Re: [PATCH] tipc: Fix uninit-value access in __tipc_nl_bearer_enable()

[Shigeru Yoshida]

On Tue, 03 Oct 2023 10:52:50 +0200, Paolo Abeni wrote:
> On Tue, 2023-09-26 at 21:51 +0900, Shigeru Yoshida wrote:
>> syzbot reported the following uninit-value access issue:
>> 
>> =====================================================
>> BUG: KMSAN: uninit-value in strscpy+0xc4/0x160
>>  strscpy+0xc4/0x160
>>  bearer_name_validate net/tipc/bearer.c:147 [inline]
>>  tipc_enable_bearer net/tipc/bearer.c:259 [inline]
>>  __tipc_nl_bearer_enable+0x634/0x2220 net/tipc/bearer.c:1043
>>  tipc_nl_bearer_enable+0x3c/0x70 net/tipc/bearer.c:1052
>>  genl_family_rcv_msg_doit net/netlink/genetlink.c:971 [inline]
>>  genl_family_rcv_msg net/netlink/genetlink.c:1051 [inline]
>>  genl_rcv_msg+0x11ec/0x1290 net/netlink/genetlink.c:1066
>>  netlink_rcv_skb+0x371/0x650 net/netlink/af_netlink.c:2545
>>  genl_rcv+0x40/0x60 net/netlink/genetlink.c:1075
>>  netlink_unicast_kernel net/netlink/af_netlink.c:1342 [inline]
>>  netlink_unicast+0xf47/0x1250 net/netlink/af_netlink.c:1368
>>  netlink_sendmsg+0x1238/0x13d0 net/netlink/af_netlink.c:1910
>>  sock_sendmsg_nosec net/socket.c:730 [inline]
>>  sock_sendmsg net/socket.c:753 [inline]
>>  ____sys_sendmsg+0x9c2/0xd60 net/socket.c:2540
>>  ___sys_sendmsg+0x28d/0x3c0 net/socket.c:2594
>>  __sys_sendmsg net/socket.c:2623 [inline]
>>  __do_sys_sendmsg net/socket.c:2632 [inline]
>>  __se_sys_sendmsg net/socket.c:2630 [inline]
>>  __x64_sys_sendmsg+0x307/0x490 net/socket.c:2630
>>  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
>> 
>> Uninit was created at:
>>  slab_post_alloc_hook+0x12f/0xb70 mm/slab.h:767
>>  slab_alloc_node mm/slub.c:3478 [inline]
>>  kmem_cache_alloc_node+0x577/0xa80 mm/slub.c:3523
>>  kmalloc_reserve+0x148/0x470 net/core/skbuff.c:559
>>  __alloc_skb+0x318/0x740 net/core/skbuff.c:644
>>  alloc_skb include/linux/skbuff.h:1286 [inline]
>>  netlink_alloc_large_skb net/netlink/af_netlink.c:1214 [inline]
>>  netlink_sendmsg+0xb34/0x13d0 net/netlink/af_netlink.c:1885
>>  sock_sendmsg_nosec net/socket.c:730 [inline]
>>  sock_sendmsg net/socket.c:753 [inline]
>>  ____sys_sendmsg+0x9c2/0xd60 net/socket.c:2540
>>  ___sys_sendmsg+0x28d/0x3c0 net/socket.c:2594
>>  __sys_sendmsg net/socket.c:2623 [inline]
>>  __do_sys_sendmsg net/socket.c:2632 [inline]
>>  __se_sys_sendmsg net/socket.c:2630 [inline]
>>  __x64_sys_sendmsg+0x307/0x490 net/socket.c:2630
>>  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
>> 
>> Bearer names must be null-terminated strings. If a bearer name which is not
>> null-terminated is passed through netlink, strcpy() and similar functions
>> can cause buffer overrun. This causes the above issue.
>> 
>> This patch fixes this issue by returning -EINVAL if a non-null-terminated
>> bearer name is passed.
>> 
>> Fixes: 0655f6a8635b ("tipc: add bearer disable/enable to new netlink api")
>> Reported-and-tested-by: syzbot+9425c47dccbcb4c17d51@xxxxxxxxxxxxxxxxxxxxxxxxx
>> Closes: https://syzkaller.appspot.com/bug?extid=9425c47dccbcb4c17d51
>> Signed-off-by: Shigeru Yoshida <syoshida@xxxxxxxxxx>
>> ---
>>  net/tipc/bearer.c | 4 ++++
>>  1 file changed, 4 insertions(+)
>> 
>> diff --git a/net/tipc/bearer.c b/net/tipc/bearer.c
>> index 2cde375477e3..62047d20e14d 100644
>> --- a/net/tipc/bearer.c
>> +++ b/net/tipc/bearer.c
>> @@ -1025,6 +1025,10 @@ int __tipc_nl_bearer_enable(struct sk_buff *skb, struct genl_info *info)
>>  
>>  	bearer = nla_data(attrs[TIPC_NLA_BEARER_NAME]);
>>  
>> +	if (bearer[strnlen(bearer,
>> +			   nla_len(attrs[TIPC_NLA_BEARER_NAME]))] != '\0')
> 
> if 'bearer' is not NULL terminated, the above will access the first
> byte after the TIPC_NLA_BEARER_NAME attribute.
> 
> I think it would cleaner and safer using nla_strscpy() instead.

Thank you so much for your comment.  I didn't notice the existence of
nla_strscpy().

> Quickly skimming over the tpic code, most TIPC_NLA_BEARER_NAME access
> looks unsafe, and possibly a similar fix should be applied in more
> places.

I've checked the usage of TIPC_NLA_BEARER_NAME accesses.  These might
cause the same issue, so the same fix should be needed, as you say.

I'll send a v2 patch.

Thanks,
Shigeru

> 
> Thanks,
> 
> Paolo
>