-) I should have asked our GIC engineers for these things.


Thanks,
zenghui