| Message ID | 20240429183643.2029108-1-edumazet@google.com (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | fff6e6accdb71e5dd36e50478bd82d5409ac32d2 |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | [net-next] ipv6: anycast: use call_rcu_hurry() in aca_put() | expand |
On 4/29/24 12:36 PM, Eric Dumazet wrote: > This is a followup of commit b5327b9a300e ("ipv6: use > call_rcu_hurry() in fib6_info_release()"). > > I had another pmtu.sh failure, and found another lazy > call_rcu() causing this failure. > > aca_free_rcu() calls fib6_info_release() which releases > devices references. > > We must not delay it too much or risk unregister_netdevice/ref_tracker > traces because references to netdev are not released in time. > > This should speedup device/netns dismantles when CONFIG_RCU_LAZY=y > > Signed-off-by: Eric Dumazet <edumazet@google.com> > --- > net/ipv6/anycast.c | 5 ++--- > 1 file changed, 2 insertions(+), 3 deletions(-) > Reviewed-by: David Ahern <dsahern@kernel.org>
Hello: This patch was applied to netdev/net-next.git (main) by David S. Miller <davem@davemloft.net>: On Mon, 29 Apr 2024 18:36:43 +0000 you wrote: > This is a followup of commit b5327b9a300e ("ipv6: use > call_rcu_hurry() in fib6_info_release()"). > > I had another pmtu.sh failure, and found another lazy > call_rcu() causing this failure. > > aca_free_rcu() calls fib6_info_release() which releases > devices references. > > [...] Here is the summary with links: - [net-next] ipv6: anycast: use call_rcu_hurry() in aca_put() https://git.kernel.org/netdev/net-next/c/fff6e6accdb7 You are awesome, thank you!
diff --git a/net/ipv6/anycast.c b/net/ipv6/anycast.c index 0f2506e3535925468dcc5fa6cc30ae0952a67ea7..0627c4c18d1a5067a668da778ad444855194cbeb 100644 --- a/net/ipv6/anycast.c +++ b/net/ipv6/anycast.c @@ -252,9 +252,8 @@ static void aca_free_rcu(struct rcu_head *h) static void aca_put(struct ifacaddr6 *ac) { - if (refcount_dec_and_test(&ac->aca_refcnt)) { - call_rcu(&ac->rcu, aca_free_rcu); - } + if (refcount_dec_and_test(&ac->aca_refcnt)) + call_rcu_hurry(&ac->rcu, aca_free_rcu); } static struct ifacaddr6 *aca_alloc(struct fib6_info *f6i,
This is a followup of commit b5327b9a300e ("ipv6: use call_rcu_hurry() in fib6_info_release()"). I had another pmtu.sh failure, and found another lazy call_rcu() causing this failure. aca_free_rcu() calls fib6_info_release() which releases devices references. We must not delay it too much or risk unregister_netdevice/ref_tracker traces because references to netdev are not released in time. This should speedup device/netns dismantles when CONFIG_RCU_LAZY=y Signed-off-by: Eric Dumazet <edumazet@google.com> --- net/ipv6/anycast.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)