I currently use 2.6.28, and don't yet want to switch to 2.6.29-rc, so I
installed vanilla 2.6.28, and compat-wireless of today on top of it
only options I changed, were I enabled rfkill in iwl3945 and crda, which
I did install.
device doesn't associate with my AP (even in open mode)
nothing suspicions in logs, just says it did associate and then it
dropped.
btw, in last versions of iwwifi.git I noticed that hardware encryption
doesn't work anymore, so I enabled swcrypto, now nether it works
Best regards,
Maxim Levitsky
On Fri, 2009-01-23 at 03:07 +0200, Maxim Levitsky wrote:
> I currently use 2.6.28, and don't yet want to switch to 2.6.29-rc, so I
> installed vanilla 2.6.28, and compat-wireless of today on top of it
> only options I changed, were I enabled rfkill in iwl3945 and crda, which
> I did install.
>
> device doesn't associate with my AP (even in open mode)
>
> nothing suspicions in logs, just says it did associate and then it
> dropped.
>
> btw, in last versions of iwwifi.git I noticed that hardware encryption
> doesn't work anymore, so I enabled swcrypto, now nether it works
>
> Best regards,
> Maxim Levitsky
>
I just installed it again with all default options, and still same thing
I also noticed wierd AP names like:
wlan0: associate with AP ffff88005463a9e0
Best regards,
Maxim Levitsky
On Thu, 2009-01-22 at 20:12 -0600, Larry Finger wrote:
> Maxim Levitsky wrote:
> > I just installed it again with all default options, and still same thing
> > I also noticed wierd AP names like:
> >
> > wlan0: associate with AP ffff88005463a9e0
>
> As you have seen, the commit with the bug that prevents a connection has been found.
>
> The funny printing of MAC addresses is due to the newer code using the %pM formatting that
> is not present in older kernels. You can ignore that.
>
> Larry
Thanks, works fine now
Best regards,
Maxim Levitsky
Maxim Levitsky wrote:
> I just installed it again with all default options, and still same thing
> I also noticed wierd AP names like:
>
> wlan0: associate with AP ffff88005463a9e0
As you have seen, the commit with the bug that prevents a connection has been found.
The funny printing of MAC addresses is due to the newer code using the %pM formatting that
is not present in older kernels. You can ignore that.
Larry
[dropped ipw3945 from cc, not intel specific, and I didn't want the
moderation bounce]
On Fri, Jan 23, 2009 at 1:27 PM, Bob Copeland <[email protected]> wrote:
> On Fri, Jan 23, 2009 at 12:51 PM, Maxim Levitsky
> <[email protected]> wrote:
>> On Fri, 2009-01-23 at 03:07 +0200, Maxim Levitsky wrote:
>>
>>
>> I noticed another wierd bug, somehow iwl3945 manages now to break
>> suspend to disk (hands on suspend) but not suspend to ram.
>
> Probably related to my mac80211 changes :( Can you boot with
> no_console_suspend=1?
Actually... I have a pretty good idea why this happens and a nice one
liner fix. Can you try changing "create_freezeable_workqueue" in
net/mac80211/main.c (there's only one) to "create_singlethread_workqueue"?
--
Bob Copeland %% http://www.bobcopeland.com
On Fri, 2009-01-23 at 17:56 -0500, Bob Copeland wrote:
> [dropped ipw3945 from cc, not intel specific, and I didn't want the
> moderation bounce]
>
> On Fri, Jan 23, 2009 at 1:27 PM, Bob Copeland <[email protected]> wrote:
> > On Fri, Jan 23, 2009 at 12:51 PM, Maxim Levitsky
> > <[email protected]> wrote:
> >> On Fri, 2009-01-23 at 03:07 +0200, Maxim Levitsky wrote:
> >>
> >>
> >> I noticed another wierd bug, somehow iwl3945 manages now to break
> >> suspend to disk (hands on suspend) but not suspend to ram.
> >
> > Probably related to my mac80211 changes :( Can you boot with
> > no_console_suspend=1?
>
> Actually... I have a pretty good idea why this happens and a nice one
> liner fix. Can you try changing "create_freezeable_workqueue" in
> net/mac80211/main.c (there's only one) to "create_singlethread_workqueue"?
>
Done, and thanks, works fine now.
Best regards,
Maxim Levitsky
On Fri, 2009-01-23 at 03:07 +0200, Maxim Levitsky wrote:
I noticed another wierd bug, somehow iwl3945 manages now to break
suspend to disk (hands on suspend) but not suspend to ram.
Also, if I suspend via pm-hibernate on console, it works, but doesn't
work when I use ubuntu's menu.
System hangs on suspend to disk, I still can switch consoles, but can't
type anything there. will need to use ssh I guess.
Maxim Levitsky
On Fri, Jan 23, 2009 at 12:51 PM, Maxim Levitsky
<[email protected]> wrote:
> On Fri, 2009-01-23 at 03:07 +0200, Maxim Levitsky wrote:
>
>
> I noticed another wierd bug, somehow iwl3945 manages now to break
> suspend to disk (hands on suspend) but not suspend to ram.
Probably related to my mac80211 changes :( Can you boot with
no_console_suspend=1?
What about direct "echo -n disk > /sys/power/state" ? I'm not sure what
ubuntu does in the menu, but usually pm-hibernate shuts down interfaces.
--
Bob Copeland %% http://www.bobcopeland.com