Intel Wi-Fi 6250 (iwldvm) fails to associate with WPA-Enterprise PEAP-MSCHAPv2 network in kernel 3.7

[Anders Kaseorg]

In raring’s kernel 3.7, I can almost never associate with the MIT SECURE wireless network, which is WPA-Enterprise with PEAP-MSCHAPv2. MIT SECURE works with no problems in quantal’s kernel 3.5 and mainline kernel 3.6.

(My home WPA-Personal network with PSK still works in all kernel versions.)

03:00.0 Network controller: Intel Corporation Centrino Advanced-N + WiMAX 6250 (rev 35)

FAILS:
raring 3.7.0-2-generic
raring 3.7.0-3-generic
raring 3.7.0-4-generic
mainline 3.7.0-030700rc1-generic
mainline 3.7.0-030700rc7-generic

WORKS:
quantal 3.5.0-18-generic
mainline 3.6.8-030608-generic

ProblemType: Bug
DistroRelease: Ubuntu 13.04
Package: linux-image-extra-3.7.0-4-generic 3.7.0-4.12
ProcVersionSignature: Ubuntu 3.7.0-4.12-generic 3.7.0-rc7
Uname: Linux 3.7.0-4-generic x86_64
ApportVersion: 2.6.2-0ubuntu5
Architecture: amd64
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: anders 2871 F.... pulseaudio
Date: Wed Nov 28 02:14:24 2012
EcryptfsInUse: Yes
HibernationDevice: RESUME=UUID=fd305e7c-c58c-4061-8105-5cda63c38849
InstallationDate: Installed on 2010-12-05 (723 days ago)
InstallationMedia: Ubuntu 11.04 "Natty Narwhal" - Alpha amd64 (20101202)
MachineType: LENOVO 4313CTO
MarkForUpload: True
ProcFB: 0 inteldrmfb
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.7.0-4-generic root=/dev/mapper/fdisk-ubuntu ro crashkernel=384M-2G:64M,2G-:128M quiet splash vt.handoff=7
RelatedPackageVersions:
 linux-restricted-modules-3.7.0-4-generic N/A
 linux-backports-modules-3.7.0-4-generic N/A
 linux-firmware 1.97
SourcePackage: linux
UpgradeStatus: Upgraded to raring on 2012-11-15 (12 days ago)
dmi.bios.date: 06/05/2012
dmi.bios.vendor: LENOVO
dmi.bios.version: 6MET91WW (1.51 )
dmi.board.name: 4313CTO
dmi.board.vendor: LENOVO
dmi.board.version: Not Available
dmi.chassis.asset.tag: No Asset Information
dmi.chassis.type: 10
dmi.chassis.vendor: LENOVO
dmi.chassis.version: Not Available
dmi.modalias: dmi:bvnLENOVO:bvr6MET91WW(1.51):bd06/05/2012:svnLENOVO:pn4313CTO:pvrThinkPadT510:rvnLENOVO:rn4313CTO:rvrNotAvailable:cvnLENOVO:ct10:cvrNotAvailable:
dmi.product.name: 4313CTO
dmi.product.version: ThinkPad T510
dmi.sys.vendor: LENOVO

[Brad Figg (brad-figg)]

This change was made by a bot.

[Anders Kaseorg (andersk)]

A bisect implicates this bad commit.

commit e7c2f967445dd2041f0f8e3179cca22bb8bb7f79
Author: Tejun Heo <email address hidden>
Date: Tue Aug 21 13:18:24 2012 -0700

    workqueue: use mod_delayed_work() instead of __cancel + queue

    Now that mod_delayed_work() is safe to call from IRQ handlers,
    __cancel_delayed_work() followed by queue_delayed_work() can be
    replaced with mod_delayed_work().

    Most conversions are straight-forward except for the following.

    * net/core/link_watch.c: linkwatch_schedule_work() was doing a quite
      elaborate dancing around its delayed_work. Collapse it such that
      linkwatch_work is queued for immediate execution if LW_URGENT and
      existing timer is kept otherwise.

    Signed-off-by: Tejun Heo <email address hidden>
    Cc: "David S. Miller" <email address hidden>
    Cc: Tomi Valkeinen <email address hidden>

 block/blk-core.c | 6 ++----
 block/blk-throttle.c | 7 +------
 drivers/block/floppy.c | 3 +--
 drivers/infiniband/core/mad.c | 14 +++++---------
 drivers/input/keyboard/qt2160.c | 3 +--
 drivers/input/mouse/synaptics_i2c.c | 7 +------
 net/core/link_watch.c | 21 ++++++---------------
 7 files changed, 17 insertions(+), 44 deletions(-)

[Joseph Salisbury (jsalisbury)]

Hi Anders,

Thanks for performing the bisect and identifying the commit that introduced this regression.

I built a Raring test kernel with that commit reverted. The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1083980/

Can you test that kernel and report back if it resolves this bug or not?

Thanks in advance!

[Anders Kaseorg (andersk)]

Yeah, the test kernel resolves the bug.

[Joseph Salisbury (jsalisbury)]

Did you happen to open an upstream bug report regarding this bug? We can have this commit reverted in Ubuntu, but it's best to notify upstream in case they would like to fix the issue another way.

I can notify upstream if you don't feel comfortable doing it.

Thanks again!

[Anders Kaseorg (andersk)]

I sent http://thread.gmane.org/gmane.linux.kernel.wireless.general/100556 yesterday, but got no responses so far.

[Anders Kaseorg (andersk)]

I got a response with a patch, and confirmed that it fixes the bug.
http://thread.gmane.org/gmane.linux.kernel.wireless.general/100556/focus=1403983

[Anders Kaseorg (andersk)]

This more general patch subsumes the first one:
http://thread.gmane.org/gmane.linux.kernel.wireless.general/100556/focus=1404114

[Anders Kaseorg (andersk)]

Pull request to Linus:
http://thread.gmane.org/gmane.linux.kernel/1404215

[Anders Kaseorg (andersk)]

Fix is in v3.7-rc8.

[Joseph Salisbury (jsalisbury)]

Thanks for the update, Anders! The fix will be pulled into Raring when we rebase to v3.7-rc8. Thanks for the help with this bug.

[Anders Kaseorg (andersk)]

Confirmed fixed in 3.7.0-5-generic.