This is the mail archive of the
java@gcc.gnu.org
mailing list for the Java project.
Re: [freenet-devl] Re: Yet another GCJ update. (a/k/a PR 1615followup)
- To: Per Bothner <per at bothner dot com>
- Subject: Re: [freenet-devl] Re: Yet another GCJ update. (a/k/a PR 1615followup)
- From: "Mark J. Roberts" <mjr at statesmean dot com>
- Date: Fri, 23 Feb 2001 18:26:36 -0500 (EST)
- cc: Bryce McKinlay <bryce at albatross dot co dot nz>, <java at gcc dot gnu dot org>
On 23 Feb 2001, Per Bothner wrote:
> I can't recall checking this in, or promising to do so.
Yeah, my bad.
> The patch seems plausible, but it is not in the gnu.math.IntNum version.
> Before I bless the patch, could you give me (again, sorry) which
> division is giving the wrong result, so I can check it against gnu.math
> as used by Kawa? (You've posted some pirs of numbers to divide, and the
> ones I tried seemed to give the correct result. So I'd like to understand
> better the need for the patch.)
That was an entirely different problem, which happened because I stupidly
broke my MPN.lshift.
This patch fixes a situation where BigInteger.divide generates bogus
remainders and quotients for certain input. I posted a test case for this,
you can see it at http://gcc.gnu.org/ml/java/2001-02/msg00293.html. This
patch fixes that bug:
mjr::chunga$ gcj -o tj Test.java --main=Test
mjr::chunga$ ./tj
367417105895394016159154787313270662212024192865627354929903779273790475448015789586251477566499147933827828993175451736204011419593248732059558975428536352923147799454938474886826728465908014291590615212928094348062380180678090626190405785338510862566219882367271848354528305975332648128732693695418736716973797590695732844596436204184266800584120723714967834374864490790434309625243649395288704661308708314268453615305198754248602412105555586454917161
--
Mark Roberts
mjr@statesmean.com