netmap: Hypervisor kernel panic when ptnet VM sending big traffic with tso/gso enabled

As mentioned in issue #387, the commands to reproduce are:

Hypervisor A:
1. qemu-system-x86_64 xenial.img -enable-kvm -smp 4 -m 4G -vga std -device ptnet-pci,netdev=data10,mac=00:AA:BB:CC:0a:06 -netdev netmap,ifname=vale1:10,id=data10,passthrough=on -vnc 0.0.0.0:1,password -monitor stdio
2. vale-ctl -a vale1:eth0 # with promisc and all offload disabled
3. In side of VM:   
netmao.ko is loaded with default ptnet_vnet_hdr=1
iperf -c 172.10.10.5 -P 2 -i 3 -t 60    # eth0 IP: 172.10.10.20 

Hypervisor B:
iperf -s  # eth0 IP: 172.10.10.5

The first feeling is the transmitting is very slow but indeed it is a kernel panic because hypervisor is dead soon. The attachment is the screenshot but unfortunately I can only get this last screen, the callstack is incomplete, just hope it can help.