Atheros (ath) MSI wireless embedded chipset fails to attach on 7.0-STABLE

[Alexander Sack]

On Tue, Jun 17, 2008 at 3:44 PM, Alexander Sack <pisymbol at gmail.com> wrote:
> On Tue, Jun 17, 2008 at 3:35 PM, Edwin L. Culp <eculp at casasponti.net> wrote:
>> "Manolis Kiagias" <sonic2000gr at gmail.com> escribió:
>>
>>> Edwin L. Culp wrote:
>>>>
>>>> "Alexander Sack" <pisymbol at gmail.com> escribió:
>>>>
>>>>> On Tue, Jun 17, 2008 at 11:31 AM, Manolis Kiagias
>>>>> <sonic2000gr at gmail.com> wrote:
>>>>>>
>>>>>> Alexander Sack wrote:
>>>>>>>
>>>>>>> Hello:
>>>>>>>
>>>>>>> I have installed FreeBSD-7.0-amd64 stable on my new AMD X2 Turon based
>>>>>>> notebook, a MSI-1710A (GX710Ax) which has a generic embedded
>>>>>>> controller.  During boot up I notice that ATH complains with:
>>>>>>>
>>>>>>> ath_rate: version 1.2 <SampleRate bit-rate selection algorithm>
>>>>>>> ath_hal: 0.9.20.3 (AR5210, AR5211, AR5212, RF5111, RF5112, RF2413,
>>>>>>> RF5413)
>>>>>>> ath0: <Atheros 5424/2424> mem 0xfd7f0000-0xfd7fffff irq 16 at device
>>>>>>> 0.0
>>>>>>> on pci2
>>>>>>> ath0: Reserved 0x10000 bytes for rid 0x10 type 3 at 0xfd7f0000
>>>>>>> ath0: [MPSAFE]
>>>>>>> ath0: [ITHREAD]
>>>>>>> ath0: unable to attach hardware; HAL status 13
>>>>>>> device_attach: ath0 attach returned 6
>>>>>>>
>>>>>>> HAL status 13 from the header file seems to indicate that the
>>>>>>> 7.0-STABLE driver doesn't support my hardware revision.  Here is my
>>>>>>> pciconf -l output:
>>>>>>>
>>>>>>
>>>>>> Maybe you could try compiling a kernel with a newer hal. This is the
>>>>>> kind of
>>>>>> hack we use on the eeepc. Have a look at this:
>>>>>>
>>>>>> http://nighthack.org/wiki/EeeBSD
>>>>>
>>>>> Thank you SO much for this link.  That's EXACTLY what I want to do
>>>>> because I realize that this is a HAL problem.  I've been searching
>>>>> like MAD where I could get an updated binary HAL for this chipset
>>>>> (PCIe based).
>>>>
>>>> That makes two of us ;)
>>>>
>>>> My dmesg is very, very similar to yours and hoped that this would work.
>>>>
>>>> ath0: <Atheros 5424/2424> mem 0xf2200000-0xf220ffff irq 19 at device 0.0
>>>> on pci5
>>>> ath0: Reserved 0x10000 bytes for rid 0x10 type 3 at 0xf2200000
>>>> ioapic0: routing intpin 19 (PCI IRQ 19) to vector 64
>>>> ath0: [MPSAFE]
>>>> ath0: [ITHREAD]
>>>> ath0: unable to attach hardware; HAL status 13
>>>> device_attach: ath0 attach returned 6
>>>>
>>>> I followed the instructions from the web page, recompiled and it made no
>>>> difference which really worries me that I must have done something wrong.
>>>>
>>>> cd madwifi-ng-r2756+ar5007/hal
>>>> cp -R * /usr/src/sys/contrib/dev/ath/
>>>>
>>>> I did not erase it previously but  am going to try that.  I made no kern
>>>> configuration changes to find that the hal is from contrib.  Is there
>>>> nothing else I should do?
>>>>
>>>> Thanks,
>>>>
>>>
>>> Well, I have only tested this on the eeepc and can confirm it works.
>>> Maybe different atheros chipset have other problems not directly related
>>> to the hal version.
>>> You do not need to do anything more that what is shown in the page: untar,
>>> replace the existing files, recompile / install kernel, reboot. If you got
>>> no errors during the kernel compilation phase, you can safely assume you did
>>> everything correctly, and the problem lies elsewhere.
>>
>> At least there was a ray of hope for the time it took to compile the kernel.
>
> Ed:
>
> I took recompiled and got the same issue.  If I use the LATEST mad
> distro I get some compile bugs (ath_desc_status was moved into
> ath_desc structure in ah_desc.h) which I can't completely work around
> (apparently the API into the HAL has changed as well).  What I'm
> trying to do is look at the Linux driver and understand the newer API
> in order to get past this compile issue and see if this works.
> Otherwise I believe we are SOL.
>
> Does anyone know if the CURRENT contains an updated ath HAL AND driver
> for support of newer PCIe based chipsets?
>
> If I get it to work I will let you know...
>

Ok the trick is not to get it from the madfi project.  Get it from the
author directly!

If you grab:

http://people.freebsd.org/~sam/ath_hal-20080528.tgz

Copy the contents into the src/sys/contrib/dev/ath/* and recompile,
you should now see ath attach properly to the your NIC card.  Thanks
go to my friend jkim for pointing this out since he has a similar
notebook/chipset and runs CURRENT successfully.  I tried using CURRENT
ath but there is to much vap support in it and it turned out the
7.0-RELEASE driver works.

Now ath attaches properly and I'm going to test it out!  (this is at
least much further than a bad attach status code from the HAL).

Let me know how it goes,

-aps