<div style="line-height:1.7;color:#000000;font-size:14px;font-family:Arial"><div>Hi Bjorn,</div><div><br></div><div>I have no environment now, need customer to do this test.</div><div>I have modified usbnet like below:</div><div><div>void usbnet_skb_return (struct usbnet *dev, struct sk_buff *skb)</div><div>{</div><div> int status;</div><div><br></div><div> if (test_bit(EVENT_RX_PAUSED, &dev->flags)) {</div><div> skb_queue_tail(&dev->rxq_pause, skb);</div><div> return;</div><div> } </div><div> <span style="color: rgb(255, 0, 0);"> if (skb->protocol != 8)</span></div><div><span style="color: rgb(255, 0, 0);"> skb->protocol = eth_type_trans (skb, dev->net);</span></div><div><br></div><div> dev->net->stats.rx_packets++;</div><div> dev->net->stats.rx_bytes += skb->len;</div><div><br></div><div> netif_dbg(dev, rx_status, dev->net, "< rx, len %zu, type 0x%x\n",</div><div> skb->len + sizeof (struct ethhdr), skb->protocol);</div><div> memset (skb->cb, 0, sizeof (struct skb_data));</div><div><br></div><div> if (skb_defer_rx_timestamp(skb))</div><div> return;</div><div><br></div><div> status = netif_rx (skb);</div><div> if (status != NET_RX_SUCCESS)</div><div> netif_dbg(dev, rx_err, dev->net,</div><div> "netif_rx status %d\n", status);</div><div>}</div></div><br><div>this can make MC7430 not enter into this logic, and for other modems, the logic is same as original.</div><div><br></div><div>I am not sure whether this is right? still wait customer's fedback.</div><br><br><div style="position:relative;zoom:1">--<br><div>
<div style="FONT-SIZE: 14px">
<div align="left"><strong><font color="#0000ff"><em>Best Regards,</em></font></strong></div>
<div><strong><font color="#0000ff"><em>Dai,Lijin</em></font></strong></div></div></div><div style="clear:both"></div></div><div id="divNeteaseMailCard"></div><br><pre><br>At 2016-04-27 15:54:23, "Bjørn Mork" <bjorn@mork.no> wrote:
>dailijin <dailijin126@126.com> writes:
>
>> For support MC7430, I have refered to
>> https://git.kernel.org/cgit/linux/kernel/git/davem/net-next.git/commit/drivers/net/usb/usbnet.c?id=81e0ce79f2919dbd5f025894d29aa806af8695c7
>> to modify usbnet driver, this will cause my system periodically reboot
>> when plugin ATT Beam modem or Sprint modem which use GobiNet
>> driver. seems there are conflict. Do you know how to fix this issue?
>
>Ouch. I wonder if you can test the attached patch, which also tries to
>capture a warning if it ever hits. Which it really shouldn't... But I
>don't see any other way that patch could cause such problems. Please let
>me know how this goes. If you get a warning then I hope that will
>explain what happens here.
>
>
>Bjørn
</pre></div><br><br><span title="neteasefooter"><p> </p></span>