Error of precision in the function pow when using the JVM

Sylvestre Ledru sylvestre.ledru at scilab.org
Tue Jul 22 00:18:03 CEST 2008


Hello,

We noticed a bug caused by the JVM causing some precision issues in the
function pow when it is used on small number.
To make it simple, the Java Virtual Machine overloads the native libc
pow function. 
I submited a bug to Sun which has been accepted on their bugtracker:
http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6727254
We have a few leads to create a workaround on this.

Sylvestre





More information about the dev mailing list