From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id AF2593A59A3; Sat, 17 Oct 2026 16:07:45 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1792253267; cv=none; b=rKde/ve8Yk5ENQm6E8dJR9dnNJZW0zGA7fnZF+HZROBtrIppje6rZdwDEilKwsx4o7Cqyc0p5Dbp5ntKvDup3OkVisCqG7JGkfOd/BM2iMfAtW9TOBHTVG7D0Zgk47R+t9Ys6bleaGbnyFB8WdkWFr2xpntrxJeRFFH9rWzU4uM= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1792253267; c=relaxed/simple; bh=9IYapL14f7wYiqBBBHUdzlECjX8Zu4lXFY6pTfzowMU=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=Fk2yAOT3wfXNxt9r5tN0HXO7BgxGgtupGXll2QsK9Dgd6STVBt0Uscb78RZ+qBA/40bjgqfTbOHbqAd61hwgepKRi50AkoPJie5nVaq+PrqCq6LsA46TK8XWQ9cYKlU/UiA1bZbmbazhT3y651wx248Blt7WpjYBE2Pp28+96G4= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linuxfoundation.org header.i=@linuxfoundation.org header.b=U4vVLDRw; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linuxfoundation.org header.i=@linuxfoundation.org header.b="U4vVLDRw" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 8179C1F000C4; Sat, 17 Oct 2026 16:07:44 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linuxfoundation.org; s=korg; t=1792253265; bh=E67BCQSGwzG/zLSzuncD239k6/cEn/wMcXj/B7+7MXA=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=U4vVLDRwOPQuHoCTu3gu+vZFzCs9kFr4YrihICi0XIN0P1qYW6weZDJ70qFCwlYBC p+8/ot36LF2xfnzjxltzDpcUPIug+AFNqF1dMAvVl/Xz9uyKhnj/ma7w4dJ5kGseht /cI26QFGGF/sicj5ygmoz6e2uUUYhm6jY6Z2cvY0= From: Greg Kroah-Hartman To: stable@vger.kernel.org Cc: Greg Kroah-Hartman , patches@lists.linux.dev, Chao Yu , Jaegeuk Kim , Sasha Levin Subject: [PATCH 6.12 060/869] f2fs: quota: fix to avoid warning in dquot_writeback_dquots() Date: Sat, 17 Oct 2026 17:48:21 +0200 Message-ID: <20261017160038.561188391@linuxfoundation.org> X-Mailer: git-send-email 2.56.0 In-Reply-To: <20261017160036.979869304@linuxfoundation.org> References: <20261017160036.979869304@linuxfoundation.org> User-Agent: quilt/0.69 X-stable: review X-Patchwork-Hint: ignore Precedence: bulk X-Mailing-List: patches@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: 8bit 6.12-stable review patch. If anyone has any objections, please let me know. ------------------ From: Chao Yu [ Upstream commit eb85c2410d6f581e957cd03a644ff6ddbe592af9 ] F2FS-fs (dm-59): checkpoint=enable has some unwritten data. ------------[ cut here ]------------ WARNING: CPU: 6 PID: 8013 at fs/quota/dquot.c:691 dquot_writeback_dquots+0x2fc/0x308 pc : dquot_writeback_dquots+0x2fc/0x308 lr : f2fs_quota_sync+0xcc/0x1c4 Call trace: dquot_writeback_dquots+0x2fc/0x308 f2fs_quota_sync+0xcc/0x1c4 f2fs_write_checkpoint+0x3d4/0x9b0 f2fs_issue_checkpoint+0x1bc/0x2c0 f2fs_sync_fs+0x54/0x150 f2fs_do_sync_file+0x2f8/0x814 __f2fs_ioctl+0x1960/0x3244 f2fs_ioctl+0x54/0xe0 __arm64_sys_ioctl+0xa8/0xe4 invoke_syscall+0x58/0x114 checkpoint and f2fs_remount may race as below, resulting triggering warning in dquot_writeback_dquots(). atomic write remount - do_remount - down_write(&sb->s_umount); - f2fs_remount - ioctl - f2fs_do_sync_file - f2fs_sync_fs - f2fs_write_checkpoint - block_operations - locked = down_read_trylock(&sbi->sb->s_umount) : fail to lock due to the write lock was held by remount - up_write(&sb->s_umount); - f2fs_quota_sync - dquot_writeback_dquots - WARN_ON_ONCE(!rwsem_is_locked(&sb->s_umount)) : trigger warning because s_umount lock was unlocked by remount If checkpoint comes from mount/umount/remount/freeze/quotactl, caller of checkpoint has already held s_umount lock, calling dquot_writeback_dquots() in the context should be safe. So let's record task to sbi->umount_lock_holder, so that checkpoint can know whether the lock has held in the context or not by checking current w/ it. In addition, in order to not misrepresent caller of checkpoint, we should not allow to trigger async checkpoint for those callers: mount/umount/remount/ freeze/quotactl. Fixes: af033b2aa8a8 ("f2fs: guarantee journalled quota data by checkpoint") Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim Signed-off-by: Sasha Levin --- fs/f2fs/checkpoint.c | 15 ++++++---- fs/f2fs/f2fs.h | 3 +- fs/f2fs/super.c | 65 ++++++++++++++++++++++++++++++++++---------- 3 files changed, 61 insertions(+), 22 deletions(-) diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index 6322bf2e16228..dc9540400caae 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -1249,7 +1249,7 @@ static int block_operations(struct f2fs_sb_info *sbi) retry_flush_quotas: f2fs_lock_all(sbi); if (__need_flush_quota(sbi)) { - int locked; + bool need_lock = sbi->umount_lock_holder != current; if (++cnt > DEFAULT_RETRY_QUOTA_FLUSH_COUNT) { set_sbi_flag(sbi, SBI_QUOTA_SKIP_FLUSH); @@ -1258,11 +1258,13 @@ static int block_operations(struct f2fs_sb_info *sbi) } f2fs_unlock_all(sbi); - /* only failed during mount/umount/freeze/quotactl */ - locked = down_read_trylock(&sbi->sb->s_umount); - f2fs_quota_sync(sbi->sb, -1); - if (locked) + /* don't grab s_umount lock during mount/umount/remount/freeze/quotactl */ + if (!need_lock) { + f2fs_do_quota_sync(sbi->sb, -1); + } else if (down_read_trylock(&sbi->sb->s_umount)) { + f2fs_do_quota_sync(sbi->sb, -1); up_read(&sbi->sb->s_umount); + } cond_resched(); goto retry_flush_quotas; } @@ -1870,7 +1872,8 @@ int f2fs_issue_checkpoint(struct f2fs_sb_info *sbi) struct cp_control cpc; cpc.reason = __get_cp_reason(sbi); - if (!test_opt(sbi, MERGE_CHECKPOINT) || cpc.reason != CP_SYNC) { + if (!test_opt(sbi, MERGE_CHECKPOINT) || cpc.reason != CP_SYNC || + sbi->umount_lock_holder == current) { int ret; f2fs_down_write(&sbi->gc_lock); diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 12ca7ba9d6dc4..81c2a566f1d42 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -1708,6 +1708,7 @@ struct f2fs_sb_info { unsigned int nquota_files; /* # of quota sysfile */ struct f2fs_rwsem quota_sem; /* blocking cp for flags */ + struct task_struct *umount_lock_holder; /* s_umount lock holder */ /* # of pages, see count_type */ atomic_t nr_pages[NR_COUNT_TYPE]; @@ -3728,7 +3729,7 @@ int f2fs_inode_dirtied(struct inode *inode, bool sync); void f2fs_inode_synced(struct inode *inode); int f2fs_dquot_initialize(struct inode *inode); int f2fs_enable_quota_files(struct f2fs_sb_info *sbi, bool rdonly); -int f2fs_quota_sync(struct super_block *sb, int type); +int f2fs_do_quota_sync(struct super_block *sb, int type); loff_t max_file_blocks(struct inode *inode); void f2fs_quota_off_umount(struct super_block *sb); void f2fs_save_errors(struct f2fs_sb_info *sbi, unsigned char flag); diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index f7acb1998b2cd..c278ce7842056 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -1736,22 +1736,28 @@ int f2fs_sync_fs(struct super_block *sb, int sync) static int f2fs_freeze(struct super_block *sb) { + struct f2fs_sb_info *sbi = F2FS_SB(sb); + if (f2fs_readonly(sb)) return 0; /* IO error happened before */ - if (unlikely(f2fs_cp_error(F2FS_SB(sb)))) + if (unlikely(f2fs_cp_error(sbi))) return -EIO; /* must be clean, since sync_filesystem() was already called */ - if (is_sbi_flag_set(F2FS_SB(sb), SBI_IS_DIRTY)) + if (is_sbi_flag_set(sbi, SBI_IS_DIRTY)) return -EINVAL; + sbi->umount_lock_holder = current; + /* Let's flush checkpoints and stop the thread. */ - f2fs_flush_ckpt_thread(F2FS_SB(sb)); + f2fs_flush_ckpt_thread(sbi); + + sbi->umount_lock_holder = NULL; /* to avoid deadlock on f2fs_evict_inode->SB_FREEZE_FS */ - set_sbi_flag(F2FS_SB(sb), SBI_IS_FREEZING); + set_sbi_flag(sbi, SBI_IS_FREEZING); return 0; } @@ -2363,6 +2369,8 @@ static int f2fs_remount(struct super_block *sb, int *flags, char *data) org_mount_opt = sbi->mount_opt; old_sb_flags = sb->s_flags; + sbi->umount_lock_holder = current; + #ifdef CONFIG_QUOTA org_mount_opt.s_jquota_fmt = F2FS_OPTION(sbi).s_jquota_fmt; for (i = 0; i < MAXQUOTAS; i++) { @@ -2593,6 +2601,8 @@ static int f2fs_remount(struct super_block *sb, int *flags, char *data) limit_reserve_root(sbi); adjust_unusable_cap_perc(sbi); *flags = (*flags & ~SB_LAZYTIME) | (sb->s_flags & SB_LAZYTIME); + + sbi->umount_lock_holder = NULL; return 0; restore_checkpoint: if (need_enable_checkpoint) { @@ -2634,6 +2644,8 @@ static int f2fs_remount(struct super_block *sb, int *flags, char *data) #endif sbi->mount_opt = org_mount_opt; sb->s_flags = old_sb_flags; + + sbi->umount_lock_holder = NULL; return err; } @@ -2950,7 +2962,7 @@ static int f2fs_quota_sync_file(struct f2fs_sb_info *sbi, int type) return ret; } -int f2fs_quota_sync(struct super_block *sb, int type) +int f2fs_do_quota_sync(struct super_block *sb, int type) { struct f2fs_sb_info *sbi = F2FS_SB(sb); struct quota_info *dqopt = sb_dqopt(sb); @@ -2998,11 +3010,21 @@ int f2fs_quota_sync(struct super_block *sb, int type) return ret; } +static int f2fs_quota_sync(struct super_block *sb, int type) +{ + int ret; + + F2FS_SB(sb)->umount_lock_holder = current; + ret = f2fs_do_quota_sync(sb, type); + F2FS_SB(sb)->umount_lock_holder = NULL; + return ret; +} + static int f2fs_quota_on(struct super_block *sb, int type, int format_id, const struct path *path) { struct inode *inode; - int err; + int err = 0; /* if quota sysfile exists, deny enabling quota with specific file */ if (f2fs_sb_has_quota_ino(F2FS_SB(sb))) { @@ -3013,31 +3035,34 @@ static int f2fs_quota_on(struct super_block *sb, int type, int format_id, if (path->dentry->d_sb != sb) return -EXDEV; - err = f2fs_quota_sync(sb, type); + F2FS_SB(sb)->umount_lock_holder = current; + + err = f2fs_do_quota_sync(sb, type); if (err) - return err; + goto out; inode = d_inode(path->dentry); err = filemap_fdatawrite(inode->i_mapping); if (err) - return err; + goto out; err = filemap_fdatawait(inode->i_mapping); if (err) - return err; + goto out; err = dquot_quota_on(sb, type, format_id, path); if (err) - return err; + goto out; inode_lock(inode); F2FS_I(inode)->i_flags |= F2FS_QUOTA_DEFAULT_FL; f2fs_set_inode_flags(inode); inode_unlock(inode); f2fs_mark_inode_dirty_sync(inode, false); - - return 0; +out: + F2FS_SB(sb)->umount_lock_holder = NULL; + return err; } static int __f2fs_quota_off(struct super_block *sb, int type) @@ -3048,7 +3073,7 @@ static int __f2fs_quota_off(struct super_block *sb, int type) if (!inode || !igrab(inode)) return dquot_quota_off(sb, type); - err = f2fs_quota_sync(sb, type); + err = f2fs_do_quota_sync(sb, type); if (err) goto out_put; @@ -3071,6 +3096,8 @@ static int f2fs_quota_off(struct super_block *sb, int type) struct f2fs_sb_info *sbi = F2FS_SB(sb); int err; + F2FS_SB(sb)->umount_lock_holder = current; + err = __f2fs_quota_off(sb, type); /* @@ -3080,6 +3107,9 @@ static int f2fs_quota_off(struct super_block *sb, int type) */ if (is_journalled_quota(sbi)) set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR); + + F2FS_SB(sb)->umount_lock_holder = NULL; + return err; } @@ -3212,7 +3242,7 @@ int f2fs_dquot_initialize(struct inode *inode) return 0; } -int f2fs_quota_sync(struct super_block *sb, int type) +int f2fs_do_quota_sync(struct super_block *sb, int type) { return 0; } @@ -4760,6 +4790,7 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent) if (err) goto free_compress_inode; + sbi->umount_lock_holder = current; #ifdef CONFIG_QUOTA /* Enable quota usage during mount */ if (f2fs_sb_has_quota_ino(sbi) && !f2fs_readonly(sb)) { @@ -4892,6 +4923,8 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent) f2fs_update_time(sbi, CP_TIME); f2fs_update_time(sbi, REQ_TIME); clear_sbi_flag(sbi, SBI_CP_DISABLED_QUICK); + + sbi->umount_lock_holder = NULL; return 0; sync_free_meta: @@ -4997,6 +5030,8 @@ static void kill_f2fs_super(struct super_block *sb) struct f2fs_sb_info *sbi = F2FS_SB(sb); if (sb->s_root) { + sbi->umount_lock_holder = current; + set_sbi_flag(sbi, SBI_IS_CLOSE); f2fs_stop_gc_thread(sbi); f2fs_stop_discard_thread(sbi); -- 2.53.0