I get the following error compiling on RHEL/CentOS 5.9 on my 64 bit dev
host...
> CC dynamic_array.lo
> dynamic_array.c: In function ‘vanessa_dynamic_array_split_str_to_int’:
> dynamic_array.c:556: warning: cast to pointer from integer of
> different size
> dynamic_array.c:565: warning: cast to pointer from integer of
> different size
This error does not occur my 32 bit dev host. Can this be safely ignored
or is there a 32/64 mismatch problem that needs to be fixed?
thanks,
David
--
David Severance
Enterprise Unix Services
Office of Information Technology
(949) 824-7552
sev(a)uci.edu
When we originally implemented perdition v1.17 on RHEL4 several years
ago we had a customized the /etc/pam.d/perdition file to use the
pam_access.so module so we could lock out accounts for
administrative/maintenance purposes. This worked well for many years.
Within the last year we upgraded to RHEL5 and perdition 1.19rc4 and then
rc5. It would appear that since this update perdition no longer seems to
pay any sort of attention to the contents of the pam file and will
always assume success. Nothing seems to be logged either unlike before.
Although we are now compiling perdition to be located in /usr/perdition
with it's conf files in /etc/perdition I don't see how that would affect
pam operations, at least I would think it shouldn't. ldd on perdition
does show pam libraries being linked. Any ideas?
thanks,
David
--
David Severance
Enterprise Unix Services
Office of Information Technology
(949) 824-7552
sev(a)uci.edu
Hello perdition users!
I really appreciate Perdition, works great also in large environment!
I see that Perdition supports STARTTLS/STLS to manage secure connections. In my
network I would like separate "good" ips and "bad" ips, where "good" can
establish a clear connection (tls_listen), and "bad" must starttls
(tls_listen_force), all on the same perdition server.
Is this possible?
It would be very useful for me if ssl_mode could be user based (for instance set
on LDAP profile of the account).
Otherwise, it could be very useful a behaviour like Postfix: local networks can
connect without encryption, and other must use STARTTLS.
Thanks a lots
Best Regards
Marco
