Messages in this thread | | | Date | Tue, 29 Jan 2008 22:54:46 +0530 | From | "Abhishek Sagar" <> | Subject | Re: [PATCH 0/3][RFC] x86: Catch stray non-kprobe breakpoints |
| |
On 1/29/08, Ananth N Mavinakayanahalli <ananth@in.ibm.com> wrote: > > May be I'm completely off the mark here, but shouldn't a small subset > > of this section simply be 'breakpoint-free' rather than 'kprobe-free'? > > Placing a breakpoint on kprobe_handler (say) can loop into a recursive > > trap without allowing the debugger's notifier chain to be invoked. > > A well heeled debugger will necessarily take care of saving contexts > (using techniques like setjmp/longjmp, etc) to help it recover from such > nested cases (See xmon for example).
Ok, but the protection/warning is not just for xmon.
> > I'm assuming that non-kprobe exception notifiers may (or even should) run > > after kprobe's notifier callback (kprobe_exceptions_notify). > > Yes, any such notifier is invoked after kprobe's callback as the kprobe > notifier is always registered with the highest priority.
Ok.
> > The WARN_ON (and not a BUG_ON) will be hit iff: > > (in_kprobes_functions(addr) && !is_jprobe_bkpt(addr)) > > But that still is unneeded dmesg clutter, IMHO.
Ok, a warning in my opinion would've been prudent since I think we cannot guarantee non kprobe breakpoint users (debuggers or anything else) from the recursive trap handling case.
> Ananth
-- Thanks, Abhishek
| |