Re: [PATCH v6 3/5] firmware: psci: Read and use vendor reset types

[Stephen Boyd <swboyd@chromium.org>]

Quoting Elliot Berman (2024-10-23 09:30:21)
> On Fri, Oct 18, 2024 at 10:42:46PM -0700, Stephen Boyd wrote:
> > Quoting Elliot Berman (2024-10-18 12:39:48)
> > > diff --git a/drivers/firmware/psci/psci.c b/drivers/firmware/psci/psci.c
> > > index 2328ca58bba6..60bc285622ce 100644
> > > --- a/drivers/firmware/psci/psci.c
> > > +++ b/drivers/firmware/psci/psci.c
> > > @@ -305,9 +315,29 @@ static int get_set_conduit_method(const struct device_node *np)
> > >         return 0;
> > >  }
> > >
> > > +static void psci_vendor_sys_reset2(unsigned long action, void *data)
> > > +{
> > > +       const char *cmd = data;
> > > +       unsigned long ret;
> > > +       size_t i;
> > > +
> > > +       for (i = 0; i < num_psci_reset_params; i++) {
> > > +               if (!strcmp(psci_reset_params[i].mode, cmd)) {
> > > +                       ret = invoke_psci_fn(PSCI_FN_NATIVE(1_1, SYSTEM_RESET2),
> > > +                                            psci_reset_params[i].reset_type,
> > > +                                            psci_reset_params[i].cookie, 0);
> > > +                       pr_err("failed to perform reset \"%s\": %ld\n",
> > > +                               cmd, (long)ret);
> >
> > Do this intentionally return? Should it be some other function that's
> > __noreturn instead and a while (1) if the firmware returns back to the
> > kernel?
> >
>
> Yes, I think it's best to make sure we fall back to the architectural
> reset (whether it's the SYSTEM_RESET or architectural SYSTEM_RESET2)
> since device would reboot then.

Ok. Please add a comment in the code so we know that it's intentional.

>
> > > +               }
> > > +       }
> > > +}
> > > +
> > >  static int psci_sys_reset(struct notifier_block *nb, unsigned long action,
> > >                           void *data)
> > >  {
> > > +       if (data && num_psci_reset_params)
> > > +               psci_vendor_sys_reset2(action, data);
> > > +

I'd add a comment here as well indicating that a fallback is used.

> > >         if ((reboot_mode == REBOOT_WARM || reboot_mode == REBOOT_SOFT) &&
> > >             psci_system_reset2_supported) {
> > >                 /*
> > > @@ -750,6 +780,68 @@ static const struct of_device_id psci_of_match[] __initconst = {
> > >         {},
[...]
> > > +                       continue;
> > > +
> > > +               num = of_property_read_variable_u32_array(np, prop->name, magic, 1, 2);
> >
> > ARRAY_SIZE(magic)?
> >
> > > +               if (num < 0) {
> >
> > Should this be less than 1?
> >
>
> of_property_read_variable_u32_array should return -EOVERFLOW (or maybe
> -ENODATA) if the array is empty. I don't see it's possible for
> of_property_read_variable_u32_array() to return a non-negative value
> that's not 1 or 2.

I think you're saying a return value of 0 is impossible? Ok. I was
mostly looking at the usage of the return value later on in this patch
and trying to understand why 0 would be allowed as a possible return
value without looking at the details of
of_property_read_variable_u32_array(). I guess the 1, 2 are the min/max
though so it's fine.