On Fri, Mar 31, 2023 at 5:47 AM Rob Herring <robh+dt@xxxxxxxxxx> wrote: > > On Thu, Mar 30, 2023 at 8:45 PM David Dai <davidai@xxxxxxxxxx> wrote: > > > > Add devicetree bindings for a virtual kvm cpufreq driver. > > > > Co-developed-by: Saravana Kannan <saravanak@xxxxxxxxxx> > > Signed-off-by: Saravana Kannan <saravanak@xxxxxxxxxx> > > Signed-off-by: David Dai <davidai@xxxxxxxxxx> > > --- > > .../bindings/cpufreq/cpufreq-virtual-kvm.yaml | 39 +++++++++++++++++++ > > 1 file changed, 39 insertions(+) > > create mode 100644 Documentation/devicetree/bindings/cpufreq/cpufreq-virtual-kvm.yaml > > > > diff --git a/Documentation/devicetree/bindings/cpufreq/cpufreq-virtual-kvm.yaml b/Documentation/devicetree/bindings/cpufreq/cpufreq-virtual-kvm.yaml > > new file mode 100644 > > index 000000000000..31e64558a7f1 > > --- /dev/null > > +++ b/Documentation/devicetree/bindings/cpufreq/cpufreq-virtual-kvm.yaml > > @@ -0,0 +1,39 @@ > > +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause > > +%YAML 1.2 > > +--- > > +$id: http://devicetree.org/schemas/cpufreq/cpufreq-virtual-kvm.yaml# > > +$schema: http://devicetree.org/meta-schemas/core.yaml# > > + > > +title: Virtual KVM CPUFreq > > + > > +maintainers: > > + - David Dai <davidai@xxxxxxxxxx> > > + > > +description: | > > + > > + KVM CPUFreq is a virtualized driver in guest kernels that sends utilization > > + of its vCPUs as a hint to the host. The host uses hint to schedule vCPU > > + threads and select CPU frequency. It enables accurate Per-Entity Load > > + Tracking for tasks running in the guest by querying host CPU frequency > > + unless a virtualized FIE exists(Like AMUs). > > + > > +properties: > > + compatible: > > + const: virtual,kvm-cpufreq > > + > > +required: > > + - compatible > > + > > +additionalProperties: false > > + > > +examples: > > + - | > > + { > > + #address-cells = <2>; > > + #size-cells = <2>; > > + > > + cpufreq { > > + compatible = "virtual,kvm-cpufreq"; > > + }; > > The same thing was tried on non-virtual h/w too. This is not a device > so it doesn't go in DT. It is just an abuse of DT as a kernel driver > instantiation mechanism. Because it has no registers it's reading and writing, right? Yeah, just went with this for now to make it easy for people to cherry pick and test it. Maybe we shouldn't have added documentation and made this look too official. In the end, I'm expecting this will be a real MMIO device. Until we move from RFC to PATCH, feel free to ignore this patch. -Saravana
