Re: [freenet-devl] Re: Yet another GCJ update. (a/k/a PR 1615followup)

["Mark J. Roberts" <mjr at statesmean dot com>]

On 23 Feb 2001, Per Bothner wrote:

> 2000-09-21  Per Bothner  <per@bothner.com>
>
> 	* IntNum.java (divide):  Zero-extend results if high-order bit set.
>
> So I believe your patch is basically correct.  However, I think it should
> be done before the "add_one" calculation, as in my patch.  Sorry for
> not noticing this change earlier - and for that matter, not submitting
> the change when I fixed the gnu.math version.

This patch breaks BigInteger.divide:

import BigInteger;

public class Test {
    public static void main(String[] args) {
        BigInteger x = new BigInteger("12345678901234567890123456789012345678901234567890");
        BigInteger y = new BigInteger("123451234561234561234561234566");
        System.out.println(x.divide(y));
    }
}


mjr::mjr$ javac Test.java BigInteger.java MPN.java ; java -cp . Test
Exception in thread "main" java.lang.ArrayIndexOutOfBoundsException: 30
        at BigInteger.divide(BigInteger.java:814)
        at BigInteger.divide(BigInteger.java:910)
        at Test.main(Test.java:39)

(same result with GCJ, just uglier stacktrace)

The exception happens at ywords[rlen] = 0; in your patch.


-- 
Mark Roberts
mjr@statesmean.com