Re: [PATCH v3 2/3] mm: khugepaged: fix folio is used after pte_unmap_unlock()

["David Hildenbrand (Arm)" <david@kernel.org>]

On 8/26/26 04:46, Vernon Yang wrote:
> On Mon, Aug 24, 2026 at 01:57:15PM +0200, David Hildenbrand (Arm) wrote:
>> On 8/24/26 11:29, Vernon Yang wrote:
>>> From: Vernon Yang <yanglincheng@kylinos.cn>
>>>
>>> After the page table lock has dropped, the folio can be freed
>>> concurrently. The trace_mm_khugepaged_scan_pmd() is left with
>>> a dangling folio pointer.
>>>
>>> So using the folio_pfn() before dropping the page table lock,
>>> closing use-after-free window.
>>>
>>> Fixes: 7d2eba0557c1 ("mm: add tracepoint for scanning pages")
>>> Cc: stable@vger.kernel.org
>>> Signed-off-by: Vernon Yang <yanglincheng@kylinos.cn>
>>> ---
>>>  include/trace/events/huge_memory.h | 6 +++---
>>>  mm/khugepaged.c                    | 5 ++++-
>>>  2 files changed, 7 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/include/trace/events/huge_memory.h b/include/trace/events/huge_memory.h
>>> index 7b526528f85b..fa828967e1fb 100644
>>> --- a/include/trace/events/huge_memory.h
>>> +++ b/include/trace/events/huge_memory.h
>>> @@ -55,10 +55,10 @@ SCAN_STATUS
>>>
>>>  TRACE_EVENT(mm_khugepaged_scan_pmd,
>>>
>>> -	TP_PROTO(struct mm_struct *mm, struct folio *folio,
>>> +	TP_PROTO(struct mm_struct *mm, unsigned long pfn,
>>>  		 int referenced, int none_or_zero, int status, int unmapped),
>>>
>>> -	TP_ARGS(mm, folio, referenced, none_or_zero, status, unmapped),
>>> +	TP_ARGS(mm, pfn, referenced, none_or_zero, status, unmapped),
>>>
>>>  	TP_STRUCT__entry(
>>>  		__field(struct mm_struct *, mm)
>>> @@ -71,7 +71,7 @@ TRACE_EVENT(mm_khugepaged_scan_pmd,
>>>
>>>  	TP_fast_assign(
>>>  		__entry->mm = mm;
>>> -		__entry->pfn = folio ? folio_pfn(folio) : -1;
>>> +		__entry->pfn = pfn;
>>>  		__entry->referenced = referenced;
>>>  		__entry->none_or_zero = none_or_zero;
>>>  		__entry->status = status;
>>> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
>>> index 00337405c0e0..4e0fca5942dd 100644
>>> --- a/mm/khugepaged.c
>>> +++ b/mm/khugepaged.c
>>> @@ -1618,6 +1618,7 @@ static enum scan_result collapse_scan_pmd(struct mm_struct *mm,
>>>  	enum scan_result result = SCAN_FAIL;
>>>  	struct page *page = NULL;
>>>  	struct folio *folio = NULL;
>>> +	unsigned long pfn = -1;
>>>  	unsigned long addr;
>>>  	unsigned long enabled_orders;
>>>  	spinlock_t *ptl;
>>> @@ -1780,6 +1781,8 @@ static enum scan_result collapse_scan_pmd(struct mm_struct *mm,
>>>  		result = SCAN_SUCCEED;
>>>  	}
>>>  out_unmap:
>>> +	if (folio)
>>> +		pfn = folio_pfn(folio);
>>
>> Should we reset the folio to NULL at the beginning of the loop? Then we really
>> only trace the PFN if it really was problematic.
> 
> Yes, this is a pre-existing bug, and I'll fix it together. Thanks!
> 
> But it is not at the beginning of the loop, it is at the __ending__ of
> the loop, for the same reason as PATCH#1.

See my reply to patch #1, hth.

-- 
Cheers,

David