[PATCH net-next,v2] bonding: use WARN_ON_ONCE instead of BUG in alb_upper_dev_walk

[PATCH net-next,v2] bonding: use WARN_ON_ONCE instead of BUG in alb_upper_dev_walk

[Zhengchao Shao]

If failed to allocate "tags" or could not find the final upper device from
start_dev's upper list in bond_verify_device_path(), only the loopback
detection of the current upper device should be affected, and the system is
no need to be panic.
Using WARN_ON_ONCE here is to avoid spamming the log if there's a lot of
macvlans above the bond.

Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com>
---
v2: use WARN_ON_ONCE instead of WARN_ON
---
 drivers/net/bonding/bond_alb.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c
index dc2c7b979656..a7bad0fff8cb 100644
--- a/drivers/net/bonding/bond_alb.c
+++ b/drivers/net/bonding/bond_alb.c
@@ -984,8 +984,10 @@ static int alb_upper_dev_walk(struct net_device *upper,
 	 */
 	if (netif_is_macvlan(upper) && !strict_match) {
 		tags = bond_verify_device_path(bond->dev, upper, 0);
-		if (IS_ERR_OR_NULL(tags))
-			BUG();
+		if (IS_ERR_OR_NULL(tags)) {
+			WARN_ON_ONCE(1);
+			return 0;
+		}
 		alb_send_lp_vid(slave, upper->dev_addr,
 				tags[0].vlan_proto, tags[0].vlan_id);
 		kfree(tags);
-- 
2.34.1

Re: [PATCH net-next,v2] bonding: use WARN_ON_ONCE instead of BUG in alb_upper_dev_walk

[Hangbin Liu]

On Tue, Nov 14, 2023 at 05:18:29PM +0800, Zhengchao Shao wrote:
> If failed to allocate "tags" or could not find the final upper device from
> start_dev's upper list in bond_verify_device_path(), only the loopback
> detection of the current upper device should be affected, and the system is
> no need to be panic.
> Using WARN_ON_ONCE here is to avoid spamming the log if there's a lot of
> macvlans above the bond.
> 
> Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com>
> ---
> v2: use WARN_ON_ONCE instead of WARN_ON
> ---
>  drivers/net/bonding/bond_alb.c | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c
> index dc2c7b979656..a7bad0fff8cb 100644
> --- a/drivers/net/bonding/bond_alb.c
> +++ b/drivers/net/bonding/bond_alb.c
> @@ -984,8 +984,10 @@ static int alb_upper_dev_walk(struct net_device *upper,
>  	 */
>  	if (netif_is_macvlan(upper) && !strict_match) {
>  		tags = bond_verify_device_path(bond->dev, upper, 0);
> -		if (IS_ERR_OR_NULL(tags))
> -			BUG();
> +		if (IS_ERR_OR_NULL(tags)) {
> +			WARN_ON_ONCE(1);
> +			return 0;

Return 0 for an error looks weird. Should we keep walking the list if
allocate "tags" failed?

Thanks
Hangbin

Re: [PATCH net-next,v2] bonding: use WARN_ON_ONCE instead of BUG in alb_upper_dev_walk

[shaozhengchao]

On 2023/11/14 18:16, Hangbin Liu wrote:
> On Tue, Nov 14, 2023 at 05:18:29PM +0800, Zhengchao Shao wrote:
>> If failed to allocate "tags" or could not find the final upper device from
>> start_dev's upper list in bond_verify_device_path(), only the loopback
>> detection of the current upper device should be affected, and the system is
>> no need to be panic.
>> Using WARN_ON_ONCE here is to avoid spamming the log if there's a lot of
>> macvlans above the bond.
>>
>> Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com>
>> ---
>> v2: use WARN_ON_ONCE instead of WARN_ON
>> ---
>>   drivers/net/bonding/bond_alb.c | 6 ++++--
>>   1 file changed, 4 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c
>> index dc2c7b979656..a7bad0fff8cb 100644
>> --- a/drivers/net/bonding/bond_alb.c
>> +++ b/drivers/net/bonding/bond_alb.c
>> @@ -984,8 +984,10 @@ static int alb_upper_dev_walk(struct net_device *upper,
>>   	 */
>>   	if (netif_is_macvlan(upper) && !strict_match) {
>>   		tags = bond_verify_device_path(bond->dev, upper, 0);
>> -		if (IS_ERR_OR_NULL(tags))
>> -			BUG();
>> +		if (IS_ERR_OR_NULL(tags)) {
>> +			WARN_ON_ONCE(1);
>> +			return 0;
> 
> Return 0 for an error looks weird. Should we keep walking the list if
> allocate "tags" failed?
> 
> Thanks
> Hangbin
> 
Hi Hangbin:
	I think minimizing the impact of a single allocation failure
is OK.

Zhengchao Shao

Re: [PATCH net-next,v2] bonding: use WARN_ON_ONCE instead of BUG in alb_upper_dev_walk

[Jay Vosburgh]

Zhengchao Shao <shaozhengchao@huawei.com> wrote:

>If failed to allocate "tags" or could not find the final upper device from
>start_dev's upper list in bond_verify_device_path(), only the loopback
>detection of the current upper device should be affected, and the system is
>no need to be panic.
>Using WARN_ON_ONCE here is to avoid spamming the log if there's a lot of
>macvlans above the bond.
>
>Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com>
>---
>v2: use WARN_ON_ONCE instead of WARN_ON
>---
> drivers/net/bonding/bond_alb.c | 6 ++++--
> 1 file changed, 4 insertions(+), 2 deletions(-)
>
>diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c
>index dc2c7b979656..a7bad0fff8cb 100644
>--- a/drivers/net/bonding/bond_alb.c
>+++ b/drivers/net/bonding/bond_alb.c
>@@ -984,8 +984,10 @@ static int alb_upper_dev_walk(struct net_device *upper,
> 	 */
> 	if (netif_is_macvlan(upper) && !strict_match) {
> 		tags = bond_verify_device_path(bond->dev, upper, 0);
>-		if (IS_ERR_OR_NULL(tags))
>-			BUG();
>+		if (IS_ERR_OR_NULL(tags)) {
>+			WARN_ON_ONCE(1);
>+			return 0;

	Ok, I know this is what I said, but on reflection, I think this
should really return non-zero to terminate the device walk.  

	-J


>+		}
> 		alb_send_lp_vid(slave, upper->dev_addr,
> 				tags[0].vlan_proto, tags[0].vlan_id);
> 		kfree(tags);
>-- 
>2.34.1

---
	-Jay Vosburgh, jay.vosburgh@canonical.com

Re: [PATCH net-next,v2] bonding: use WARN_ON_ONCE instead of BUG in alb_upper_dev_walk

[shaozhengchao]

On 2023/11/15 3:40, Jay Vosburgh wrote:
> Zhengchao Shao <shaozhengchao@huawei.com> wrote:
> 
>> If failed to allocate "tags" or could not find the final upper device from
>> start_dev's upper list in bond_verify_device_path(), only the loopback
>> detection of the current upper device should be affected, and the system is
>> no need to be panic.
>> Using WARN_ON_ONCE here is to avoid spamming the log if there's a lot of
>> macvlans above the bond.
>>
>> Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com>
>> ---
>> v2: use WARN_ON_ONCE instead of WARN_ON
>> ---
>> drivers/net/bonding/bond_alb.c | 6 ++++--
>> 1 file changed, 4 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c
>> index dc2c7b979656..a7bad0fff8cb 100644
>> --- a/drivers/net/bonding/bond_alb.c
>> +++ b/drivers/net/bonding/bond_alb.c
>> @@ -984,8 +984,10 @@ static int alb_upper_dev_walk(struct net_device *upper,
>> 	 */
>> 	if (netif_is_macvlan(upper) && !strict_match) {
>> 		tags = bond_verify_device_path(bond->dev, upper, 0);
>> -		if (IS_ERR_OR_NULL(tags))
>> -			BUG();
>> +		if (IS_ERR_OR_NULL(tags)) {
>> +			WARN_ON_ONCE(1);
>> +			return 0;
> 
> 	Ok, I know this is what I said, but on reflection, I think this
> should really return non-zero to terminate the device walk.
> 
> 	-J
> 
	After this failure, there is a high probability that walk
process will still fail. Therefore, it is OK to exit directly. Thanks.
I will send V3.

Zhengchao Shao
> 
>> +		}
>> 		alb_send_lp_vid(slave, upper->dev_addr,
>> 				tags[0].vlan_proto, tags[0].vlan_id);
>> 		kfree(tags);
>> -- 
>> 2.34.1
> 
> ---
> 	-Jay Vosburgh, jay.vosburgh@canonical.com