From nobody Thu Apr 30 08:30:47 2026 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 8BB02C433EF for ; Wed, 13 Jul 2022 12:48:17 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231861AbiGMMsP (ORCPT ); Wed, 13 Jul 2022 08:48:15 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:33988 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229774AbiGMMsN (ORCPT ); Wed, 13 Jul 2022 08:48:13 -0400 Received: from szxga08-in.huawei.com (szxga08-in.huawei.com [45.249.212.255]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 75882224; Wed, 13 Jul 2022 05:48:11 -0700 (PDT) Received: from dggemv711-chm.china.huawei.com (unknown [172.30.72.55]) by szxga08-in.huawei.com (SkyGuard) with ESMTP id 4LjcmG5Tvnz1L96G; Wed, 13 Jul 2022 20:45:34 +0800 (CST) Received: from kwepemm600013.china.huawei.com (7.193.23.68) by dggemv711-chm.china.huawei.com (10.1.198.66) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2375.24; Wed, 13 Jul 2022 20:48:07 +0800 Received: from huawei.com (10.175.127.227) by kwepemm600013.china.huawei.com (7.193.23.68) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2375.24; Wed, 13 Jul 2022 20:48:06 +0800 From: Zhihao Cheng To: , , , , , CC: , , , Subject: [PATCH v5] proc: Fix a dentry lock race between release_task and lookup Date: Wed, 13 Jul 2022 21:00:29 +0800 Message-ID: <20220713130029.4133533-1-chengzhihao1@huawei.com> X-Mailer: git-send-email 2.31.1 MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.175.127.227] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600013.china.huawei.com (7.193.23.68) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Commit 7bc3e6e55acf06 ("proc: Use a list of inodes to flush from proc") moved proc_flush_task() behind __exit_signal(). Then, process systemd can take long period high cpu usage during releasing task in following concurrent processes: systemd ps kernel_waitid stat(/proc/tgid) do_wait filename_lookup wait_consider_task lookup_fast release_task __exit_signal __unhash_process detach_pid __change_pid // remove task->pid_links d_revalidate -> pid_revalidate // 0 d_invalidate(/proc/tgid) shrink_dcache_parent(/proc/tgid) d_walk(/proc/tgid) spin_lock_nested(/proc/tgid/fd) // iterating opened fd proc_flush_pid | d_invalidate (/proc/tgid/fd) | shrink_dcache_parent(/proc/tgid/fd) | shrink_dentry_list(subdirs) =E2=86=93 shrink_lock_dentry(/proc/tgid/fd) --> race on dentry lock Function d_invalidate() will remove dentry from hash firstly, but why does proc_flush_pid() process dentry '/proc/tgid/fd' before dentry '/proc/tgid'? That's because proc_pid_make_inode() adds proc inode in reverse order by invoking hlist_add_head_rcu(). But proc should not add any inodes under '/proc/tgid' except '/proc/tgid/task/pid', fix it by adding inode into 'pid->inodes' only if the inode is /proc/tgid or /proc/tgid/task/pid. Performance regression: Create 200 tasks, each task open one file for 50,000 times. Kill all tasks when opened files exceed 10,000,000 (cat /proc/sys/fs/file-nr). Before fix: $ time killall -wq aa real 4m40.946s # During this period, we can see 'ps' and 'systemd' taking high cpu usage. After fix: $ time killall -wq aa real 1m20.732s # During this period, we can see 'systemd' taking high cpu usage. Fixes: 7bc3e6e55acf06 ("proc: Use a list of inodes to flush from proc") Link: https://bugzilla.kernel.org/show_bug.cgi?id=3D216054 Signed-off-by: Zhihao Cheng Signed-off-by: Zhang Yi Suggested-by: Brian Foster Reviewed-by: Brian Foster --- v1->v2: Add new helper proc_pid_make_base_inode that performs the extra work of adding to the pid->list. v2->v3: Add performance regression in commit message. v3->v4: Make proc_pid_make_base_inode() static v4->v5: Add notes to explain what proc_pid_make_base_inode() does fs/proc/base.c | 46 ++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 38 insertions(+), 8 deletions(-) diff --git a/fs/proc/base.c b/fs/proc/base.c index 8dfa36a99c74..93f7e3d971e4 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -1885,7 +1885,7 @@ void proc_pid_evict_inode(struct proc_inode *ei) put_pid(pid); } =20 -struct inode *proc_pid_make_inode(struct super_block * sb, +struct inode *proc_pid_make_inode(struct super_block *sb, struct task_struct *task, umode_t mode) { struct inode * inode; @@ -1914,11 +1914,6 @@ struct inode *proc_pid_make_inode(struct super_block= * sb, =20 /* Let the pid remember us for quick removal */ ei->pid =3D pid; - if (S_ISDIR(mode)) { - spin_lock(&pid->lock); - hlist_add_head_rcu(&ei->sibling_inodes, &pid->inodes); - spin_unlock(&pid->lock); - } =20 task_dump_owner(task, 0, &inode->i_uid, &inode->i_gid); security_task_to_inode(task, inode); @@ -1931,6 +1926,39 @@ struct inode *proc_pid_make_inode(struct super_block= * sb, return NULL; } =20 +/* + * Generating an inode and adding it into @pid->inodes, so that task will + * invalidate inode's dentry before being released. + * + * This helper is used for creating dir-type entries under '/proc' and + * '/proc//task'. Other entries(eg. fd, stat) under '/proc/' + * can be released by invalidating '/proc/' dentry. + * In theory, dentries under '/proc//task' can also be released by + * invalidating '/proc/' dentry, we reserve it to handle single + * thread exiting situation: Any one of threads should invalidate its + * '/proc//task/' dentry before released. + */ +static struct inode *proc_pid_make_base_inode(struct super_block *sb, + struct task_struct *task, umode_t mode) +{ + struct inode *inode; + struct proc_inode *ei; + struct pid *pid; + + inode =3D proc_pid_make_inode(sb, task, mode); + if (!inode) + return NULL; + + /* Let proc_flush_pid find this directory inode */ + ei =3D PROC_I(inode); + pid =3D ei->pid; + spin_lock(&pid->lock); + hlist_add_head_rcu(&ei->sibling_inodes, &pid->inodes); + spin_unlock(&pid->lock); + + return inode; +} + int pid_getattr(struct user_namespace *mnt_userns, const struct path *path, struct kstat *stat, u32 request_mask, unsigned int query_flags) { @@ -3369,7 +3397,8 @@ static struct dentry *proc_pid_instantiate(struct den= try * dentry, { struct inode *inode; =20 - inode =3D proc_pid_make_inode(dentry->d_sb, task, S_IFDIR | S_IRUGO | S_I= XUGO); + inode =3D proc_pid_make_base_inode(dentry->d_sb, task, + S_IFDIR | S_IRUGO | S_IXUGO); if (!inode) return ERR_PTR(-ENOENT); =20 @@ -3671,7 +3700,8 @@ static struct dentry *proc_task_instantiate(struct de= ntry *dentry, struct task_struct *task, const void *ptr) { struct inode *inode; - inode =3D proc_pid_make_inode(dentry->d_sb, task, S_IFDIR | S_IRUGO | S_I= XUGO); + inode =3D proc_pid_make_base_inode(dentry->d_sb, task, + S_IFDIR | S_IRUGO | S_IXUGO); if (!inode) return ERR_PTR(-ENOENT); =20 --=20 2.31.1