Re: [PATCH 4/5] USB Kconfig: Source "drivers/usb/gadget/Kconfig" for ARCH=arm

Re: [PATCH 4/5] USB Kconfig: Source "drivers/usb/gadget/Kconfig" for ARCH=arm

[Al Boldi]

From: David Brownell <david-b@pacbell.net>

Source "drivers/usb/gadget/Kconfig" for ARCH=arm.


Cc: David Brownell <david-b@pacbell.net>
Cc: Greg KH <greg@kroah.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Boldi <a1426z@gawab.com>

---

--- 23.a/arch/arm/Kconfig
+++ 23.b/arch/arm/Kconfig
@@ -1049,6 +1049,8 @@ source "drivers/hid/Kconfig"
 
 source "drivers/usb/Kconfig"
 
+source "drivers/usb/gadget/Kconfig"
+
 source "drivers/mmc/Kconfig"
 
 source "drivers/rtc/Kconfig"

Re: [PATCH 4/5] USB Kconfig: Source "drivers/usb/gadget/Kconfig" for ARCH=arm

[David Brownell]

> --- 23.a/arch/arm/Kconfig
> +++ 23.b/arch/arm/Kconfig
> @@ -1049,6 +1049,8 @@ source "drivers/hid/Kconfig"
>  
>  source "drivers/usb/Kconfig"
>  
> +source "drivers/usb/gadget/Kconfig"
> +
>  source "drivers/mmc/Kconfig"
>  
>  source "drivers/rtc/Kconfig"
> 


Better IMO to include such updates with the patch that reorganizes
that part of the menu ... it's conceptually a single change, and
fragmenting it will cause bisection problems.

In fact, moving the gadget Kconfig in a separate step would make
your main changes more clear too.
 

Can ARM use drivers/Kconfig [Was: Kconfig: Source "drivers/usb/gadget/Kconfig" for ARCH=arm]

[Sam Ravnborg]

On Sat, Jan 05, 2008 at 06:42:52PM +0300, Al Boldi wrote:
> 
> From: David Brownell <david-b@pacbell.net>
> 
> Source "drivers/usb/gadget/Kconfig" for ARCH=arm.
> 
> 
> Cc: David Brownell <david-b@pacbell.net>
> Cc: Greg KH <greg@kroah.com>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> Signed-off-by: Al Boldi <a1426z@gawab.com>
> 
> ---
> 
> --- 23.a/arch/arm/Kconfig
> +++ 23.b/arch/arm/Kconfig
> @@ -1049,6 +1049,8 @@ source "drivers/hid/Kconfig"
>  
>  source "drivers/usb/Kconfig"
>  
> +source "drivers/usb/gadget/Kconfig"
> +
>  source "drivers/mmc/Kconfig"
>  
>  source "drivers/rtc/Kconfig"

What is preventing ARM from using drivers/Kconfig?
That a better fix IMO.

	Sam

Re: Can ARM use drivers/Kconfig [Was: Kconfig: Source "drivers/usb/gadget/Kconfig" for ARCH=arm]

[Russell King]

On Sat, Jan 05, 2008 at 08:20:00PM +0100, Sam Ravnborg wrote:
> On Sat, Jan 05, 2008 at 06:42:52PM +0300, Al Boldi wrote:
> > 
> > From: David Brownell <david-b@pacbell.net>
> > 
> > Source "drivers/usb/gadget/Kconfig" for ARCH=arm.
> > 
> > 
> > Cc: David Brownell <david-b@pacbell.net>
> > Cc: Greg KH <greg@kroah.com>
> > Cc: Andrew Morton <akpm@linux-foundation.org>
> > Signed-off-by: Al Boldi <a1426z@gawab.com>
> > 
> > ---
> > 
> > --- 23.a/arch/arm/Kconfig
> > +++ 23.b/arch/arm/Kconfig
> > @@ -1049,6 +1049,8 @@ source "drivers/hid/Kconfig"
> >  
> >  source "drivers/usb/Kconfig"
> >  
> > +source "drivers/usb/gadget/Kconfig"
> > +
> >  source "drivers/mmc/Kconfig"
> >  
> >  source "drivers/rtc/Kconfig"
> 
> What is preventing ARM from using drivers/Kconfig?
> That a better fix IMO.

Not everything in drivers/ is suitable for every ARM configuration.  It
was felt at the time better for ARM to remain separate because people
didn't want to pollute drivers/Kconfig with the ARM specific conditionals.

(mtd is conditional because it's not supported on ARM CPUs without
alignment trap support.)

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:

Re: Can ARM use drivers/Kconfig [Was: Kconfig: Source "drivers/usb/gadget/Kconfig" for ARCH=arm]

[Sam Ravnborg]

On Sat, Jan 05, 2008 at 08:31:24PM +0000, Russell King wrote:
> On Sat, Jan 05, 2008 at 08:20:00PM +0100, Sam Ravnborg wrote:
> > On Sat, Jan 05, 2008 at 06:42:52PM +0300, Al Boldi wrote:
> > > 
> > > From: David Brownell <david-b@pacbell.net>
> > > 
> > > Source "drivers/usb/gadget/Kconfig" for ARCH=arm.
> > > 
> > > 
> > > Cc: David Brownell <david-b@pacbell.net>
> > > Cc: Greg KH <greg@kroah.com>
> > > Cc: Andrew Morton <akpm@linux-foundation.org>
> > > Signed-off-by: Al Boldi <a1426z@gawab.com>
> > > 
> > > ---
> > > 
> > > --- 23.a/arch/arm/Kconfig
> > > +++ 23.b/arch/arm/Kconfig
> > > @@ -1049,6 +1049,8 @@ source "drivers/hid/Kconfig"
> > >  
> > >  source "drivers/usb/Kconfig"
> > >  
> > > +source "drivers/usb/gadget/Kconfig"
> > > +
> > >  source "drivers/mmc/Kconfig"
> > >  
> > >  source "drivers/rtc/Kconfig"
> > 
> > What is preventing ARM from using drivers/Kconfig?
> > That a better fix IMO.
> 
> Not everything in drivers/ is suitable for every ARM configuration.  It
> was felt at the time better for ARM to remain separate because people
> didn't want to pollute drivers/Kconfig with the ARM specific conditionals.
> 
Today we would have used HAVE_ for the problematic parts and thus
it would have been easy to not select what ARM does not support.
But unfortunately thats not how it got designed.


It would be nice to revist one day to see how much is actually needed
to cleanly supprt ARM in drivers/Kconfig.

	Sam

Re: Can ARM use drivers/Kconfig [Was: Kconfig: Source "drivers/usb/gadget/Kconfig" for ARCH=arm]

[Russell King]

On Sat, Jan 05, 2008 at 09:54:24PM +0100, Sam Ravnborg wrote:
> On Sat, Jan 05, 2008 at 08:31:24PM +0000, Russell King wrote:
> > Not everything in drivers/ is suitable for every ARM configuration.  It
> > was felt at the time better for ARM to remain separate because people
> > didn't want to pollute drivers/Kconfig with the ARM specific conditionals.
> > 
> Today we would have used HAVE_ for the problematic parts and thus
> it would have been easy to not select what ARM does not support.
> But unfortunately thats not how it got designed.
> 
> 
> It would be nice to revist one day to see how much is actually needed
> to cleanly supprt ARM in drivers/Kconfig.

You could diff the two Kconfig files - I try to keep them so they list
the same Kconfigs in the same order to keep the diff size down.

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:

Re: Can ARM use drivers/Kconfig [Was: Kconfig: Source "drivers/usb/gadget/Kconfig" for ARCH=arm]

[Arnd Bergmann]

On Saturday 05 January 2008, Russell King wrote:
> Not everything in drivers/ is suitable for every ARM configuration.  It
> was felt at the time better for ARM to remain separate because people
> didn't want to pollute drivers/Kconfig with the ARM specific conditionals.

We made drivers/Kconfig work for s390 in the meantime, which is far more
special than arm in this regard, so I think you should try to give it
another go.

What are the main obstacles on arm that prevent you from building these
drivers?

	Arnd <><