[PATCH] net:sched fix array-index-out-of-bounds in taprio_change

[PATCH] net:sched fix array-index-out-of-bounds in taprio_change

[tcs.kernel]

From: Haimin Zhang <tcs_kernel@tencent.com>

syzbot report an array-index-out-of-bounds in taprio_change
index 16 is out of range for type '__u16 [16]'
that's because mqprio->num_tc is lager than TC_MAX_QUEUE,so we check
the return value of netdev_set_num_tc.

Reported-by: syzbot+2b3e5fb6c7ef285a94f6@syzkaller.appspotmail.com
Signed-off-by: Haimin Zhang <tcs_kernel@tencent.com>
---
 net/sched/sch_taprio.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c
index 9c79374..1ab2fc9 100644
--- a/net/sched/sch_taprio.c
+++ b/net/sched/sch_taprio.c
@@ -1513,7 +1513,9 @@ static int taprio_change(struct Qdisc *sch, struct nlattr *opt,
 	taprio_set_picos_per_byte(dev, q);
 
 	if (mqprio) {
-		netdev_set_num_tc(dev, mqprio->num_tc);
+		err = netdev_set_num_tc(dev, mqprio->num_tc);
+		if (err)
+			goto free_sched;
 		for (i = 0; i < mqprio->num_tc; i++)
 			netdev_set_tc_queue(dev, i,
 					    mqprio->count[i],
-- 
1.8.3.1

Re: [PATCH] net:sched fix array-index-out-of-bounds in taprio_change

[Eric Dumazet]

On 8/11/21 7:10 AM, tcs.kernel@gmail.com wrote:
> From: Haimin Zhang <tcs_kernel@tencent.com>
> 
> syzbot report an array-index-out-of-bounds in taprio_change
> index 16 is out of range for type '__u16 [16]'
> that's because mqprio->num_tc is lager than TC_MAX_QUEUE,so we check
> the return value of netdev_set_num_tc.
> 
> Reported-by: syzbot+2b3e5fb6c7ef285a94f6@syzkaller.appspotmail.com
> Signed-off-by: Haimin Zhang <tcs_kernel@tencent.com>
> ---
>  net/sched/sch_taprio.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c
> index 9c79374..1ab2fc9 100644
> --- a/net/sched/sch_taprio.c
> +++ b/net/sched/sch_taprio.c
> @@ -1513,7 +1513,9 @@ static int taprio_change(struct Qdisc *sch, struct nlattr *opt,
>  	taprio_set_picos_per_byte(dev, q);
>  
>  	if (mqprio) {
> -		netdev_set_num_tc(dev, mqprio->num_tc);
> +		err = netdev_set_num_tc(dev, mqprio->num_tc);
> +		if (err)
> +			goto free_sched;
>  		for (i = 0; i < mqprio->num_tc; i++)
>  			netdev_set_tc_queue(dev, i,
>  					    mqprio->count[i],
> 

When was the bug added ?

Hint: Please provide a Fixes: tag

taprio_parse_mqprio_opt() already checks :

/* Verify num_tc is not out of max range */
if (qopt->num_tc > TC_MAX_QUEUE) {
    NL_SET_ERR_MSG(extack, "Number of traffic classes is outside valid range");
    return -EINVAL;
}

So what is happening exactly ?

Re: [PATCH] net:sched fix array-index-out-of-bounds in taprio_change

[Pavel Skripkin]

On 8/11/21 10:44 AM, Eric Dumazet wrote:
> 
> 
> On 8/11/21 7:10 AM, tcs.kernel@gmail.com wrote:
>> From: Haimin Zhang <tcs_kernel@tencent.com>
>> 
>> syzbot report an array-index-out-of-bounds in taprio_change
>> index 16 is out of range for type '__u16 [16]'
>> that's because mqprio->num_tc is lager than TC_MAX_QUEUE,so we check
>> the return value of netdev_set_num_tc.
>> 
>> Reported-by: syzbot+2b3e5fb6c7ef285a94f6@syzkaller.appspotmail.com
>> Signed-off-by: Haimin Zhang <tcs_kernel@tencent.com>
>> ---
>>  net/sched/sch_taprio.c | 4 +++-
>>  1 file changed, 3 insertions(+), 1 deletion(-)
>> 
>> diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c
>> index 9c79374..1ab2fc9 100644
>> --- a/net/sched/sch_taprio.c
>> +++ b/net/sched/sch_taprio.c
>> @@ -1513,7 +1513,9 @@ static int taprio_change(struct Qdisc *sch, struct nlattr *opt,
>>  	taprio_set_picos_per_byte(dev, q);
>>  
>>  	if (mqprio) {
>> -		netdev_set_num_tc(dev, mqprio->num_tc);
>> +		err = netdev_set_num_tc(dev, mqprio->num_tc);
>> +		if (err)
>> +			goto free_sched;
>>  		for (i = 0; i < mqprio->num_tc; i++)
>>  			netdev_set_tc_queue(dev, i,
>>  					    mqprio->count[i],
>> 
> 
> When was the bug added ?
> 
> Hint: Please provide a Fixes: tag
> 
> taprio_parse_mqprio_opt() already checks :
> 
> /* Verify num_tc is not out of max range */
> if (qopt->num_tc > TC_MAX_QUEUE) {
>      NL_SET_ERR_MSG(extack, "Number of traffic classes is outside valid range");
>      return -EINVAL;
> }
> 
> So what is happening exactly ?
> 
> 

Hi, Eric!

I've looked into this bug, but I decided to write a reproducer for it 
first. Unfortunately, I didn't finish it yesterday, but I have an idea 
about what happened:

taprio_parse_mqprio_opt() may return 0 before qopt->num_tc check:

	/* If num_tc is already set, it means that the user already
	 * configured the mqprio part
	 */
	if (dev->num_tc)
		return 0;

Then taprio_mqprio_cmp() fails here:

	if (!mqprio || mqprio->num_tc != dev->num_tc)
		return -1;

That's why we won't get shift-out-of-bound in taprio_mqprio_cmp().

And finally taprio_change() gets to buggy for with wrong mqprio->num_tc.
I don't know how to reproduce it, but I'll try to finish my reproducer 
this evening.


Does above makes any sense?



With regards,
Pavel Skripkin

Re: [PATCH] net:sched fix array-index-out-of-bounds in taprio_change

[Pavel Skripkin]

On 8/11/21 11:30 AM, Pavel Skripkin wrote:
> On 8/11/21 10:44 AM, Eric Dumazet wrote:
>> 
>> 
>> On 8/11/21 7:10 AM, tcs.kernel@gmail.com wrote:
>>> From: Haimin Zhang <tcs_kernel@tencent.com>
>>> 
>>> syzbot report an array-index-out-of-bounds in taprio_change
>>> index 16 is out of range for type '__u16 [16]'
>>> that's because mqprio->num_tc is lager than TC_MAX_QUEUE,so we check
>>> the return value of netdev_set_num_tc.
>>> 
>>> Reported-by: syzbot+2b3e5fb6c7ef285a94f6@syzkaller.appspotmail.com
>>> Signed-off-by: Haimin Zhang <tcs_kernel@tencent.com>
>>> ---
>>>  net/sched/sch_taprio.c | 4 +++-
>>>  1 file changed, 3 insertions(+), 1 deletion(-)
>>> 
>>> diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c
>>> index 9c79374..1ab2fc9 100644
>>> --- a/net/sched/sch_taprio.c
>>> +++ b/net/sched/sch_taprio.c
>>> @@ -1513,7 +1513,9 @@ static int taprio_change(struct Qdisc *sch, struct nlattr *opt,
>>>  	taprio_set_picos_per_byte(dev, q);
>>>  
>>>  	if (mqprio) {
>>> -		netdev_set_num_tc(dev, mqprio->num_tc);
>>> +		err = netdev_set_num_tc(dev, mqprio->num_tc);
>>> +		if (err)
>>> +			goto free_sched;
>>>  		for (i = 0; i < mqprio->num_tc; i++)
>>>  			netdev_set_tc_queue(dev, i,
>>>  					    mqprio->count[i],
>>> 
>> 
>> When was the bug added ?
>> 
>> Hint: Please provide a Fixes: tag
>> 
>> taprio_parse_mqprio_opt() already checks :
>> 
>> /* Verify num_tc is not out of max range */
>> if (qopt->num_tc > TC_MAX_QUEUE) {
>>      NL_SET_ERR_MSG(extack, "Number of traffic classes is outside valid range");
>>      return -EINVAL;
>> }
>> 
>> So what is happening exactly ?
>> 
>> 
> 
> Hi, Eric!
> 
> I've looked into this bug, but I decided to write a reproducer for it
> first. Unfortunately, I didn't finish it yesterday, but I have an idea
> about what happened:
> 
> taprio_parse_mqprio_opt() may return 0 before qopt->num_tc check:
> 
> 	/* If num_tc is already set, it means that the user already
> 	 * configured the mqprio part
> 	 */
> 	if (dev->num_tc)
> 		return 0;
> 
> Then taprio_mqprio_cmp() fails here:
> 
> 	if (!mqprio || mqprio->num_tc != dev->num_tc)
> 		return -1;
> 
> That's why we won't get shift-out-of-bound in taprio_mqprio_cmp().

			  ^^^^^^^^^^^^^^^^^

			array-index-out-of-bounds

Sorry for confusion



With regards,
Pavel Skripkin

Re: [PATCH] net:sched fix array-index-out-of-bounds in taprio_change

[Haimin Zhang]

在 2021/8/11 15:44, Eric Dumazet 写道:
> 
> 
> On 8/11/21 7:10 AM, tcs.kernel@gmail.com wrote:
>> From: Haimin Zhang <tcs_kernel@tencent.com>
>>
>> syzbot report an array-index-out-of-bounds in taprio_change
>> index 16 is out of range for type '__u16 [16]'
>> that's because mqprio->num_tc is lager than TC_MAX_QUEUE,so we check
>> the return value of netdev_set_num_tc.
>>
>> Reported-by: syzbot+2b3e5fb6c7ef285a94f6@syzkaller.appspotmail.com
>> Signed-off-by: Haimin Zhang <tcs_kernel@tencent.com>
>> ---
>>   net/sched/sch_taprio.c | 4 +++-
>>   1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c
>> index 9c79374..1ab2fc9 100644
>> --- a/net/sched/sch_taprio.c
>> +++ b/net/sched/sch_taprio.c
>> @@ -1513,7 +1513,9 @@ static int taprio_change(struct Qdisc *sch, struct nlattr *opt,
>>   	taprio_set_picos_per_byte(dev, q);
>>   
>>   	if (mqprio) {
>> -		netdev_set_num_tc(dev, mqprio->num_tc);
>> +		err = netdev_set_num_tc(dev, mqprio->num_tc);
>> +		if (err)
>> +			goto free_sched;
>>   		for (i = 0; i < mqprio->num_tc; i++)
>>   			netdev_set_tc_queue(dev, i,
>>   					    mqprio->count[i],
>>
> 
> When was the bug added ?
> 
> Hint: Please provide a Fixes: tag
> 
> taprio_parse_mqprio_opt() already checks :
> 
> /* Verify num_tc is not out of max range */
> if (qopt->num_tc > TC_MAX_QUEUE) {
>      NL_SET_ERR_MSG(extack, "Number of traffic classes is outside valid range");
>      return -EINVAL;
> }
> 
> So what is happening exactly ?
> 
> 
> 
> 
syzkaller reported this problem,the log shows mqprio->count[16] is accessed.
here is the log
https://syzkaller.appspot.com/bug?id=3a3677d4e7539ec5e671a81e32882ad40b5f7b64

the added check logic is hurtlessness,and netdev_set_num_tc does have a 
return value.

Re: [PATCH] net:sched fix array-index-out-of-bounds in taprio_change

[Jakub Kicinski]

On Wed, 11 Aug 2021 13:10:09 +0800 tcs.kernel@gmail.com wrote:
> From: Haimin Zhang <tcs_kernel@tencent.com>
> 
> syzbot report an array-index-out-of-bounds in taprio_change
> index 16 is out of range for type '__u16 [16]'
> that's because mqprio->num_tc is lager than TC_MAX_QUEUE,so we check
> the return value of netdev_set_num_tc.
> 
> Reported-by: syzbot+2b3e5fb6c7ef285a94f6@syzkaller.appspotmail.com
> Signed-off-by: Haimin Zhang <tcs_kernel@tencent.com>
> ---
>  net/sched/sch_taprio.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c
> index 9c79374..1ab2fc9 100644
> --- a/net/sched/sch_taprio.c
> +++ b/net/sched/sch_taprio.c
> @@ -1513,7 +1513,9 @@ static int taprio_change(struct Qdisc *sch, struct nlattr *opt,
>  	taprio_set_picos_per_byte(dev, q);
>  
>  	if (mqprio) {
> -		netdev_set_num_tc(dev, mqprio->num_tc);
> +		err = netdev_set_num_tc(dev, mqprio->num_tc);
> +		if (err)
> +			goto free_sched;

taprio_set_picos_per_byte() already got called and applied some of the
changes. It seems like the early return from taprio_parse_mqprio_opt()
if dev->num_tc is non-zero is incorrect. That function is supposed to
validate that mqprio_opt() is correct AFAIU. That would mean:

Fixes: a3d43c0d56f1 ("taprio: Add support adding an admin schedule")

Vinicius - WDYT?

>  		for (i = 0; i < mqprio->num_tc; i++)
>  			netdev_set_tc_queue(dev, i,
>  					    mqprio->count[i],

Re: [PATCH] net:sched fix array-index-out-of-bounds in taprio_change

[Haimin Zhang]

hi，i wonder to know what‘s going on with this patch ?
are you plannig to merge it ?


在 2021/8/11 16:36, Haimin Zhang 写道:
> 
> 
> 在 2021/8/11 15:44, Eric Dumazet 写道:
>>
>>
>> On 8/11/21 7:10 AM, tcs.kernel@gmail.com wrote:
>>> From: Haimin Zhang <tcs_kernel@tencent.com>
>>>
>>> syzbot report an array-index-out-of-bounds in taprio_change
>>> index 16 is out of range for type '__u16 [16]'
>>> that's because mqprio->num_tc is lager than TC_MAX_QUEUE,so we check
>>> the return value of netdev_set_num_tc.
>>>
>>> Reported-by: syzbot+2b3e5fb6c7ef285a94f6@syzkaller.appspotmail.com
>>> Signed-off-by: Haimin Zhang <tcs_kernel@tencent.com>
>>> ---
>>>   net/sched/sch_taprio.c | 4 +++-
>>>   1 file changed, 3 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c
>>> index 9c79374..1ab2fc9 100644
>>> --- a/net/sched/sch_taprio.c
>>> +++ b/net/sched/sch_taprio.c
>>> @@ -1513,7 +1513,9 @@ static int taprio_change(struct Qdisc *sch, 
>>> struct nlattr *opt,
>>>       taprio_set_picos_per_byte(dev, q);
>>>       if (mqprio) {
>>> -        netdev_set_num_tc(dev, mqprio->num_tc);
>>> +        err = netdev_set_num_tc(dev, mqprio->num_tc);
>>> +        if (err)
>>> +            goto free_sched;
>>>           for (i = 0; i < mqprio->num_tc; i++)
>>>               netdev_set_tc_queue(dev, i,
>>>                           mqprio->count[i],
>>>
>>
>> When was the bug added ?
>>
>> Hint: Please provide a Fixes: tag
>>
>> taprio_parse_mqprio_opt() already checks :
>>
>> /* Verify num_tc is not out of max range */
>> if (qopt->num_tc > TC_MAX_QUEUE) {
>>      NL_SET_ERR_MSG(extack, "Number of traffic classes is outside 
>> valid range");
>>      return -EINVAL;
>> }
>>
>> So what is happening exactly ?
>>
>>
>>
>>
> syzkaller reported this problem,the log shows mqprio->count[16] is 
> accessed.
> here is the log
> https://syzkaller.appspot.com/bug?id=3a3677d4e7539ec5e671a81e32882ad40b5f7b64 
> 
> 
> the added check logic is hurtlessness,and netdev_set_num_tc does have a 
> return value.

Re: [PATCH] net:sched fix array-index-out-of-bounds in taprio_change

[Jakub Kicinski]

On Mon, 30 Aug 2021 10:09:38 +0800 Haimin Zhang wrote:
> hi，i wonder to know what‘s going on with this patch ?
> are you plannig to merge it ?

Please don't top post. As I told you I think that fix belongs in
taprio_parse_mqprio_opt() which is validating inputs, not in the 
middle of the code applying the changes.

[PATCH] net:sched fix array-index-out-of-bounds in taprio_change

[tcs.kernel]

From: Haimin Zhang <tcs_kernel@tencent.com>

Reported-by: syzbot+2b3e5fb6c7ef285a94f6@syzkaller.appspotmail.com
Signed-off-by: Haimin Zhang <tcs_kernel@tencent.com>
---
 net/sched/sch_taprio.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c
index 9c79374..1ab2fc9 100644
--- a/net/sched/sch_taprio.c
+++ b/net/sched/sch_taprio.c
@@ -1513,7 +1513,9 @@ static int taprio_change(struct Qdisc *sch, struct nlattr *opt,
 	taprio_set_picos_per_byte(dev, q);
 
 	if (mqprio) {
-		netdev_set_num_tc(dev, mqprio->num_tc);
+		err = netdev_set_num_tc(dev, mqprio->num_tc);
+		if (err)
+			goto free_sched;
 		for (i = 0; i < mqprio->num_tc; i++)
 			netdev_set_tc_queue(dev, i,
 					    mqprio->count[i],
-- 
1.8.3.1