Re: rtl8187b Problem with tx level

[Greg KH <greg@kroah.com>]

On Mon, Aug 31, 2009 at 04:56:11PM -0700, Luis R. Rodriguez wrote:
> 2009/8/31 Gábor Stefanik <netrolller.3d@gmail.com>:
> > On Tue, Sep 1, 2009 at 1:37 AM, Luis R. Rodriguez<mcgrof@gmail.com> wrote:
> >> 2009/8/31 Gábor Stefanik <netrolller.3d@gmail.com>:
> >>> On Mon, Aug 31, 2009 at 11:10 PM, Larry Finger<Larry.Finger@lwfinger.net> wrote:
> >>>> Hin-Tak Leung wrote:
> >>>>>
> >>>>> I just took a look at the history - there aren't obvious bug fixes, a
> >>>>> bunch of work queue-related changes from you & Johannes; feature-wise,
> >>>>> Larry added LED blinking after 2.6.30, and Herton added rfkill support
> >>>>> quite recently after 2.6.31-rc7 . But no, there isn't any change that
> >>>>> would affect tx level beyond 2.6.31-rc7, I think.
> >>>>
> >>>> Your analysis matches my recollection.
> >>>>
> >>>> Apparently RealTek is supplying a closed-source driver for RTL8187B
> >>>> these days. There are no sources for the B on their web site. In
> >>>> addition, the one that Tobias loaded tainted his kernel.
> >>>>
> >>>> Larry
> >>>
> >>> The driver is not actually closed-source; they just don't provide any
> >>> drivers on their site anymore, instead they send their reference
> >>> drivers to OEMs, expecting the OEMs to release their own versions
> >>> (with sources - most vendors who do release Linux drivers in fact only
> >>> release sources, without binaries). The reason for this appears to be
> >>> that Realtek now encourages vendors to sell RTL81xx wireless chips
> >>> with modified USB IDs. (Or are you seeing a binary-only driver on
> >>> realtek.com.tw? That would definitely be a GPL violation, given that
> >>> they use a modified version of the GPLed libipw stack.)
> >>
> >> Gabor, do you speak to RealkTek folks? It seems what may be best for
> >> OEMs and RealTek themselves is to post drivers for inclusion into
> >> staging first and then for interested parties to port them properly.
> >>
> >>  Luis
> >
> > Not yet, I haven't tried contacting them. I simply based my comments
> > on the fact that OEMs are releasing sources. However, contacting them
> > about work to include their drivers in the kernel is a good idea -
> > apparently there is nothing keeping Realtek from taking community
> > contributions (unlike e.g. Zydas - the vendor driver is based on
> > Andrea Merello's old r8180-sa2400 driver, and was not developed
> > in-house by Realtek).
> 
> If anyone has contacts that does indeed seem reasonable to do, seems
> like it could help them and their customers.

I do have a few contacts there, and have been trying to work with them
to do this, but it is slow going :(

thanks,

greg k-h