Gcj binaries performance (arm-linux)

Andrew John Hughes gnu_andrew@member.fsf.org
Fri May 9 09:11:00 GMT 2008


2008/5/9 Andrew Haley <aph@redhat.com>:
> Andrew John Hughes wrote:
>  > 2008/4/29 ffileppo <ffileppo@libero.it>:
>  >>  >
>  >>  > aph@debian:~$  /home/aph/gcc-4.3.0-RC-20080301-install/bin/gcj MyPanel.class -O --main=MyPanel
>  >>  > aph@debian:~$ ./a.out
>  >>  > Test "Bubble Sort" (Num loop: 100)
>  >>  > First loop time: 22
>  >>  > Total time elapsed: 2197
>  >>  >
>  >>  > aph@debian:~$ /home/aph/gcc-4.3.0-RC-20080301-install/bin/gcj MyPanel.class -O2 --main=MyPanel
>  >>  > aph@debian:~$ ./a.out
>  >>  > Test "Bubble Sort" (Num loop: 100)
>  >>  > First loop time: 20
>  >>  > Total time elapsed: 1943
>
>  >>
>
> >>  Thanks for your help Andrew,
>  >>  I'm getting good performance on target using -O / -O2
>
>  >
>
> > So does gcj default to no optimisation? This is what these results
>  > suggest, as the -O version is about 18 times quicker.
>
>  Yes, of course.  That's standard for UNIX, as far as I'm aware: I don't
>  think I've ever come across a compiler that optimized unless it was
>  asked to.
>
>  Andrew.
>

That makes sense, especially as something like -O2 will make debugging harder
and probably has effect on some of the reflection properties of Java.

I guess this is what you get from Java people who tend to expect
everything to work
best with no options... ;) I always pass plenty when hacking C or C++ code.
-- 
Andrew :-)

Support Free Java!
Contribute to GNU Classpath and the OpenJDK
http://www.gnu.org/software/classpath
http://openjdk.java.net

PGP Key: 94EFD9D8 (http://subkeys.pgp.net)
Fingerprint: F8EF F1EA 401E 2E60 15FA 7927 142C 2591 94EF D9D8



More information about the Java mailing list