[PATCH net 0/2] Mellanox driver fixes

[PATCH net 0/2] Mellanox driver fixes

[Or Gerlitz]

Hi Dave, 

Two small fixes, please apply to net.

Both patches should go to 3.19.y too.

thanks, 

Or.

Ido Shamay (1):
  net/mlx4_en: Disbale GRO for incoming loopback/selftest packets

Or Gerlitz (1):
  net/mlx4_core: Fix wrong mask and error flow for the update-qp command

 drivers/net/ethernet/mellanox/mlx4/en_selftest.c   |    8 +++++++-
 drivers/net/ethernet/mellanox/mlx4/qp.c            |    1 -
 .../net/ethernet/mellanox/mlx4/resource_tracker.c  |    9 ++++++---
 include/linux/mlx4/qp.h                            |    2 +-
 4 files changed, 14 insertions(+), 6 deletions(-)

[PATCH net 1/2] net/mlx4_core: Fix wrong mask and error flow for the update-qp command

[Or Gerlitz]

The bit mask for currently supported driver features (MLX4_UPDATE_QP_SUPPORTED_ATTRS)
of the update-qp command was defined twice (using enum value and pre-processor
define directive) and wrong.

The return value of the call to mlx4_update_qp() from within the SRIOV
resource-tracker was wrongly voided down.

Fix both issues.

Fixes: 09e05c3f78e9 ('net/mlx4: Set vlan stripping policy by the right command')
Fixes: ce8d9e0d6746 ('net/mlx4_core: Add UPDATE_QP SRIOV wrapper support')
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx4/qp.c            |    1 -
 .../net/ethernet/mellanox/mlx4/resource_tracker.c  |    9 ++++++---
 include/linux/mlx4/qp.h                            |    2 +-
 3 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx4/qp.c b/drivers/net/ethernet/mellanox/mlx4/qp.c
index 2bb8553..eda29db 100644
--- a/drivers/net/ethernet/mellanox/mlx4/qp.c
+++ b/drivers/net/ethernet/mellanox/mlx4/qp.c
@@ -412,7 +412,6 @@ err_icm:
 
 EXPORT_SYMBOL_GPL(mlx4_qp_alloc);
 
-#define MLX4_UPDATE_QP_SUPPORTED_ATTRS MLX4_UPDATE_QP_SMAC
 int mlx4_update_qp(struct mlx4_dev *dev, u32 qpn,
 		   enum mlx4_update_qp_attr attr,
 		   struct mlx4_update_qp_params *params)
diff --git a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c
index 486e3d2..d97ca88 100644
--- a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c
+++ b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c
@@ -713,7 +713,7 @@ static int update_vport_qp_param(struct mlx4_dev *dev,
 	struct mlx4_vport_oper_state *vp_oper;
 	struct mlx4_priv *priv;
 	u32 qp_type;
-	int port;
+	int port, err = 0;
 
 	port = (qpc->pri_path.sched_queue & 0x40) ? 2 : 1;
 	priv = mlx4_priv(dev);
@@ -738,7 +738,9 @@ static int update_vport_qp_param(struct mlx4_dev *dev,
 			} else {
 				struct mlx4_update_qp_params params = {.flags = 0};
 
-				mlx4_update_qp(dev, qpn, MLX4_UPDATE_QP_VSD, &params);
+				err = mlx4_update_qp(dev, qpn, MLX4_UPDATE_QP_VSD, &params);
+				if (err)
+					goto out;
 			}
 		}
 
@@ -773,7 +775,8 @@ static int update_vport_qp_param(struct mlx4_dev *dev,
 		qpc->pri_path.feup |= MLX4_FSM_FORCE_ETH_SRC_MAC;
 		qpc->pri_path.grh_mylmc = (0x80 & qpc->pri_path.grh_mylmc) + vp_oper->mac_idx;
 	}
-	return 0;
+out:
+	return err;
 }
 
 static int mpt_mask(struct mlx4_dev *dev)
diff --git a/include/linux/mlx4/qp.h b/include/linux/mlx4/qp.h
index 2bbc62a..551f854 100644
--- a/include/linux/mlx4/qp.h
+++ b/include/linux/mlx4/qp.h
@@ -427,7 +427,7 @@ struct mlx4_wqe_inline_seg {
 
 enum mlx4_update_qp_attr {
 	MLX4_UPDATE_QP_SMAC		= 1 << 0,
-	MLX4_UPDATE_QP_VSD		= 1 << 2,
+	MLX4_UPDATE_QP_VSD		= 1 << 1,
 	MLX4_UPDATE_QP_SUPPORTED_ATTRS	= (1 << 2) - 1
 };
 
-- 
1.7.1

[PATCH net 2/2] net/mlx4_en: Disbale GRO for incoming loopback/selftest packets

[Or Gerlitz]

From: Ido Shamay <idos@mellanox.com>

Packets which are sent from the selftest (ethtool) flow, should not 
be passed to the GRO stack but rather dropped by the driver after validation. 
To achieve that, we disable GRO for the duration of the selftest.

Fixes: dd65beac48a5 ("net/mlx4_en: Extend usage of napi_gro_frags")
Reported-by: Carol Soto <clsoto@linux.vnet.ibm.com>
Signed-off-by: Ido Shamay <idos@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx4/en_selftest.c |    8 +++++++-
 1 files changed, 7 insertions(+), 1 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx4/en_selftest.c b/drivers/net/ethernet/mellanox/mlx4/en_selftest.c
index 2d8ee66..ac8870c 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_selftest.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_selftest.c
@@ -81,12 +81,14 @@ static int mlx4_en_test_loopback(struct mlx4_en_priv *priv)
 {
 	u32 loopback_ok = 0;
 	int i;
-
+	bool gro_enabled;
 
         priv->loopback_ok = 0;
 	priv->validate_loopback = 1;
+	gro_enabled = (priv->dev->features & NETIF_F_GRO);
 
 	mlx4_en_update_loopback_state(priv->dev, priv->dev->features);
+	priv->dev->features &= ~NETIF_F_GRO;
 
 	/* xmit */
 	if (mlx4_en_test_loopback_xmit(priv)) {
@@ -108,6 +110,10 @@ static int mlx4_en_test_loopback(struct mlx4_en_priv *priv)
 mlx4_en_test_loopback_exit:
 
 	priv->validate_loopback = 0;
+
+	if (gro_enabled)
+		priv->dev->features |= NETIF_F_GRO;
+
 	mlx4_en_update_loopback_state(priv->dev, priv->dev->features);
 	return !loopback_ok;
 }
-- 
1.7.1

Re: [PATCH net 2/2] net/mlx4_en: Disbale GRO for incoming loopback/selftest packets

[Sergei Shtylyov]

Hello.

On 3/1/2015 5:56 PM, Or Gerlitz wrote:

> From: Ido Shamay <idos@mellanox.com>

> Packets which are sent from the selftest (ethtool) flow, should not
> be passed to the GRO stack but rather dropped by the driver after validation.
> To achieve that, we disable GRO for the duration of the selftest.

> Fixes: dd65beac48a5 ("net/mlx4_en: Extend usage of napi_gro_frags")
> Reported-by: Carol Soto <clsoto@linux.vnet.ibm.com>
> Signed-off-by: Ido Shamay <idos@mellanox.com>
> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
> ---
>   drivers/net/ethernet/mellanox/mlx4/en_selftest.c |    8 +++++++-
>   1 files changed, 7 insertions(+), 1 deletions(-)

> diff --git a/drivers/net/ethernet/mellanox/mlx4/en_selftest.c b/drivers/net/ethernet/mellanox/mlx4/en_selftest.c
> index 2d8ee66..ac8870c 100644
> --- a/drivers/net/ethernet/mellanox/mlx4/en_selftest.c
> +++ b/drivers/net/ethernet/mellanox/mlx4/en_selftest.c
> @@ -81,12 +81,14 @@ static int mlx4_en_test_loopback(struct mlx4_en_priv *priv)
>   {
>   	u32 loopback_ok = 0;
>   	int i;
> -
> +	bool gro_enabled;
>
>           priv->loopback_ok = 0;
>   	priv->validate_loopback = 1;
> +	gro_enabled = (priv->dev->features & NETIF_F_GRO);

    Parens not necessary here.

[...]

WBR, Sergei

Re: [PATCH net 2/2] net/mlx4_en: Disbale GRO for incoming loopback/selftest packets

[Ido Shamay]

On 3/1/2015 7:24 PM, Sergei Shtylyov wrote:
> Hello.
>
> On 3/1/2015 5:56 PM, Or Gerlitz wrote:
>
>> From: Ido Shamay <idos@mellanox.com>
>
>> Packets which are sent from the selftest (ethtool) flow, should not
>> be passed to the GRO stack but rather dropped by the driver after
>> validation.
>> To achieve that, we disable GRO for the duration of the selftest.
>
>> Fixes: dd65beac48a5 ("net/mlx4_en: Extend usage of napi_gro_frags")
>> Reported-by: Carol Soto <clsoto@linux.vnet.ibm.com>
>> Signed-off-by: Ido Shamay <idos@mellanox.com>
>> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
>> ---
>>   drivers/net/ethernet/mellanox/mlx4/en_selftest.c |    8 +++++++-
>>   1 files changed, 7 insertions(+), 1 deletions(-)
>
>> diff --git a/drivers/net/ethernet/mellanox/mlx4/en_selftest.c
>> b/drivers/net/ethernet/mellanox/mlx4/en_selftest.c
>> index 2d8ee66..ac8870c 100644
>> --- a/drivers/net/ethernet/mellanox/mlx4/en_selftest.c
>> +++ b/drivers/net/ethernet/mellanox/mlx4/en_selftest.c
>> @@ -81,12 +81,14 @@ static int mlx4_en_test_loopback(struct
>> mlx4_en_priv *priv)
>>   {
>>       u32 loopback_ok = 0;
>>       int i;
>> -
>> +    bool gro_enabled;
>>
>>           priv->loopback_ok = 0;
>>       priv->validate_loopback = 1;
>> +    gro_enabled = (priv->dev->features & NETIF_F_GRO);
>
>     Parens not necessary here.
>
Will be fixed in V1, thanks.
Any chance checkpatch can identify such cases?

> [...]
>
> WBR, Sergei
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>