aMule Forum

English => Backtraces => Topic started by: Guest on November 27, 2003, 07:17:55 PM

Title: Backtrace 1.2.0
Post by: Guest on November 27, 2003, 07:17:55 PM

[code:1]
Program received signal SIGPIPE, Broken pipe.
[Switching to Thread 16384 (LWP 19316)]
0x405ee4d8 in send () from /lib/libpthread.so.0
(gdb) bt full
#0  0x405ee4d8 in send () from /lib/libpthread.so.0
No symbol table info available.
#1  0x403a23d5 in _GSocket_Send_Stream () from /usr/lib/libwx_gtk-2.4.so
No symbol table info available.
#2  0x403a1e2a in GSocket_Write () from /usr/lib/libwx_gtk-2.4.so
No symbol table info available.
#3  0x4035b9a7 in wxSocketBase::_Write(void const*, unsigned) () from /usr/lib/libwx_gtk-2.4.so
No symbol table info available.
#4  0x4035b87a in wxSocketBase::Write(void const*, unsigned) () from /usr/lib/libwx_gtk-2.4.so
No symbol table info available.
#5  0x080cabac in CEMSocket::Send(char*, int, int) (this=0x8f169c8, lpBuf=0xbfffef30 >\016\001>, nBufLen=-32,
    nFlags=0) at EMSocket.cpp:342
        tosend = 58
        result = 4294967264
#6  0x080ca910 in CEMSocket::SendPacket(Packet*, bool, bool) (this=0x8f169c8, packet=0x917a848, controlpacket=true)
    at packets.h:47
        delpacket = Cannot access memory at address 0x0
[/code:1]

Title: Re: Backtrace 1.2.0
Post by: pure_ascii on November 27, 2003, 09:32:39 PM

I remember from lMule times that SIGPIPE should not hurt. So you can tell gdb not to care about it. Just enter [code:1]ha SIGPIPE nostop noprint pass[/code:1] inside gdb before calling "run".

HTH, PA.

Title: Re: Backtrace 1.2.0
Post by: Guest on November 27, 2003, 11:22:54 PM

but it's crashing? when i'm not running gdb it always crashes after some minutes.
i will try your hint, maybe it's something different that makes it crash