Re: [PATCH-next 3/6] cgroup/cpuset: Fix incorrect effective_cpus bug in update_cpumasks_hier()

[Waiman Long <longman@redhat.com>]

On 10/23/26 3:07 AM, Ridong Chen wrote:
>
>
> On 10/22/2026 11:16 AM, Waiman Long wrote:
>> With the introduction of cpuset.cpus.exclusive, the computation of
>> effective_cpus differs depending on if a cpuset is a valid partition
>> root or not.
>>
>> In update_cpumasks_hier(), effective_cpus is computed early
>> to check if the current subtree can be skipped. Later on
>> update_parent_effective_cpumask() may be called which can change the
>> partition state from valid to invalid or vice versa. If the partition
>> state changes, the previously computed effective_cpus may no longer be
>> valid. Fix that by recomputing effective_cpus again when that happens.
>>
>> Also add a new test case into test_cpuset_prs.sh to verify this change.
>>
>> Fixes: e2ffe502ba45 ("cgroup/cpuset: Add cpuset.cpus.exclusive for v2")
>> Signed-off-by: Waiman Long <longman@redhat.com>
>> ---
>>   kernel/cgroup/cpuset.c                            | 10 ++++++++--
>>   tools/testing/selftests/cgroup/test_cpuset_prs.sh |  3 +++
>>   2 files changed, 11 insertions(+), 2 deletions(-)
>>
>> diff --git a/kernel/cgroup/cpuset.c b/kernel/cgroup/cpuset.c
>> index db6e4663f4be..7b9062326f35 100644
>> --- a/kernel/cgroup/cpuset.c
>> +++ b/kernel/cgroup/cpuset.c
>> @@ -2316,10 +2316,16 @@ static void update_cpumasks_hier(struct 
>> cpuset *cs, struct tmpmasks *tmp,
>>           if (update_parent) {
>>               update_parent_effective_cpumask(cp, partcmd_update, 
>> NULL, tmp);
>>               /*
>> -             * The cpuset partition_root_state may become
>> -             * invalid. Capture it.
>> +             * Capture current partition_root_state as it may have
>> +             * changed. Recompute effective cpumask if changed.
>>                */
>>               new_prs = cp->partition_root_state;
>> +            if (new_prs != old_prs) {
>> +                if (new_prs > 0)
>> +                    compute_partition_effective_cpumask(cp, 
>> tmp->new_cpus);
>> +                else
>> +                    compute_effective_cpumask(tmp->new_cpus, cp, 
>> parent);
>> +            }
>>           }
>
> If cs is a partition (remote or local), we call 
> compute_partition_effective_cpumask to compute the effective cpumask; 
> otherwise we should call compute_effective_cpumask. Wouldn't it be 
> better to call compute_partition_effective_cpumask whenever cs is a 
> valid partition, so we can eliminate the branch outside?
>
> ```
> static void compute_effective_cpumask(struct cpumask *new_cpus,
>                       struct cpuset *cs, struct cpuset *parent)
> {
>     bool has_cpus;
>
>     if (is_remote_partition (cs) ||
>         (is_partition_valid(cs) && is_partition_valid(parent)))
>         return compute_partition_effective_cpumask(cs, &new_cpus);
>
>     has_cpus = cpumask_and(new_cpus, cs->cpus_allowed, 
> parent->effective_cpus);
>     if (!has_cpus && is_in_v2_mode())
>         cpumask_copy(new_cpus, parent->effective_cpus);
> }
> ```
>
Good point. I will send a separate patch to do the consolidation.

Cheers,
Longman

>> spin_lock_irq(&callback_lock);
>> diff --git a/tools/testing/selftests/cgroup/test_cpuset_prs.sh 
>> b/tools/testing/selftests/cgroup/test_cpuset_prs.sh
>> index 47791591002f..8595b6b4b7bf 100755
>> --- a/tools/testing/selftests/cgroup/test_cpuset_prs.sh
>> +++ b/tools/testing/selftests/cgroup/test_cpuset_prs.sh
>> @@ -364,6 +364,9 @@ TEST_MATRIX=(
>>       " C1-4:X2-4 C2-4:X3-4:P1 C3-4:X4:P2 \
>>                      .      .      X5      .      .    0 
>> A1:1-4|A2:2-4|A3:3-4 \
>>                                          A1:P0|A2:P-1|A3:P-2 ."
>> +    " C1-5:X3-5:P2 C3-5:X4-5:P2 C4-5:X5:P2 \
>> +                   .    X1-4     .       .      .    0 
>> A1:1-3|A2:4|A3:4 \
>> +                                       A1:P2|A2:P2|A3:P-2 1-4"
>>       " C1-4:X2-4 C1-4:X3-4:P1 C3-4:X4:P2 \
>>                      .      .      .      X1      .    0 
>> A1:1-2|A2:3-4|A3:3-4 \
>>                                          A1:P0|A2:P1|A3:P-2 ."
>