Re: emacs-26 eec71eb: Speed up replace-buffer-contents

[Paul Eggert]

Eli Zaretskii wrote:
> Profiling the optimized build shows that 60% of the time is spent
> inside buffer_chars_equal, and another 26% inside compareseq.
> rarely_quit and maybe_quit are nowhere in sight in the profile.

That's what I would expect: rarely_quit is inlined, so you don't see it in the 
profile (even though it's there and has some cost), and maybe_quit is called so 
rarely that you don't see it in the profile.

> So I think we have no performance problem we need to fix, right?

It sounds like we never had a performance problem in optimized builds that we 
needed to fix. I.e., the recent patch was needed only for unoptimized builds. 
Which is unfortunate, as we shouldn't have to hand-optimize unoptimized builds. 
I'll keep that in mind if I ever find time to look into the performance problem.