On Mon, 2008-10-20 at 02:42 -0600, Otto Solares wrote:
> 
> Kamikaze is in 2.6.25.17

Right.

> but it
> doesn't use the in-kernel mac80211 or b43 driver,

Correct again.

> both are compiled
> "out-of-band" ie they pull from the wireless-tree,

Indeed.

> as I said in the
> other mail b43 is working perfectly in latest Kamikaze,

Not here it isn't anyway, not on a WRT54GS.  I have had others confirm
my findings on the openwrt-users list but nobody claim to have it all
working perfectly fine.

> your problem
> could be something phisical (location?).

So you think that something "environmental" is responsible for the high
softirq that OpenWrt is clocking up?

> First check the antennas and the location,

Are you being serious?  You think that the antennas (i.e. not being
pointed in the right direction?) are causing high amounts of softirqs
and that if I fix the antennas, the softirqs will go away?  Can you
explain the technical details of how this can be?  I'm having a hard
time drawing a conclusion between high softirq counts and antennas.

> if that  doesn't solve it
> we could see other things.

I'm game.  I really doubt antennas, or anything else "physical".  I've
already successfully backported b43 from 2.6.27.2 and it did not solve
the problem.  Unfortunately mac80211 will not backport so easily.

b.

Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
Bcm43xx-dev mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev

Reply via email to