From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by smtp.lore.kernel.org (Postfix) with ESMTP id 9CC4DC2BBCA for ; Tue, 25 Jun 2024 11:40:38 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id 2C4DC6B02E8; Tue, 25 Jun 2024 07:40:38 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id 226D56B02E9; Tue, 25 Jun 2024 07:40:38 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 079D36B02EA; Tue, 25 Jun 2024 07:40:37 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0010.hostedemail.com [216.40.44.10]) by kanga.kvack.org (Postfix) with ESMTP id CD77E6B02E8 for ; Tue, 25 Jun 2024 07:40:37 -0400 (EDT) Received: from smtpin04.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay09.hostedemail.com (Postfix) with ESMTP id 510498184E for ; Tue, 25 Jun 2024 11:40:37 +0000 (UTC) X-FDA: 82269218514.04.0ECDC86 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by imf19.hostedemail.com (Postfix) with ESMTP id 43A9B1A000B for ; Tue, 25 Jun 2024 11:40:35 +0000 (UTC) Authentication-Results: imf19.hostedemail.com; dkim=none; spf=pass (imf19.hostedemail.com: domain of ryan.roberts@arm.com designates 217.140.110.172 as permitted sender) smtp.mailfrom=ryan.roberts@arm.com; dmarc=pass (policy=none) header.from=arm.com ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=hostedemail.com; s=arc-20220608; t=1719315627; h=from:from:sender:reply-to:subject:subject:date:date: message-id:message-id:to:to:cc:cc:mime-version:mime-version: content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=U/b5d/KzdyPujWEdCq+y/6XuBFApuf0+Mg5iq7ZA5Yw=; b=N1j169gooco9DrXTHYTpBkSHjM06CInOBumCRylNWCfN6oFctqncZJ8jD1aSmtdzZmXLPd r0y0ZantBrXoa1qUJw+JtZj6TQ5UWdXWpz6Yeoz/0tjEKgyvrRtukUOiZgMYy7Tcmk0XNg ZqtBQzzjawiEfHQT2f7ri6HP09AkF2c= ARC-Authentication-Results: i=1; imf19.hostedemail.com; dkim=none; spf=pass (imf19.hostedemail.com: domain of ryan.roberts@arm.com designates 217.140.110.172 as permitted sender) smtp.mailfrom=ryan.roberts@arm.com; dmarc=pass (policy=none) header.from=arm.com ARC-Seal: i=1; s=arc-20220608; d=hostedemail.com; t=1719315627; a=rsa-sha256; cv=none; b=NkvQ5ItoUlJLRp/kYj90+O40rznMKcNX0o5UHLOw9FAkqkFxBMDSuOdylG+hMnesvh6cZQ vxSaYnpWPlQI/OOKgqxBYS1e5Y1ibwB3M8G0pdHgtMwjR/2onG0bfX2GZ5285WIp5rvHTF zu0WkduQJHtx0WWms1PFwgjW6BBX6Hk= Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 0FB97339; Tue, 25 Jun 2024 04:40:54 -0700 (PDT) Received: from [10.1.39.170] (XHFQ2J9959.cambridge.arm.com [10.1.39.170]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPSA id CB9883F73B; Tue, 25 Jun 2024 04:40:25 -0700 (PDT) Message-ID: Date: Tue, 25 Jun 2024 12:40:24 +0100 MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH v6 18/18] arm64/mm: Automatically fold contpte mappings Content-Language: en-GB To: Baolin Wang , Kefeng Wang , Catalin Marinas , Will Deacon , Ard Biesheuvel , Marc Zyngier , James Morse , Andrey Ryabinin , Andrew Morton , Matthew Wilcox , Mark Rutland , David Hildenbrand , John Hubbard , Zi Yan , Barry Song <21cnbao@gmail.com>, Alistair Popple , Yang Shi , Thomas Gleixner , Ingo Molnar , Borislav Petkov , Dave Hansen , "H. Peter Anvin" , "Yin, Fengwei" Cc: linux-arm-kernel@lists.infradead.org, x86@kernel.org, linuxppc-dev@lists.ozlabs.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org References: <20240215103205.2607016-1-ryan.roberts@arm.com> <20240215103205.2607016-19-ryan.roberts@arm.com> <1285eb59-fcc3-4db8-9dd9-e7c4d82b1be0@huawei.com> <8d57ed0d-fdd0-4fc6-b9f1-a6ac11ce93ce@arm.com> <018b5e83-789e-480f-82c8-a64515cdd14a@huawei.com> From: Ryan Roberts In-Reply-To: Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-Rspamd-Server: rspam11 X-Rspamd-Queue-Id: 43A9B1A000B X-Stat-Signature: mftabr4ozdemgrk6gzpfhj3h3qqo5p1c X-Rspam-User: X-HE-Tag: 1719315635-630750 X-HE-Meta: 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 lx+COGcc iPFNL1xEAqZoJINBXggAIDF8jHoVnvJiD/aLZdf6kxRRazTW+aWFTo2oV2p3iD/Xz12N1X1NsTmvIyip8hZBEERPVbKAqDeIhvgfM/CtsAvpASoqxM+i+VGysE1SvgnEH2GolGwWHaB+4sy50UNb+bWv3f+yXyNMsOL4jkcA2dyCAzlgb1jGpbk+bFBGQAzpjzg2Yjr+9bOKq1T5vAbPn/27CTl/Yi17G9nsoB5srx/RQ4gEcri37hxqnLYQDANkVUXwc6IFOsMawZfJsIIsOjuQgD+mdyzCNsjgKvU7vLjJ0PBQt83fNypibBz5nouZ4hWGzuSSUjHq7FEomPLizr3lBic7xnAlwcUO4VYESFnN3T0AWL6npjucYgm0GqQ97D4AYxX/fs8BS1iV6lLZk7MxrMVcSyg4U9F8ffVT6ia/m1kPlcltBA/DKeA== X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: List-Subscribe: List-Unsubscribe: On 25/06/2024 08:23, Baolin Wang wrote: > > > On 2024/6/25 11:16, Kefeng Wang wrote: >> >> >> On 2024/6/24 23:56, Ryan Roberts wrote: >>> + Baolin Wang and Yin Fengwei, who maybe able to help with this. >>> >>> >>> Hi Kefeng, >>> >>> Thanks for the report! >>> >>> >>> On 24/06/2024 15:30, Kefeng Wang wrote: >>>> Hi Ryan, >>>> >>>> A big regression on page-fault3("Separate file shared mapping page >>>> fault") testcase from will-it-scale on arm64, no issue on x86, >>>> >>>> ./page_fault3_processes -t 128 -s 5 >>> >>> I see that this program is mkstmp'ing a file at "/tmp/willitscale.XXXXXX". Based >>> on your description, I'm inferring that /tmp is backed by ext4 with your large >>> folio patches enabled? >> >> Yes, mount /tmp by ext4, sorry to forget to mention that. >> >>> >>>> >>>> 1) large folio disabled on ext4: >>>>     92378735 >>>> 2) large folio  enabled on ext4 +  CONTPTE enabled >>>>     16164943 >>>> 3) large folio  enabled on ext4 +  CONTPTE disabled >>>>     80364074 >>>> 4) large folio  enabled on ext4 +  CONTPTE enabled + large folio mapping >>>> enabled >>>> in finish_fault()[2] >>>>     299656874 >>>> >>>> We found *contpte_convert* consume lots of CPU(76%) in case 2), >>> >>> contpte_convert() is expensive and to be avoided; In this case I expect it is >>> repainting the PTEs with the PTE_CONT bit added in, and to do that it needs to >>> invalidate the tlb for the virtual range. The code is there to mop up user space >>> patterns where each page in a range is temporarily made RO, then later changed >>> back. In this case, we want to re-fold the contpte range once all pages have >>> been serviced in RO mode. >>> >>> Of course this path is only intended as a fallback, and the more optimium >>> approach is to set_ptes() the whole folio in one go where possible - kind of >>> what you are doing below. >>> >>>> and disappeared >>>> by following change[2], it is easy to understood the different between case 2) >>>> and case 4) since case 2) always map one page >>>> size, but always try to fold contpte mappings, which spend a lot of >>>> time. Case 4) is a workaround, any other better suggestion? >>> >>> See below. >>> >>>> >>>> Thanks. >>>> >>>> [1] https://github.com/antonblanchard/will-it-scale >>>> [2] enable large folio mapping in finish_fault() >>>> >>>> diff --git a/mm/memory.c b/mm/memory.c >>>> index 00728ea95583..5623a8ce3a1e 100644 >>>> --- a/mm/memory.c >>>> +++ b/mm/memory.c >>>> @@ -4880,7 +4880,7 @@ vm_fault_t finish_fault(struct vm_fault *vmf) >>>>           * approach also applies to non-anonymous-shmem faults to avoid >>>>           * inflating the RSS of the process. >>>>           */ >>>> -       if (!vma_is_anon_shmem(vma) || unlikely(userfaultfd_armed(vma))) { >>>> +       if (unlikely(userfaultfd_armed(vma))) { >>> >>> The change to make finish_fault() handle multiple pages in one go are new; added >>> by Baolin Wang at [1]. That extra conditional that you have removed is there to >>> prevent RSS reporting bloat. See discussion that starts at [2]. >>> >>> Anyway, it was my vague understanding that the fault around mechanism >>> (do_fault_around()) would ensure that (by default) 64K worth of pages get mapped >>> together in a single set_ptes() call, via filemap_map_pages() -> >>> filemap_map_folio_range(). Looking at the code, I guess fault around only >>> applies to read faults. This test is doing a write fault. >>> >>> I guess we need to do a change a bit like what you have done, but also taking >>> into account fault_around configuration? > > For the writable mmap() of tmpfs, we will use mTHP interface to control the size > of folio to allocate, as discussed in previous meeting [1], so I don't think > fault_around configuration will be helpful for tmpfs. Yes agreed. But we are talking about ext4 here. > > For other filesystems, like ext4, I did not found the logic to determin what > size of folio to allocate in writable mmap() path Yes I'd be keen to understand this to. When I was doing contpte, page cache would only allocate large folios for readahead. So that's why I wouldn't have seen this. > (Kefeng, please correct me if > I missed something). If there is a control like mTHP, we can rely on that > instead of 'fault_around'? Page cache doesn't currently expose any controls for folio allocation size. Personally, I'd like to see some in future becaudse I suspect it will be neccessary to limit physical fragmentation. But that is another conversation... > > [1] https://lore.kernel.org/all/f1783ff0-65bd-4b2b-8952-52b6822a0835@redhat.com/ > >> Yes, the current changes is not enough, I hint some issue and still debugging, >> so our direction is trying to map large folio for do_shared_fault(), right? We just need to make sure that if finish_fault() has a (non-shmem) large folio, it never maps more than fault_around_pages, and it does it in a way that is naturally aligned in virtual space (like do_fault_around() does). do_fault_around() actually tries to get other folios from the page cache to map. We don't want to do that; we just want to make sure that we don't inflate a process's RSS by mapping unbounded large folios. Another (orthogonal, longer term) strategy would be to optimize contpte_convert(). arm64 has a feature called "BBM level 2"; we could potentially elide the TLBIs for systems that support this. But ultimately its best to avoid the need for folding in the first place. Thanks, Ryan > > I think this is the right direction to do. I add this '!vma_is_anon_shmem(vma)' > conditon to gradually implement support for large folio mapping buidling, > especially for writable mmap() support in tmpfs. > >>> [1] >>> https://lore.kernel.org/all/3a190892355989d42f59cf9f2f98b94694b0d24d.1718090413.git.baolin.wang@linux.alibaba.com/ >>> [2] >>> https://lore.kernel.org/linux-mm/13939ade-a99a-4075-8a26-9be7576b7e03@arm.com/ >>> >>> >>>>                  nr_pages = 1; >>>>          } else if (nr_pages > 1) { >>>>                  pgoff_t idx = folio_page_idx(folio, page); >>>> >>>> >>>> On 2024/2/15 18:32, Ryan Roberts wrote: >>>>> There are situations where a change to a single PTE could cause the >>>>> contpte block in which it resides to become foldable (i.e. could be >>>>> repainted with the contiguous bit). Such situations arise, for example, >>>>> when user space temporarily changes protections, via mprotect, for >>>>> individual pages, such can be the case for certain garbage collectors. >>>>> >>>>> We would like to detect when such a PTE change occurs. However this can >>>>> be expensive due to the amount of checking required. Therefore only >>>>> perform the checks when an indiviual PTE is modified via mprotect >>>>> (ptep_modify_prot_commit() -> set_pte_at() -> set_ptes(nr=1)) and only >>>>> when we are setting the final PTE in a contpte-aligned block. >>>>> >>>>> Signed-off-by: Ryan Roberts >>>>> --- >>>>>    arch/arm64/include/asm/pgtable.h | 26 +++++++++++++ >>>>>    arch/arm64/mm/contpte.c          | 64 ++++++++++++++++++++++++++++++++ >>>>>    2 files changed, 90 insertions(+) >>>>> >>>>> diff --git a/arch/arm64/include/asm/pgtable.h >>>>> b/arch/arm64/include/asm/pgtable.h >>>>> index 8310875133ff..401087e8a43d 100644 >>>>> --- a/arch/arm64/include/asm/pgtable.h >>>>> +++ b/arch/arm64/include/asm/pgtable.h >>>>> @@ -1185,6 +1185,8 @@ extern void ptep_modify_prot_commit(struct >>>>> vm_area_struct *vma, >>>>>     * where it is possible and makes sense to do so. The PTE_CONT bit is >>>>> considered >>>>>     * a private implementation detail of the public ptep API (see below). >>>>>     */ >>>>> +extern void __contpte_try_fold(struct mm_struct *mm, unsigned long addr, >>>>> +                pte_t *ptep, pte_t pte); >>>>>    extern void __contpte_try_unfold(struct mm_struct *mm, unsigned long addr, >>>>>                    pte_t *ptep, pte_t pte); >>>>>    extern pte_t contpte_ptep_get(pte_t *ptep, pte_t orig_pte); >>>>> @@ -1206,6 +1208,29 @@ extern int contpte_ptep_set_access_flags(struct >>>>> vm_area_struct *vma, >>>>>                    unsigned long addr, pte_t *ptep, >>>>>                    pte_t entry, int dirty); >>>>>    +static __always_inline void contpte_try_fold(struct mm_struct *mm, >>>>> +                unsigned long addr, pte_t *ptep, pte_t pte) >>>>> +{ >>>>> +    /* >>>>> +     * Only bother trying if both the virtual and physical addresses are >>>>> +     * aligned and correspond to the last entry in a contig range. The core >>>>> +     * code mostly modifies ranges from low to high, so this is the likely >>>>> +     * the last modification in the contig range, so a good time to fold. >>>>> +     * We can't fold special mappings, because there is no associated folio. >>>>> +     */ >>>>> + >>>>> +    const unsigned long contmask = CONT_PTES - 1; >>>>> +    bool valign = ((addr >> PAGE_SHIFT) & contmask) == contmask; >>>>> + >>>>> +    if (unlikely(valign)) { >>>>> +        bool palign = (pte_pfn(pte) & contmask) == contmask; >>>>> + >>>>> +        if (unlikely(palign && >>>>> +            pte_valid(pte) && !pte_cont(pte) && !pte_special(pte))) >>>>> +            __contpte_try_fold(mm, addr, ptep, pte); >>>>> +    } >>>>> +} >>>>> + >>>>>    static __always_inline void contpte_try_unfold(struct mm_struct *mm, >>>>>                    unsigned long addr, pte_t *ptep, pte_t pte) >>>>>    { >>>>> @@ -1286,6 +1311,7 @@ static __always_inline void set_ptes(struct mm_struct >>>>> *mm, unsigned long addr, >>>>>        if (likely(nr == 1)) { >>>>>            contpte_try_unfold(mm, addr, ptep, __ptep_get(ptep)); >>>>>            __set_ptes(mm, addr, ptep, pte, 1); >>>>> +        contpte_try_fold(mm, addr, ptep, pte); >>>>>        } else { >>>>>            contpte_set_ptes(mm, addr, ptep, pte, nr); >>>>>        } >>>>> diff --git a/arch/arm64/mm/contpte.c b/arch/arm64/mm/contpte.c >>>>> index 50e0173dc5ee..16788f07716d 100644 >>>>> --- a/arch/arm64/mm/contpte.c >>>>> +++ b/arch/arm64/mm/contpte.c >>>>> @@ -73,6 +73,70 @@ static void contpte_convert(struct mm_struct *mm, unsigned >>>>> long addr, >>>>>        __set_ptes(mm, start_addr, start_ptep, pte, CONT_PTES); >>>>>    } >>>>>    +void __contpte_try_fold(struct mm_struct *mm, unsigned long addr, >>>>> +            pte_t *ptep, pte_t pte) >>>>> +{ >>>>> +    /* >>>>> +     * We have already checked that the virtual and pysical addresses are >>>>> +     * correctly aligned for a contpte mapping in contpte_try_fold() so the >>>>> +     * remaining checks are to ensure that the contpte range is fully >>>>> +     * covered by a single folio, and ensure that all the ptes are valid >>>>> +     * with contiguous PFNs and matching prots. We ignore the state of the >>>>> +     * access and dirty bits for the purpose of deciding if its a contiguous >>>>> +     * range; the folding process will generate a single contpte entry which >>>>> +     * has a single access and dirty bit. Those 2 bits are the logical OR of >>>>> +     * their respective bits in the constituent pte entries. In order to >>>>> +     * ensure the contpte range is covered by a single folio, we must >>>>> +     * recover the folio from the pfn, but special mappings don't have a >>>>> +     * folio backing them. Fortunately contpte_try_fold() already checked >>>>> +     * that the pte is not special - we never try to fold special mappings. >>>>> +     * Note we can't use vm_normal_page() for this since we don't have the >>>>> +     * vma. >>>>> +     */ >>>>> + >>>>> +    unsigned long folio_start, folio_end; >>>>> +    unsigned long cont_start, cont_end; >>>>> +    pte_t expected_pte, subpte; >>>>> +    struct folio *folio; >>>>> +    struct page *page; >>>>> +    unsigned long pfn; >>>>> +    pte_t *orig_ptep; >>>>> +    pgprot_t prot; >>>>> + >>>>> +    int i; >>>>> + >>>>> +    if (!mm_is_user(mm)) >>>>> +        return; >>>>> + >>>>> +    page = pte_page(pte); >>>>> +    folio = page_folio(page); >>>>> +    folio_start = addr - (page - &folio->page) * PAGE_SIZE; >>>>> +    folio_end = folio_start + folio_nr_pages(folio) * PAGE_SIZE; >>>>> +    cont_start = ALIGN_DOWN(addr, CONT_PTE_SIZE); >>>>> +    cont_end = cont_start + CONT_PTE_SIZE; >>>>> + >>>>> +    if (folio_start > cont_start || folio_end < cont_end) >>>>> +        return; >>>>> + >>>>> +    pfn = ALIGN_DOWN(pte_pfn(pte), CONT_PTES); >>>>> +    prot = pte_pgprot(pte_mkold(pte_mkclean(pte))); >>>>> +    expected_pte = pfn_pte(pfn, prot); >>>>> +    orig_ptep = ptep; >>>>> +    ptep = contpte_align_down(ptep); >>>>> + >>>>> +    for (i = 0; i < CONT_PTES; i++) { >>>>> +        subpte = pte_mkold(pte_mkclean(__ptep_get(ptep))); >>>>> +        if (!pte_same(subpte, expected_pte)) >>>>> +            return; >>>>> +        expected_pte = pte_advance_pfn(expected_pte, 1); >>>>> +        ptep++; >>>>> +    } >>>>> + >>>>> +    pte = pte_mkcont(pte); >>>>> +    contpte_convert(mm, addr, orig_ptep, pte); >>>>> +} >>>>> +EXPORT_SYMBOL(__contpte_try_fold); >>>>> + >>>>>    void __contpte_try_unfold(struct mm_struct *mm, unsigned long addr, >>>>>                pte_t *ptep, pte_t pte) >>>>>    { >>> >>>