linux-next: manual merge of the mfd tree with the ux500-core tree

linux-next: manual merge of the mfd tree with the ux500-core tree

[Stephen Rothwell]

Hi Samuel,

Today's linux-next merge of the mfd tree got a conflict in
drivers/mfd/tc35892.c between commit
f4e8afdc7ab1b5a0962be02a9dd15d29a81f4c53 ("mfd/tc35892: rename tc35892
core driver to tc3589x") from the ux500-core tree and commit
9716d7717db1eb7b407e67e576a8ca515836d0da ("mfd: Convert tc35892 to new
irq_ APIs") from the mfd tree.

The former renamed the file modified by the latter.  I have applied the
patch from the mfd tree to the new file drivers/mfd/tc3589x.c (see below).
-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

From: Mark Brown <broonie@opensource.wolfsonmicro.com>
Date: Sun, 12 Dec 2010 12:26:09 +0000
Subject: [PATCH] mfd: Convert tc35892 to new irq_ APIs

The genirq core is being converted to pass struct irq_data to irq_chip
operations rather than an IRQ number. Update the tc35892 driver to the
new APIs.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Rabin Vincent <rabin.vincent@stericsson.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
---
 drivers/mfd/tc3589x.c |    8 ++++----
 1 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/mfd/tc3589x.c b/drivers/mfd/tc3589x.c
index 32291fe..dde2d858 100644
--- a/drivers/mfd/tc3589x.c
+++ b/drivers/mfd/tc3589x.c
@@ -170,15 +170,15 @@ again:
 	return IRQ_HANDLED;
 }
 
-static void tc3589x_irq_dummy(unsigned int irq)
+static void tc3589x_irq_dummy(struct irq_data *data)
 {
 	/* No mask/unmask at this level */
 }
 
 static struct irq_chip tc3589x_irq_chip = {
-	.name	= "tc3589x",
-	.mask	= tc3589x_irq_dummy,
-	.unmask	= tc3589x_irq_dummy,
+	.name		= "tc3589x",
+	.irq_mask	= tc3589x_irq_dummy,
+	.irq_unmask	= tc3589x_irq_dummy,
 };
 
 static int tc3589x_irq_init(struct tc3589x *tc3589x)
-- 
1.7.2.3

Re: linux-next: manual merge of the mfd tree with the ux500-core tree

[Paul Mundt]

On Mon, Dec 20, 2010 at 03:21:48PM +1100, Stephen Rothwell wrote:
> The genirq core is being converted to pass struct irq_data to irq_chip
> operations rather than an IRQ number. Update the tc35892 driver to the
> new APIs.
> 
> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
> Acked-by: Rabin Vincent <rabin.vincent@stericsson.com>
> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>

Sorry to hijack the thread..

> diff --git a/drivers/mfd/tc3589x.c b/drivers/mfd/tc3589x.c
> index 32291fe..dde2d858 100644
> --- a/drivers/mfd/tc3589x.c
> +++ b/drivers/mfd/tc3589x.c
> @@ -170,15 +170,15 @@ again:
>  	return IRQ_HANDLED;
>  }
>  
> -static void tc3589x_irq_dummy(unsigned int irq)
> +static void tc3589x_irq_dummy(struct irq_data *data)
>  {
>  	/* No mask/unmask at this level */
>  }
>  
>  static struct irq_chip tc3589x_irq_chip = {
> -	.name	= "tc3589x",
> -	.mask	= tc3589x_irq_dummy,
> -	.unmask	= tc3589x_irq_dummy,
> +	.name		= "tc3589x",
> +	.irq_mask	= tc3589x_irq_dummy,
> +	.irq_unmask	= tc3589x_irq_dummy,
>  };
>  
>  static int tc3589x_irq_init(struct tc3589x *tc3589x)

Is there some reason you don't just kill this off and use dummy_irq_chip?

Re: linux-next: manual merge of the mfd tree with the ux500-core tree

[Samuel Ortiz]

On Mon, Dec 20, 2010 at 01:23:44PM +0900, Paul Mundt wrote:
> On Mon, Dec 20, 2010 at 03:21:48PM +1100, Stephen Rothwell wrote:
> > The genirq core is being converted to pass struct irq_data to irq_chip
> > operations rather than an IRQ number. Update the tc35892 driver to the
> > new APIs.
> > 
> > Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
> > Acked-by: Rabin Vincent <rabin.vincent@stericsson.com>
> > Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
> 
> Sorry to hijack the thread..
> 
> > diff --git a/drivers/mfd/tc3589x.c b/drivers/mfd/tc3589x.c
> > index 32291fe..dde2d858 100644
> > --- a/drivers/mfd/tc3589x.c
> > +++ b/drivers/mfd/tc3589x.c
> > @@ -170,15 +170,15 @@ again:
> >  	return IRQ_HANDLED;
> >  }
> >  
> > -static void tc3589x_irq_dummy(unsigned int irq)
> > +static void tc3589x_irq_dummy(struct irq_data *data)
> >  {
> >  	/* No mask/unmask at this level */
> >  }
> >  
> >  static struct irq_chip tc3589x_irq_chip = {
> > -	.name	= "tc3589x",
> > -	.mask	= tc3589x_irq_dummy,
> > -	.unmask	= tc3589x_irq_dummy,
> > +	.name		= "tc3589x",
> > +	.irq_mask	= tc3589x_irq_dummy,
> > +	.irq_unmask	= tc3589x_irq_dummy,
> >  };
> >  
> >  static int tc3589x_irq_init(struct tc3589x *tc3589x)
> 
> Is there some reason you don't just kill this off and use dummy_irq_chip?
Good point.
I removed this patch from my branch and sent a fix to LinusW so that he
carries it from his ux500-core branch.

Cheers,
Samuel.

-- 
Intel Open Source Technology Centre
http://oss.intel.com/

Re: linux-next: manual merge of the mfd tree with the ux500-core tree

[Mark Brown]

On Mon, Dec 20, 2010 at 01:23:44PM +0900, Paul Mundt wrote:
> On Mon, Dec 20, 2010 at 03:21:48PM +1100, Stephen Rothwell wrote:

> > -static void tc3589x_irq_dummy(unsigned int irq)
> > +static void tc3589x_irq_dummy(struct irq_data *data)
> >  {
> >  	/* No mask/unmask at this level */
> >  }
> >  
> >  static struct irq_chip tc3589x_irq_chip = {
> > -	.name	= "tc3589x",
> > -	.mask	= tc3589x_irq_dummy,
> > -	.unmask	= tc3589x_irq_dummy,
> > +	.name		= "tc3589x",
> > +	.irq_mask	= tc3589x_irq_dummy,
> > +	.irq_unmask	= tc3589x_irq_dummy,
> >  };

> >  static int tc3589x_irq_init(struct tc3589x *tc3589x)

> Is there some reason you don't just kill this off and use dummy_irq_chip?

Mainly that I didn't take the time to read through the code and figure
out if it was doing anything interesting with the struct elsewhere
which meant this was actually sensible (one of the other drivers was
doing something surprising).