Re: [Question]: major faults are still triggered after mlockall when numa balancing

["Yin, Fengwei" <fengwei.yin@intel.com>]

On 11/10/2023 11:39 AM, Kefeng Wang wrote:
> 
> 
> On 2023/11/10 9:57, Yin, Fengwei wrote:
>>
>>
>> On 11/10/2023 6:54 AM, Yang Shi wrote:
>>> On Thu, Nov 9, 2023 at 5:48 AM zhangpeng (AS) <zhangpeng362@huawei.com> wrote:
>>>>
>>>> Hi everyone,
>>>>
>>>> There is a performance issue that has been bothering us recently.
>>>> This problem can reproduce in the latest mainline version (Linux 6.6).
>>>>
>>>> We use mlockall(MCL_CURRENT | MCL_FUTURE) in the user mode process
>>>> to avoid performance problems caused by major fault.
>>>>
>>>> There is a stage in numa fault which will set pte as 0 in do_numa_page() :
>>>> ptep_modify_prot_start() will clear the vmf->pte, until
>>>> ptep_modify_prot_commit() assign a value to the vmf->pte.
>>>>
>>>> For the data segment of the user-mode program, the global variable area
>>>> is a private mapping. After the pagecache is loaded, the private
>>>> anonymous page is generated after the COW is triggered. Mlockall can
>>>> lock COW pages (anonymous pages), but the original file pages cannot
>>>> be locked and may be reclaimed. If the global variable (private anon page)
>>>> is accessed when vmf->pte is zero which is concurrently set by numa fault,
>>>> a file page fault will be triggered.
>>>>
>>>> At this time, the original private file page may have been reclaimed.
>>>> If the page cache is not available at this time, a major fault will be
>>>> triggered and the file will be read, causing additional overhead.
>>>>
>>>> Our problem scenario is as follows:
>>>>
>>>> task 1                      task 2
>>>> ------                      ------
>>>> /* scan global variables */
>>>> do_numa_page()
>>>>     spin_lock(vmf->ptl)
>>>>     ptep_modify_prot_start()
>>>>     /* set vmf->pte as null */
>>>>                               /* Access global variables */
>>>>                               handle_pte_fault()
>>>>                                 /* no pte lock */
>>>>                                 do_pte_missing()
>>>>                                   do_fault()
>>>>                                     do_read_fault()
>>>>     ptep_modify_prot_commit()
>>>>     /* ptep update done */
>>>>     pte_unmap_unlock(vmf->pte, vmf->ptl)
>>>>                                       do_fault_around()
>>>>                                       __do_fault()
>>>>                                         filemap_fault()
>>>>                                           /* page cache is not available
>>>>                                           and a major fault is triggered */
>>>>                                           do_sync_mmap_readahead()
>>>>                                           /* page_not_uptodate and goto
>>>>                                           out_retry. */
>>>>
>>>> Is there any way to avoid such a major fault?
>>>
>>> IMHO I don't think it is a bug. The man page quoted by Willy says "All
>>> mapped pages are guaranteed to be resident in RAM when the call
>>> returns successfully", but the later COW already made the file page
>>> unmapped, right? The PTE pointed to the COW'ed anon page.
>>> Hypothetically if we kept the file page mlocked and unmapped,
>>> munlock() would have not munlocked the file page at all, it would be
>>> mlocked in memory forever.
>> But in this case, even the COW page is mlocked. There is small window
>> that PTE is set to null in do_numa_page(). data segment access (it's to
>> COW page which has nothing to do with original page cache) happens in
>> this small window will trigger filemap_fault() to fault in original
>> page cache.
>>
>> I had thought to do double check whether vmf->pte is NULL in do_read_fault().
>> But it's not reliable enough.
>>
>> Matthew's idea to use protnone to block both hardware accessing and
>> do_pte_missing() looks more promising to me.
> 
> Actual， we could revert the following patch to avoid this issue,
> but this workaroud from ppc...
> 
> commit cee216a696b2004017a5ecb583366093d90b1568
> Author: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
> Date:   Fri Feb 24 14:59:13 2017 -0800
> 
>     mm/autonuma: don't use set_pte_at when updating protnone ptes
> 
>     Architectures like ppc64, use privilege access bit to mark pte non
>     accessible.  This implies that kernel can do a copy_to_user to an
>     address marked for numa fault.  This also implies that there can be a
>     parallel hardware update for the pte.  set_pte_at cannot be used in such
>     scenarios.  Hence switch the pte update to use ptep_get_and_clear and
>     set_pte_at combination.
Oh. This means the protnone doesn't work for PPC.

> 
>>
>>
>> Regards
>> Yin, Fengwei
>>
>>>
>>>>
>>>> -- 
>>>> Best Regards,
>>>> Peng
>>>>