Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <df1ae3ff-3d16-9e82-8234-82913e054444@gmail.com>
Date:   Fri, 9 Sep 2022 10:02:41 +0800
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.11.0
Subject: Re: [PATCH] tty: vt: add a bounds checking in vt_do_kdgkb_ioctl()
Content-Language: en-US
To:     Jiri Slaby <jirislaby@kernel.org>, gregkh@linuxfoundation.org,
        changlianzhi@uniontech.com, dmitry.torokhov@gmail.com
Cc:     linux-kernel@vger.kernel.org
References: <20220908075403.27930-1-hbh25y@gmail.com>
 <6236da54-c651-9dc7-f5ce-824be96b3e9d@kernel.org>
From:   Hangyu Hua <hbh25y@gmail.com>
In-Reply-To: <6236da54-c651-9dc7-f5ce-824be96b3e9d@kernel.org>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

On 8/9/2022 16:10, Jiri Slaby wrote:
> On 08. 09. 22, 9:54, Hangyu Hua wrote:
>> As array_index_nospec's comments indicate，a bounds checking need to add
>> before calling array_index_nospec.
>>
>> Signed-off-by: Hangyu Hua <hbh25y@gmail.com>
>> ---
>>   drivers/tty/vt/keyboard.c | 3 +++
>>   1 file changed, 3 insertions(+)
>>
>> diff --git a/drivers/tty/vt/keyboard.c b/drivers/tty/vt/keyboard.c
>> index be8313cdbac3..b9845455df79 100644
>> --- a/drivers/tty/vt/keyboard.c
>> +++ b/drivers/tty/vt/keyboard.c
>> @@ -2067,6 +2067,9 @@ int vt_do_kdgkb_ioctl(int cmd, struct kbsentry 
>> __user *user_kdgkb, int perm)
>>       if (get_user(kb_func, &user_kdgkb->kb_func))
>>           return -EFAULT;
>> +    if (kb_func >= MAX_NR_FUNC)
> 
> kb_func is unsigned char and MAX_NR_FUNC is 256. So this should be 
> eliminated by the compiler anyway.
> 
> But the check might be a good idea if we ever decide to support more 
> keys. But will/can we? I am not so sure, so adding it right now is kind 
> of superfluous. In any way we'd need to introduce a completely different 
> iterface/ioctls.

If you say so, I think greg should be right. We don't need any bounds 
checking here. It may be a good idea to delete redundant
array_index_nospec. Do i need to make a new patch?

> 
>> +        return -EFAULT;
> 
> EINVAL would be more appropriate, IMO.
> 
>> +
>>       kb_func = array_index_nospec(kb_func, MAX_NR_FUNC);
>>       switch (cmd) {
> 
> thanks,

Thanks,
Hangyu