Re: [PATCH 1/4] move capable() to capability.h

[Tim Schmielau]

On Sat, 7 Jan 2006, Randy.Dunlap wrote:

> (nothing to do with inlining here)
> 
> From: Randy Dunlap <rdunlap@xxxxxxxxxxxx>
> 
> headers + core:
> - Move capable() from sched.h to capability.h;
> - Use <linux/capability.h> where capable() is used
> 	(in include/, block/, ipc/, kernel/, a few drivers/,
> 	mm/, security/, & sound/;
> 	many more drivers/ to go)
> 
> Signed-off-by: Randy Dunlap <rdunlap@xxxxxxxxxxxx>

> --- linux-2615-g3.orig/include/linux/capability.h
> +++ linux-2615-g3/include/linux/capability.h
> @@ -43,6 +43,7 @@ typedef struct __user_cap_data_struct {
>  #ifdef __KERNEL__
>  
>  #include <linux/spinlock.h>
> +#include <asm/current.h>
>  
>  /* #define STRICT_CAP_T_TYPECHECKS */
>  
> @@ -356,6 +357,20 @@ static inline kernel_cap_t cap_invert(ke
>  
>  #define cap_is_fs_cap(c)     (CAP_TO_MASK(c) & CAP_FS_MASK)
>  
> +#ifdef CONFIG_SECURITY
> +/* code is in security.c */
> +extern int capable(int cap);
> +#else
> +static inline int capable(int cap)
> +{
> +	if (cap_raised(current->cap_effective, cap)) {
> +		current->flags |= PF_SUPERPRIV;
> +		return 1;
> +	}
> +	return 0;
> +}
> +#endif

I wonder how this can actually work. For dereferencing current, it is not 
enough to include <asm/current.h>. The actual layout of struct task_struct
needs to be known to the compiler, which is given in <linux/sched.h>.

Maybe you were just lucky with your .config and every file using capable()
just by chance also included <linux/sched.h>?

(Chances are not bad since currently about every other .c file includes 
sched.h. However, I have patches pending to reduce this number to ~500..1000)

Uninlining capable() might indeed help us here.

Tim
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/