Re: Elisp native profiler

[Eli Zaretskii]

> From: Stefan Monnier <address@hidden>
> Cc: Tomohiro Matsuyama <address@hidden>,  address@hidden
> Date: Fri, 28 Sep 2012 08:43:00 -0400
> 
> > If the overhead can be avoided when not profiling (e.g., depending on
> > some global variable), that would be acceptable, I think.
> 
> No, I think some of the overhead can't be chosen at runtime (e.g. the
> `struct backtrace' will always use up stack space)

I thought about CPU overhead, not stack size overhead.

> and adding a test on a global variable to functions like `car'
> doesn't sound like a good idea.

Why not?  Profiling is a global state, affecting all of Emacs, isn't
it?