Re: [PATCH v4 1/2] mm/memory-failure: fix missing ->mf_stats count in hugetlb poison

[jane.chu@oracle.com]

On 1/14/2026 7:37 AM, Chris Mason wrote:
> On Tue, 13 Jan 2026 01:07:50 -0700 Jane Chu <jane.chu@oracle.com> wrote:
> 
>> When a newly poisoned subpage ends up in an already poisoned hugetlb
>> folio, 'num_poisoned_pages' is incremented, but the per node ->mf_stats
>> is not. Fix the inconsistency by designating action_result() to update
>> them both.
>>
>> While at it, define __get_huge_page_for_hwpoison() return values in terms
>> of symbol names for better readibility. Also rename
>> folio_set_hugetlb_hwpoison() to hugetlb_update_hwpoison() since the
>> function does more than the conventional bit setting and the fact
>> three possible return values are expected.
>>
> 
> Hi everyone,
> 
> I ran mm-new through the AI patch review prompts, and this commit was flagged.
> The review below looks right to me:
> 
>> diff --git a/mm/memory-failure.c b/mm/memory-failure.c
>> --- a/mm/memory-failure.c
>> +++ b/mm/memory-failure.c
> 
> [ ... ]
> 
>> @@ -2029,22 +2037,29 @@ static int try_memory_failure_hugetlb(unsigned long pfn, int flags, int *hugetlb)
>>   	*hugetlb = 1;
>>   retry:
>>   	res = get_huge_page_for_hwpoison(pfn, flags, &migratable_cleared);
>> -	if (res == 2) { /* fallback to normal page handling */
>> +	switch (res) {
>> +	case -EINVAL:	/* fallback to normal page handling */
>>   		*hugetlb = 0;
>>   		return 0;
>> -	} else if (res == -EHWPOISON) {
>> -		if (flags & MF_ACTION_REQUIRED) {
>> -			folio = page_folio(p);
>> -			res = kill_accessing_process(current, folio_pfn(folio), flags);
>> -		}
>> -		action_result(pfn, MF_MSG_ALREADY_POISONED, MF_FAILED);
>> -		return res;
>> -	} else if (res == -EBUSY) {
>> +	case -EBUSY:
>>   		if (!(flags & MF_NO_RETRY)) {
>>   			flags |= MF_NO_RETRY;
>>   			goto retry;
>>   		}
>>   		return action_result(pfn, MF_MSG_GET_HWPOISON, MF_IGNORED);
>> +	case MF_HUGETLB_FOLIO_PRE_POISONED:
>> +	case MF_HUGETLB_PAGE_PRE_POISON:
>> +		if (flags & MF_ACTION_REQUIRED) {
>> +			folio = page_folio(p);
>> +			res = kill_accessing_process(current, folio_pfn(folio), flags);
>> +		}
>> +		if (res == MF_HUGETLB_FOLIO_PRE_POISONED)
> 
> When MF_ACTION_REQUIRED is set, kill_accessing_process() overwrites res
> before this check. kill_accessing_process() returns either 0 or
> -EHWPOISON, so the check for MF_HUGETLB_FOLIO_PRE_POISONED (which is 3)
> will always be false in that path.
> 
> This means action_result() will be called with MF_MSG_HUGE instead of
> MF_MSG_ALREADY_POISONED. Looking at action_result():
> 
>      if (type != MF_MSG_ALREADY_POISONED && type != MF_MSG_PFN_MAP) {
>          num_poisoned_pages_inc(pfn);
>          update_per_node_mf_stats(pfn, result);
>      }
> 
> Does this cause incorrect counter increments for already-poisoned pages
> when MF_ACTION_REQUIRED is set? The original code called action_result()
> unconditionally with MF_MSG_ALREADY_POISONED regardless of whether
> kill_accessing_process() was called.
> 
> Should a separate variable preserve the original res value before the
> call to kill_accessing_process()?

Good catch!  William Roche has also pointed out the same issue to me in 
private.  I will fix it in v5 soon.

thanks,
-jane


> 
>> +			action_result(pfn, MF_MSG_ALREADY_POISONED, MF_FAILED);
>> +		else
>> +			action_result(pfn, MF_MSG_HUGE, MF_FAILED);
>> +		return res;
> 
> 
>