linux-kernel - Re: [PATCH v2 12/13] iommu/rockchip: Add runtime PM support

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, 17 Jan 2018 16:48:46 +0800
From:   JeffyChen <jeffy.chen@...k-chips.com>
To:     Tomasz Figa <tfiga@...omium.org>
CC:     linux-kernel@...r.kernel.org, Ricky Liang <jcliang@...omium.org>,
        Robin Murphy <robin.murphy@....com>,
        simon xue <xxm@...k-chips.com>,
        Heiko Stuebner <heiko@...ech.de>,
        "open list:ARM/Rockchip SoC..." <linux-rockchip@...ts.infradead.org>,
        "list@....net:IOMMU DRIVERS" <iommu@...ts.linux-foundation.org>,
        Joerg Roedel <joro@...tes.org>,
        linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH v2 12/13] iommu/rockchip: Add runtime PM support

Hi Tomasz,

On 01/17/2018 03:52 PM, JeffyChen wrote:
> Hi Tomasz,
>
> On 01/17/2018 03:38 PM, Tomasz Figa wrote:
>>>> >>Don't we need to check here (and in _shutdown() too) if we have a
>>>> >>domain attached?
>>> >
>>> >hmmm, right, the startup might been called by resume, so should check
>>> >iommu->domain here.
>>> >
>>> >but the shutdown would be called at the end of detach or suspend, it
>>> could
>>> >be not attached or attached.
>> If startup might be called by resume, without domain attached, what
>> prevents shutdown from being called by suspend after that resume,
>> still without domain attached? Is it guaranteed that if resume is
>> called, someone will attach a domain before suspend is called?
>>
> no, the shutdown would be called by:
> 1/ end of detach_dev
> so it would be not attached at that time
>
> 2/ suspend
> so it could be attached, and also could be not attached
>
>
> anyway, i think the shutdown would work without domain attached(just
> disable paging and clear the iommu bases) ;)
>
hmmm, i see the problem.

so we need to:
1/ move shutdown a little earlier in detach_dev, so it could still see 
the iommu->domain

2/ check iommu->domain in shutdown, to prevent unnecessary shutdown


or maybe just add iommu->domain check in suspend and resume.
>> Best regards,
>> Tomasz
>>
>>
>>
>