diff mbox series

[v3] net/vhost: fix potential memory leak

Message ID 20200305025450.5498-1-oda@valinux.co.jp (mailing list archive)
State Accepted, archived
Delegated to: Maxime Coquelin
Headers
Series [v3] net/vhost: fix potential memory leak |

Checks

Context Check Description
ci/checkpatch success coding style OK
ci/iol-testing success Testing PASS
ci/iol-mellanox-Performance success Performance Testing PASS
ci/Intel-compilation success Compilation OK
ci/travis-robot success Travis build: passed

Commit Message

Itsuro Oda March 5, 2020, 2:54 a.m. UTC 
  If a vhost device is closed before eth_dev_configure is done
to the device, internal resources allocated to the device
would not be freed. This patch fixes it.

Fixes: 3d01b759d267 ("net/vhost: delay driver setup")
Cc: stable@dpdk.org

Signed-off-by: Itsuro Oda <oda@valinux.co.jp>
Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>
---
v2:
- fix commit message

v3:
- fix spell error of Reviewed-by

 drivers/net/vhost/rte_eth_vhost.c | 16 +++++++---------
 1 file changed, 7 insertions(+), 9 deletions(-)

Comments

Kevin Traynor March 19, 2020, 5:15 p.m. UTC | #1 
Hi Itsuro,

On 05/03/2020 02:54, Itsuro Oda wrote:
> If a vhost device is closed before eth_dev_configure is done
> to the device, internal resources allocated to the device
> would not be freed. This patch fixes it.
> 
> Fixes: 3d01b759d267 ("net/vhost: delay driver setup")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Itsuro Oda <oda@valinux.co.jp>
> Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>

This fixes an issue with the patch you backported for 18.11. Is the
issue also present in the backported version?

If so, this patch is not in upstream dpdk or gone through validation. So
the choices are,

1. revert your patches from 18.11
2. go ahead on stable without this patch
3. delay until this patch is in master (but not until validated) and
then backport to stable

Itsuro/Maxime, what do you think?

thanks,
Kevin.

> ---
> v2:
> - fix commit message
> 
> v3:
> - fix spell error of Reviewed-by
> 
>  drivers/net/vhost/rte_eth_vhost.c | 16 +++++++---------
>  1 file changed, 7 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/net/vhost/rte_eth_vhost.c b/drivers/net/vhost/rte_eth_vhost.c
> index 458ed58f5..1ed977e9b 100644
> --- a/drivers/net/vhost/rte_eth_vhost.c
> +++ b/drivers/net/vhost/rte_eth_vhost.c
> @@ -1065,16 +1065,14 @@ eth_dev_close(struct rte_eth_dev *dev)
>  
>  	eth_dev_stop(dev);
>  
> -	rte_vhost_driver_unregister(internal->iface_name);
> -
>  	list = find_internal_resource(internal->iface_name);
> -	if (!list)
> -		return;
> -
> -	pthread_mutex_lock(&internal_list_lock);
> -	TAILQ_REMOVE(&internal_list, list, next);
> -	pthread_mutex_unlock(&internal_list_lock);
> -	rte_free(list);
> +	if (list) {
> +		rte_vhost_driver_unregister(internal->iface_name);
> +		pthread_mutex_lock(&internal_list_lock);
> +		TAILQ_REMOVE(&internal_list, list, next);
> +		pthread_mutex_unlock(&internal_list_lock);
> +		rte_free(list);
> +	}
>  
>  	if (dev->data->rx_queues)
>  		for (i = 0; i < dev->data->nb_rx_queues; i++)
>
Maxime Coquelin March 20, 2020, 4:23 p.m. UTC | #2 
On 3/19/20 6:15 PM, Kevin Traynor wrote:
> Hi Itsuro,
> 
> On 05/03/2020 02:54, Itsuro Oda wrote:
>> If a vhost device is closed before eth_dev_configure is done
>> to the device, internal resources allocated to the device
>> would not be freed. This patch fixes it.
>>
>> Fixes: 3d01b759d267 ("net/vhost: delay driver setup")
>> Cc: stable@dpdk.org
>>
>> Signed-off-by: Itsuro Oda <oda@valinux.co.jp>
>> Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>
> 
> This fixes an issue with the patch you backported for 18.11. Is the
> issue also present in the backported version?
> 
> If so, this patch is not in upstream dpdk or gone through validation. So
> the choices are,
> 
> 1. revert your patches from 18.11
> 2. go ahead on stable without this patch
> 3. delay until this patch is in master (but not until validated) and
> then backport to stable
> 
> Itsuro/Maxime, what do you think?

I think you should drop Itsuro patches for now, as long as this patch is
not in master. Secondary process was broken for several revisions in
Vhost, so it is better to keep it broken for now than risking
regressions on primary process support.

If this patch is in master before the next 18.11 is done, then we can
pick all the patches.

Thanks,
Maxime
> thanks,
> Kevin.
> 
>> ---
>> v2:
>> - fix commit message
>>
>> v3:
>> - fix spell error of Reviewed-by
>>
>>  drivers/net/vhost/rte_eth_vhost.c | 16 +++++++---------
>>  1 file changed, 7 insertions(+), 9 deletions(-)
>>
>> diff --git a/drivers/net/vhost/rte_eth_vhost.c b/drivers/net/vhost/rte_eth_vhost.c
>> index 458ed58f5..1ed977e9b 100644
>> --- a/drivers/net/vhost/rte_eth_vhost.c
>> +++ b/drivers/net/vhost/rte_eth_vhost.c
>> @@ -1065,16 +1065,14 @@ eth_dev_close(struct rte_eth_dev *dev)
>>  
>>  	eth_dev_stop(dev);
>>  
>> -	rte_vhost_driver_unregister(internal->iface_name);
>> -
>>  	list = find_internal_resource(internal->iface_name);
>> -	if (!list)
>> -		return;
>> -
>> -	pthread_mutex_lock(&internal_list_lock);
>> -	TAILQ_REMOVE(&internal_list, list, next);
>> -	pthread_mutex_unlock(&internal_list_lock);
>> -	rte_free(list);
>> +	if (list) {
>> +		rte_vhost_driver_unregister(internal->iface_name);
>> +		pthread_mutex_lock(&internal_list_lock);
>> +		TAILQ_REMOVE(&internal_list, list, next);
>> +		pthread_mutex_unlock(&internal_list_lock);
>> +		rte_free(list);
>> +	}
>>  
>>  	if (dev->data->rx_queues)
>>  		for (i = 0; i < dev->data->nb_rx_queues; i++)
>>
>
Kevin Traynor March 20, 2020, 7:25 p.m. UTC | #3 
On 20/03/2020 16:23, Maxime Coquelin wrote:
> 
> 
> On 3/19/20 6:15 PM, Kevin Traynor wrote:
>> Hi Itsuro,
>>
>> On 05/03/2020 02:54, Itsuro Oda wrote:
>>> If a vhost device is closed before eth_dev_configure is done
>>> to the device, internal resources allocated to the device
>>> would not be freed. This patch fixes it.
>>>
>>> Fixes: 3d01b759d267 ("net/vhost: delay driver setup")
>>> Cc: stable@dpdk.org
>>>
>>> Signed-off-by: Itsuro Oda <oda@valinux.co.jp>
>>> Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>
>>
>> This fixes an issue with the patch you backported for 18.11. Is the
>> issue also present in the backported version?
>>
>> If so, this patch is not in upstream dpdk or gone through validation. So
>> the choices are,
>>
>> 1. revert your patches from 18.11
>> 2. go ahead on stable without this patch
>> 3. delay until this patch is in master (but not until validated) and
>> then backport to stable
>>
>> Itsuro/Maxime, what do you think?
> 
> I think you should drop Itsuro patches for now, as long as this patch is
> not in master. Secondary process was broken for several revisions in
> Vhost, so it is better to keep it broken for now than risking
> regressions on primary process support.
> 

I agree, vhost primary is too well used to take risks on regression
while there are still some outstanding issues being resolved with these
patches for secondary process.

Patches reverted.

> If this patch is in master before the next 18.11 is done, then we can
> pick all the patches.
> 

Yes, we can do that.

thanks,
Kevin.

> Thanks,
> Maxime
>> thanks,
>> Kevin.
>>
>>> ---
>>> v2:
>>> - fix commit message
>>>
>>> v3:
>>> - fix spell error of Reviewed-by
>>>
>>>  drivers/net/vhost/rte_eth_vhost.c | 16 +++++++---------
>>>  1 file changed, 7 insertions(+), 9 deletions(-)
>>>
>>> diff --git a/drivers/net/vhost/rte_eth_vhost.c b/drivers/net/vhost/rte_eth_vhost.c
>>> index 458ed58f5..1ed977e9b 100644
>>> --- a/drivers/net/vhost/rte_eth_vhost.c
>>> +++ b/drivers/net/vhost/rte_eth_vhost.c
>>> @@ -1065,16 +1065,14 @@ eth_dev_close(struct rte_eth_dev *dev)
>>>  
>>>  	eth_dev_stop(dev);
>>>  
>>> -	rte_vhost_driver_unregister(internal->iface_name);
>>> -
>>>  	list = find_internal_resource(internal->iface_name);
>>> -	if (!list)
>>> -		return;
>>> -
>>> -	pthread_mutex_lock(&internal_list_lock);
>>> -	TAILQ_REMOVE(&internal_list, list, next);
>>> -	pthread_mutex_unlock(&internal_list_lock);
>>> -	rte_free(list);
>>> +	if (list) {
>>> +		rte_vhost_driver_unregister(internal->iface_name);
>>> +		pthread_mutex_lock(&internal_list_lock);
>>> +		TAILQ_REMOVE(&internal_list, list, next);
>>> +		pthread_mutex_unlock(&internal_list_lock);
>>> +		rte_free(list);
>>> +	}
>>>  
>>>  	if (dev->data->rx_queues)
>>>  		for (i = 0; i < dev->data->nb_rx_queues; i++)
>>>
>>
Maxime Coquelin April 10, 2020, 2:44 p.m. UTC | #4 
On 3/5/20 3:54 AM, Itsuro Oda wrote:
> If a vhost device is closed before eth_dev_configure is done
> to the device, internal resources allocated to the device
> would not be freed. This patch fixes it.
> 
> Fixes: 3d01b759d267 ("net/vhost: delay driver setup")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Itsuro Oda <oda@valinux.co.jp>
> Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>
> ---
> v2:
> - fix commit message
> 
> v3:
> - fix spell error of Reviewed-by
> 
>  drivers/net/vhost/rte_eth_vhost.c | 16 +++++++---------
>  1 file changed, 7 insertions(+), 9 deletions(-)

Applied to dpdk-next-virtio/master.

Thanks,
Maxime
diff mbox series

Patch

diff --git a/drivers/net/vhost/rte_eth_vhost.c b/drivers/net/vhost/rte_eth_vhost.c
index 458ed58f5..1ed977e9b 100644
--- a/drivers/net/vhost/rte_eth_vhost.c
+++ b/drivers/net/vhost/rte_eth_vhost.c
@@ -1065,16 +1065,14 @@  eth_dev_close(struct rte_eth_dev *dev)
 
 	eth_dev_stop(dev);
 
-	rte_vhost_driver_unregister(internal->iface_name);
-
 	list = find_internal_resource(internal->iface_name);
-	if (!list)
-		return;
-
-	pthread_mutex_lock(&internal_list_lock);
-	TAILQ_REMOVE(&internal_list, list, next);
-	pthread_mutex_unlock(&internal_list_lock);
-	rte_free(list);
+	if (list) {
+		rte_vhost_driver_unregister(internal->iface_name);
+		pthread_mutex_lock(&internal_list_lock);
+		TAILQ_REMOVE(&internal_list, list, next);
+		pthread_mutex_unlock(&internal_list_lock);
+		rte_free(list);
+	}
 
 	if (dev->data->rx_queues)
 		for (i = 0; i < dev->data->nb_rx_queues; i++)