[Bug c/25332] New: 64 bit multiplication incorrectly folded.

2005-12-09 Thread jyates at netezza dot com 
/*
 * Compile -O0 (no optimization)
 */

#include 

typedef int int32;
typedef long long int64;


int main ()
{
int32 i = 1;/* hence (2*i)-1 below is 1... */
#define K 0x80003LL /* 64 bit constant; low 32 are 3; high 32 are 8
*/

printf("%016llx (gcc 2.96 bad:  0003)\n", K * ( (2*i) -
   1  ));
printf("%016llx (gcc 2.96 bad:  0003)\n", K * ( (2*i) -
   1LL));
printf("%016llx (gcc 2.96 bad:  0003)\n", K * ( (2*i) -
(int64)1  ));

printf("%016llx (gcc 2.96 bad:  fff80003)\n", K * (  (int64)(2*i) -
   1  ));
printf("%016llx (gcc 2.96 bad:  fff80003)\n", K * (  (int64)(2*i) -
   1LL));
printf("%016llx (gcc 2.96 bad:  fff80003)\n", K * (  (int64)(2*i) -
(int64)1  ));

printf("%016llx (gcc 2.96 good: 00080003)\n", K * (2*(int64)(  i) -
   1  ));
}


-- 
   Summary: 64 bit multiplication incorrectly folded.
   Product: gcc
   Version: 2.96
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: c
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: jyates at netezza dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25332

[Bug c/25332] 64 bit multiplication incorrectly folded.

2005-12-09 Thread jyates at netezza dot com 


--- Comment #2 from jyates at netezza dot com  2005-12-09 22:42 ---
Subject: RE:  64 bit multiplication incorrectly folded.

Is it possible to determine what commit between 3.3.3 and 3.4.0
actually fixed the bug?  I have a very strong motivation to see
if I can propagate the fix back to the 2.96 code base.

/john

-Original Message-
From: pinskia at gcc dot gnu dot org [mailto:[EMAIL PROTECTED]
Sent: Friday, December 09, 2005 4:38 PM
To: John Yates
Subject: [Bug c/25332] 64 bit multiplication incorrectly folded.




--- Comment #1 from pinskia at gcc dot gnu dot org  2005-12-09 21:37
---
Fixed in 3.4.0 and above, since 3.3.x and below are no longer being maintained
closing as fixed.


-- 

pinskia at gcc dot gnu dot org changed:

   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
   Keywords||wrong-code
  Known to fail||3.3.3 3.2.3 3.0.4
  Known to work||3.4.0 4.0.0 4.1.0 2.95.3
 Resolution||FIXED
   Target Milestone|--- |3.4.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25332

--- You are receiving this mail because: ---
You reported the bug, or are watching the reporter.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25332