[MPlayer-users] Possible bug [FreeBSD 5.0-CURRENT]

[Per-Arne Holtmon Akø]

On Wed, Jul 03, 2002 at 23:20:47 +0000, Felix Buenemann wrote:
> [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> On Wednesday 03 July 2002 21:13, Per-Arne Holtmon Akø wrote:
> > > > > what cpu do you have and which values were used for -march, -mcpu?
> > > >
> > > > athlon-xp for both. I have Athlon MP's.
> > >
> > > afaik Atnlon XP and Athlon MP (that's the SMP model if I haven't mixed it
> > > up), are pretty much the same core, so I guess you are right and it's a
> > > compiler bug. You could use gdb + disass to trace the faulty
> > > instruction...
> >
> > If someone would mind holding my hand through the process I'm more than
> > willing to try this.
> DOCS/bugreports.html is your friend.

Ok, I did what was said to do for a reproducible crash, and I've
attached the output from the whole gdb session.

> > > For now you should use -march=i686 -mcpu=i686 as it will be faster, than
> > > omitting them at all.
> >
> > And that works as well. Not that I really care about my CPU usage when
> > I'm watching divx's, since the machine rarely do something else then.
> >
> > And you're right, the MP's are the SMP version, though the core isn't
> > exactly the same. I don't recall the differences right now, but they
> > exist, and it makes the MP's a bit faster in database and similiar
> > stuff.
> I doubt it's much more then bigger cache and some different registers, that 
> identify the CPU as SMP capable. At least that's how it works with Intel CPUs 
> (PIII vs PIII Xeon etc).

Might be, but I know that the Anandtech guys had a rather thorough and
nice review of the 760MP chipset and the MP's last summer, though I
ended up reading both the Parhelia-512 articles they had instead :)

-- 
vrak AKA Per-Arne Holtmon Akø
E-mail: perarneh at online.no
ICQ UIN: 9455554
HP: http://home.online.no/~perarneh/
-------------- next part --------------
Script started on Thu Jul  4 09:36:55 2002
GNU gdb 4.18 (FreeBSD)
Copyright 1998 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or 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.
This GDB was configured as "i386-unknown-freebsd"...Deprecated bfd_read called at /usr/src/contrib/gdb/gdb/dwarf2read.c line 3049 in dwarf2_read_section


Dwarf Error: Cannot handle DW_FORM_strp in DWARF reader.

(gdb) run -v -osdlevel 1 /stash/anime/dilldall/azumanga_op.avi
Starting program: /usr/local/src/mplayer/main/./mplayer -v -osdlevel 1 /stash/anime/dilldall/azumanga_op.avi
(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...
(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...
(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...
(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...
(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...
(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...


MPlayer CVS-020704-08:26-3.1 (C) 2000-2002 Arpad Gereoffy (see DOCS!)

CPU: Advanced Micro Devices Athlon 4 PM Palomino/Athlon MP Multiprocessor/Athlon XP eXtreme Performance (Family: 6, Stepping: 1)
3DNow supported but disabled
3DNowExt supported but disabled
CPUflags:  MMX: 1 MMX2: 1 3DNow: 0 3DNow2: 0 SSE: 0 SSE2: 0
Compiled for x86 CPU with extensions: MMX MMX2

/home/vrak/.mplayer/config(4): option: xineramascreen
/home/vrak/.mplayer/config(4): parameter: 0
(no debugging symbols found)...
Program received signal SIGILL, Illegal instruction.
0x806a47a in ?? ()
(gdb) br t
#0  0x806a47a in ?? ()
#1  0x806ad2c in ?? ()
#2  0x806b2d7 in ?? ()
#3  0x8062246 in ?? ()
#4  0x80626d1 in ?? ()
#5  0x80618f9 in ?? ()
(gdb) diss ass # $eip-32 $eip- +32
Dump of assembler code from 0x806a45a to 0x806a49a:
0x806a45a:	loopne 0x806a3dc
0x806a45c:	cmp    %al,(%eax)
0x806a45e:	jne    0x806a50b
0x806a464:	mov    0xffffffd4(%ebp),%eax
0x806a467:	mov    0xffffffb4(%ebp),%ecx
0x806a46a:	shl    $0x3,%eax
0x806a46d:	sub    0xffffffd4(%ebp),%eax
0x806a470:	shl    $0x2,%eax
0x806a473:	testb  $0x1,0xc(%ecx,%eax,1)
0x806a478:	je     0x806a492
0x806a47a:	cvtsi2ss %edx,%xmm0
0x806a47e:	flds   0x10(%ecx,%eax,1)
0x806a482:	movss  %xmm0,0xffffff94(%ebp)
0x806a487:	flds   0xffffff94(%ebp)
0x806a48a:	fxch   %st(1)
0x806a48c:	fucomip %st(1),%st
0x806a48e:	fstp   %st(0)
0x806a490:	ja     0x806a4f4
0x806a492:	mov    0xffffffd4(%ebp),%eax
0x806a495:	mov    0xffffffb4(%ebp),%ecx
0x806a498:	shl    $0x3,%eax
End of assembler dump.
(gdb) quit
The program is running.  Exit anyway? (y or n) y

Script done on Thu Jul  4 09:37:37 2002