[rt2x00-users] rt2800 usb crash with hostapd

Gertjan van Wingerde gwingerde at gmail.com
Wed Oct 19 20:43:55 AEDT 2011


Hi Olivier,

On Wed, Oct 12, 2011 at 6:07 PM, Olivier Gandit
<olivier.gandit at gmail.com> wrote:
> Hi rt2x00 users,
>
> I have built the stable release of the rt2x00 drivers for use with
> hostapd on an ARM system with kernel version 2.6.35.5 and a rt2800 usb
> chipset.
>
> Everything starts fine but when a client tries to connect to the
> access point, it fails with the following message:
>
> BUG: scheduling while atomic: hostapd/1089/0x00000002
> Modules linked in: rt2800usb rt2800lib rt2x00usb rt2x00lib mac80211
> cfg80211 kfifo(P) compat lib80211_crypt_wep lib80211_crypt_tkip
> lib80211_crypt_ccmp ppp_async ppp_generic slhc
> [<c002872c>] (unwind_backtrace+0x0/0xf4) from [<c02a95b0>]
> (schedule+0x274/0x34c)
> [<c02a95b0>] (schedule+0x274/0x34c) from [<c02a9dc8>]
> (schedule_timeout+0x13c/0x1d8)
> [<c02a9dc8>] (schedule_timeout+0x13c/0x1d8) from [<c02a9af4>]
> (wait_for_common+0xe8/0x1c4)
> [<c02a9af4>] (wait_for_common+0xe8/0x1c4) from [<c01d6240>]
> (usb_start_wait_urb+0x9c/0xc4)
> [<c01d6240>] (usb_start_wait_urb+0x9c/0xc4) from [<c01d6450>]
> (usb_control_msg+0xd4/0xf8)
> [<c01d6450>] (usb_control_msg+0xd4/0xf8) from [<bf138de0>]
> (rt2x00usb_vendor_request+0xa8/0x130 [rt2x00usb])
> [<bf138de0>] (rt2x00usb_vendor_request+0xa8/0x130 [rt2x00usb]) from
> [<bf138f08>] (rt2x00usb_vendor_req_buff_lock+0x68/0x114 [rt2x00usb])
> [<bf138f08>] (rt2x00usb_vendor_req_buff_lock+0x68/0x114 [rt2x00usb])
> from [<bf13911c>] (rt2x00usb_vendor_request_buff+0x68/0xbc
> [rt2x00usb])
> [<bf13911c>] (rt2x00usb_vendor_request_buff+0x68/0xbc [rt2x00usb])
> from [<bf151534>] (rt2x00usb_register_multiread+0x44/0x4c [rt2800usb])
> [<bf151534>] (rt2x00usb_register_multiread+0x44/0x4c [rt2800usb]) from
> [<bf141148>] (rt2800_get_tkip_seq+0x34/0x4c [rt2800lib])
> [<bf141148>] (rt2800_get_tkip_seq+0x34/0x4c [rt2800lib]) from
> [<bf0f5c74>] (ieee80211_get_key+0x198/0x1e8 [mac80211])
> [<bf0f5c74>] (ieee80211_get_key+0x198/0x1e8 [mac80211]) from
> [<bf0bb79c>] (nl80211_get_key+0x284/0x328 [cfg80211])
> [<bf0bb79c>] (nl80211_get_key+0x284/0x328 [cfg80211]) from
> [<bf0a37ec>] (nl_doit_wrapper+0x130/0x158 [compat])
> [<bf0a37ec>] (nl_doit_wrapper+0x130/0x158 [compat]) from [<c023754c>]
> (genl_rcv_msg+0x1e8/0x210)
> [<c023754c>] (genl_rcv_msg+0x1e8/0x210) from [<c0235bb8>]
> (netlink_rcv_skb+0xbc/0xd8)
> [<c0235bb8>] (netlink_rcv_skb+0xbc/0xd8) from [<c0237358>] (genl_rcv+0x18/0x24)
> [<c0237358>] (genl_rcv+0x18/0x24) from [<c023579c>]
> (netlink_unicast+0x2d0/0x32c)
> [<c023579c>] (netlink_unicast+0x2d0/0x32c) from [<c02360dc>]
> (netlink_sendmsg+0x190/0x270)
> [<c02360dc>] (netlink_sendmsg+0x190/0x270) from [<c021073c>]
> (sock_sendmsg+0x8c/0xb0)
> [<c021073c>] (sock_sendmsg+0x8c/0xb0) from [<c0210c8c>]
> (sys_sendmsg+0x130/0x254)
> [<c0210c8c>] (sys_sendmsg+0x130/0x254) from [<c0022ee0>]
> (ret_fast_syscall+0x0/0x2c)
>
> Does anyone has already seen this?
> Do you know any solution?
>

Am I right in my understanding that you are trying to use TKIP keys?
Can you check what happens if you use CCMP or AES keys?

Regardless it would still be an issue that needs to be fixed, but it
would allow us to get closer to the issue,
and would allow you run without these warnings being generated.

---
Gertjan




More information about the users mailing list