Message-ID: <ad7f2b88-d483-47d9-9f62-12b55805e1a3@bytedance.com>
Date:   Fri, 20 Oct 2023 21:51:18 +0800
From:   Qi Zheng <zhengqi.arch@...edance.com>
To:     Matthew Wilcox <willy@...radead.org>
Cc:     Petr Mladek <pmladek@...e.com>,
        Zhang Zhiyu <zhiyuzhang999@...il.com>,
        linux-kernel@...r.kernel.org,
        Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: KASAN: slab-use-after-free Read in radix_tree_lookup in&after
 Linux Kernel 6.4-rc6

Hi all,

On 2023/10/20 20:34, Matthew Wilcox wrote:
> On Fri, Oct 20, 2023 at 10:26:31AM +0200, Petr Mladek wrote:
>> Adding Matthew into Cc in the hope that he is still familiar with the
>> code. Also adding Andrew who accepts patches.
> 
> oh joy.  i love dealing with cves.
> 
>>>> I agree, this issue looks to be in kernel-core radix tree code in ./lib/radix-tree.c in two of any places.
> 
> the radix tree code is the victim here.  maybe also the perpetrator, but
> it's rather hard to say.
> 
> shrink_slab_memcg()
> 	down_read_trylock(&shrinker_rwsem)
> 	shrinker = idr_find(&shrinker_idr, i);
> 
> i assume is the path to this bug.  the reporter didn't run the
> stacktrace through scripts/decode_stacktrace.sh so it's less useful than
> we might want.
> 
> prealloc_memcg_shrinker() calls idr_alloc and idr_remove under
> shrinker_rwsem in write mode, so that should be fine.
> 
> unregister_memcg_shrinker() calls idr_remove after asserting &shrinker_rwsem
> is held (although not asserting that it's held for write ... hmm ... but
> both callers appear to hold it for write anyway)
> 
> so i don't see why we'd get a UAF here.
> 
> anyway, adding Qi Zheng to the cc since they're responsible for the
> shrinker code.

Thanks for CC'ing me, I'd be happy to troubleshoot any issues that may
be shrinker related.

Between v6.4-rc1 and v6.4 versions, we briefly implemented lockless slab
shrink using the SRCU method. In these versions, we call idr_alloc and
idr_remove under shrinker_mutex, and idr_find under srcu_read_lock.

These are all legitimate uses of the IDR APIs and the shrinker_idr
will never be destroyed, so at a quick glance I didn't see why it would
cause UAF here.

Anyway I will keep working on this issue, and it would be nice if
there was a way to reproduce it.

Thanks,
Qi

> 
>>>> 817 void *radix_tree_lookup(const struct radix_tree_root *root, unsigned long index)
>>>> 818 {
>>>> 819 return __radix_tree_lookup(root, index, NULL, NULL);==>
>>>> 820 }
>>>>
>>>>
>>>> ==> 747 void *__radix_tree_lookup(const struct radix_tree_root *root,
>>>> 748 unsigned long index, struct radix_tree_node **nodep,
>>>> 749 void __rcu ***slotp)
>>>> 750 {
>>>> 751 struct radix_tree_node *node, *parent;
>>>> 752 unsigned long maxindex;
>>>> 753 void __rcu **slot;
>>>> 754
>>>> 755 restart:
>>>> 756 parent = NULL;
>>>> 757 slot = (void __rcu **)&root->xa_head;
>>>> 758 radix_tree_load_root(root, &node, &maxindex);
>>>> 759 if (index > maxindex)
>>>> 760 return NULL;
>>>> 761
>>>> 762 while (radix_tree_is_internal_node(node)) {
>>>> ^..may be (1)
>>>> 763 unsigned offset;
>>>> 764
>>>> 765 parent = entry_to_node(node);
>>>> 766 offset = radix_tree_descend(parent, &node, index);
>>>> 767 slot = parent->slots + offset;
>>>> 768 if (node == RADIX_TREE_RETRY)
>>>> 769 goto restart;
>>>> 770 if (parent->shift == 0)
>>>> 771 break;
>>>> 772 }
>>>> 773
>>>> 774 if (nodep)
>>>> 775 *nodep = parent;
>>>> 776 if (slotp)
>>>> 777 *slotp = slot;
>>>> 778 return node;
>>>> 779 }
>>>>
>>>> looking at the backtrack:
>>>>
>>>> 1)
>>>>
>>>> [ 2351.169619][ T88] ==================================================================
>>>> [ 2351.170501][ T88] BUG: KASAN: slab-use-after-free in radix_tree_lookup+0x12d/0x290
>>>> [ 2351.171365][ T88] Read of size 8 at addr ffff8880675ab1a8 by task kswapd0/88
>>>> [ 2351.172129][ T88]
>>>> [ 2351.172393][ T88] CPU: 0 PID: 88 Comm: kswapd0 Not tainted 6.4.0-rc6 #1
>>>> [ 2351.173124][ T88] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1ubuntu1 04/01/2014
>>>> [ 2351.174222][ T88] Call Trace:
>>>> [ 2351.174593][ T88]
>>>> [ 2351.174930][ T88] dump_stack_lvl+0x1ae/0x2a0
>>>> [ 2351.175483][ T88] ? show_regs_print_info+0x20/0x20
>>>> [ 2351.176077][ T88] ? _printk+0xc0/0x100
>>>> [ 2351.176571][ T88] ? vprintk_emit+0x109/0x1e0
>>>> [ 2351.177112][ T88] ? __wake_up_klogd+0xcc/0x100
>>>> [ 2351.177697][ T88] ? log_buf_vmcoreinfo_setup+0x4a0/0x4a0
>>>> [ 2351.178358][ T88] ? _printk+0xc0/0x100
>>>> [ 2351.178830][ T88] print_address_description+0x78/0x390
>>>> [ 2351.179465][ T88] print_report+0x107/0x1e0
>>>> [ 2351.179982][ T88] ? __virt_addr_valid+0x21b/0x2d0
>>>> [ 2351.180581][ T88] ? __phys_addr+0xb5/0x160
>>>> [ 2351.181107][ T88] ? radix_tree_lookup+0x12d/0x290
>>>> [ 2351.181682][ T88] kasan_report+0xcd/0x100
>>>> [ 2351.182171][ T88] ? radix_tree_lookup+0x12d/0x290
>>>> [ 2351.182721][ T88] radix_tree_lookup+0x12d/0x290
>>>> [ 2351.183256][ T88] shrink_slab_memcg+0x369/0x7c0
>>>> [ 2351.183791][ T88] ? shrink_slab+0x3c0/0x3c0
>>>> [ 2351.184657][ T88] ? lockdep_hardirqs_on_prepare+0x3e2/0x750
>>>> [ 2351.186670][ T88] shrink_slab+0xbd/0x3c0
>>>> [ 2351.188672][ T88] ? _raw_spin_unlock_irqrestore+0x8b/0x120
>>>> [ 2351.190625][ T88] ? lockdep_hardirqs_on+0x8d/0x130
>>>> [ 2351.191796][ T88] ? free_shrinker_info_rcu+0x20/0x20
>>>> [ 2351.192381][ T88] shrink_one+0x30e/0x750
>>>> [ 2351.192853][ T88] shrink_many+0x3ce/0x6d0
>>>> [ 2351.193354][ T88] lru_gen_shrink_node+0x3c1/0x5c0
>>>> [ 2351.193907][ T88] ? shrink_node+0xb6/0x1040
>>>> [ 2351.194406][ T88] ? shrink_node+0x1040/0x1040
>>>> [ 2351.194926][ T88] ? pgdat_balanced+0x1e8/0x210
>>>> [ 2351.195465][ T88] balance_pgdat+0xca7/0x1ac0
>>>>
>>>> 2)
>>>> Or could be in
>>>>
>>>> 747 void *__radix_tree_lookup(const struct radix_tree_root *root,
>>>> 748 unsigned long index, struct radix_tree_node **nodep,
>>>> 749 void __rcu ***slotp)
>>>> 750 {
>>>> ..
>>>> 762 while (radix_tree_is_internal_node(node)) {
>>>> 763 unsigned offset;
>>>> 764
>>>> 765 parent = entry_to_node(node);
>>>> 766 offset = radix_tree_descend(parent, &node, index);------- (2)
>>>> ^....//Here
>>>>
>>>>
>>>> CPU: 0 PID: 88 Comm: kswapd0 Not tainted 6.4.0-rc6 #1
>>>> Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1ubuntu1 04/01/2014
>>>> Call Trace:
>>>> __dump_stack lib/dump_stack.c:88 [inline]
>>>> dump_stack_lvl+0x1ae/0x2a0 lib/dump_stack.c:106
>>>> print_address_description+0x78/0x390 mm/kasan/report.c:351
>>>> print_report+0x107/0x1e0 mm/kasan/report.c:462
>>>> kasan_report+0xcd/0x100 mm/kasan/report.c:572
>>>> radix_tree_descend lib/radix-tree.c:87 [inline]
>>>> ^....................................................//here (2)
>>>> __radix_tree_lookup lib/radix-tree.c:764 [inline]
>>>> radix_tree_lookup+0x12d/0x290 lib/radix-tree.c:817
>>>> shrink_slab_memcg+0x369/0x7c0 mm/vmscan.c:970
>>>> shrink_slab+0xbd/0x3c0 mm/vmscan.c:1062
>>>> shrink_one+0x30e/0x750 mm/vmscan.c:5380
>>>> shrink_many+0x3ce/0x6d0 mm/vmscan.c:5430
>>>> lru_gen_shrink_node+0x3c1/0x5c0 mm/vmscan.c:5547
>>>> kswapd_shrink_node mm/vmscan.c:7288 [inline]
>>>> balance_pgdat+0xca7/0x1ac0 mm/vmscan.c:7478
>>>> kswapd+0x2a5/0x540 mm/vmscan.c:7738
>>>> kthread+0x2eb/0x310 kernel/kthread.c:379
>>>> ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:308
>>>>
>>>> So, I have assigned CVE-2023-4610 to this defect.
>>>>
>>>> We will appreciate if this is notified to the upstream and if there are any patch suggestions identified.
>>>>
>>>> Regards,
>>>> Rohit
>>>
>>> Message-ID: <CALf2hKs61RkLpoMNb9b2Zcq-wHYnQceb6+s6sdG6E+a2Egx1CA@...l.gmail.com>
>>>> Hi Rohit,
>>>>
>>>> Thanks for you attention to this issue. May I ask that is
>>>> CVE-2023-4610 an internal cve of RedHat or a cve recognized by Mitre?
>>>>
>>>> Also, I would like to know if the latest status of this issue has been
>>>> reported to Upstream? If not, do I need to forward all the previous
>>>> messages to Upstream such as linux-kernel@...r.kernel.org?
>>>>
>>>> Best Regards,
>>>> Zhiyu
>>>
>>> Message-ID: <29561486.28558.1693805923362@...130155.ycg3.service-now.com>
>>>> Hi Zhiyu,
>>>>
>>>>> May I ask that is CVE-2023-4610 an internal cve of RedHat or a cve recognized by Mitre?
>>>>
>>>> It is recognized by Mitre, so it is regular CVE for everyone. When you will discuss it with Upstream, please refer to this existing CVE num.
>>>>
>>>>> Also, I would like to know if the latest status of this issue has been reported to Upstream?
>>>>> If not, do I need to forward all the previous messages to Upstream such as linux-kernel@...r.kernel.org?
>>>>
>>>> No, from Red Hat we didn't inform Upstream instead of you.
>>>> We keep this one as embargoed currently (not visible outside of Red Hat).
>>>> Please forward all the details to the Upstream ( I think linux-kernel@...r.kernel.org good to start with).
>>>>
>>>> Thank you,
>>>> - Alexander Larkin
>>>> Senior Product Security Engineer
>>>
>>> Hope these would help you check and fix the problem.
>>>
>>> Best regards,
>>> Zhiyu Zhang
>>>
>>>
>>> Zhang Zhiyu <zhiyuzhang999@...il.com> 于2023年8月7日周一 23:54写道：
>>>>
>>>> Hi,
>>>>
>>>> I found a KASAN: slab-use-after-free Read in radix_tree_lookup while
>>>> fuzzing Linux kernel 6.4-rc6 with my modified syzkaller in 24 July.
>>>>
>>>> The report, log, and config can be downloaded from:
>>>> https://drive.google.com/file/d/1KiZCUHEyp-_Mbq8wdXvjPLs6KU-12JwM/view?usp=sharing
>>>>
>>>> Here is the bug-related key info:
>>>>
>>>> BUG: KASAN: slab-use-after-free in radix_tree_descend
>>>> lib/radix-tree.c:87 [inline]
>>>> BUG: KASAN: slab-use-after-free in __radix_tree_lookup
>>>> lib/radix-tree.c:764 [inline]
>>>> BUG: KASAN: slab-use-after-free in radix_tree_lookup+0x12d/0x290
>>>> lib/radix-tree.c:817
>>>> Read of size 8 at addr ffff8880675ab1a8 by task kswapd0/88
>>>>
>>>> I have preliminarily anlyzed the root cause. The suspected UAF Read is
>>>> located in the __radix_tree_lookup function, which is part of the
>>>> Radix Tree implementation in the Linux kernel. The condition check
>>>> radix_tree_is_internal_node(node) in the while loop is used to verify
>>>> if the node is an internal node. However, this check does not fully
>>>> guarantee that the passed pointer node is always valid. If an invalid
>>>> node pointer is passed to this function, it could lead to undefined
>>>> behavior, potentially including a Use-After-Free Read.
>>>>
>>>> As this is a data race uaf, the syzkaller* cannot easily generate POC.
>>>> I am still trying to construct a POC. By comparing the code of
>>>> radix-tree.c, the vulnerability affects versions ranging from 6.4-rc6
>>>> to the latest mainline.
>>>>
>>>> I recommend to patch it by adding invalid check of node in
>>>> __radix_tree_lookup, if it is validated as a bug.
>>>>
>>>> Best. Have a good day!
>>>> Zhiyu Zhang
>>>

Powered by blists - more mailing lists