-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On 03/11/2010 10:05 AM, Sumit Bose wrote:
On Thu, Mar 11, 2010 at 07:19:13AM -0500, Stephen Gallagher wrote:
On 03/10/2010 11:18 AM, Sumit Bose wrote:
>>> Hi,
>>>
>>> this is an attempt to fix bz#571896 aka trac#415. A new parameter
>>> krb5_kadmin is introduced together with a new failover service in the
>>> Kerberos provider. If the change password service is not available the
>>> backend is not switched to offline mode unless authentication isn't
>>> possible too.
>>>
>>> I haven't add to option to the IPA provider, because I think here the
>>> kpasswd always runs together with the KDC.
>>>
>>> bye,
>>> Sumit
>>>
To sum up a conversation we had on IRC yesterday:
Nack, please use krb5_kpasswd instead of krb5_kadmin.
Additionally, this patch brought to mind that we need to handle
non-standard ports for both krb5_kdcip and krb5_kpasswd.
> new version attached.
> bye,
> Sumit
Nack (nitpicks).
Using strtol() to assign a value to an int (rather than a long) can be
unsafe, as on some platforms int == short.
I'd rather you used a uint16_t for the sssd_ctx->kdc_port and
sssd_ctx->kpasswd_port. That's the only acceptable range in any case.
Same for the port in the sssd_krb5_locator_lookup(). And other places.
Otherwise, the code looks sound to me.
- --
Stephen Gallagher
RHCE 804006346421761
Delivering value year after year.
Red Hat ranks #1 in value among software vendors.
http://www.redhat.com/promo/vendor/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Fedora -
http://enigmail.mozdev.org/
iEYEARECAAYFAkuZE7oACgkQeiVVYja6o6OT3wCdEVRiDXbj+K3jXvAwp0fAaJhc
xAUAoJ2GwOMAWIrMOnnfyNJRTV4/dg+A
=2uDl
-----END PGP SIGNATURE-----