From kde-optimize Mon Nov 14 15:02:09 2005 From: Lubos Lunak Date: Mon, 14 Nov 2005 15:02:09 +0000 To: kde-optimize Subject: Re: normalized signature of signals/slots Message-Id: <200511141602.09965.l.lunak () suse ! cz> X-MARC-Message: https://marc.info/?l=kde-optimize&m=113198053819752 On Monday 14 of November 2005 15:40, Olivier Goffart wrote:> Le Lundi 14 Novembre 2005 14:56, Koos Vriezen a écrit :> > On Mon, Nov 14, 2005 at 01:44:14PM +0100, Olivier Goffart wrote:> > > As result i obtain> > >> > > normalized: 2469ms> > > usual: 5323ms ratio: 2.15593> >> > 2.6ms/connect still doesn't say much actually, does sound quite slow> > (but maybe you ran it on a 66 MHz ARM CPU :)>> 5323/(20000*3) = less than 0.1 ms per connect/disconnect pair> It was on a 1,8Ghz Pentium M>> Yes, it's quite few. that is why i ask if it would be interesting to write> a script that normalize the signature of signals in the code. There's benchmarking for finding out things like that. > I don't know how many signals/slots are connected at kde start static int cnt = 0; qDebug( "CONNECTS: %d", ++cnt ); > , but it > should be much.> is gaining 0.02ms per connection interesting enough ? Profiling kwrite startup with sysprof here says about 1-2% is spent in connect() but it's so fast that it's reaching samples resolution and as such I wouldn't trust it so much. You might want to try with cachegrind or a slower machine. -- Lubos LunakKDE developer---------------------------------------------------------------------SuSE CR, s.r.o. e-mail: l.lunak@suse.cz , l.lunak@kde.orgDrahobejlova 27 tel: +420 2 9654 2373190 00 Praha 9 fax: +420 2 9654 2374Czech Republic http://www.suse.cz/_______________________________________________Kde-optimize mailing listKde-optimize@kde.orghttps://mail.kde.org/mailman/listinfo/kde-optimize