Re: [PATCH] sched/core: Mask out offline CPUs when user_cpus_ptr is used

[Waiman Long <llong@redhat.com>]

On 7/17/25 10:42 PM, Chen Ridong wrote:
>
> On 2025/7/15 23:58, Waiman Long wrote:
>> Chen Ridong reported that cpuset could report a kernel warning for a task
>> due to set_cpus_allowed_ptr() returning failure in the corner case that:
>>
>> 1) the task used sched_setaffinity(2) to set its CPU affinity mask to
>>     be the same as the cpuset.cpus of its cpuset,
>> 2) all the CPUs assigned to that cpuset were taken offline, and
>> 3) cpuset v1 is in use and the task had to be migrated to the top cpuset.
>>
>> Due to the fact that CPU affinity of the tasks in the top cpuset are
>> not updated when a CPU hotplug online/offline event happens, offline
>> CPUs are included in CPU affinity of those tasks. It is possible
>> that further masking with user_cpus_ptr set by sched_setaffinity(2)
>> in __set_cpus_allowed_ptr() will leave only offline CPUs in the new
>> mask causing the subsequent call to __set_cpus_allowed_ptr_locked()
>> to return failure with an empty CPU affinity.
>>
>> Fix this failure by masking out offline CPUs when user_cpus_ptr masking
>> has to be done and fall back to ignoring user_cpus_ptr if the resulting
>> cpumask is empty.
>>
>> Reported-by: Chen Ridong <chenridong@huaweicloud.com>
>> Closes: https://lore.kernel.org/lkml/20250714032311.3570157-1-chenridong@huaweicloud.com/
>> Fixes: da019032819a ("sched: Enforce user requested affinity")
>> Signed-off-by: Waiman Long <longman@redhat.com>
>> ---
>>   kernel/sched/core.c | 7 ++++++-
>>   1 file changed, 6 insertions(+), 1 deletion(-)
>>
>> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
>> index 81c6df746df1..4cf25dd8827f 100644
>> --- a/kernel/sched/core.c
>> +++ b/kernel/sched/core.c
>> @@ -3172,10 +3172,15 @@ int __set_cpus_allowed_ptr(struct task_struct *p, struct affinity_context *ctx)
>>   	/*
>>   	 * Masking should be skipped if SCA_USER or any of the SCA_MIGRATE_*
>>   	 * flags are set.
>> +	 *
>> +	 * Even though the given new_mask must have at least one online CPU,
>> +	 * masking with user_cpus_ptr may strip out all online CPUs causing
>> +	 * failure. So offline CPUs have to be masked out too.
>>   	 */
>>   	if (p->user_cpus_ptr &&
>>   	    !(ctx->flags & (SCA_USER | SCA_MIGRATE_ENABLE | SCA_MIGRATE_DISABLE)) &&
>> -	    cpumask_and(rq->scratch_mask, ctx->new_mask, p->user_cpus_ptr))
>> +	    cpumask_and(rq->scratch_mask, ctx->new_mask, p->user_cpus_ptr) &&
>> +	    cpumask_and(rq->scratch_mask, rq->scratch_mask, cpu_active_mask))
>>   		ctx->new_mask = rq->scratch_mask;
>>   
>>   	return __set_cpus_allowed_ptr_locked(p, ctx, rq, &rf);
> Hi, Waiman,
> Would the following modification make more sense?
>
>    	if (p->user_cpus_ptr &&
>    	    !(ctx->flags & (SCA_USER | SCA_MIGRATE_ENABLE | SCA_MIGRATE_DISABLE)) &&
>   -	    cpumask_and(rq->scratch_mask, ctx->new_mask, p->user_cpus_ptr))
>   +	    cpumask_and(rq->scratch_mask, ctx->new_mask, p->user_cpus_ptr) &&
>   +	    cpumask_intersects(rq->scratch_mask, cpu_active_mask))
>    		ctx->new_mask = rq->scratch_mask;
>
> This can preserve user intent as much as possible.

I realized that I should have used cpumask_intersects() instead after 
sending out this patch. It looks like you have come to the same 
conclusion. I will send out a v2 to update that.

Thanks,
Longman