normalized signature of signals/slots

[Olivier Goffart]

Le Lundi 14 Novembre 2005 13:05, Koos Vriezen a écrit :
> On Saturday 12 November 2005 11:30, Olivier Goffart wrote:
> > Hello
>
> ...
>
> > After some test (attached) I noticed that the normalized signature is
> > about twice times faster than the not normalized one
>
> Can you give some numbers (I tried to compile it but apparently that's
> not a 'g++ test_signals.cpp -lqt-mt' one ..)

I used 
gcc -I/opt/qt4/include/ -I/opt/qt4/include/QtCore/ -L/opt/qt4/lib/ 
-lQtCore_debug -O2 test_signals.cpp       


As result i obtain

normalized: 2469ms
usual: 5323ms    ratio: 2.15593

I repeated it several times, the ratio is always between 1.9 and 2.2

> > (This may i guess be even more in the case of bigger signatures, or class
> > with more signals and slots)
>
> Isn't it a hashtable?

No, it's all linear

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://mail.kde.org/pipermail/kde-optimize/attachments/20051114/cd4c4133/attachment.pgp