-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On 09/07/2010 11:52 AM, Stephen Gallagher wrote:
The new patch attached addresses this recommendation as well as Jakub's
comment about initializing enum_cached.
I also noticed that there was a parentage bug with setent_add_ref() in
two places, where the reference was being allocated as a child of the
getent_ctx (gctx or pctx) instead of as a child of the cli_ctx. In this
case, if the client disappeared whild setpwent was still processing, we
would have been accessing freed memory when calling the notifications.
This new patch obsoletes the patches in the thread "Add a setent result
object timeout" as those patches were squashed into this one and
replaced the reference count.
Thank you for the reviews.
Jakub reported that the patch did not apply cleanly atop master, so here
is a rebased patch.
- --
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/
iEYEARECAAYFAkyHqqcACgkQeiVVYja6o6NSXQCeOFa59VKVcyfbOtTrecoCRQgk
IsEAn3+CfuOrQ/pGlglc73sDIbBMpPS7
=O87t
-----END PGP SIGNATURE-----