From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH] netfilter: use per-cpu spinlock rather than RCU Date: Mon, 13 Apr 2009 19:40:24 +0200 Message-ID: <49E37908.2080903@cosmosbay.com> References: <20090411174801.GG6822@linux.vnet.ibm.com> <18913.53699.544083.320542@cargo.ozlabs.ibm.com> <20090412173108.GO6822@linux.vnet.ibm.com> <20090412.181330.23529546.davem@davemloft.net> <20090413040413.GQ6822@linux.vnet.ibm.com> <20090413095309.631cf395@nehalam> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: paulmck@linux.vnet.ibm.com, David Miller , paulus@samba.org, mingo@elte.hu, torvalds@linux-foundation.org, laijs@cn.fujitsu.com, jeff.chua.linux@gmail.com, jengelh@medozas.de, kaber@trash.net, r000n@r000n.net, linux-kernel@vger.kernel.org, netfilter-devel@vger.kernel.org, netdev@vger.kernel.org, benh@kernel.crashing.org To: Stephen Hemminger Return-path: In-Reply-To: <20090413095309.631cf395@nehalam> Sender: netfilter-devel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org Stephen Hemminger a =E9crit : > This is an alternative version of ip/ip6/arp tables locking using > per-cpu locks. This avoids the overhead of synchronize_net() during > update but still removes the expensive rwlock in earlier versions. >=20 > The idea for this came from an earlier version done by Eric Duzamet. > Locking is done per-cpu, the fast path locks on the current cpu > and updates counters. The slow case involves acquiring the locks on > all cpu's. >=20 > The mutex that was added for 2.6.30 in xt_table is unnecessary since > there already is a mutex for xt[af].mutex that is held. >=20 > Tested basic functionality (add/remove/list), but don't have test cas= es > for stress, ip6tables or arptables. >=20 > Signed-off-by: Stephen Hemminger Patch seems good to me, but apparently xt_replace_table() misses the "acquiring the locks on all cpus" you mentioned in ChangeLog= ? I am still off-computers until tomorrow so cannot provide a patch for t= his, sorry. Some form of local_bh_disable(); for_each_possible_cpu(cpu) spin_lock(&per_cpu(ip_tables_lock, cpu)); oldinfo =3D private; /* do the substitution */ table->private =3D newinfo; newinfo->initial_entries =3D oldinfo->initial_entries; for_each_possible_cpu(cpu) spin_unlock(&per_cpu(ip_tables_lock, cpu)); local_bh_enable(); But I wonder if this could hit a limit of max spinlocks held by this cp= u, say on a 4096 cpu machine ? -- To unsubscribe from this list: send the line "unsubscribe netfilter-dev= el" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html