www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.bugs - Large integer division

reply Deewiant <deewiant.doesnotlike.spam gmail.com> writes:
Dividing really big ulongs by each other doesn't seem to work at all.

---
void main() {
	ulong a = 10_000_000_000_000_000,
	      b =  1_000_000_000_000_000;

	assert (a / b == 10);
}
---

a / b is, according to writef, 9223372036854775818.
Dec 06 2005
next sibling parent reply "Kris" <fu bar.com> writes:
Confirmed:  _ULDIV appears to be faulty.


"Deewiant" <deewiant.doesnotlike.spam gmail.com> wrote in message 
news:dn3old$hhr$1 digitaldaemon.com...

 Dividing really big ulongs by each other doesn't seem to work at all.

 ---
 void main() {
 ulong a = 10_000_000_000_000_000,
       b =  1_000_000_000_000_000;

 assert (a / b == 10);
 }
 ---

 a / b is, according to writef, 9223372036854775818.
Dec 06 2005
parent Dave <Dave_member pathlink.com> writes:
In article <dn4l2u$1joa$1 digitaldaemon.com>, Kris says...

Confirmed:  _ULDIV appears to be faulty.


"Deewiant" <deewiant.doesnotlike.spam gmail.com> wrote in message 
news:dn3old$hhr$1 digitaldaemon.com...

 Dividing really big ulongs by each other doesn't seem to work at all.

 ---
 void main() {
 ulong a = 10_000_000_000_000_000,
       b =  1_000_000_000_000_000;

 assert (a / b == 10);
 }
 ---

 a / b is, according to writef, 9223372036854775818. 




There looks to be a problem with longs as well (__LDIV__).
Dec 06 2005
prev sibling next sibling parent reply =?ISO-8859-15?Q?Anders_F_Bj=F6rklund?= <afb algonet.se> writes:
Deewiant wrote:


 Dividing really big ulongs by each other doesn't seem to work at all.
 
 ---
 void main() {
 	ulong a = 10_000_000_000_000_000,
 	      b =  1_000_000_000_000_000;
 
 	assert (a / b == 10);
 }
 ---
 
 a / b is, according to writef, 9223372036854775818.


FWIW, Both "long" and "ulong" works fine with GDC

import std.stdio;

int main() {
         ulong a = 10_000_000_000_000_000,
               b =  1_000_000_000_000_000;

         assert (a / b == 10);
         debug writefln(a / b);

         long c = -42_000_000_000_000_000,
              d =   1_000_000_000_000_000;

         assert (c / d == -42);
         debug writefln(c / d);

         return 0;
}

prints:
10
42

--anders
Dec 07 2005
parent =?ISO-8859-15?Q?Anders_F_Bj=F6rklund?= <afb algonet.se> writes:
 prints:
 10
 42


<duh> -42

--anders
Dec 07 2005
prev sibling next sibling parent "Walter Bright" <newshound digitalmars.com> writes:
"Deewiant" <deewiant.doesnotlike.spam gmail.com> wrote in message
news:dn3old$hhr$1 digitaldaemon.com...

 Dividing really big ulongs by each other doesn't seem to work at all.


Criminy! You're right. I have it fixed, it'll go out in the next update.
Dec 08 2005
prev sibling parent Thomas Kuehne <thomas-dloop kuehne.cn> writes:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Deewiant schrieb am 2005-12-06:

 Dividing really big ulongs by each other doesn't seem to work at all.

 ---
 void main() {
 	ulong a = 10_000_000_000_000_000,
 	      b =  1_000_000_000_000_000;

 	assert (a / b == 10);
 }
 ---

 a / b is, according to writef, 9223372036854775818.


Added to DStress as
http://dstress.kuehne.cn/run/o/opDiv_13_A1.d
http://dstress.kuehne.cn/run/o/opDiv_13_A2.d
http://dstress.kuehne.cn/run/o/opDiv_13_B1.d
http://dstress.kuehne.cn/run/o/opDiv_13_B2.d
http://dstress.kuehne.cn/run/o/opDiv_13_C1.d
http://dstress.kuehne.cn/run/o/opDiv_13_C2.d
http://dstress.kuehne.cn/run/o/opDiv_13_D1.d
http://dstress.kuehne.cn/run/o/opDiv_13_D2.d

Thomas


-----BEGIN PGP SIGNATURE-----

iD8DBQFDmlgE3w+/yD4P9tIRAjmDAJ44kwQO0J1E8V2rPCbUHjUpKa/3kQCgz2cc
GW1Cne6oEZg5maUsuSEt8WU=
=gB7+
-----END PGP SIGNATURE-----
Dec 09 2005