Re: [RFC PATCH] cpufreq/hotplug: Fix cpu-hotplug cpufreq race conditions

[Preeti U Murthy <preeti@linux.vnet.ibm.com>]

On 06/02/2015 11:41 AM, Viresh Kumar wrote:
> On 02-06-15, 11:33, Preeti U Murthy wrote:
>> No, dbs_data is a governor wide data structure and not a policy wide
> 
> Yeah, that's the common part which I was referring to. But normally
> its just read for policies in START/STOP, they just update per-cpu
> data for policy->cpus.
> 
>> one, which is manipulated in START/STOP calls for drivers where the
>> CPUFREQ_HAVE_GOVERNOR_PER_POLICY is not set.
>>
>> So even if we assume that we hold per-policy locks, the following race
>> is still present. Assume that we have just two cpus which do not have a
>> governor-per-policy set.
>>
>> CPU0                        CPU1
>>
>>  store*                     store*
>>
>> lock(policy 1)              lock(policy 2)
>> cpufreq_set_policy()       cpufreq_set_policy()
>> EXIT() :
>> dbs-data->usage_count--
>>
>> INIT()
>> dbs_data exists
> 
> You missed the usage_count++ here.

Ok, sorry about that. How about the below ?
> 
>> so return
>>                             EXIT()
>>                             dbs_data->usage_count -- = 0
>>                             kfree(dbs_data)
> 
>                                 And so this shouldn't happen. Else we
>                                 are missing locking in governor's
>                                 code, rather than cpufreq.c
> 
 CPU0                        CPU1

store*                     store*

lock(policy 1)              lock(policy 2)
cpufreq_set_policy()       cpufreq_set_policy()
EXIT() :
dbs-data->usage_count--

INIT()                      EXIT()
dbs_data exists             dbs_data->usage_count -- = 0
                            kfree(dbs_data)
dbs-data->usage_count++
*NULL dereference*

The point is there are potential race conditions. Its just a matter of
interleaving ?

Regards
Preeti U Murthy