linux-kernel - Re: [PATCH] pstore: Revert pmsg

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 for Android: free password hash cracker in your pocket

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

Date:   Thu, 2 Mar 2023 16:56:13 -0500
From:   Steven Rostedt <rostedt@...dmis.org>
To:     John Stultz <jstultz@...gle.com>
Cc:     LKML <linux-kernel@...r.kernel.org>, Wei Wang <wvw@...gle.com>,
        Midas Chien <midaschieh@...gle.com>,
        Kees Cook <keescook@...omium.org>,
        Anton Vorontsov <anton@...msg.org>,
        "Guilherme G. Piccoli" <gpiccoli@...lia.com>,
        Tony Luck <tony.luck@...el.com>, kernel-team@...roid.com,
        Thomas Gleixner <tglx@...utronix.de>,
        Peter Zijlstra <peterz@...radead.org>,
        Sebastian Andrzej Siewior <bigeasy@...utronix.de>
Subject: Re: [PATCH] pstore: Revert pmsg_lock back to a normal mutex

On Thu, 2 Mar 2023 16:36:03 -0500
Steven Rostedt <rostedt@...dmis.org> wrote:

> I just noticed there's an rcu_read_lock() here around the loop.
> I'm guessing that's to keep this race from happening.
> Would have been nice to have a comment there stating such.

Knowing that rcu_read_lock() keeps the tasks safe, I made the optimization
to only grab the spinlock (and disable interrupts) once, or whenever the
top waiter changes.

-- Steve

diff --git a/kernel/locking/rtmutex.c b/kernel/locking/rtmutex.c
index 010cf4e6d0b8..37820331857b 100644
--- a/kernel/locking/rtmutex.c
+++ b/kernel/locking/rtmutex.c
@@ -1399,8 +1399,12 @@ static bool rtmutex_spin_on_owner(struct rt_mutex_base *lock,
 				  struct rt_mutex_waiter *waiter,
 				  struct task_struct *owner)
 {
+	struct rt_mutex_waiter *top_waiter;
+	struct rt_mutex_waiter *last_waiter = NULL;
+	struct task_struct *top_task = NULL;
 	bool res = true;
 
+	/* rcu_read_lock keeps task_structs around */
 	rcu_read_lock();
 	for (;;) {
 		/* If owner changed, trylock again. */
@@ -1421,11 +1425,27 @@ static bool rtmutex_spin_on_owner(struct rt_mutex_base *lock,
 		 *    for CONFIG_PREEMPT_RCU=y)
 		 *  - the VCPU on which owner runs is preempted
 		 */
-		if (!owner_on_cpu(owner) || need_resched() ||
-		    !rt_mutex_waiter_is_top_waiter(lock, waiter)) {
+		if (!owner_on_cpu(owner) || need_resched()) {
 			res = false;
 			break;
 		}
+		top_waiter = rt_mutex_top_waiter(lock);
+		if (top_waiter != waiter) {
+			if (top_waiter != last_waiter) {
+				raw_spin_lock_irq(&lock->wait_lock);
+				last_waiter = rt_mutex_top_waiter(lock);
+				top_task = last_waiter->task;
+				raw_spin_unlock_irq(&lock->wait_lock);
+			}
+			trace_printk("spin on owner! %s:%d\n",
+				     top_task->comm,
+				     top_task->pid);
+
+			if (!owner_on_cpu(top_task)) {
+				res = false;
+				break;
+			}
+		}
 		cpu_relax();
 	}
 	rcu_read_unlock();