Re: [PATCH v2 4/4] arm64: dts: qcom: Add PSCI SYSTEM_RESET2 types for qcm6490-idp

[Elliot Berman <quic_eberman@quicinc.com>]

On Mon, Apr 15, 2024 at 09:42:40PM +0200, Konrad Dybcio wrote:
> 
> 
> On 4/15/24 02:32, Elliot Berman wrote:
> > On Mon, Apr 15, 2024 at 02:13:29AM +0300, Dmitry Baryshkov wrote:
> > > On Sun, 14 Apr 2024 at 22:32, Elliot Berman <quic_eberman@quicinc.com> wrote:
> > > > 
> > > > Add nodes for the vendor-defined system resets. "bootloader" will cause
> > > > device to reboot and stop in the bootloader's fastboot mode. "edl" will
> > > > cause device to reboot into "emergency download mode", which permits
> > > > loading images via the Firehose protocol.
> > > > 
> > > > Co-developed-by: Shivendra Pratap <quic_spratap@quicinc.com>
> > > > Signed-off-by: Shivendra Pratap <quic_spratap@quicinc.com>
> > > > Signed-off-by: Elliot Berman <quic_eberman@quicinc.com>
> > > > ---
> > > >   arch/arm64/boot/dts/qcom/qcm6490-idp.dts | 5 +++++
> > > >   1 file changed, 5 insertions(+)
> > > > 
> > > > diff --git a/arch/arm64/boot/dts/qcom/qcm6490-idp.dts b/arch/arm64/boot/dts/qcom/qcm6490-idp.dts
> > > > index e4bfad50a669..a966f6c8dd7c 100644
> > > > --- a/arch/arm64/boot/dts/qcom/qcm6490-idp.dts
> > > > +++ b/arch/arm64/boot/dts/qcom/qcm6490-idp.dts
> > > > @@ -126,6 +126,11 @@ debug_vm_mem: debug-vm@d0600000 {
> > > >                  };
> > > >          };
> > > > 
> > > > +       psci {
> > > 
> > > Please use a label instead. Otherwise it looks as if you are adding
> > > new device node.
> > > 
> > 
> > Right. Fixed for the next revision.
> 
> Are you guys planning to make this sorta ABI-like?
> 
> If so (which would be greatly appreciated by the way..), perhaps you
> could stick these magic values in dt-bindings and give them cool names
> 
> FWIW DEN0022 (my second-favorite book) suggests these values are almost
> totally vendor-defined, so if I were Qualcomm, I'd take the creative
> liberty to come up with a set of numbers and never ever ever change
> them

This is my goal as well. I'd like to keep the magic values out of
dt-bindings until we get the vendor SYSTEM_RESET2 spread across more
devices, as things might need a bit of settling. Since having stable
vendor reset2 is (IMO) primarily a benefit to Qualcomm, I expect this
will happen naturally.