*, /, and .im of complex numbers is wrong at compile time.

DMD 0.138, Windows.
It works for add + subtract, but returns 0 (with the correct sign) whenever * or / is involved (BUG 1-4).
I think the root cause lies with ireal.im, which behaves _very_ oddly.
(BUG 5 and BUG 6) Looks like it's grabbing the wrong part of the syntax tree.

BUG 1-------------------------------
static assert(1i * 1i == -1);
bug.d(1): static assert  (-0 == -1) is false

BUG 2-------------------------------
static assert(3.0i /1.0i == 3);
bug.d(1): static assert  (0 == 3) is false

BUG 3-------------------------------
static assert(3.0i *2 == 6i);
bug.d(1): static assert  (0i == 6i) is false

BUG 4-------------------------------
static assert((2.0i).im == 2.0);
bug.d(1): static assert  (2i == 2) is false

BUG 5------------------------------
static assert((3+2.0i).im == 2.0);
bug.d(1): static assert  (3 == 2) is false

November 12, 2005

Re: *, /, and .im of complex numbers is wrong at compile time.

Posted by Thomas Kuehne
in reply to Don Clugston

Permalink

Thomas Kuehne

Posted in reply to Don Clugston

Attachments:

pgp.sig

Permalink

Don Clugston schrieb am 2005-11-08:
> DMD 0.138, Windows.
> It works for add + subtract, but returns 0 (with the correct sign)
> whenever * or / is involved (BUG 1-4).
> I think the root cause lies with ireal.im, which behaves _very_ oddly.
> (BUG 5 and BUG 6) Looks like it's grabbing the wrong part of the syntax
> tree.
>
> BUG 1-------------------------------
> static assert(1i * 1i == -1);
> bug.d(1): static assert  (-0 == -1) is false
>
> BUG 2-------------------------------
> static assert(3.0i /1.0i == 3);
> bug.d(1): static assert  (0 == 3) is false
>
> BUG 3-------------------------------
> static assert(3.0i *2 == 6i);
> bug.d(1): static assert  (0i == 6i) is false
>
> BUG 4-------------------------------
> static assert((2.0i).im == 2.0);
> bug.d(1): static assert  (2i == 2) is false
>
> BUG 5------------------------------
> static assert((3+2.0i).im == 2.0);
> bug.d(1): static assert  (3 == 2) is false

Added to DStress as http://dstress.kuehne.cn/run/c/cdouble_02_A.d http://dstress.kuehne.cn/run/c/cdouble_02_B.d http://dstress.kuehne.cn/run/c/cdouble_02_C.d http://dstress.kuehne.cn/run/c/cfloat_03_A.d http://dstress.kuehne.cn/run/c/cfloat_03_B.d http://dstress.kuehne.cn/run/c/cfloat_03_C.d http://dstress.kuehne.cn/run/c/creal_27_A.d http://dstress.kuehne.cn/run/c/creal_27_B.d http://dstress.kuehne.cn/run/c/creal_27_C.d http://dstress.kuehne.cn/run/o/opAdd_07_A.d http://dstress.kuehne.cn/run/o/opAdd_07_B.d http://dstress.kuehne.cn/run/o/opDiv_12_A.d http://dstress.kuehne.cn/run/o/opDiv_12_B.d http://dstress.kuehne.cn/run/o/opMul_08_A.d http://dstress.kuehne.cn/run/o/opMul_08_B.d http://dstress.kuehne.cn/run/o/opMul_08_C.d http://dstress.kuehne.cn/run/o/opMul_08_D.d http://dstress.kuehne.cn/run/o/opMul_08_E.d http://dstress.kuehne.cn/run/o/opSub_07_A.d http://dstress.kuehne.cn/run/o/opSub_07_B.d

Thomas

Forums