Re: [PATCH net-next 2/2] r8152: check the status before submittingrx

[David Miller <davem@davemloft.net>]

From: Hayes Wang <hayeswang@realtek.com>
Date: Mon, 22 Dec 2014 02:53:42 +0000

>  David Miller [mailto:davem@davemloft.net] 
>> Sent: Saturday, December 20, 2014 4:44 AM
> [...]
>> > Don't submit the rx if the device is unplugged, linking down,
>> > or stopped.
>>  ...
>> > @@ -1789,6 +1789,11 @@ int r8152_submit_rx(struct r8152 
>> *tp, struct rx_agg *agg, gfp_t mem_flags)
>> >  {
>> >  	int ret;
>> >  
>> > +	/* The rx would be stopped, so skip submitting */
>> > +	if (test_bit(RTL8152_UNPLUG, &tp->flags) ||
>> > +	    !test_bit(WORK_ENABLE, &tp->flags) || !(tp->speed & LINK_STATUS))
>> > +		return 0;
>> > +
>> 
>> I think netif_carrier_off() should always be true in all three of those
>> situations, and would be a much simpler test than what you've coded
>> here.
> 
> When the device is unplugged or stopped, the linking status
> may be true, so I add additional checks to avoid the submission.
> 
> Besides, in set_carrier() I set netif_carrier_on() after
> ops.enable() to avoid any transmission before I finish
> starting the tx/rx.
> 
> 	tp->rtl_ops.enable(tp);
> 	set_bit(RTL8152_SET_RX_MODE, &tp->flags);
> 	netif_carrier_on(netdev);
> 
> However, the r8152_submit_rx() would be called in ops.enable(),
> and the check of netif_carrier_ok() would be always false. That
> is why I use tp->speed, not netif_carrier_ok(), to check the
> linking stauts.

I stil think your check is way too complicated for this fast path so I
would ask that you arrange things such that the simpler
netif_carrier_off() test works.

Especially because that is what the core networking stack uses
to decide whether to send packets to us as well.