linux-wireless.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Stefan Monnier <monnier@iro.umontreal.ca>
To: linux-wireless@vger.kernel.org
Subject: b43 terribly slow
Date: Thu, 08 Jan 2009 20:47:51 -0500	[thread overview]
Message-ID: <jwvvdspbjv6.fsf-monnier+gmane.linux.kernel.wireless.general@gnu.org> (raw)

I'm using the b43 driver with a OpenWRT's 2.6.25 kernel (which includes
changes from compat-wireless).  I'm happy to say that it works on my WPA
network, but the bandwidth it offers is cometimes terrible.
(like less than 50kB/s over NFS).

The affected machine (an asus WL-700gE) is not overloaded at all, and
while it serves those files over NFS, `top' says it's 94% idle.

Has anybody seen this or have some idea what might be the trouble?

It's not always like that, so it might be due to interference, but 
I have an old WL-500g right next to this box (i.e. similar b43
hardware, but running a 2.4 kernel with broadcom's proprietary driver),
and I always get around 700kB/s bandwidth, so the problem is likely to
be on the driver side.

This afternoon, the problem lasted for a while, so I experimented a bit.
While investigating, I noticed that a likely culprit is that `iwconfig'
says it's rate is set to 1M.  So I tried to change the rate.

Typical link quality during this test:
Link Quality=70/100  Signal level:-40 dBm  Noise level=-63 dBm
tho I've seen link quality reported as high as 105/100
(and as low as 60/100).

- with a 1M transfer speed, I see a bandwidth around 80kB/s
- setting the rate manually to 2M improved the transfer speed to
  around 140kB/s.
- setting the rate to 5.5M improved the speed to around 280kB/s,
  but with periods where nothing gets through.
- with 11M, the periods where nothing get through are much more
  significant, so connections are becoming very difficult.
- with 54M, same thing.
- if I "iwconfig wlan0 rate auto" it goes back down to 1M fairly quickly.

Any idea what might be going on?  Or what I could do to try and figure
it out?  It seems that the rate choice is a bit conservative (2M works
as well as 1M, just faster), but it's probably not the culprit.
So could someone explain to me why I see these drops in connectivity
even though the WL-500g that sits right next stays connected "full
blast" (at least 500kB/s, so at least 11M I guess)?


        Stefan


                 reply	other threads:[~2009-01-09  1:48 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=jwvvdspbjv6.fsf-monnier+gmane.linux.kernel.wireless.general@gnu.org \
    --to=monnier@iro.umontreal.ca \
    --cc=linux-wireless@vger.kernel.org \
    /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;
as well as URLs for NNTP newsgroup(s).