Bug 43038

Summary: Unable to authenticate to AP since 3.3.1
Product: Networking Reporter: Christian Casteyde (casteyde.christian)
Component: WirelessAssignee: networking_wireless (networking_wireless)
Status: RESOLVED UNREPRODUCIBLE    
Severity: high CC: brittman914, linville, shafi.wireless, v
Priority: P1    
Hardware: All   
OS: Linux   
Kernel Version: 3.3.1 Subsystem:
Regression: Yes Bisected commit-id:
Attachments: Dmesg output

Description Christian Casteyde 2012-04-03 19:46:15 UTC
Acer Aspire 7750G
Slacwkare 64 13.37
Kernel built from vanilla source
ath9k wireless driver
Last known working kernel : 3.3.0

Since kernel 3.3.1, I cannot authenticate anymore to any access point. dmesg output looks like this:

eth0: authenticate with ee:7d:78:eb:fc:64 (try 1)
eth0: authenticate with ee:7d:78:eb:fc:64 (try 2)
eth0: authenticate with ee:7d:78:eb:fc:64 (try 3)
eth0: authentication with ee:7d:78:eb:fc:64 timed out
eth0: authenticate with ee:7d:78:eb:fc:64 (try 1)
eth0: authenticate with ee:7d:78:eb:fc:64 (try 2)
eth0: authenticate with ee:7d:78:eb:fc:64 (try 3)
eth0: authentication with ee:7d:78:eb:fc:64 timed out
eth0: authenticate with ee:7d:78:eb:fc:66 (try 1)
eth0: authenticate with ee:7d:78:eb:fc:66 (try 2)
eth0: authenticate with ee:7d:78:eb:fc:66 (try 3)
eth0: authentication with ee:7d:78:eb:fc:66 timed out
eth0: authenticate with ee:7d:78:eb:fc:66 (try 1)
eth0: authenticate with ee:7d:78:eb:fc:66 (try 2)
eth0: authenticate with ee:7d:78:eb:fc:66 (try 3)
eth0: authentication with ee:7d:78:eb:fc:66 timed out
...
and fills up, as wpa_supplicant tries to authenticates.

The used hardware is a laptop wireless adapter, dmesg says that at boot:

ath: EEPROM regdomain: 0x65
ath: EEPROM indicates we should expect a direct regpair map
ath: Country alpha2 being used: 00
ath: Regpair used: 0x65
ieee80211 phy0: Selected rate control algorithm 'ath9k_rate_control'
Registered led device: ath9k-phy0
ieee80211 phy0: Atheros AR9287 Rev:2 mem=0xffffc90011b20000, irq=17

lspci -v shows this:

03:00.0 Network controller: Atheros Communications Inc. AR9287 Wireless Network Adapter (PCI-Express) (rev 01)
        Subsystem: Foxconn International, Inc. Device e034
        Flags: bus master, fast devsel, latency 0, IRQ 17
        Memory at c0100000 (64-bit, non-prefetchable) [size=64K]
        Capabilities: [40] Power Management version 3
        Capabilities: [50] MSI: Enable- Count=1/1 Maskable- 64bit-
        Capabilities: [60] Express Legacy Endpoint, MSI 00
        Capabilities: [100] Advanced Error Reporting
        Capabilities: [140] Virtual Channel
        Capabilities: [160] Device Serial Number 00-15-17-ff-ff-24-14-12
        Capabilities: [170] Power Budgeting <?>
        Kernel driver in use: ath9k

The strange thing is that scanning indeed works, and I can see many AP (I'm in a town).

This is a regression since 3.3.0 works fine, as I'm reporting from this version.
Comment 1 shafi 2012-04-04 07:29:47 UTC
(In reply to comment #0)
> Acer Aspire 7750G
> Slacwkare 64 13.37
> Kernel built from vanilla source
> ath9k wireless driver
> Last known working kernel : 3.3.0
> 
> Since kernel 3.3.1, I cannot authenticate anymore to any access point. dmesg
> output looks like this:
> 
> eth0: authenticate with ee:7d:78:eb:fc:64 (try 1)
> eth0: authenticate with ee:7d:78:eb:fc:64 (try 2)
> eth0: authenticate with ee:7d:78:eb:fc:64 (try 3)
> eth0: authentication with ee:7d:78:eb:fc:64 timed out
> eth0: authenticate with ee:7d:78:eb:fc:64 (try 1)
> eth0: authenticate with ee:7d:78:eb:fc:64 (try 2)
> eth0: authenticate with ee:7d:78:eb:fc:64 (try 3)
> eth0: authentication with ee:7d:78:eb:fc:64 timed out
> eth0: authenticate with ee:7d:78:eb:fc:66 (try 1)
> eth0: authenticate with ee:7d:78:eb:fc:66 (try 2)
> eth0: authenticate with ee:7d:78:eb:fc:66 (try 3)
> eth0: authentication with ee:7d:78:eb:fc:66 timed out
> eth0: authenticate with ee:7d:78:eb:fc:66 (try 1)
> eth0: authenticate with ee:7d:78:eb:fc:66 (try 2)
> eth0: authenticate with ee:7d:78:eb:fc:66 (try 3)
> eth0: authentication with ee:7d:78:eb:fc:66 timed out
> ...
> and fills up, as wpa_supplicant tries to authenticates.
> 
> The used hardware is a laptop wireless adapter, dmesg says that at boot:
> 
> ath: EEPROM regdomain: 0x65
> ath: EEPROM indicates we should expect a direct regpair map
> ath: Country alpha2 being used: 00
> ath: Regpair used: 0x65
> ieee80211 phy0: Selected rate control algorithm 'ath9k_rate_control'
> Registered led device: ath9k-phy0
> ieee80211 phy0: Atheros AR9287 Rev:2 mem=0xffffc90011b20000, irq=17
> 
> lspci -v shows this:
> 
> 03:00.0 Network controller: Atheros Communications Inc. AR9287 Wireless
> Network
> Adapter (PCI-Express) (rev 01)
>         Subsystem: Foxconn International, Inc. Device e034
>         Flags: bus master, fast devsel, latency 0, IRQ 17
>         Memory at c0100000 (64-bit, non-prefetchable) [size=64K]
>         Capabilities: [40] Power Management version 3
>         Capabilities: [50] MSI: Enable- Count=1/1 Maskable- 64bit-
>         Capabilities: [60] Express Legacy Endpoint, MSI 00
>         Capabilities: [100] Advanced Error Reporting
>         Capabilities: [140] Virtual Channel
>         Capabilities: [160] Device Serial Number 00-15-17-ff-ff-24-14-12
>         Capabilities: [170] Power Budgeting <?>
>         Kernel driver in use: ath9k
> 
> The strange thing is that scanning indeed works, and I can see many AP (I'm
> in
> a town).
> 
> This is a regression since 3.3.0 works fine, as I'm reporting from this
> version.

a similiar thread with a similiar issue is reported, let me try to recreate it
Comment 3 shafi 2012-04-10 10:08:50 UTC
http://comments.gmane.org/gmane.linux.kernel.wireless.general/88543

the above revert patch from Sujith should fix, got to fix unassociated idle power save case
Comment 4 Vadim Dyadkin 2012-04-13 12:19:53 UTC
The same for me. The bug 43077 seems to duplicate this bug.
In my dmesg also there is a warning

Apr  8 23:25:12 kato kernel: wlan0: deauthenticating from c6:8e:37:e4:ab:4a by
local choice (reason=3)
Apr  8 23:25:12 kato kernel: ------------[ cut here ]------------
Apr  8 23:25:12 kato kernel: WARNING: at net/wireless/mlme.c:309
cfg80211_send_auth_timeout+0x6d/0xd0 [cfg80211]()
Apr  8 23:25:12 kato kernel: Hardware name: VPCCW2S1R
Apr  8 23:25:12 kato kernel: Modules linked in: ipt_MASQUERADE iptable_nat
nf_nat xt_recent xt_tcpudp nf_conntrack_i
pv4 nf_defrag_ipv4 xt_state nf_conntrack ipv6 iptable_filter ip_tables x_tables
fuse vboxnetflt(O) vboxnetadp(O) pci
_stub vboxpci(O) vboxdrv(O) nfs auth_rpcgss lockd sunrpc nvidia_bl(O) usbhid
hid uvcvideo videobuf2_vmalloc videobuf
2_memops videobuf2_core videodev v4l2_compat_ioctl32 nvidia(PO) btusb bluetooth
sr_mod snd_hda_codec_hdmi cdrom arc4
 snd_hda_codec_realtek ath9k ath9k_common ath9k_hw ath mac80211 cfg80211
ehci_hcd usbcore snd_hda_intel snd_hda_code
c snd_hwdep snd_pcm snd_page_alloc snd_timer usb_common sdhci_pci i2c_i801
sdhci snd battery video i2c_core psmouse
evdev sony_laptop sky2 mmc_core ac rtc_cmos button backlight rfkill soundcore
unix
Apr  8 23:25:12 kato kernel: Pid: 522, comm: kworker/u:5 Tainted: P           O
3.3.1-gentoo #1
Apr  8 23:25:12 kato kernel: Call Trace:
Apr  8 23:25:12 kato kernel: [<ffffffff8102eb5b>] ?
warn_slowpath_common+0x7b/0xc0
Apr  8 23:25:12 kato kernel: [<ffffffffa01aab0d>] ?
cfg80211_send_auth_timeout+0x6d/0xd0 [cfg80211]
Apr  8 23:25:12 kato kernel: [<ffffffff81056a7e>] ? try_to_wake_up+0x1de/0x290
Apr  8 23:25:12 kato kernel: [<ffffffffa0271e21>] ? ath9k_config+0x281/0x530
[ath9k]
Apr  8 23:25:12 kato kernel: [<ffffffffa01d5858>] ?
ieee80211_probe_auth_done+0xf8/0x110 [mac80211]
Apr  8 23:25:12 kato kernel: [<ffffffff8133d83a>] ?
__mutex_unlock_slowpath+0x3a/0x50
Apr  8 23:25:12 kato kernel: [<ffffffffa01d8f7e>] ?
ieee80211_work_work+0x4ee/0x14f0 [mac80211]
Apr  8 23:25:12 kato kernel: [<ffffffff812a46ae>] ? skb_dequeue+0x5e/0x80
Apr  8 23:25:12 kato kernel: [<ffffffffa01d8a90>] ? free_work+0x20/0x20
[mac80211]
Apr  8 23:25:12 kato kernel: [<ffffffff81043d6f>] ?
process_one_work+0x10f/0x380
Apr  8 23:25:12 kato kernel: [<ffffffff8104667e>] ? worker_thread+0x15e/0x330
Apr  8 23:25:12 kato kernel: [<ffffffff81052280>] ? __wake_up_common+0x50/0x80
Apr  8 23:25:12 kato kernel: [<ffffffff81046520>] ?
manage_workers.clone.22+0x230/0x230
Apr  8 23:25:12 kato kernel: [<ffffffff81046520>] ?
manage_workers.clone.22+0x230/0x230
Apr  8 23:25:12 kato kernel: [<ffffffff8104aace>] ? kthread+0x9e/0xb0
Apr  8 23:25:12 kato kernel: [<ffffffff811a0000>] ?
aead_geniv_alloc+0x2c0/0x350
Apr  8 23:25:12 kato kernel: [<ffffffff81341c14>] ?
kernel_thread_helper+0x4/0x10
Apr  8 23:25:12 kato kernel: [<ffffffff8104aa30>] ?
kthread_freezable_should_stop+0x60/0x60
Apr  8 23:25:12 kato kernel: [<ffffffff81341c10>] ? gs_change+0xb/0xb
Apr  8 23:25:12 kato kernel: 
---[ end trace 3a88f303b66412b6 ]---
Comment 5 Britton Dodd 2012-04-19 23:58:11 UTC
https://bugzilla.redhat.com/show_bug.cgi?id=814011

I added a bug for Fedora 16, however I feel information should go here too
Comment 6 Britton Dodd 2012-04-20 00:00:00 UTC
Created attachment 72994 [details]
Dmesg output
Comment 7 Britton Dodd 2012-04-20 00:01:06 UTC
uname -r 
3.3.1-5.fc16.x86_64
Comment 8 Christian Casteyde 2012-04-22 15:20:40 UTC
I confirm the problem is not present anymore in 3.3.2 (commit was reverted), and is not present in 3.4-rc4.