Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 14 Dec 2004 22:28:35 +0300
From:      "Andrew P." <infofarmer@mail.ru>
To:        Toomas Aas <toomas.aas@raad.tartu.ee>
Cc:        questions@freebsd.org
Subject:   Re: ld-elf.so.1: Shared object"libintl.so.6" not found
Message-ID:  <41BF3EE3.9080706@mail.ru>
In-Reply-To: <41BF30FB.1020309@raad.tartu.ee>
References:  <41BF2C32.3080808@mail.ru> <41BF30FB.1020309@raad.tartu.ee>

next in thread | previous in thread | raw e-mail | index | archive | help
Toomas Aas wrote:
> Andrew P. wrote:
> 
>>>> I got this at startup:
>>>>
>>>> Dec 11 03:32:42 satbsd /kernel:
>>>>   Starting ppp as "root"
>>>> Dec 11 03:32:42 satbsd /kernel:
>>>>   /usr/libexec/ld-elf.so.1:
>>>> Dec 11 03:32:42 satbsd /kernel:
>>>>   Shared object "libintl.so.6" not found
> 
> 
>>> The real question you should be asking is "why does ppp (a system
>>> binary) depend on libintl (not a system library)?"  You've probably
>>> replaced your ppp(8) with something else, with poor consequences. 
>>
>>
>> Can you tell me how to figure that out? 
> 
> 
> It's hard to give a definitive guide here.
> 
> First find out which startup script starts ppp at boot time. I can't 
> look it up for you because I have FreeBSD 5.3 here where the startup 
> scripts are different from 4.10. Has the script been modified. Is it 
> still trying to run /usr/sbin/ppp or is it trying to run something else?

/etc/rc.network starts ppp with the following line:
	"/usr/sbin/ppp -quiet -ddial default"
I am pretty sure both rc and rc.network are the original ones

> Look at the output of 'ls -l /usr/sbin'. Is the timestamp of ppp close 
> to other files in that directory (that would be the time when you 
> built/installed the world)? If it's not, then ppp has been replaced with 
> something else. The timestamp of /usr/sbin/ppp might give you an idea 
> when it happened. Try to remember what you were doing around this time.

On a system which was never rebuilt since 4.10 install it's like this:
-r-xr-xr-x  1 root  wheel      3712 May 26 01:28:37 2004 pmap_set
-r-xr-xr-x  1 root  wheel      9868 May 26 01:28:37 2004 pnpinfo
-r-xr-xr-x  1 root  wheel     11916 May 26 01:28:37 2004 portmap
-r-sr-xr--  1 root  network  336204 May 26 01:28:37 2004 ppp
-r-xr-xr-x  1 root  wheel     10088 May 26 01:28:37 2004 pppctl
-r-sr-x---  1 root  dialer    95376 May 26 01:28:37 2004 pppd
-r-xr-xr-x  1 root  wheel      9228 May 26 01:28:38 2004 pppstats

On a freshly rebuilt system it's this:
-r-xr-xr-x  1 root  wheel      3744 Dec 10 13:32:36 2004 pmap_set
-r-xr-xr-x  1 root  wheel      9868 Dec 10 13:32:36 2004 pnpinfo
-r-xr-xr-x  1 root  wheel     12044 Dec 10 13:32:36 2004 portmap
-r-sr-xr--  1 root  network  340876 Dec 10 13:32:36 2004 ppp
-r-xr-xr-x  1 root  wheel     10152 Dec 10 13:32:37 2004 pppctl
-r-sr-x---  1 root  dialer    96656 Dec 10 13:32:37 2004 pppd
-r-xr-xr-x  1 root  wheel      9260 Dec 10 13:32:37 2004 pppstats

> I don't want to cry wolf, but it might even be some kind of trojan. If 
> you're sure it's not a trojan (some accident during portupgrade or 
> such), the easiest way to recover would probably be to rebuild and 
> reinstall the world. If it's a trojan, the only safe way is to wipe the 
> entire system, reinstall and restore the data.

I'm pretty sure it could not be a trojan. I've tried to rebuild and 
reinstall the world and the kernel, following the current version of 
handbook. I know that if I reinstall FreeBSD, it'll probably solve the 
problem. But it's the M$ way, and the thing I love UNIX for is that I 
have a chance to know precisely why a system sneezes when it sneezes.

So I would be glad and grateful if you could give me an advice like 
"read this book on gdb, and the developer's handbook, parse these 10Mb 
of logs manually, rebuild 21 times with these and these make.conf's, 
backtrace everything you see" - and I'd do everything is possible, if 
only it could lead to a success.

Thanks,
Andrew P.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?41BF3EE3.9080706>