| Message ID | 20150327211422.2465.50276.stgit@ahduyck-vm-fedora22 |
|---|---|
| State | Accepted, archived |
| Delegated to: | David Miller |
| Headers | show |
From: Alexander Duyck <alexander.h.duyck@redhat.com> Date: Fri, 27 Mar 2015 14:14:22 -0700 > While fixing a recent issue I noticed that we are doing some unnecessary > work inside the loop for ip_fib_net_exit. As such I am pulling out the > initialization to NULL for the locally stored fib_local, fib_main, and > fib_default. > > In addition I am restoring the original code for flushing the table as > there is no need to split up the fib_table_flush and hlist_del work since > the code for packing the tnodes with multiple key vectors was dropped. > > Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Applied. I also agree with you that we shouldn't try to get too fancy here as Cong is suggesting, lest ->delete() have other side effects in the future. Thanks. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c index 767120111d90..718b0a16ea40 100644 --- a/net/ipv4/fib_frontend.c +++ b/net/ipv4/fib_frontend.c @@ -1176,35 +1176,20 @@ static void ip_fib_net_exit(struct net *net) rtnl_lock(); +#ifdef CONFIG_IP_MULTIPLE_TABLES + RCU_INIT_POINTER(net->ipv4.fib_local, NULL); + RCU_INIT_POINTER(net->ipv4.fib_main, NULL); + RCU_INIT_POINTER(net->ipv4.fib_default, NULL); +#endif + for (i = 0; i < FIB_TABLE_HASHSZ; i++) { struct hlist_head *head = &net->ipv4.fib_table_hash[i]; struct hlist_node *tmp; struct fib_table *tb; - /* this is done in two passes as flushing the table could - * cause it to be reallocated in order to accommodate new - * tnodes at the root as the table shrinks. - */ - hlist_for_each_entry_safe(tb, tmp, head, tb_hlist) - fib_table_flush(tb); - hlist_for_each_entry_safe(tb, tmp, head, tb_hlist) { -#ifdef CONFIG_IP_MULTIPLE_TABLES - switch (tb->tb_id) { - case RT_TABLE_LOCAL: - RCU_INIT_POINTER(net->ipv4.fib_local, NULL); - break; - case RT_TABLE_MAIN: - RCU_INIT_POINTER(net->ipv4.fib_main, NULL); - break; - case RT_TABLE_DEFAULT: - RCU_INIT_POINTER(net->ipv4.fib_default, NULL); - break; - default: - break; - } -#endif hlist_del(&tb->tb_hlist); + fib_table_flush(tb); fib_free_table(tb); } }
While fixing a recent issue I noticed that we are doing some unnecessary work inside the loop for ip_fib_net_exit. As such I am pulling out the initialization to NULL for the locally stored fib_local, fib_main, and fib_default. In addition I am restoring the original code for flushing the table as there is no need to split up the fib_table_flush and hlist_del work since the code for packing the tnodes with multiple key vectors was dropped. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> --- net/ipv4/fib_frontend.c | 29 +++++++---------------------- 1 file changed, 7 insertions(+), 22 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html