www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.bugs - mistake in the calculation when compiling

reply Hiroshi Sakurai <Hiroshi_member pathlink.com> writes:
Hi.
This topic writen U in Japanese D Language Wiki.
http://f17.aaa.livedoor.jp/~labamba/?BugTrack%2F19

foo(-288*1000/360) is not -800.



void main(char[][] arg) {
printf("ok %d\n",  3/3);// 1
printf("ok %d\n", -(3/3));// -1

printf("ng %d\n", -3/3);// 1431655764
int a = cast(int)(-3/3);
printf("ng %d\n", a);// 1431655764
printf("ng %d\n", (-3)/3);// 1431655764
}
Aug 24 2005
next sibling parent reply zwang <nehzgnaw gmail.com> writes:
Hiroshi Sakurai wrote:

 Hi.
 This topic writen U in Japanese D Language Wiki.
 http://f17.aaa.livedoor.jp/~labamba/?BugTrack%2F19
 
 foo(-288*1000/360) is not -800.
 
 
 
 void main(char[][] arg) {
 printf("ok %d\n",  3/3);// 1
 printf("ok %d\n", -(3/3));// -1
 
 printf("ng %d\n", -3/3);// 1431655764
 int a = cast(int)(-3/3);
 printf("ng %d\n", a);// 1431655764
 printf("ng %d\n", (-3)/3);// 1431655764
 }
 
 


Already reported. See D.bugs:4750.
Aug 24 2005
parent reply Hasan Aljudy <hasan.aljudy gmail.com> writes:
zwang wrote:

 Hiroshi Sakurai wrote:
 

 Hi.
 This topic writen U in Japanese D Language Wiki.
 http://f17.aaa.livedoor.jp/~labamba/?BugTrack%2F19

 foo(-288*1000/360) is not -800.



 void main(char[][] arg) {
 printf("ok %d\n",  3/3);// 1
 printf("ok %d\n", -(3/3));// -1

 printf("ng %d\n", -3/3);// 1431655764
 int a = cast(int)(-3/3);
 printf("ng %d\n", a);// 1431655764
 printf("ng %d\n", (-3)/3);// 1431655764
 }


 
 Already reported. See D.bugs:4750.


but this is kind of different, it's not treated as positive (or atleast 
doesn't seem like that)

funny thing is, -2/2 works fine!!
and so does -3/1

it's kind of odd .. in some cases it works, some cases it doesn't .. 
maybe an optamization bug?

<bug.d>
void main(char[][] arg)
{
     printf("-5/3 prints: %d\n", -5/3);
     printf("-5/2 prints: %d\n", -5/2);
     printf("-7/3 prints: %d\n", -7/3);
     printf("-7/4 prints: %d\n", -7/4);
     printf("-7/7 prints: %d\n", -7/7);
     printf("-8/7 prints: %d\n", -8/7);
     printf("-9/7 prints: %d\n", -9/7);
     printf("-11/8 prints: %d\n", -11/8);
     printf("-7/9 prints: %d\n", -7/9);
}
</bug.d>


output:

D:\Documents and Settings\Aljudy\temp>dmd bug
E:\dmd\dmd\bin\..\..\dm\bin\link.exe bug,,,user32+kernel32/noi;

D:\Documents and Settings\Aljudy\temp>bug
-5/3 prints: 1431655763
-5/2 prints: -3
-7/3 prints: 1431655763
-7/4 prints: -2
-7/7 prints: -1840700271
-8/7 prints: -1840700271
-9/7 prints: -1840700271
-11/8 prints: -2
-7/9 prints: 1908874353
Aug 24 2005
parent Hiroshi Sakurai <Hiroshi_member pathlink.com> writes:
zwang wrote:

 Hiroshi Sakurai wrote:
 

 Hi.
 This topic writen U in Japanese D Language Wiki.
 http://f17.aaa.livedoor.jp/~labamba/?BugTrack%2F19

 foo(-288*1000/360) is not -800.



 void main(char[][] arg) {
 printf("ok %d\n",  3/3);// 1
 printf("ok %d\n", -(3/3));// -1

 printf("ng %d\n", -3/3);// 1431655764
 int a = cast(int)(-3/3);
 printf("ng %d\n", a);// 1431655764
 printf("ng %d\n", (-3)/3);// 1431655764
 }


 
 Already reported. See D.bugs:4750.


but this is kind of different, it's not treated as positive (or atleast 
doesn't seem like that)

funny thing is, -2/2 works fine!!
and so does -3/1

it's kind of odd .. in some cases it works, some cases it doesn't .. 
maybe an optamization bug?

<bug.d>
void main(char[][] arg)
{
     printf("-5/3 prints: %d\n", -5/3);
     printf("-5/2 prints: %d\n", -5/2);
     printf("-7/3 prints: %d\n", -7/3);
     printf("-7/4 prints: %d\n", -7/4);
     printf("-7/7 prints: %d\n", -7/7);
     printf("-8/7 prints: %d\n", -8/7);
     printf("-9/7 prints: %d\n", -9/7);
     printf("-11/8 prints: %d\n", -11/8);
     printf("-7/9 prints: %d\n", -7/9);
}
</bug.d>


output:

D:\Documents and Settings\Aljudy\temp>dmd bug
E:\dmd\dmd\bin\..\..\dm\bin\link.exe bug,,,user32+kernel32/noi;

D:\Documents and Settings\Aljudy\temp>bug
-5/3 prints: 1431655763
-5/2 prints: -3
-7/3 prints: 1431655763
-7/4 prints: -2
-7/7 prints: -1840700271
-8/7 prints: -1840700271
-9/7 prints: -1840700271
-11/8 prints: -2
-7/9 prints: 1908874353


Thank you for having already taught what the bug has reported. 
I am not reading all Newsgroups. 
Therefore, I did not notice it. 

I thought that it was interesting that -2/2 had moved well.
I also thought that it might have happened while the bug had optimized the code.


Hiroshi Sakurai.
Aug 27 2005
prev sibling parent =?UTF-8?B?VGhvbWFzIEvDvGhuZQ==?= <thomas-dloop kuehne.cn> writes:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hiroshi Sakurai schrieb:


 Hi.
 This topic writen U in Japanese D Language Wiki.
 http://f17.aaa.livedoor.jp/~labamba/?BugTrack%2F19
 
 foo(-288*1000/360) is not -800.
 
 
 
 void main(char[][] arg) {
 printf("ok %d\n",  3/3);// 1
 printf("ok %d\n", -(3/3));// -1
 
 printf("ng %d\n", -3/3);// 1431655764
 int a = cast(int)(-3/3);
 printf("ng %d\n", a);// 1431655764
 printf("ng %d\n", (-3)/3);// 1431655764
 }


Added to DStress as
http://dstress.kuehne.cn/run/o/opDiv_10_A.d
http://dstress.kuehne.cn/run/o/opDiv_10_B.d

Thomas

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

iD4DBQFDEJss3w+/yD4P9tIRAiTUAJi0XU4grbgT6BRLeg7ew3+WVTviAKCWEyBI
1G6VcZpxQGO4+qnUN0s3Og==
=MlOX
-----END PGP SIGNATURE-----
Aug 27 2005