On Thu, 2010-02-04 at 12:02 +0100, michael schuster wrote:
> Daniel Stenberg wrote:
> > On Thu, 4 Feb 2010, michael schuster wrote:
> >
> >>> I edited away the glib-related details from the curltest.c example
> >>> and when running as instructed I can see the curltest process use the
> >>> same amount of memory all the time...
> >>
> >> I'd be interested in your rationale for removing glib details - can
> >> you please expand?
> >
> > Primarily I wanted to remove them as I don't have any glib devel package
> > installed and I didn't want to bother with that, but then I also thought
> > that the test might even get better without glib since I then remove a
> > 3rd party thing that we don't want to risk having anything to do with
> > the problem at hand.
>
> it might be interesting to see whether you can detect any difference with
> and without glib - I don't (not significantly, anyway).
>
> thx
> Michael

FWIW, glib has some known issues with at least appearing to leak memory.

I don't know if this bug is relevant to your code since I am very much
not a glib developer.

https://bugzilla.gnome.org/show_bug.cgi?id=64096

http://mail.gnome.org/archives/gtk-devel-list/2001-November/msg00541.html

See also:
http://library.gnome.org/devel/gtk-faq/stable/x702.html

Long story short, I'd definitely remove glib from the equation in order
to track the leak down.

--Chris

-------------------------------------------------------------------
List admin: http://cool.haxx.se/list/listinfo/curl-library
Etiquette: http://curl.haxx.se/mail/etiquette.html
Received on 2010-02-04