linux-mips
[Top] [All Lists]

Re: [PATCH 1/8] MIPS: Alchemy: abstract USB block control register acces

To: Manuel Lauss <manuel.lauss@googlemail.com>
Subject: Re: [PATCH 1/8] MIPS: Alchemy: abstract USB block control register access
From: Greg KH <gregkh@suse.de>
Date: Fri, 12 Aug 2011 09:33:25 -0700
Cc: Linux-MIPS <linux-mips@linux-mips.org>, Ralf Baechle <ralf@linux-mips.org>, linux-usb@vger.kernel.org
In-reply-to: <CAOLZvyFNoxjkU9bcdRtQA9Ey2wnsEN0MbRL3vmNeYhJZHMWoVQ@mail.gmail.com>
References: <1313141985-5830-1-git-send-email-manuel.lauss@googlemail.com> <1313141985-5830-2-git-send-email-manuel.lauss@googlemail.com> <20110812160454.GA11898@suse.de> <CAOLZvyFNoxjkU9bcdRtQA9Ey2wnsEN0MbRL3vmNeYhJZHMWoVQ@mail.gmail.com>
Sender: linux-mips-bounce@linux-mips.org
User-agent: Mutt/1.5.21 (2010-09-15)
On Fri, Aug 12, 2011 at 06:20:48PM +0200, Manuel Lauss wrote:
> On Fri, Aug 12, 2011 at 6:04 PM, Greg KH <gregkh@suse.de> wrote:
> > On Fri, Aug 12, 2011 at 11:39:38AM +0200, Manuel Lauss wrote:
> >> Alchemy chips have one or more registers which control access
> >> to the usb blocks as well as PHY configuration.  I don't want
> >> the OHCI/EHCI glues to know about the different registers and bits;
> >> new arch code hides the gory details of USB configuration from them.
> >>
> >> Cc: linux-usb@vger.kernel.org
> >> Cc: Greg Kroah-Hartman <gregkh@suse.de>  (USB glue parts)
> >> Signed-off-by: Manuel Lauss <manuel.lauss@googlemail.com>
> >> ---
> >> CC'ed Greg for an Ack on the USB glue parts.  I'd like this to go through 
> >> the
> >>  MIPS tree since other changes in it depend on it.
> >
> > Fine with me on the USB portions, you are just deleting code, which I
> > like :)
> 
> Thanks!
> 
> 
> > But should the "common" USB code really live under arch/mips/alchemy/ ?
> > The goal is to move driver code out of arch/ and into drivers/.  Why are
> > you moving stuff backwards here?
> 
> This is just chip-dependent code for usb block and phy management, which
> varies wich chip subtype, the OHCI and EHCI controllers are identical on all
> of them.  At the time moving the chip-depedent code to the other 
> chip-dependent
> code seemed like a good idea...

But that's what drivers/* is, chip-dependent code.  Please move it there
instead.

Has the MIPS developers learned nothing from the recent ARM mess? :)

greg k-h

<Prev in Thread] Current Thread [Next in Thread>