Re: macOS metal rendering engine in mac port

[Alan Third]

On Sat, May 29, 2021 at 01:52:35AM -0700, Aaron Jensen wrote:
> 
> emacs -Q never really felt slow in terms of typing latency. In terms
> of the scroll benchmark, I got:
> 
> w/o patch: 26.8s first run, 3.84s second run
> w/ patch: 25.1s first run, 3.6s second run
> 
> With my emacs config that has, with everything loaded, 1156 faces, I get:
> 
> w/o patch: 18.6s on second run (at least 9s of which are attributed to
> TAGGEDP and CONSP from lface_from_face_name_no_resolve)
> w/ patch: 6.1s on second run
> 
> The impact is drastic.

It makes me wonder if we're doing something very silly with faces in
the NS port, but after a quick look through I don't see anything
obvious. I don't really know what I'd be looking for though.

One thing I know that's odd in the NS port is that we work out which
face we need on the fly, whereas the other terms seem to have a
central function that works out the faces early. But I don't believe
that should have any effect as it all appears to be fairly simple
lookups.
-- 
Alan Third