[driverloader] Re: Broadcom 14e4:4320, work well with 1.38!

[James Horey]

On Monday 10 November 2003 11:19 am, Jan Slupski wrote:
> On Mon, 10 Nov 2003, Marc Boucher wrote:
> > I believe that we have now found and fixed the problem whereby
> > init consumes 100% of CPU under 2.6.0-test. Please upgrade
> > to version 1.38 (which will appear on our website shortly)
> > and let me know if it works fine or if you still have difficulties.
>
> Yes!
>
> I confirm that init problem is solved with 1.38 driverloader.
>
> Thanks a lot!

I can also confirm that the init problem is gone using 1.38. Woohoo!


> After loading driverloader (or, I think, after sending first packet via
> wireless device after loading driverloader):
>
> I'm observing something that I could name 'remote terminal effect'
>
> - I'm holding one key, and it is written in some batches, eg
>   10 letters, 250ms pause, 10 letter, etc.
> - Any fast typing is just as you would work on a remote terminal (delays)
> - Problem persist even after I unload the driverloader
>
> - If this is not strange enough, I can get rid of this problem
>   by issuing
>    # service network restart
>   (standard RedHat command, it doesn't up wifi interface, only ethernet)
>

I also have this problem. When I manually do an "ifdown eth0" to shutdown the 
ethernet card, all keyboard input becomes incredibly slow. However, if I 
restart ethernet and allow ethernet to fail (because nothing is connected to 
the card or by Ctrl+C), then the keyboard becomes normal again. 

>
> I will try to see what 'service network restart' exactly do, and when
> problem disapears, and report here.

I think "service network restart" is similiar to "/etc/init.d/network 
restart", which can be found on debian systems. It basically just calls "ifup 
eth0". 

-- 
James Horey
Email:jhorey at cs.unm.edu
Webpage:www.cs.unm.edu/~jhorey
Office:FEC 301D