unofficial mirror of bug-gnu-emacs@gnu.org 
 help / color / mirror / code / Atom feed
From: Arnold Noronha <arnold@tdrhq.com>
To: 41029@debbugs.gnu.org
Subject: bug#41029: Improve ido-switch-buffer performance when many buffers are open
Date: Sat, 2 May 2020 08:40:55 -0700	[thread overview]
Message-ID: <20200502154055.GA11238@tdrhq.com> (raw)

[-- Attachment #1: Type: text/plain, Size: 465 bytes --]


Many of the buffer locals (in particular I think it's the
SYMBOL_FORWARDED types --- my terminoly may be off), appear to be O(n)
to bind, where n is the number of live buffers.

I tried to see if I can optimize that binding process directly, but
couldn't figure out a way to do it without changing current behavior,
so I opted to just optimize ido-switch-buffer directly.

After this patch, the performance is quite bearable at ~6000 buffers,
but still not ideal.

[-- Attachment #2: 0001-Avoid-binding-CASE-FOLD-SEARCH-in-tight-loop.patch --]
[-- Type: text/x-diff, Size: 3944 bytes --]

From 3bc4eb78c365f4b4a387451f0b55264076a97c4b Mon Sep 17 00:00:00 2001
From: Arnold Noronha <arnold@tdrhq.com>
Date: Sat, 2 May 2020 08:30:47 -0700
Subject: [PATCH] Avoid binding CASE-FOLD-SEARCH in tight loop

IDO-IGNORE-ITEM-P is called for every buffer when switching
buffers. However, that function binds CASE-FOLD-SEARCH, which
appears to be an O(n) operation where n is (length (buffer-list)).
So we avoid binding it if we know we don't need to as a workaround.

To test this, this this:

(loop for i from 0 to 6000
      do
      (find-file (format "/tmp/xy%d.txt" i)))

Then try switching buffers with and without this patch.
---
 lisp/ido.el | 67 ++++++++++++++++++++++++++++-------------------------
 1 file changed, 36 insertions(+), 31 deletions(-)

diff --git a/lisp/ido.el b/lisp/ido.el
index 81883402ad..d7a24ae1d7 100644
--- a/lisp/ido.el
+++ b/lisp/ido.el
@@ -3408,7 +3408,7 @@ ido-make-merged-file-list
 (defun ido-make-buffer-list-1 (&optional frame visible)
   "Return list of non-ignored buffer names."
   (delq nil
-	(mapcar
+       (mapcar
 	 (lambda (x)
 	   (let ((name (buffer-name x)))
 	     (if (not (or (ido-ignore-item-p name ido-ignore-buffers) (member name visible)))
@@ -3823,36 +3823,41 @@ ido-ignore-item-p
       (and
        ido-process-ignore-lists re-list
        (save-match-data
-	 (let ((ext-list (and ignore-ext ido-ignore-extensions
-			      completion-ignored-extensions))
-	       (case-fold-search ido-case-fold)
-	       ignorep nextstr
-	       (flen (length name)) slen)
-	   (while ext-list
-	     (setq nextstr (car ext-list))
-	     (if (cond
-		  ((stringp nextstr)
-		   (and (>= flen (setq slen (length nextstr)))
-			(string-equal (substring name (- flen slen)) nextstr)))
-		  ((functionp nextstr) (funcall nextstr name))
-		  (t nil))
-		 (setq ignorep t
-		       ext-list nil
-		       re-list nil)
-	       (setq ext-list (cdr ext-list))))
-	   (while re-list
-	     (setq nextstr (car re-list))
-	     (if (cond
-		  ((stringp nextstr) (string-match nextstr name))
-		  ((functionp nextstr) (funcall nextstr name))
-		  (t nil))
-		 (setq ignorep t
-		       re-list nil)
-	       (setq re-list (cdr re-list))))
-	   ;; return the result
-	   (if ignorep
-	       (setq ido-ignored-list (cons name ido-ignored-list)))
-	   ignorep)))))
+         (flet ((inner ()
+                       (let ((ext-list (and ignore-ext ido-ignore-extensions
+			                    completion-ignored-extensions))
+	                     ignorep nextstr
+	                     (flen (length name)) slen)
+	                 (while ext-list
+	                   (setq nextstr (car ext-list))
+	                   (if (cond
+		                ((stringp nextstr)
+		                 (and (>= flen (setq slen (length nextstr)))
+			              (string-equal (substring name (- flen slen)) nextstr)))
+		                ((functionp nextstr) (funcall nextstr name))
+		                (t nil))
+		               (setq ignorep t
+		                     ext-list nil
+		                     re-list nil)
+	                     (setq ext-list (cdr ext-list))))
+	                 (while re-list
+	                   (setq nextstr (car re-list))
+	                   (if (cond
+		                ((stringp nextstr) (string-match nextstr name))
+		                ((functionp nextstr) (funcall nextstr name))
+		                (t nil))
+		               (setq ignorep t
+		                     re-list nil)
+	                     (setq re-list (cdr re-list))))
+	                 ;; return the result
+	                 (if ignorep
+	                     (setq ido-ignored-list (cons name ido-ignored-list)))
+	                 ignorep)))
+
+           (if (eq case-fold-search ido-case-fold)
+               (inner)
+             (let ((case-fold-search ido-case-fold))
+               (inner))))))))
 
 ;; Private variable used by `ido-word-matching-substring'.
 (defvar ido-change-word-sub)
-- 
2.20.1


             reply	other threads:[~2020-05-02 15:40 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-02 15:40 Arnold Noronha [this message]
     [not found] ` <handler.41029.B.158843545619280.ack@debbugs.gnu.org>
2020-05-02 16:30   ` bug#41029: Acknowledgement (Improve ido-switch-buffer performance when many buffers are open) Arnold Noronha
2020-05-02 17:05 ` bug#41029: Improve ido-switch-buffer performance when many buffers are open Arnold Noronha
2020-05-02 18:26   ` Arnold Noronha
2020-05-03  2:03     ` Dmitry Gutov
2020-05-28 23:36       ` Dmitry Gutov

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

  List information: https://www.gnu.org/software/emacs/

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

  git send-email \
    --in-reply-to=20200502154055.GA11238@tdrhq.com \
    --to=arnold@tdrhq.com \
    --cc=41029@debbugs.gnu.org \
    /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 public inbox

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

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).