Re: [RFC/RFT] rtl8187: Improve wireless statistics for RTL8187B

[Herton Ronaldo Krzesinski <herton@mandriva.com.br>]

Em Tuesday 22 July 2008 19:25:20 Hin-Tak Leung escreveu:
> On Tue, Jul 22, 2008 at 8:13 PM, Herton Ronaldo Krzesinski
> <herton@mandriva.com.br> wrote:
> <snipped>
>
> > I got some time to take at a look at it again, I did the bisect and found
> > commit 8f87dd7e540d455f8e7f11478133b85edc969c67 ("mac80211: power
> > management wext hooks") to cause the issue here (non working interface).
> > But with it, making "iwconfig wlan* power off" and making interface go up
> > again, it starts to work. May be something related to the power
> > management needs to be implemented in driver now? or keep it disabled
> > with rtl8187...
>
> Disabling power management doesn't improve the situation with latest
> wireless testing
> for me. I had a look at the change and it seems simple enough.

Yes, I also looked at the commit, it's simple and I can't see how it's 
affecting transmission in my case. It's bizarre and I must be doing something 
wrong.

>
> The latest vendor driver has a lot of suspend/resume code in it. I
> have had a go at
> trying to get it running against a current kernel. modprobe works but
> ifconfig up results in a lot
>
> rtl8187: EE:cannot submit RX command. URB_STATUS 0
>
> Hin-Tak

-- 
[]'s
Herton