Received: by 2002:a05:6602:18e:0:0:0:0 with SMTP id m14csp4305464ioo; Tue, 31 May 2022 01:08:52 -0700 (PDT) X-Google-Smtp-Source: ABdhPJyGht1dDPMv/fBqa3LL9kZsy9REugfauBDx59owLbSemUFaPMS86UjumIxveAM9w7nvRYbP X-Received: by 2002:a17:907:3e07:b0:6fe:fdf9:2c2e with SMTP id hp7-20020a1709073e0700b006fefdf92c2emr33221834ejc.378.1653984532348; Tue, 31 May 2022 01:08:52 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1653984532; cv=none; d=google.com; s=arc-20160816; b=p6Vyixl5yXx6U5DeiCOsBoC92TBYMqx9Bv8wPue4RQAXJ6Xvv17ECSiZiAL6UpNsPU ktuBJV9U2foZAwUICuXAALMw9nGmQV2HexobhwqUs/08plGnibdwJpRqcjd3BWFX8/JU UyoqVMd4d/6GRPNyRnhqiK2cJH9yWuM5ECtSjJ+dXK6vk/emUOI0/qRWavzT8/tjwWmL JtATqxg453QsYgPaTxVm0NF+aXiF+UaivsYz7Eokgh2nkCPBSPSeALyR98tCqO98cflm RlpjrXZneoxXSx7mj7+zDV7Xa9xcDqLQbCsFYqKkljLDRX4e0EY+Ln5e9+4cURmq8HP6 OhHQ== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:content-transfer-encoding:in-reply-to:from :references:cc:to:content-language:subject:user-agent:mime-version :date:message-id:dkim-signature; bh=v5gI6W4viTWql8Uwhb50aMFlObiKdy2vCUTCbMVtQ14=; b=nG5LhZ6X3HYJt8/eVB1YLKhOIMDmkWnr7tYlKZPtUxT3SoNOkUwYXzwnaA+Wc81ggG WHLGpML2azDtvyAS3W+P7zhK/AXIm+ALEJSUPBuWfgzBC1nR2TJLRaUzcDHAytj7LxcM /lph+fxjyp/U6C7WNEefrSG4zZkTrmZht4u1mArYkz56+4Ht8VP36gRAY3aQf3eqdKF/ UOHZXCGYfY1xirpsM4tUYZGXeHvdiGo3t9PkNKdLWu/dDV7z6qs0z3Poviu261a5lm+r UCuEI0dXUvr7S44pAnWON+hD2N+qav6+LNG4Jv6Wn9mnUuC3jF8QFQ4pghkINCPBhO3X j6EA== ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@kernel.org header.s=k20201202 header.b=hnncicVx; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=kernel.org Return-Path: Received: from out1.vger.email (out1.vger.email. [2620:137:e000::1:20]) by mx.google.com with ESMTP id jt8-20020a170906ca0800b006ff47216539si8523523ejb.537.2022.05.31.01.08.26; Tue, 31 May 2022 01:08:52 -0700 (PDT) Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20; Authentication-Results: mx.google.com; dkim=pass header.i=@kernel.org header.s=k20201202 header.b=hnncicVx; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S244698AbiEaIBK (ORCPT + 99 others); Tue, 31 May 2022 04:01:10 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:50626 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S235885AbiEaIBJ (ORCPT ); Tue, 31 May 2022 04:01:09 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [145.40.68.75]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4529C7220A for ; Tue, 31 May 2022 01:01:08 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id F2238B80E20 for ; Tue, 31 May 2022 08:01:06 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 14DB1C34119; Tue, 31 May 2022 08:01:03 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1653984065; bh=jDuI1OezqWtRhCv5nxYOoLBrXfUkFW5cSU8f621yeiU=; h=Date:Subject:To:Cc:References:From:In-Reply-To:From; b=hnncicVxHW86XI1aJOefhoU8I5MQu4f342W0L31+J4wjyHdqUCm7PYgbR4SI6I0rK J6ZlOAdYhe1wIVE/1+qT9gMtP0Z3yb9IBug/M+NY/Gc0q18XW+d2P1zf5XzrHeg6cs mNE+hrAXSeAeVwo2Hd0xCNf+hF9mhCDLSr7YQGUAC+iyYIKF6KMdcHTAvTPdUoDfK+ U2MRDCoQmoJHbmr7QPzDS7hBWRH1XOJudAL/nr3qM/9FIvTqvTV1Z948JqSHtiyI4l L1JC+LKawmNUq1WmeLQpy3uxugeDZItIS8CIUooyGhvO5qqLxPSBwNQNxBkykHV3dZ skS+W0hmvpODg== Message-ID: Date: Tue, 31 May 2022 16:01:01 +0800 MIME-Version: 1.0 User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101 Thunderbird/91.9.1 Subject: Re: [PATCH v4] f2fs: separate NOCoW and pinfile semantics Content-Language: en-US To: Jaegeuk Kim Cc: linux-f2fs-devel@lists.sourceforge.net, linux-kernel@vger.kernel.org, Chao Yu References: <20220517032410.3564033-1-chao@kernel.org> From: Chao Yu In-Reply-To: Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Status: No, score=-10.5 required=5.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF,NICE_REPLY_A, RCVD_IN_DNSWL_HI,SPF_HELO_NONE,SPF_PASS,T_SCC_BODY_TEXT_LINE autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on lindbergh.monkeyblade.net Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 2022/5/18 1:21, Jaegeuk Kim wrote: > On 05/17, Chao Yu wrote: >> Pinning a file is heavy, because skipping pinned files make GC >> running with heavy load or no effect. >> >> So that this patch proposes to separate nocow and pinfile semantics: >> - NOCoW flag can only be set on regular file. >> - NOCoW file will only trigger IPU at common writeback/flush. >> - NOCow file will do OPU during GC. > > How about adding > - NOCow file will allocate 2MB-aligned space via fallocate. > > So, it'd be same as file pinning except allowing GCs. wdyt? Well, it seems the segment-aligned allocation feature should not couple with pin_file or NoCow flag, what about introducing another ioctl for that? something like: F2FS_IOC_ALIGNMENT_PREALLOCATION w/ arg.alignment_start, arg.alignment_len, arg.total_len? Thanks, > >> >> This flag can satisfying the demand of: >> 1) avoiding fragment of file's physical block >> 2) userspace doesn't want to pin file's physical address >> >> After commit 5d539245cb18 ("f2fs: export FS_NOCOW_FL flag to user"), >> Pin_file and NOCoW flags have already been twined closely. e.g. >> once we set pinfile flag in file, nocow flag will be shown; and after >> clearing pinfile flag, nocow flag will disappear. >> >> So, in order to keep backward compatibility, let use below semantics: >> >> f2fs_ioc_set_pin_file/f2fs_fileattr_set logic: >> pinfile nocow >> set set pinfile | nocow set nocow >> clear clear pinfile | nocow clear nocow >> >> File Behaviors: >> w/ pinfile, w/ nocow: use pinfile semantics >> w/ pinfile, w/o nocow: use pinfile semantics >> w/o pinfile, w/ nocow: use nocow semantics >> w/o pinfile, w/o nocow: no pinfile or nocow semantics >> >> NOCoW can also be set on directory, and it will have no effect on >> directory, however, new files created in nocow directory will have the >> flag set. >> >> Signed-off-by: Chao Yu >> --- >> v4: >> - allow IPU only for NoCowed regular inode. >> fs/f2fs/data.c | 3 +++ >> fs/f2fs/f2fs.h | 13 +++++++++++-- >> fs/f2fs/file.c | 18 +++++++++++++++++- >> 3 files changed, 31 insertions(+), 3 deletions(-) >> >> diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c >> index 54a7a8ad994d..42d95ac6b508 100644 >> --- a/fs/f2fs/data.c >> +++ b/fs/f2fs/data.c >> @@ -2498,6 +2498,9 @@ bool f2fs_should_update_inplace(struct inode *inode, struct f2fs_io_info *fio) >> if (f2fs_is_pinned_file(inode)) >> return true; >> >> + if (S_ISREG(inode->i_mode) && F2FS_I(inode)->i_flags & F2FS_NOCOW_FL) >> + return true; >> + >> /* if this is cold file, we should overwrite to avoid fragmentation */ >> if (file_is_cold(inode)) >> return true; >> diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h >> index 492af5b96de1..5c67736000a7 100644 >> --- a/fs/f2fs/f2fs.h >> +++ b/fs/f2fs/f2fs.h >> @@ -2916,13 +2916,15 @@ static inline void f2fs_change_bit(unsigned int nr, char *addr) >> #define F2FS_NOCOMP_FL 0x00000400 /* Don't compress */ >> #define F2FS_INDEX_FL 0x00001000 /* hash-indexed directory */ >> #define F2FS_DIRSYNC_FL 0x00010000 /* dirsync behaviour (directories only) */ >> +#define F2FS_NOCOW_FL 0x00800000 /* Do not cow file */ >> #define F2FS_PROJINHERIT_FL 0x20000000 /* Create with parents projid */ >> #define F2FS_CASEFOLD_FL 0x40000000 /* Casefolded file */ >> >> /* Flags that should be inherited by new inodes from their parent. */ >> #define F2FS_FL_INHERITED (F2FS_SYNC_FL | F2FS_NODUMP_FL | F2FS_NOATIME_FL | \ >> F2FS_DIRSYNC_FL | F2FS_PROJINHERIT_FL | \ >> - F2FS_CASEFOLD_FL | F2FS_COMPR_FL | F2FS_NOCOMP_FL) >> + F2FS_CASEFOLD_FL | F2FS_COMPR_FL | F2FS_NOCOMP_FL | \ >> + F2FS_NOCOW_FL) >> >> /* Flags that are appropriate for regular files (all but dir-specific ones). */ >> #define F2FS_REG_FLMASK (~(F2FS_DIRSYNC_FL | F2FS_PROJINHERIT_FL | \ >> @@ -2954,9 +2956,16 @@ static inline void __mark_inode_dirty_flag(struct inode *inode, >> fallthrough; >> case FI_DATA_EXIST: >> case FI_INLINE_DOTS: >> - case FI_PIN_FILE: >> case FI_COMPRESS_RELEASED: >> f2fs_mark_inode_dirty_sync(inode, true); >> + break; >> + case FI_PIN_FILE: >> + if (set) >> + F2FS_I(inode)->i_flags |= F2FS_NOCOW_FL; >> + else >> + F2FS_I(inode)->i_flags &= ~F2FS_NOCOW_FL; >> + f2fs_mark_inode_dirty_sync(inode, true); >> + break; >> } >> } >> >> diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c >> index 216081ea8c81..9e71ce8601f9 100644 >> --- a/fs/f2fs/file.c >> +++ b/fs/f2fs/file.c >> @@ -1851,6 +1851,20 @@ static int f2fs_setflags_common(struct inode *inode, u32 iflags, u32 mask) >> if (IS_NOQUOTA(inode)) >> return -EPERM; >> >> + if ((iflags ^ masked_flags) & F2FS_NOCOW_FL) { >> + int ret; >> + >> + if (!S_ISREG(inode->i_mode) && !S_ISDIR(inode->i_mode)) >> + return -EINVAL; >> + if (S_ISREG(inode->i_mode)) { >> + if (f2fs_should_update_outplace(inode, NULL)) >> + return -EINVAL; >> + ret = f2fs_convert_inline_inode(inode); >> + if (ret) >> + return ret; >> + } >> + } >> + >> if ((iflags ^ masked_flags) & F2FS_CASEFOLD_FL) { >> if (!f2fs_sb_has_casefold(F2FS_I_SB(inode))) >> return -EOPNOTSUPP; >> @@ -1926,6 +1940,7 @@ static const struct { >> { F2FS_NOCOMP_FL, FS_NOCOMP_FL }, >> { F2FS_INDEX_FL, FS_INDEX_FL }, >> { F2FS_DIRSYNC_FL, FS_DIRSYNC_FL }, >> + { F2FS_NOCOW_FL, FS_NOCOW_FL }, >> { F2FS_PROJINHERIT_FL, FS_PROJINHERIT_FL }, >> { F2FS_CASEFOLD_FL, FS_CASEFOLD_FL }, >> }; >> @@ -1957,7 +1972,8 @@ static const struct { >> FS_NOCOMP_FL | \ >> FS_DIRSYNC_FL | \ >> FS_PROJINHERIT_FL | \ >> - FS_CASEFOLD_FL) >> + FS_CASEFOLD_FL | \ >> + FS_NOCOW_FL) >> >> /* Convert f2fs on-disk i_flags to FS_IOC_{GET,SET}FLAGS flags */ >> static inline u32 f2fs_iflags_to_fsflags(u32 iflags) >> -- >> 2.25.1