From: Arun Kumar Prakash <redacted>
Commit ec8ee714736e ("ionic: stretch heartbeat detection") notes that
firmware can take more than 2 seconds to service a device or admin
queue request, and is only expected to complete it within the ionic
command timeout. The RDMA admin queue watchdog still uses 2 seconds,
and when it fires it resets the RDMA LIF and every later verb fails
with -ENODEV. Use the ionic command timeout, 5 seconds, instead. The
AH create/destroy busy-wait keeps its fixed 2 second bound.
Signed-off-by: Arun Kumar Prakash <redacted>
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Acked-by: Allen Hubbe <allen.hubbe@amd.com>
Signed-off-by: Nikhil P. Rao <redacted>
---
drivers/infiniband/hw/ionic/ionic_admin.c | 9 ++++++---
1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/drivers/infiniband/hw/ionic/ionic_admin.c b/drivers/infiniband/hw/ionic/ionic_admin.c
index 0ccd869ad8e6..b463ad8b4e80 100644
--- a/drivers/infiniband/hw/ionic/ionic_admin.c
+++ b/drivers/infiniband/hw/ionic/ionic_admin.c
@@ -18,8 +18,6 @@
#define IONIC_ADMIN_BUSY_RETRY_COUNT 2000
#define IONIC_ADMIN_BUSY_RETRY_MS 1
-/* admin queue will be considered failed if a command takes longer */
-#define IONIC_ADMIN_TIMEOUT (HZ * 2)
#define IONIC_ADMIN_WARN (HZ / 8)
/* will poll for admin cq to tolerate and report from missed event */
@@ -28,6 +26,11 @@
/* work queue for polling the event queue and admin cq */
struct workqueue_struct *ionic_evt_workq;
+static unsigned long ionic_admin_timeout(struct ionic_ibdev *dev)
+{
+ return HZ * ionic_cmd_timeout(dev->lif_cfg.lif);
+}
+
static void ionic_admin_timedout(struct ionic_aq *aq)
{
struct ionic_ibdev *dev = aq->dev;@@ -294,7 +297,7 @@ static void ionic_admin_dwork(struct work_struct *ws)
}
if (time_is_after_eq_jiffies(aq->stamp +
- IONIC_ADMIN_TIMEOUT)) {
+ ionic_admin_timeout(dev))) {
/* Timeout threshold not met */
ibdev_dbg(&dev->ibdev, "no progress after %ums\n",
(u32)jiffies_to_msecs(jiffies - aq->stamp));
--
2.43.0