2013-09-11 07:14:39

by Lan Tianyu

Subject: [PATCH] Cpufreq: Acquire read lock in the cpufreq_policy_restore() rather than write lock

From: Lan Tianyu <[email protected]>

In the cpufreq_policy_restore(), policy before system suspend is read from
from percpu's cpufreq_cpu_data_fallback. It's read operation rather than
write operation. So convert write lock to read lock

Signed-off-by: Lan Tianyu <[email protected]>
---
drivers/cpufreq/cpufreq.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
index 81ceea6..b762f9b 100644
--- a/drivers/cpufreq/cpufreq.c
+++ b/drivers/cpufreq/cpufreq.c
@@ -912,11 +912,11 @@ static struct cpufreq_policy *cpufreq_policy_restore(unsigned int cpu)
struct cpufreq_policy *policy;
unsigned long flags;

- write_lock_irqsave(&cpufreq_driver_lock, flags);
+ read_lock_irqsave(&cpufreq_driver_lock, flags);

policy = per_cpu(cpufreq_cpu_data_fallback, cpu);

- write_unlock_irqrestore(&cpufreq_driver_lock, flags);
+ read_unlock_irqrestore(&cpufreq_driver_lock, flags);

return policy;
}
--
1.8.4.rc0.1.g8f6a3e5.dirty

2013-09-11 08:13:36

by Srivatsa S. Bhat

Subject: Re: [PATCH] Cpufreq: Acquire read lock in the cpufreq_policy_restore() rather than write lock

On 09/11/2013 12:35 PM, [email protected] wrote:
> From: Lan Tianyu <[email protected]>
>
> In the cpufreq_policy_restore(), policy before system suspend is read from
> from percpu's cpufreq_cpu_data_fallback. It's read operation rather than
> write operation. So convert write lock to read lock
>
> Signed-off-by: Lan Tianyu <[email protected]>

Reviewed-by: Srivatsa S. Bhat <[email protected]>

I noticed it too yesterday, while looking at something else. Thanks for
fixing this!

> ---
> drivers/cpufreq/cpufreq.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> index 81ceea6..b762f9b 100644
> --- a/drivers/cpufreq/cpufreq.c
> +++ b/drivers/cpufreq/cpufreq.c
> @@ -912,11 +912,11 @@ static struct cpufreq_policy *cpufreq_policy_restore(unsigned int cpu)
> struct cpufreq_policy *policy;
> unsigned long flags;
>
> - write_lock_irqsave(&cpufreq_driver_lock, flags);
> + read_lock_irqsave(&cpufreq_driver_lock, flags);
>
> policy = per_cpu(cpufreq_cpu_data_fallback, cpu);
>
> - write_unlock_irqrestore(&cpufreq_driver_lock, flags);
> + read_unlock_irqrestore(&cpufreq_driver_lock, flags);
>
> return policy;
> }
>

Regards,
Srivatsa S. Bhat

2013-09-11 08:44:53

by Viresh Kumar

Subject: Re: [PATCH] Cpufreq: Acquire read lock in the cpufreq_policy_restore() rather than write lock

On 11 September 2013 12:35, <[email protected]> wrote:
> From: Lan Tianyu <[email protected]>
>
> In the cpufreq_policy_restore(), policy before system suspend is read from
> from percpu's cpufreq_cpu_data_fallback. It's read operation rather than
> write operation. So convert write lock to read lock
>
> Signed-off-by: Lan Tianyu <[email protected]>
> ---
> drivers/cpufreq/cpufreq.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> index 81ceea6..b762f9b 100644
> --- a/drivers/cpufreq/cpufreq.c
> +++ b/drivers/cpufreq/cpufreq.c
> @@ -912,11 +912,11 @@ static struct cpufreq_policy *cpufreq_policy_restore(unsigned int cpu)
> struct cpufreq_policy *policy;
> unsigned long flags;
>
> - write_lock_irqsave(&cpufreq_driver_lock, flags);
> + read_lock_irqsave(&cpufreq_driver_lock, flags);
>
> policy = per_cpu(cpufreq_cpu_data_fallback, cpu);
>
> - write_unlock_irqrestore(&cpufreq_driver_lock, flags);
> + read_unlock_irqrestore(&cpufreq_driver_lock, flags);
>
> return policy;
> }

Acked-by: Viresh Kumar <[email protected]>

2013-09-11 22:45:04

by Rafael J. Wysocki

Subject: Re: [PATCH] Cpufreq: Acquire read lock in the cpufreq_policy_restore() rather than write lock

On Wednesday, September 11, 2013 03:05:05 PM [email protected] wrote:
> From: Lan Tianyu <[email protected]>
>
> In the cpufreq_policy_restore(), policy before system suspend is read from
> from percpu's cpufreq_cpu_data_fallback. It's read operation rather than
> write operation. So convert write lock to read lock
>
> Signed-off-by: Lan Tianyu <[email protected]>

Applied, thanks Tianyu!

> ---
> drivers/cpufreq/cpufreq.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> index 81ceea6..b762f9b 100644
> --- a/drivers/cpufreq/cpufreq.c
> +++ b/drivers/cpufreq/cpufreq.c
> @@ -912,11 +912,11 @@ static struct cpufreq_policy *cpufreq_policy_restore(unsigned int cpu)
> struct cpufreq_policy *policy;
> unsigned long flags;
>
> - write_lock_irqsave(&cpufreq_driver_lock, flags);
> + read_lock_irqsave(&cpufreq_driver_lock, flags);
>
> policy = per_cpu(cpufreq_cpu_data_fallback, cpu);
>
> - write_unlock_irqrestore(&cpufreq_driver_lock, flags);
> + read_unlock_irqrestore(&cpufreq_driver_lock, flags);
>
> return policy;
> }
>
--
I speak only for myself.
Rafael J. Wysocki, Intel Open Source Technology Center.