linux-kernel - Re: [PATCH v2] platform/x86: intel-hid: Add support for Device Specific Methods

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:   Sat, 07 Jul 2018 16:37:52 -0700
From:   Srinivas Pandruvada <srinivas.pandruvada@...ux.intel.com>
To:     Andy Shevchenko <andy.shevchenko@...il.com>,
        Mario Limonciello <Mario.Limonciello@...l.com>
Cc:     Darren Hart <dvhart@...radead.org>,
        Alex Hung <alex.hung@...onical.com>,
        Andy Shevchenko <andy@...radead.org>,
        Platform Driver <platform-driver-x86@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        "Rafael J. Wysocki" <rjw@...ysocki.net>
Subject: Re: [PATCH v2] platform/x86: intel-hid: Add support for Device
 Specific Methods

On Sat, 2018-07-07 at 19:24 +0300, Andy Shevchenko wrote:
> On Sat, Jul 7, 2018 at 6:48 AM,  <Mario.Limonciello@...l.com> wrote:
> > > I strongly advocate for vendors to have more control over their
> > > drivers,
> > > but this scenario really frustrates me. I don't think I can
> > > justify this
> > > to Linus as a fix. But before we just say "no" (because hey, I
> > > want
> > > these fixes available as early as possible too), let's ask Rafael
> > > if he
> > > has an opinion or if there is precedent for this in his
> > > experience with
> > > ACPI drivers in general:
> > 
> > Full disclosure - an updated FW has since been rolled out that
> > reverted this
> > behavior back to previous FW behavior due to lack of Linux support
> > for the
> > new _DSM.  There is desire to use the new interface (as it did fix
> > actual
> > problems with the old one) so at some point it may return.  When
> > that happens
> > it would be ideal that people who are (for example) running an LTS
> > kernel
> > or distro kernel that tracks stable can pick it up too.
> 
> I am all for the better user experience and bugs especially in FW
> should be fixed, but I think to support all kind of behaviour the new
> FW release should always provide a knob by which we can check if some
> feature is supported or not.
> 
> Above sounds to me as mistake in design of the fix.
The fix here takes care of all the combination. If the new _DSM is
available, it checks what functionality the _DSM interface can provide
(Fn 0). For those functionality it gets using _DSM matching new FW
interface. If there is no _DSM or _DSM doesn't provide the
functionality or _DSM provided method fails, it still revert to old FW
interface style using standalone method. Eventually because of
configurable 5-button array, all platforms will be using new FW
interface with consideration of backward compatibility.

Thanks,
Srinivas