OSDN Git Service

netvsc: Deal with rescinded channels correctly
authorK. Y. Srinivasan <kys@microsoft.com>
Wed, 19 Apr 2017 20:53:49 +0000 (13:53 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Sat, 24 Mar 2018 10:00:13 +0000 (11:00 +0100)
[ Upstream commit 73e64fa4f417b22d8d5521999a631ced8e2d442e ]

We will not be able to send packets over a channel that has been
rescinded. Make necessary adjustments so we can properly cleanup
even when the channel is rescinded. This issue can be trigerred
in the NIC hot-remove path.

Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <alexander.levin@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/net/hyperv/netvsc.c

index c2ac39a..14f58b6 100644 (file)
@@ -151,6 +151,13 @@ static void netvsc_destroy_buf(struct hv_device *device)
                                       sizeof(struct nvsp_message),
                                       (unsigned long)revoke_packet,
                                       VM_PKT_DATA_INBAND, 0);
+               /* If the failure is because the channel is rescinded;
+                * ignore the failure since we cannot send on a rescinded
+                * channel. This would allow us to properly cleanup
+                * even when the channel is rescinded.
+                */
+               if (device->channel->rescind)
+                       ret = 0;
                /*
                 * If we failed here, we might as well return and
                 * have a leak rather than continue and a bugchk
@@ -211,6 +218,15 @@ static void netvsc_destroy_buf(struct hv_device *device)
                                       sizeof(struct nvsp_message),
                                       (unsigned long)revoke_packet,
                                       VM_PKT_DATA_INBAND, 0);
+
+               /* If the failure is because the channel is rescinded;
+                * ignore the failure since we cannot send on a rescinded
+                * channel. This would allow us to properly cleanup
+                * even when the channel is rescinded.
+                */
+               if (device->channel->rescind)
+                       ret = 0;
+
                /* If we failed here, we might as well return and
                 * have a leak rather than continue and a bugchk
                 */