Re: The Emacs master is much slower than the emacs-27 branch.

[Eli Zaretskii]

> From: Stephen Berman <stephen.berman@gmx.net>
> Cc: Óscar Fuentes <ofv@wanadoo.es>,  emacs-devel@gnu.org
> Date: Fri, 04 Dec 2020 22:37:59 +0100
> 
> > But I see no significant difference in my unoptimized builds of both
> > branches.
> 
> Doesn't -O0 mean unoptimized?

It does.

> But Gregory Heytings said that -O0 suffers the same performance hit
> as -Og in emacs-27 but not im master, which your timings seem to
> contradict.  What am I missing?

You aren't missing anything.  If Gregory has timings of -O0 builds of
both branches that show a significant effect of this issue, and which
disappear at the bisection point, then there is indeed a strange
contradiction between his and my timings.