While I don't necessarily object to this change, it is not necessary
because this function is only called in situations where the link will
not have been made:
* When an adapter or domain is assigned to the vfio_ap mdev, in which
case no queue with the APID of the adaper or the APQI of the domain will
have been linked.
* When a queue device is probed, in which case the vfio_ap_queue object
is created and linked if the its APQN is assigned to the vfio_ap mdev.
In any case, it certainly doesn't hurt and if a future change is made
such that this could come into play, the code is already there. So I'll
leave it up to you if you want to keep this; if so, you already have my r-b.
On 3/6/24 9:08 AM, Jason J. Herne wrote:
vfio_ap_mdev_link_queue is changed to detect if a matrix_mdev has
already linked the given queue. If so, it bails out.
Signed-off-by: Jason J. Herne <jjherne@xxxxxxxxxxxxx>
Reviewed-by: Tony Krowiak <akrowiak@xxxxxxxxxxxxx>
---
drivers/s390/crypto/vfio_ap_ops.c | 9 +++++----
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c
index b1c1dc0233e1..259130347d00 100644
--- a/drivers/s390/crypto/vfio_ap_ops.c
+++ b/drivers/s390/crypto/vfio_ap_ops.c
@@ -781,10 +781,11 @@ static int vfio_ap_mdev_probe(struct mdev_device *mdev)
static void vfio_ap_mdev_link_queue(struct ap_matrix_mdev *matrix_mdev,
struct vfio_ap_queue *q)
{
- if (q) {
- q->matrix_mdev = matrix_mdev;
- hash_add(matrix_mdev->qtable.queues, &q->mdev_qnode, q->apqn);
- }
+ if (!q || vfio_ap_mdev_get_queue(matrix_mdev, q->apqn))
+ return;
+
+ q->matrix_mdev = matrix_mdev;
+ hash_add(matrix_mdev->qtable.queues, &q->mdev_qnode, q->apqn);
}
static void vfio_ap_mdev_link_apqn(struct ap_matrix_mdev *matrix_mdev, int apqn)
