[dpdk-dev] [PATCH] mlx5: fix get link settings call for speed and duplex values

[dpdk-dev] [PATCH] mlx5: fix get link settings call for speed and duplex values

[Gowrishankar]

From: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

mlx5 pmd does not correctly retrieve link speed and duplex values
in linksetting struct through ETHTOOL_GLINKSETTINGS ioctl call.
Due to which pktgen application could not correctly calculate tx/rx
rate and hence, very few packets (of count 32) only sent all the time.

Link settings can be derived through ETHTOOL_GLINKSETTINGS ioctl
only after updating ethtool_link_settings struct with appropriate
link mode bitmask words returned by its kernel side implementation.
With this patch, pktgen app is able to tx/rx packets on mlx5 ports.

Signed-off-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
---
 drivers/net/mlx5/mlx5_ethdev.c | 27 +++++++++++++++++++--------
 1 file changed, 19 insertions(+), 8 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_ethdev.c b/drivers/net/mlx5/mlx5_ethdev.c
index b70b7b9..003cef3 100644
--- a/drivers/net/mlx5/mlx5_ethdev.c
+++ b/drivers/net/mlx5/mlx5_ethdev.c
@@ -119,6 +119,7 @@ struct ethtool_link_settings {
 #define ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT 38
 #define ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT 39
 #endif
+#define ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32	(SCHAR_MAX)
 
 /**
  * Return private structure associated with an Ethernet device.
@@ -807,9 +808,11 @@ struct priv *
 mlx5_link_update_unlocked_gs(struct rte_eth_dev *dev, int wait_to_complete)
 {
 	struct priv *priv = mlx5_get_priv(dev);
-	struct ethtool_link_settings edata = {
-		.cmd = ETHTOOL_GLINKSETTINGS,
-	};
+	struct {
+		struct ethtool_link_settings edata;
+		__u32 link_mode_data[3 * ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32];
+	} ecmd;
+
 	struct ifreq ifr;
 	struct rte_eth_link dev_link;
 	uint64_t sc;
@@ -822,15 +825,23 @@ struct priv *
 	memset(&dev_link, 0, sizeof(dev_link));
 	dev_link.link_status = ((ifr.ifr_flags & IFF_UP) &&
 				(ifr.ifr_flags & IFF_RUNNING));
-	ifr.ifr_data = (void *)&edata;
+	memset(&ecmd, 0, sizeof(ecmd));
+	ecmd.edata.cmd = ETHTOOL_GLINKSETTINGS;
+	ifr.ifr_data = (void *)&ecmd;
 	if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) {
 		DEBUG("ioctl(SIOCETHTOOL, ETHTOOL_GLINKSETTINGS) failed: %s",
 		      strerror(errno));
 		return -1;
 	}
-	dev_link.link_speed = edata.speed;
-	sc = edata.link_mode_masks[0] |
-		((uint64_t)edata.link_mode_masks[1] << 32);
+	ecmd.edata.link_mode_masks_nwords = -ecmd.edata.link_mode_masks_nwords;
+	if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) {
+		DEBUG("ioctl(SIOCETHTOOL, ETHTOOL_GLINKSETTINGS) failed: %s",
+		      strerror(errno));
+		return -1;
+	}
+	dev_link.link_speed = ecmd.edata.speed;
+	sc = ecmd.edata.link_mode_masks[0] |
+		((uint64_t)ecmd.edata.link_mode_masks[1] << 32);
 	priv->link_speed_capa = 0;
 	if (sc & ETHTOOL_LINK_MODE_Autoneg_BIT)
 		priv->link_speed_capa |= ETH_LINK_SPEED_AUTONEG;
@@ -866,7 +877,7 @@ struct priv *
 		  ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT |
 		  ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT))
 		priv->link_speed_capa |= ETH_LINK_SPEED_100G;
-	dev_link.link_duplex = ((edata.duplex == DUPLEX_HALF) ?
+	dev_link.link_duplex = ((ecmd.edata.duplex == DUPLEX_HALF) ?
 				ETH_LINK_HALF_DUPLEX : ETH_LINK_FULL_DUPLEX);
 	dev_link.link_autoneg = !(dev->data->dev_conf.link_speeds &
 				  ETH_LINK_SPEED_FIXED);
-- 
1.9.1

Re: [dpdk-dev] [PATCH] mlx5: fix get link settings call for speed and duplex values

[Nélio Laranjeiro]

Hi Gowrishankar,

Sorry, it took me a little time to test this patch, nice catch.

Small format points, the title should be net/mlx5 and there is a
complain with checkpatch see below.

On Tue, Jul 18, 2017 at 11:58:34AM +0530, Gowrishankar wrote:
> From: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
> 
> mlx5 pmd does not correctly retrieve link speed and duplex values
> in linksetting struct through ETHTOOL_GLINKSETTINGS ioctl call.

I would suggest to make it a little shorter by adding the following fact:

 This returns an inconsistent link status.

It does not only impact traffic generator but anything relying on the
link status.

> Due to which pktgen application could not correctly calculate tx/rx
> rate and hence, very few packets (of count 32) only sent all the time.
> 
> Link settings can be derived through ETHTOOL_GLINKSETTINGS ioctl
> only after updating ethtool_link_settings struct with appropriate
> link mode bitmask words returned by its kernel side implementation.
> With this patch, pktgen app is able to tx/rx packets on mlx5 ports.

As it is fixing an issue it should have the fixes tag line, in this case
add the following one:

 Fixes: 188408719888 ("net/mlx5: fix support for newer link speeds")
 Cc: stable@dpdk.org

> Signed-off-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
> ---
>  drivers/net/mlx5/mlx5_ethdev.c | 27 +++++++++++++++++++--------
>  1 file changed, 19 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/net/mlx5/mlx5_ethdev.c b/drivers/net/mlx5/mlx5_ethdev.c
> index b70b7b9..003cef3 100644
> --- a/drivers/net/mlx5/mlx5_ethdev.c
> +++ b/drivers/net/mlx5/mlx5_ethdev.c
> @@ -119,6 +119,7 @@ struct ethtool_link_settings {
>  #define ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT 38
>  #define ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT 39
>  #endif
> +#define ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32	(SCHAR_MAX)
>  
>  /**
>   * Return private structure associated with an Ethernet device.
> @@ -807,9 +808,11 @@ struct priv *
>  mlx5_link_update_unlocked_gs(struct rte_eth_dev *dev, int wait_to_complete)
>  {
>  	struct priv *priv = mlx5_get_priv(dev);
> -	struct ethtool_link_settings edata = {
> -		.cmd = ETHTOOL_GLINKSETTINGS,
> -	};
> +	struct {
> +		struct ethtool_link_settings edata;
> +		__u32 link_mode_data[3 * ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32];

Checkpatch complains about the line size here.

Thanks,

-- 
Nélio Laranjeiro
6WIND

[dpdk-dev] [PATCH v2] net/mlx5: fix get link settings call for speed and duplex values

[Gowrishankar]

From: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

ETHTOOL_GLINKSETTINGS ioctl call in mlx5 pmd returns inconsistent
link status due to which any application relying on it would not
function correctly.

Changes:
 v2 - coding style update

Fixes: 188408719888 ("net/mlx5: fix support for newer link speeds")
Cc: stable@dpdk.org

Signed-off-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
---
 drivers/net/mlx5/mlx5_ethdev.c | 28 ++++++++++++++++++++--------
 1 file changed, 20 insertions(+), 8 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_ethdev.c b/drivers/net/mlx5/mlx5_ethdev.c
index b70b7b9..3ceb13a 100644
--- a/drivers/net/mlx5/mlx5_ethdev.c
+++ b/drivers/net/mlx5/mlx5_ethdev.c
@@ -119,6 +119,7 @@ struct ethtool_link_settings {
 #define ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT 38
 #define ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT 39
 #endif
+#define ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32	(SCHAR_MAX)
 
 /**
  * Return private structure associated with an Ethernet device.
@@ -807,9 +808,12 @@ struct priv *
 mlx5_link_update_unlocked_gs(struct rte_eth_dev *dev, int wait_to_complete)
 {
 	struct priv *priv = mlx5_get_priv(dev);
-	struct ethtool_link_settings edata = {
-		.cmd = ETHTOOL_GLINKSETTINGS,
-	};
+	struct {
+		struct ethtool_link_settings edata;
+		__u32 link_mode_data[
+			3 * ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32];
+	} ecmd;
+
 	struct ifreq ifr;
 	struct rte_eth_link dev_link;
 	uint64_t sc;
@@ -822,15 +826,23 @@ struct priv *
 	memset(&dev_link, 0, sizeof(dev_link));
 	dev_link.link_status = ((ifr.ifr_flags & IFF_UP) &&
 				(ifr.ifr_flags & IFF_RUNNING));
-	ifr.ifr_data = (void *)&edata;
+	memset(&ecmd, 0, sizeof(ecmd));
+	ecmd.edata.cmd = ETHTOOL_GLINKSETTINGS;
+	ifr.ifr_data = (void *)&ecmd;
 	if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) {
 		DEBUG("ioctl(SIOCETHTOOL, ETHTOOL_GLINKSETTINGS) failed: %s",
 		      strerror(errno));
 		return -1;
 	}
-	dev_link.link_speed = edata.speed;
-	sc = edata.link_mode_masks[0] |
-		((uint64_t)edata.link_mode_masks[1] << 32);
+	ecmd.edata.link_mode_masks_nwords = -ecmd.edata.link_mode_masks_nwords;
+	if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) {
+		DEBUG("ioctl(SIOCETHTOOL, ETHTOOL_GLINKSETTINGS) failed: %s",
+		      strerror(errno));
+		return -1;
+	}
+	dev_link.link_speed = ecmd.edata.speed;
+	sc = ecmd.edata.link_mode_masks[0] |
+		((uint64_t)ecmd.edata.link_mode_masks[1] << 32);
 	priv->link_speed_capa = 0;
 	if (sc & ETHTOOL_LINK_MODE_Autoneg_BIT)
 		priv->link_speed_capa |= ETH_LINK_SPEED_AUTONEG;
@@ -866,7 +878,7 @@ struct priv *
 		  ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT |
 		  ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT))
 		priv->link_speed_capa |= ETH_LINK_SPEED_100G;
-	dev_link.link_duplex = ((edata.duplex == DUPLEX_HALF) ?
+	dev_link.link_duplex = ((ecmd.edata.duplex == DUPLEX_HALF) ?
 				ETH_LINK_HALF_DUPLEX : ETH_LINK_FULL_DUPLEX);
 	dev_link.link_autoneg = !(dev->data->dev_conf.link_speeds &
 				  ETH_LINK_SPEED_FIXED);
-- 
1.9.1

[dpdk-dev] [PATCH] net/mlx5: fix inconsistent link status query

[Gowrishankar]

From: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

ETHTOOL_GLINKSETTINGS ioctl call in mlx5 pmd returns inconsistent
link status due to which any application relying on it would not
function correctly.

Fixes: 188408719888 ("net/mlx5: fix support for newer link speeds")
Cc: stable@dpdk.org

Signed-off-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
---
v3:
 - stick with C standard for uint32_t
 
 drivers/net/mlx5/mlx5_ethdev.c | 28 ++++++++++++++++++++--------
 1 file changed, 20 insertions(+), 8 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_ethdev.c b/drivers/net/mlx5/mlx5_ethdev.c
index b70b7b9..01f881e 100644
--- a/drivers/net/mlx5/mlx5_ethdev.c
+++ b/drivers/net/mlx5/mlx5_ethdev.c
@@ -119,6 +119,7 @@ struct ethtool_link_settings {
 #define ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT 38
 #define ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT 39
 #endif
+#define ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32	(SCHAR_MAX)
 
 /**
  * Return private structure associated with an Ethernet device.
@@ -807,9 +808,12 @@ struct priv *
 mlx5_link_update_unlocked_gs(struct rte_eth_dev *dev, int wait_to_complete)
 {
 	struct priv *priv = mlx5_get_priv(dev);
-	struct ethtool_link_settings edata = {
-		.cmd = ETHTOOL_GLINKSETTINGS,
-	};
+	struct {
+		struct ethtool_link_settings edata;
+		uint32_t link_mode_data[3 *
+					ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32];
+	} ecmd;
+
 	struct ifreq ifr;
 	struct rte_eth_link dev_link;
 	uint64_t sc;
@@ -822,15 +826,23 @@ struct priv *
 	memset(&dev_link, 0, sizeof(dev_link));
 	dev_link.link_status = ((ifr.ifr_flags & IFF_UP) &&
 				(ifr.ifr_flags & IFF_RUNNING));
-	ifr.ifr_data = (void *)&edata;
+	memset(&ecmd, 0, sizeof(ecmd));
+	ecmd.edata.cmd = ETHTOOL_GLINKSETTINGS;
+	ifr.ifr_data = (void *)&ecmd;
 	if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) {
 		DEBUG("ioctl(SIOCETHTOOL, ETHTOOL_GLINKSETTINGS) failed: %s",
 		      strerror(errno));
 		return -1;
 	}
-	dev_link.link_speed = edata.speed;
-	sc = edata.link_mode_masks[0] |
-		((uint64_t)edata.link_mode_masks[1] << 32);
+	ecmd.edata.link_mode_masks_nwords = -ecmd.edata.link_mode_masks_nwords;
+	if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) {
+		DEBUG("ioctl(SIOCETHTOOL, ETHTOOL_GLINKSETTINGS) failed: %s",
+		      strerror(errno));
+		return -1;
+	}
+	dev_link.link_speed = ecmd.edata.speed;
+	sc = ecmd.edata.link_mode_masks[0] |
+		((uint64_t)ecmd.edata.link_mode_masks[1] << 32);
 	priv->link_speed_capa = 0;
 	if (sc & ETHTOOL_LINK_MODE_Autoneg_BIT)
 		priv->link_speed_capa |= ETH_LINK_SPEED_AUTONEG;
@@ -866,7 +878,7 @@ struct priv *
 		  ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT |
 		  ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT))
 		priv->link_speed_capa |= ETH_LINK_SPEED_100G;
-	dev_link.link_duplex = ((edata.duplex == DUPLEX_HALF) ?
+	dev_link.link_duplex = ((ecmd.edata.duplex == DUPLEX_HALF) ?
 				ETH_LINK_HALF_DUPLEX : ETH_LINK_FULL_DUPLEX);
 	dev_link.link_autoneg = !(dev->data->dev_conf.link_speeds &
 				  ETH_LINK_SPEED_FIXED);
-- 
1.9.1

Re: [dpdk-dev] [PATCH] net/mlx5: fix inconsistent link status query

[Nélio Laranjeiro]

On Tue, Jul 25, 2017 at 01:37:39PM +0530, Gowrishankar wrote:
> From: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
> 
> ETHTOOL_GLINKSETTINGS ioctl call in mlx5 pmd returns inconsistent
> link status due to which any application relying on it would not
> function correctly.
> 
> Fixes: 188408719888 ("net/mlx5: fix support for newer link speeds")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
> ---
> v3:
>  - stick with C standard for uint32_t
>  
>  drivers/net/mlx5/mlx5_ethdev.c | 28 ++++++++++++++++++++--------
>  1 file changed, 20 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/net/mlx5/mlx5_ethdev.c b/drivers/net/mlx5/mlx5_ethdev.c
> index b70b7b9..01f881e 100644
> --- a/drivers/net/mlx5/mlx5_ethdev.c
> +++ b/drivers/net/mlx5/mlx5_ethdev.c
> @@ -119,6 +119,7 @@ struct ethtool_link_settings {
>  #define ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT 38
>  #define ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT 39
>  #endif
> +#define ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32	(SCHAR_MAX)

As I have another comment about compilation stuff, I would also ask to
remove those extra spaces.

>  /**
>   * Return private structure associated with an Ethernet device.
> @@ -807,9 +808,12 @@ struct priv *
>  mlx5_link_update_unlocked_gs(struct rte_eth_dev *dev, int wait_to_complete)
>  {
>  	struct priv *priv = mlx5_get_priv(dev);
> -	struct ethtool_link_settings edata = {
> -		.cmd = ETHTOOL_GLINKSETTINGS,
> -	};
> +	struct {

Misses the __extension__ before struct here.
struct ethtool_link_settings is finishing with a flexible array (e.g.
[0])  which avoids it to be included in another structure as it breaks
the standard.  

> +		struct ethtool_link_settings edata;
> +		uint32_t link_mode_data[3 *
> +					ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32];
> +	} ecmd;

Please add this __extension__ and it will be good ;)

Thanks,

-- 
Nélio Laranjeiro
6WIND

[dpdk-dev] [PATCH v4] net/mlx5: fix inconsistent link status query

[Gowrishankar]

From: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

ETHTOOL_GLINKSETTINGS ioctl call in mlx5 pmd returns inconsistent
link status due to which any application relying on it would not
function correctly.

Fixes: 188408719888 ("net/mlx5: fix support for newer link speeds")
Cc: stable@dpdk.org

Signed-off-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
---
v4:
 - use __extension__ for GNU C standard (struct in struct)
 
 drivers/net/mlx5/mlx5_ethdev.c | 28 ++++++++++++++++++++--------
 1 file changed, 20 insertions(+), 8 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_ethdev.c b/drivers/net/mlx5/mlx5_ethdev.c
index b70b7b9..1644546 100644
--- a/drivers/net/mlx5/mlx5_ethdev.c
+++ b/drivers/net/mlx5/mlx5_ethdev.c
@@ -119,6 +119,7 @@ struct ethtool_link_settings {
 #define ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT 38
 #define ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT 39
 #endif
+#define ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32 (SCHAR_MAX)
 
 /**
  * Return private structure associated with an Ethernet device.
@@ -807,9 +808,12 @@ struct priv *
 mlx5_link_update_unlocked_gs(struct rte_eth_dev *dev, int wait_to_complete)
 {
 	struct priv *priv = mlx5_get_priv(dev);
-	struct ethtool_link_settings edata = {
-		.cmd = ETHTOOL_GLINKSETTINGS,
-	};
+	__extension__ struct {
+		struct ethtool_link_settings edata;
+		uint32_t link_mode_data[3 *
+					ETHTOOL_LINK_MODE_MASK_MAX_KERNEL_NU32];
+	} ecmd;
+
 	struct ifreq ifr;
 	struct rte_eth_link dev_link;
 	uint64_t sc;
@@ -822,15 +826,23 @@ struct priv *
 	memset(&dev_link, 0, sizeof(dev_link));
 	dev_link.link_status = ((ifr.ifr_flags & IFF_UP) &&
 				(ifr.ifr_flags & IFF_RUNNING));
-	ifr.ifr_data = (void *)&edata;
+	memset(&ecmd, 0, sizeof(ecmd));
+	ecmd.edata.cmd = ETHTOOL_GLINKSETTINGS;
+	ifr.ifr_data = (void *)&ecmd;
 	if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) {
 		DEBUG("ioctl(SIOCETHTOOL, ETHTOOL_GLINKSETTINGS) failed: %s",
 		      strerror(errno));
 		return -1;
 	}
-	dev_link.link_speed = edata.speed;
-	sc = edata.link_mode_masks[0] |
-		((uint64_t)edata.link_mode_masks[1] << 32);
+	ecmd.edata.link_mode_masks_nwords = -ecmd.edata.link_mode_masks_nwords;
+	if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) {
+		DEBUG("ioctl(SIOCETHTOOL, ETHTOOL_GLINKSETTINGS) failed: %s",
+		      strerror(errno));
+		return -1;
+	}
+	dev_link.link_speed = ecmd.edata.speed;
+	sc = ecmd.edata.link_mode_masks[0] |
+		((uint64_t)ecmd.edata.link_mode_masks[1] << 32);
 	priv->link_speed_capa = 0;
 	if (sc & ETHTOOL_LINK_MODE_Autoneg_BIT)
 		priv->link_speed_capa |= ETH_LINK_SPEED_AUTONEG;
@@ -866,7 +878,7 @@ struct priv *
 		  ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT |
 		  ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT))
 		priv->link_speed_capa |= ETH_LINK_SPEED_100G;
-	dev_link.link_duplex = ((edata.duplex == DUPLEX_HALF) ?
+	dev_link.link_duplex = ((ecmd.edata.duplex == DUPLEX_HALF) ?
 				ETH_LINK_HALF_DUPLEX : ETH_LINK_FULL_DUPLEX);
 	dev_link.link_autoneg = !(dev->data->dev_conf.link_speeds &
 				  ETH_LINK_SPEED_FIXED);
-- 
1.9.1

Re: [dpdk-dev] [PATCH v4] net/mlx5: fix inconsistent link status query

[Nélio Laranjeiro]

On Tue, Jul 25, 2017 at 07:08:07PM +0530, Gowrishankar wrote:
> From: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
> 
> ETHTOOL_GLINKSETTINGS ioctl call in mlx5 pmd returns inconsistent
> link status due to which any application relying on it would not
> function correctly.
> 
> Fixes: 188408719888 ("net/mlx5: fix support for newer link speeds")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>

-- 
Nélio Laranjeiro
6WIND

Re: [dpdk-dev] [PATCH v4] net/mlx5: fix inconsistent link status query

[Ferruh Yigit]

On 7/25/2017 3:01 PM, Nélio Laranjeiro wrote:
> On Tue, Jul 25, 2017 at 07:08:07PM +0530, Gowrishankar wrote:
>> From: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
>>
>> ETHTOOL_GLINKSETTINGS ioctl call in mlx5 pmd returns inconsistent
>> link status due to which any application relying on it would not
>> function correctly.
>>
>> Fixes: 188408719888 ("net/mlx5: fix support for newer link speeds")
>> Cc: stable@dpdk.org
>>
>> Signed-off-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
> 
> Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>

Applied to dpdk-next-net/master, thanks.