Messages in this thread | | | Subject | Re: [PATCH] Bluetooth: Fix UAF for hci_chan | From | "Ziyang Xuan (William)" <> | Date | Sun, 8 Oct 2023 19:49:23 +0800 |
| |
> Hi, > > la, 2023-10-07 kello 17:35 +0800, Ziyang Xuan kirjoitti: >> Syzbot reports a UAF problem as follows: >> >> BUG: KASAN: slab-use-after-free in hci_send_acl+0x48/0xe70 net/bluetooth/hci_core.c:3231 >> ... >> Call Trace: >> <TASK> >> __dump_stack lib/dump_stack.c:88 [inline] >> dump_stack_lvl+0x1e7/0x2d0 lib/dump_stack.c:106 >> print_address_description mm/kasan/report.c:364 [inline] >> print_report+0x163/0x540 mm/kasan/report.c:475 >> kasan_report+0x175/0x1b0 mm/kasan/report.c:588 >> hci_send_acl+0x48/0xe70 net/bluetooth/hci_core.c:3231 >> l2cap_send_conn_req net/bluetooth/l2cap_core.c:1286 [inline] >> l2cap_start_connection+0x465/0x620 net/bluetooth/l2cap_core.c:1514 >> l2cap_conn_start+0xbf3/0x1130 net/bluetooth/l2cap_core.c:1661 >> process_one_work kernel/workqueue.c:2630 [inline] >> process_scheduled_works+0x90f/0x1400 kernel/workqueue.c:2703 >> worker_thread+0xa5f/0xff0 kernel/workqueue.c:2784 >> kthread+0x2d3/0x370 kernel/kthread.c:388 >> ret_from_fork+0x48/0x80 arch/x86/kernel/process.c:147 >> ret_from_fork_asm+0x11/0x20 arch/x86/entry/entry_64.S:304 >> </TASK> > > Note this looks like the same issue as in the following thread, where > there's a suggestion from maintainer: > > https://lore.kernel.org/linux-bluetooth/fd9c0f0eddb8a7d73e9797568dc3cf2bc9b8ad62.1696077070.git.pav@iki.fi/T/#u > > > Is it the same issue?
Yes, Thank you for your reminder.
I think use ida to make handle unique can solve the problem more easily.
I will rebase my patch.
Ziyang Xuan > >> >> Allocated by task 27110: >> kasan_save_stack mm/kasan/common.c:45 [inline] >> kasan_set_track+0x4f/0x70 mm/kasan/common.c:52 >> ____kasan_kmalloc mm/kasan/common.c:374 [inline] >> __kasan_kmalloc+0x98/0xb0 mm/kasan/common.c:383 >> kmalloc include/linux/slab.h:599 [inline] >> kzalloc include/linux/slab.h:720 [inline] >> hci_chan_create+0xc8/0x300 net/bluetooth/hci_conn.c:2714 >> l2cap_conn_add+0x69/0xc10 net/bluetooth/l2cap_core.c:7841 >> l2cap_chan_connect+0x61f/0xeb0 net/bluetooth/l2cap_core.c:8053 >> bt_6lowpan_connect net/bluetooth/6lowpan.c:894 [inline] >> lowpan_control_write+0x55e/0x850 net/bluetooth/6lowpan.c:1129 >> full_proxy_write+0x113/0x1c0 fs/debugfs/file.c:236 >> vfs_write+0x286/0xaf0 fs/read_write.c:582 >> ksys_write+0x1a0/0x2c0 fs/read_write.c:637 >> do_syscall_x64 arch/x86/entry/common.c:50 [inline] >> do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80 >> entry_SYSCALL_64_after_hwframe+0x63/0xcd >> >> Freed by task 5067: >> kasan_save_stack mm/kasan/common.c:45 [inline] >> kasan_set_track+0x4f/0x70 mm/kasan/common.c:52 >> kasan_save_free_info+0x28/0x40 mm/kasan/generic.c:522 >> ____kasan_slab_free+0xd6/0x120 mm/kasan/common.c:236 >> kasan_slab_free include/linux/kasan.h:164 [inline] >> slab_free_hook mm/slub.c:1800 [inline] >> slab_free_freelist_hook mm/slub.c:1826 [inline] >> slab_free mm/slub.c:3809 [inline] >> __kmem_cache_free+0x25f/0x3b0 mm/slub.c:3822 >> hci_chan_list_flush net/bluetooth/hci_conn.c:2754 [inline] >> hci_conn_cleanup net/bluetooth/hci_conn.c:152 [inline] >> hci_conn_del+0x4f8/0xc40 net/bluetooth/hci_conn.c:1156 >> hci_abort_conn_sync+0xa45/0xfe0 >> hci_cmd_sync_work+0x228/0x400 net/bluetooth/hci_sync.c:306 >> process_one_work kernel/workqueue.c:2630 [inline] >> process_scheduled_works+0x90f/0x1400 kernel/workqueue.c:2703 >> worker_thread+0xa5f/0xff0 kernel/workqueue.c:2784 >> kthread+0x2d3/0x370 kernel/kthread.c:388 >> ret_from_fork+0x48/0x80 arch/x86/kernel/process.c:147 >> ret_from_fork_asm+0x11/0x20 arch/x86/entry/entry_64.S:304 >> >> There are two main reasons for this: >> 1. In the case of hci_conn_add() concurrency, the handle of >> hci_conn allocated through hci_conn_hash_alloc_unset() is not unique. >> 2. The processes related to hci_abort_conn() and hci_connect_xxx() >> can be concurrent, but the two processes lack synchronization. >> >> To fix this, use ida to manage the allocation of conn->handle, and >> add the HCI_CONN_DISC flag to achieve synchronization between >> hci_abort_conn() and hci_connect_xxx() related processes. >> >> Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") >> Signed-off-by: Ziyang Xuan <william.xuanziyang@huawei.com>
| |