Re: [PATCH RFC V4 09/10] x86/jump_label: use GENERIC_NOP5_ATOMICinstead of jmp5 +0

[Steven Rostedt]

On Fri, 2011-10-14 at 14:53 -0700, H. Peter Anvin wrote:
> On 10/13/2011 11:37 AM, Steven Rostedt wrote:
> > On Thu, 2011-10-13 at 09:57 -0700, Jeremy Fitzhardinge wrote:
> >> On 10/13/2011 08:40 AM, H. Peter Anvin wrote:
> >>> On 10/12/2011 05:08 PM, Jeremy Fitzhardinge wrote:
> >>>> From: Jeremy Fitzhardinge <jeremy.fitzhardinge@xxxxxxxxxx>
> >>>>
> >>>> GENERIC_NOP5_ATOMIC should make a better nop.
> >>>>
> >>> On 32 bits, yes.  On 64 bits you should use P6_NOP5_ATOMIC.
> >>
> >> Something like this?
> >>
> > 
> > No, we only care about 5byte nops here. Look at ideal_nop in
> > alterative.c
> > 
> 
> He needs a compile-time alternative.

I didn't explain this well. I actually just meant something like:

#ifdef CONFIG_X86_64
# define IDEAL_NOP5	P6_NOP5_ATOMIC
#else
# define IDEAL_NOP5	GENERIC_NOP5_ATOMIC
#endif

That is, we know we are using this for a 5byte jump. Just keep it simple
for that. Do you envision needing other nops too? Well, we could do 2
byte nops, but I think we could use the same for both. Can't we?

-- Steve


--
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/