linux-kernel - Re: [RFC 2/7] domainsctrl: Introduce domains controller framework

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]

Date:   Fri, 18 Jan 2019 09:26:45 +0000
From:   Benjamin GAIGNARD <benjamin.gaignard@...com>
To:     Mark Brown <broonie@...nel.org>
CC:     "robh@...nel.org" <robh@...nel.org>,
        "arnd@...db.de" <arnd@...db.de>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        Loic PALLARDY <loic.pallardy@...com>,
        "benjamin.gaignard@...aro.org" <benjamin.gaignard@...aro.org>
Subject: Re: [RFC 2/7] domainsctrl: Introduce domains controller framework

On 1/17/19 6:27 PM, Mark Brown wrote:
> On Mon, Jan 14, 2019 at 03:41:57PM +0100, Benjamin Gaignard wrote:
>
>> Configurations could be applied with functions like
>> domainsctrl_set_config_by_index() or domainsctrl_set_config_by_name().
> Do you have any clients in the works for this?  It seems fairly likely
> that everything is fine in terms of setting the mode but it'd be good to
> confirm that's the case.  The main thing I can think of that might be a
> problem here is how you'd handle a case where we were talking to another
> processor that owns the permissions, we'd probably want more ways to
> query state there but I'm thinking there'd likely be some other higher
> level way to talk to the other processor there so perhaps it's moot.

Patch 3 use those functions to apply the default configuration before probing
a driver (or after unbind it). I have in mind that drivers could ask to apply
a configuration like it is done for pinctrl in resume/suspend functions.
An example of that could be to start the hardware block on the main processor
and, when going to sleep, change the configuration to grant the access to lower
power processor.

I think that talking to a remote processor is another problem already addressed
remoteproc or secure monitor calls. Domains controllers drivers should be created
for those cases.

>
> Otherwise this all looks pretty clean and simple, there's some handling
> for probe deferral in there which is the only slightly complex thing I
> noticed.