mbox series

[0/6] CPPC_CPUFREQ improvements for Tegra241

Message ID 20230418113459.12860-1-sumitg@nvidia.com
Headers show
Series CPPC_CPUFREQ improvements for Tegra241 | expand

Message

Sumit Gupta April 18, 2023, 11:34 a.m. UTC 
This patch series contains improvements and changes needed to get a
stable value of current CPU frequency from "cpuinfo_cur_freq" sysfs
node in Tegra241 SoC. All the CPU's in Tegra241 implement the ARM
Activity Monitor Unit (AMU). On reading the sysfs node, the frequency
value is re-constructed using AMU counters.

Sanjay Chandrashekara (1):
  cpufreq: use correct unit when verify cur freq

Sumit Gupta (5):
  cpufreq: CPPC: make workaround apply code generic
  irqchip/gicv3: Export arm_smccc_get_soc_id_xx funcs
  cpufreq: CPPC: update sampling window for Tegra241
  arm64: cpufeature: Export get_cpu_with_amu_feat func
  cpufreq: CPPC: use wq to read amu counters on target cpu

 arch/arm64/kernel/cpufeature.c |  1 +
 drivers/cpufreq/cppc_cpufreq.c | 99 +++++++++++++++++++++++++++++-----
 drivers/cpufreq/cpufreq.c      |  2 +-
 drivers/firmware/smccc/smccc.c |  2 +
 4 files changed, 89 insertions(+), 15 deletions(-)

Comments

Rafael J. Wysocki April 18, 2023, 12:57 p.m. UTC | #1 
On Tue, Apr 18, 2023 at 1:35 PM Sumit Gupta <sumitg@nvidia.com> wrote:
>
> From: Sanjay Chandrashekara <sanjayc@nvidia.com>
>
> cpufreq_verify_current_freq checks if the frequency returned by
> the hardware has a slight delta with the valid frequency value
> last set and returns "policy->cur" if the delta is within "1 MHz".
> In the comparison, "policy->cur" is in "kHz" but it's compared
> against HZ_PER_MHZ. So, the comparison range becomes "1 GHz".
> Fix this by comparing against KHZ_PER_MHZ instead of HZ_PER_MHZ.
>
> Fixes: f55ae08c8987 ("cpufreq: Avoid unnecessary frequency updates due to mismatch")
> Signed-off-by: Sanjay Chandrashekara <sanjayc@nvidia.com>
> [ sumit gupta: Commit message update ]
> Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
> ---
>  drivers/cpufreq/cpufreq.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> index 8b0509f89f1b..6b52ebe5a890 100644
> --- a/drivers/cpufreq/cpufreq.c
> +++ b/drivers/cpufreq/cpufreq.c
> @@ -1732,7 +1732,7 @@ static unsigned int cpufreq_verify_current_freq(struct cpufreq_policy *policy, b
>                  * MHz. In such cases it is better to avoid getting into
>                  * unnecessary frequency updates.
>                  */
> -               if (abs(policy->cur - new_freq) < HZ_PER_MHZ)
> +               if (abs(policy->cur - new_freq) < KHZ_PER_MHZ)
>                         return policy->cur;
>
>                 cpufreq_out_of_sync(policy, new_freq);
> --

So this is a fix that can be applied separately from the rest of the
series, isn't it?
Rafael J. Wysocki April 18, 2023, 3:47 p.m. UTC | #2 
On Tue, Apr 18, 2023 at 3:32 PM Sumit Gupta <sumitg@nvidia.com> wrote:
>
>
>
> On 18/04/23 18:27, Rafael J. Wysocki wrote:
> > External email: Use caution opening links or attachments
> >
> >
> > On Tue, Apr 18, 2023 at 1:35 PM Sumit Gupta <sumitg@nvidia.com> wrote:
> >>
> >> From: Sanjay Chandrashekara <sanjayc@nvidia.com>
> >>
> >> cpufreq_verify_current_freq checks if the frequency returned by
> >> the hardware has a slight delta with the valid frequency value
> >> last set and returns "policy->cur" if the delta is within "1 MHz".
> >> In the comparison, "policy->cur" is in "kHz" but it's compared
> >> against HZ_PER_MHZ. So, the comparison range becomes "1 GHz".
> >> Fix this by comparing against KHZ_PER_MHZ instead of HZ_PER_MHZ.
> >>
> >> Fixes: f55ae08c8987 ("cpufreq: Avoid unnecessary frequency updates due to mismatch")
> >> Signed-off-by: Sanjay Chandrashekara <sanjayc@nvidia.com>
> >> [ sumit gupta: Commit message update ]
> >> Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
> >> ---
> >>   drivers/cpufreq/cpufreq.c | 2 +-
> >>   1 file changed, 1 insertion(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> >> index 8b0509f89f1b..6b52ebe5a890 100644
> >> --- a/drivers/cpufreq/cpufreq.c
> >> +++ b/drivers/cpufreq/cpufreq.c
> >> @@ -1732,7 +1732,7 @@ static unsigned int cpufreq_verify_current_freq(struct cpufreq_policy *policy, b
> >>                   * MHz. In such cases it is better to avoid getting into
> >>                   * unnecessary frequency updates.
> >>                   */
> >> -               if (abs(policy->cur - new_freq) < HZ_PER_MHZ)
> >> +               if (abs(policy->cur - new_freq) < KHZ_PER_MHZ)
> >>                          return policy->cur;
> >>
> >>                  cpufreq_out_of_sync(policy, new_freq);
> >> --
> >
> > So this is a fix that can be applied separately from the rest of the
> > series, isn't it?
>
> Yes.

So applied as 6.4 material.