2014-06-29 00:19:57

by Sam Edwards

[permalink] [raw]
Subject: BUG: iwlwifi microcode errors on Intel 7260HMW

Hello!

I've been working for a week to track down the cause of this bug.
Sadly, I have no background in wireless development and no knowledge
of the Intel microcode format, so I've gotten stuck.

I'm trying to use an Intel 7260HMW with OpenWrt to act as an AP. I'm
using a recent snapshot ("Barrier Breaker") build of OpenWrt, on the
Linux 3.10.36 kernel. The iwlwifi driver is built from the Linux
driver backports package; a diff reveals that it is fully up-to-date
with the latest source in the upstream kernel repository.

When trying to run hostapd, my kernel log repeats output similar to
this: http://pastebin.com/1UuUuMMg
The problems are not limited to trying to use the device as an AP.
When trying to associate to an existing network, I am given similar
output: http://pastebin.com/qxwbdHyu

The wireless device is, however, capable of scanning for nearby APs,
but so far I have not been able to get it to do anything else.

I've been unable to get the non-backported (i.e. 3.10.36) version of
the driver to load correctly, and I cannot easily test this on the
latest kernel either. However, if I should try a different version,
please let me know.

Thanks,
Sam


2014-06-29 04:55:24

by Emmanuel Grumbach

[permalink] [raw]
Subject: Re: BUG: iwlwifi microcode errors on Intel 7260HMW

> Hello!
>
> I've been working for a week to track down the cause of this bug.
> Sadly, I have no background in wireless development and no knowledge
> of the Intel microcode format, so I've gotten stuck.
>
> I'm trying to use an Intel 7260HMW with OpenWrt to act as an AP. I'm
> using a recent snapshot ("Barrier Breaker") build of OpenWrt, on the
> Linux 3.10.36 kernel. The iwlwifi driver is built from the Linux
> driver backports package; a diff reveals that it is fully up-to-date
> with the latest source in the upstream kernel repository.
>
> When trying to run hostapd, my kernel log repeats output similar to
> this: http://pastebin.com/1UuUuMMg
> The problems are not limited to trying to use the device as an AP.
> When trying to associate to an existing network, I am given similar
> output: http://pastebin.com/qxwbdHyu
>
> The wireless device is, however, capable of scanning for nearby APs,
> but so far I have not been able to get it to do anything else.
>
> I've been unable to get the non-backported (i.e. 3.10.36) version of
> the driver to load correctly, and I cannot easily test this on the
> latest kernel either. However, if I should try a different version,
> please let me know.
>

Please open a bug on bugzilla.kernel.org and copy [email protected]
to the bug.
This is most likely a firmware bug.

Thank you.

2014-06-29 04:57:37

by Sam Edwards

[permalink] [raw]
Subject: Re: BUG: iwlwifi microcode errors on Intel 7260HMW

Will do so immediately, thanks!

On Sat, Jun 28, 2014 at 9:55 PM, Emmanuel Grumbach <[email protected]> wrote:
>> Hello!
>>
>> I've been working for a week to track down the cause of this bug.
>> Sadly, I have no background in wireless development and no knowledge
>> of the Intel microcode format, so I've gotten stuck.
>>
>> I'm trying to use an Intel 7260HMW with OpenWrt to act as an AP. I'm
>> using a recent snapshot ("Barrier Breaker") build of OpenWrt, on the
>> Linux 3.10.36 kernel. The iwlwifi driver is built from the Linux
>> driver backports package; a diff reveals that it is fully up-to-date
>> with the latest source in the upstream kernel repository.
>>
>> When trying to run hostapd, my kernel log repeats output similar to
>> this: http://pastebin.com/1UuUuMMg
>> The problems are not limited to trying to use the device as an AP.
>> When trying to associate to an existing network, I am given similar
>> output: http://pastebin.com/qxwbdHyu
>>
>> The wireless device is, however, capable of scanning for nearby APs,
>> but so far I have not been able to get it to do anything else.
>>
>> I've been unable to get the non-backported (i.e. 3.10.36) version of
>> the driver to load correctly, and I cannot easily test this on the
>> latest kernel either. However, if I should try a different version,
>> please let me know.
>>
>
> Please open a bug on bugzilla.kernel.org and copy [email protected]
> to the bug.
> This is most likely a firmware bug.
>
> Thank you.