[PATCH] cpufreq: intel_pstate: Use raw_smp_processor_id() in hwp_get_cpu_scaling()

[PATCH] cpufreq: intel_pstate: Use raw_smp_processor_id() in hwp_get_cpu_scaling()

[Xi Ruoyao]

Use raw_smp_processor_id() instead of plain smp_processor_id() in
hwp_get_cpu_scaling(), otherwise we get some errors on a Lenovo Thinkpad
T14P Gen 2:

    BUG: using smp_processor_id() in preemptible [00000000] code: swapper/0/1
    caller is hwp_get_cpu_scaling+0x7f/0xc0

Fixes: b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to get cpu-type")
Signed-off-by: Xi Ruoyao <xry111@xry111.site>
---
 drivers/cpufreq/intel_pstate.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c
index 4aad79d26c64..bfc20b978240 100644
--- a/drivers/cpufreq/intel_pstate.c
+++ b/drivers/cpufreq/intel_pstate.c
@@ -2209,7 +2209,7 @@ static int knl_get_turbo_pstate(int cpu)
 static int hwp_get_cpu_scaling(int cpu)
 {
 	if (hybrid_scaling_factor) {
-		struct cpuinfo_x86 *c = &cpu_data(smp_processor_id());
+		struct cpuinfo_x86 *c = &cpu_data(raw_smp_processor_id());
 		u8 cpu_type = c->topo.intel_type;
 
 		/*
-- 
2.49.0

Re: [PATCH] cpufreq: intel_pstate: Use raw_smp_processor_id() in hwp_get_cpu_scaling()

[srinivas pandruvada]

On Sat, 2025-04-12 at 18:34 +0800, Xi Ruoyao wrote:
> Use raw_smp_processor_id() instead of plain smp_processor_id() in
> hwp_get_cpu_scaling(), otherwise we get some errors on a Lenovo
> Thinkpad
> T14P Gen 2:
> 
>     BUG: using smp_processor_id() in preemptible [00000000] code:
> swapper/0/1
>     caller is hwp_get_cpu_scaling+0x7f/0xc0
> 
> Fixes: b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to get
> cpu-type")
> Signed-off-by: Xi Ruoyao <xry111@xry111.site>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>

> ---
>  drivers/cpufreq/intel_pstate.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/cpufreq/intel_pstate.c
> b/drivers/cpufreq/intel_pstate.c
> index 4aad79d26c64..bfc20b978240 100644
> --- a/drivers/cpufreq/intel_pstate.c
> +++ b/drivers/cpufreq/intel_pstate.c
> @@ -2209,7 +2209,7 @@ static int knl_get_turbo_pstate(int cpu)
>  static int hwp_get_cpu_scaling(int cpu)
>  {
>  	if (hybrid_scaling_factor) {
> -		struct cpuinfo_x86 *c =
> &cpu_data(smp_processor_id());
> +		struct cpuinfo_x86 *c =
> &cpu_data(raw_smp_processor_id());
>  		u8 cpu_type = c->topo.intel_type;
>  
>  		/*

Re: [PATCH] cpufreq: intel_pstate: Use raw_smp_processor_id() in hwp_get_cpu_scaling()

[Rafael J. Wysocki]

On Sunday, April 13, 2025 4:44:56 PM CEST srinivas pandruvada wrote:
> On Sat, 2025-04-12 at 18:34 +0800, Xi Ruoyao wrote:
> > Use raw_smp_processor_id() instead of plain smp_processor_id() in
> > hwp_get_cpu_scaling(), otherwise we get some errors on a Lenovo
> > Thinkpad
> > T14P Gen 2:
> > 
> >     BUG: using smp_processor_id() in preemptible [00000000] code:
> > swapper/0/1
> >     caller is hwp_get_cpu_scaling+0x7f/0xc0
> > 
> > Fixes: b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to get
> > cpu-type")
> > Signed-off-by: Xi Ruoyao <xry111@xry111.site>
>
> Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>

It's still broken after this patch though because the function should
use the cpu_data() of the target CPU and not of the CPU running the code.

The patch below should fix it.

===
From: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Subject: [PATCH v1] cpufreq: intel_pstate: Fix hwp_get_cpu_scaling()

Commit b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to get
cpu-type") introduced two issues into hwp_get_cpu_scaling().  First,
it made that function use the CPU type of the CPU running the code
even though the target CPU is passed as the argument to it and second,
it used smp_processor_id() for that even though hwp_get_cpu_scaling()
runs in preemptible context.

Fix both of these problems by simply passing "cpu" to cpu_data().

Fixes: b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to get cpu-type")
Link: https://lore.kernel.org/linux-pm/20250412103434.5321-1-xry111@xry111.site/
Reported-by: Xi Ruoyao <xry111@xry111.site>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
---
 drivers/cpufreq/intel_pstate.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

--- a/drivers/cpufreq/intel_pstate.c
+++ b/drivers/cpufreq/intel_pstate.c
@@ -2336,7 +2336,7 @@
 static int hwp_get_cpu_scaling(int cpu)
 {
 	if (hybrid_scaling_factor) {
-		struct cpuinfo_x86 *c = &cpu_data(smp_processor_id());
+		struct cpuinfo_x86 *c = &cpu_data(cpu);
 		u8 cpu_type = c->topo.intel_type;
 
 		/*

Re: [PATCH] cpufreq: intel_pstate: Use raw_smp_processor_id() in hwp_get_cpu_scaling()

[srinivas pandruvada]

On Mon, 2025-04-14 at 17:19 +0200, Rafael J. Wysocki wrote:
> On Sunday, April 13, 2025 4:44:56 PM CEST srinivas pandruvada wrote:
> > On Sat, 2025-04-12 at 18:34 +0800, Xi Ruoyao wrote:
> > > Use raw_smp_processor_id() instead of plain smp_processor_id() in
> > > hwp_get_cpu_scaling(), otherwise we get some errors on a Lenovo
> > > Thinkpad
> > > T14P Gen 2:
> > > 
> > >     BUG: using smp_processor_id() in preemptible [00000000] code:
> > > swapper/0/1
> > >     caller is hwp_get_cpu_scaling+0x7f/0xc0
> > > 
> > > Fixes: b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to
> > > get
> > > cpu-type")
> > > Signed-off-by: Xi Ruoyao <xry111@xry111.site>
> > 
> > Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
> 
> It's still broken after this patch though because the function should
> use the cpu_data() of the target CPU and not of the CPU running the
> code.

That is correct. The below patch should work.

Thanks,
Srinivas

> 
> The patch below should fix it.
> 
> ===
> From: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> Subject: [PATCH v1] cpufreq: intel_pstate: Fix hwp_get_cpu_scaling()
> 
> Commit b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to get
> cpu-type") introduced two issues into hwp_get_cpu_scaling().  First,
> it made that function use the CPU type of the CPU running the code
> even though the target CPU is passed as the argument to it and
> second,
> it used smp_processor_id() for that even though hwp_get_cpu_scaling()
> runs in preemptible context.
> 
> Fix both of these problems by simply passing "cpu" to cpu_data().
> 
> Fixes: b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to get
> cpu-type")
> Link:
> https://lore.kernel.org/linux-pm/20250412103434.5321-1-xry111@xry111.site/
> Reported-by: Xi Ruoyao <xry111@xry111.site>
> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> ---
>  drivers/cpufreq/intel_pstate.c |    2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> --- a/drivers/cpufreq/intel_pstate.c
> +++ b/drivers/cpufreq/intel_pstate.c
> @@ -2336,7 +2336,7 @@
>  static int hwp_get_cpu_scaling(int cpu)
>  {
>  	if (hybrid_scaling_factor) {
> -		struct cpuinfo_x86 *c =
> &cpu_data(smp_processor_id());
> +		struct cpuinfo_x86 *c = &cpu_data(cpu);
>  		u8 cpu_type = c->topo.intel_type;
>  
>  		/*
> 
> 
> 
> 

Re: [PATCH] cpufreq: intel_pstate: Use raw_smp_processor_id() in hwp_get_cpu_scaling()

[Rafael J. Wysocki]

On Mon, Apr 14, 2025 at 5:26 PM srinivas pandruvada
<srinivas.pandruvada@linux.intel.com> wrote:
>
> On Mon, 2025-04-14 at 17:19 +0200, Rafael J. Wysocki wrote:
> > On Sunday, April 13, 2025 4:44:56 PM CEST srinivas pandruvada wrote:
> > > On Sat, 2025-04-12 at 18:34 +0800, Xi Ruoyao wrote:
> > > > Use raw_smp_processor_id() instead of plain smp_processor_id() in
> > > > hwp_get_cpu_scaling(), otherwise we get some errors on a Lenovo
> > > > Thinkpad
> > > > T14P Gen 2:
> > > >
> > > >     BUG: using smp_processor_id() in preemptible [00000000] code:
> > > > swapper/0/1
> > > >     caller is hwp_get_cpu_scaling+0x7f/0xc0
> > > >
> > > > Fixes: b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to
> > > > get
> > > > cpu-type")
> > > > Signed-off-by: Xi Ruoyao <xry111@xry111.site>
> > >
> > > Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
> >
> > It's still broken after this patch though because the function should
> > use the cpu_data() of the target CPU and not of the CPU running the
> > code.
>
> That is correct. The below patch should work.

OK, I'll add an ACK from you to it then if you don't mind.

> >
> > The patch below should fix it.
> >
> > ===
> > From: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> > Subject: [PATCH v1] cpufreq: intel_pstate: Fix hwp_get_cpu_scaling()
> >
> > Commit b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to get
> > cpu-type") introduced two issues into hwp_get_cpu_scaling().  First,
> > it made that function use the CPU type of the CPU running the code
> > even though the target CPU is passed as the argument to it and
> > second,
> > it used smp_processor_id() for that even though hwp_get_cpu_scaling()
> > runs in preemptible context.
> >
> > Fix both of these problems by simply passing "cpu" to cpu_data().
> >
> > Fixes: b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to get
> > cpu-type")
> > Link:
> > https://lore.kernel.org/linux-pm/20250412103434.5321-1-xry111@xry111.site/
> > Reported-by: Xi Ruoyao <xry111@xry111.site>
> > Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> > ---
> >  drivers/cpufreq/intel_pstate.c |    2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > --- a/drivers/cpufreq/intel_pstate.c
> > +++ b/drivers/cpufreq/intel_pstate.c
> > @@ -2336,7 +2336,7 @@
> >  static int hwp_get_cpu_scaling(int cpu)
> >  {
> >       if (hybrid_scaling_factor) {
> > -             struct cpuinfo_x86 *c =
> > &cpu_data(smp_processor_id());
> > +             struct cpuinfo_x86 *c = &cpu_data(cpu);
> >               u8 cpu_type = c->topo.intel_type;
> >
> >               /*
> >
> >
> >
> >
>
>

Re: [PATCH] cpufreq: intel_pstate: Use raw_smp_processor_id() in hwp_get_cpu_scaling()

[Pawan Gupta]

On Mon, Apr 14, 2025 at 05:19:04PM +0200, Rafael J. Wysocki wrote:
> On Sunday, April 13, 2025 4:44:56 PM CEST srinivas pandruvada wrote:
> > On Sat, 2025-04-12 at 18:34 +0800, Xi Ruoyao wrote:
> > > Use raw_smp_processor_id() instead of plain smp_processor_id() in
> > > hwp_get_cpu_scaling(), otherwise we get some errors on a Lenovo
> > > Thinkpad
> > > T14P Gen 2:
> > > 
> > >     BUG: using smp_processor_id() in preemptible [00000000] code:
> > > swapper/0/1
> > >     caller is hwp_get_cpu_scaling+0x7f/0xc0
> > > 
> > > Fixes: b52aaeeadfac ("cpufreq: intel_pstate: Avoid SMP calls to get
> > > cpu-type")
> > > Signed-off-by: Xi Ruoyao <xry111@xry111.site>
> >
> > Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
> 
> It's still broken after this patch though because the function should
> use the cpu_data() of the target CPU and not of the CPU running the code.

Sorry for missing that.

I noticed that find_hybrid_pmu_for_cpu() doesn't take the cpu argument.
Does it suffer from the same problem?

init_hybrid_pmu(int cpu)
{
	struct cpu_hw_events *cpuc = &per_cpu(cpu_hw_events, cpu);
	struct x86_hybrid_pmu *pmu = find_hybrid_pmu_for_cpu();