linux-pm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] cpufreq: schedutil: Fix covert rate_limit_us to freq_update_delay_ns overflow
@ 2019-07-08  8:46 ZhangXiaoxu
  2019-07-08  9:37 ` Viresh Kumar
  0 siblings, 1 reply; 3+ messages in thread
From: ZhangXiaoxu @ 2019-07-08  8:46 UTC (permalink / raw)
  To: rjw, viresh.kumar, mingo, peterz, linux-pm, zhangxiaoxu5

When covert rate_limit_us to freq_update_delay_ns, it maybe overflow
and lead an undefined behavior.

So, limit the rate_limit_us to UINT_MAX / NSEC_PER_USEC.

Signed-off-by: ZhangXiaoxu <zhangxiaoxu5@huawei.com>
---
 kernel/sched/cpufreq_schedutil.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c
index 962cf34..01e05f3 100644
--- a/kernel/sched/cpufreq_schedutil.c
+++ b/kernel/sched/cpufreq_schedutil.c
@@ -590,6 +590,9 @@ rate_limit_us_store(struct gov_attr_set *attr_set, const char *buf, size_t count
 	if (kstrtouint(buf, 10, &rate_limit_us))
 		return -EINVAL;
 
+	if (rate_limit_us > UINT_MAX / NSEC_PER_USEC)
+		return -EINVAL;
+
 	tunables->rate_limit_us = rate_limit_us;
 
 	list_for_each_entry(sg_policy, &attr_set->policy_list, tunables_hook)
-- 
2.7.4


^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH] cpufreq: schedutil: Fix covert rate_limit_us to freq_update_delay_ns overflow
  2019-07-08  8:46 [PATCH] cpufreq: schedutil: Fix covert rate_limit_us to freq_update_delay_ns overflow ZhangXiaoxu
@ 2019-07-08  9:37 ` Viresh Kumar
  2019-07-08 10:47   ` zhangxiaoxu (A)
  0 siblings, 1 reply; 3+ messages in thread
From: Viresh Kumar @ 2019-07-08  9:37 UTC (permalink / raw)
  To: ZhangXiaoxu; +Cc: rjw, mingo, peterz, linux-pm

On 08-07-19, 16:46, ZhangXiaoxu wrote:
> When covert rate_limit_us to freq_update_delay_ns, it maybe overflow
> and lead an undefined behavior.

freq_update_delay_ns is s64, still overflow can happen ?

> So, limit the rate_limit_us to UINT_MAX / NSEC_PER_USEC.
> 
> Signed-off-by: ZhangXiaoxu <zhangxiaoxu5@huawei.com>
> ---
>  kernel/sched/cpufreq_schedutil.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c
> index 962cf34..01e05f3 100644
> --- a/kernel/sched/cpufreq_schedutil.c
> +++ b/kernel/sched/cpufreq_schedutil.c
> @@ -590,6 +590,9 @@ rate_limit_us_store(struct gov_attr_set *attr_set, const char *buf, size_t count
>  	if (kstrtouint(buf, 10, &rate_limit_us))
>  		return -EINVAL;
>  
> +	if (rate_limit_us > UINT_MAX / NSEC_PER_USEC)
> +		return -EINVAL;
> +
>  	tunables->rate_limit_us = rate_limit_us;
>  
>  	list_for_each_entry(sg_policy, &attr_set->policy_list, tunables_hook)
> -- 
> 2.7.4

-- 
viresh

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] cpufreq: schedutil: Fix covert rate_limit_us to freq_update_delay_ns overflow
  2019-07-08  9:37 ` Viresh Kumar
@ 2019-07-08 10:47   ` zhangxiaoxu (A)
  0 siblings, 0 replies; 3+ messages in thread
From: zhangxiaoxu (A) @ 2019-07-08 10:47 UTC (permalink / raw)
  To: Viresh Kumar; +Cc: rjw, mingo, peterz, linux-pm



在 2019/7/8 17:37, Viresh Kumar 写道:
> On 08-07-19, 16:46, ZhangXiaoxu wrote:
>> When covert rate_limit_us to freq_update_delay_ns, it maybe overflow
>> and lead an undefined behavior.
> 
> freq_update_delay_ns is s64, still overflow can happen ?
It will not happened. sorry for my misunderstander :(
> 
>> So, limit the rate_limit_us to UINT_MAX / NSEC_PER_USEC.
>>
>> Signed-off-by: ZhangXiaoxu <zhangxiaoxu5@huawei.com>
>> ---
>>   kernel/sched/cpufreq_schedutil.c | 3 +++
>>   1 file changed, 3 insertions(+)
>>
>> diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c
>> index 962cf34..01e05f3 100644
>> --- a/kernel/sched/cpufreq_schedutil.c
>> +++ b/kernel/sched/cpufreq_schedutil.c
>> @@ -590,6 +590,9 @@ rate_limit_us_store(struct gov_attr_set *attr_set, const char *buf, size_t count
>>   	if (kstrtouint(buf, 10, &rate_limit_us))
>>   		return -EINVAL;
>>   
>> +	if (rate_limit_us > UINT_MAX / NSEC_PER_USEC)
>> +		return -EINVAL;
>> +
>>   	tunables->rate_limit_us = rate_limit_us;
>>   
>>   	list_for_each_entry(sg_policy, &attr_set->policy_list, tunables_hook)
>> -- 
>> 2.7.4
> 


^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2019-07-08 10:47 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-07-08  8:46 [PATCH] cpufreq: schedutil: Fix covert rate_limit_us to freq_update_delay_ns overflow ZhangXiaoxu
2019-07-08  9:37 ` Viresh Kumar
2019-07-08 10:47   ` zhangxiaoxu (A)

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).