Re: Bug#321561: dlerror: /lib/libresolv.so.2: symbol __res_maybe_init, version GLIBC_PRIVATE
Daniel Jacobowitz <dan@debian.org> writes:
> On Wed, Aug 10, 2005 at 10:49:47PM -0700, Russ Allbery wrote:
>> As the co-maintainer of pam_krb5.so, I'd love some more information
>> here. I certainly do believe you, but nothing about that error message
>> helps me figure out where the bug is in pam_krb5.so. It looks
>> completely internal to the linkage between libresolv.so.2 and
>> libc.so.6.
> Yes - my fault, I believe I misread the error. This one would probably
> be fixed by restarting the service.
Ah, yes, cron appears to be at fault in this case.
>> Is there a standard cause of this bug or a standard thing that pam_krb5.so
>> should be doing that it's not? This just showed up with the new glibc
>> version. I'm also seeing:
>>
>> Aug 10 22:53:01 rivendell CRON[30864]: PAM [dlerror: /lib/libpthread.so.0: symbol __libc_stack_end, version GLIBC_2.1 not defined in file ld-linux.so.2 with link time reference]
>>
>> with another PAM module, similar to another reported bug, but restarting
>> the software doesn't help.
> How did you get a libpthread.so.0 that does not match your ld.so?
Sorry, false alarm, cron just happened to start at the same time that I
was doing testing and I failed to notice that the error messages were from
cron rather than what I was testing. After restarting cron, everything
now appears to be happy.
--
Russ Allbery (rra@stanford.edu) <http://www.eyrie.org/~eagle/>
Reply to: