NetBSD Problem Report #5459

Received: (qmail 23742 invoked from network); 17 May 1998 19:09:51 -0000
Message-Id: <199805171909.VAA02693@arresum.inka.de>
Date: Sun, 17 May 1998 21:09:51 +0200 (MEST)
From: Bernd Ernesti <pr199801@veego.de>
Reply-To: pr199801@veego.de
To: gnats-bugs@gnats.netbsd.org, bernd@arresum.inka.de
Subject: Failures from the rgression tests on an amiga with a 68040 CPU
X-Send-Pr-Version: 3.95

>Number:         5459
>Category:       port-amiga
>Synopsis:       Failures from the rgression tests on an amiga with a 68040 CPU
>Confidential:   no
>Severity:       critical
>Priority:       medium
>Responsible:    port-amiga-maintainer
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sun May 17 12:20:00 +0000 1998
>Closed-Date:    Mon Oct 27 03:11:57 +0000 2008
>Last-Modified:  Mon Oct 27 03:11:57 +0000 2008
>Originator:     Bernd Ernesti
>Release:        NetBSD-current 17-May-98
>Organization:

>Environment:

System: NetBSD arresum 1.3E NetBSD 1.3E (GENERIC) #1: Sun May 17 19:59:41 MEST 1998 bernd@arresum:/home/work/obj_amiga/src/sys/arch/amiga/compile/GENERIC amiga


>Description:
	A make regress dumpes core on an amiga with a 68040.
>How-To-Repeat:
1st core dump:

/src/regress/lib/libc/ieeefp> make regress
regress ===> except
./except
*** Signal 4
Stop.
*** Error code 1
Stop.

/src/regress/lib/libc/ieeefp/except> gdb except except.core
GDB is free software and you are welcome to distribute copies of it
 under certain conditions; type "show copying" to see the conditions.
There is absolutely no warranty for GDB; type "show warranty" for details.
GDB 4.16 (m68k-netbsd), Copyright 1996 Free Software Foundation, Inc...
Core was generated by `except'.
Program terminated with signal 4, Illegal instruction.
Reading symbols from /usr/libexec/ld.so...done.
Reading symbols from /usr/lib/libc.so.12.26...done.
#0  0x2aa4 in main () at /src/regress/lib/libc/ieeefp/except/except.c:59
59              x = one / zero;

2nd core dump:

/src/regress/sys/arch/m68k/060sp> make regress
Segmentation fault - core dumped
Testing 68060 ISP started:
        64-bit multiply...3 failed
        64-bit divide...passed
        cmp2,chk2...passed
        movep...*** Error code 1
Stop.

/src/regress/sys/arch/m68k/060sp> gdb dotest dotest.core
GDB is free software and you are welcome to distribute copies of it
 under certain conditions; type "show copying" to see the conditions.
There is absolutely no warranty for GDB; type "show warranty" for details.
GDB 4.16 (m68k-netbsd), Copyright 1996 Free Software Foundation, Inc...
Core was generated by `dotest'.
Program terminated with signal 11, Segmentation fault.
Reading symbols from /usr/libexec/ld.so...done.
Reading symbols from /usr/lib/libc.so.12.26...done.
#0  0x44f6 in 060ISP_TEST ()
(gdb) where
#0  0x44f6 in 060ISP_TEST ()
#1  0x414c in 060ISP_TEST ()
#2  0x280c in main () at /src/regress/sys/arch/m68k/060sp/dotest.c:31

>Fix:
	unknown.
>Release-Note:
>Audit-Trail:

From: Ignatios Souvatzis <is@jocelyn.rhein.de>
To: gnats-bugs@netbsd.org
Cc: bernd@arresum.inka.de
Subject: port-amiga/5459
Date: Tue, 16 Jun 1998 21:58:34 +0200 (MEST)

 The 2nd half of this has been fixed by not running the regression
 tests for the M68060 software support package installation if not
 executing on a M68060 cpu. [This affected all m68k ports.]

 (Apparently, the 68030 64bit divider and the 68040 64bit multiplier
 set something different from the 68060 emulation software, movep
 has some side effect on the 68040 that SEGVs the test program, and
 the software FPU rounding algorithm produces some loworder bits
 differently from the hardware.)
Responsible-Changed-From-To: gnats-admin->port-amiga-maintainer 
Responsible-Changed-By: fair 
Responsible-Changed-When: Mon Dec 28 09:34:32 PST 1998 
Responsible-Changed-Why:  
This PR is the responsibility of the portmaster, 
not the GNATS database administrator. 
State-Changed-From-To: open->feedback 
State-Changed-By: tv 
State-Changed-When: Sun Nov 5 20:29:17 PST 2000 
State-Changed-Why:  
This PR is from ages ago.  Does this still fail in regress? 
State-Changed-From-To: feedback->open 
State-Changed-By: is 
State-Changed-When: Tue Nov 21 12:22:28 PST 2000 
State-Changed-Why:  
yes. Well, during 1.5_ALPHA time, and there was no time to hunt it down. 
State-Changed-From-To: open->feedback
State-Changed-By: dholland@NetBSD.org
State-Changed-When: Thu, 27 Mar 2008 15:47:32 +0000
State-Changed-Why:
Is this still current?


From: Martin Husemann <martin@duskware.de>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: port-amiga/5459: Failures from the rgression tests on an amiga with a 68040 CPU
Date: Mon, 26 May 2008 14:06:44 +0200

 FWIW: a mac68k with 68040 passes all libc/ieeefp tests in -current.
 I think this PR can be closed.

 Martin

State-Changed-From-To: feedback->closed
State-Changed-By: dholland@NetBSD.org
State-Changed-When: Mon, 27 Oct 2008 03:11:57 +0000
State-Changed-Why:
Believed fixed; close at recommendation of martin@.


>Unformatted:

Home	PR Database Search