[Click] polling extension for madwifi drivers

[Nicola Arnoldi]

The question came from the fact that, while running click on embedded  
systems (small nodes like meraki nodes, or other common accesspoints)  
it took the 99% of the CPU.
I guessed that it was because it issues a lot of interrupts to the  
CPU to poll packets.

I have to point out that I run a sligthly modified version of roofnet  
which uses devices in ad-hoc mode instead of monitoring mode, which  
may be the cause...maybe the roofnet guys can tell us more about that.

Can it be that running in ad-hoc instead of monitoring makes click  
more CPU-hungry???

Nicola


Il giorno 19/gen/07, alle ore 18:49, Eddie Kohler ha scritto:

> (But I could be wrong.)
>
>
> Nicola Arnoldi wrote:
>> Any way to have the device polling capabilities for madwifi drivers?
>> Or have they been already implemented?
>> Il giorno ven, 12/01/2007 alle 11.48 -0700, Qinghua(Kevin) Ye ha
>> scritto:
>>> Hi, All,
>>> We did some work on the extension of polling support on released  
>>> broadcom tg3 driver in Linux 2.6.16.13 source packet, and seems  
>>> work well on our AMD opteron box. Please refer to the attached  
>>> packet. Any comments and improvements are welcome.
>>> Regards,
>>> Qinghua(Kevin) Ye
>>> _______________________________________________ click mailing  
>>> list click at amsterdam.lcs.mit.edu https://amsterdam.lcs.mit.edu/ 
>>> mailman/listinfo/click
>> _______________________________________________
>> click mailing list
>> click at amsterdam.lcs.mit.edu
>> https://amsterdam.lcs.mit.edu/mailman/listinfo/click