[RFT PATCH v2 11/12] drm/radeon: Call drm_helper_force_disable_all() at shutdown/remove time

[Doug Anderson]

Hi,

On Mon, Sep 25, 2023 at 8:49 AM Deucher, Alexander
<Alexander.Deucher at amd.com> wrote:
>
> [Public]
>
> > -----Original Message-----
> > From: Douglas Anderson <dianders at chromium.org>
> > Sent: Thursday, September 21, 2023 3:27 PM
> > To: dri-devel at lists.freedesktop.org; Maxime Ripard <mripard at kernel.org>
> > Cc: Douglas Anderson <dianders at chromium.org>; Pan, Xinhui
> > <Xinhui.Pan at amd.com>; airlied at gmail.com; Deucher, Alexander
> > <Alexander.Deucher at amd.com>; amd-gfx at lists.freedesktop.org; Koenig,
> > Christian <Christian.Koenig at amd.com>; daniel at ffwll.ch; linux-
> > kernel at vger.kernel.org
> > Subject: [RFT PATCH v2 11/12] drm/radeon: Call
> > drm_helper_force_disable_all() at shutdown/remove time
> >
> > Based on grepping through the source code, this driver appears to be missing
> > a call to drm_atomic_helper_shutdown(), or in this case the non-atomic
> > equivalent drm_helper_force_disable_all(), at system shutdown time and at
> > driver remove time. This is important because
> > drm_helper_force_disable_all() will cause panels to get disabled cleanly which
> > may be important for their power sequencing. Future changes will remove any
> > custom powering off in individual panel drivers so the DRM drivers need to
> > start getting this right.
> >
> > The fact that we should call drm_atomic_helper_shutdown(), or in this case
> > the non-atomic equivalent drm_helper_force_disable_all(), in the case of OS
> > shutdown/restart comes straight out of the kernel doc "driver instance
> > overview" in drm_drv.c.
> >
> > NOTE: in order to get things inserted in the right place, I had to replace the
> > old/deprecated drm_put_dev() function with the equivalent new calls.
> >
> > Suggested-by: Maxime Ripard <mripard at kernel.org>
> > Reviewed-by: Maxime Ripard <mripard at kernel.org>
> > Signed-off-by: Douglas Anderson <dianders at chromium.org>
> > ---
> > I honestly have no idea if I got this patch right. The shutdown() function
> > already had some special case logic for PPC, Loongson, and VMs and I don't
> > 100% for sure know how this interacts with those. Everything here is just
> > compile tested.
>
> I think the reason for most of this funniness is to reduce shutdown time.  Lots of users complain if driver takes a while to shutdown and there is a point to be made that if the system is going into power down, there is not much reason to spend a lot of time messing with the hardware.

Sure, you don't want to do too much at shutdown time. That's the whole
reason that "shutdown" doesn't do a full remove / uninitialization of
all drivers. ...but drm_atomic_helper_shutdown() is documented to do
the things that are important for shutdown. Specifically, it cleanly
disables all of the displays. Depending on the display, this could
avoid temporary garbage on the display at reboot time or it could even
be important for the long term health of the panel.

-Doug