Re: [PATCH] Add vt6656 driver to drivers/staging.

[Olivier Blin <blino@mandriva.com>]

Forest Bond <forest@alittletooquiet.net> writes:

> Hi Olivier,
>
> On Mon, Jun 08, 2009 at 07:08:14PM +0200, Olivier Blin wrote:
>> By the way, are you sure about the hook below about wpactl.c in commit
>> 8f2c32b86b3ac16ba4c93bf0c2b766089644d9d1 ?
>> 
>> If I do the same in vt6656, it is freezing the kernel, or causing random
>> crashes. Removing the netdev priv assignation makes the driver ok.
>> I don't think the wpa priv was really useful, but I fail to see where it
>> was used, and why it caused the hang.
>> 
>> Don't you have the same issue with vt6655?
>
> I'm embarrassed to say that I haven't actually tested this change on real
> hardware yet.  I've been trying to make changes that do not introduce any
> difference in functionality, but that fix the compile errors resulting from
> changing kernel APIs.
>
> Let me restate that I'm not really a kernel hacker, so much of my work is
> best-effort material.  TBH, I don't really understand the wpadev stuff, so it's
> not clear to me why wpadev->priv can be removed.  If you have insight here that
> I lack, I'd love to hear it.

Well, same here. My only insight is that I have tested the code on
vt6656 hardware, and that duplicating the device in netdev_priv of
wpadev makes the driver crash the system randomly.

> I should be able to get some real hardware to test on sometime this week.
>
>> BTW, I have some patches to send for vt6656, what is the proper place to
>> submit them since vt6656 is not in staging for now?
>
> I have a similar set of patches for vt6656 mostly prepared that I was going to
> submit to Greg.  Hopefully I can do that soon.  I suspect your patches are
> roughly the same; if they look anything like the vt6655 patches, they probably
> are.

My first patches are the same, and I ended up reusing your tree.
But I have a few more cleanups.

-- 
Olivier Blin (blino) - Mandriva