Re: net: use-after-free in worker_thread
From: Eric Dumazet
Date: Sat Dec 03 2016 - 08:50:00 EST
On Sat, 2016-12-03 at 14:05 +0100, Andrey Konovalov wrote:
> On Sat, Dec 3, 2016 at 1:58 PM, Andrey Konovalov <andreyknvl@xxxxxxxxxx> wrote:
> > +syzkaller@xxxxxxxxxxxxxxxx
> >
> > On Sat, Dec 3, 2016 at 1:56 PM, Andrey Konovalov <andreyknvl@xxxxxxxxxx> wrote:
> >> Hi!
> >>
> >> I'm seeing lots of the following error reports while running the
> >> syzkaller fuzzer.
> >>
> >> Reports appeared when I updated to 3c49de52 (Dec 2) from 2caceb32 (Dec 1).
> >>
> >> ==================================================================
> >> BUG: KASAN: use-after-free in worker_thread+0x17d8/0x18a0
> >> Read of size 8 at addr ffff880067f3ecd8 by task kworker/3:1/774
> >>
> >> page:ffffea00019fce00 count:1 mapcount:0 mapping: (null)
> >> index:0xffff880067f39c10 compound_mapcount: 0
> >> flags: 0x500000000004080(slab|head)
> >> page dumped because: kasan: bad access detected
> >>
> >> CPU: 3 PID: 774 Comm: kworker/3:1 Not tainted 4.9.0-rc7+ #66
> >> Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
> >> ffff88006c267838 ffffffff81f882da ffffffff6c25e338 1ffff1000d84ce9a
> >> ffffed000d84ce92 ffff88006c25e340 0000000041b58ab3 ffffffff8541e198
> >> ffffffff81f88048 0000000100000000 0000000041b58ab3 ffffffff853d3ee8
> >> Call Trace:
> >> [< inline >] __dump_stack lib/dump_stack.c:15
> >> [<ffffffff81f882da>] dump_stack+0x292/0x398 lib/dump_stack.c:51
> >> [< inline >] describe_address mm/kasan/report.c:262
> >> [<ffffffff817e50d1>] kasan_report_error+0x121/0x560 mm/kasan/report.c:368
> >> [< inline >] kasan_report mm/kasan/report.c:390
> >> [<ffffffff817e560e>] __asan_report_load8_noabort+0x3e/0x40
> >> mm/kasan/report.c:411
> >> [<ffffffff81329b88>] worker_thread+0x17d8/0x18a0 kernel/workqueue.c:2228
> >> [<ffffffff8133ebf3>] kthread+0x323/0x3e0 kernel/kthread.c:209
> >> [<ffffffff84a2a22a>] ret_from_fork+0x2a/0x40 arch/x86/entry/entry_64.S:433
> >>
> >> The buggy address belongs to the object at ffff880067f3e6d0
> >> which belongs to the cache kmalloc-2048 of size 2048
> >> The buggy address ffff880067f3ecd8 is located 1544 bytes inside
> >> of 2048-byte region [ffff880067f3e6d0, ffff880067f3eed0)
> >>
> >> Freed by task 0:
> >> [<ffffffff81203526>] save_stack_trace+0x16/0x20 arch/x86/kernel/stacktrace.c:57
> >> [<ffffffff817e4173>] save_stack+0x43/0xd0 mm/kasan/kasan.c:495
> >> [< inline >] set_track mm/kasan/kasan.c:507
> >> [<ffffffff817e4a53>] kasan_slab_free+0x73/0xc0 mm/kasan/kasan.c:571
> >> [< inline >] slab_free_hook mm/slub.c:1352
> >> [< inline >] slab_free_freelist_hook mm/slub.c:1374
> >> [< inline >] slab_free mm/slub.c:2951
> >> [<ffffffff817e0eb7>] kfree+0xe7/0x2b0 mm/slub.c:3871
> >> [< inline >] sk_prot_free net/core/sock.c:1372
> >> [<ffffffff831ea1c7>] __sk_destruct+0x5c7/0x6e0 net/core/sock.c:1445
> >> [<ffffffff831f3517>] sk_destruct+0x47/0x80 net/core/sock.c:1453
> >> [<ffffffff831f35a7>] __sk_free+0x57/0x230 net/core/sock.c:1461
> >> [<ffffffff831f37a3>] sk_free+0x23/0x30 net/core/sock.c:1472
> >> [< inline >] sock_put include/net/sock.h:1591
> >> [<ffffffff8348ca9c>] deferred_put_nlk_sk+0x2c/0x40 net/netlink/af_netlink.c:671
> >> [< inline >] __rcu_reclaim kernel/rcu/rcu.h:118
> >> [<ffffffff8146d42f>] rcu_do_batch.isra.67+0x8ff/0xc50 kernel/rcu/tree.c:2776
> >> [< inline >] invoke_rcu_callbacks kernel/rcu/tree.c:3040
> >> [< inline >] __rcu_process_callbacks kernel/rcu/tree.c:3007
> >> [<ffffffff8146e097>] rcu_process_callbacks+0x2b7/0xba0 kernel/rcu/tree.c:3024
> >> [<ffffffff84a2d08b>] __do_softirq+0x2fb/0xb63 kernel/softirq.c:284
> >>
> >> Allocated by task 10748:
> >> [<ffffffff81203526>] save_stack_trace+0x16/0x20 arch/x86/kernel/stacktrace.c:57
> >> [<ffffffff817e4173>] save_stack+0x43/0xd0 mm/kasan/kasan.c:495
> >> [< inline >] set_track mm/kasan/kasan.c:507
> >> [<ffffffff817e43fd>] kasan_kmalloc+0xad/0xe0 mm/kasan/kasan.c:598
> >> [<ffffffff817e0050>] __kmalloc+0xa0/0x2d0 mm/slub.c:3734
> >> [< inline >] kmalloc include/linux/slab.h:495
> >> [<ffffffff831e4c01>] sk_prot_alloc+0x101/0x2a0 net/core/sock.c:1333
> >> [<ffffffff831efd15>] sk_alloc+0x105/0x1000 net/core/sock.c:1389
> >> [<ffffffff8348ad46>] __netlink_create+0x66/0x1d0 net/netlink/af_netlink.c:588
> >> [<ffffffff8348cdab>] netlink_create+0x2fb/0x500 net/netlink/af_netlink.c:647
> >> [<ffffffff831dd1d6>] __sock_create+0x4f6/0x880 net/socket.c:1168
> >> [< inline >] sock_create net/socket.c:1208
> >> [< inline >] SYSC_socket net/socket.c:1238
> >> [<ffffffff831dd799>] SyS_socket+0xf9/0x230 net/socket.c:1218
> >> [<ffffffff84a29fc1>] entry_SYSCALL_64_fastpath+0x1f/0xc2
> >>
> >> Memory state around the buggy address:
> >> ffff880067f3eb80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> >> ffff880067f3ec00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> >>>ffff880067f3ec80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> >> ^
> >> ffff880067f3ed00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> >> ffff880067f3ed80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> >> ==================================================================
>
> Here is another report that looks related:
>
> ==================================================================
> BUG: KASAN: use-after-free in __list_add+0x236/0x2c0
> Read of size 8 at addr ffff880068854780 by task ksoftirqd/2/20
>
> page:ffffea0001a21400 count:1 mapcount:0 mapping: (null)
> index:0x0 compound_mapcount: 0
> flags: 0x500000000004080(slab|head)
> page dumped because: kasan: bad access detected
>
> CPU: 2 PID: 20 Comm: ksoftirqd/2 Not tainted 4.9.0-rc7+ #66
> Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
> ffff88006daf6578 ffffffff81f882da ffffffff6daf62a0 1ffff1000db5ec42
> ffffed000db5ec3a dffffc0000000000 0000000041b58ab3 ffffffff8541e198
> ffffffff81f88048 ffff88006dac3610 ffff88006daf6300 0000000000000802
> Call Trace:
> [< inline >] __dump_stack lib/dump_stack.c:15
> [<ffffffff81f882da>] dump_stack+0x292/0x398 lib/dump_stack.c:51
> [< inline >] describe_address mm/kasan/report.c:262
> [<ffffffff817e50d1>] kasan_report_error+0x121/0x560 mm/kasan/report.c:368
> [< inline >] kasan_report mm/kasan/report.c:390
> [<ffffffff817e560e>] __asan_report_load8_noabort+0x3e/0x40
> mm/kasan/report.c:411
> [<ffffffff8200c166>] __list_add+0x236/0x2c0 lib/list_debug.c:30
> [< inline >] list_add_tail include/linux/list.h:77
> [<ffffffff8131e295>] insert_work+0x175/0x4b0 kernel/workqueue.c:1298
> [<ffffffff8131eb52>] __queue_work+0x582/0x11e0 kernel/workqueue.c:1459
> [<ffffffff81320c21>] queue_work_on+0x231/0x240 kernel/workqueue.c:1484
> [< inline >] queue_work include/linux/workqueue.h:474
> [< inline >] schedule_work include/linux/workqueue.h:532
> [<ffffffff8348c8cc>] netlink_sock_destruct+0x23c/0x2d0
> net/netlink/af_netlink.c:361
> [<ffffffff831e9ce1>] __sk_destruct+0xe1/0x6e0 net/core/sock.c:1423
> [<ffffffff831f3517>] sk_destruct+0x47/0x80 net/core/sock.c:1453
> [<ffffffff831f35a7>] __sk_free+0x57/0x230 net/core/sock.c:1461
> [<ffffffff831f37a3>] sk_free+0x23/0x30 net/core/sock.c:1472
> [< inline >] sock_put include/net/sock.h:1591
> [<ffffffff8348ca9c>] deferred_put_nlk_sk+0x2c/0x40 net/netlink/af_netlink.c:671
> [< inline >] __rcu_reclaim kernel/rcu/rcu.h:118
> [<ffffffff8146d42f>] rcu_do_batch.isra.67+0x8ff/0xc50 kernel/rcu/tree.c:2776
> [< inline >] invoke_rcu_callbacks kernel/rcu/tree.c:3040
> [< inline >] __rcu_process_callbacks kernel/rcu/tree.c:3007
> [<ffffffff8146e097>] rcu_process_callbacks+0x2b7/0xba0 kernel/rcu/tree.c:3024
> [<ffffffff84a2d08b>] __do_softirq+0x2fb/0xb63 kernel/softirq.c:284
> [<ffffffff812d38c0>] run_ksoftirqd+0x20/0x60 kernel/softirq.c:676
> [<ffffffff81350132>] smpboot_thread_fn+0x562/0x860 kernel/smpboot.c:163
> [<ffffffff8133ebf3>] kthread+0x323/0x3e0 kernel/kthread.c:209
> [<ffffffff84a2a22a>] ret_from_fork+0x2a/0x40 arch/x86/entry/entry_64.S:433
>
> The buggy address belongs to the object at ffff880068854170
> which belongs to the cache kmalloc-2048 of size 2048
> The buggy address ffff880068854780 is located 1552 bytes inside
> of 2048-byte region [ffff880068854170, ffff880068854970)
>
> Freed by task 20:
> [<ffffffff81203526>] save_stack_trace+0x16/0x20 arch/x86/kernel/stacktrace.c:57
> [<ffffffff817e4173>] save_stack+0x43/0xd0 mm/kasan/kasan.c:495
> [< inline >] set_track mm/kasan/kasan.c:507
> [<ffffffff817e4a53>] kasan_slab_free+0x73/0xc0 mm/kasan/kasan.c:571
> [< inline >] slab_free_hook mm/slub.c:1352
> [< inline >] slab_free_freelist_hook mm/slub.c:1374
> [< inline >] slab_free mm/slub.c:2951
> [<ffffffff817e0eb7>] kfree+0xe7/0x2b0 mm/slub.c:3871
> [< inline >] sk_prot_free net/core/sock.c:1372
> [<ffffffff831ea1c7>] __sk_destruct+0x5c7/0x6e0 net/core/sock.c:1445
> [<ffffffff831f3517>] sk_destruct+0x47/0x80 net/core/sock.c:1453
> [<ffffffff831f35a7>] __sk_free+0x57/0x230 net/core/sock.c:1461
> [<ffffffff831f37a3>] sk_free+0x23/0x30 net/core/sock.c:1472
> [< inline >] sock_put include/net/sock.h:1591
> [<ffffffff8348ca9c>] deferred_put_nlk_sk+0x2c/0x40 net/netlink/af_netlink.c:671
> [< inline >] __rcu_reclaim kernel/rcu/rcu.h:118
> [<ffffffff8146d42f>] rcu_do_batch.isra.67+0x8ff/0xc50 kernel/rcu/tree.c:2776
> [< inline >] invoke_rcu_callbacks kernel/rcu/tree.c:3040
> [< inline >] __rcu_process_callbacks kernel/rcu/tree.c:3007
> [<ffffffff8146e097>] rcu_process_callbacks+0x2b7/0xba0 kernel/rcu/tree.c:3024
> [<ffffffff84a2d08b>] __do_softirq+0x2fb/0xb63 kernel/softirq.c:284
>
> Allocated by task 9480:
> [<ffffffff81203526>] save_stack_trace+0x16/0x20 arch/x86/kernel/stacktrace.c:57
> [<ffffffff817e4173>] save_stack+0x43/0xd0 mm/kasan/kasan.c:495
> [< inline >] set_track mm/kasan/kasan.c:507
> [<ffffffff817e43fd>] kasan_kmalloc+0xad/0xe0 mm/kasan/kasan.c:598
> [<ffffffff817e0050>] __kmalloc+0xa0/0x2d0 mm/slub.c:3734
> [< inline >] kmalloc include/linux/slab.h:495
> [<ffffffff831e4c01>] sk_prot_alloc+0x101/0x2a0 net/core/sock.c:1333
> [<ffffffff831efd15>] sk_alloc+0x105/0x1000 net/core/sock.c:1389
> [<ffffffff8348ad46>] __netlink_create+0x66/0x1d0 net/netlink/af_netlink.c:588
> [<ffffffff8348cdab>] netlink_create+0x2fb/0x500 net/netlink/af_netlink.c:647
> [<ffffffff831dd1d6>] __sock_create+0x4f6/0x880 net/socket.c:1168
> [< inline >] sock_create net/socket.c:1208
> [< inline >] SYSC_socket net/socket.c:1238
> [<ffffffff831dd799>] SyS_socket+0xf9/0x230 net/socket.c:1218
> [<ffffffff84a29fc1>] entry_SYSCALL_64_fastpath+0x1f/0xc2
>
> Memory state around the buggy address:
> ffff880068854680: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> ffff880068854700: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> >ffff880068854780: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> ^
> ffff880068854800: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> ffff880068854880: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> ==================================================================
Hi Andrey. Please give us some rest during the week end ;)
This looks like the bug I mentioned earlier for which I have a pending
patch ? Can you try it ?
The RCU conversion done by Thomas was quite buggy.
Thanks.
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 602e5ebe9db39ec6c72708628bc48efad9f0e680..c348c4a5ea4ecc05dcc9e2afbc069ab65a1a57fe 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -475,8 +475,8 @@ static struct sock *netlink_lookup(struct net *net, int protocol, u32 portid)
rcu_read_lock();
sk = __netlink_lookup(table, portid, net);
- if (sk)
- sock_hold(sk);
+ if (sk && !atomic_inc_not_zero(&sk->sk_refcnt))
+ sk = NULL;
rcu_read_unlock();
return sk;
@@ -600,6 +600,7 @@ static int __netlink_create(struct net *net, struct socket *sock,
}
init_waitqueue_head(&nlk->wait);
+ sock_set_flag(sk, SOCK_RCU_FREE);
sk->sk_destruct = netlink_sock_destruct;
sk->sk_protocol = protocol;
return 0;
@@ -664,13 +665,6 @@ static int netlink_create(struct net *net, struct socket *sock, int protocol,
goto out;
}
-static void deferred_put_nlk_sk(struct rcu_head *head)
-{
- struct netlink_sock *nlk = container_of(head, struct netlink_sock, rcu);
-
- sock_put(&nlk->sk);
-}
-
static int netlink_release(struct socket *sock)
{
struct sock *sk = sock->sk;
@@ -743,7 +737,7 @@ static int netlink_release(struct socket *sock)
local_bh_disable();
sock_prot_inuse_add(sock_net(sk), &netlink_proto, -1);
local_bh_enable();
- call_rcu(&nlk->rcu, deferred_put_nlk_sk);
+ sock_put(sk);
return 0;
}
diff --git a/net/netlink/af_netlink.h b/net/netlink/af_netlink.h
index 4fdb3831897775547f77c069a8018c0d2a253c8c..988d1a02487e37b7efd4872dd0ab6d230e5a2021 100644
--- a/net/netlink/af_netlink.h
+++ b/net/netlink/af_netlink.h
@@ -33,7 +33,6 @@ struct netlink_sock {
struct module *module;
struct rhash_head node;
- struct rcu_head rcu;
struct work_struct work;
};