NetBSD Problem Report #18769

Received: (qmail 25033 invoked by uid 605); 22 Oct 2002 19:17:39 -0000
Message-Id: <20021022191738.147AD7B14@yeah-baby.shagadelic.org>
Date: Tue, 22 Oct 2002 12:17:38 -0700 (PDT)
From: thorpej@shagadelic.org
Sender: gnats-bugs-owner@netbsd.org
Reply-To: thorpej@shagadelic.org
To: gnats-bugs@gnats.netbsd.org
Subject: IEEE FP rounding bug NetBSD/alpha
X-Send-Pr-Version: 3.95

>Number:         18769
>Category:       port-alpha
>Synopsis:       IEEE FP rounding bug NetBSD/alpha
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    thorpej
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Tue Oct 22 19:18:00 +0000 2002
>Closed-Date:    Sat Apr 17 23:33:32 +0000 2021
>Last-Modified:  Sat Apr 17 23:33:32 +0000 2021
>Originator:     Jason R Thorpe
>Release:        NetBSD 1.6I
>Organization:
Wasabi Systems, Inc.
>Environment:


System: NetBSD frau-farbissina.shagadelic.org 1.6I NetBSD 1.6I (FRAU-FARBISSINA) #149: Sat Oct 12 12:39:17 PDT 2002     thorpej@frau-farbissina.shagadelic.org:/u2/netbsd/src/sys/arch/alpha/compile/FRAU-FARBISSINA alpha
Architecture: alpha
Machine: alpha
>Description:
	NetBSD/alpha appears to have an IEEE FP rounding bug in the
	kernel FP completion code.  This bug causes failures in the
	GCC 3.3 testsuite.

>How-To-Repeat:
	Compile the following program with "cc -mieee".  The optimization
	level is irrelevant.

double
d (unsigned long long k)
{
  double x;

  x = (double) k;
  return x;
}

int
main (int argc, char *argv[])
{
  unsigned long long k;
  double x;

  k = 0x8693ba6d7d220401ULL;
  x = d (k);
  k = (unsigned long long) x;
  if (k != 0x8693ba6d7d220800ULL)
    abort ();
}

	This code is taken from the relevant GCC testcase.  The
	testcase passes when compiled *without* "cc -mieee".

>Fix:
	Not known at this time.  (More analysis of the problem needs
	to be done.)
>Release-Note:
>Audit-Trail:
From: Adam Hoka <adam.hoka@gmail.com>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: port-alpha/18769
Date: Sat, 19 Jan 2008 19:02:15 +0100

 The test code still fails on NetBSD 4.0 RC1.

From: Felix Deichmann <m4j0rd0m0@gmail.com>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: port-alpha/18769
Date: Fri, 16 Dec 2016 18:34:51 +0100

 FWIW, the test still fails on NetBSD-current 7.99.50:
 gcc (nb1 20160606) 5.4.0

State-Changed-From-To: open->feedback
State-Changed-By: maya@NetBSD.org
State-Changed-When: Tue, 13 Oct 2020 03:57:45 +0000
State-Changed-Why:
This sounds a lot like 52520. Is this still an issue?


Responsible-Changed-From-To: port-alpha-maintainer->thorpej
Responsible-Changed-By: thorpej@NetBSD.org
Responsible-Changed-When: Tue, 13 Oct 2020 15:45:07 +0000
Responsible-Changed-Why:
Take.


State-Changed-From-To: feedback->open
State-Changed-By: maya@NetBSD.org
State-Changed-When: Sat, 17 Apr 2021 23:09:58 +0000
State-Changed-Why:
Not going to keep pingign about this.


State-Changed-From-To: open->closed
State-Changed-By: thorpej@NetBSD.org
State-Changed-When: Sat, 17 Apr 2021 23:33:32 +0000
State-Changed-Why:
I'll just assume this is a dup of 52520, and push "run GCC test suite"
onto the project queue.


>Unformatted:

Home	PR Database Search