linux-kernel - [tip: core/rcu] rcu: Add WRITE_ONCE() to rcu_node ->boost

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <158923078527.390.9732610515449854619.tip-bot2@tip-bot2>
Date:   Mon, 11 May 2020 20:59:45 -0000
From:   "tip-bot2 for Paul E. McKenney" <tip-bot2@...utronix.de>
To:     linux-tip-commits@...r.kernel.org
Cc:     "Paul E. McKenney" <paulmck@...nel.org>, x86 <x86@...nel.org>,
        LKML <linux-kernel@...r.kernel.org>
Subject: [tip: core/rcu] rcu: Add WRITE_ONCE() to rcu_node ->boost_tasks

The following commit has been merged into the core/rcu branch of tip:

Commit-ID:     5822b8126ff01e0baaf7d5168adc4ac8aeae088c
Gitweb:        https://git.kernel.org/tip/5822b8126ff01e0baaf7d5168adc4ac8aeae088c
Author:        Paul E. McKenney <paulmck@...nel.org>
AuthorDate:    Sat, 04 Jan 2020 10:44:41 -08:00
Committer:     Paul E. McKenney <paulmck@...nel.org>
CommitterDate: Mon, 27 Apr 2020 11:01:16 -07:00

rcu: Add WRITE_ONCE() to rcu_node ->boost_tasks

The rcu_node structure's ->boost_tasks field is read locklessly, so this
commit adds the WRITE_ONCE() to an update in order to provide proper
documentation and READ_ONCE()/WRITE_ONCE() pairing.

This data race was reported by KCSAN.  Not appropriate for backporting
due to failure being unlikely.

Signed-off-by: Paul E. McKenney <paulmck@...nel.org>
---
 kernel/rcu/tree_plugin.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h
index ed6bb46..664d0aa 100644
--- a/kernel/rcu/tree_plugin.h
+++ b/kernel/rcu/tree_plugin.h
@@ -505,7 +505,7 @@ rcu_preempt_deferred_qs_irqrestore(struct task_struct *t, unsigned long flags)
 			/* Snapshot ->boost_mtx ownership w/rnp->lock held. */
 			drop_boost_mutex = rt_mutex_owner(&rnp->boost_mtx) == t;
 			if (&t->rcu_node_entry == rnp->boost_tasks)
-				rnp->boost_tasks = np;
+				WRITE_ONCE(rnp->boost_tasks, np);
 		}
 
 		/*
@@ -1082,7 +1082,7 @@ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags)
 	     rnp->qsmask == 0 &&
 	     (ULONG_CMP_GE(jiffies, rnp->boost_time) || rcu_state.cbovld))) {
 		if (rnp->exp_tasks == NULL)
-			rnp->boost_tasks = rnp->gp_tasks;
+			WRITE_ONCE(rnp->boost_tasks, rnp->gp_tasks);
 		raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
 		rcu_wake_cond(rnp->boost_kthread_task,
 			      READ_ONCE(rnp->boost_kthread_status));