From mboxrd@z Thu Jan 1 00:00:00 1970 From: Simon Horman Subject: Re: [slab poison overwritten] Re: [GIT] Networking Date: Tue, 22 Mar 2011 10:18:01 +0900 Message-ID: <20110322011801.GE27019@verge.net.au> References: <1300714346.2884.284.camel@edumazet-laptop> <20110321161528.GA28580@elte.hu> <20110321164238.GA5303@elte.hu> <20110321173941.GB3892@elte.hu> <1300730838.2884.578.camel@edumazet-laptop> <1300738540.2837.5.camel@edumazet-laptop> <20110321221357.GF22625@verge.net.au> <20110321232921.GG22625@verge.net.au> <20110322000130.GC27019@verge.net.au> <20110322001706.GD27019@verge.net.au> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Ingo Molnar , David Miller , torvalds@linux-foundation.org, akpm@linux-foundation.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Peter Zijlstra , Thomas Gleixner , Arnd Bergmann , Pekka Enberg , Julian Anastasov , Hans Schillstrom To: Eric Dumazet Return-path: Content-Disposition: inline In-Reply-To: <20110322001706.GD27019@verge.net.au> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Tue, Mar 22, 2011 at 09:17:07AM +0900, Simon Horman wrote: > On Tue, Mar 22, 2011 at 09:01:33AM +0900, Simon Horman wrote: > > On Tue, Mar 22, 2011 at 08:29:21AM +0900, Simon Horman wrote: > > > On Tue, Mar 22, 2011 at 07:13:58AM +0900, Simon Horman wrote: > > > > On Mon, Mar 21, 2011 at 09:15:40PM +0100, Eric Dumazet wrote: > > > > > Le lundi 21 mars 2011 =C3=A0 19:07 +0100, Eric Dumazet a =C3=A9= crit : > > > > > > Le lundi 21 mars 2011 =C3=A0 18:39 +0100, Ingo Molnar a =C3= =A9crit : > > > > > > > here's the same but with kallsyms enabled. > > > > > > >=20 > > > > > > > Thanks, > > > > > > >=20 > > > > > > > Ingo > > > > > > >=20 > > > > > > > [ 9.585627] initcall 0xffffffff81d5b806 returned 0 aft= er 0 usecs > > > > > > > [ 9.588960] calling 0xffffffff81d5b9da @ 1 > > > > > > > [ 9.592303] IPVS: Creating netns size=3D1272 id=3D0 > > > > > > > [ 9.595646] IPVS: __ip_vs_control_init(): alloc_percpu= =2E > > > > > > > [ 9.602298] IPVS: cannot register namespace. > > > > > > > [ 9.605627] IPVS: can't setup control > > > > > >=20 > > > > > > It seems IPVS is busted in case of memory allocation error = in=20 > > > > > > __ip_vs_control_init() > > > > > >=20 > > > > > > IPVS deinits its "struct netns_ipvs" space, but something (= in IPVS) uses > > > > > > it after free. > > > > > >=20 > > > > > > __ip_vs_init() seems to be called before ip_vs_init() compl= etes > > > > > > correctly. We then keep in net->ipvs a pointer to some free= d memory. > > > > > >=20 > > > > > > Commit 14e405461e664b7 did some changes in this area > > > > > >=20 > > > > > > Simon, any idea ? > > > > > >=20 > > > > > >=20 > > > > >=20 > > > > > For the time being, we can avoid the false memory allocation = error (and > > > > > leak) > > > >=20 > > > > Sorry, that typo is my work. > > >=20 > > > With your patch applied I now see the following > > >=20 > > > ffff880003bbf1a0 corresponds to &ipvs->app_key in __ip_vs_app_ini= t(). > > > I'll continue looking into this. > > >=20 > > > [ 12.610000] IPVS: Creating netns size=3D2456 id=3D0 > > > [ 12.630000] IPVS: Registered protocols (TCP, UDP, SCTP, AH, ES= P) > > > [ 12.640000] BUG: key ffff880003bbf1a0 not in .data! > > > [ 12.640000] ------------[ cut here ]------------ > > > [ 12.640000] WARNING: at kernel/lockdep.c:2701 > > > lockdep_init_map+0x37b/0x570() > > > [ 12.640000] Hardware name: Bochs > > > [ 12.640000] Pid: 1, comm: swapper Tainted: G W > > > 2.6.38-kexec-06330-g69b7efe-dirty #122 > > > [ 12.650000] Call Trace: > > > [ 12.650000] [] warn_slowpath_common+0x75/0x= b0 > > > [ 12.650000] [] warn_slowpath_null+0x15/0x20 > > > [ 12.650000] [] lockdep_init_map+0x37b/0x570 > > > [ 12.650000] [] ? trace_hardirqs_on+0xd/0x10 > > > [ 12.650000] [] debug_mutex_init+0x38/0x50 > > > [ 12.650000] [] __mutex_init+0x5c/0x70 > > > [ 12.650000] [] __ip_vs_app_init+0x64/0x86 > > > [ 12.660000] [] ? ip_vs_init+0x0/0xff > > > [ 12.660000] [] T.620+0x43/0x170 > > > [ 12.660000] [] ? register_pernet_subsys+0x1= a/0x40 > > > [ 12.660000] [] ? ip_vs_init+0x0/0xff > > > [ 12.660000] [] ? ip_vs_init+0x0/0xff > > > [ 12.660000] [] register_pernet_operations+0= x57/0xb0 > > > [ 12.660000] [] ? ip_vs_init+0x0/0xff > > > [ 12.670000] [] register_pernet_subsys+0x29/= 0x40 > > > [ 12.670000] [] ip_vs_app_init+0x10/0x12 > > > [ 12.670000] [] ip_vs_init+0x4c/0xff > > > [ 12.670000] [] do_one_initcall+0x7a/0x12e > > > [ 12.670000] [] kernel_init+0x13e/0x1c2 > > > [ 12.670000] [] kernel_thread_helper+0x4/0x1= 0 > > > [ 12.670000] [] ? restore_args+0x0/0x30 > > > [ 12.680000] [] ? kernel_init+0x0/0x1c2 > > > [ 12.680000] [] ? kernel_thread_helper+0x0/0= x10 > > > [ 12.680000] ---[ end trace 4eaa2a86a8e2da23 ]--- > >=20 > > It seems that the problem above was introduced by > > ab8a5e8408c3 ("IPVS: netns awareness to ip_vs_app"). > > I assume the hungs are the cause: >=20 > s/hungs/hunks below/ >=20 > I am a little unsure of what to do about this. >=20 > The problem seems to be that ipvs->app_key is not in static storage. > But I'm not sure how to resolve that given that the struct netns_ipvs= is > per-network namespace. So I guess that the locking needs to be re-wor= ked. > Again, I'm a little unsure of what the best way forward is. I had an idea for a fix over breakfast. IPVS: Use global mutex in ip_vs_app.c As part of the work to make IPVS network namespace aware __ip_vs_app_mutex was replaced by a per-namespace lock, ipvs->app_mutex. ipvs->app_key is also supplied for debugging purposes. Unfortunately this implementation results in ipvs->app_key residing in non-static storage which at the very least causes a lockdep warning. This patch takes the rather heavy-handed approach of reinstating __ip_vs_app_mutex which will cover access to the ipvs->list_head of all network namespaces. [ 12.610000] IPVS: Creating netns size=3D2456 id=3D0 [ 12.630000] IPVS: Registered protocols (TCP, UDP, SCTP, AH, ESP) [ 12.640000] BUG: key ffff880003bbf1a0 not in .data! [ 12.640000] ------------[ cut here ]------------ [ 12.640000] WARNING: at kernel/lockdep.c:2701 lockdep_init_map+0x37b= /0x570() [ 12.640000] Hardware name: Bochs [ 12.640000] Pid: 1, comm: swapper Tainted: G W 2.6.38-kexec-0= 6330-g69b7efe-dirty #122 [ 12.650000] Call Trace: [ 12.650000] [] warn_slowpath_common+0x75/0xb0 [ 12.650000] [] warn_slowpath_null+0x15/0x20 [ 12.650000] [] lockdep_init_map+0x37b/0x570 [ 12.650000] [] ? trace_hardirqs_on+0xd/0x10 [ 12.650000] [] debug_mutex_init+0x38/0x50 [ 12.650000] [] __mutex_init+0x5c/0x70 [ 12.650000] [] __ip_vs_app_init+0x64/0x86 [ 12.660000] [] ? ip_vs_init+0x0/0xff [ 12.660000] [] T.620+0x43/0x170 [ 12.660000] [] ? register_pernet_subsys+0x1a/0x40 [ 12.660000] [] ? ip_vs_init+0x0/0xff [ 12.660000] [] ? ip_vs_init+0x0/0xff [ 12.660000] [] register_pernet_operations+0x57/0x= b0 [ 12.660000] [] ? ip_vs_init+0x0/0xff [ 12.670000] [] register_pernet_subsys+0x29/0x40 [ 12.670000] [] ip_vs_app_init+0x10/0x12 [ 12.670000] [] ip_vs_init+0x4c/0xff [ 12.670000] [] do_one_initcall+0x7a/0x12e [ 12.670000] [] kernel_init+0x13e/0x1c2 [ 12.670000] [] kernel_thread_helper+0x4/0x10 [ 12.670000] [] ? restore_args+0x0/0x30 [ 12.680000] [] ? kernel_init+0x0/0x1c2 [ 12.680000] [] ? kernel_thread_helper+0x0/0x1glob= al0 Signed-off-by: Simon Horman Cc: Ingo Molnar Cc: Eric Dumazet Cc: Julian Anastasov Cc: Hans Schillstrom --- include/net/ip_vs.h | 2 -- net/netfilter/ipvs/ip_vs_app.c | 23 ++++++++++------------- 2 files changed, 10 insertions(+), 15 deletions(-) diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index 272f593..30b49ed 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -801,8 +801,6 @@ struct netns_ipvs { struct list_head rs_table[IP_VS_RTAB_SIZE]; /* ip_vs_app */ struct list_head app_list; - struct mutex app_mutex; - struct lock_class_key app_key; /* mutex debuging */ =20 /* ip_vs_proto */ #define IP_VS_PROTO_TAB_SIZE 32 /* must be power of 2 */ diff --git a/net/netfilter/ipvs/ip_vs_app.c b/net/netfilter/ipvs/ip_vs_= app.c index 5c48ffb..2dc6de1 100644 --- a/net/netfilter/ipvs/ip_vs_app.c +++ b/net/netfilter/ipvs/ip_vs_app.c @@ -43,6 +43,8 @@ EXPORT_SYMBOL(register_ip_vs_app); EXPORT_SYMBOL(unregister_ip_vs_app); EXPORT_SYMBOL(register_ip_vs_app_inc); =20 +static DEFINE_MUTEX(__ip_vs_app_mutex); + /* * Get an ip_vs_app object */ @@ -167,14 +169,13 @@ int register_ip_vs_app_inc(struct net *net, struct ip_vs_app *app, __u16 p= roto, __u16 port) { - struct netns_ipvs *ipvs =3D net_ipvs(net); int result; =20 - mutex_lock(&ipvs->app_mutex); + mutex_lock(&__ip_vs_app_mutex); =20 result =3D ip_vs_app_inc_new(net, app, proto, port); =20 - mutex_unlock(&ipvs->app_mutex); + mutex_unlock(&__ip_vs_app_mutex); =20 return result; } @@ -189,11 +190,11 @@ int register_ip_vs_app(struct net *net, struct ip= _vs_app *app) /* increase the module use count */ ip_vs_use_count_inc(); =20 - mutex_lock(&ipvs->app_mutex); + mutex_lock(&__ip_vs_app_mutex); =20 list_add(&app->a_list, &ipvs->app_list); =20 - mutex_unlock(&ipvs->app_mutex); + mutex_unlock(&__ip_vs_app_mutex); =20 return 0; } @@ -205,10 +206,9 @@ int register_ip_vs_app(struct net *net, struct ip_= vs_app *app) */ void unregister_ip_vs_app(struct net *net, struct ip_vs_app *app) { - struct netns_ipvs *ipvs =3D net_ipvs(net); struct ip_vs_app *inc, *nxt; =20 - mutex_lock(&ipvs->app_mutex); + mutex_lock(&__ip_vs_app_mutex); =20 list_for_each_entry_safe(inc, nxt, &app->incs_list, a_list) { ip_vs_app_inc_release(net, inc); @@ -216,7 +216,7 @@ void unregister_ip_vs_app(struct net *net, struct i= p_vs_app *app) =20 list_del(&app->a_list); =20 - mutex_unlock(&ipvs->app_mutex); + mutex_unlock(&__ip_vs_app_mutex); =20 /* decrease the module use count */ ip_vs_use_count_dec(); @@ -501,7 +501,7 @@ static void *ip_vs_app_seq_start(struct seq_file *s= eq, loff_t *pos) struct net *net =3D seq_file_net(seq); struct netns_ipvs *ipvs =3D net_ipvs(net); =20 - mutex_lock(&ipvs->app_mutex); + mutex_lock(&__ip_vs_app_mutex); =20 return *pos ? ip_vs_app_idx(ipvs, *pos - 1) : SEQ_START_TOKEN; } @@ -535,9 +535,7 @@ static void *ip_vs_app_seq_next(struct seq_file *se= q, void *v, loff_t *pos) =20 static void ip_vs_app_seq_stop(struct seq_file *seq, void *v) { - struct netns_ipvs *ipvs =3D net_ipvs(seq_file_net(seq)); - - mutex_unlock(&ipvs->app_mutex); + mutex_unlock(&__ip_vs_app_mutex); } =20 static int ip_vs_app_seq_show(struct seq_file *seq, void *v) @@ -583,7 +581,6 @@ static int __net_init __ip_vs_app_init(struct net *= net) struct netns_ipvs *ipvs =3D net_ipvs(net); =20 INIT_LIST_HEAD(&ipvs->app_list); - __mutex_init(&ipvs->app_mutex, "ipvs->app_mutex", &ipvs->app_key); proc_net_fops_create(net, "ip_vs_app", 0, &ip_vs_app_fops); return 0; } --=20 1.7.2.3