Comments on: GObject performance work

By: alexl

alexl — Tue, 15 Sep 2009 12:24:12 +0000

Karl:
Its not that wrong. The construction speed is not the construction time. The construction speed is the number of constructed object per second, i.e. what is printed out by the performance test app, and this doubled.

By: Karl Lattimer

Karl Lattimer — Tue, 15 Sep 2009 09:52:49 +0000

“Construction speed for simple objects more than doubled” is bad English.

Essentially if construction speed was 1/2 a second, this sentence implies that it’s now 1 second.

What you should have said is;

“Construction time for simple objects more than halved” – which makes more sense.

By: 14.09.2009 OSiM 2009 – Blog of Tim Janik

14.09.2009 OSiM 2009 – Blog of Tim Janik — Tue, 15 Sep 2009 07:55:29 +0000

[...] Yes, I’ve seen Alex recent work and will take a look once I’m back from the [...]

By: GNOME Commit-Digest » Blog Archive » Issue 49

GNOME Commit-Digest » Blog Archive » Issue 49 — Sun, 13 Sep 2009 11:51:39 +0000

[...] Larsson did some GObject performance work (in the glib gobject-performance branch), improving the speed of instance creation for classes that [...]

By: Havoc

Havoc — Sat, 12 Sep 2009 01:57:35 +0000

Nice, that was about 9 years overdue

Might be fun if the performance suite had a simple gobject vs. hand-rolled struct comparison in it, just to quantify that (people will still think the hand-rolled struct is much faster until they see numbers I bet)

By: pvanhoof

pvanhoof — Fri, 11 Sep 2009 08:26:01 +0000

Thanks for picking up the patches. Sad that it had to take so much time for somebody to pick it up, the bug is almost a year old, and the same patch was proposed immediately after opening the bug. Nice that you further finetuned by taking a look at the interface-lookup problems too.

By: Stefan Kost

Stefan Kost — Fri, 11 Sep 2009 05:53:47 +0000

Forgot to mentioned the bugs:
* bug #536939 has a testapplication and some discussion
* bug #418970 has some more discussion with the outcome that the notify queue list should use g_slice

By: Jeffrey Stedfast

Jeffrey Stedfast — Thu, 10 Sep 2009 20:39:28 +0000

Woot! Thank you Alex! This is awesome! GObject creation performance was the one thing that really hurt GMime’s performance in the move from 2.2 to 2.4 (because I made everything into a GObject instead of a custom struct for a lot of things).

By: Stefan Kost

Stefan Kost — Thu, 10 Sep 2009 19:26:00 +0000

Cool. So what about getting rid of the global paramspec pool next

By: jegHegy

jegHegy — Thu, 10 Sep 2009 14:23:17 +0000

Thanks for the replies! Keep up the good work, everyone.