Re: [PATCH 1/1] ath5k: fix hw rate index condition

From: Bob Copeland <me@bobcopeland.com>
To: Dhaval Giani <dhaval@linux.vnet.ibm.com>
Cc: Jiri Slaby <jirislaby@gmail.com>,
	Dhaval Giani <dhaval.lists@thegianis.in>,
	linville@tuxdriver.com, davem@davemloft.net,
	linux-wireless@vger.kernel.org, ath5k-devel@venema.h4ckr.net,
	Nick Kossifidis <mickflemm@gmail.com>,
	"Luis R. Rodriguez" <lrodriguez@atheros.com>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 1/1] ath5k: fix hw rate index condition
Date: Tue, 31 Mar 2009 08:23:48 -0400	[thread overview]
Message-ID: <20090331122348.GA31324@hash.localnet> (raw)
In-Reply-To: <20090331035140.GC4199@linux.vnet.ibm.com>

On Tue, Mar 31, 2009 at 09:21:40AM +0530, Dhaval Giani wrote:
> Am not sure what the PID controller is, and google gave me a number of
> results, which did not make too much sense in the context.

CONFIG_MAC80211_RC_PID -- unfortunately I recall having to jump through
a few config hoops to enable it.

> One way I have found of reproducing it is to connect to open networks,
> but it does not happen always. At home, when my network is set to open,
> I do not see this issue, whereas at the airport, kaboom.

Ok - that is a useful data point.  Perhaps something to do with the rates
the peer supports; it would help if you could grab a scan next time you
are in the area.  Turn off auto-connect to open networks, then do:

 # iw dev wlan0 scan trigger
 # iw dev wlan0 scan dump >> dump.out    # do this a few times

Then if a particular peer triggers the problem, we can look at the 
advertised rates to see if anything jumps out.

-- 
Bob Copeland %% www.bobcopeland.com