On Sun, 2012-03-11 at 22:03:26 +0100, Witold Baryluk wrote:
> Package: dpkg
> Version: 1.16.1.2
> Severity: important
> File: /sbin/start-stop-daemon
> I'm not sure if this is s-s-d bug or eglibc one, here is gdb session:
>
> a02:~# gdb --args start-stop-daemon --start --quiet --oknodo --pidfile
> /var/run/rrdcached.pid --exec /usr/bin/rrdcached -- -l
> unix:/var/run/rrdcached.sock -j /var/lib/rrdcached/journal/ -F -b
> /var/lib/rrdcached/db/ -B -p /var/run/rrdcached.pid
[...]
> (gdb) r
> Starting program: /sbin/start-stop-daemon --start --quiet --oknodo --pidfile
> /var/run/rrdcached.pid --exec /usr/bin/rrdcached -- -l
> unix:/var/run/rrdcached.sock -j /var/lib/rrdcached/journal/ -F -b
> /var/lib/rrdcached/db/ -B -p /var/run/rrdcached.pid
> process 16599 is executing new program: /usr/bin/rrdcached
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib/alpha-linux-gnu/libthread_db.so.1".
>
> Program received signal SIGSEGV, Segmentation fault.
> strlen () at ../ports/sysdeps/alpha/strlen.S:45
> 45 ../ports/sysdeps/alpha/strlen.S: Nie ma takiego pliku ani katalogu.
> (gdb) bt full
> #0 strlen () at ../ports/sysdeps/alpha/strlen.S:45
> No locals.
> #1 0x000002000083bda0 in *__GI___strdup (s=0x0) at strdup.c:42
> len = <optimized out>
> #2 0x0000000120004534 in ?? ()
> No symbol table info available.
> #3 0x0000000120002c64 in ?? ()
> No symbol table info available.
> #4 0x00000200007ca84c in __libc_start_main (main=<optimized out>,
> argc=<optimized out>, ubp_av=<optimized out>, init=0x12000ad40,
> fini=<optimized out>,
> rtld_fini=<optimized out>, stack_end=0x11fdeb640) at libc-start.c:228
> self = 0x0
> result = <optimized out>
> unwind_buf = {cancel_jmp_buf = {{jmp_buf = {4831882560, 0,
> 4831970936, 0, 4831970936, 4831970936, -7573792870751129764,
> -7573795069782489268,
> -7573795073543097844, 0, 0, 0, 0, 0, 0, 0, 0}, mask_was_saved
> = 0}}, priv = {pad = {0x0, 0x0, 0x0, 0x0}, data = {prev = 0x0, cleanup = 0x0,
> canceltype = 0}}}
> #5 0x0000000120003e18 in ?? ()
> No symbol table info available.
> Backtrace stopped: previous frame inner to this frame (corrupt stack?)
Hmm the backtrace is not very helpful.
> I'm not sure, why this only happens rrdcached initscript.
>
> I do not have valgrind available (as it is not available for alpha yet,
> or anymore), so debuging is harder. I belive strlen problems was
> discussed some time ago on debian-alpha mailing list.
Could you build dpkg with DEB_BUILD_OPTIONS="noopt nostrip" and try to
get a backtrace again? I'm inclined to think this is a problem in
alpha's libc, but w/o more information it's hard to tell.
thanks,
guillem
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
