[MPlayer-users] [BUG] Realmedia crash/premature exit
Glyn
glyn.kennington at hertford.oxford.ac.uk
Wed Jun 19 13:39:01 CEST 2002
I know that the realmedia support is still alpha, and people are
reporting problems left right and centre, so I don't know how redundant
this bugreport may be, but there seems to be some mis-parsing of the
packet sizes, that causes some .rm files to exit before reaching the end
of the stream. Two specific ones are:
http://www.queenadreena.com/www2/mutability/movement/heavenlyT1.rm
http://www.queenadreena.com/www2/mutability/movement/prettypollyT1.rm
(The same fault occurs with all -vo and -ao that I've tried, including
null.) Because I think the problem is to do with the packets, I've
recorded logs using -v -v, so the resulting files are huge. I've
uploaded them to incoming/queenadreena/{heavenly,prettypolly}.log.gz .
prettypollyT1.rm causes a segfault, so here's the gdb session:
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1024 (LWP 4139)]
0x4039b197 in memcpy () from /lib/libc.so.6
(gdb) bt
#0 0x4039b197 in memcpy () from /lib/libc.so.6
#1 0x08125c92 in demux_real_fill_buffer (demuxer=0x82c0630) at stream.h:141
#2 0x0811cd09 in ds_fill_buffer (ds=0x8280608) at demuxer.c:342
#3 0x0811d07d in ds_get_packet (ds=0x8280608, start=0xbfffe8a0) at demuxer.c:416
#4 0x0811681c in video_read_frame (sh_video=0x827f248, frame_time_ptr=0xbfffe8c4, start=0xbfffe8a0, force_fps=0) at video.c:269
#5 0x08064e80 in main (argc=4, argv=0xbffffa84, envp=0xbffffa98) at mplayer.c:1572
(gdb) disass $eip-32 $eip+32
Dump of assembler code from 0x4039b177 to 0x4039b1b7:
0x4039b177 <memcpy+7>: or %cl,0xd7891045(%ebx)
0x4039b17d <memcpy+13>: mov 0xc(%ebp),%esi
0x4039b180 <memcpy+16>: cmp $0x7,%eax
0x4039b183 <memcpy+19>: jbe 0x4039b1a0 <memcpy+48>
0x4039b185 <memcpy+21>: mov %edx,%ecx
0x4039b187 <memcpy+23>: neg %ecx
0x4039b189 <memcpy+25>: and $0x3,%ecx
0x4039b18c <memcpy+28>: sub %ecx,%eax
0x4039b18e <memcpy+30>: cld
0x4039b18f <memcpy+31>: repz movsb %ds:(%esi),%es:(%edi)
0x4039b191 <memcpy+33>: mov %eax,%ecx
0x4039b193 <memcpy+35>: shr $0x2,%ecx
0x4039b196 <memcpy+38>: cld
0x4039b197 <memcpy+39>: repz movsl %ds:(%esi),%es:(%edi)
0x4039b199 <memcpy+41>: and $0x3,%eax
0x4039b19c <memcpy+44>: lea 0x0(%esi,1),%esi
0x4039b1a0 <memcpy+48>: mov %eax,%ecx
0x4039b1a2 <memcpy+50>: cld
0x4039b1a3 <memcpy+51>: repz movsb %ds:(%esi),%es:(%edi)
0x4039b1a5 <memcpy+53>: mov %edx,%eax
0x4039b1a7 <memcpy+55>: pop %esi
0x4039b1a8 <memcpy+56>: pop %edi
0x4039b1a9 <memcpy+57>: leave
0x4039b1aa <memcpy+58>: ret
0x4039b1ab <memcpy+59>: nop
0x4039b1ac <memcpy+60>: lea 0x0(%esi,1),%esi
0x4039b1b0 <memcpy+64>: push %ebp
0x4039b1b1 <memcpy+65>: mov %esp,%ebp
0x4039b1b3 <memcpy+67>: push %edi
0x4039b1b4 <memcpy+68>: push %esi
0x4039b1b5 <memcpy+69>: push %ebx
0x4039b1b6 <memcpy+70>: call 0x4039b1bb <memcpy+75>
End of assembler dump.
(gdb)
System info:
Distro - Debian Woody
$ uname -a
Linux corrosive 2.4.18 #1 Wed Mar 6 21:49:14 GMT 2002 i586 unknown
$ ls -l /lib/libc[.-]*
-rwxr-xr-x 1 root root 1153784 Apr 28 10:57 /lib/libc-2.2.5.so
lrwxrwxrwx 1 root root 13 May 12 17:07 /lib/libc.so.6 -> libc-2.2.5.so
$ gcc -v
Reading specs from /usr/lib/gcc-lib/i386-linux/3.0.4/specs
Configured with: ../src/configure -v --enable-languages=c,c++,java,f77,proto,objc --prefix=/usr --infodir=/share/info --mandir=/share/man --enable-shared --with-gnu-as --with-gnu-ld --with-system-zlib --enable-long-long --enable-nls --without-included-gettext --disable-checking --enable-threads=posix --enable-java-gc=boehm --with-cpp-install-dir=bin --enable-objc-gc i386-linux
Thread model: posix
gcc version 3.0.4
$ ld -v
GNU ld version 2.12.90.0.1 20020307 Debian/GNU Linux
$ as --version
GNU assembler 2.12.90.0.1 20020307 Debian/GNU Linux
Copyright 2002 Free Software Foundation, Inc.
This program is free software; you may redistribute it under the terms of
the GNU General Public License. This program has absolutely no warranty.
This assembler was configured for a target of `i386-linux'.
--
There are pavements on New Inn Hall Street. Down both sides, in fact.
More information about the MPlayer-users
mailing list