Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
From:   Pingfan Liu <kernelfans@gmail.com>
To:     linux-kernel@vger.kernel.org
Cc:     Pingfan Liu <kernelfans@gmail.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        Valentin Schneider <valentin.schneider@arm.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Frederic Weisbecker <frederic@kernel.org>,
        Baokun Li <libaokun1@huawei.com>,
        Mark Rutland <mark.rutland@arm.com>,
        Steven Price <steven.price@arm.com>,
        "Jason A. Donenfeld" <Jason@zx2c4.com>,
        Yuan ZhaoXiong <yuanzhaoxiong@baidu.com>,
        YueHaibing <yuehaibing@huawei.com>,
        Randy Dunlap <rdunlap@infradead.org>
Subject: [PATCH 7/9] irq: remove needless lock in takedown_cpu()
Date:   Wed, 20 Apr 2022 22:05:19 +0800
Message-Id: <20220420140521.45361-8-kernelfans@gmail.com>
In-Reply-To: <20220420140521.45361-1-kernelfans@gmail.com>
References: <20220420140521.45361-1-kernelfans@gmail.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

Any allocation or release of irq_desc should fall into the interface of
__irq_alloc_descs() or irq_free_descs(). And both of them hold the mutex
sparse_irq_lock. This preemptability contradicts the preempt-disabled
context when dispatching fn in cpu_stopper_thread(). So allocation or
free of irq_desc can not be demanded in cpu_stopper_thread().

On the other hand, for the safety of access to irq_desc, rcu still keeps
watching the dying cpu until the last minute rcu_report_dead(), so each
cpu_stop_fn_t can safely access irq_desc.

As a result, in takedown_cpu() irq_lock_sparse()/irq_unlock_sparse() can
be safely removed.

Signed-off-by: Pingfan Liu <kernelfans@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Frederic Weisbecker <frederic@kernel.org>
Cc: Baokun Li <libaokun1@huawei.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Steven Price <steven.price@arm.com>
Cc: "Jason A. Donenfeld" <Jason@zx2c4.com>
Cc: Yuan ZhaoXiong <yuanzhaoxiong@baidu.com>
Cc: YueHaibing <yuehaibing@huawei.com>
Cc: Randy Dunlap <rdunlap@infradead.org>
To: linux-kernel@vger.kernel.org
---
 kernel/cpu.c | 11 +++--------
 1 file changed, 3 insertions(+), 8 deletions(-)

diff --git a/kernel/cpu.c b/kernel/cpu.c
index d0a9aa0b42e8..94a6b512c26d 100644
--- a/kernel/cpu.c
+++ b/kernel/cpu.c
@@ -1033,18 +1033,16 @@ static int takedown_cpu(unsigned int cpu)
 	kthread_park(st->thread);
 
 	/*
-	 * Prevent irq alloc/free while the dying cpu reorganizes the
-	 * interrupt affinities.
+	 * RCU keeps watching 'cpu' until do_idle()->rcu_report_dead().
+	 * And cpu_stopper's fn is dispatched with preemption disabled.
+	 * So it can not occur to release a irq_desc.
 	 */
-	irq_lock_sparse();
 
 	/*
 	 * So now all preempt/rcu users must observe !cpu_active().
 	 */
 	err = stop_machine_cpuslocked(take_cpu_down, NULL, cpumask_of(cpu));
 	if (err) {
-		/* CPU refused to die */
-		irq_unlock_sparse();
 		/* Unpark the hotplug thread so we can rollback there */
 		kthread_unpark(st->thread);
 		return err;
@@ -1061,9 +1059,6 @@ static int takedown_cpu(unsigned int cpu)
 	wait_for_ap_thread(st, false);
 	BUG_ON(st->state != CPUHP_AP_IDLE_DEAD);
 
-	/* Interrupts are moved away from the dying cpu, reenable alloc/free */
-	irq_unlock_sparse();
-
 	hotplug_cpu__broadcast_tick_pull(cpu);
 	/* This actually kills the CPU. */
 	__cpu_die(cpu);
-- 
2.31.1