Re: [PATCH v3 1/2] mm: migration: fix migration of huge PMD shared pages

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 08/23/2018 01:21 AM, Kirill A. Shutemov wrote:
> On Thu, Aug 23, 2018 at 09:30:35AM +0200, Michal Hocko wrote:
>> On Wed 22-08-18 09:48:16, Mike Kravetz wrote:
>>> On 08/22/2018 05:28 AM, Michal Hocko wrote:
>>>> On Tue 21-08-18 18:10:42, Mike Kravetz wrote:
>>>> [...]
>>>>> diff --git a/mm/rmap.c b/mm/rmap.c
>>>>> index eb477809a5c0..8cf853a4b093 100644
>>>>> --- a/mm/rmap.c
>>>>> +++ b/mm/rmap.c
>>>>> @@ -1362,11 +1362,21 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
>>>>>  	}
>>>>>  
>>>>>  	/*
>>>>> -	 * We have to assume the worse case ie pmd for invalidation. Note that
>>>>> -	 * the page can not be free in this function as call of try_to_unmap()
>>>>> -	 * must hold a reference on the page.
>>>>> +	 * For THP, we have to assume the worse case ie pmd for invalidation.
>>>>> +	 * For hugetlb, it could be much worse if we need to do pud
>>>>> +	 * invalidation in the case of pmd sharing.
>>>>> +	 *
>>>>> +	 * Note that the page can not be free in this function as call of
>>>>> +	 * try_to_unmap() must hold a reference on the page.
>>>>>  	 */
>>>>>  	end = min(vma->vm_end, start + (PAGE_SIZE << compound_order(page)));
>>>>> +	if (PageHuge(page)) {
>>>>> +		/*
>>>>> +		 * If sharing is possible, start and end will be adjusted
>>>>> +		 * accordingly.
>>>>> +		 */
>>>>> +		(void)huge_pmd_sharing_possible(vma, &start, &end);
>>>>> +	}
>>>>>  	mmu_notifier_invalidate_range_start(vma->vm_mm, start, end);
>>>>
>>>> I do not get this part. Why don't we simply unconditionally invalidate
>>>> the whole huge page range?
>>>
>>> In this routine, we are only unmapping a single page.  The existing code
>>> is limiting the invalidate range to that page size: 4K or 2M.  With shared
>>> PMDs, we have the possibility of unmapping a PUD_SIZE area: 1G.  I don't
>>> think we want to unconditionally invalidate 1G.  Is that what you are asking?
>>
>> But we know that huge_pmd_unshare unmapped a shared pte so we know when
>> to flush 2MB or 1GB. I really do not like how huge_pmd_sharing_possible
>> a) duplicates some checks and b) it updates start/stop out of line.
> 
> My reading on this is that mmu_notifier_invalidate_range_start() has to be
> called from sleepable context on the full range that *can* be invalidated
> before following mmu_notifier_invalidate_range_end().
> 
> In this case huge_pmd_unshare() may unmap aligned PUD_SIZE around the PMD
> page that effectively enlarge range that has to be covered by
> mmu_notifier_invalidate_range_start(). We cannot yet know if there's any
> shared page tables in the range, so we need to go with worst case
> scenario.

Yes, that is a good summary.  We can not know for sure if there is PMD
sharing until we hold the page table lock.  So, we don't know if we should
invalidate/flush 2M or 1G.  Yet, we need to call
mmu_notifier_invalidate_range_start before taking the lock.  And, the notifiers
need to know the range of the worst possible case.  The best approach I came up
with is to adjust the range if sharing is 'possible'.

> 
> I don't see conceptually better solution than what is proposed.
> 

I have updated the patches based on Kirill's previous comments and will send out
a new version later today.

-- 
Mike Kravetz




[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux