Re: x86/kprobes: kretprobe fails to triggered if kprobe at function entry is not optimized (trigger by int3 breakpoint)

[Masami Hiramatsu]

On Tue, 25 Aug 2020 16:10:58 +0200
peterz@xxxxxxxxxxxxx wrote:

> On Tue, Aug 25, 2020 at 03:30:05PM +0200, peterz@xxxxxxxxxxxxx wrote:
> > On Tue, Aug 25, 2020 at 10:15:55PM +0900, Masami Hiramatsu wrote:
> 
> > > OK, this looks good to me too.
> > > I'll make a series to rewrite kretprobe based on this patch, OK?
> > 
> > Please, I'll send the fix along when I have it.
> 
> One approach that I think might work nicely is trying to pull
> trampoline_handler() into core code (with a few arch helpers). Then we
> can replace that loop once, instead of having to go fix each
> architectures one by one.

Yes, that is what I'm trying (I had done it, but lost the code... let's do it again)

Thank you,

> 
> They're all basically the same loop after all.


-- 
Masami Hiramatsu <mhiramat@xxxxxxxxxx>