[patch] net/input: fix net/rfkill/rfkill-input.c bug on 64-bit systems

[patch] net/input: fix net/rfkill/rfkill-input.c bug on 64-bit systems

[Ingo Molnar]

From: Ingo Molnar <mingo@elte.hu>
Subject: [patch] net/input: fix net/rfkill/rfkill-input.c bug on 64-bit systems

this recent commit:

 commit cf4328cd949c2086091c62c5685f1580fe9b55e4
 Author: Ivo van Doorn <IvDoorn@gmail.com>
 Date:   Mon May 7 00:34:20 2007 -0700

     [NET]: rfkill: add support for input key to control wireless radio

added this 64-bit bug:

        ....
	unsigned int flags;
 
 	spin_lock_irqsave(&task->lock, flags);
        ....

irq 'flags' must be unsigned long, not unsigned int. The -rt tree has 
strict checks about this on 64-bit so this triggered a build failure. 

For -stable too i suspect.

Signed-off-by: Ingo Molnar <mingo@elte.hu>

Index: linux-rt.q/net/rfkill/rfkill-input.c
===================================================================
--- linux-rt.q.orig/net/rfkill/rfkill-input.c
+++ linux-rt.q/net/rfkill/rfkill-input.c
@@ -55,7 +55,7 @@ static void rfkill_task_handler(struct w
 
 static void rfkill_schedule_toggle(struct rfkill_task *task)
 {
-	unsigned int flags;
+	unsigned long flags;
 
 	spin_lock_irqsave(&task->lock, flags);
 

Re: [patch] net/input: fix net/rfkill/rfkill-input.c bug on 64-bit systems

[Dmitry Torokhov]

On 7/12/07, Ingo Molnar <mingo@elte.hu> wrote:
> From: Ingo Molnar <mingo@elte.hu>
> Subject: [patch] net/input: fix net/rfkill/rfkill-input.c bug on 64-bit systems
>
> this recent commit:
>
>  commit cf4328cd949c2086091c62c5685f1580fe9b55e4
>  Author: Ivo van Doorn <IvDoorn@gmail.com>
>  Date:   Mon May 7 00:34:20 2007 -0700
>
>     [NET]: rfkill: add support for input key to control wireless radio
>
> added this 64-bit bug:
>
>        ....
>        unsigned int flags;
>
>        spin_lock_irqsave(&task->lock, flags);
>        ....
>
> irq 'flags' must be unsigned long, not unsigned int. The -rt tree has
> strict checks about this on 64-bit so this triggered a build failure.
>

Yep, my fault.

> For -stable too i suspect.
>

I don't think anyone uses rfkill in mainline yet. People were asking
to add it so that they can develop against it but I guess I should
have waited for the real user before pushing rfkill to David...
ANyway, I see it is being used in wireless-dev so it is not completely
useless ;)

-- 
Dmitry

Re: [patch] net/input: fix net/rfkill/rfkill-input.c bug on 64-bit systems

[David Miller]

From: Ingo Molnar <mingo@elte.hu>
Date: Thu, 12 Jul 2007 14:45:03 +0200

> From: Ingo Molnar <mingo@elte.hu>
> Subject: [patch] net/input: fix net/rfkill/rfkill-input.c bug on 64-bit systems
> 
> this recent commit:
> 
>  commit cf4328cd949c2086091c62c5685f1580fe9b55e4
>  Author: Ivo van Doorn <IvDoorn@gmail.com>
>  Date:   Mon May 7 00:34:20 2007 -0700
> 
>      [NET]: rfkill: add support for input key to control wireless radio
> 
> added this 64-bit bug:
> 
>         ....
> 	unsigned int flags;
>  
>  	spin_lock_irqsave(&task->lock, flags);
>         ....
> 
> irq 'flags' must be unsigned long, not unsigned int. The -rt tree has 
> strict checks about this on 64-bit so this triggered a build failure. 
> 
> For -stable too i suspect.
> 
> Signed-off-by: Ingo Molnar <mingo@elte.hu>

Applied, thanks Ingo.

I'll push this to -stable.