Re: [PATCH 4/4] cpufreq/amd-pstate: Remove the redundant amd_pstate_set_driver() call

[Mario Limonciello]

On 10/26/24 06:58, Klara Modin wrote:
> Hi,
>
> On 2024-10-17 12:05, Dhananjay Ugwekar wrote:
> > amd_pstate_set_driver() is called twice, once in amd_pstate_init() and 
> > once
> > as part of amd_pstate_register_driver(). Move around code and eliminate
> > the redundancy.
> >
> > Signed-off-by: Dhananjay Ugwekar <Dhananjay.Ugwekar@xxxxxxx>
> > ---
> >   drivers/cpufreq/amd-pstate.c | 12 ++++--------
> >   1 file changed, 4 insertions(+), 8 deletions(-)
> >
> > diff --git a/drivers/cpufreq/amd-pstate.c b/drivers/cpufreq/amd-pstate.c
> > index 13ee5cac901d..6f6d961879cc 100644
> > --- a/drivers/cpufreq/amd-pstate.c
> > +++ b/drivers/cpufreq/amd-pstate.c
> > @@ -1848,9 +1848,11 @@ static int __init amd_pstate_init(void)
> >           return -ENODEV;
> >       }
> > -    ret = amd_pstate_set_driver(cppc_state);
> > -    if (ret)
> > +    ret = amd_pstate_register_driver(cppc_state);
> > +    if (ret) {
> > +        pr_err("failed to register with return %d\n", ret);
> >           return ret;
> > +    }
> >       /* capability check */
> >       if (cpu_feature_enabled(X86_FEATURE_CPPC)) {
> > @@ -1870,12 +1872,6 @@ static int __init amd_pstate_init(void)
> >               return ret;
> >       }
> > -    ret = amd_pstate_register_driver(cppc_state);
> > -    if (ret) {
> > -        pr_err("failed to register with return %d\n", ret);
> > -        return ret;
> > -    }
> > -
> >       dev_root = bus_get_dev_root(&cpu_subsys);
> >       if (dev_root) {
> >           ret = sysfs_create_group(&dev_root->kobj, 
> > &amd_pstate_global_attr_group);
>
>
> This seems to break boot on my Zen 2 desktop (my Zen 4 laptop is fine, 
> however). I don't see any messages on the console and nothing shows up 
> in the pstore either.
>
> I'll attach the kernel log from a normal boot (with the patch reverted) 
> in case it helps.
>
> Please let me know if there's anything else you need.
>
> Regards,
> Klara Modin

Hi,

Thanks for reporting.  I think this might be a regression on shared 
memory designs specifically because static calls weren't updated yet.

Can you try this below diff?

diff --git a/drivers/cpufreq/amd-pstate.c b/drivers/cpufreq/amd-pstate.c
index 206725219d8c9..2f0e29b05963d 100644
--- a/drivers/cpufreq/amd-pstate.c
+++ b/drivers/cpufreq/amd-pstate.c
@@ -1857,12 +1857,6 @@ static int __init amd_pstate_init(void)
                return -ENODEV;
        }

-       ret = amd_pstate_register_driver(cppc_state);
-       if (ret) {
-               pr_err("failed to register with return %d\n", ret);
-               return ret;
-       }
-
        /* capability check */
        if (cpu_feature_enabled(X86_FEATURE_CPPC)) {
                pr_debug("AMD CPPC MSR based functionality is 
supported\n");
@@ -1881,6 +1875,12 @@ static int __init amd_pstate_init(void)
                        return ret;
        }

+       ret = amd_pstate_register_driver(cppc_state);
+       if (ret) {
+               pr_err("failed to register with return %d\n", ret);
+               return ret;
+       }
+
        dev_root = bus_get_dev_root(&cpu_subsys);
        if (dev_root) {
                ret = sysfs_create_group(&dev_root->kobj, 
&amd_pstate_global_attr_group);