Re: 2.6.38-rc7, iwlagn warning

[Johannes Berg]

FWIW, I'm changing the subject since "crash" bothers me, it's not even
close to a crash.

On Sat, 2011-03-05 at 23:45 +0000, Mathias BurÃn wrote:

> I got it again, with the newly compiled kernel. However, I've been
> trying for 20 minutes now and when I load iwlagn with the debug
> parameter I can't seem to get the warning. I was monitoring dmesg with
> "watch --interval=1 "dmesg|tail -n40"" but I never saw the warning. As
> soon as I removed the module, and inserted it again without the debug
> option I could reproduce the warning. Here are some dmesg logs:

Curious. 

> no debug, crash 1: http://www.privatepaste.com/download/06539404a8
> no debug, crash 2: http://www.privatepaste.com/download/a6be5f5c81

And I'd prefer if you also stopped using the word crash. Nothing
crashed. All you got was a stack trace because I put a warning there.
Really, not the end of the world, and in fact it probably has very
little or even no effect on functionality.


However ... actually looking at the log files I'm totally confused what
it is that you're doing. Something appears to be continually switching
between interface modes, even using wireless extensions, yet it really
wants to be IBSS all the time. What's causing all this pointless
switching? It'd be useful to figure that out, but I'm not sure how. What
applications are you running that manage wireless? Are you, by any
chance, running wpa_supplicant manually in addition to network-manager?

Irregardless of that though, there's clearly some kind of bug in the
driver since the TX power setting times out, and then as a result we get
the warning.

Would you be willing to enable tracing and capture a trace using
trace-cmd, by running 'trace-cmd record -e mac80211 -e iwlwifi'?

johannes

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/