Re: [RFC PATCH -next v2 3/4] arm64/ftrace: support dynamically allocated trampolines

[Wangshaobo (bobo)]

在 2022/4/21 22:06, Steven Rostedt 写道:
> On Thu, 21 Apr 2022 14:10:04 +0100
> Mark Rutland <mark.rutland@xxxxxxx> wrote:
>
> > On Wed, Mar 16, 2022 at 06:01:31PM +0800, Wang ShaoBo wrote:
> > > From: Cheng Jian <cj.chengjian@xxxxxxxxxx>
> > >
> > > When tracing multiple functions customly, a list function is called
> > > in ftrace_(regs)_caller, which makes all the other traced functions
> > > recheck the hash of the ftrace_ops when tracing happend, apparently
> > > it is inefficient.
> > ... and when does that actually matter? Who does this and why?
> I don't think it was explained properly. What dynamically allocated
> trampolines give you is this.
>
> Let's say you have 10 ftrace_ops registered (with bpf and kprobes this can
> be quite common). But each of these ftrace_ops traces a function (or
> functions) that are not being traced by the other ftrace_ops. That is, each
> ftrace_ops has its own unique function(s) that they are tracing. One could
> be tracing schedule, the other could be tracing ksoftirqd_should_run
> (whatever).
>
> Without this change, because the arch does not support dynamically
> allocated trampolines, it means that all these ftrace_ops will be
> registered to the same trampoline. That means, for every function that is
> traced, it will loop through all 10 of theses ftrace_ops and check their
> hashes to see if their callback should be called or not.
>
> With dynamically allocated trampolines, each ftrace_ops will have their own
> trampoline, and that trampoline will be called directly if the function
> is only being traced by the one ftrace_ops. This is much more efficient.
>
> If a function is traced by more than one ftrace_ops, then it falls back to
> the loop.
>
> -- Steve
> .

yes, this explanation is easier to understand, I will update commit 
description according to this.

-- Wang ShaoBo