Kernel panic with skb_tstamp_tx and sock_queue_err_skb in backtrace

Issue

• Kernel panic with backtrace similar to the following:

    [exception RIP: tcp_recv_skb+0x6d]
...
 #9 [ffff880061c43c28] tcp_read_sock at ffffffff814d5445
#10 [ffff880061c43c98] xs_tcp_data_ready at ffffffffa03e3c5a [sunrpc]
#11 [ffff880061c43ce8] sock_queue_err_skb at ffffffff8147b8b4
#12 [ffff880061c43d08] skb_tstamp_tx at ffffffff8147da67
#13 [ffff880061c43d38] mlx4_en_free_tx_desc at ffffffffa00d0fd2 [mlx4_en]
#14 [ffff880061c43da8] mlx4_en_poll_tx_cq at ffffffffa00d11bb [mlx4_en]
#15 [ffff880061c43e58] net_rx_action at ffffffff8148ee69
#16 [ffff880061c43eb8] __do_softirq at ffffffff810899aa
#17 [ffff880061c43f38] call_softirq at ffffffff8156591c
#18 [ffff880061c43f50] do_softirq at ffffffff8100e535
#19 [ffff880061c43f70] irq_exit at ffffffff8108963d
#20 [ffff880061c43f80] do_IRQ at ffffffff81566795