Re: [PATCH v4 07/13] mm: khugepaged: collapse_pte_mapped_thp() use pte_offset_map_rw_nolock()

[Muchun Song <muchun.song@linux.dev>]

> On Sep 24, 2024, at 16:57, Qi Zheng <zhengqi.arch@bytedance.com> wrote:
> 
> 
> 
> On 2024/9/24 16:52, Muchun Song wrote:
>>> On Sep 24, 2024, at 15:29, Qi Zheng <zhengqi.arch@bytedance.com> wrote:
>>> 
>>> 
>>> 
>>> On 2024/9/24 15:14, Muchun Song wrote:
>>>>> On Sep 24, 2024, at 14:11, Qi Zheng <zhengqi.arch@bytedance.com> wrote:
>>>>> In collapse_pte_mapped_thp(), we may modify the pte and pmd entry after
>>>>> acquring the ptl, so convert it to using pte_offset_map_rw_nolock(). At
>>>>> this time, the pte_same() check is not performed after the PTL held. So we
>>>>> should get pgt_pmd and do pmd_same() check after the ptl held.
>>>>> 
>>>>> Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
>>>>> ---
>>>>> mm/khugepaged.c | 14 +++++++++++---
>>>>> 1 file changed, 11 insertions(+), 3 deletions(-)
>>>>> 
>>>>> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
>>>>> index 6498721d4783a..8ab79c13d077f 100644
>>>>> --- a/mm/khugepaged.c
>>>>> +++ b/mm/khugepaged.c
>>>>> @@ -1605,7 +1605,7 @@ int collapse_pte_mapped_thp(struct mm_struct *mm, unsigned long addr,
>>>>>    if (userfaultfd_armed(vma) && !(vma->vm_flags & VM_SHARED))
>>>>>        pml = pmd_lock(mm, pmd);
>>>>> 
>>>>> -    start_pte = pte_offset_map_nolock(mm, pmd, haddr, &ptl);
>>>>> +    start_pte = pte_offset_map_rw_nolock(mm, pmd, haddr, &pgt_pmd, &ptl);
>>>>>    if (!start_pte)        /* mmap_lock + page lock should prevent this */
>>>>>        goto abort;
>>>>>    if (!pml)
>>>>> @@ -1613,6 +1613,9 @@ int collapse_pte_mapped_thp(struct mm_struct *mm, unsigned long addr,
>>>>>    else if (ptl != pml)
>>>>>        spin_lock_nested(ptl, SINGLE_DEPTH_NESTING);
>>>>> 
>>>>> +    if (unlikely(!pmd_same(pgt_pmd, pmdp_get_lockless(pmd))))
>>>>> +        goto abort;
>>>>> +
>>>>>    /* step 2: clear page table and adjust rmap */
>>>>>    for (i = 0, addr = haddr, pte = start_pte;
>>>>>         i < HPAGE_PMD_NR; i++, addr += PAGE_SIZE, pte++) {
>>>>> @@ -1645,7 +1648,6 @@ int collapse_pte_mapped_thp(struct mm_struct *mm, unsigned long addr,
>>>>>        nr_ptes++;
>>>>>    }
>>>>> 
>>>>> -    pte_unmap(start_pte);
>>>>>    if (!pml)
>>>>>        spin_unlock(ptl);
>>>>> 
>>>>> @@ -1658,13 +1660,19 @@ int collapse_pte_mapped_thp(struct mm_struct *mm, unsigned long addr,
>>>>>    /* step 4: remove empty page table */
>>>>>    if (!pml) {
>>>>>        pml = pmd_lock(mm, pmd);
>>>>> -        if (ptl != pml)
>>>>> +        if (ptl != pml) {
>>>>>            spin_lock_nested(ptl, SINGLE_DEPTH_NESTING);
>>>>> +            if (unlikely(!pmd_same(pgt_pmd, pmdp_get_lockless(pmd)))) {
>>>>> +                spin_unlock(pml);
>>>>> +                goto abort;
>>>> Drop the reference of folio and the mm counter twice at the label of abort and the step 3.
>>> 
>>> My bad, should set nr_ptes to 0 and call flush_tlb_mm() here, right?
>> Or add a new label "out" just below the "abort". Then go to out.
> 
> For this way, we also need to call flush_tlb_mm() first, like the
> following:
> 
> if (unlikely(!pmd_same(pgt_pmd, pmdp_get_lockless(pmd)))) {
> 	spin_unlock(pml);
> 	flush_tlb_mm(mm);
> 	goto out;
> }

Fine.

> 
>>> 
>>>>> +            }
>>>>> +        }
>>>>>    }
>>>>>    pgt_pmd = pmdp_collapse_flush(vma, haddr, pmd);
>>>>>    pmdp_get_lockless_sync();
>>>>>    if (ptl != pml)
>>>>>        spin_unlock(ptl);
>>>>> +    pte_unmap(start_pte);
>>>>>    spin_unlock(pml);
>>>> Why not?
>>>> pte_unmap_unlock(start_pte, ptl);
>>>> if (pml != ptl)
>>>>         spin_unlock(pml);
>>> 
>>> Both are fine, will do.
>>> 
>>> Thanks,
>>> Qi
>>> 
>>>>> 
>>>>>    mmu_notifier_invalidate_range_end(&range);
>>>>> --
>>>>> 2.20.1