Received-SPF: pass (google.com: domain of linux-kernel+bounces-94849-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Message-ID: <f5df24df-00b4-154c-9805-4b6232f17825@huawei.com>
Date: Thu, 7 Mar 2024 10:05:56 +0800
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:102.0) Gecko/20100101
 Thunderbird/102.9.0
Subject: Re: [PATCH v4] filemap: avoid unnecessary major faults in
 filemap_fault()
Content-Language: en-US
To: David Hildenbrand <david@redhat.com>, <linux-mm@kvack.org>,
	<linux-kernel@vger.kernel.org>
CC: <akpm@linux-foundation.org>, <willy@infradead.org>,
	<ying.huang@intel.com>, <fengwei.yin@intel.com>,
	<aneesh.kumar@linux.ibm.com>, <shy828301@gmail.com>, <hughd@google.com>,
	<wangkefeng.wang@huawei.com>, <sunnanyong@huawei.com>
References: <20240306083809.1236634-1-zhangpeng362@huawei.com>
 <5691dff3-a944-42f2-9848-ea7287a89192@redhat.com>
From: "zhangpeng (AS)" <zhangpeng362@huawei.com>
In-Reply-To: <5691dff3-a944-42f2-9848-ea7287a89192@redhat.com>
Content-Type: text/plain; charset="UTF-8"; format=flowed
Content-Transfer-Encoding: 8bit

On 2024/3/6 18:24, David Hildenbrand wrote:

> On 06.03.24 09:38, Peng Zhang wrote:
>> From: ZhangPeng <zhangpeng362@huawei.com>
>>
>> The major fault occurred when using mlockall(MCL_CURRENT | MCL_FUTURE)
>> in application, which leading to an unexpected issue[1].
>>
>> This caused by temporarily cleared PTE during a read+clear/modify/write
>> update of the PTE, eg, do_numa_page()/change_pte_range().
>>
>> 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 zeroed in 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.
>>
>> This issue affects our traffic analysis service. The inbound traffic is
>> heavy. If a major fault occurs, the I/O schedule is triggered and the
>> original I/O is suspended. Generally, the I/O schedule is 0.7 ms. If
>> other applications are operating disks, the system needs to wait for
>> more than 10 ms. However, the inbound traffic is heavy and the NIC 
>> buffer
>> is small. As a result, packet loss occurs. But the traffic analysis 
>> service
>> can't tolerate packet loss.
>>
>> Fix this by holding PTL and rechecking the PTE in filemap_fault() before
>> triggering a major fault. We do this check only if vma is VM_LOCKED to
>> reduce the performance impact in common scenarios.
>>
>> In our product environment, there were 7 major faults every 12 hours.
>> After the patch is applied, no major fault have been triggered.
>>
>> Testing file page read and write page fault performance in ext4 and
>> ramdisk using will-it-scale[2] on a x86 physical machine. The data is
>> the average change compared with the mainline after the patch is 
>> applied.
>> The test results are within the range of fluctuation. We do this check
>> only if vma is VM_LOCKED, therefore, no performance regressions is 
>> caused
>> for most common cases.
>>
>> The test results are as follows:
>>                            processes processes_idle  threads 
>> threads_idle
>> ext4    private file write:  0.22%    0.26%           1.21% -0.15%
>> ext4    private file  read:  0.03%    1.00%           1.39% 0.34%
>> ext4    shared  file write: -0.50%   -0.02%          -0.14% -0.02%
>> ramdisk private file write:  0.07%    0.02%           0.53% 0.04%
>> ramdisk private file  read:  0.01%    1.60%          -0.32% -0.02%
>>
>> [1] 
>> https://lore.kernel.org/linux-mm/9e62fd9a-bee0-52bf-50a7-498fa17434ee@huawei.com/
>> [2] https://github.com/antonblanchard/will-it-scale/
>>
>> Suggested-by: "Huang, Ying" <ying.huang@intel.com>
>> Suggested-by: David Hildenbrand <david@redhat.com>
>> Signed-off-by: ZhangPeng <zhangpeng362@huawei.com>
>> Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
>> ---
>> v3->v4:
>> - Update the performance data and commit message
>> - Check PTE without lock firstly per Huang, Ying
>> - Update comments for recheck function per David Hildenbrand
>> - Simply return 0 to make it easier to read per David Hildenbrand
>> - Check !FAULT_FLAG_ORIG_PTE_VALID instead of pmd_none()
>>
>> v2->v3:
>> - Do this check only if vma is VM_LOCKED per David Hildenbrand
>> - Hold PTL and recheck the PTE
>> - Place the recheck code in a new function filemap_fault_recheck_pte()
>>
>> v1->v2:
>> - Add more test results per Huang, Ying
>> - Add more comments before check PTE per Huang, Ying, David Hildenbrand
>>    and Yin Fengwei
>> - Change pte_offset_map_nolock to pte_offset_map as the PTL won't
>>    be used
>>
>> RFC->v1:
>> - Add error handling when ptep == NULL per Huang, Ying and Matthew
>>    Wilcox
>> - Check the PTE without acquiring PTL in filemap_fault(), suggested by
>>    Huang, Ying and Yin Fengwei
>> - Add pmd_none() check before PTE map
>> - Update commit message and add performance test information
>>
>>   mm/filemap.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++
>>   1 file changed, 46 insertions(+)
>>
>> diff --git a/mm/filemap.c b/mm/filemap.c
>> index b4858d89f1b1..31ab455c4537 100644
>> --- a/mm/filemap.c
>> +++ b/mm/filemap.c
>> @@ -3181,6 +3181,48 @@ static struct file 
>> *do_async_mmap_readahead(struct vm_fault *vmf,
>>       return fpin;
>>   }
>>   +static vm_fault_t filemap_fault_recheck_pte_none(struct vm_fault 
>> *vmf)
>> +{
>> +    struct vm_area_struct *vma = vmf->vma;
>> +    vm_fault_t ret = 0;
>> +    pte_t *ptep;
>> +
>> +    /*
>> +     * We might have COW'ed a pagecache folio and might now have an 
>> mlocked
>> +     * anon folio mapped. The original pagecache folio is not 
>> mlocked and
>> +     * might have been evicted. During a read+clear/modify/write 
>> update of
>> +     * the PTE, such as done in do_numa_page()/change_pte_range(), we
>> +     * temporarily clear the PTE under PT lock and might detect it 
>> here as
>> +     * "none" when not holding the PT lock.
>> +     *
>> +     * Not rechecking the PTE under PT lock could result in an 
>> unexpected
>> +     * major fault in an mlock'ed region. Recheck only for this special
>> +     * scenario while holding the PT lock, to not degrade non-mlocked
>> +     * scenarios. Recheck the PTE without PT lock firstly, thereby 
>> reducing
>> +     * the number of times we hold PT lock.
>> +     */
>> +    if (!(vma->vm_flags & VM_LOCKED))
>> +        return 0;
>> +
>> +    if (!(vmf->flags & FAULT_FLAG_ORIG_PTE_VALID))
>> +        return 0;
>> +
>> +    ptep = pte_offset_map(vmf->pmd, vmf->address);
>> +    if (unlikely(!ptep))
>> +        return VM_FAULT_NOPAGE;
>> +
>> +    if (unlikely(!pte_none(ptep_get_lockless(ptep)))) {
>> +        ret = VM_FAULT_NOPAGE;
>> +    } else {
>> +        spin_lock(vmf->ptl);
>
> I assume the assumption is that vmf->ptl is still set from 
> handle_pte_fault(). But in the meantime, we did a pte_unmap(vmf->pte).
>
> Looking at other users of vmf->ptl, this here sticks out.
>
> I would suggest replacing the pte_offset_map() above by a 
> pte_offset_map_nolock() that gives us the ptl.
>
>
I'll replace it in the next version.

Thanks!

> With that
>
> Reviewed-by: David Hildenbrand <david@redhat.com>
>
-- 
Best Regards,
Peng