From: Dmitry Antipov <dmantipov@yandex.ru>
To: emacs-devel@gnu.org
Subject: Re: Emacs bzr memory footprint
Date: Fri, 21 Oct 2011 18:30:28 +0400 [thread overview]
Message-ID: <4EA18204.4010604@yandex.ru> (raw)
In-Reply-To: <87y5weo6jc.fsf@spindle.srvr.nix>
On 10/21/2011 04:47 PM, Nix wrote:
> Unfortunately it won't release memory back to the OS in the situation I
> described. mmap() is expensive, so only large allocations (by
> default>128Kb or>512Kb depending on glibc version and arch) get
> mmap()ed.
On GNU/Linux system, Emacs tweaks this default with mallopt (M_MMAP_THRESHOLD, 64*1024).
Nevertheless, glibc malloc design assumes that smaller (below threshold above) chunks may
be still allocated with mmap(). Run emacs with:
strace -e signal=none -e trace=mmap,mremap,munmap emacs
and you will see a lot of allocation requests of very different sizes.
On the other side, lisp data is newer mapped, and in some cases this is the main reason
which causes sbrk()'ed heap fragmentation. Don't beat toolkits, don't beat system libraries.
Now I'm observing the case where 4 GCs reduces amount of lisp data (live + free lists) from
30M to 8M, and mallinfo().fordblks raises from 1M to 22M. This 21M of fragmented heap
is 'locked' into malloc freelists; they can be re-used, but will never released back to OS.
Dmitry
next prev parent reply other threads:[~2011-10-21 14:30 UTC|newest]
Thread overview: 113+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-10-13 9:24 Emacs bzr memory footprint Carsten Mattner
2011-10-13 10:01 ` Eli Zaretskii
2011-10-13 10:33 ` Andreas Schwab
2011-10-13 10:46 ` Eli Zaretskii
2011-10-13 13:08 ` Carsten Mattner
2011-10-13 13:38 ` Eli Zaretskii
2011-10-13 15:22 ` Miles Bader
2011-10-13 15:30 ` Eli Zaretskii
2011-10-13 16:16 ` Stefan Monnier
2011-10-13 17:33 ` Dmitry Antipov
2011-10-13 18:42 ` Eli Zaretskii
2011-10-14 3:09 ` Dmitry Antipov
2011-10-13 16:46 ` Stephen J. Turnbull
2011-10-13 16:59 ` Carsten Mattner
2011-10-13 13:13 ` Carsten Mattner
2011-10-13 13:43 ` Eli Zaretskii
2011-10-13 14:10 ` Dave Abrahams
2011-10-13 14:46 ` Eli Zaretskii
2011-10-13 15:05 ` Dave Abrahams
2011-10-13 14:22 ` Stefan Monnier
2011-10-13 16:55 ` Carsten Mattner
2011-10-13 17:09 ` Stefan Monnier
2011-10-14 12:11 ` Carsten Mattner
2011-10-14 21:02 ` John Wiegley
2011-10-15 1:22 ` Stefan Monnier
2011-10-13 17:22 ` Richard Riley
2011-10-13 16:12 ` Chong Yidong
2011-10-13 20:18 ` chad
2011-10-14 12:13 ` Carsten Mattner
2011-10-14 12:18 ` Carsten Mattner
2011-10-14 15:49 ` Sven Joachim
2011-10-14 16:30 ` Ted Zlatanov
2011-10-14 16:48 ` Sven Joachim
2011-10-14 17:16 ` Carsten Mattner
2011-10-15 18:12 ` Sven Joachim
2011-10-20 13:35 ` Nix
2011-10-20 23:02 ` John Wiegley
2011-10-21 0:19 ` Nix
2011-10-21 8:19 ` Eli Zaretskii
2011-10-21 12:05 ` Nix
2011-10-21 12:36 ` Eli Zaretskii
2011-10-21 12:47 ` Nix
2011-10-21 14:30 ` Dmitry Antipov [this message]
2011-10-21 15:22 ` Nix
2011-10-21 18:29 ` Dmitry Antipov
2011-10-21 18:36 ` Nix
2011-10-21 13:30 ` Stefan Monnier
2011-10-21 15:02 ` Nix
2011-10-21 17:52 ` Stefan Monnier
2011-10-21 19:12 ` Nix
2011-10-21 15:34 ` Dmitry Antipov
2011-10-21 18:00 ` Stefan Monnier
2011-10-21 18:27 ` Óscar Fuentes
2011-10-21 18:58 ` Eli Zaretskii
2011-10-21 19:11 ` Nix
2011-10-21 20:18 ` Eli Zaretskii
2011-10-21 20:21 ` Nix
2011-10-22 7:34 ` Carsten Mattner
2011-10-22 8:30 ` Stephen J. Turnbull
2011-10-22 12:03 ` Carsten Mattner
2011-10-22 16:08 ` Stephen J. Turnbull
2011-10-27 22:10 ` Ted Zlatanov
2011-10-28 6:40 ` Eli Zaretskii
2011-10-28 9:34 ` Stephen J. Turnbull
2011-10-28 9:47 ` Eli Zaretskii
2011-10-28 10:22 ` Stephen J. Turnbull
2011-10-28 10:36 ` Eli Zaretskii
2011-10-28 11:52 ` valgrind warnings [Re: Emacs bzr memory footprint] Dan Nicolaescu
2011-10-28 14:08 ` Stephen J. Turnbull
2011-10-28 14:38 ` Dan Nicolaescu
2011-10-28 15:17 ` Chong Yidong
2011-10-28 15:35 ` Eli Zaretskii
2011-10-28 15:27 ` Eli Zaretskii
2011-10-28 16:34 ` Dan Nicolaescu
2011-10-28 18:36 ` Eli Zaretskii
2011-10-28 18:51 ` Eli Zaretskii
2011-10-28 18:52 ` Dan Nicolaescu
2011-10-28 19:03 ` Eli Zaretskii
2011-10-28 15:08 ` Ted Zlatanov
2011-12-04 21:11 ` Florian Weimer
2011-10-28 20:20 ` Dan Nicolaescu
2011-10-29 3:40 ` Chong Yidong
2011-10-29 7:04 ` Andreas Schwab
2011-10-29 16:25 ` Stefan Monnier
2011-11-04 12:51 ` Dan Nicolaescu
2011-11-04 13:08 ` Andreas Schwab
2011-11-04 15:50 ` valgrind warnings Paul Eggert
2011-11-04 17:37 ` valgrind warnings [Re: Emacs bzr memory footprint] Stefan Monnier
2011-11-04 19:10 ` Dan Nicolaescu
2011-11-04 20:56 ` Paul Eggert
2011-11-05 22:08 ` Dan Nicolaescu
2011-11-06 2:34 ` Paul Eggert
2011-11-06 12:37 ` Dan Nicolaescu
2011-11-07 1:13 ` Paul Eggert
2011-10-29 9:10 ` Eli Zaretskii
2011-10-22 8:44 ` Emacs bzr memory footprint Sven Joachim
2011-10-22 11:57 ` Carsten Mattner
2011-10-23 18:04 ` Stefan Monnier
2011-10-25 17:35 ` Nix
2011-10-25 20:23 ` Stefan Monnier
2011-10-25 20:34 ` Nix
2011-10-26 14:52 ` gnutls memory leak [Was: Re: Emacs bzr memory footprint] Chong Yidong
2011-10-27 6:10 ` Chong Yidong
2011-10-27 22:01 ` Ted Zlatanov
2011-10-28 12:04 ` Ted Zlatanov
2011-10-28 17:29 ` Nix
2011-10-27 22:03 ` Ted Zlatanov
2011-10-13 21:23 ` Emacs bzr memory footprint Dan Nicolaescu
2011-10-14 13:04 ` Carsten Mattner
2011-10-14 18:13 ` James Cloos
2011-10-17 20:52 ` Juri Linkov
2011-10-17 23:34 ` James Cloos
2011-10-18 7:24 ` Andreas Schwab
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=4EA18204.4010604@yandex.ru \
--to=dmantipov@yandex.ru \
--cc=emacs-devel@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).