From: Chengfeng Ye dg573847474@gmail.com
[ Upstream commit dd64f80587190265ca8a0f4be6c64c2fda6d3ac2 ]
As &qedi_percpu->p_work_lock is acquired by hard IRQ qedi_msix_handler(), other acquisitions of the same lock under process context should disable IRQ, otherwise deadlock could happen if the IRQ preempts the execution while the lock is held in process context on the same CPU.
qedi_cpu_offline() is one such function which acquires the lock in 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 deadlocks.
The tentative patch fix the potential deadlock by spin_lock_irqsave() under process context.
Signed-off-by: Chengfeng Ye dg573847474@gmail.com Link: https://lore.kernel.org/r/20230726125655.4197-1-dg573847474@gmail.com Acked-by: Manish Rangankar mrangankar@marvell.com Signed-off-by: Martin K. Petersen martin.petersen@oracle.com Signed-off-by: Sasha Levin sashal@kernel.org --- 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 09f57ef35990c..b8b177018031c 100644 --- a/drivers/scsi/qedi/qedi_main.c +++ b/drivers/scsi/qedi/qedi_main.c @@ -1669,8 +1669,9 @@ 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;
@@ -1681,7 +1682,7 @@ 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;