RE: Class loader optimization patch

[Mark Wielaard]

Hi,

On Tue, 2003-03-04 at 10:44, Jeroen Frijters wrote:
> Mark Wielaard wrote:
> > Note that this is a very nice speed tests for the different VMs. It
> > takes minutes before the main Eclipse Window shows up even on a fast
> > machine. Maybe we can have a price for the first VM that 
> > manages to load Eclipse under one minute on some standard PC
> > (say 1Ghz Athlon/Pentium with 512MB).
> 
> 1 minute 12 seconds is the best I can do at the moment. This is with
> Eclipse 2.1 RC1 with no windows open on an 850 MHz Pentium III with
> 512MB. Getting to the point where I can edit Java code takes a couple
> minutes more though.

Impressive. I get the following timings for Eclipse RC2, gij 3.4 (CVS
HEAD, verifier disabled), starting up and using the mouse to select the
exit item as soon as possible (it then takes a couple of seconds to
"Save the workspace"):

Complete fresh start (two splash screens):
real    3m41.889s
user    3m28.310s
sys     0m3.320s

Second startup (one splash screen):
real    2m44.068s
user    2m33.670s
sys     0m2.430s

This is on an Athlon XP 1.4Ghz with 768MB.

The bottleneck seems to be class loading and exception handling.

> For reference, starting eclipse.exe (using Sun's JRE 1.4.1) takes 11
> seconds.

Nah. Proprietary implementations don't count. I assume your timings are
with Mono on Wine. None of that proprietary MS stuff I hope :)

Cheers,

Mark