Re: [ipwave] Patching Linux Kernel with OCB

[Sara el hamdani <[e-mail]>]

Hello Alex,

Many thanks for sharing your insights.

Indeed, we knew during the hackathon that we should better patch the
kernel in its newest version 5.1.12. However, we couldn't create a new
patch because of the lack of time as mentioned Mr, Nabil. In fact, we did
not have the right adapter for the physical OCB Card we had. So, we spent
the whole first day trying to install the card on the desktop including
creating the adequate antenas manually . Then, we had  in the next day only
six hours to capture IPv6 packets, patch and recompile the kernel ...

We have recommended to the organizers to reserve more time to this truck in
future hackathons. We think also that creating a new patch for the current
linux of the kernel is worth to be the object of an independent truck in
the hackathon.

Concerning the parameters, we have tested the connectivity using ath9k
driver successfully, and we were about to do the same for ath10k but it was
already the presentation time (as we aforementioned, a lack of time).

For the patch creator, we don't know honestly who he is, but we have found
the link mentioned on a previous email of you in the link:
https://drive.google.com/file/d/0BxK6WTQZ97QVWF9tRURjOGhBU2c/view. Thus, we
decided to fix it and to make it work.

Besides, even if we arrived to patch the kernel with ath9k, the report in
the terminal showed that it was not completely successful (not 100%) and it
seems that this is what happens usually. So based on your big experience,
Mr. Alex, in creating and patching the kernel, we would like to know if it
is common for you to patch the kernel successfully 100%?

Le lun. 24 juin 2019 à 00:32, Sara el hamdani <[e-mail]> a
écrit :

> Hello Alex,
>
> Many thanks for sharing your insights.
>
> Indeed, we knew during the hackathon that we should better patch the
> kernel in its newest version 5.1.12. However, we couldn't create a new
> patch because of the lack of time as mentioned Mr, Nabil. In fact, we
> didn't have the right adapter for the physical OCBcard we had. So, we whole
> first day spent the the first day trying to install the card on the desktop
> including creating manually the adequate antenas.  Then, we had  only six
> hours in the next day to capture IPv6 packets patching and recompiling the
> kernel ect.
>
> We have recommended to the organizers the reserve more time to this truck
> in future hackathons. We think also that creating a new patch for the
> current linux of the kernel is worth to be the object of an independent
> truck in the hackathon.
>
> Concerning the parameters, we have tested the connectivity using ath9k
> driver successfully, and we were about to do the same for ath10k but it was
> the already presentation time (as we aforementioned a lack of time).
>
> For the patch creator, we don't know honestly who he is, but we have found
> the link mentioned on a previous email of you in the link:
> https://drive.google.com/file/d/0BxK6WTQZ97QVWF9tRURjOGhBU2c/view. Thus,
> we decided to fix it and to make it work.
>
> Besides, even if we arrived to patch the kernel with ath9k, the report in
> the terminal showed that it was not completely successful (not 100%) and it
> seems that this is what happens usually. So based on your big in you
> experience, Mr. Alex, on creating and patching the kernel, we would know if
> it is common for you to patch the kernel successfully 100%?
>
> Sara
>
>


> Le ven. 21 juin 2019 à 12:41, Nabil Benamar <[e-mail]> a
> écrit :
>
>> Hi Alex,
>>
>> Thank you for your comments.
>> This is exactly what the hackathon participants put in future work!
>>
>> 2 days were not enough to build the testbed and test different drivers
>> and parameters...
>>
>> On Fri, Jun 21, 2019, 14:01 Alexandre Petrescu <
>> [e-mail]> wrote:
>>
>>> Hello Mr. Seur,
>>>
>>> I would like to ask you whether you can apply that patch to ath10k also?
>>> (instead of ath9k).
>>>
>>> If you do, you risk being the first to achieve 1Gbps on OCB, hopefully
>>> with IPv6.
>>>
>>> Alex
>>>
>>> Le 20/06/2019 à 15:25, John Seur a écrit :
>>> > Hello IPWavers,
>>> >
>>> > During the Hackathon@AIS2019
>>> > (https://hackathon.internetsummitafrica.org/), the team in IPWave
>>> track
>>> > (led by Nabil Benamar) was patching the Linux kernel 4.9.61 with OCB
>>> > mode and discovered the following bugs in the patch file
>>> > (https://drive.google.com/file/d/0BxK6WTQZ97QVWF9tRURjOGhBU2c/view).
>>> >    drivers/net/wireless/ath/ath9k/main.c:961:2: error: duplicate case
>>> value
>>> >    case NL80211_IFTYPE_OCB:
>>> >    drivers/net/wireless/ath/ath9k/hw.c:1241:2: error: duplicate case
>>> value
>>> >    case NL80211_IFTYPE_OCB:
>>> >
>>> > The duplicated lines of code were commented out and the patch file
>>> works!
>>> >
>>> > _______________________________________________
>>> > its mailing list
>>> > [e-mail]
>>> > https://www.ietf.org/mailman/listinfo/its
>>> >
>>>
>>> _______________________________________________
>>> its mailing list
>>> [e-mail]
>>> https://www.ietf.org/mailman/listinfo/its
>>>
>> _______________________________________________
>> its mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/its
>>
>
>
> --
> *Best regards*
>
>
> *Sara EL HAMDANI*
> *Phd student -Umi University.*
>


-- 
*Best regards*


*Sara EL HAMDANI*
*Phd student -Umi University.*