RDMA: Fix potential deadlock on &dev->rdi.pending_lock
PATCH 1/1 RDMA: Fix potential deadlock on &dev->rdi.pending_lock
qib_7322intr() could introduce double locks on &dev->rdi.pending_lock if it preempts other execution units requiring the same locks.
<Deadlock #1>qib_notify_error_qp()
--> spin_lock(&dev->rdi.pending_lock)
<interrupt>
--> qib_7322intr()
--> qib_ib_piobufavail()
--> spin_lock_irqsave(&dev->rdi.pending_lock)
<Deadlock #2>qib_flush_qp_waiters()
--> spin_lock(&dev->rdi.pending_lock)
<interrupt>
--> qib_7220intr()
--> qib_ib_piobufavail()
--> spin_lock_irqsave(&dev->rdi.pending_lock)
This flaw was found by an experimental static analysis tool I am developing for irq-related deadlock.
To prevent the potential deadlock, the patch uses spin_lock_irqsave() on &dev->rdi.pending_lock inside qib_notify_error_qp() and qib_flush_qp_waiters() to prevent the possible deadlock scenario.
--- drivers/infiniband/hw/qib/qib_qp.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-)
drivers/infiniband/hw/qib/qib_qp.c +6 -4
@@ -241,13 +241,14 @@ void qib_notify_error_qp(struct rvt_qp *qp) {struct qib_qp_priv *priv = qp->priv;struct qib_ibdev *dev =to_idev (qp->ibqp.device); +unsigned long flags; -spin_lock (&dev->rdi.pending_lock); +spin_lock_irqsave (&dev->rdi.pending_lock, flags);if (!list_empty (&priv->iowait) && !(qp->s_flags &RVT_S_BUSY )) { qp->s_flags &= ~RVT_S_ANY_WAIT_IO ;list_del_init (&priv->iowait); } -spin_unlock (&dev->rdi.pending_lock); +spin_unlock_irqrestore (&dev->rdi.pending_lock, flags);if (!(qp->s_flags &RVT_S_BUSY )) { qp->s_hdrwords =0 ;
@@ -367,11 +368,12 @@ void qib_flush_qp_waiters(struct rvt_qp *qp) {struct qib_qp_priv *priv = qp->priv;struct qib_ibdev *dev =to_idev (qp->ibqp.device); +unsigned long flags; -spin_lock (&dev->rdi.pending_lock); +spin_lock_irqsave (&dev->rdi.pending_lock, flags);if (!list_empty (&priv->iowait))list_del_init (&priv->iowait); -spin_unlock (&dev->rdi.pending_lock); +spin_unlock_irqrestore (&dev->rdi.pending_lock, flags); }/**
Signature · 1 line
2.17.1
I just noticed the two callbacks already called with irq disabled, then the patch is unnecessary, sorry for the interruption.
Regards,
Chengfeng