Linux-Security-Module Archive on lore.kernel.org
 help / color / Atom feed
From: Stephen Brennan <stephen.s.brennan@oracle.com>
To: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Stephen Brennan <stephen.s.brennan@oracle.com>,
	James Morris <jmorris@namei.org>,
	"Serge E. Hallyn" <serge@hallyn.com>,
	linux-security-module@vger.kernel.org,
	Paul Moore <paul@paul-moore.com>,
	Stephen Smalley <stephen.smalley.work@gmail.com>,
	Eric Paris <eparis@parisplace.org>,
	selinux@vger.kernel.org, Casey Schaufler <casey@schaufler-ca.com>,
	Eric Biederman <ebiederm@xmission.com>,
	Alexander Viro <viro@zeniv.linux.org.uk>,
	linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
	Matthew Wilcox <willy@infradead.org>
Subject: [PATCH RESEND v5] proc: Allow pid_revalidate() during LOOKUP_RCU
Date: Thu,  4 Feb 2021 09:17:19 -0800
Message-ID: <20210204171719.1021547-1-stephen.s.brennan@oracle.com> (raw)

The pid_revalidate() function drops from RCU into REF lookup mode. When
many threads are resolving paths within /proc in parallel, this can
result in heavy spinlock contention on d_lockref as each thread tries to
grab a reference to the /proc dentry (and drop it shortly thereafter).

Investigation indicates that it is not necessary to drop RCU in
pid_revalidate(), as no RCU data is modified and the function never
sleeps. So, remove the LOOKUP_RCU check.

Signed-off-by: Stephen Brennan <stephen.s.brennan@oracle.com>
---

When running running ~100 parallel instances of "TZ=/etc/localtime ps -fe
>/dev/null" on a 100CPU machine, the %sys utilization reaches 90%, and perf
shows the following code path as being responsible for heavy contention on
the d_lockref spinlock:

      walk_component()
        lookup_fast()
          d_revalidate()
            pid_revalidate() // returns -ECHILD
          unlazy_child()
            lockref_get_not_dead(&nd->path.dentry->d_lockref) <-- contention

By applying this patch, %sys utilization falls to around 60% under the same
workload. Although this particular workload is a bit contrived, we have
seen some monitoring scripts which produced similarly high %sys time due to
this contention.

As a result this patch, several procfs methods which were only called in
ref-walk mode could now be called from RCU mode. To ensure that this patch
is safe, I audited all the inode get_link and permission() implementations,
as well as dentry d_revalidate() implementations, in fs/proc. These methods
are called in the following ways:

* get_link() receives a NULL dentry pointer when called in RCU mode.
* permission() receives MAY_NOT_BLOCK in the mode parameter when called
  from RCU.
* d_revalidate() receives LOOKUP_RCU in flags.

There were generally three groups I found. Group (1) are functions which
contain a check at the top of the function and return -ECHILD, and so
appear to be trivially RCU safe (although this is by dropping out of RCU
completely). Group (2) are functions which have no explicit check, but
on my audit, I was confident that there were no sleeping function calls,
and thus were RCU safe as is. However, I would appreciate any additional
review if possible. Group (3) are functions which might be be unsafe for some
reason or another.

Group (1):
 proc_ns_get_link()
 proc_pid_get_link()
 map_files_d_revalidate()
 proc_misc_d_revalidate()
 tid_fd_revalidate()

Group (2):
 proc_get_link()
 proc_self_get_link()
 proc_thread_self_get_link()
 proc_fd_permission()

Group (3):
 pid_revalidate()            -- addressed by my patch
 proc_pid_permission()       -- addressed by commits by Al
 proc_map_files_get_link()   -- calls capable() which could audit

I believe proc_map_files_get_link() is safe despite calling into the audit
framework, however I'm not confident and so I did not include it in Group 2.
proc_pid_permission() calls into the audit code, and is not safe with
LSM_AUDIT_DATA_DENTRY or LSM_AUDIT_DATA_INODE. Al's commits[1] address
these issues. This patch is tested and stable on the workload described
at the beginning of this cover letter, on a system with selinux enabled.

[1]: 23d8f5b684fc ("make dump_common_audit_data() safe to be called from
     RCU pathwalk") and 2 previous

Changes in v5:
- Al's commits are now in linux-next, resolving proc_pid_permission() issue.
- Add NULL check after d_inode_rcu(dentry), because inode may become NULL if
  we do not hold a reference.
Changes in v4:
- Simplify by unconditionally calling pid_update_inode() from pid_revalidate,
  and removing the LOOKUP_RCU check.
Changes in v3:
- Rather than call pid_update_inode() with flags, create
  proc_inode_needs_update() to determine whether the call can be skipped.
- Restore the call to the security hook (see next patch).
Changes in v2:
- Remove get_pid_task_rcu_user() and get_proc_task_rcu(), since they were
  unnecessary.
- Remove the call to security_task_to_inode().

 fs/proc/base.c | 18 ++++++++++--------
 1 file changed, 10 insertions(+), 8 deletions(-)

diff --git a/fs/proc/base.c b/fs/proc/base.c
index ebea9501afb8..3e105bd05801 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -1830,19 +1830,21 @@ static int pid_revalidate(struct dentry *dentry, unsigned int flags)
 {
 	struct inode *inode;
 	struct task_struct *task;
+	int ret = 0;
 
-	if (flags & LOOKUP_RCU)
-		return -ECHILD;
-
-	inode = d_inode(dentry);
-	task = get_proc_task(inode);
+	rcu_read_lock();
+	inode = d_inode_rcu(dentry);
+	if (!inode)
+		goto out;
+	task = pid_task(proc_pid(inode), PIDTYPE_PID);
 
 	if (task) {
 		pid_update_inode(task, inode);
-		put_task_struct(task);
-		return 1;
+		ret = 1;
 	}
-	return 0;
+out:
+	rcu_read_unlock();
+	return ret;
 }
 
 static inline bool proc_inode_is_dead(struct inode *inode)
-- 
2.27.0


             reply index

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-02-04 17:17 Stephen Brennan [this message]
2021-02-17 18:11 ` Stephen Brennan
2021-04-16  1:16 Stephen Brennan
2021-05-19 20:31 ` Stephen Brennan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210204171719.1021547-1-stephen.s.brennan@oracle.com \
    --to=stephen.s.brennan@oracle.com \
    --cc=adobriyan@gmail.com \
    --cc=casey@schaufler-ca.com \
    --cc=ebiederm@xmission.com \
    --cc=eparis@parisplace.org \
    --cc=jmorris@namei.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-security-module@vger.kernel.org \
    --cc=paul@paul-moore.com \
    --cc=selinux@vger.kernel.org \
    --cc=serge@hallyn.com \
    --cc=stephen.smalley.work@gmail.com \
    --cc=viro@zeniv.linux.org.uk \
    --cc=willy@infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

Linux-Security-Module Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/linux-security-module/0 linux-security-module/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 linux-security-module linux-security-module/ https://lore.kernel.org/linux-security-module \
		linux-security-module@vger.kernel.org
	public-inbox-index linux-security-module

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.linux-security-module


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git