I'm a bit curious what does

ldd /lib64/libsepol.so.1

report?

Also, you may want to try:

rpm -Va

That will take loads of time, but may report something interesting.

Pat

On 06/25/2014 07:44 AM, Werf, C.G. van der (Carel) wrote:
> Running an strace still offers no clues to this problem...
>
>
> When running:
> # strace yum --help 2>&1 | cat > yum-tracelog
>
>
> I see this at the end of the log.
>
> open("/lib64/libsepol.so.1", O_RDONLY)  = 6
> read(6, "\177ELF\2\1\1\0\0\0\0\0\0\0\0\0\3\0>\0\1\0\0\0\0=\300C7\0\0\0"..., 832) = 832
> fstat(6, {st_mode=S_IFREG|0755, st_size=247496, ...}) = 0
> mmap(0x3743c00000, 2383136, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 6, 0) = 0x3743c00000
> mprotect(0x3743c3b000, 2097152, PROT_NONE) = 0
> mmap(0x3743e3b000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 6, 0x3b000) = 0x3743e3b000
> mmap(0x3743e3c000, 40224, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x3743e3c000
> close(6)                                = 0
> --- SIGSEGV (Segmentation fault) @ 0 (0) ---
> +++ killed by SIGSEGV +++
> <<
>
> So after closing /lib64/libsepol.so.1, the segfault is triggered.
>
> But, since I disabled selinux, why is there a call to libsepol.so.1 ?
>
>
> On a my mirror-server, the same call to # yum --help" does work without segfault.
> An strace in this situation, will show that after closing the /lib64/libespol.so.1 the next commands are traced ...
>
> mprotect(0x319a607000, 4096, PROT_READ) = 0
> access("/etc/selinux/", F_OK)           = 0
> open("/etc/selinux/config", O_RDONLY)   = 6
> fstat(6, {st_mode=S_IFREG|0644, st_size=511, ...}) = 0
> mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x2b9d45eb4000
> read(6, "# This file controls the state o"..., 4096) = 511
> read(6, "", 4096)                       = 0
> close(6)
> <
>
> /etc/selinux/config is identical on both systems.
>
>
> Anyone a clue ?
>
> Regards,
> Carel
>
>
> -----Original Message-----
> From: Zhi-Wei Lu [mailto:[log in to unmask]]
> Sent: dinsdag 24 juni 2014 14:18
> To: Werf, C.G. van der (Carel); [log in to unmask]
> Subject: RE: yum segfaulting
>
> I had similar problem before, someone changed the stock system libz.so with a newer version libz.so, which yum didn't like it!
>
> Zhi-Wei Lu
> IET-CR-Network Operations Center
> University of California, Davis
> (530) 752-0155
>
> -----Original Message-----
> From: [log in to unmask] [mailto:[log in to unmask]] On Behalf Of Werf, C.G. van der (Carel)
> Sent: Tuesday, June 24, 2014 12:23 AM
> To: [log in to unmask]
> Subject: yum segfaulting
>
> Hi,
>
> I have two identical SL 5.3 fileservers, who function as a DRBD-pair.
> One of them was recently completely replaced with identical hardware, so I had to image the old one, and install OS-image on "new" server.
>
> But now, when I run yum on the new server, it returns a segmentation fault (even a simple: # yum --help).
>
> Googling this, a lot of pages hint for a memory error. But, running a memtest did not show any error.
>
> So far, only yum returns the segmentation fault.
>
> Does anyone have any clue for this ?
>
> Regards,
> Carel van der Werf


-- 
Pat Riehecky

Scientific Linux developer
http://www.scientificlinux.org/