linux-kernel - Re: [PATCH] mm: filemap: avoid unnecessary major faults in filemap

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <fedd15a2-1c22-598d-beb7-ab514886ef35@huawei.com>
Date:   Fri, 24 Nov 2023 11:31:53 +0800
From:   "zhangpeng (AS)" <zhangpeng362@...wei.com>
To:     <linux-mm@...ck.org>, <linux-kernel@...r.kernel.org>
CC:     <akpm@...ux-foundation.org>, <willy@...radead.org>,
        <fengwei.yin@...el.com>, <ying.huang@...el.com>,
        <aneesh.kumar@...ux.ibm.com>, <shy828301@...il.com>,
        <hughd@...gle.com>, <david@...hat.com>,
        <wangkefeng.wang@...wei.com>, <sunnanyong@...wei.com>
Subject: Re: [PATCH] mm: filemap: avoid unnecessary major faults in
 filemap_fault()

please ignore this...

On 2023/11/24 10:31, Peng Zhang wrote:

> From: ZhangPeng <zhangpeng362@...wei.com>
>
> The major fault occurred when using mlockall(MCL_CURRENT | MCL_FUTURE)
> in application, which leading to an unexpected performance issue[1].
>
> This caused by temporarily cleared pte during a read/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.
>
> Fix this by rechecking the pte by holding ptl in filemap_fault() before
> triggering a major fault.
>
> We tested the performance of file private mapping page fault
> (page_fault2.c of will-it-scale [2]) and file shared mapping page fault
> (page_fault3.c of will-it-scale). The difference in performance (in
> operations per second) before and after patch applied is about 0.7% on a
> x86 physical machine.
>
> [1] https://lore.kernel.org/linux-mm/9e62fd9a-bee0-52bf-50a7-498fa17434ee@huawei.com/
> [2] https://github.com/antonblanchard/will-it-scale/tree/master
>
> Suggested-by: "Huang, Ying" <ying.huang@...el.com>
> Signed-off-by: ZhangPeng <zhangpeng362@...wei.com>
> Signed-off-by: Kefeng Wang <wangkefeng.wang@...wei.com>
> Reviewed-by: Yin Fengwei <fengwei.yin@...el.com>
> ---
> RFC->v1: - Update commit message and add RB from Yin Fengwei
>           - Add error handling when ptep == NULL per Huang, Ying and
>             Matthew Wilcox
>
>   mm/filemap.c | 14 ++++++++++++++
>   1 file changed, 14 insertions(+)
>
> diff --git a/mm/filemap.c b/mm/filemap.c
> index 71f00539ac00..f3dcabdbc810 100644
> --- a/mm/filemap.c
> +++ b/mm/filemap.c
> @@ -3226,6 +3226,20 @@ vm_fault_t filemap_fault(struct vm_fault *vmf)
>   			mapping_locked = true;
>   		}
>   	} else {
> +		pte_t *ptep = pte_offset_map_lock(vmf->vma->vm_mm, vmf->pmd,
> +						  vmf->address, &vmf->ptl);
> +		if (unlikely(!ptep))
> +			return VM_FAULT_NOPAGE;
> +		/*
> +		 * Recheck pte with ptl locked as the pte can be cleared
> +		 * temporarily during a read/modify/write update.
> +		 */
> +		if (unlikely(!pte_none(ptep_get(ptep))))
> +			ret = VM_FAULT_NOPAGE;
> +		pte_unmap_unlock(ptep, vmf->ptl);
> +		if (unlikely(ret))
> +			return ret;
> +
>   		/* No page in the page cache at all */
>   		count_vm_event(PGMAJFAULT);
>   		count_memcg_event_mm(vmf->vma->vm_mm, PGMAJFAULT);

-- 
Best Regards,
Peng