[PATCH v3 3/3] arm64: kprobe: disable probe of fault prone ptrauth instruction

[Mark Rutland]

On Mon, Jun 22, 2020 at 06:13:46PM +0100, Mark Brown wrote:
> On Mon, Jun 22, 2020 at 05:57:49PM +0100, Mark Rutland wrote:
> > On Mon, Jun 22, 2020 at 03:40:26PM +0100, Dave Martin wrote:
> > > On Thu, Jun 18, 2020 at 10:40:29AM +0530, Amit Daniel Kachhap wrote:
> 
> > > > -		 * The HINT instruction is is problematic when single-stepping,
> > > > +		 * The HINT instruction is problematic when single-stepping,
> 
> > > Nit: doesn't matter too much, but ideally this should be a separate
> > > patch (or just don't bother).
> 
> > Agreed. I also think if we change this at all we should drop the comment
> > entirely: it's somewhat misleading given it implies NOP is the only HINT
> > space instruction that can be sfely stepped, and doesn't suggest *why*
> > hints might be proboematic.
> 
> Given the rename of _is_nop() to _is_steppable_hint() we should probably
> push an explanation to _is_steppable_hint() rather than having something
> here but I do think it's worth having something more than just the plain
> code.  I do agree that the current comment really doesn't say much
> though.

That's fair. It's clearly material for a separate patch, regardless!

Mark.