dmaengine: idxd: fix calling wq quiesce inside spinlock
authorDave Jiang <dave.jiang@intel.com>
Wed, 17 Nov 2021 17:03:51 +0000 (10:03 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 22 Dec 2021 08:32:38 +0000 (09:32 +0100)
[ Upstream commit fa51b16d05583c7aebbc06330afb50276243d198 ]

Dan reports that smatch has found idxd_wq_quiesce() is being called inside
the idxd->dev_lock. idxd_wq_quiesce() calls wait_for_completion() and
therefore it can sleep. Move the call outside of the spinlock as it does
not need device lock.

Fixes: 5b0c68c473a1 ("dmaengine: idxd: support reporting of halt interrupt")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/163716858508.1721911.15051495873516709923.stgit@djiang5-desk3.ch.intel.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
drivers/dma/idxd/irq.c

index 3261ea2..6d6af0d 100644 (file)
@@ -137,10 +137,10 @@ halt:
                        INIT_WORK(&idxd->work, idxd_device_reinit);
                        queue_work(idxd->wq, &idxd->work);
                } else {
-                       spin_lock(&idxd->dev_lock);
                        idxd->state = IDXD_DEV_HALTED;
                        idxd_wqs_quiesce(idxd);
                        idxd_wqs_unmap_portal(idxd);
+                       spin_lock(&idxd->dev_lock);
                        idxd_device_clear_state(idxd);
                        dev_err(&idxd->pdev->dev,
                                "idxd halted, need %s.\n",