Re: [PATCH 0/5][RFC] kprobes/ftrace: Have kprobes use ftrace on ftrace nops

[Steven Rostedt <rostedt@goodmis.org>]

On Sat, 2011-08-13 at 19:09 +0900, Masami Hiramatsu wrote:
> (2011/08/12 22:08), Steven Rostedt wrote:
> > On Fri, 2011-08-12 at 11:57 +0900, Masami Hiramatsu wrote:
> >
> >> I don't think it won't work. It can work but on a long way.
> >> Could you tell me your "bigger ideas"? Perhaps, we are on the different
> >> way but aim to same goal.
> >
> > Part of the bigger ideas is to have things like function graph tracing
> > use this, as it will simplify the entry.S code. There's other things
> > that may come out of this too.
> 
> Hmm, I think that the current function graph tracing implementation
> is more scalable than kretprobes, because kretprobe requires
> spinlock on every hit. Moreover, you can't probe NMI handler with
> kprobe, and kprobes on irq-handler are also possible to fail
> because of recursive-call.
> So I don't recommend using kretprobe for function-graph tracer :-(

Sorry for the confusion. My idea is not to use kretprobe with function
graph tracer, but to use the ftrace hooks with the pt_regs and friends
for function graph tracer instead of what it does today, which is to add
function graph code directly into entry.S.

The point I was making is, if I need to get ftrace function tracing
being good enough for function graph tracer, then it should work with
kprobes without any issues. If I need to do the work anyway (for
function graph tracing) then why not use it directly with kprobes
instead of doing more hooks just in the kprobe_trace?

-- Steve