Re: wireless

On Mon, Jul 23, 2007 at 09:42:43PM -0500, Craig Boston wrote:
On Mon, Jul 23, 2007 at 11:48:21PM +1200, Andrew Thompson wrote:
You can test this WIP.

Hi,

This patch definitely is a big step in the right direction.

The first time I booted with the patch, there were so many debug
messages spamming the console it made the machine completely unusable
(VESA raster console couldn't keep up, heh).

After setting debug.ipw=0 things seem to be working much better. I can
manually associate with an open AP just fine. Don't have any in range
with WEP to try at the moment.

I diffed it straight from my dev box and forgot that I had bumped up the
debugging quite high. You can change ipw_debug near the top of the file
back to zero.

The big news however, which I've been holding my breath about since
skimming the patch this morning but didn't want to be too hopeful until
I actually tried it, is WPA (TKIP) works! That's something that ipw
hasn't supported at all until now, so I'm quite excited about it.

The first time when I ran wpa_supplicant by hand it seemed to work for a
minute, but lost connectivity while dhclient was requesting a lease. It
worked long enough to get a NAK response for an old address, but never
saw any response to discover requests after that. Manually setting an
address didn't help either. ifconfig still showed Associated but no
packets were being received. There was an "ipw0: scan suck" message on
the console right about this time but nothing else to indicate a
problem.

After rebooting and letting it happen automatically with the rc scripts
it seems to be stable now. I'm not 100% sure how often TKIP changes the
key, but I'm certain this has been up and running long enough for it to
have rotate a few times at least.

Will have to do some more testing to see if the stuck scan was a fluke
or if I can reproduce it.

Thanks for the report. The reason I hadnt sent it out until now was that
I am also bumping into these problems on my test system. I have been
sorting out some ndis issues lately but will get back to testing ipw
this week.


cheers,
Andrew
_______________________________________________
freebsd-current@xxxxxxxxxxx mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscribe@xxxxxxxxxxx"

Relevant Pages

  • Re: Patch to framebuffer
    ... this is my first time posting to this list so please forgive me ... > framebuffer code to modify its behavior a bit. ... there is only 1 active console at one time. ... It will also make your patch smaller, ...
    (Linux-Kernel)
  • Duragesic Patches - Generic Options?
    ... This is the first time that I have ever posted to an NG ... the patch. ... with Mylan brand because my insurance would not pay for Janssen's ... I have noticed that the Mylans are not working nearly as ...
    (alt.support.chronic-pain)
  • Re: Which Mac soundcard works with the Mute key?
    ... but the M-Audio 2496 ignores it for ... What am I doing that I risk my speakers? ... patch 34 sounds normal, patch 35 tears the walls off. ... Sure, it's a problem only the first time, but so far I had, like, a ...
    (rec.audio.pro)
  • Re: wireless
    ... This patch definitely is a big step in the right direction. ... The first time I booted with the patch, ... messages spamming the console it made the machine completely unusable ... (VESA raster console couldn't keep up, ...
    (freebsd-current)
  • [patch 1/1] x86_64: remove duplicated sys_time64
    ... I'm resending this patch again for 4th time since it wasn't merged nor it is ... The first time it didn't look right because of Andi looking at an ... fixed if needed by adding a separate __ARCH_WANT_SYS_STIME macro, and defining ... send the line "unsubscribe linux-kernel" in ...
    (Linux-Kernel)