stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Wu Bo <wubo40@huawei.com>,
	Corey Minyard <cminyard@mvista.com>,
	Sasha Levin <sashal@kernel.org>
Subject: [PATCH 5.4 11/47] ipmi: Fix UAF when uninstall ipmi_si and ipmi_msghandler module
Date: Mon, 27 Dec 2021 16:30:47 +0100	[thread overview]
Message-ID: <20211227151321.178297410@linuxfoundation.org> (raw)
In-Reply-To: <20211227151320.801714429@linuxfoundation.org>

From: Wu Bo <wubo40@huawei.com>

[ Upstream commit ffb76a86f8096a8206be03b14adda6092e18e275 ]

Hi,

When testing install and uninstall of ipmi_si.ko and ipmi_msghandler.ko,
the system crashed.

The log as follows:
[  141.087026] BUG: unable to handle kernel paging request at ffffffffc09b3a5a
[  141.087241] PGD 8fe4c0d067 P4D 8fe4c0d067 PUD 8fe4c0f067 PMD 103ad89067 PTE 0
[  141.087464] Oops: 0010 [#1] SMP NOPTI
[  141.087580] CPU: 67 PID: 668 Comm: kworker/67:1 Kdump: loaded Not tainted 4.18.0.x86_64 #47
[  141.088009] Workqueue: events 0xffffffffc09b3a40
[  141.088009] RIP: 0010:0xffffffffc09b3a5a
[  141.088009] Code: Bad RIP value.
[  141.088009] RSP: 0018:ffffb9094e2c3e88 EFLAGS: 00010246
[  141.088009] RAX: 0000000000000000 RBX: ffff9abfdb1f04a0 RCX: 0000000000000000
[  141.088009] RDX: 0000000000000000 RSI: 0000000000000246 RDI: 0000000000000246
[  141.088009] RBP: 0000000000000000 R08: ffff9abfffee3cb8 R09: 00000000000002e1
[  141.088009] R10: ffffb9094cb73d90 R11: 00000000000f4240 R12: ffff9abfffee8700
[  141.088009] R13: 0000000000000000 R14: ffff9abfdb1f04a0 R15: ffff9abfdb1f04a8
[  141.088009] FS:  0000000000000000(0000) GS:ffff9abfffec0000(0000) knlGS:0000000000000000
[  141.088009] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  141.088009] CR2: ffffffffc09b3a30 CR3: 0000008fe4c0a001 CR4: 00000000007606e0
[  141.088009] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  141.088009] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[  141.088009] PKRU: 55555554
[  141.088009] Call Trace:
[  141.088009]  ? process_one_work+0x195/0x390
[  141.088009]  ? worker_thread+0x30/0x390
[  141.088009]  ? process_one_work+0x390/0x390
[  141.088009]  ? kthread+0x10d/0x130
[  141.088009]  ? kthread_flush_work_fn+0x10/0x10
[  141.088009]  ? ret_from_fork+0x35/0x40] BUG: unable to handle kernel paging request at ffffffffc0b28a5a
[  200.223240] PGD 97fe00d067 P4D 97fe00d067 PUD 97fe00f067 PMD a580cbf067 PTE 0
[  200.223464] Oops: 0010 [#1] SMP NOPTI
[  200.223579] CPU: 63 PID: 664 Comm: kworker/63:1 Kdump: loaded Not tainted 4.18.0.x86_64 #46
[  200.224008] Workqueue: events 0xffffffffc0b28a40
[  200.224008] RIP: 0010:0xffffffffc0b28a5a
[  200.224008] Code: Bad RIP value.
[  200.224008] RSP: 0018:ffffbf3c8e2a3e88 EFLAGS: 00010246
[  200.224008] RAX: 0000000000000000 RBX: ffffa0799ad6bca0 RCX: 0000000000000000
[  200.224008] RDX: 0000000000000000 RSI: 0000000000000246 RDI: 0000000000000246
[  200.224008] RBP: 0000000000000000 R08: ffff9fe43fde3cb8 R09: 00000000000000d5
[  200.224008] R10: ffffbf3c8cb53d90 R11: 00000000000f4240 R12: ffff9fe43fde8700
[  200.224008] R13: 0000000000000000 R14: ffffa0799ad6bca0 R15: ffffa0799ad6bca8
[  200.224008] FS:  0000000000000000(0000) GS:ffff9fe43fdc0000(0000) knlGS:0000000000000000
[  200.224008] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  200.224008] CR2: ffffffffc0b28a30 CR3: 00000097fe00a002 CR4: 00000000007606e0
[  200.224008] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  200.224008] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[  200.224008] PKRU: 55555554
[  200.224008] Call Trace:
[  200.224008]  ? process_one_work+0x195/0x390
[  200.224008]  ? worker_thread+0x30/0x390
[  200.224008]  ? process_one_work+0x390/0x390
[  200.224008]  ? kthread+0x10d/0x130
[  200.224008]  ? kthread_flush_work_fn+0x10/0x10
[  200.224008]  ? ret_from_fork+0x35/0x40
[  200.224008] kernel fault(0x1) notification starting on CPU 63
[  200.224008] kernel fault(0x1) notification finished on CPU 63
[  200.224008] CR2: ffffffffc0b28a5a
[  200.224008] ---[ end trace c82a412d93f57412 ]---

The reason is as follows:
T1: rmmod ipmi_si.
    ->ipmi_unregister_smi()
        -> ipmi_bmc_unregister()
            -> __ipmi_bmc_unregister()
                -> kref_put(&bmc->usecount, cleanup_bmc_device);
                    -> schedule_work(&bmc->remove_work);

T2: rmmod ipmi_msghandler.
    ipmi_msghander module uninstalled, and the module space
    will be freed.

T3: bmc->remove_work doing cleanup the bmc resource.
    -> cleanup_bmc_work()
        -> platform_device_unregister(&bmc->pdev);
            -> platform_device_del(pdev);
                -> device_del(&pdev->dev);
                    -> kobject_uevent(&dev->kobj, KOBJ_REMOVE);
                        -> kobject_uevent_env()
                            -> dev_uevent()
                                -> if (dev->type && dev->type->name)

   'dev->type'(bmc_device_type) pointer space has freed when uninstall
    ipmi_msghander module, 'dev->type->name' cause the system crash.

drivers/char/ipmi/ipmi_msghandler.c:
2820 static const struct device_type bmc_device_type = {
2821         .groups         = bmc_dev_attr_groups,
2822 };

Steps to reproduce:
Add a time delay in cleanup_bmc_work() function,
and uninstall ipmi_si and ipmi_msghandler module.

2910 static void cleanup_bmc_work(struct work_struct *work)
2911 {
2912         struct bmc_device *bmc = container_of(work, struct bmc_device,
2913                                               remove_work);
2914         int id = bmc->pdev.id; /* Unregister overwrites id */
2915
2916         msleep(3000);   <---
2917         platform_device_unregister(&bmc->pdev);
2918         ida_simple_remove(&ipmi_bmc_ida, id);
2919 }

Use 'remove_work_wq' instead of 'system_wq' to solve this issues.

Fixes: b2cfd8ab4add ("ipmi: Rework device id and guid handling to catch changing BMCs")
Signed-off-by: Wu Bo <wubo40@huawei.com>
Message-Id: <1640070034-56671-1-git-send-email-wubo40@huawei.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/char/ipmi/ipmi_msghandler.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/char/ipmi/ipmi_msghandler.c b/drivers/char/ipmi/ipmi_msghandler.c
index 8a93d5a494fff..435e9459a859f 100644
--- a/drivers/char/ipmi/ipmi_msghandler.c
+++ b/drivers/char/ipmi/ipmi_msghandler.c
@@ -2938,7 +2938,7 @@ cleanup_bmc_device(struct kref *ref)
 	 * with removing the device attributes while reading a device
 	 * attribute.
 	 */
-	schedule_work(&bmc->remove_work);
+	queue_work(remove_work_wq, &bmc->remove_work);
 }
 
 /*
-- 
2.34.1




  parent reply	other threads:[~2021-12-27 15:37 UTC|newest]

Thread overview: 54+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-12-27 15:30 [PATCH 5.4 00/47] 5.4.169-rc1 review Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 01/47] net: usb: lan78xx: add Allied Telesis AT29M2-AF Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 02/47] serial: 8250_fintek: Fix garbled text for console Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 03/47] HID: holtek: fix mouse probing Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 04/47] arm64: dts: allwinner: orangepi-zero-plus: fix PHY mode Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 05/47] spi: change clk_disable_unprepare to clk_unprepare Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 06/47] IB/qib: Fix memory leak in qib_user_sdma_queue_pkts() Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 07/47] netfilter: fix regression in looped (broad|multi)casts MAC handling Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 08/47] qlcnic: potential dereference null pointer of rx_queue->page_ring Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 09/47] net: accept UFOv6 packages in virtio_net_hdr_to_skb Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 10/47] net: skip virtio_net_hdr_set_proto if protocol already set Greg Kroah-Hartman
2021-12-27 15:30 ` Greg Kroah-Hartman [this message]
2021-12-27 15:30 ` [PATCH 5.4 12/47] bonding: fix ad_actor_system option setting to default Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 13/47] fjes: Check for error irq Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 14/47] drivers: net: smc911x: " Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 15/47] sfc: falcon: Check null pointer of rx_queue->page_ring Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 16/47] Input: elantech - fix stack out of bound access in elantech_change_report_id() Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 17/47] hwmon: (lm90) Fix usage of CONFIG2 register in detect function Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 18/47] hwmon: (lm90) Add max6654 support to lm90 driver Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 19/47] hwmon: (lm90) Add basic support for TI TMP461 Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 20/47] hwmon: (lm90) Introduce flag indicating extended temperature support Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 21/47] hwmon: (lm90) Drop critical attribute support for MAX6654 Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 22/47] ALSA: jack: Check the return value of kstrdup() Greg Kroah-Hartman
2021-12-27 15:30 ` [PATCH 5.4 23/47] ALSA: drivers: opl3: Fix incorrect use of vp->state Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 24/47] ALSA: hda/realtek: Amp init fixup for HP ZBook 15 G6 Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 25/47] Input: atmel_mxt_ts - fix double free in mxt_read_info_block Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 26/47] ipmi: bail out if init_srcu_struct fails Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 27/47] ipmi: ssif: initialize ssif_info->client early Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 28/47] ipmi: fix initialization when workqueue allocation fails Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 29/47] parisc: Correct completer in lws start Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 30/47] x86/pkey: Fix undefined behaviour with PKRU_WD_BIT Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 31/47] pinctrl: stm32: consider the GPIO offset to expose all the GPIO lines Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 32/47] mmc: sdhci-tegra: Fix switch to HS400ES mode Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 33/47] mmc: core: Disable card detect during shutdown Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 34/47] ARM: 9169/1: entry: fix Thumb2 bug in iWMMXt exception handling Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 35/47] tee: optee: Fix incorrect page free bug Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 36/47] f2fs: fix to do sanity check on last xattr entry in __f2fs_setxattr() Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 37/47] usb: gadget: u_ether: fix race in setting MAC address in setup phase Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 38/47] KVM: VMX: Fix stale docs for kvm-intel.emulate_invalid_guest_state Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 39/47] mm: mempolicy: fix THP allocations escaping mempolicy restrictions Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 40/47] Input: i8042 - enable deferred probe quirk for ASUS UM325UA Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 41/47] pinctrl: mediatek: fix global-out-of-bounds issue Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 42/47] hwmom: (lm90) Fix citical alarm status for MAX6680/MAX6681 Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 43/47] hwmon: (lm90) Do not report busy status bit as alarm Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 44/47] ax25: NPD bug when detaching AX25 device Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 45/47] hamradio: defer ax25 kfree after unregister_netdev Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 46/47] hamradio: improve the incomplete fix to avoid NPD Greg Kroah-Hartman
2021-12-27 15:31 ` [PATCH 5.4 47/47] phonet/pep: refuse to enable an unbound pipe Greg Kroah-Hartman
2021-12-27 17:43 ` [PATCH 5.4 00/47] 5.4.169-rc1 review Florian Fainelli
2021-12-28  0:56 ` Samuel Zou
2021-12-28 10:21 ` Naresh Kamboju
2021-12-28 13:20 ` Sudip Mukherjee
2021-12-28 17:07 ` Guenter Roeck
2021-12-28 21:28 ` Shuah Khan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20211227151321.178297410@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=cminyard@mvista.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=sashal@kernel.org \
    --cc=stable@vger.kernel.org \
    --cc=wubo40@huawei.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).