[Aide] Segmentation fault on s390x architecture [0.11]

Christoph Ehret christoph.ehret at swisscom.com
Tue Feb 28 09:13:48 EET 2006


Hi Pablo,

Here are the different informations you asked for : 

Core was generated by `./aide --init -c ./aide.conf'.
Program terminated with signal 11, Segmentation fault.
#0  0x0000000080012944 in mutils_word32nswap (x=0x8012ed08, n=13,
    destructive=false) at stdfns.c:182
182     stdfns.c: No such file or directory.
        in stdfns.c


And here the backtrace :

(gdb) bt
#0  0x0000000080012944 in mutils_word32nswap (x=0x8012ed08, n=13,
    destructive=false) at stdfns.c:182
#1  0x00000000800143a6 in MD5Final (ctx=0x8012ecf0, digest=0x3fffffff144
"")
    at md5.c:125
#2  0x0000000080012066 in mhash_deinit (td=0x8012ecb0,
result=0x3fffffff144)
    at mhash.c:365
#3  0x0000000080003a48 in close_md (md=0x3fffffff140) at md.c:225
#4  0x000000008000d84e in calc_md (old_fs=Variable "old_fs" is not
available.
) at do_md.c:219
#5  0x000000008000c4ae in get_file_attrs (filename=0x8012c120
"/.autofsck",
    attr=Variable "attr" is not available.
) at gen_list.c:1289
#6  0x0000000080009140 in db_readline_disk (db=Variable "db" is not
available.
) at db_disk.c:251
#7  0x000000008000c056 in populate_tree (tree=0x8012b0e0) at
gen_list.c:1471
#8  0x00000000800116ec in main (argc=Variable "argc" is not available.
) at aide.c:535



Do you have any idea ?

Thanks for your help.

Chris



Le lundi 27 février 2006 à 18:04 +0200, Pablo Virolainen a écrit :
> On Mon, 27 Feb 2006, Christoph Ehret wrote:
> 
> > Hi,
> >
> > When running aide on a s390x architecture, I get a segmentation fault. I
> > first saw this when running "make check". I then tried to run "aide
> > --init" with my own aide.conf file, but again I had a segmentation
> > fault. If we have a look at was happened when I ran aide --init with my
> > own conf file, we can see that some data was written to "aide.db.new"
> > before the segmentation fault happened :
> >
> > @@begin_db
> > # This file was generated by Aide, version 0.9.4
> > # Time of generation was 2006-02-27 16:19:27
> > @@db_spec name lname attr perm uid gid size atime mtime ctime inode
> > lcount md5 sha1 rmd160 tiger
> > /sbin 0 67111933 40755 0 0 4096 MTE0MTA0MTgyNA== MTEzNzUxMTE4Ng==
> > MTEzNzUxMTE4Ng== 86 2 0 0 0 0
> > /scratch 0 67111933 40777 0 0 3976 MTE0MTA1MDgxOA== MTE0MDYxODE4NQ==
> > MTE0MDYxODE4NQ== 2 111 0 0 0 0
> > /tmp 0 67111933 41777 0 0 4096 MTE0MTA1MzE2MA== MTE0MTA1MzE1OQ==
> > MTE0MTA1MzE1OQ== 2 3 0 0 0 0
> > /sys 0 67111933 40755 0 0 0 MTE0MTAwOTYxNA== MTE0MDU4NzUwMA==
> > MTE0MDU4NzUwMA== 1 8 0 0 0 0
> > /dev 0 67111933 40755 0 0 3000 MTE0MTAzMDg5OA== MTE0MDU4NzUxOQ==
> > MTE0MDU4NzUxOQ== 204 8 0 0 0 0
> > /srv 0 67111933 40755 0 0 4096 MTE0MTAzMDg5OA== MTA5MjMzMDEzMA==
> > MTEyMzE1NTg0MA== 87 2 0 0 0 0
> > /media 0 67111933 40755 0 0 4096 MTE0MTAzMDg5OA== MTA5MjMzMDEzMA==
> > MTEyMzE1NTg0MA== 84 2 0 0 0 0
> > /mnt 0 67111933 40755 0 0 4096 MTE0MTAzMDg5OA== MTA5MjMzMDEzMA==
> > MTEyMzE1NTg0MA== 85 2 0 0 0 0
> >
> >
> > And here the end of the strace output when I run again the aide --init :
> >
> >
> > write(3, "/media 0 67111933 40755 0 0 4096"..., 97) = 97
> > lstat("/mnt", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
> > gettimeofday({1141053567, 607180}, NULL) = 0
> > write(3, "/mnt 0 67111933 40755 0 0 4096 M"..., 95) = 95
> > lstat("/.autofsck", {st_mode=S_IFREG|0644, st_size=0, ...}) = 0
> > gettimeofday({1141053567, 607710}, NULL) = 0
> > open("/.autofsck", O_RDONLY)            = 6
> > fstat(6, {st_mode=S_IFREG|0644, st_size=0, ...}) = 0
> > --- SIGSEGV (Segmentation fault) @ 0 (0) ---
> > +++ killed by SIGSEGV +++
> >
> >
> > Is this due to the s390x architecture or is there any other reason to
> > this segmentation fault ?
> >
> > I really thank you for your help.
> 
> Could you send backtrace of the core file also?
> 
> (ulimit -c unlimited; <run aide> ; gdb aide core and command bt)
> 
> Pablo Virolainen
> 
-- 

Christoph Ehret

Swisscom AG
Linux Engineering
Zentweg 46
CH-3050 Bern

Internet: www.swisscom.com/it-services


More information about the Aide mailing list