2010-01-14 16:51:40

by Roel Kluin

Subject: [PATCH] sparc: leds_resource.end assigned to itself in clock_board_probe()

These statements appear redundant.

Signed-off-by: Roel Kluin <[email protected]>
---
Is there any reason to assign this to itself?

diff --git a/arch/sparc/kernel/central.c b/arch/sparc/kernel/central.c
index f3b5466..773caf4 100644
--- a/arch/sparc/kernel/central.c
+++ b/arch/sparc/kernel/central.c
@@ -99,7 +99,6 @@ static int __devinit clock_board_probe(struct of_device *op,

p->leds_resource.start = (unsigned long)
(p->clock_regs + CLOCK_CTRL);
- p->leds_resource.end = p->leds_resource.end;
p->leds_resource.name = "leds";

p->leds_pdev.name = "sunfire-clockboard-leds";
@@ -194,7 +193,6 @@ static int __devinit fhc_probe(struct of_device *op,
if (!p->central) {
p->leds_resource.start = (unsigned long)
(p->pregs + FHC_PREGS_CTRL);
- p->leds_resource.end = p->leds_resource.end;
p->leds_resource.name = "leds";

p->leds_pdev.name = "sunfire-fhc-leds";

2010-01-14 17:31:00

by Marc Zyngier

Subject: Re: [PATCH] sparc: leds_resource.end assigned to itself in clock_board_probe()

On Thu, 14 Jan 2010 17:55:55 +0100
Roel Kluin <[email protected]> wrote:

> These statements appear redundant.
>
> Signed-off-by: Roel Kluin <[email protected]>
> ---
> Is there any reason to assign this to itself?
>
> diff --git a/arch/sparc/kernel/central.c b/arch/sparc/kernel/central.c
> index f3b5466..773caf4 100644
> --- a/arch/sparc/kernel/central.c
> +++ b/arch/sparc/kernel/central.c
> @@ -99,7 +99,6 @@ static int __devinit clock_board_probe(struct
> of_device *op,
> p->leds_resource.start = (unsigned long)
> (p->clock_regs + CLOCK_CTRL);
> - p->leds_resource.end = p->leds_resource.end;

I have the feeling that the right fix is to have:

p->leds_resource.end = p->leds_resource.start;

hence describing a 1 byte region.

M.
--
I'm the slime oozin' out from your TV set...

2010-01-14 19:48:55

by Roel Kluin

Subject: Re: [PATCH] sparc: leds_resource.end assigned to itself in clock_board_probe()

It should be a 1 byte region.

Signed-off-by: Roel Kluin <[email protected]>
---
>> Is there any reason to assign this to itself?

>> @@ -99,7 +99,6 @@ static int __devinit clock_board_probe(struct
>> of_device *op,
>> p->leds_resource.start = (unsigned long)
>> (p->clock_regs + CLOCK_CTRL);
>> - p->leds_resource.end = p->leds_resource.end;
>
> I have the feeling that the right fix is to have:
>
> p->leds_resource.end = p->leds_resource.start;
>
> hence describing a 1 byte region.

diff --git a/arch/sparc/kernel/central.c b/arch/sparc/kernel/central.c
index f3b5466..4589ca3 100644
--- a/arch/sparc/kernel/central.c
+++ b/arch/sparc/kernel/central.c
@@ -99,7 +99,7 @@ static int __devinit clock_board_probe(struct of_device *op,

p->leds_resource.start = (unsigned long)
(p->clock_regs + CLOCK_CTRL);
- p->leds_resource.end = p->leds_resource.end;
+ p->leds_resource.end = p->leds_resource.start;
p->leds_resource.name = "leds";

p->leds_pdev.name = "sunfire-clockboard-leds";
@@ -194,7 +194,7 @@ static int __devinit fhc_probe(struct of_device *op,
if (!p->central) {
p->leds_resource.start = (unsigned long)
(p->pregs + FHC_PREGS_CTRL);
- p->leds_resource.end = p->leds_resource.end;
+ p->leds_resource.end = p->leds_resource.start;
p->leds_resource.name = "leds";

p->leds_pdev.name = "sunfire-fhc-leds";

2010-01-15 00:58:39

by David Miller

Subject: Re: [PATCH] sparc: leds_resource.end assigned to itself in clock_board_probe()

From: Roel Kluin <[email protected]>
Date: Thu, 14 Jan 2010 17:55:55 +0100

> These statements appear redundant.
>
> Signed-off-by: Roel Kluin <[email protected]>
> ---
> Is there any reason to assign this to itself?

It's a bug, let me try and figure out what this code is trying
to do.

2010-01-15 09:27:56

by David Miller

Subject: Re: [PATCH] sparc: leds_resource.end assigned to itself in clock_board_probe()

From: Roel Kluin <[email protected]>
Date: Thu, 14 Jan 2010 20:53:14 +0100

> It should be a 1 byte region.
>
> Signed-off-by: Roel Kluin <[email protected]>

Yes, that was the intention.

Applied, thanks Roel!