[PATCH] Kprobes for ppc64

[Ananth N Mavinakayanahalli]

On Sat, Oct 23, 2004 at 06:19:00PM +1000, Paul Mackerras wrote:
> Ananth N Mavinakayanahalli writes:
> 
> > 2. arch_prepare_kprobe() now returns an int. I have made the necessary
> >    changes to i386 and sparc64 kprobes files, but is untested.
> 
> Are you going to send this upstream?

Prasanna has a set of changes which he will be pushing to akpm shortly.
This will be part of the set.

> > + * Interrupts are disabled on entry as trap3 is an interrupt gate and they
> > + * remain disabled thorough out this function.
> > + */
> > +static inline int kprobe_handler(struct pt_regs *regs)
> 
> Comments about "trap3" and "interrupt gate" don't help me understand
> this function on ppc64. :)  At present interrupts are enabled in a
> program check exception handler but disabled in a single-step handler.
> When does this function get called?

Ah, I missed the comment .. my bad :(
kprobe_handler() gets invoked from ProgramCheckException(). 
 
> > @@ -96,6 +97,9 @@ int do_page_fault(struct pt_regs *regs, 
> >  	BUG_ON((trap == 0x380) || (trap == 0x480));
> >  
> >  	if (trap == 0x300) {
> > +		if (notify_die(DIE_PAGE_FAULT, "page_fault", regs, error_code,
> > +					11, SIGSEGV) == NOTIFY_STOP)
> > +			return 0;
> 
> Hmmm, this seems a bit heavyweight for adding to the page fault path.
> Have you done any benchmarks with vs. without kprobes?

Hmm no, not yet.
 
> On the whole the patch looks OK.  I haven't checked the kprobe_handler
> code to see if I think it's all SMP- and preempt-safe, but I assume
> you have done it similarly on x86 and checked it there.

Yes - the port is based off the initial x86 code. It is SMP and
preempt safe.

Thanks for your comments! I will rework the patch a bit and post it
soon.

Thanks,
Ananth