Nouveau - Oct 2019 - [PATCH v3] pci: prevent putting nvidia GPUs into lower device states on certain intel bridges

On Mon, Oct 21, 2019 at 02:00:46PM +0200, Karol Herbst
wrote:
> On Mon, Oct 21, 2019 at 1:40 PM Mika Westerberg
> <mika.westerberg at intel.com> wrote:
> >
> > Hi Karol,
> >
> > Sorry for commenting late, I just came back from vacation.
> >
> > On Wed, Oct 16, 2019 at 04:44:49PM +0200, Karol Herbst wrote:
> > > Fixes state transitions of Nvidia Pascal GPUs from D3cold into
higher device
> > > states.
> > >
> > > v2: convert to pci_dev quirk
> > >     put a proper technical explanation of the issue as a in-code
comment
> > > v3: disable it only for certain combinations of intel and nvidia
hardware
> > >
> > > Signed-off-by: Karol Herbst <kherbst at redhat.com>
> > > Cc: Bjorn Helgaas <bhelgaas at google.com>
> > > Cc: Lyude Paul <lyude at redhat.com>
> > > Cc: Rafael J. Wysocki <rjw at rjwysocki.net>
> > > Cc: Mika Westerberg <mika.westerberg at intel.com>
> > > Cc: linux-pci at vger.kernel.org
> > > Cc: linux-pm at vger.kernel.org
> > > Cc: dri-devel at lists.freedesktop.org
> > > Cc: nouveau at lists.freedesktop.org
> > > ---
> > >  drivers/pci/pci.c    | 11 ++++++++++
> > >  drivers/pci/quirks.c | 52
++++++++++++++++++++++++++++++++++++++++++++
> >
> > I may be missing something but why you can't do this in the
nouveau
> > driver itself?
> 
> What do you mean precisely? Move the quirk into nouveau, but keep the
> changes to pci core?
No, just block runtime PM from the device in nouveau driver.

On Mon, Oct 21, 2019 at 2:06 PM Mika Westerberg
<mika.westerberg at intel.com> wrote:
>
> On Mon, Oct 21, 2019 at 02:00:46PM +0200, Karol Herbst wrote:
> > On Mon, Oct 21, 2019 at 1:40 PM Mika Westerberg
> > <mika.westerberg at intel.com> wrote:
> > >
> > > Hi Karol,
> > >
> > > Sorry for commenting late, I just came back from vacation.
> > >
> > > On Wed, Oct 16, 2019 at 04:44:49PM +0200, Karol Herbst wrote:
> > > > Fixes state transitions of Nvidia Pascal GPUs from D3cold
into higher device
> > > > states.
> > > >
> > > > v2: convert to pci_dev quirk
> > > >     put a proper technical explanation of the issue as a
in-code comment
> > > > v3: disable it only for certain combinations of intel and
nvidia hardware
> > > >
> > > > Signed-off-by: Karol Herbst <kherbst at redhat.com>
> > > > Cc: Bjorn Helgaas <bhelgaas at google.com>
> > > > Cc: Lyude Paul <lyude at redhat.com>
> > > > Cc: Rafael J. Wysocki <rjw at rjwysocki.net>
> > > > Cc: Mika Westerberg <mika.westerberg at intel.com>
> > > > Cc: linux-pci at vger.kernel.org
> > > > Cc: linux-pm at vger.kernel.org
> > > > Cc: dri-devel at lists.freedesktop.org
> > > > Cc: nouveau at lists.freedesktop.org
> > > > ---
> > > >  drivers/pci/pci.c    | 11 ++++++++++
> > > >  drivers/pci/quirks.c | 52
++++++++++++++++++++++++++++++++++++++++++++
> > >
> > > I may be missing something but why you can't do this in the
nouveau
> > > driver itself?
> >
> > What do you mean precisely? Move the quirk into nouveau, but keep the
> > changes to pci core?
>
> No, just block runtime PM from the device in nouveau driver.
but that's not what the patch does. It only skips the PCI PM reg
write, but still let the ACPI method be invoked to put the device into
D3cold

On Mon, Oct 21, 2019 at 03:02:23PM +0200, Karol Herbst
wrote:
> > No, just block runtime PM from the device in nouveau driver.
> 
> but that's not what the patch does. It only skips the PCI PM reg
> write, but still let the ACPI method be invoked to put the device into
> D3cold
Oh, indeed it does. I did not realize that.

Which makes me wonder whether ACPI _OFF() expects the device to be in D0
for some reason.