unofficial mirror of emacs-devel@gnu.org 
 help / color / mirror / code / Atom feed
From: joakim@verona.se
To: Tassilo Horn <tassilo@member.fsf.org>
Cc: Emacs-devel@gnu.org, Jashy <nanjunjie@gmail.com>,
	Miles Bader <miles@gnu.org>
Subject: Re: Using the ImageMagick backend seems to leak memory
Date: Tue, 11 Jan 2011 19:11:48 +0100	[thread overview]
Message-ID: <m3r5cjl38b.fsf@verona.se> (raw)
In-Reply-To: <87zkr7nxo8.fsf@member.fsf.org> (Tassilo Horn's message of "Tue,  11 Jan 2011 18:43:35 +0100")

Tassilo Horn <tassilo@member.fsf.org> writes:

> joakim@verona.se writes:
>
>>> ClearMagickWand() clears resources associated with the wand.
>>> The format of the ClearMagickWand method is:
>>>
>>>   void ClearMagickWand(MagickWand *wand)
>>>
>>> A description of each parameter follows:
>>> wand
>>>   the magick wand.
>>>
>>> In the emacs sources, that function's never called.  So that's
>>> probably the source of the leak.
>>
>> Hmm. We can try.
>
> Andreas already added DestroyMagickWand() calls that imply clearing, and
> I locally added MagickWandTerminus() calls.  Right now, I presume the
> leak is not on the IM side.
>
> Your patch in the other mail (I'll try it later) seems to support that
> guess.

I've looked at this a little bit further.

The error is that the image read with ReadImage is cloned by
NewMagickWandFromImage not refered as I believed when I wrote the code.

The code was written that way to avoid another memory problem I
experienced with image bundles. If the current strategy remains,
DestroyImage(im_image); should be the right fix.

However, the strategy should probably be changed to avoid the
cloning. This will result in a more complex code path.


>
>> Anyway i asked the IM team:
>> http://www.imagemagick.org/discourse-server/viewtopic.php?f=1&t=17819>
>> Some other things:
>> - Do you see the issue only when scaling?
>
> No, always.
>
>> - Only with PNG? Or also JPG?
>
> I tested with PNG and TIF, and both have the problem.
>
> Bye,
> Tassilo

-- 
Joakim Verona



  reply	other threads:[~2011-01-11 18:11 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-01-06  7:51 Using the ImageMagick backend seems to leak memory Tassilo Horn
2011-01-06  9:02 ` joakim
2011-01-06 10:20   ` Tassilo Horn
2011-01-07  4:17 ` Jashy
2011-01-07  4:59   ` Miles Bader
2011-01-07  6:27     ` Stephen J. Turnbull
2011-01-07  7:25       ` Miles Bader
2011-01-07  8:28     ` Tassilo Horn
2011-01-07 22:39       ` Andreas Schwab
2011-01-10 10:34         ` Tassilo Horn
2011-01-10 13:28           ` Jashy
2011-01-10 13:56             ` Tassilo Horn
2011-01-10 14:36               ` Jashy
2011-01-11 10:14                 ` Jashy
2011-01-11 10:59           ` joakim
2011-01-11 14:17       ` joakim
2011-01-11 14:38         ` joakim
2011-01-11 17:43         ` Tassilo Horn
2011-01-11 18:11           ` joakim [this message]
2011-01-11 19:10             ` Tassilo Horn
2011-01-11 19:35               ` joakim
2011-01-11 19:55                 ` Tassilo Horn
2011-01-12  1:47               ` Stefan Monnier
2011-01-12  7:37                 ` Tassilo Horn
2011-01-10 21:47 ` Tassilo Horn
2011-01-11  8:47   ` Tassilo Horn
2011-01-11  9:35     ` joakim
2011-01-11  9:53       ` Tassilo Horn
2011-01-11  9:54       ` joakim

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=m3r5cjl38b.fsf@verona.se \
    --to=joakim@verona.se \
    --cc=Emacs-devel@gnu.org \
    --cc=miles@gnu.org \
    --cc=nanjunjie@gmail.com \
    --cc=tassilo@member.fsf.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).