Jacob Keller 94877768cf fm10k: wait for queues to drain if stop_hw() fails once
It turns out that sometimes during a reset the Tx queues will be
temporarily stuck longer than .stop_hw() expects. Work around this issue
by attempting to .stop_hw() first. If it tails, wait a number of
attempts until the Tx queues appear to be drained. After this, attempt
stop_hw() again. This ensures that we avoid waiting if we don't need to,
such as during the first initialization of a VF, and give the proper
amount of time necessary to recover from most situations. It is possible
that the hardware is actually stuck. For PFs, this is usually fixed by
a datapath reset. Unfortunately the VF cannot request a similar reset
for itself.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2016-07-20 15:22:12 -07:00
..
2016-07-01 15:31:48 -07:00
2016-06-18 06:04:01 -10:00
2016-05-24 11:00:20 -07:00
2016-05-19 11:27:09 -07:00
2016-05-31 22:17:05 +05:30
2016-06-09 22:38:16 +02:00
2016-05-18 16:38:59 -07:00
2016-06-23 00:29:32 +02:00
2016-06-09 14:28:28 -06:00
2016-05-27 15:26:11 -07:00
2016-04-30 13:56:46 -07:00
2016-06-15 09:51:35 +01:00
2016-05-20 21:26:15 -07:00
2016-05-29 22:32:27 -07:00
2016-05-20 18:33:37 -07:00
2016-06-07 15:56:31 -07:00
2016-04-30 14:15:56 -07:00
2016-07-06 22:32:15 -07:00
2016-05-01 14:11:12 -07:00
2016-05-31 21:25:52 -06:00
2016-07-01 05:32:17 -04:00
2016-05-31 08:20:43 +03:00
2016-05-23 11:18:01 -07:00