netdev - Re: [PATCH] netfilter: use per-cpu spinlock rather than RCU (v3)

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Tue, 14 Apr 2009 23:40:05 +0200
From:	Eric Dumazet <dada1@...mosbay.com>
To:	Stephen Hemminger <shemminger@...tta.com>
CC:	Jeff Chua <jeff.chua.linux@...il.com>,
	Patrick McHardy <kaber@...sh.net>, paulmck@...ux.vnet.ibm.com,
	David Miller <davem@...emloft.net>, paulus@...ba.org,
	mingo@...e.hu, torvalds@...ux-foundation.org, laijs@...fujitsu.com,
	jengelh@...ozas.de, r000n@...0n.net, linux-kernel@...r.kernel.org,
	netfilter-devel@...r.kernel.org, netdev@...r.kernel.org,
	benh@...nel.crashing.org
Subject: Re: [PATCH] netfilter: use per-cpu spinlock rather than RCU (v3)

Stephen Hemminger a écrit :
> 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.
> 
> The idea for this came from an earlier version done by Eric Dumazet.
> 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.
> 
> 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.
> 
> Signed-off-by: Stephen Hemminger <shemminger@...tta.com>
> 
> ---
>  include/linux/netfilter/x_tables.h |    5 -
>  net/ipv4/netfilter/arp_tables.c    |  112 +++++++++------------------------
>  net/ipv4/netfilter/ip_tables.c     |  123 +++++++++++--------------------------
>  net/ipv6/netfilter/ip6_tables.c    |  119 +++++++++++------------------------
>  net/netfilter/x_tables.c           |   28 --------
>  5 files changed, 110 insertions(+), 277 deletions(-)
> 

Tested successfuly on my dev machine, thanks Stephen.

Signed-off-by: Eric Dumazet <dada1@...mosbay.com>


"strace -tt -T iptables -L" on this 8 cpu machine show quite fast operations now
(sub 1/HZ) 

23:37:26.629489 getsockopt(3, SOL_IP, 0x40 /* IP_??? */, "filter\0\300\\\236@\365\271\231\"\300`\225T\367\1\0\0\0\0vm\300`\225T\367\
16"..., [84]) = 0 <0.000008>
23:37:26.629579 brk(0)                  = 0x8054000 <0.000006>
23:37:26.629613 brk(0x8075000)          = 0x8075000 <0.000007>
23:37:26.629660 getsockopt(3, SOL_IP, 0x41 /* IP_??? */, "filter\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0L"..., [10608])
= 0 <0.000031>
23:37:26.629772 fstat64(1, {st_mode=S_IFCHR|0620, st_rdev=makedev(136, 0), ...}) = 0 <0.000007>

same fast operation for an "iptables -A INPUT" operation


23:37:02.180313 close(4)                = 0 <0.000006>
23:37:02.180382 setsockopt(3, SOL_IP, 0x40 /* IP_??? */, "filter\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\16"..., 10664)
= 0 <0.000114>
23:37:02.180552 setsockopt(3, SOL_IP, 0x41 /* IP_??? */, "filter\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\20"..., 292) =
0 <0.000015>
23:37:02.180635 close(3)                = 0 <0.000010>

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html