| From bippy-5f407fcff5a0 Mon Sep 17 00:00:00 2001 |
| From: Greg Kroah-Hartman <gregkh@linuxfoundation.org> |
| To: <linux-cve-announce@vger.kernel.org> |
| Reply-to: <cve@kernel.org>, <linux-kernel@vger.kernel.org> |
| Subject: CVE-2022-49505: NFC: NULL out the dev->rfkill to prevent UAF |
| |
| Description |
| =========== |
| |
| In the Linux kernel, the following vulnerability has been resolved: |
| |
| NFC: NULL out the dev->rfkill to prevent UAF |
| |
| Commit 3e3b5dfcd16a ("NFC: reorder the logic in nfc_{un,}register_device") |
| assumes the device_is_registered() in function nfc_dev_up() will help |
| to check when the rfkill is unregistered. However, this check only |
| take effect when device_del(&dev->dev) is done in nfc_unregister_device(). |
| Hence, the rfkill object is still possible be dereferenced. |
| |
| The crash trace in latest kernel (5.18-rc2): |
| |
| [ 68.760105] ================================================================== |
| [ 68.760330] BUG: KASAN: use-after-free in __lock_acquire+0x3ec1/0x6750 |
| [ 68.760756] Read of size 8 at addr ffff888009c93018 by task fuzz/313 |
| [ 68.760756] |
| [ 68.760756] CPU: 0 PID: 313 Comm: fuzz Not tainted 5.18.0-rc2 #4 |
| [ 68.760756] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.14.0-0-g155821a1990b-prebuilt.qemu.org 04/01/2014 |
| [ 68.760756] Call Trace: |
| [ 68.760756] <TASK> |
| [ 68.760756] dump_stack_lvl+0x57/0x7d |
| [ 68.760756] print_report.cold+0x5e/0x5db |
| [ 68.760756] ? __lock_acquire+0x3ec1/0x6750 |
| [ 68.760756] kasan_report+0xbe/0x1c0 |
| [ 68.760756] ? __lock_acquire+0x3ec1/0x6750 |
| [ 68.760756] __lock_acquire+0x3ec1/0x6750 |
| [ 68.760756] ? lockdep_hardirqs_on_prepare+0x410/0x410 |
| [ 68.760756] ? register_lock_class+0x18d0/0x18d0 |
| [ 68.760756] lock_acquire+0x1ac/0x4f0 |
| [ 68.760756] ? rfkill_blocked+0xe/0x60 |
| [ 68.760756] ? lockdep_hardirqs_on_prepare+0x410/0x410 |
| [ 68.760756] ? mutex_lock_io_nested+0x12c0/0x12c0 |
| [ 68.760756] ? nla_get_range_signed+0x540/0x540 |
| [ 68.760756] ? _raw_spin_lock_irqsave+0x4e/0x50 |
| [ 68.760756] _raw_spin_lock_irqsave+0x39/0x50 |
| [ 68.760756] ? rfkill_blocked+0xe/0x60 |
| [ 68.760756] rfkill_blocked+0xe/0x60 |
| [ 68.760756] nfc_dev_up+0x84/0x260 |
| [ 68.760756] nfc_genl_dev_up+0x90/0xe0 |
| [ 68.760756] genl_family_rcv_msg_doit+0x1f4/0x2f0 |
| [ 68.760756] ? genl_family_rcv_msg_attrs_parse.constprop.0+0x230/0x230 |
| [ 68.760756] ? security_capable+0x51/0x90 |
| [ 68.760756] genl_rcv_msg+0x280/0x500 |
| [ 68.760756] ? genl_get_cmd+0x3c0/0x3c0 |
| [ 68.760756] ? lock_acquire+0x1ac/0x4f0 |
| [ 68.760756] ? nfc_genl_dev_down+0xe0/0xe0 |
| [ 68.760756] ? lockdep_hardirqs_on_prepare+0x410/0x410 |
| [ 68.760756] netlink_rcv_skb+0x11b/0x340 |
| [ 68.760756] ? genl_get_cmd+0x3c0/0x3c0 |
| [ 68.760756] ? netlink_ack+0x9c0/0x9c0 |
| [ 68.760756] ? netlink_deliver_tap+0x136/0xb00 |
| [ 68.760756] genl_rcv+0x1f/0x30 |
| [ 68.760756] netlink_unicast+0x430/0x710 |
| [ 68.760756] ? memset+0x20/0x40 |
| [ 68.760756] ? netlink_attachskb+0x740/0x740 |
| [ 68.760756] ? __build_skb_around+0x1f4/0x2a0 |
| [ 68.760756] netlink_sendmsg+0x75d/0xc00 |
| [ 68.760756] ? netlink_unicast+0x710/0x710 |
| [ 68.760756] ? netlink_unicast+0x710/0x710 |
| [ 68.760756] sock_sendmsg+0xdf/0x110 |
| [ 68.760756] __sys_sendto+0x19e/0x270 |
| [ 68.760756] ? __ia32_sys_getpeername+0xa0/0xa0 |
| [ 68.760756] ? fd_install+0x178/0x4c0 |
| [ 68.760756] ? fd_install+0x195/0x4c0 |
| [ 68.760756] ? kernel_fpu_begin_mask+0x1c0/0x1c0 |
| [ 68.760756] __x64_sys_sendto+0xd8/0x1b0 |
| [ 68.760756] ? lockdep_hardirqs_on+0xbf/0x130 |
| [ 68.760756] ? syscall_enter_from_user_mode+0x1d/0x50 |
| [ 68.760756] do_syscall_64+0x3b/0x90 |
| [ 68.760756] entry_SYSCALL_64_after_hwframe+0x44/0xae |
| [ 68.760756] RIP: 0033:0x7f67fb50e6b3 |
| ... |
| [ 68.760756] RSP: 002b:00007f67fa91fe90 EFLAGS: 00000293 ORIG_RAX: 000000000000002c |
| [ 68.760756] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f67fb50e6b3 |
| [ 68.760756] RDX: 000000000000001c RSI: 0000559354603090 RDI: 0000000000000003 |
| [ 68.760756] RBP: 00007f67fa91ff00 R08: 00007f67fa91fedc R09: 000000000000000c |
| [ 68.760756] R10: 0000000000000000 R11: 0000000000000293 R12: 00007ffe824d496e |
| [ 68.760756] R13: 00007ffe824d496f R14: 00007f67fa120000 R15: 0000000000000003 |
| |
| [ 68.760756] </TASK> |
| [ 68.760756] |
| [ 68.760756] Allocated by task 279: |
| [ 68.760756] kasan_save_stack+0x1e/0x40 |
| [ 68.760756] __kasan_kmalloc+0x81/0xa0 |
| [ 68.760756] rfkill_alloc+0x7f/0x280 |
| [ 68.760756] nfc_register_device+0xa3/0x1a0 |
| [ 68.760756] nci_register_device+0x77a/0xad0 |
| [ 68.760756] nfcmrvl_nci_register_dev+0x20b/0x2c0 |
| [ 68.760756] nfcmrvl_nci_uart_open+0xf2/0x1dd |
| [ 68.760756] nci_uart_tty_ioctl+0x2c3/0x4a0 |
| [ 68.760756] tty_ioctl+0x764/0x1310 |
| [ 68.760756] __x64_sys_ioctl+0x122/0x190 |
| [ 68.760756] do_syscall_64+0x3b/0x90 |
| [ 68.760756] entry_SYSCALL_64_after_hwframe+0x44/0xae |
| [ 68.760756] |
| [ 68.760756] Freed by task 314: |
| [ 68.760756] kasan_save_stack+0x1e/0x40 |
| [ 68.760756] kasan_set_track+0x21/0x30 |
| [ 68.760756] kasan_set_free_info+0x20/0x30 |
| [ 68.760756] __kasan_slab_free+0x108/0x170 |
| [ 68.760756] kfree+0xb0/0x330 |
| [ 68.760756] device_release+0x96/0x200 |
| [ 68.760756] kobject_put+0xf9/0x1d0 |
| [ 68.760756] nfc_unregister_device+0x77/0x190 |
| [ 68.760756] nfcmrvl_nci_unregister_dev+0x88/0xd0 |
| [ 68.760756] nci_uart_tty_close+0xdf/0x180 |
| [ 68.760756] tty_ldisc_kill+0x73/0x110 |
| [ 68.760756] tty_ldisc_hangup+0x281/0x5b0 |
| [ 68.760756] __tty_hangup.part.0+0x431/0x890 |
| [ 68.760756] tty_release+0x3a8/0xc80 |
| [ 68.760756] __fput+0x1f0/0x8c0 |
| [ 68.760756] task_work_run+0xc9/0x170 |
| [ 68.760756] exit_to_user_mode_prepare+0x194/0x1a0 |
| [ 68.760756] syscall_exit_to_user_mode+0x19/0x50 |
| [ 68.760756] do_syscall_64+0x48/0x90 |
| [ 68.760756] entry_SYSCALL_64_after_hwframe+0x44/0xae |
| |
| This patch just add the null out of dev->rfkill to make sure such |
| dereference cannot happen. This is safe since the device_lock() already |
| protect the check/write from data race. |
| |
| The Linux kernel CVE team has assigned CVE-2022-49505 to this issue. |
| |
| |
| Affected and fixed versions |
| =========================== |
| |
| Issue introduced in 4.9.291 with commit ff169909eac9e00bf1aa0af739ba6ddfb1b1d135 and fixed in 4.9.318 with commit a8e03bcad52dc9afabf650fdbad84f739cec9efa |
| Issue introduced in 4.14.256 with commit 47244ac0b65bd74cc70007d8e1bac68bd2baad19 and fixed in 4.14.283 with commit f81270125b50532624400063281e6611ecd61ddf |
| Issue introduced in 4.19.218 with commit c45cea83e13699bdfd47842e04d09dd43af4c371 and fixed in 4.19.247 with commit 6abfaca8711803d0d7cc8c0fac1070a88509d463 |
| Issue introduced in 5.4.162 with commit 307d2e6cebfca9d92f86c8e2c8e3dd4a8be46ba6 and fixed in 5.4.198 with commit fbf9c4c714d3cdeb98b6a18e4d057f931cad1d81 |
| Issue introduced in 5.10.82 with commit 73a0d12114b4bc1a9def79a623264754b9df698e and fixed in 5.10.121 with commit 2a1b5110c95e4d49c8c3906270dfcde680a5a7be |
| Issue introduced in 5.15.5 with commit 8a9c61c3ef187d8891225f9b932390670a43a0d3 and fixed in 5.15.46 with commit 1632be63862f183cd5cf1cc094e698e6ec005dfd |
| Issue introduced in 5.16 with commit 3e3b5dfcd16a3e254aab61bd1e8c417dd4503102 and fixed in 5.17.14 with commit 4a68938f43b7c2663e4c90bb9bbe29ac8b9a42a0 |
| Issue introduced in 5.16 with commit 3e3b5dfcd16a3e254aab61bd1e8c417dd4503102 and fixed in 5.18.3 with commit 4f5d71930f41be78557f9714393179025baacd65 |
| Issue introduced in 5.16 with commit 3e3b5dfcd16a3e254aab61bd1e8c417dd4503102 and fixed in 5.19 with commit 1b0e81416a24d6e9b8c2341e22e8bf48f8b8bfc9 |
| Issue introduced in 4.4.293 with commit 5ef16d2d172ee56714cff37cd005b98aba08ef5a |
| |
| Please see https://www.kernel.org for a full list of currently supported |
| kernel versions by the kernel community. |
| |
| Unaffected versions might change over time as fixes are backported to |
| older supported kernel versions. The official CVE entry at |
| https://cve.org/CVERecord/?id=CVE-2022-49505 |
| will be updated if fixes are backported, please check that for the most |
| up to date information about this issue. |
| |
| |
| Affected files |
| ============== |
| |
| The file(s) affected by this issue are: |
| net/nfc/core.c |
| |
| |
| Mitigation |
| ========== |
| |
| The Linux kernel CVE team recommends that you update to the latest |
| stable kernel version for this, and many other bugfixes. Individual |
| changes are never tested alone, but rather are part of a larger kernel |
| release. Cherry-picking individual commits is not recommended or |
| supported by the Linux kernel community at all. If however, updating to |
| the latest release is impossible, the individual changes to resolve this |
| issue can be found at these commits: |
| https://git.kernel.org/stable/c/a8e03bcad52dc9afabf650fdbad84f739cec9efa |
| https://git.kernel.org/stable/c/f81270125b50532624400063281e6611ecd61ddf |
| https://git.kernel.org/stable/c/6abfaca8711803d0d7cc8c0fac1070a88509d463 |
| https://git.kernel.org/stable/c/fbf9c4c714d3cdeb98b6a18e4d057f931cad1d81 |
| https://git.kernel.org/stable/c/2a1b5110c95e4d49c8c3906270dfcde680a5a7be |
| https://git.kernel.org/stable/c/1632be63862f183cd5cf1cc094e698e6ec005dfd |
| https://git.kernel.org/stable/c/4a68938f43b7c2663e4c90bb9bbe29ac8b9a42a0 |
| https://git.kernel.org/stable/c/4f5d71930f41be78557f9714393179025baacd65 |
| https://git.kernel.org/stable/c/1b0e81416a24d6e9b8c2341e22e8bf48f8b8bfc9 |
