Hi,
ma, 2023-12-11 kello 22:40 +0800, Chaoyuan Peng kirjoitti:
> In __sco_sock_close(), the sco_conn_lock doesn't properly guard the access
> to 'conn'. This leads to the UAF where __sco_sock_close() access the freed
> 'sco_pi(sk)->conn' which is destroyed by hci_abort_conn_sync() on another
> thread.
>
> Such is the case in the following call trace:
>
> BUG: KASAN: slab-use-after-free in __sco_sock_close+0x2d7/0x6b0 net/bluetooth/sco.c:444
> Write of size 4 at addr ffff88801c47a010 by task syz-executor402/10616
What branch is this crash seen in?
sco_disconn_cfm and sco_conn_del should have been called from
hci_abort_conn_sync. In particular sco_conn_del should set
sco_pi(sk)->conn = NULL.
Now, it's possible those are racing here, and maybe it's possible to
make the locking is correct if it's not.
It's also possible there's still some bug that hci_abort_conn_sync gets
called with reason == 0. Most of these should have been fixed by
181a42edddf51d5d9697ec
Adding WARN_ON(!reason) in hci_abort_conn_sync should tell if it's
still that bug.
>
> Call Trace:
> <TASK>
> __dump_stack lib/dump_stack.c:88 [inline]
> dump_stack_lvl+0xd9/0x150 lib/dump_stack.c:106
> print_address_description mm/kasan/report.c:364 [inline]
> print_report+0xc1/0x5e0 mm/kasan/report.c:475
> kasan_report+0xbe/0xf0 mm/kasan/report.c:588
> check_region_inline mm/kasan/generic.c:181 [inline]
> kasan_check_range+0xf0/0x190 mm/kasan/generic.c:187
> instrument_atomic_read_write ./include/linux/instrumented.h:96 [inline]
> atomic_dec_and_test ./include/linux/atomic/atomic-instrumented.h:1375 [inline]
> hci_conn_drop ./include/net/bluetooth/hci_core.h:1523 [inline]
> __sco_sock_close+0x2d7/0x6b0 net/bluetooth/sco.c:444
> sco_sock_close net/bluetooth/sco.c:469 [inline]
> sco_sock_release+0x7b/0x2d0 net/bluetooth/sco.c:1246
> __sock_release+0xae/0x260 net/socket.c:659
> sock_close+0x1c/0x20 net/socket.c:1419
> __fput+0x406/0xa70 fs/file_table.c:384
> __fput_sync+0x45/0x50 fs/file_table.c:465
> __do_sys_close fs/open.c:1572 [inline]
> __se_sys_close fs/open.c:1557 [inline]
> __x64_sys_close+0x8b/0x110 fs/open.c:1557
> do_syscall_x64 arch/x86/entry/common.c:50 [inline]
> do_syscall_64+0x39/0xb0 arch/x86/entry/common.c:80
> entry_SYSCALL_64_after_hwframe+0x63/0xcd
> RIP: 0033:0x7f338be5659b
> Code: 03 00 00 00 0f 05 48 3d 00 f0 ff ff 77 41 c3 48 83 ec 18 89 7c 24 0c e8 d3 eb 02 00 8b 7c 24 0c 41 89 c0 b8 03 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 35 44 89 c7 89 44 24 0c e8 21 ec 02 00 8b 44
> RSP: 002b:00007ffd1c844f50 EFLAGS: 00000293 ORIG_RAX: 0000000000000003
> RAX: ffffffffffffffda RBX: 0000000000000007 RCX: 00007f338be5659b
> RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000006
> RBP: 0000000000000032 R08: 0000000000000000 R09: 000000000487598e
> R10: 0000000000000000 R11: 0000000000000293 R12: 00007f338bef32fc
> R13: 00007ffd1c844fa0 R14: 00007ffd1c844fc0 R15: 00007f338be0a5f0
> </TASK>
>
> Allocated by task 10619:
> kasan_save_stack+0x22/0x40 mm/kasan/common.c:45
> kasan_set_track+0x25/0x30 mm/kasan/common.c:52
> ____kasan_kmalloc mm/kasan/common.c:374 [inline]
> ____kasan_kmalloc mm/kasan/common.c:333 [inline]
> __kasan_kmalloc+0xa2/0xb0 mm/kasan/common.c:383
> kmalloc ./include/linux/slab.h:599 [inline]
> kzalloc ./include/linux/slab.h:720 [inline]
> hci_conn_add+0xc2/0x1850 net/bluetooth/hci_conn.c:957
> hci_connect_sco+0x3bb/0x1050 net/bluetooth/hci_conn.c:1701
> sco_connect net/bluetooth/sco.c:266 [inline]
> sco_sock_connect+0x2df/0xa90 net/bluetooth/sco.c:591
> __sys_connect_file+0x15f/0x1a0 net/socket.c:2050
> __sys_connect+0x165/0x1a0 net/socket.c:2067
> __do_sys_connect net/socket.c:2077 [inline]
> __se_sys_connect net/socket.c:2074 [inline]
> __x64_sys_connect+0x72/0xb0 net/socket.c:2074
> do_syscall_x64 arch/x86/entry/common.c:50 [inline]
> do_syscall_64+0x39/0xb0 arch/x86/entry/common.c:80
> entry_SYSCALL_64_after_hwframe+0x63/0xcd
>
> Freed by task 8146:
> kasan_save_stack+0x22/0x40 mm/kasan/common.c:45
> kasan_set_track+0x25/0x30 mm/kasan/common.c:52
> kasan_save_free_info+0x2b/0x40 mm/kasan/generic.c:522
> ____kasan_slab_free mm/kasan/common.c:236 [inline]
> ____kasan_slab_free+0x15e/0x1c0 mm/kasan/common.c:200
> kasan_slab_free ./include/linux/kasan.h:164 [inline]
> slab_free_hook mm/slub.c:1800 [inline]
> slab_free_freelist_hook+0x95/0x1d0 mm/slub.c:1826
> slab_free mm/slub.c:3809 [inline]
> __kmem_cache_free+0xb8/0x2e0 mm/slub.c:3822
> device_release+0xa3/0x240 drivers/base/core.c:2484
> kobject_cleanup lib/kobject.c:682 [inline]
> kobject_release lib/kobject.c:716 [inline]
> kref_put ./include/linux/kref.h:65 [inline]
> kobject_put+0x1a2/0x3e0 lib/kobject.c:733
> put_device+0x1f/0x30 drivers/base/core.c:3732
> hci_abort_conn_sync+0x566/0xde0 net/bluetooth/hci_sync.c:5428
> abort_conn_sync+0x188/0x3a0 net/bluetooth/hci_conn.c:2910
> hci_cmd_sync_work+0x1c5/0x430 net/bluetooth/hci_sync.c:306
> process_one_work+0x876/0x15a0 kernel/workqueue.c:2630
> process_scheduled_works kernel/workqueue.c:2703 [inline]
> worker_thread+0x855/0x11f0 kernel/workqueue.c:2784
> kthread+0x346/0x450 kernel/kthread.c:388
> ret_from_fork+0x45/0x80 arch/x86/kernel/process.c:147
> ret_from_fork_asm+0x11/0x20 arch/x86/entry/entry_64.S:304
>
> The buggy address belongs to the object at ffff88801c47a000
> which belongs to the cache kmalloc-4k of size 4096
> The buggy address is located 16 bytes inside of
> freed 4096-byte region [ffff88801c47a000, ffff88801c47b000)
>
> The buggy address belongs to the physical page:
> page:ffffea0000711e00 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x1c478
> head:ffffea0000711e00 order:3 entire_mapcount:0 nr_pages_mapped:0 pincount:0
> anon flags: 0xfff00000000840(slab|head|node=0|zone=1|lastcpupid=0x7ff)
> page_type: 0xffffffff()
> raw: 00fff00000000840 ffff888012843040 0000000000000000 dead000000000001
> raw: 0000000000000000 0000000000040004 00000001ffffffff 0000000000000000
> page dumped because: kasan: bad access detected
> page_owner tracks the page as allocated
> page last allocated via order 3, migratetype Unmovable, gfp_mask 0xd20c0(__GFP_IO|__GFP_FS|__GFP_NOWARN|__GFP_NORETRY|__GFP_COMP|__GFP_NOMEMALLOC), pid 1, tgid 1 (swapper/0), ts 6803836749, free_ts 0
> set_page_owner ./include/linux/page_owner.h:31 [inline]
> post_alloc_hook+0x2d8/0x350 mm/page_alloc.c:1536
> prep_new_page mm/page_alloc.c:1543 [inline]
> get_page_from_freelist+0xf09/0x2c50 mm/page_alloc.c:3170
> __alloc_pages+0x1cb/0x4a0 mm/page_alloc.c:4426
> alloc_page_interleave+0x1e/0x250 mm/mempolicy.c:2130
> alloc_pages+0x233/0x270 mm/mempolicy.c:2292
> alloc_slab_page mm/slub.c:1870 [inline]
> allocate_slab+0x261/0x390 mm/slub.c:2017
> new_slab mm/slub.c:2070 [inline]
> ___slab_alloc+0xbda/0x15e0 mm/slub.c:3223
> __slab_alloc.constprop.0+0x56/0xa0 mm/slub.c:3322
> __slab_alloc_node mm/slub.c:3375 [inline]
> slab_alloc_node mm/slub.c:3468 [inline]
> __kmem_cache_alloc_node+0x13a/0x350 mm/slub.c:3517
> kmalloc_node_trace+0x21/0xd0 mm/slab_common.c:1130
> kmalloc_node ./include/linux/slab.h:615 [inline]
> kzalloc_node ./include/linux/slab.h:731 [inline]
> bdi_alloc+0x47/0x180 mm/backing-dev.c:932
> __alloc_disk_node+0xa2/0x650 block/genhd.c:1337
> __blk_alloc_disk+0x37/0x90 block/genhd.c:1393
> brd_alloc.part.0+0x256/0x770 drivers/block/brd.c:338
> brd_alloc drivers/block/brd.c:458 [inline]
> brd_init+0x1b5/0x2a0 drivers/block/brd.c:425
> do_one_initcall+0x105/0x620 init/main.c:1232
> page_owner free stack trace missing
>
> Memory state around the buggy address:
> ffff88801c479f00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> ffff88801c479f80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> > ffff88801c47a000: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> ^
> ffff88801c47a080: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> ffff88801c47a100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> ==================================================================
>
> Signed-off-by: Chaoyuan Peng <hedonistsmith@xxxxxxxxx>
> ---
> net/bluetooth/sco.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/net/bluetooth/sco.c b/net/bluetooth/sco.c
> index c736186ab..46e158bdd 100644
> --- a/net/bluetooth/sco.c
> +++ b/net/bluetooth/sco.c
> @@ -296,6 +296,7 @@ static int sco_connect(struct sock *sk)
> sco_sock_set_timer(sk, sk->sk_sndtimeo);
> }
>
> + hci_conn_get(hcon);
> release_sock(sk);
>
> unlock:
> @@ -438,12 +439,13 @@ static void __sco_sock_close(struct sock *sk)
> case BT_CONNECTED:
> case BT_CONFIG:
> if (sco_pi(sk)->conn->hcon) {
> + struct hci_conn *hcon = sco_pi(sk)->conn->hcon;
> sk->sk_state = BT_DISCONN;
> sco_sock_set_timer(sk, SCO_DISCONN_TIMEOUT);
> sco_conn_lock(sco_pi(sk)->conn);
> - hci_conn_drop(sco_pi(sk)->conn->hcon);
> sco_pi(sk)->conn->hcon = NULL;
> sco_conn_unlock(sco_pi(sk)->conn);
> + hci_conn_put(hcon);
I think this reference is leaked if sco_conn_del gets called first,
which occurs when the hcon is being deleted.
> } else
> sco_chan_del(sk, ECONNRESET);
> break;
--
Pauli Virtanen
