Re: NULL pointer dereference in pick_next_task_fair

[Valentin Schneider]

On 08/11/2019 11:02, Quentin Perret wrote:
>> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
>> index a14487462b6c..6b983214e00f 100644
>> --- a/kernel/sched/fair.c
>> +++ b/kernel/sched/fair.c
>> @@ -6746,10 +6746,18 @@ done: __maybe_unused;
>>  	return NULL;
>>  }
>>
>> +static int balance_fair(struct rq *rq, struct task_struct *prev, struct rq_flags *rf)
>> +{
>> +	if (rq->cfs.nr_running)
>> +		return 1;
>> +
>> +	return newidle_balance(rq, rf) != 0;
> 
> And you can ignore the RETRY_TASK case here under the assumption that
> we must have tried to pull from RT/DL before ending up here ?
> 

I think we can ignore RETRY_TASK because this happens before the picking loop,
so we'll observe any new DL/RT task that got enqueued while newidle released
the lock. This also means we can safely break the balance loop in
pick_next_task() when we get RETRY_TASK, because we've got something to pick
(some new RT/DL task). This wants a comment though, methinks.

Other than that I agree with Quentin, it's a much cleaner approach and I quite
like it.

> Thanks,
> Quentin
>