Bug 161293 - [iwn] hang at startup when starting network
Summary: [iwn] hang at startup when starting network
Status: Open
Alias: None
Product: Base System
Classification: Unclassified
Component: wireless (show other bugs)
Version: 9.0-BETA3
Hardware: Any Any
: Normal Affects Only Me
Assignee: freebsd-wireless (Nobody)
URL:
Keywords:
Depends on:
Blocks: 278175
  Show dependency treegraph
 
Reported: 2011-10-05 06:30 UTC by Sebastian Chmielewski
Modified: 2024-04-04 21:13 UTC (History)
0 users

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Sebastian Chmielewski 2011-10-05 06:30:10 UTC
With iwn0 wi-fi driver system hangs at startup, first following line is output in kernel output:

Oct  5 06:46:58 chulak kernel: wlan0: Ethernet address: e8:9a:8f:13:10:96
Oct  5 06:46:58 chulak kernel: iwn0: radio is disabled by hardware switch
Oct  5 06:46:58 chulak kernel: iwn0: radio is disabled by hardware switch
Oct  5 06:46:58 chulak kernel: taskqueue_drainiwn0: radio is disabled by hardware switch

which is *NOT* true - radio switch is enabled

then following LOR is print:
Oct  5 06:46:58 chulak kernel: with the following non-sleepable locks held:
Oct  5 06:46:58 chulak kernel: exclusive rw if_lagg rwlock (if_lagg rwlock) r = 0 (0xfffffe001e5c8808) locked @ /work/opensource/OS/FreeBSD/stable9-gem/sys/modules/if_lagg/../../net/if_lagg.c:1024
Oct  5 06:46:58 chulak kernel: KDB: stack backtrace:
Oct  5 06:46:58 chulak kernel: db_trace_self_wrapper() at db_trace_self_wrapper+0x2a
Oct  5 06:46:58 chulak kernel: kdb_backtrace() at kdb_backtrace+0x37
Oct  5 06:46:58 chulak kernel: _witness_debugger() at _witness_debugger+0x2e
Oct  5 06:46:58 chulak kernel: witness_warn() at witness_warn+0x2c4
Oct  5 06:46:58 chulak kernel: taskqueue_drain() at taskqueue_drain+0xfd
Oct  5 06:46:58 chulak kernel: ieee80211_waitfor_parent() at ieee80211_waitfor_parent+0x2b
Oct  5 06:46:58 chulak kernel: ieee80211_ioctl() at ieee80211_ioctl+0x161
Oct  5 06:46:58 chulak kernel: lagg_port_ioctl() at lagg_port_ioctl+0x1fe
Oct  5 06:46:58 chulak kernel: if_setflag() at if_setflag+0x111
Oct  5 06:46:58 chulak kernel: ifpromisc() at ifpromisc+0x34
Oct  5 06:46:58 chulak kernel: lagg_setflag() at lagg_setflag+0x99
Oct  5 06:46:58 chulak kernel: lagg_setflags() at lagg_setflags+0x43
Oct  5 06:46:58 chulak kernel: lagg_ioctl() at lagg_ioctl+0x879
Oct  5 06:46:58 chulak kernel: if_setflag() at if_setflag+0x111
Oct  5 06:46:58 chulak kernel: ifpromisc() at ifpromisc+0x34
Oct  5 06:46:58 chulak kernel: bridge_ioctl_add() at bridge_ioctl_add+0x591
Oct  5 06:46:58 chulak kernel: bridge_ioctl() at bridge_ioctl+0x25a
Oct  5 06:46:58 chulak kernel: in_control() at in_control+0x1ff
Oct  5 06:46:58 chulak kernel: ifioctl() at ifioctl+0x751
Oct  5 06:46:58 chulak kernel: kern_ioctl() at kern_ioctl+0xcd
Oct  5 06:46:58 chulak kernel: sys_ioctl() at sys_ioctl+0xfd
Oct  5 06:46:58 chulak kernel: amd64_syscall() at amd64_syscall+0x3d8
Oct  5 06:46:58 chulak kernel: Xfast_syscall() at Xfast_syscall+0xf7

My network configuration is following:

ifconfig_re0="ether xxxx up"
ifconfig_iwn0="ether xxxx up"
wlans_iwn0="wlan0"
ifconfig_wlan0="ssid xxx bssid xxxx mode 11g channel 11 country PL WPA"
cloned_interfaces="lagg0 bridge0 tap0"
ifconfig_lagg0="laggproto failover laggport re0 laggport wlan0 DHCP"

ifconfig_bridge0="addm lagg0 addm tap0 up"

The problem didn't occur with BETA1 or BETA2 or with 8.x.
The problem doesn't occur when radio switch is really disabled.

How-To-Repeat: restart the system with radio switch enabled.
Comment 1 Mark Linimon freebsd_committer freebsd_triage 2011-10-10 06:05:55 UTC
Responsible Changed
From-To: freebsd-bugs->freebsd-wireless

Over to maintainer(s).
Comment 2 Eitan Adler freebsd_committer freebsd_triage 2018-05-28 19:45:54 UTC
batch change:

For bugs that match the following
-  Status Is In progress 
AND
- Untouched since 2018-01-01.
AND
- Affects Base System OR Documentation

DO:

Reset to open status.


Note:
I did a quick pass but if you are getting this email it might be worthwhile to double check to see if this bug ought to be closed.