Re: [stable-6.6.y] mm: khugepaged refuses to freeze

[Baolin Wang <baolin.wang@linux.alibaba.com>]

On 2/10/26 6:21 PM, David Hildenbrand (Arm) wrote:
> On 2/10/26 11:07, Baolin Wang wrote:
>>
>>
>> On 2/10/26 11:21 AM, Sergey Senozhatsky wrote:
>>> On (26/02/06 10:00), David Hildenbrand (Arm) wrote:
>>>>
>>>> Right, mimicking what kswapd does sound reasonable!
>>>
>>> I may be missing something, as I'm not seeing dev_pm_ops in vmscan code.
>>> Would something like this work?
>>>
>>> ---
>>>
>>> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
>>> index fa6a018b20a8..c5d89ec223d3 100644
>>> --- a/mm/khugepaged.c
>>> +++ b/mm/khugepaged.c
>>> @@ -394,8 +394,12 @@ static inline int 
>>> hpage_collapse_test_exit(struct mm_struct *mm)
>>>   static inline int hpage_collapse_test_exit_or_disable(struct 
>>> mm_struct *mm)
>>>   {
>>> +    bool was_frozen;
>>> +    int ret = kthread_freezable_should_stop(&was_frozen);
>>> +
>>>       return hpage_collapse_test_exit(mm) ||
>>> -        mm_flags_test(MMF_DISABLE_THP_COMPLETELY, mm);
>>> +        mm_flags_test(MMF_DISABLE_THP_COMPLETELY, mm) ||
>>> +        was_frozen || ret;
>>>   }
>>
>> Since the hpage_collapse_test_exit_or_disable() can be called by 
>> madvise_callapse(), which is not a kernel thread. 
> 
> Which raises the question whether we should forward that context 
> (khugepaged vs. madvise) to hpage_collapse_test_exit_or_disable().

Passing in the 'cc' pointer looks fine to me. Something like:

static inline int hpage_collapse_test_exit_or_disable(struct mm_struct *mm,
                                                 struct collapse_control 
*cc)
{
         bool was_frozen = false;

         if (cc->is_khugepaged && 
unlikely(kthread_freezable_should_stop(&was_frozen)))
                 return 1;

         return hpage_collapse_test_exit(mm) ||
                 mm_flags_test(MMF_DISABLE_THP_COMPLETELY, mm) ||
                 was_frozen;
}

Sergey, could you submit a formal patch for review?