linux-kernel - Re: [PATCH] drm/i915: Switch TGL-H DP-IN to dGFX when it's supported

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:   Wed, 24 Aug 2022 22:22:10 +0800
From:   Kai-Heng Feng <kai.heng.feng@...onical.com>
To:     Karol Herbst <kherbst@...hat.com>
Cc:     Lukas Wunner <lukas@...ner.de>,
        Jani Nikula <jani.nikula@...ux.intel.com>,
        tvrtko.ursulin@...ux.intel.com,
        Heikki Krogerus <heikki.krogerus@...ux.intel.com>,
        David Airlie <airlied@...ux.ie>,
        intel-gfx@...ts.freedesktop.org, linux-kernel@...r.kernel.org,
        dri-devel@...ts.freedesktop.org,
        Hans de Goede <hdegoede@...hat.com>, rodrigo.vivi@...el.com,
        Zenghui Yu <yuzenghui@...wei.com>
Subject: Re: [PATCH] drm/i915: Switch TGL-H DP-IN to dGFX when it's supported

On Sat, Aug 20, 2022 at 1:01 AM Karol Herbst <kherbst@...hat.com> wrote:
>
> On Thu, Aug 18, 2022 at 2:09 PM Lukas Wunner <lukas@...ner.de> wrote:
> >
> > On Tue, Aug 16, 2022 at 11:06:18AM +0300, Jani Nikula wrote:
> > > On Tue, 16 Aug 2022, Kai-Heng Feng <kai.heng.feng@...onical.com> wrote:
> > > > On mobile workstations like HP ZBook Fury G8, iGFX's DP-IN can switch to
> > > > dGFX so external monitors are routed to dGFX, and more monitors can be
> > > > supported as result.
> > > >
> > > > To switch the DP-IN to dGFX, the driver needs to invoke _DSM function 20
> > > > on intel_dsm_guid2. This method is described in Intel document 632107.
> > >
> > > Is this the policy decision that we want to unconditionally make,
> > > though?
> >
> > In general, we handle switching of outputs between GPUs in vga_switcheroo.c
> > upon a request from user space via sysfs (well, debugfs currently).
> > It's up to users to decide which policy suits their needs best.
> >
> > That said, we never grew support to allow different switching policies for
> > the built-in panel and external outputs.  Laptops supporting this are
> > rare.  Older MacBook Pros introduced between 2008 and 2010 are among them:
> > They have separate muxes for the panel and external DP port.  Our policy
> > is documented in a code comment in drivers/platform/x86/apple-gmux.c:
> >
> >  * The external DP port is only fully switchable on the first two unibody
> >  * MacBook Pro generations, MBP5 2008/09 and MBP6 2010. This is done by an
> >  * `NXP CBTL06141`_ which is controlled by gmux.
> >  [...]
> >  * Our switching policy for the external port is that on those generations
> >  * which are able to switch it fully, the port is switched together with the
> >  * panel when IGD / DIS commands are issued to vga_switcheroo. It is thus
> >  * possible to drive e.g. a beamer on battery power with the integrated GPU.
> >  * The user may manually switch to the discrete GPU if more performance is
> >  * needed.
> >  *
> >  * On all newer generations, the external port can only be driven by the
> >  * discrete GPU. If a display is plugged in while the panel is switched to
> >  * the integrated GPU, *both* GPUs will be in use for maximum performance.
> >  * To decrease power consumption, the user may manually switch to the
> >  * discrete GPU, thereby suspending the integrated GPU.
> >
> > In other words, on these older MacBook Pros, we switch the panel and
> > external DP port in unison, thus always allowing one of the two GPUs
> > to runtime suspend and save power.
> >
> > Thanks,
> >
> > Lukas
> >
>
> sure, but this is changing now. I do have a laptop with a muxable
> internal display. But this is considered to be a dynamic on demand
> switching thing not a boot time switch.
>
> Anyway, I am still not convinced that doing that unconditionally is
> what we want, especially as userspace has to support dynamic switching
> regardless.
>

According to the doc, there's no MUX in TGL-H DP-IN. The dGPU outputs
are routed through TGL-H TCSS directly.
That's probably the reason why it can't be dynamic.

Kai-Heng