acpi-cpufreq scales the aperf/mperf measurements by max_freq, but HWP
needs to scale by base frequency.  Settings max_freq to base_freq
"works" but the code is not obvious, and returning values to userspace
is tricky.  Add an additonal perf_freq member which is used for scaling
aperf/mperf measurements.

Signed-off-by: Jason Andryuk <jandryuk@gmail.com>
---
I don't like this, but it seems the best way to re-use the common
aperf/mperf code.  The other option would be to add wrappers that then
do the acpi vs. hwp scaling.
---
 xen/arch/x86/acpi/cpufreq/cpufreq.c | 2 +-
 xen/drivers/cpufreq/utility.c       | 1 +
 xen/include/acpi/cpufreq/cpufreq.h  | 3 +++
 3 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/xen/arch/x86/acpi/cpufreq/cpufreq.c b/xen/arch/x86/acpi/cpufreq/cpufreq.c
index f1f3c6923f..5eac2f7321 100644
--- a/xen/arch/x86/acpi/cpufreq/cpufreq.c
+++ b/xen/arch/x86/acpi/cpufreq/cpufreq.c
@@ -317,7 +317,7 @@ unsigned int get_measured_perf(unsigned int cpu, unsigned int flag)
     else
         perf_percent = 0;
 
-    return policy->cpuinfo.max_freq * perf_percent / 100;
+    return policy->cpuinfo.perf_freq * perf_percent / 100;
 }
 
 static unsigned int get_cur_freq_on_cpu(unsigned int cpu)
diff --git a/xen/drivers/cpufreq/utility.c b/xen/drivers/cpufreq/utility.c
index b93895d4dd..788929e079 100644
--- a/xen/drivers/cpufreq/utility.c
+++ b/xen/drivers/cpufreq/utility.c
@@ -236,6 +236,7 @@ int cpufreq_frequency_table_cpuinfo(struct cpufreq_policy *policy,
 
     policy->min = policy->cpuinfo.min_freq = min_freq;
     policy->max = policy->cpuinfo.max_freq = max_freq;
+    policy->cpuinfo.perf_freq = max_freq;
     policy->cpuinfo.second_max_freq = second_max_freq;
 
     if (policy->min == ~0)
diff --git a/xen/include/acpi/cpufreq/cpufreq.h b/xen/include/acpi/cpufreq/cpufreq.h
index 56df5eebed..b91859ce5d 100644
--- a/xen/include/acpi/cpufreq/cpufreq.h
+++ b/xen/include/acpi/cpufreq/cpufreq.h
@@ -37,6 +37,9 @@ extern struct acpi_cpufreq_data *cpufreq_drv_data[NR_CPUS];
 struct cpufreq_cpuinfo {
     unsigned int        max_freq;
     unsigned int        second_max_freq;    /* P1 if Turbo Mode is on */
+    unsigned int        perf_freq; /* Scaling freq for aperf/mpref.
+                                      acpi-cpufreq uses max_freq, but HWP uses
+                                      base_freq.*/
     unsigned int        min_freq;
     unsigned int        transition_latency; /* in 10^(-9) s = nanoseconds */
 };
-- 
2.30.2

On 03.05.2021 21:27, Jason Andryuk wrote:
> acpi-cpufreq scales the aperf/mperf measurements by max_freq, but HWP
> needs to scale by base frequency.  Settings max_freq to base_freq
> "works" but the code is not obvious, and returning values to userspace
> is tricky.  Add an additonal perf_freq member which is used for scaling
> aperf/mperf measurements.
> 
> Signed-off-by: Jason Andryuk <jandryuk@gmail.com>
> ---
> I don't like this, but it seems the best way to re-use the common
> aperf/mperf code.  The other option would be to add wrappers that then
> do the acpi vs. hwp scaling.

Not sure I understand what you mean by "wrappers". I would assume that
for hwp you simply install a different getavg hook? Or else I guess
I'd need to see at least an outline of what you see as the alternative.

Jan

On Wed, May 26, 2021 at 9:24 AM Jan Beulich <jbeulich@suse.com> wrote:
>
> On 03.05.2021 21:27, Jason Andryuk wrote:
> > acpi-cpufreq scales the aperf/mperf measurements by max_freq, but HWP
> > needs to scale by base frequency.  Settings max_freq to base_freq
> > "works" but the code is not obvious, and returning values to userspace
> > is tricky.  Add an additonal perf_freq member which is used for scaling
> > aperf/mperf measurements.
> >
> > Signed-off-by: Jason Andryuk <jandryuk@gmail.com>
> > ---
> > I don't like this, but it seems the best way to re-use the common
> > aperf/mperf code.  The other option would be to add wrappers that then
> > do the acpi vs. hwp scaling.
>
> Not sure I understand what you mean by "wrappers". I would assume that
> for hwp you simply install a different getavg hook? Or else I guess
> I'd need to see at least an outline of what you see as the alternative.

Something like a common get_measured_perf() returning just the
aperf/mperf ratio with cpufreq_driver specific getavg() scaling as
appropriate.

Regards,
Jason