The hash table that is used to cache netgroups is allocated with low-level hash_create() and the freed in a destructor. The entries themselves are talloc pointers with destructors which results in a funky behaviour during shutdown, often a segfault.
I think the correct thing to do is just use sss_hash_create() so that the table is deallocated when the responder goes away.
On Fri, 2012-02-03 at 15:10 +0100, Jakub Hrozek wrote:
The hash table that is used to cache netgroups is allocated with low-level hash_create() and the freed in a destructor. The entries themselves are talloc pointers with destructors which results in a funky behaviour during shutdown, often a segfault.
I think the correct thing to do is just use sss_hash_create() so that the table is deallocated when the responder goes away.
ACK
Simo.
On Fri, 2012-02-03 at 14:40 -0500, Simo Sorce wrote:
On Fri, 2012-02-03 at 15:10 +0100, Jakub Hrozek wrote:
The hash table that is used to cache netgroups is allocated with low-level hash_create() and the freed in a destructor. The entries themselves are talloc pointers with destructors which results in a funky behaviour during shutdown, often a segfault.
I think the correct thing to do is just use sss_hash_create() so that the table is deallocated when the responder goes away.
ACK
Pushed to master
sssd-devel@lists.fedorahosted.org
-
Jakub Hrozek -
Simo Sorce -
Stephen Gallagher