[PATCH v5 05/16] ARM: use a function table for determining instruction interpreter action

[Jon Medhurst (Tixy)]

On Thu, 2014-01-23 at 15:05 -0500, David Long wrote:
> From: "David A. Long" <dave.long at linaro.org>
> 
> Make the instruction interpreter call back to semantic action functions
> through a function pointer array provided by the invoker.  The interpreter
> decodes the instructions into groups and uses the group number to index
> into the supplied array.  kprobes and uprobes code will each supply their
> own array of functions.
> 
> Signed-off-by: David A. Long <dave.long at linaro.org>
> ---

[...]

> --- a/arch/arm/kernel/probes.c
> +++ b/arch/arm/kernel/probes.c
> @@ -378,10 +378,11 @@ static const int decode_struct_sizes[NUM_DECODE_TYPES] = {
>   */
>  int __kprobes
>  kprobe_decode_insn(kprobe_opcode_t insn, struct arch_specific_insn *asi,
> -		   const union decode_item *table, bool thumb)
> +		   const union decode_item *table, bool thumb,
> +		   const union decode_action *actions)
>  {
> -	const struct decode_header *h = (struct decode_header *)table;
> -	const struct decode_header *next;
> +	struct decode_header *h = (struct decode_header *)table;
> +	struct decode_header *next;

The decode tables are fixed structures which nothing should want to
modify, so I think the const's above should be kept. I believe that
you've had to resort to changing them because the following typedef
lacks a 'const' on the final argument....

[...]
> diff --git a/arch/arm/kernel/probes.h b/arch/arm/kernel/probes.h
> index c610fa9..81b6e61 100644
[...]
>  
> +typedef enum kprobe_insn (probes_custom_decode_t)(kprobe_opcode_t,
> +						  struct arch_specific_insn *,
> +						  struct decode_header *);
> +

Adding 'const' above will also have the knock on effect of requiring
const on all the 'custom decode' functions as well.

-- 
Tixy