Re: [RFC PATCH v4 7/8] hugetlb: create hugetlb_unmap_file_folio to unmap single file folio

[Miaohe Lin <linmiaohe@huawei.com>]

On 2022/7/30 2:11, Mike Kravetz wrote:
> On 07/29/22 10:02, Miaohe Lin wrote:
>> On 2022/7/7 4:23, Mike Kravetz wrote:
>>> Create the new routine hugetlb_unmap_file_folio that will unmap a single
>>> file folio.  This is refactored code from hugetlb_vmdelete_list.  It is
>>> modified to do locking within the routine itself and check whether the
>>> page is mapped within a specific vma before unmapping.
>>>
>>> This refactoring will be put to use and expanded upon in a subsequent
>>> patch adding vma specific locking.
>>>
>>> Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
>>> ---
>>>  fs/hugetlbfs/inode.c | 124 +++++++++++++++++++++++++++++++++----------
>>>  1 file changed, 95 insertions(+), 29 deletions(-)
>>>
>>> diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c
>>> index 31bd4325fce5..0eac0ea2a245 100644
>>> --- a/fs/hugetlbfs/inode.c
>>> +++ b/fs/hugetlbfs/inode.c
>>> @@ -396,6 +396,94 @@ static int hugetlbfs_write_end(struct file *file, struct address_space *mapping,
>>>  	return -EINVAL;
>>>  }
>>>  
>>> +/*
>>> + * Called with i_mmap_rwsem held for inode based vma maps.  This makes
>>> + * sure vma (and vm_mm) will not go away.  We also hold the hugetlb fault
>>> + * mutex for the page in the mapping.  So, we can not race with page being
>>> + * faulted into the vma.
>>> + */
>>> +static bool hugetlb_vma_maps_page(struct vm_area_struct *vma,
>>> +				unsigned long addr, struct page *page)
>>> +{
>>> +	pte_t *ptep, pte;
>>> +
>>> +	ptep = huge_pte_offset(vma->vm_mm, addr,
>>> +			huge_page_size(hstate_vma(vma)));
>>> +
>>> +	if (!ptep)
>>> +		return false;
>>> +
>>> +	pte = huge_ptep_get(ptep);
>>> +	if (huge_pte_none(pte) || !pte_present(pte))
>>> +		return false;
>>> +
>>> +	if (pte_page(pte) == page)
>>> +		return true;
>>> +
>>> +	return false;	/* WTH??? */
>>
>> I'm sorry but what does WTH means? IIUC, this could happen if pte_page is a COW-ed private page?
>> vma_interval_tree_foreach doesn't exclude the private mapping even after cow?
> 
> My apologies, I left that comment in during development and should have removed
> it.  WTH is an acronym for 'What the Heck?".  I added it because I did not
> think we should ever get to this return statement.
> 

That's all right. Thanks for your hard work.

> I am not sure if your COW of a private page would get us to this return
> statement.  In any case, if we get there we need to return false.
> 
> Thank you for your analysis and comments!
>