[PATCH] unicore32: Remove unneeded Kconfig entry NO_IOPORT_MAP

[PATCH] unicore32: Remove unneeded Kconfig entry NO_IOPORT_MAP

[Paul Bolle]

Architectures only need a Kconfig entry for NO_IOPORT_MAP if it is
possible that its value will be 'y'. For unicore32 its value will always
be 'n', making it pointless. Remove it.

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
---
Tested by playing with arch/unicore32/configs/unicore32_defconfig. This
patch made no difference whatsoever to the generated .config file.
Please note that it has
    CONFIG_HAS_IOPORT_MAP=y

set after invoking "make oldconfig" both before and after this patch.

 arch/unicore32/Kconfig | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/arch/unicore32/Kconfig b/arch/unicore32/Kconfig
index 928237a7b9ca..2322cc87e7cb 100644
--- a/arch/unicore32/Kconfig
+++ b/arch/unicore32/Kconfig
@@ -27,9 +27,6 @@ config UNICORE32
 config GENERIC_CSUM
 	def_bool y
 
-config NO_IOPORT_MAP
-	bool
-
 config STACKTRACE_SUPPORT
 	def_bool y
 
-- 
1.9.3

Re: [PATCH] unicore32: Remove unneeded Kconfig entry NO_IOPORT_MAP

[Xuetao Guan]

Sorry for late reply.

I checked this config, and it's only used for HAS_IOPORT_MAP in lib/Kconfig
Sure, removing it means no different for .config file.
I think a better way is reserving it or moving it into arch/Kconfig

Cc: linux-arch@vger.kernel.org

Xuetao Guan

----- Paul Bolle <pebolle@tiscali.nl> 写道：
> Architectures only need a Kconfig entry for NO_IOPORT_MAP if it is
> possible that its value will be 'y'. For unicore32 its value will always
> be 'n', making it pointless. Remove it.
> 
> Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
> ---
> Tested by playing with arch/unicore32/configs/unicore32_defconfig. This
> patch made no difference whatsoever to the generated .config file.
> Please note that it has
>     CONFIG_HAS_IOPORT_MAP=y
> 
> set after invoking "make oldconfig" both before and after this patch.
> 
>  arch/unicore32/Kconfig | 3 ---
>  1 file changed, 3 deletions(-)
> 
> diff --git a/arch/unicore32/Kconfig b/arch/unicore32/Kconfig
> index 928237a7b9ca..2322cc87e7cb 100644
> --- a/arch/unicore32/Kconfig
> +++ b/arch/unicore32/Kconfig
> @@ -27,9 +27,6 @@ config UNICORE32
>  config GENERIC_CSUM
>  	def_bool y
>  
> -config NO_IOPORT_MAP
> -	bool
> -
>  config STACKTRACE_SUPPORT
>  	def_bool y
>  
> -- 
> 1.9.3
> 

Re: [PATCH] unicore32: Remove unneeded Kconfig entry NO_IOPORT_MAP

[Paul Bolle]

On Sun, 2014-09-28 at 09:49 +0800, Xuetao Guan wrote:
> Sorry for late reply.

You replied within four days. That's rather quick, actually.

> I checked this config, and it's only used for HAS_IOPORT_MAP in lib/Kconfig
> Sure, removing it means no different for .config file.
> I think a better way is reserving it

I'm not sure I get what you're saying here: what do mean with "reserving
it"?

> or moving it into arch/Kconfig

Without a (treewide) patch I find it hard to say whether that's worth
the trouble. But even if that move would be done, wouldn't it include
dropping this entry from arch/unicore32 anyway?


Paul Bolle

回复： Re: [PATCH] unicore32: Remove unneeded Kconfig entry NO_IOPORT_MAP

[Xuetao Guan]

----- Paul Bolle <pebolle@tiscali.nl> 写道：
> On Sun, 2014-09-28 at 09:49 +0800, Xuetao Guan wrote:
> > Sorry for late reply.
> 
> You replied within four days. That's rather quick, actually.
> 
> > I checked this config, and it's only used for HAS_IOPORT_MAP in lib/Kconfig
> > Sure, removing it means no different for .config file.
> > I think a better way is reserving it
> 
> I'm not sure I get what you're saying here: what do mean with "reserving
> it"?

I mean, it's better to keep the code unchanged.

> 
> > or moving it into arch/Kconfig
> 
> Without a (treewide) patch I find it hard to say whether that's worth
> the trouble. But even if that move would be done, wouldn't it include
> dropping this entry from arch/unicore32 anyway?

IMHO, this config could be removed from kernel code.

> 
> 
> Paul Bolle
>