all messages for Emacs-related lists mirrored at yhetil.org
 help / color / mirror / code / Atom feed
From: Lars Ingebrigtsen <larsi@gnus.org>
To: "Óscar Fuentes" <ofv@wanadoo.es>
Cc: emacs-devel@gnu.org
Subject: Re: Enabling native compilation by default when libgccjit is present
Date: Sun, 05 Dec 2021 20:40:43 +0100	[thread overview]
Message-ID: <87y24yrgas.fsf@gnus.org> (raw)
In-Reply-To: <87mtlfasg4.fsf@telefonica.net> ("Óscar Fuentes"'s message of "Sun, 05 Dec 2021 18:10:19 +0100")

Óscar Fuentes <ofv@wanadoo.es> writes:

>> I'd be curious to know where people are seeing speedups.
>
> In my experience nativecomp is somewhat noticeable while using lsp-mode.
>
> Otherwise, not much, if at all.

In some micro-benchmarks, I see significant speedups.  Let's see...
what did I test last...

Right:

---
;; (benchmark-run 1000 (benchy))

(defun benchy ()
  (let ((syms 0))
    (mapatoms (lambda (sym)
		(setq syms (+ syms (length (symbol-name sym)))))
	      obarray)
    syms))
---

This is 50% faster with nativecomp than without.

It's easy to be fooled when trying to benchmark, since an "emacs -Q"
doesn't native-comp the bits you're trying to benchmark, necessarily.
(Unless you've done a full AOT build, which most people don't.)

But still, even when that's done correctly, the nativecomp speedups are
pretty elusive in real world code.  For instance, I'd expect shr
rendering to be significantly speeded up when doing table-based layouts,
because it's doing a lot of...  stuff.  But:

(benchmark-run 10 (eww-open-file "/tmp/foo.html"))

stubbornly takes the same amount of time with or without nativecomp,
and it's the same way with every real bit of code I've tested (like byte
compilation, which I'd also expect to be faster with nativecomp).

So I don't know whether something is defeating nativecomp things in more
complex situations.

> The reason for this is not hard to understand for anyone with a little
> bit of knowledge about binary code optimization. For instance, if your
> Elisp code consists on invoking opaque C primitives, the gain of
> compiling it to machine instructions will be mostly irrelevant.

My micro-benchmark up there is almost all opaque C primitives, but is
helped a lot.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no



  parent reply	other threads:[~2021-12-05 19:40 UTC|newest]

Thread overview: 91+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-12-02 19:25 Enabling native compilation by default when libgccjit is present Andrea Corallo
2021-12-02 19:39 ` Eli Zaretskii
2021-12-03 16:52   ` Lars Ingebrigtsen
2021-12-03 18:30     ` Eli Zaretskii
2021-12-03 21:27       ` Lars Ingebrigtsen
2021-12-03 18:53     ` Stefan Kangas
2021-12-03 21:30       ` Lars Ingebrigtsen
2021-12-03 22:39         ` Stefan Kangas
2021-12-04  2:48           ` Lars Ingebrigtsen
2021-12-04  7:48             ` Eli Zaretskii
2021-12-04 19:28             ` Andrea Corallo
2021-12-04 19:41               ` Lars Ingebrigtsen
2021-12-04 19:57                 ` Andrea Corallo
2021-12-05  5:04                 ` Richard Stallman
2021-12-05  5:09                   ` Lars Ingebrigtsen
2021-12-05  5:29                     ` Po Lu
2021-12-05  5:31                       ` Lars Ingebrigtsen
2021-12-05  5:46                         ` Po Lu
2021-12-05  5:48                           ` Lars Ingebrigtsen
2021-12-05  6:00                             ` Po Lu
2021-12-05  5:54                         ` Tim Cross
2021-12-05  8:20                           ` Eli Zaretskii
2021-12-05 11:27                           ` Stefan Kangas
2021-12-05 14:15                           ` Jose A. Ortega Ruiz
2021-12-05 16:52                             ` Arthur Miller
2021-12-05 17:42                               ` Jose A. Ortega Ruiz
2021-12-05 20:48                                 ` Arthur Miller
2021-12-05 17:10                             ` Óscar Fuentes
2021-12-05 19:25                               ` Jose A. Ortega Ruiz
2021-12-05 20:55                                 ` Óscar Fuentes
2021-12-05 22:59                                   ` Stefan Monnier
2021-12-05 19:40                               ` Lars Ingebrigtsen [this message]
2021-12-05 20:02                                 ` Eli Zaretskii
2021-12-05 20:32                                   ` Lars Ingebrigtsen
2021-12-05 23:03                                     ` Stefan Monnier
2021-12-05 23:05                                       ` Lars Ingebrigtsen
2021-12-05 23:27                                         ` Stefan Monnier
2021-12-05 20:37                                 ` Óscar Fuentes
2021-12-05 21:44                                 ` Arthur Miller
2021-12-05 21:48                                   ` Lars Ingebrigtsen
2021-12-05 22:58                                     ` Arthur Miller
2021-12-05 21:15                               ` Arthur Miller
2021-12-06  4:33                               ` Richard Stallman
2021-12-06 12:59                                 ` Eli Zaretskii
2021-12-06 13:57                                   ` Stefan Monnier
2021-12-06 14:08                                     ` Eli Zaretskii
2021-12-07 16:02                                       ` Matt Armstrong
2021-12-07 16:55                                         ` tomas
2021-12-07 17:02                                         ` T.V Raman
2021-12-05 20:00                             ` Jim Myhrberg
2021-12-06  1:03                               ` Dmitry Gutov
2021-12-06  2:50                                 ` Jim Myhrberg
2021-12-06 12:59                                   ` Dmitry Gutov
2021-12-07  0:34                           ` João Pedro de Amorim Paula
2021-12-08  4:34                             ` Richard Stallman
2021-12-09 17:00                           ` David Koppelman
2021-12-06  4:33                     ` Richard Stallman
2021-12-03 22:34 ` Michael Heerdegen
2021-12-04  7:26   ` Eli Zaretskii
2021-12-07  4:38     ` Michael Heerdegen
2021-12-07  5:02       ` Michael Heerdegen
2021-12-07 13:59         ` Eli Zaretskii
2021-12-08  2:31           ` Michael Heerdegen
2021-12-08  9:55             ` Robert Pluim
2021-12-09  0:56               ` Michael Heerdegen
2021-12-08 12:52             ` Eli Zaretskii
2021-12-08  8:16           ` Andreas Schwab
2021-12-08 13:21             ` Eli Zaretskii
2021-12-09  0:29               ` Michael Heerdegen
2021-12-09  4:49                 ` Stefan Monnier
2021-12-09  5:27                   ` Michael Heerdegen
2021-12-09 10:22                     ` Andrea Corallo
2021-12-09 13:14                     ` Stefan Monnier
2021-12-09 22:28                       ` Michael Heerdegen
2021-12-09 22:50                         ` Michael Heerdegen
2021-12-10  8:06                           ` Eli Zaretskii
2021-12-11  3:11                             ` Michael Heerdegen
2021-12-07 13:53       ` Eli Zaretskii
2021-12-08  2:25         ` Michael Heerdegen
2021-12-08  2:40           ` Po Lu
2021-12-08  3:39             ` Michael Heerdegen
2021-12-08 12:55               ` Eli Zaretskii
2021-12-08 17:08               ` T.V Raman
2021-12-08 13:52             ` Stefan Monnier
2021-12-08 14:31               ` Eli Zaretskii
2021-12-09  7:19                 ` Óscar Fuentes
2021-12-08 12:51           ` Eli Zaretskii
2021-12-09  0:44             ` Michael Heerdegen
2021-12-09  4:13           ` Richard Stallman
2021-12-09  9:27             ` Eli Zaretskii
  -- strict thread matches above, loose matches on Subject: below --
2021-12-05 17:24 Angelo Graziosi

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87y24yrgas.fsf@gnus.org \
    --to=larsi@gnus.org \
    --cc=emacs-devel@gnu.org \
    --cc=ofv@wanadoo.es \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this external index

	https://git.savannah.gnu.org/cgit/emacs.git
	https://git.savannah.gnu.org/cgit/emacs/org-mode.git

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.