[nhc-bugs] nhc98 bignum div/mod gives wrong answers or segmentation faults

[Martin Guy]

Hi!
   I'm in the first month of translating my brain from mira to hs,
and seem to have come across an hnc98 bug: my app is the
famous infinite-precision scientific math library - see
http://freaknet.org/bignum - and while the hs version currently
has no known bugs under hugs, when I compile with nhc98
it gives me pi as 3.157..... while sqrt(2) gives a couple of
thousand correct decimal places and then segmentation faults.
Other calculations run ok.
   The problem only seems to occur in code that uses the
div/mod on Integers, so I'm assuming the problem is in that.

   I've trimmed down the sqrt2 example to a small source file
which is attached and obligingly continues to work under hugs
and dump core with nhc98. Unfortunately I don't have another
architecture than Linux/x86 to try it on - would someone check
whether it does the same on another system?
   I'm afraid that the wrong-answer of pi is rather hard to extract
into a self-contained fragment as it uses the infinite-precision
trig functions and they call in the rest of the world.

The system is Slackware Linux 9.0 (GCC 3.2.2) on Intel Pentium 3
(Coppermine) with 128MB RAM.
Hugs is hugs98 - November 2003 and nhc is nhc98-1.16
both compiled from source tarballs with no problems.

I'm online every day or so if there are turther tests I can run
to help out.

Blessings

   Martin Guy
-------------- next part --------------
A non-text attachment was scrubbed...
Name: s.hs
Type: application/octet-stream
Size: 1736 bytes
Desc: not available
Url : http://www.haskell.org//pipermail/nhc-bugs/attachments/20041028/5c9f1f10/s.obj