[bf1942] 1.5 linux server is a no go :(

[Steven Hartland]

Ok patched truss to print the args of linux_access giving:
linux_access("/home/mpukgame/.terminfo/v/vt100",0x4,04) ERR#2 'No such file or directory'
linux_access("/usr/share/terminfo/v/vt100",0x4,04) ERR#2 'No such file or directory'
Tracked this down to debian base symlinking:
/usr/share/terminfo/v/vt100 -> /etc/terminfo/v/vt100
but since these are /compat/linux based looks like link resolving is not quite up
to scratch.
Fixed by replacing it with:
ln -s /usr/compat/linux/etc/terminfo/v/vt100 /usr/compat/linux/usr/share/terminfo/v/vt100

Now it gets passed that but then segv's :(

Core was generated by `bf1942_lnxded.stati'.
Program terminated with signal 6, Aborted.
#0  0x08467ed7 in sigsuspend ()
(gdb) bt
#0  0x08467ed7 in sigsuspend ()
#1  0xbfbfe520 in ?? ()
#2  0x0845ed3c in __pthread_wait_for_restart_signal ()
#3  0x0845e73f in pthread_create ()
#4  0x083029d2 in dice::ref2::io::NetServer::init(std::string const&, unsigned, bool) ()
Cannot access memory at address 0x8f7e

I started without +statusMonitor 1 just to check its a no go :(

Here's the truss of the crash:
read(0x69,0x9401000,0x10000)                     = 29677 (0x73ed)       
read(0x69,0x9401000,0x10000)                     = 0 (0x0)              
close(105)                                       = 0 (0x0)              
gettimeofday(0x8788798,0x0)                      = 0 (0x0)              
linux_socketcall(0x1,0xbfbfe5e0)                 = 105 (0x69)           
linux_socketcall(0xe,0xbfbfe5d0)                 = 0 (0x0)              
linux_socketcall(0x2,0xbfbfe5e0)                 = 0 (0x0)              
linux_fcntl64(0x69,0x3,0x0)                      = 2 (0x2)              
linux_fcntl64(0x69,0x4,0x802)                    = 0 (0x0)              
linux_getrlimit(0x3,0xbfbfe500)                  = 0 (0x0)              
linux_pipe(0xbfbfe518)                           = 0 (0x0)              
linux_clone(0xf00,0x9468dd8)                     = 74036 (0x12134)      
SIGNAL 6
SIGNAL 6
Process stopped because of:  16
process exit, rval = 6

Is the use of clone new in this build or is there something new about
the data being passed which is some how bad?

    Steve / K
----- Original Message ----- 
From: "Steven Hartland" <killing at barrysworld.com>
To: <bf1942 at icculus.org>
Sent: Wednesday, October 29, 2003 11:27 PM
Subject: [bf1942] 1.5 linux server is a no go :(


> When run via screen just quits with:
> Error opening terminal: 
> When run from a vt100 ssh session:
> Error opening terminal: vt100
> 
> truss doesnt give much clue to whats going on:
> linux_brk(0x8637000)                             = 140734464 (0x8637000)
> linux_time(0x0)                                  = 1067469932 (0x3fa04c6c)
> linux_access(0xbfbfe0c0,0x4)                     ERR#2 'No such file or directory'
> linux_access(0xbfbfe0c0,0x4)                     ERR#2 'No such file or directory'
> write(2,0xbfbfcb60,31)                           = 31 (0x1f)
> exit(0x1)
> process exit, rval = 256
> 
> N.B. OS = FreeBSD
> 
>     Steve / K
> 
> 
> 
>