[Pnet-developers] [bugs #10483] Modulus (%) borked

[Rhys Weatherley]

This mail is an automated notification from the bugs tracker
 of the project: DotGNU Portable.NET.

/**************************************************************************/
[bugs #10483] Latest Modifications:

Changes by: 
                Rhys Weatherley <address@hidden>
'Date: 
                Thu 09/23/2004 at 22:26 (GMT)

            What     | Removed                   | Added
---------------------------------------------------------------------------
          Resolution | None                      | Fixed
              Status | Open                      | Closed


------------------ Additional Follow-up Comments ----------------------------
"fprem" is what should be used in any case, as I discovered
in libjit recently.  I copied "fprem1" from an older version 
of Mono a long time ago, and they must have gotten it wrong.  
They have since fixed it, but I didn't notice their fix.
Fix committed to CVS - 24 Sep 2004.






/**************************************************************************/
[bugs #10483] Full Item Snapshot:

URL: <http://savannah.gnu.org/bugs/?func=detailitem&item_id=10483>
Project: DotGNU Portable.NET
Submitted by: Carl-Adam Brengesjo
On: Thu 09/23/2004 at 16:13

Category:  None
Severity:  7 - Major
Item Group:  None
Resolution:  Fixed
Privacy:  Public
Assigned to:  None
Status:  Closed


Summary:  Modulus (%) borked

Original Submission:  This is is really borked operation...
A simple mod operation of 3 and 2 results in -1 !!

test.cs:
using System;

class Foo
{
  static void Main()
  {
    Double x1 = 3, x2 = 2, x3 = x1 % x2;
    Console.WriteLine("x3 = {0} = {1} % {2}", x3, x1, x2);
  }
}

pnet output:
x3 = -1 = 3 % 2

expected:
x3 = 1 = 3 % 2

Follow-up Comments
------------------


-------------------------------------------------------
Date: Thu 09/23/2004 at 22:26       By: Rhys Weatherley <rweather>
"fprem" is what should be used in any case, as I discovered
in libjit recently.  I copied "fprem1" from an older version 
of Mono a long time ago, and they must have gotten it wrong.  
They have since fixed it, but I didn't notice their fix.
Fix committed to CVS - 24 Sep 2004.

-------------------------------------------------------
Date: Thu 09/23/2004 at 17:44       By: Gopal.V <t3rmin4t0r>
I'm attaching my debug results and strangely this looks like a CPU bug ?.

See the float.log for the gdb logs which are clearly strange.
and the md_x86.c.hack for the changes need to start debugging
this problem

Using fprem instead of the IEEE fprem1 fixed the issue, but it's a sad day when 
FREM doesn't behave like IEEE float specs.








File Attachments
-------------------

-------------------------------------------------------
Date: Thu 09/23/2004 at 17:45  Name: md_x86.c.hack  Size: 424B   By: t3rmin4t0r
md_x86.c debug hacks
http://savannah.gnu.org/bugs/download.php?item_id=10483&amp;item_file_id=1699

-------------------------------------------------------
Date: Thu 09/23/2004 at 17:44  Name: float.log  Size: 4.81KB   By: t3rmin4t0r
gdb log
http://savannah.gnu.org/bugs/download.php?item_id=10483&amp;item_file_id=1698






For detailed info, follow this link:
<http://savannah.gnu.org/bugs/?func=detailitem&item_id=10483>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/