Re: [gmime-devel] GMime, now faster than ever.

[Jeffrey Stedfast <fejj novell com>]

On Sat, 2008-09-13 at 08:33 -0400, Jeffrey Stedfast wrote:
[snip]
> Still a full second slower than 2.2 svn :(
> 
> 
> Part of the improvement over yesterday's run-times for 2.3.x were
> because I unset my G_SLICE=always-malloc debug environment, which helped
> a little, but most of the rest of it is tricks to keep signal emission
> to a minimum and elimination of hash table lookups/maintenance.
> 
> What I've discovered is that GMime is spending a huge amount of time in
> g_object_new() (and children) and signal handler maintenance (such as
> g_signal_handlers_disconnect_matched) called from the dtors.
> 
> I'll keep trying to think of other ways to improve performance...

Decided to try writing my own simple event system to handle my simple
needs internally in GMime and I've gained back nearly all of the
performance I had lost in the switch to GObject:

[fejj serenity tests]$ time ./test-mbox

real	0m6.701s
user	0m5.704s
sys	0m0.888s
[fejj serenity tests]$ time ./test-mbox

real	0m6.705s
user	0m5.780s
sys	0m0.808s
[fejj serenity tests]$ time ./test-mbox

real	0m6.688s
user	0m5.732s
sys	0m0.852s


On this ancient/slow machine, 2.3.x is now only 0.1s slower than 2.2.23,
although on my more modern laptop, 2.3.x is now slightly faster than
2.2.23.

I guess I can close the case now ;-)

Jeff