first test

Thu Jul 15 11:43:00 GMT 2004

Andrew Haley wrote:
> > public class Loop {
> > 
> > public static void main(String[] args) {
> > long before = System.currentTimeMillis();
> > for (int i=0; i < 1000000; i++){
> > int a = 27 + 3 * i;
> > String b = String.valueOf(a) + "hello";
> > }
> > long after = System.currentTimeMillis();
> > System.out.println("time: " + (after-before));
> > }
> > 
> > }
> > 
> > running it on Mac OS X 10.3.4 1.33 GHz G4:
> > 
> > java Loop
> > time: 2269
> > 
> > gcj --main=Loop -o Loop Loop.java
> > ./Loop
> > time: 4369
> > 
> > ??? the app compiled with gcj is twice as much slow as the one running 
> > through the VM ???
> > Redoing the test doesn't change much.
>> This isn't a test of the compiler, because it spends almost of of its
> time inside the library calls. It's really a test of the performance
> of the different Java runtime libraries.
>> However, I get 3860 ms for the JIT compiler and 4613 ms for gcj. That
> means that gcj's implementation of strings and/or memory allocation is
> slightly slower.
>> I don't know why there is such a disparity in your case. It is
> possible that on Mac OS X you have a fast implementation of Strings.

FWIW here are some more Mac OS X results (darwin kernel 7.4.0, 1GHz G4,
gcj 3.5.0 20040709):
time java Loop
time: 2066
real 0m2.803s
user 0m2.270s
sys 0m0.220s
gcj --main=Loop -o Loop Loop.java
time ./Loop
time: 3466
real 0m4.340s
user 0m3.920s
sys 0m0.100s
gcj --main=Loop -o Loop -O2 Loop.java
time ./Loop
time: 3372
real 0m4.230s
user 0m3.900s
sys 0m0.050s
-- 
Chris Burdess