iwlwifi still hangs, clunky workaround

John W. Linville linville at redhat.com
Fri Apr 27 15:05:15 UTC 2012


On Thu, 2012-04-26 at 21:27 -0400, Paul Wouters wrote:
> I'm still seeing the wifi lockup issues on  3.3.2-1.fc17.x86_64.
> 
> The logs show:
> 
> Apr 26 21:02:53 thinkpad kernel: [262822.181984] iwlwifi 0000:02:00.0: Queue 2 stuck for 2000 ms.
> Apr 26 21:02:53 thinkpad kernel: [262822.181992] iwlwifi 0000:02:00.0: Current S W read_ptr 214 write_ptr 218
> Apr 26 21:02:53 thinkpad kernel: [262822.182046] iwlwifi 0000:02:00.0: Current H W read_ptr 214 write_ptr 218
> Apr 26 21:02:53 thinkpad kernel: [262822.182051] iwlwifi 0000:02:00.0: On demand firmware reload
> Apr 26 21:02:53 thinkpad kernel: [262822.182826] ieee80211 phy0: Hardware restar t was requested
> Apr 26 21:02:53 thinkpad kernel: [262822.182908] iwlwifi 0000:02:00.0: L1 Enable d; Disabling L0S
> Apr 26 21:02:53 thinkpad kernel: [262822.183495] iwlwifi 0000:02:00.0: Radio typ e=0x0-0x3-0x1

This is a firmware failure/crash on the wireless device.  Unfortunately,
we don't have a lot of visibility into that.  The Intel folks will need
to be involved.

I'll copy Wey-yi and Johannes.  Hopefully they can give this some
attention.  You might also open a bug at bugzilla.intellinuxwireless.org
for their convenience.

We've been getting a lot of complaints about iwlwifi stability in
general lately.  We may need to flood their inboxes until they can spare
us some cycles... :-)

> A workaround that makes the wifi bearable (but not pleasant) is:
> 
> modprobe iwlwifi wd_disable=1
> 
> I can still see the failures, but at least it recovers:
> 
> 64 bytes from 8.8.8.8: icmp_req=617 ttl=54 time=42.9 ms
> 64 bytes from 8.8.8.8: icmp_req=618 ttl=54 time=46.1 ms
> 64 bytes from 8.8.8.8: icmp_req=619 ttl=54 time=50.9 ms
> 64 bytes from 8.8.8.8: icmp_req=620 ttl=54 time=46.3 ms
> 64 bytes from 8.8.8.8: icmp_req=621 ttl=54 time=48.1 ms
> 64 bytes from 8.8.8.8: icmp_req=622 ttl=54 time=41.4 ms
> 64 bytes from 8.8.8.8: icmp_req=623 ttl=54 time=46.3 ms
> 64 bytes from 8.8.8.8: icmp_req=625 ttl=54 time=2209 ms
> 64 bytes from 8.8.8.8: icmp_req=626 ttl=54 time=1213 ms
> 64 bytes from 8.8.8.8: icmp_req=627 ttl=54 time=213 ms
> 64 bytes from 8.8.8.8: icmp_req=628 ttl=54 time=44.2 ms
> 64 bytes from 8.8.8.8: icmp_req=629 ttl=54 time=43.3 ms
> 64 bytes from 8.8.8.8: icmp_req=630 ttl=54 time=44.6 ms
> 64 bytes from 8.8.8.8: icmp_req=631 ttl=54 time=45.0 ms
> 64 bytes from 8.8.8.8: icmp_req=632 ttl=54 time=44.3 ms
> 64 bytes from 8.8.8.8: icmp_req=633 ttl=54 time=44.0 ms
> 
> I still get one of these lock ups every 10 to 30 seconds,
> but at least I dont have to throw the hardware switch
> back and forth on the laptop to recover.
> 
> If there is more debug info I can get someone, let me know
> 
> Paul
> -- 
> devel mailing list
> devel at lists.fedoraproject.org
> https://admin.fedoraproject.org/mailman/listinfo/devel

-- 
John W. Linville                The water won't run clean until you get
linville at redhat.com                        the pigs out of the creek.




More information about the devel mailing list