[Soot-list] instruction scheduling

[Hiroshi Yamauchi]

Hi all,

I am wondering if static instruction scheduling *on the Java bytecode  
level*
using a tool like Soot would matter in terms of performance.

Has anyone heard of any work on that topic? Even if it is indeed not  
a good idea or
does not contribute to the runtime performance at all, it would be  
nice to have that
confirmed by some measurement results.

Best regards,
Hiroshi