On Tue, Oct 01, 2019 at 01:45:30PM +0100, Robin Murphy wrote: > On 01/10/2019 09:59, Sudeep Holla wrote: > > On Mon, Sep 30, 2019 at 12:46:33PM -0500, Rob Herring wrote: > > > On Mon, Sep 30, 2019 at 10:25 AM Robin Murphy <robin.murphy@xxxxxxx> wrote: > > > > > > > > Since we now have bindings for Mali Midgard GPUs, let's use them to > > > > describe Juno's GPU subsystem, if only because we can. Juno sports a > > > > Mali-T624 integrated behind an MMU-400 (as a gesture towards > > > > virtualisation), in their own dedicated power domain with DVFS > > > > controlled by the SCP. > > > > > > > > CC: Liviu Dudau <liviu.dudau@xxxxxxx> > > > > CC: Sudeep Holla <sudeep.holla@xxxxxxx> > > > > CC: Lorenzo Pieralisi <lorenzo.pieralisi@xxxxxxx> > > > > Signed-off-by: Robin Murphy <robin.murphy@xxxxxxx> > > > > --- > > > > .../bindings/gpu/arm,mali-midgard.yaml | 5 +++- > > > > arch/arm64/boot/dts/arm/juno-base.dtsi | 27 +++++++++++++++++++ > > > > 2 files changed, 31 insertions(+), 1 deletion(-) > > > > > > Reviewed-by: Rob Herring <robh@xxxxxxxxxx> > > > > If you plan to take it along with driver change, > > > > Acked-by: Sudeep Holla <sudeep.holla@xxxxxxx> > > > > If not, please let us know. I can take it for v5.5 > > The driver change is debatable and not strictly necessary, so it probably > makes more sense to take this one through the VExpress tree on its own. I > wouldn't suggest flipping the status to "enabled" just yet, but it seems Sure, make sense. > worth putting the basic description in place as a jumping-off point for > folks to hack on (e.g. it's another 'interesting' case for devfreq OPP > stuff). IIUC, the devfreq support in panfrost driver should work fine with SCPI as it has clock provider for GPU DVFS. With SCMI, we don't want to hook to clock framework, but I am yet to come up with a sane generic way to add SCMI devfreq and integrate it with the panfrost devfreq. -- Regards, Sudeep
