Commit 9eedb509 authored Jan 09, 2017 by Tim Sell Committed by Greg Kroah-Hartman Jan 10, 2017

staging: unisys: remove redundant unlocks of visornic_devdata.priv_lock



These redundant unlocks of visornic_devdata.priv_lock would result in
the RHEL 7.2 guests hanging during service partition recovery testing.

__Testing__

* An scp of a large file was started from a remote host TO the RHEL 7.2
  Linux guest.

* During the scp transfer, s-Par service partition recovery was forced
  twice. After each occasion, I verified that the guest recovered
  completely (all s-Par guest devices), and that the file transfer
  resumed.

* Within the RHEL 7.2 guest environment, copied the large file to
  another location in the local filesystem.

* During the copy, s-Par service partition recovery was again forced
  twice. After each occasion, I verified that the guest recovered
  completely (all s-Par guest devices), and that the copy resumed.

* An scp of the new copy of the large file was started FROM the RHEL 7.2
  guest to a remote host.

* During the scp transfer, s-Par service partition recovery was forced
  twice. After each occasion, I verified that the guest recovered
  completely (all s-Par guest devices), and that the file transfer
  resumed.

* Used cmp to verify that the large file had successfully survived the
  round-trip without becoming corrupted.

Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

parent ef44146c

drivers/staging/unisys/visornic/visornic_main.c

+0 −2

Original line number	Diff line number	Diff line
		@@ -461,7 +461,6 @@ visornic_disable_with_timeout(struct net_device *netdev, const int timeout)
		if (devdata->enab_dis_acked)
		break;
		if (devdata->server_down \|\| devdata->server_change_state) {
		spin_unlock_irqrestore(&devdata->priv_lock, flags);
		dev_dbg(&netdev->dev, "%s server went away\n",
		__func__);
		break;
		@@ -605,7 +604,6 @@ visornic_enable_with_timeout(struct net_device *netdev, const int timeout)
		if (devdata->enab_dis_acked)
		break;
		if (devdata->server_down \|\| devdata->server_change_state) {
		spin_unlock_irqrestore(&devdata->priv_lock, flags);
		dev_dbg(&netdev->dev, "%s server went away\n",
		__func__);
		break;

Admin message