[PATCH] security/keys: fix slab-out-of-bounds in key_task_permission

[Jarkko Sakkinen]

Hi,

Revisit...

On Fri, 2024-09-13 at 07:09 +0000, Chen Ridong wrote:
> We meet the same issue with the LINK, which reads memory out of
> bounds:

Never ever use pronoun "we" in a commit message in any possible
sentence. Instead always use passive imperative.

What you probably want to say is:

"KASAN reports an out of bounds read:"

Right?

> BUG: KASAN: slab-out-of-bounds in __kuid_val
> include/linux/uidgid.h:36
> BUG: KASAN: slab-out-of-bounds in uid_eq include/linux/uidgid.h:63
> [inline]
> BUG: KASAN: slab-out-of-bounds in key_task_permission+0x394/0x410
> security/keys/permission.c:54
> Read of size 4 at addr ffff88813c3ab618 by task stress-ng/4362
> 
> CPU: 2 PID: 4362 Comm: stress-ng Not tainted 5.10.0-14930-
> gafbffd6c3ede #15
> Call Trace:
>  __dump_stack lib/dump_stack.c:82 [inline]
>  dump_stack+0x107/0x167 lib/dump_stack.c:123
>  print_address_description.constprop.0+0x19/0x170
> mm/kasan/report.c:400
>  __kasan_report.cold+0x6c/0x84 mm/kasan/report.c:560
>  kasan_report+0x3a/0x50 mm/kasan/report.c:585
>  __kuid_val include/linux/uidgid.h:36 [inline]
>  uid_eq include/linux/uidgid.h:63 [inline]
>  key_task_permission+0x394/0x410 security/keys/permission.c:54
>  search_nested_keyrings+0x90e/0xe90 security/keys/keyring.c:793

Snip all below away:

>  keyring_search_rcu+0x1b6/0x310 security/keys/keyring.c:922
>  search_cred_keyrings_rcu+0x111/0x2e0
> security/keys/process_keys.c:459
>  search_process_keyrings_rcu+0x1d/0x310
> security/keys/process_keys.c:544
>  lookup_user_key+0x782/0x12e0 security/keys/process_keys.c:762
>  keyctl_invalidate_key+0x20/0x190 security/keys/keyctl.c:434
>  __do_sys_keyctl security/keys/keyctl.c:1978 [inline]
>  __se_sys_keyctl+0x1de/0x5b0 security/keys/keyctl.c:1880
>  do_syscall_64+0x30/0x40 arch/x86/entry/common.c:46
>  entry_SYSCALL_64_after_hwframe+0x67/0xd1

Remember to cut only the relevant part of the stack trace to make this
commit message more compact and readable.

> 
> However, we can't reproduce this issue.
> After our analysis, it can make this issue by following steps.
> 1.As syzkaller reported, the memory is allocated for struct

"1. "

>   assoc_array_shortcut in the assoc_array_insert_into_terminal_node
>   functions.
> 2.In the search_nested_keyrings, when we go through the slots in a
> node,
>   (bellow tag ascend_to_node), and the slot ptr is meta and
>   node->back_pointer != NULL, we will proceed to  descend_to_node.
>   However, there is an exception. If node is the root, and one of the
>   slots points to a shortcut, it will be treated as a keyring.
> 3.Whether the ptr is keyring decided by keyring_ptr_is_keyring
> function.
>   However, KEYRING_PTR_SUBTYPE is 0x2UL, the same as
>   ASSOC_ARRAY_PTR_SUBTYPE_MASK,
> 4.As mentioned above, If a slot of the root is a shortcut, it may be
>   mistakenly be transferred to a key*, leading to an read out-of-
> bounds
>   read.

Delete the whole list and write a description of the problem and why
your change resolves it.

As per code change, let's layout it something more readable first:

/* Traverse branches into depth: */
if (assoc_array_ptr_is_meta(ptr)) {
	if (node->back_pointer || assoc_array_ptr_is_shortcut(ptr))
		goto descend_to_node;
}

So one thing that should be explained just to make the description
rigid is why 'ptr' is passed to assoc_array_ptr_is_shortcut() and
not 'node'. I'm actually 100% sure about that part, which kind
of supports my view here, right? :-)

The first part of the if-statement obviously filters out everything
that is not root (when it comes to 'node'). Explain the second part.
At that point it is know that node is a root node, so continue from
there.

BR, Jarkko