Impossible condition at dhc6.c:4109.

[Cathy Almond]

On 22/05/2014 22:51, John Newlin wrote:
> Version: dhcp-4.1-ESV-R5
> OS: Linux
> Kernel version: 2.6.32.11
> libc:  ucLibc xxx (not sure the exact version)
> Log:  http://pastebin.com/cVBE8za1
> 
> 
> The dhcp server was set with very short preferred lifetimes, 5 minutes
> (not exactly sure).
> DHCP server changed pools, so old address would depref or expire and
> we'd pick up a new address.
> 
> What seems to happen in dhc6.c is:
>  old address id DEPREF'd
>  the client state goes back to INIT
>  do_check_times gets called and hits this case because there is an
> address that isn't marked as expired, it is marked as DEPREF.
> 
> Has anyone seen this before, and know if it's fixed in a more recent image?
> 

I can't see anything similar that's been reported, or that's been fixed
between DHCP 4.1-ESV-R5 and the latest 4.1-ESV-R9.

If you wouldn't mind, could you confirm that it still exists on current
versions of DHCP, and then report it to us officially (email to
dhcp-bugs at isc.org or use the bug reporting form on our website:
http://www.isc.org/community/report-bug/).

If you can reproduce it reliably, please give us the steps.

If it's not readily reproducible, then we'll need a core dump, binary
and libs and anything else that you have to try to track down the root
cause.

More details on what to collect/how to submit etc.. can be found here:

https://kb.isc.org/article/AA-00340/0/What-to-do-if-your-BIND-or-DHCP-server-has-crashed.html

Thanks!

Cathy