[PATCH 1/1] arch/unicore32/mm/ioremap.c: convert printk/warn_on to warn(1

[PATCH 1/1] arch/unicore32/mm/ioremap.c: convert printk/warn_on to warn(1

[Fabian Frederick]

+coalesce formats

(This is untested)

Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Fabian Frederick <fabf@skynet.be>
---
 arch/unicore32/mm/ioremap.c | 9 ++-------
 1 file changed, 2 insertions(+), 7 deletions(-)

diff --git a/arch/unicore32/mm/ioremap.c b/arch/unicore32/mm/ioremap.c
index 13068ee..5d9568c 100644
--- a/arch/unicore32/mm/ioremap.c
+++ b/arch/unicore32/mm/ioremap.c
@@ -143,13 +143,8 @@ void __iomem *__uc32_ioremap_pfn_caller(unsigned long pfn,
 	/*
 	 * Don't allow RAM to be mapped
 	 */
-	if (pfn_valid(pfn)) {
-		printk(KERN_WARNING "BUG: Your driver calls ioremap() on\n"
-			"system memory.  This leads to architecturally\n"
-			"unpredictable behaviour, and ioremap() will fail in\n"
-			"the next kernel release. Please fix your driver.\n");
-		WARN_ON(1);
-	}
+	if (pfn_valid(pfn))
+		WARN(1, "BUG: Your driver calls ioremap() on\nsystem memory.  This leads to architecturally\nunpredictable behaviour, and ioremap() will fail in\nthe next kernel release. Please fix your driver.\n");
 
 	type = get_mem_type(mtype);
 	if (!type)
-- 
1.8.4.5

Re: [PATCH 1/1] arch/unicore32/mm/ioremap.c: convert printk/warn_on to warn(1

[Andrew Morton]

On Tue, 6 May 2014 19:07:34 +0200 Fabian Frederick <fabf@skynet.be> wrote:

> +coalesce formats
> 
> (This is untested)
> 
> Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> Signed-off-by: Fabian Frederick <fabf@skynet.be>
> ---
>  arch/unicore32/mm/ioremap.c | 9 ++-------
>  1 file changed, 2 insertions(+), 7 deletions(-)
> 
> diff --git a/arch/unicore32/mm/ioremap.c b/arch/unicore32/mm/ioremap.c
> index 13068ee..5d9568c 100644
> --- a/arch/unicore32/mm/ioremap.c
> +++ b/arch/unicore32/mm/ioremap.c
> @@ -143,13 +143,8 @@ void __iomem *__uc32_ioremap_pfn_caller(unsigned long pfn,
>  	/*
>  	 * Don't allow RAM to be mapped
>  	 */
> -	if (pfn_valid(pfn)) {
> -		printk(KERN_WARNING "BUG: Your driver calls ioremap() on\n"
> -			"system memory.  This leads to architecturally\n"
> -			"unpredictable behaviour, and ioremap() will fail in\n"
> -			"the next kernel release. Please fix your driver.\n");
> -		WARN_ON(1);
> -	}
> +	if (pfn_valid(pfn))
> +		WARN(1, "BUG: Your driver calls ioremap() on\nsystem memory.  This leads to architecturally\nunpredictable behaviour, and ioremap() will fail in\nthe next kernel release. Please fix your driver.\n");
>  

I don't think that was an improvement.  Particularly with all
those \n's in there.

checkpatch is a just a guide.  It can and should be overruled by common
sense and taste.

--- a/arch/unicore32/mm/ioremap.c~arch-unicore32-mm-ioremapc-convert-printk-warn_on-to-warn1-fix
+++ a/arch/unicore32/mm/ioremap.c
@@ -144,7 +144,10 @@ void __iomem *__uc32_ioremap_pfn_caller(
 	 * Don't allow RAM to be mapped
 	 */
 	if (pfn_valid(pfn))
-		WARN(1, "BUG: Your driver calls ioremap() on\nsystem memory.  This leads to architecturally\nunpredictable behaviour, and ioremap() will fail in\nthe next kernel release. Please fix your driver.\n");
+		WARN(1, "BUG: Your driver calls ioremap() on\n"
+			"system memory.  This leads to architecturally\n"
+			"unpredictable behaviour, and ioremap() will fail in\n"
+			"the next kernel release. Please fix your driver.\n");
 
 	type = get_mem_type(mtype);
 	if (!type)
_

回复： Re: [PATCH 1/1] arch/unicore32/mm/ioremap.c: convert printk/warn_on to warn(1

[管雪涛]

----- Andrew Morton <akpm@linux-foundation.org> 写道：
> On Tue, 6 May 2014 19:07:34 +0200 Fabian Frederick <fabf@skynet.be> wrote:
> 
> > +coalesce formats
> > 
> > (This is untested)
> > 
> > Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
> > Cc: Andrew Morton <akpm@linux-foundation.org>
> > Signed-off-by: Fabian Frederick <fabf@skynet.be>
> > ---
> >  arch/unicore32/mm/ioremap.c | 9 ++-------
> >  1 file changed, 2 insertions(+), 7 deletions(-)
> > 
> > diff --git a/arch/unicore32/mm/ioremap.c b/arch/unicore32/mm/ioremap.c
> > index 13068ee..5d9568c 100644
> > --- a/arch/unicore32/mm/ioremap.c
> > +++ b/arch/unicore32/mm/ioremap.c
> > @@ -143,13 +143,8 @@ void __iomem *__uc32_ioremap_pfn_caller(unsigned long pfn,
> >  	/*
> >  	 * Don't allow RAM to be mapped
> >  	 */
> > -	if (pfn_valid(pfn)) {
> > -		printk(KERN_WARNING "BUG: Your driver calls ioremap() on\n"
> > -			"system memory.  This leads to architecturally\n"
> > -			"unpredictable behaviour, and ioremap() will fail in\n"
> > -			"the next kernel release. Please fix your driver.\n");
> > -		WARN_ON(1);
> > -	}
> > +	if (pfn_valid(pfn))
> > +		WARN(1, "BUG: Your driver calls ioremap() on\nsystem memory.  This leads to architecturally\nunpredictable behaviour, and ioremap() will fail in\nthe next kernel release. Please fix your driver.\n");
> >  
> 
> I don't think that was an improvement.  Particularly with all
> those \n's in there.
> 
> checkpatch is a just a guide.  It can and should be overruled by common
> sense and taste.
> 
> --- a/arch/unicore32/mm/ioremap.c~arch-unicore32-mm-ioremapc-convert-printk-warn_on-to-warn1-fix
> +++ a/arch/unicore32/mm/ioremap.c
> @@ -144,7 +144,10 @@ void __iomem *__uc32_ioremap_pfn_caller(
>  	 * Don't allow RAM to be mapped
>  	 */
>  	if (pfn_valid(pfn))
> -		WARN(1, "BUG: Your driver calls ioremap() on\nsystem memory.  This leads to architecturally\nunpredictable behaviour, and ioremap() will fail in\nthe next kernel release. Please fix your driver.\n");
> +		WARN(1, "BUG: Your driver calls ioremap() on\n"
> +			"system memory.  This leads to architecturally\n"
> +			"unpredictable behaviour, and ioremap() will fail in\n"
> +			"the next kernel release. Please fix your driver.\n");

A 'return NULL' is also needed to terminate this call.
Febian, could you submit another version?
Thanks.

>  
>  	type = get_mem_type(mtype);
>  	if (!type)
> _
>