Re: [PATCH 2/2] staging: ath6kl: use native methods from kernel library

[Greg KH <gregkh@suse.de>]

On Tue, Sep 07, 2010 at 07:29:30PM -0700, Vipin Mehta wrote:
> 
> 
> > -----Original Message-----
> > From: Greg KH [mailto:gregkh@suse.de]
> > Sent: Tuesday, September 07, 2010 5:04 PM
> > To: Vipin Mehta
> > Cc: Andy Shevchenko; linux-kernel@vger.kernel.org;
> > devel@driverdev.osuosl.org
> > Subject: Re: [PATCH 2/2] staging: ath6kl: use native methods from kernel
> > library
> > 
> > On Tue, Sep 07, 2010 at 02:18:02PM -0700, Vipin Mehta wrote:
> > > Andy,
> > > Thanks for the patch. Although the driver supports Linux but it has
> > > been architected to support other OSes if desired. The code is,
> > > therefore, organized into generic and OS specific components with the
> > > latter part of the code captured into the os/linux directory. As such,
> > > the patch will be more useful if we can move out the inclusion of
> > > linux specific header files <linux/kernel.h> and <linux/ctype.h> into
> > > a header file in os/linux/include directory.
> > 
> > Ick, no.
> > 
> > The driver is now in the main kernel tree, the "compatiblity" wrappers
> > are to now be removed as the kernel code does not need to run on any
> > other operating systems.
> > 
> > That whole subdir needs to be deleted entirely.
> > 
> > > You can refer to
> > > os/linux/include/osapi_linux.h for example. A macro could be defined
> > > which would use linux specific APIs for Linux and keep the original
> > > definitions for any other OS.
> > 
> > Again, no, this is not ok at all.  If you are going to insist on trying
> > to do this for the driver, I will just delete it from the tree right now
> > as that is now how kernel development happens, sorry.
> 
> Got it. I think we can discard my suggestion in that case. The patch
> looks good otherwise. It is ok from my side and could be pushed in to
> the driver.

Thanks, will do.

greg k-h