Linux wireless drivers development
 help / color / mirror / Atom feed
From: Dan Williams <dcbw@redhat.com>
To: timg@tpi.com
Cc: linux-wireless@vger.kernel.org
Subject: Re: [PATCH] ipw2200: Change driver default policies
Date: Tue, 14 Oct 2008 12:35:37 -0400	[thread overview]
Message-ID: <1224002137.1799.0.camel@localhost.localdomain> (raw)
In-Reply-To: <48F4C7E8.4020201@tpi.com>

On Tue, 2008-10-14 at 10:25 -0600, Tim Gardner wrote:
> Dan Williams wrote:
> > On Tue, 2008-10-14 at 08:57 -0600, Tim Gardner wrote:
> >> From 41804698de6f2ce121c1452943b9ad2b8a54988b Mon Sep 17 00:00:00 2001
> >> From: Tim Gardner <tim.gardner@canonical.com>
> >> Date: Tue, 14 Oct 2008 08:16:09 -0600
> >> Subject: [PATCH] ipw2200: change default policies for auto-associate, auto-create, and auto-roaming
> >>
> >> We now have applications available to set policy. This patch changes the driver defaults to:
> >>
> >> 1) associate=0 - do not automatically associate to an SSID.
> > 
> > This one is fine.
> > 
> >> 2) auto_create=0 - do not automatically create an ad-hoc network
> > 
> > This will break adhoc network creation, thus NAK.  I think we should
> > just remove the module parameter and default to 1 anyway.  This is the
> > way adhoc _should_ work in the first place, if the adhoc network you're
> > attempting to join does not yet exist, the driver should create that
> > network.  This is the way it works for all other drivers as well.
> > 
> > Disabling this effectively means you can only join existing adhoc
> > networks, not start your own.
> > 
> 
> Agreed - I missed the part where its qualified by IW_MODE_ADHOC.
> 
> >> 3) roaming=0 - do not automatically roam to another AP with the same SSID.
> > 
> > Not really sure why this should be changed?  Were there reported issues
> > with this behavior?
> > 
> 
> Its been my expereince that wireless drivers rarely make the correct
> decision about when or where to roam. It also seems like a policy
> decision that ought to be made elsewhere.
> 
> Frankly, I'm not that concerned about the roaming option. The default
> for 'associate' was the one I _really_ wanted changed. I'll submit a v2
> patch with just that change.

Sure.  That's something I've wanted to turn off for quite a while, it's
sort of a security problem.  Nothing should be auto-associating until
told to do so to a specific SSID or SSID/BSSID combination.

Dan


      parent reply	other threads:[~2008-10-14 16:36 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-10-14 14:57 [PATCH] ipw2200: Change driver default policies Tim Gardner
2008-10-14 15:43 ` Dan Williams
2008-10-14 16:25   ` Tim Gardner
2008-10-14 16:35     ` Johannes Berg
2008-10-14 16:44       ` Tim Gardner
2008-10-14 16:47       ` Dan Williams
2008-10-14 16:35     ` Dan Williams [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1224002137.1799.0.camel@localhost.localdomain \
    --to=dcbw@redhat.com \
    --cc=linux-wireless@vger.kernel.org \
    --cc=timg@tpi.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox