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 88371C3A5A7 for ; Tue, 6 Dec 2022 09:38:13 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id EA2E88E0003; Tue, 6 Dec 2022 04:38:12 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id E2C5B8E0001; Tue, 6 Dec 2022 04:38:12 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id CCDB18E0003; Tue, 6 Dec 2022 04:38:12 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0014.hostedemail.com [216.40.44.14]) by kanga.kvack.org (Postfix) with ESMTP id B9A1E8E0001 for ; Tue, 6 Dec 2022 04:38:12 -0500 (EST) Received: from smtpin04.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay09.hostedemail.com (Postfix) with ESMTP id 4491C80C6D for ; Tue, 6 Dec 2022 09:38:12 +0000 (UTC) X-FDA: 80211380424.04.927625A Received: from szxga01-in.huawei.com (szxga01-in.huawei.com [45.249.212.187]) by imf10.hostedemail.com (Postfix) with ESMTP id 7F8AEC0006 for ; Tue, 6 Dec 2022 09:38:09 +0000 (UTC) Authentication-Results: imf10.hostedemail.com; dkim=none; dmarc=pass (policy=quarantine) header.from=huawei.com; spf=pass (imf10.hostedemail.com: domain of linmiaohe@huawei.com designates 45.249.212.187 as permitted sender) smtp.mailfrom=linmiaohe@huawei.com ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=hostedemail.com; s=arc-20220608; t=1670319491; 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=HLh5rne3oG9hjiYZQZDfmVuWR8OrWpuI65w3FGQxnVU=; b=KXUq6D5BICYUV5E2PtnRm2g8fOE1IteEbHFS4LdNyuTbsUMdKnDswRgNrhQmKrPlZXt8E8 V3gazSoFz4JpDE/b0WYiHtPZRMgK9kqPxr/vFjxRzg2A6hxkCWtqDRdo41iO02usQ30qMO ogQIvE4Ul0mYLiFgudP74AmYdwt/D0Q= ARC-Authentication-Results: i=1; imf10.hostedemail.com; dkim=none; dmarc=pass (policy=quarantine) header.from=huawei.com; spf=pass (imf10.hostedemail.com: domain of linmiaohe@huawei.com designates 45.249.212.187 as permitted sender) smtp.mailfrom=linmiaohe@huawei.com ARC-Seal: i=1; s=arc-20220608; d=hostedemail.com; t=1670319491; a=rsa-sha256; cv=none; b=zCZhIeuySanABLNv1zNkSs8UzlOWmA5Yz+Fomh14V0KjwTTMBk4szHCLxEhrFubrabNFE1 ycxqhDmDmmG6z3+zw0tVNtsQJODaNIeRrw+yk/2SLUWK1b6IN1O/sVWch01ZtFyaR0eDDf 8rEhgyl8W2LBJ2FgU49NY5T2+JBJ4YY= Received: from canpemm500002.china.huawei.com (unknown [172.30.72.56]) by szxga01-in.huawei.com (SkyGuard) with ESMTP id 4NRFgc2bW4zmWPR; Tue, 6 Dec 2022 17:37:16 +0800 (CST) Received: from [10.174.151.185] (10.174.151.185) by canpemm500002.china.huawei.com (7.192.104.244) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2375.31; Tue, 6 Dec 2022 17:37:31 +0800 Subject: Re: [PATCH v4 12/17] mm: remember exclusively mapped anonymous pages with PG_anon_exclusive To: David Hildenbrand , CC: Andrew Morton , Hugh Dickins , Linus Torvalds , David Rientjes , Shakeel Butt , John Hubbard , Jason Gunthorpe , Mike Kravetz , Mike Rapoport , Yang Shi , "Kirill A . Shutemov" , Matthew Wilcox , Vlastimil Babka , Jann Horn , Michal Hocko , Nadav Amit , Rik van Riel , Roman Gushchin , Andrea Arcangeli , Peter Xu , Donald Dutile , Christoph Hellwig , Oleg Nesterov , Jan Kara , Liang Zhang , Pedro Gomes , Oded Gabbay , References: <20220428083441.37290-1-david@redhat.com> <20220428083441.37290-13-david@redhat.com> <90dd6a93-4500-e0de-2bf0-bf522c311b0c@huawei.com> <3c7fd5da-b3f8-5562-45a9-f83d7dbcdd7d@redhat.com> From: Miaohe Lin Message-ID: Date: Tue, 6 Dec 2022 17:37:31 +0800 User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:78.0) Gecko/20100101 Thunderbird/78.6.0 MIME-Version: 1.0 In-Reply-To: <3c7fd5da-b3f8-5562-45a9-f83d7dbcdd7d@redhat.com> Content-Type: text/plain; charset="utf-8" Content-Language: en-US Content-Transfer-Encoding: 8bit X-Originating-IP: [10.174.151.185] X-ClientProxiedBy: dggems704-chm.china.huawei.com (10.3.19.181) To canpemm500002.china.huawei.com (7.192.104.244) X-CFilter-Loop: Reflected X-Rspam-User: X-Rspamd-Server: rspam02 X-Rspamd-Queue-Id: 7F8AEC0006 X-Stat-Signature: pp79wfdjpn78r7wtezy7xh49wikdck8a X-Spamd-Result: default: False [-3.31 / 9.00]; BAYES_HAM(-5.11)[98.01%]; SUSPICIOUS_RECIPS(1.50)[]; SUBJECT_HAS_UNDERSCORES(1.00)[]; DMARC_POLICY_ALLOW(-0.50)[huawei.com,quarantine]; R_SPF_ALLOW(-0.20)[+ip4:45.249.212.187/29]; MIME_GOOD(-0.10)[text/plain]; RCVD_NO_TLS_LAST(0.10)[]; RCPT_COUNT_TWELVE(0.00)[30]; FROM_EQ_ENVFROM(0.00)[]; R_DKIM_NA(0.00)[]; MIME_TRACE(0.00)[0:+]; TO_MATCH_ENVRCPT_SOME(0.00)[]; ARC_SIGNED(0.00)[hostedemail.com:s=arc-20220608:i=1]; MID_RHS_MATCH_FROM(0.00)[]; HAS_XOIP(0.00)[]; RCVD_COUNT_THREE(0.00)[3]; FROM_HAS_DN(0.00)[]; TAGGED_RCPT(0.00)[]; TO_DN_SOME(0.00)[]; ARC_NA(0.00)[] X-HE-Tag: 1670319489-780187 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: On 2022/12/6 16:43, David Hildenbrand wrote: >>> >> >> Hi David, sorry for the late respond and a possible inconsequential question. :) > > Better late than never! Thanks for the review, independently at which time it happens :) > >> >> >> >>> diff --git a/mm/hugetlb.c b/mm/hugetlb.c >>> index 7a71ed679853..5add8bbd47cd 100644 >>> --- a/mm/hugetlb.c >>> +++ b/mm/hugetlb.c >>> @@ -4772,7 +4772,7 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src, >>>                       is_hugetlb_entry_hwpoisoned(entry))) { >>>               swp_entry_t swp_entry = pte_to_swp_entry(entry); >>>   -            if (is_writable_migration_entry(swp_entry) && cow) { >>> +            if (!is_readable_migration_entry(swp_entry) && cow) { >>>                   /* >>>                    * COW mappings require pages in both >>>                    * parent and child to be set to read. >>> @@ -5172,6 +5172,8 @@ static vm_fault_t hugetlb_cow(struct mm_struct *mm, struct vm_area_struct *vma, >>>           set_huge_ptep_writable(vma, haddr, ptep); >>>           return 0; >>>       } >>> +    VM_BUG_ON_PAGE(PageAnon(old_page) && PageAnonExclusive(old_page), >>> +               old_page); >>>         /* >>>        * If the process that created a MAP_PRIVATE mapping is about to >>> @@ -6169,12 +6171,17 @@ unsigned long hugetlb_change_protection(struct vm_area_struct *vma, >>>           } >>>           if (unlikely(is_hugetlb_entry_migration(pte))) { >>>               swp_entry_t entry = pte_to_swp_entry(pte); >>> +            struct page *page = pfn_swap_entry_to_page(entry); >>>   -            if (is_writable_migration_entry(entry)) { >>> +            if (!is_readable_migration_entry(entry)) { >> >> In hugetlb_change_protection(), is_writable_migration_entry() is changed to !is_readable_migration_entry(), >> but >> >>>                   pte_t newpte; >>>   -                entry = make_readable_migration_entry( >>> -                            swp_offset(entry)); >>> +                if (PageAnon(page)) >>> +                    entry = make_readable_exclusive_migration_entry( >>> +                                swp_offset(entry)); >>> +                else >>> +                    entry = make_readable_migration_entry( >>> +                                swp_offset(entry)); >>>                   newpte = swp_entry_to_pte(entry); >>>                   set_huge_swap_pte_at(mm, address, ptep, >>>                                newpte, huge_page_size(h)); >> >> >> >>> diff --git a/mm/mprotect.c b/mm/mprotect.c >>> index b69ce7a7b2b7..56060acdabd3 100644 >>> --- a/mm/mprotect.c >>> +++ b/mm/mprotect.c >>> @@ -152,6 +152,7 @@ static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd, >>>               pages++; >>>           } else if (is_swap_pte(oldpte)) { >>>               swp_entry_t entry = pte_to_swp_entry(oldpte); >>> +            struct page *page = pfn_swap_entry_to_page(entry); >>>               pte_t newpte; >>>                 if (is_writable_migration_entry(entry)) { >> >> In change_pte_range(), is_writable_migration_entry() is not changed to !is_readable_migration_entry(). > > Yes, and also in change_huge_pmd(), is_writable_migration_entry() stays unchanged. > >> Is this done intentionally? Could you tell me why there's such a difference? I'm confused. It's very >> kind of you if you can answer my puzzle. > > For change protection, the only relevant part is to convert writable -> readable or writable -> readable_exclusive. > > If an entry is already readable or readable_exclusive, there is nothing to do. The only issues would be when turning a readable one into a readable_exclusive one or a readable_exclusive one into a readable one. > > > In hugetlb_change_protection(), the "!is_readable_migration_entry" could in fact be turned into a "is_writable_migration_entry()". Right now, it would convert writable -> readable or writable -> readable_exclusive AND readable -> readable AND readable_exclusive -> readable_exclusive, which isn't necessary but also shouldn't hurt either. Many thanks for your explanation. It's really helpful. :) > > > So yeah, it's not consistent but shouldn't be problematic. Do you see an issue with that? No, I don't see any issue with that. I just wonder whether we can change "!is_readable_migration_entry" to "is_writable_migration_entry()" to make code more consistent and avoid possible future puzzle. Also we can further remove this harmless unnecessary migration entry conversion. But this should be a separate cleanup patch anyway. > > It would be great to extend the "selftest/vm cow" test to also cover migration entries, however, that requires slightly more work and "luck" to fork() while migration is happening. > > Thanks! Many thanks for your work! Thanks, Miaohe Lin >