[Aldor-l] Bigint bug(!)

[Stephen Watt]

Such a huge list of bugs is very surprising.   Which C compiler are you using?
It may be as Gaby suggested that more aggressive optimization is
changing the behaviour of code when ints wrap around.  -- Stephen

2008/8/16  <pip88nl at gmail.com>:
> On Sat, Aug 16, 2008 at 01:39:03AM +0200, Ralf Hemmecke wrote:
>> I wonder what a Aldor program might be that hits this bug.
>
> Many.. here is a list of tests that trigger it:
>
> bugs/bug969.as
> bugs/bug1021.as
> bugs/bug1022.as
> bugs/bug1027.as
> bugs/bug1044.as
> bugs/bug1055.as
> bugs/bug1061.as
> bugs/bug1063.as
> bugs/bug1077.as
> bugs/bug1090.as
> bugs/bug1110.as
> bugs/bug1121.as
> bugs/bug1130.as
> bugs/bug1178.as
> bugs/bug1211.as
> bugs/bug1220.as
> bugs/bug1224.as
> bugs/bug1250.as
> library/axllib/bigmand.as
> library/axllib/bug969.as
> library/axllib/bug1165.as
> library/axllib/bug1172.as
> library/axllib/bug1278.as
> library/axllib/cascade4.as
> library/axllib/collect1.as
> library/axllib/condapply2.as
> library/axllib/const3.as
> library/axllib/const4.as
> library/axllib/const5.as
> library/axllib/cycle0.as
> library/axllib/defarg9.as
> library/axllib/float1.as
> library/axllib/float3.as
> library/axllib/float4.as
> library/axllib/format1.as
> library/axllib/format2.as
> library/axllib/funlist1.as
> library/axllib/genops.as
> library/axllib/grok.as
> library/axllib/has2.as
> library/axllib/imod0.as
> library/axllib/inline3.as
> library/axllib/object0.as
> library/axllib/opt0.as
> library/axllib/rawrec2.as
> library/axllib/rawrec3.as
> library/axllib/rawrec4.as
> library/axllib/scobind4.as
> library/axllib/scope11.as
> library/axllib/select0.as
> library/axllib/select1.as
> library/axllib/swap1.as
> library/axllib/swap.as
> library/axllib/t1059.as
> library/axllib/t1145.as
> library/axllib/table1.as
> library/axllib/tinfer0.as
> library/axllib/type1.as
> library/axllib/type2.as
> library/axllib/where1.as
>
> This list shows that the code that triggers this bug is not inside the
> aldor program itself. bug969.as doesn't use big integer arithmetic, only
> SingleInteger. The bug occurs somewhere else, deeper in the runtime. Oh
> well.. more about this later.
>
> Pippijn
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.6 (GNU/Linux)
>
> iD4DBQFIplJDJc+zqGNdDgoRAq4jAJ9FwLkLp4B+azLnhEF+2QGRlcOU9QCY9xF0
> pV1tx7LfYj1wUof2TyAcdQ==
> =m/J1
> -----END PGP SIGNATURE-----
>
> _______________________________________________
> Aldor-l mailing list
> Aldor-l at aldor.org
> http://aldor.org/mailman/listinfo/aldor-l_aldor.org
>
>