Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Chengfeng Ye <dg573847474@gmail.com>
To:     njavali@marvell.com, mrangankar@marvell.com,
        GR-QLogic-Storage-Upstream@marvell.com, jejb@linux.ibm.com,
        martin.petersen@oracle.com
Cc:     linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org,
        Chengfeng Ye <dg573847474@gmail.com>
Subject: [PATCH] scsi: qedi: Fix potential deadlock on &qedi_percpu->p_work_lock
Date:   Wed, 26 Jul 2023 12:56:55 +0000
Message-Id: <20230726125655.4197-1-dg573847474@gmail.com>
Precedence: bulk

As &qedi_percpu->p_work_lock is acquired by hard irq qedi_msix_handler(),
other acquisition of the same lock under process context should disable
irq, otherwise deadlock could happen if the irq preempt the execution
while the lock is held in process context on the same CPU.

qedi_cpu_offline() is one such function acquires the lock on process
context.

[Deadlock Scenario]
qedi_cpu_offline()
    ->spin_lock(&p->p_work_lock)
        <irq>
        ->qedi_msix_handler()
        ->edi_process_completions()
        ->spin_lock_irqsave(&p->p_work_lock, flags); (deadlock here)

This flaw was found by an experimental static analysis tool I am
developing for irq-related deadlock.

The tentative patch fix the potential deadlock by spin_lock_irqsave()
under process context.

Signed-off-by: Chengfeng Ye <dg573847474@gmail.com>
---
 drivers/scsi/qedi/qedi_main.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/scsi/qedi/qedi_main.c b/drivers/scsi/qedi/qedi_main.c
index 450522b204d6..77a56a136678 100644
--- a/drivers/scsi/qedi/qedi_main.c
+++ b/drivers/scsi/qedi/qedi_main.c
@@ -1976,8 +1976,9 @@ static int qedi_cpu_offline(unsigned int cpu)
 	struct qedi_percpu_s *p = this_cpu_ptr(&qedi_percpu);
 	struct qedi_work *work, *tmp;
 	struct task_struct *thread;
+	unsigned long flags;
 
-	spin_lock_bh(&p->p_work_lock);
+	spin_lock_irqsave(&p->p_work_lock, flags);
 	thread = p->iothread;
 	p->iothread = NULL;
 
@@ -1988,7 +1989,7 @@ static int qedi_cpu_offline(unsigned int cpu)
 			kfree(work);
 	}
 
-	spin_unlock_bh(&p->p_work_lock);
+	spin_unlock_irqrestore(&p->p_work_lock, flags);
 	if (thread)
 		kthread_stop(thread);
 	return 0;
-- 
2.17.1