Skip to contentSkip to conversation

net/vmxnet3: Added mtu_set() function to allow setting MTU.

PATCH Superseded Myers, Charles 5 participants 14 messages Aug 17, 2019 – Apr 1, 2020

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net/vmxnet3: Added mtu_set() function to allow setting MTU.

When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500.

This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS.

---
 drivers/net/vmxnet3/vmxnet3_ethdev.c | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)
drivers/net/vmxnet3/vmxnet3_ethdev.c +22 -0
index 57feb37..c32c92d 100644
@@ -88,6 +88,7 @@ static void vmxnet3_dev_info_get(struct rte_eth_dev *dev,
 				 struct rte_eth_dev_info *dev_info);
 static const uint32_t *
 vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev);
+static int vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu);
 static int vmxnet3_dev_vlan_filter_set(struct rte_eth_dev *dev,
 				       uint16_t vid, int on);
 static int vmxnet3_dev_vlan_offload_set(struct rte_eth_dev *dev, int mask);
@@ -125,6 +126,7 @@ static const struct eth_dev_ops vmxnet3_eth_dev_ops = {
 	.mac_addr_set         = vmxnet3_mac_addr_set,
 	.dev_infos_get        = vmxnet3_dev_info_get,
 	.dev_supported_ptypes_get = vmxnet3_dev_supported_ptypes_get,
+	.mtu_set              = vmxnet3_dev_mtu_set,
 	.vlan_filter_set      = vmxnet3_dev_vlan_filter_set,
 	.vlan_offload_set     = vmxnet3_dev_vlan_offload_set,
 	.rx_queue_setup       = vmxnet3_dev_rx_queue_setup,
@@ -1161,6 +1163,8 @@ vmxnet3_dev_info_get(struct rte_eth_dev *dev __rte_unused,
 	dev_info->max_tx_queues = VMXNET3_MAX_TX_QUEUES;
 	dev_info->min_rx_bufsize = 1518 + RTE_PKTMBUF_HEADROOM;
 	dev_info->max_rx_pktlen = 16384; /* includes CRC, cf MAXFRS register */
+	dev_info->min_mtu = VMXNET3_MIN_MTU;
+	dev_info->max_mtu = VMXNET3_MAX_MTU;
 	dev_info->speed_capa = ETH_LINK_SPEED_10G;
 	dev_info->max_mac_addrs = VMXNET3_MAX_MAC_ADDRS;
 
@@ -1205,6 +1209,24 @@ vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev)
 }
 
 static int
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
+{
+	if (mtu < VMXNET3_MIN_MTU || mtu > VMXNET3_MAX_MTU) {
+		PMD_DRV_LOG(ERR, "MTU should be between %d and %d",
+			     VMXNET3_MIN_MTU, VMXNET3_MAX_MTU);
+		return -EINVAL;
+	}
+
+	if (dev->data->dev_started) {
+		PMD_DRV_LOG(ERR, "Port %d must be stopped to configure MTU",
+			    dev->data->port_id);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+static int
 vmxnet3_mac_addr_set(struct rte_eth_dev *dev, struct rte_ether_addr *mac_addr)
 {
 	struct vmxnet3_hw *hw = dev->data->dev_private;
Signature · 1 line
2.7.4
#1 lore raw

PATCH v2 1/1 net/vmxnet3: Added mtu_set() function to allow setting MTU.

From: Charles Myers <[email protected]>

When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500.

This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS.

---
SMTP server overwrote the From address display name in previous e-mail. Hopefully this resubmission fixes it.

 drivers/net/vmxnet3/vmxnet3_ethdev.c | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)
drivers/net/vmxnet3/vmxnet3_ethdev.c +22 -0
index 57feb37..c32c92d 100644
@@ -88,6 +88,7 @@ static void vmxnet3_dev_info_get(struct rte_eth_dev *dev,
 				 struct rte_eth_dev_info *dev_info);
 static const uint32_t *
 vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev);
+static int vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu);
 static int vmxnet3_dev_vlan_filter_set(struct rte_eth_dev *dev,
 				       uint16_t vid, int on);
 static int vmxnet3_dev_vlan_offload_set(struct rte_eth_dev *dev, int mask);
@@ -125,6 +126,7 @@ static const struct eth_dev_ops vmxnet3_eth_dev_ops = {
 	.mac_addr_set         = vmxnet3_mac_addr_set,
 	.dev_infos_get        = vmxnet3_dev_info_get,
 	.dev_supported_ptypes_get = vmxnet3_dev_supported_ptypes_get,
+	.mtu_set              = vmxnet3_dev_mtu_set,
 	.vlan_filter_set      = vmxnet3_dev_vlan_filter_set,
 	.vlan_offload_set     = vmxnet3_dev_vlan_offload_set,
 	.rx_queue_setup       = vmxnet3_dev_rx_queue_setup,
@@ -1161,6 +1163,8 @@ vmxnet3_dev_info_get(struct rte_eth_dev *dev __rte_unused,
 	dev_info->max_tx_queues = VMXNET3_MAX_TX_QUEUES;
 	dev_info->min_rx_bufsize = 1518 + RTE_PKTMBUF_HEADROOM;
 	dev_info->max_rx_pktlen = 16384; /* includes CRC, cf MAXFRS register */
+	dev_info->min_mtu = VMXNET3_MIN_MTU;
+	dev_info->max_mtu = VMXNET3_MAX_MTU;
 	dev_info->speed_capa = ETH_LINK_SPEED_10G;
 	dev_info->max_mac_addrs = VMXNET3_MAX_MAC_ADDRS;
 
@@ -1205,6 +1209,24 @@ vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev)
 }
 
 static int
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
+{
+	if (mtu < VMXNET3_MIN_MTU || mtu > VMXNET3_MAX_MTU) {
+		PMD_DRV_LOG(ERR, "MTU should be between %d and %d",
+			     VMXNET3_MIN_MTU, VMXNET3_MAX_MTU);
+		return -EINVAL;
+	}
+
+	if (dev->data->dev_started) {
+		PMD_DRV_LOG(ERR, "Port %d must be stopped to configure MTU",
+			    dev->data->port_id);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+static int
 vmxnet3_mac_addr_set(struct rte_eth_dev *dev, struct rte_ether_addr *mac_addr)
 {
 	struct vmxnet3_hw *hw = dev->data->dev_private;
Signature · 1 line
2.7.4
#2 lore raw

drivers/net/vmxnet3/vmxnet3_ethdev.c vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev) · line 1219

Quoted 9 lines · Charles

static int

drivers/net/vmxnet3/vmxnet3_ethdev.c +8 -0
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
+{
+	if (mtu < VMXNET3_MIN_MTU || mtu > VMXNET3_MAX_MTU) {
+		PMD_DRV_LOG(ERR, "MTU should be between %d and %d",
+			     VMXNET3_MIN_MTU, VMXNET3_MAX_MTU);
+		return -EINVAL;
+	}
+

This is not the best way to handle checking validity of MTU value.

The device should report min/max MTU and then rte_eth_dev_set_mtu will do enforcement there.

This allows application to know what MTU to use, and eliminates guess/check behavior.

#3 lore raw
In reply to Myers, Charles

-----Original Message-----
From: "Myers, Charles" <[email protected]>
Date: Monday, August 19, 2019 at 9:07 PM
To: Yong Wang <[email protected]>
Cc: "[email protected]" <[email protected]>, "Myers, Charles" <[email protected]>
Subject: [PATCH v2] net/vmxnet3: Added mtu_set() function to allow setting MTU.

    From: Charles Myers <[email protected]>
    
    When the mtu_set() function is not implemented, rte_eth_dev_set_mtu()
    fails with -ENOTSUP and mtu is not stored in the mtu field in the
    rte_eth_dev_data.  This causes the mtu in Vmxnet3_MiscConf which is
    shared with hypervisor to always be set to 1500.
    
    This may cause issues receiving jumbo frames on Enhanced Data Path
    N-VDS.
    
    Signed-off-by: Charles Myers <[email protected]>
    ---
    SMTP server overwrote the From address display name in previous e-mail.
    Hopefully this resubmission fixes it.
    
     drivers/net/vmxnet3/vmxnet3_ethdev.c | 22 ++++++++++++++++++++++
     1 file changed, 22 insertions(+)
    
    diff --git a/drivers/net/vmxnet3/vmxnet3_ethdev.c b/drivers/net/vmxnet3/vmxnet3_ethdev.c
    index 57feb37..c32c92d 100644
    --- a/drivers/net/vmxnet3/vmxnet3_ethdev.c
    +++ b/drivers/net/vmxnet3/vmxnet3_ethdev.c
    @@ -88,6 +88,7 @@ static void vmxnet3_dev_info_get(struct rte_eth_dev *dev,
     				 struct rte_eth_dev_info *dev_info);
     static const uint32_t *
     vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev);
    +static int vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu);
     static int vmxnet3_dev_vlan_filter_set(struct rte_eth_dev *dev,
     				       uint16_t vid, int on);
     static int vmxnet3_dev_vlan_offload_set(struct rte_eth_dev *dev, int mask);
    @@ -125,6 +126,7 @@ static const struct eth_dev_ops vmxnet3_eth_dev_ops = {
     	.mac_addr_set         = vmxnet3_mac_addr_set,
     	.dev_infos_get        = vmxnet3_dev_info_get,
     	.dev_supported_ptypes_get = vmxnet3_dev_supported_ptypes_get,
    +	.mtu_set              = vmxnet3_dev_mtu_set,
     	.vlan_filter_set      = vmxnet3_dev_vlan_filter_set,
     	.vlan_offload_set     = vmxnet3_dev_vlan_offload_set,
     	.rx_queue_setup       = vmxnet3_dev_rx_queue_setup,
    @@ -1161,6 +1163,8 @@ vmxnet3_dev_info_get(struct rte_eth_dev *dev __rte_unused,
     	dev_info->max_tx_queues = VMXNET3_MAX_TX_QUEUES;
     	dev_info->min_rx_bufsize = 1518 + RTE_PKTMBUF_HEADROOM;
     	dev_info->max_rx_pktlen = 16384; /* includes CRC, cf MAXFRS register */
    +	dev_info->min_mtu = VMXNET3_MIN_MTU;
    +	dev_info->max_mtu = VMXNET3_MAX_MTU;
     	dev_info->speed_capa = ETH_LINK_SPEED_10G;
     	dev_info->max_mac_addrs = VMXNET3_MAX_MAC_ADDRS;
     
    @@ -1205,6 +1209,24 @@ vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev)
     }
     
     static int
    +vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
    +{
    +	if (mtu < VMXNET3_MIN_MTU || mtu > VMXNET3_MAX_MTU) {
    +		PMD_DRV_LOG(ERR, "MTU should be between %d and %d",
    +			     VMXNET3_MIN_MTU, VMXNET3_MAX_MTU);
    +		return -EINVAL;
    +	}
    +

It looks like the above check is redundant as rte_eth_dev_set_mtu() already checks the same based on dev_info->min_mtu and max_mtu. But seems many other drivers are doing the same check so probably a separate patch to clean this up.

    +	if (dev->data->dev_started) {
    +		PMD_DRV_LOG(ERR, "Port %d must be stopped to configure MTU",
    +			    dev->data->port_id);
    +		return -EBUSY;
    +	}
    +

Can you add the following for the vmxnet3 backend to see the change:

	dev->data->mtu = mtu;  /* this is needed because vmxnet3_dev_start use this to update the backend but  rte_eth_dev_set_mtu() set it after the driver callback */

	/* changing mtu for vmxnet3 pmd does not require a restart                                                                                                                                               
	 * We stop and restart the device here                                                                                                                                              
	 * just to pass the mtu info to the backend. */                                                                                                                                                          
	vmxnet3_dev_stop(dev);                                                                                                                                                                                   
	vmxnet3_dev_start(dev);      

    +	return 0;
    +}
    +
    +static int
     vmxnet3_mac_addr_set(struct rte_eth_dev *dev, struct rte_ether_addr *mac_addr)
     {
     	struct vmxnet3_hw *hw = dev->data->dev_private;
    -- 
    2.7.4
#4 lore raw

PATCH v3 1/1 net/vmxnet3: Added mtu_set() function to allow setting MTU.

In reply to Myers, Charles

From: Charles Myers <[email protected]>

When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500.

This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS.

---
Removed redundant MTU check in vmxnet3_dev_mtu_set() which should already be done in rte_eth_dev_set_mtu().

 drivers/net/vmxnet3/vmxnet3_ethdev.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)
drivers/net/vmxnet3/vmxnet3_ethdev.c +16 -0
index 57feb37..c12d70b 100644
@@ -88,6 +88,7 @@ static void vmxnet3_dev_info_get(struct rte_eth_dev *dev,
 				 struct rte_eth_dev_info *dev_info);
 static const uint32_t *
 vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev);
+static int vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu);
 static int vmxnet3_dev_vlan_filter_set(struct rte_eth_dev *dev,
 				       uint16_t vid, int on);
 static int vmxnet3_dev_vlan_offload_set(struct rte_eth_dev *dev, int mask);
@@ -125,6 +126,7 @@ static const struct eth_dev_ops vmxnet3_eth_dev_ops = {
 	.mac_addr_set         = vmxnet3_mac_addr_set,
 	.dev_infos_get        = vmxnet3_dev_info_get,
 	.dev_supported_ptypes_get = vmxnet3_dev_supported_ptypes_get,
+	.mtu_set              = vmxnet3_dev_mtu_set,
 	.vlan_filter_set      = vmxnet3_dev_vlan_filter_set,
 	.vlan_offload_set     = vmxnet3_dev_vlan_offload_set,
 	.rx_queue_setup       = vmxnet3_dev_rx_queue_setup,
@@ -1161,6 +1163,8 @@ vmxnet3_dev_info_get(struct rte_eth_dev *dev __rte_unused,
 	dev_info->max_tx_queues = VMXNET3_MAX_TX_QUEUES;
 	dev_info->min_rx_bufsize = 1518 + RTE_PKTMBUF_HEADROOM;
 	dev_info->max_rx_pktlen = 16384; /* includes CRC, cf MAXFRS register */
+	dev_info->min_mtu = VMXNET3_MIN_MTU;
+	dev_info->max_mtu = VMXNET3_MAX_MTU;
 	dev_info->speed_capa = ETH_LINK_SPEED_10G;
 	dev_info->max_mac_addrs = VMXNET3_MAX_MAC_ADDRS;
 
@@ -1205,6 +1209,18 @@ vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev)
 }
 
 static int
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
+{
+	if (dev->data->dev_started) {
+		PMD_DRV_LOG(ERR, "Port %d must be stopped to configure MTU",
+			    dev->data->port_id);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+static int
 vmxnet3_mac_addr_set(struct rte_eth_dev *dev, struct rte_ether_addr *mac_addr)
 {
 	struct vmxnet3_hw *hw = dev->data->dev_private;
Signature · 1 line
2.7.4
#5 lore raw

drivers/net/vmxnet3/vmxnet3_ethdev.c vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev) · line 1222

Quoted 13 lines · Charles

static int

drivers/net/vmxnet3/vmxnet3_ethdev.c +11 -0
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
+{
+	if (dev->data->dev_started) {
+		PMD_DRV_LOG(ERR, "Port %d must be stopped to configure MTU",
+			    dev->data->port_id);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+

Don't you need to reset the rx ring to change mtu on the fly? At a minimum you need to communicate this value to host through the shared driver page.

#6 lore raw

MTU change is currently only allowed when device is stopped.

Current code in master branch already sets MTU in shared data correctly when starting the device:

static int vmxnet3_setup_driver_shared(struct rte_eth_dev *dev)

{
        struct rte_eth_conf port_conf = dev->data->dev_conf;
        struct vmxnet3_hw *hw = dev->data->dev_private;
        uint32_t mtu = dev->data->mtu;
...
        devRead->misc.mtu = rte_le_to_cpu_32(mtu);

...
}

However current code in master branch does not implement a mtu_set() function so MTU in the dev->data is never changed from default of 1500:

int
rte_eth_dev_set_mtu(uint16_t port_id, uint16_t mtu)

{
        int ret;
        struct rte_eth_dev_info dev_info;
        struct rte_eth_dev *dev;

        RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
        dev = &rte_eth_devices[port_id];
        RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->mtu_set, -ENOTSUP);

        /*
         * Check if the device supports dev_infos_get, if it does not
         * skip min_mtu/max_mtu validation here as this requires values
         * that are populated within the call to rte_eth_dev_info_get()
         * which relies on dev->dev_ops->dev_infos_get.
         */
        if (*dev->dev_ops->dev_infos_get != NULL) {
                rte_eth_dev_info_get(port_id, &dev_info);
                if (mtu < dev_info.min_mtu || mtu > dev_info.max_mtu)
                        return -EINVAL;
        }

        ret = (*dev->dev_ops->mtu_set)(dev, mtu);
        if (!ret)
                dev->data->mtu = mtu;

        return eth_err(port_id, ret);
}

The vmxnet3 driver just needs to implement mtu_set() so that rte_eth_dev_set_mtu() allows changing the MTU and puts the value in dev->data->mtu .

-----Original Message-----
From: Stephen Hemminger <[email protected]>
Sent: Tuesday, August 20, 2019 6:43 PM
To: Myers, Charles <[email protected]>
Cc: Yong Wang <[email protected]>; [email protected]
Subject: Re: [PATCH v3] net/vmxnet3: Added mtu_set() function to allow setting MTU.

drivers/net/vmxnet3/vmxnet3_ethdev.c vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev) · line 1222

Quoted 12 lines · Charles

static int

drivers/net/vmxnet3/vmxnet3_ethdev.c +10 -0
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu) {
+	if (dev->data->dev_started) {
+		PMD_DRV_LOG(ERR, "Port %d must be stopped to configure MTU",
+			    dev->data->port_id);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+

Don't you need to reset the rx ring to change mtu on the fly?
At a minimum you need to communicate this value to host through the shared driver page.

Reviewed-by on v3 1/1 Applied
#7 lore raw
In reply to Myers, Charles
Quoted 11 lines · Charles

From: Charles Myers <[email protected]>

When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500.

This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS.

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

drivers/net/vmxnet3/vmxnet3_ethdev.c vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev) · line 1213

Quoted 6 lines
drivers/net/vmxnet3/vmxnet3_ethdev.c +2 -0
@@ -1205,6 +1209,18 @@ vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev)
 }
 
 static int
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
+{

'mtu' is unused and giving build error:
error: unused parameter ‘mtu’ [-Werror=unused-parameter]

Fixed while merging:
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu __rte_unused)

#8 lore raw
Quoted 8 lines · Ferruh Yigit
Quoted 7 lines · Charles

From: Charles Myers <[email protected]>

When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500.

Quoted 8 lines
Quoted 4 lines

This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS.

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

It seems there is waiting change request from the maintainer that I missed, dropping the patch from the tree and updating the patchwork state.

Charles, can you please sync with Yong if the change request is not clear?

Quoted 16 lines

drivers/net/vmxnet3/vmxnet3_ethdev.c vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev) · line 1213

Quoted 8 lines
drivers/net/vmxnet3/vmxnet3_ethdev.c +2 -0
@@ -1205,6 +1209,18 @@ vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev)
 }
 
 static int
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
+{

'mtu' is unused and giving build error:
error: unused parameter ‘mtu’ [-Werror=unused-parameter]

Fixed while merging:
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu __rte_unused)

#9 lore raw

PATCH v4 1/1 net/vmxnet3: Added mtu_set() function to allow setting MTU.

In reply to Myers, Charles

(Picked up from @Charles Myers patch https://patchwork.dpdk.org/patch/57771/)

When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500.

This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS.

---
 drivers/net/vmxnet3/vmxnet3_ethdev.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)
drivers/net/vmxnet3/vmxnet3_ethdev.c +16 -0
index 705e976..40e81a4 100644
@@ -87,6 +87,7 @@ static int vmxnet3_dev_info_get(struct rte_eth_dev *dev,
 				struct rte_eth_dev_info *dev_info);
 static const uint32_t *
 vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev);
+static int vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu);
 static int vmxnet3_dev_vlan_filter_set(struct rte_eth_dev *dev,
 				       uint16_t vid, int on);
 static int vmxnet3_dev_vlan_offload_set(struct rte_eth_dev *dev, int mask);
@@ -124,6 +125,7 @@ static const struct eth_dev_ops vmxnet3_eth_dev_ops = {
 	.mac_addr_set         = vmxnet3_mac_addr_set,
 	.dev_infos_get        = vmxnet3_dev_info_get,
 	.dev_supported_ptypes_get = vmxnet3_dev_supported_ptypes_get,
+	.mtu_set              = vmxnet3_dev_mtu_set,
 	.vlan_filter_set      = vmxnet3_dev_vlan_filter_set,
 	.vlan_offload_set     = vmxnet3_dev_vlan_offload_set,
 	.rx_queue_setup       = vmxnet3_dev_rx_queue_setup,
@@ -1166,6 +1168,8 @@ vmxnet3_dev_info_get(struct rte_eth_dev *dev,
 	dev_info->max_tx_queues = VMXNET3_MAX_TX_QUEUES;
 	dev_info->min_rx_bufsize = 1518 + RTE_PKTMBUF_HEADROOM;
 	dev_info->max_rx_pktlen = 16384; /* includes CRC, cf MAXFRS register */
+	dev_info->min_mtu = VMXNET3_MIN_MTU;
+	dev_info->max_mtu = VMXNET3_MAX_MTU;
 	dev_info->speed_capa = ETH_LINK_SPEED_10G;
 	dev_info->max_mac_addrs = VMXNET3_MAX_MAC_ADDRS;
 
@@ -1212,6 +1216,18 @@ vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev)
 }
 
 static int
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, __rte_unused uint16_t mtu )
+{
+	if (dev->data->dev_started) {
+		PMD_DRV_LOG(ERR, "Port %d must be stopped to configure MTU",
+			    dev->data->port_id);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+static int
 vmxnet3_mac_addr_set(struct rte_eth_dev *dev, struct rte_ether_addr *mac_addr)
 {
 	struct vmxnet3_hw *hw = dev->data->dev_private;
Signature · 1 line
2.7.4
#10 lore raw

PATCH v5 1/1 net/vmxnet3: Added mtu_set() function to allow setting MTU.

(Picked up from @Charles Myers patch https://patchwork.dpdk.org/patch/57771/)

When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500.

This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS.

---
 drivers/net/vmxnet3/vmxnet3_ethdev.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)
drivers/net/vmxnet3/vmxnet3_ethdev.c +16 -0
index 705e976..c6e11ad 100644
@@ -87,6 +87,7 @@ static int vmxnet3_dev_info_get(struct rte_eth_dev *dev,
 				struct rte_eth_dev_info *dev_info);
 static const uint32_t *
 vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev);
+static int vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu);
 static int vmxnet3_dev_vlan_filter_set(struct rte_eth_dev *dev,
 				       uint16_t vid, int on);
 static int vmxnet3_dev_vlan_offload_set(struct rte_eth_dev *dev, int mask);
@@ -124,6 +125,7 @@ static const struct eth_dev_ops vmxnet3_eth_dev_ops = {
 	.mac_addr_set         = vmxnet3_mac_addr_set,
 	.dev_infos_get        = vmxnet3_dev_info_get,
 	.dev_supported_ptypes_get = vmxnet3_dev_supported_ptypes_get,
+	.mtu_set              = vmxnet3_dev_mtu_set,
 	.vlan_filter_set      = vmxnet3_dev_vlan_filter_set,
 	.vlan_offload_set     = vmxnet3_dev_vlan_offload_set,
 	.rx_queue_setup       = vmxnet3_dev_rx_queue_setup,
@@ -1166,6 +1168,8 @@ vmxnet3_dev_info_get(struct rte_eth_dev *dev,
 	dev_info->max_tx_queues = VMXNET3_MAX_TX_QUEUES;
 	dev_info->min_rx_bufsize = 1518 + RTE_PKTMBUF_HEADROOM;
 	dev_info->max_rx_pktlen = 16384; /* includes CRC, cf MAXFRS register */
+	dev_info->min_mtu = VMXNET3_MIN_MTU;
+	dev_info->max_mtu = VMXNET3_MAX_MTU;
 	dev_info->speed_capa = ETH_LINK_SPEED_10G;
 	dev_info->max_mac_addrs = VMXNET3_MAX_MAC_ADDRS;
 
@@ -1212,6 +1216,18 @@ vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev)
 }
 
 static int
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, __rte_unused uint16_t mtu)
+{
+	if (dev->data->dev_started) {
+		PMD_DRV_LOG(ERR, "Port %d must be stopped to configure MTU",
+			    dev->data->port_id);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+static int
 vmxnet3_mac_addr_set(struct rte_eth_dev *dev, struct rte_ether_addr *mac_addr)
 {
 	struct vmxnet3_hw *hw = dev->data->dev_private;
Signature · 1 line
2.7.4
#11 lore raw

PATCH v6 1/1 net/vmxnet3: Added mtu_set() function to allow setting MTU.

(Picked up from @Charles Myers patch
https://patchwork.dpdk.org/patch/57771/)

When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500.

This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS.

---
 drivers/net/vmxnet3/vmxnet3_ethdev.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)
drivers/net/vmxnet3/vmxnet3_ethdev.c +16 -0
index 705e976..c6e11ad 100644
@@ -87,6 +87,7 @@ static int vmxnet3_dev_info_get(struct rte_eth_dev *dev,
 				struct rte_eth_dev_info *dev_info);
 static const uint32_t *
 vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev);
+static int vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu);
 static int vmxnet3_dev_vlan_filter_set(struct rte_eth_dev *dev,
 				       uint16_t vid, int on);
 static int vmxnet3_dev_vlan_offload_set(struct rte_eth_dev *dev, int mask);
@@ -124,6 +125,7 @@ static const struct eth_dev_ops vmxnet3_eth_dev_ops = {
 	.mac_addr_set         = vmxnet3_mac_addr_set,
 	.dev_infos_get        = vmxnet3_dev_info_get,
 	.dev_supported_ptypes_get = vmxnet3_dev_supported_ptypes_get,
+	.mtu_set              = vmxnet3_dev_mtu_set,
 	.vlan_filter_set      = vmxnet3_dev_vlan_filter_set,
 	.vlan_offload_set     = vmxnet3_dev_vlan_offload_set,
 	.rx_queue_setup       = vmxnet3_dev_rx_queue_setup,
@@ -1166,6 +1168,8 @@ vmxnet3_dev_info_get(struct rte_eth_dev *dev,
 	dev_info->max_tx_queues = VMXNET3_MAX_TX_QUEUES;
 	dev_info->min_rx_bufsize = 1518 + RTE_PKTMBUF_HEADROOM;
 	dev_info->max_rx_pktlen = 16384; /* includes CRC, cf MAXFRS register */
+	dev_info->min_mtu = VMXNET3_MIN_MTU;
+	dev_info->max_mtu = VMXNET3_MAX_MTU;
 	dev_info->speed_capa = ETH_LINK_SPEED_10G;
 	dev_info->max_mac_addrs = VMXNET3_MAX_MAC_ADDRS;
 
@@ -1212,6 +1216,18 @@ vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev)
 }
 
 static int
+vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, __rte_unused uint16_t mtu)
+{
+	if (dev->data->dev_started) {
+		PMD_DRV_LOG(ERR, "Port %d must be stopped to configure MTU",
+			    dev->data->port_id);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+static int
 vmxnet3_mac_addr_set(struct rte_eth_dev *dev, struct rte_ether_addr *mac_addr)
 {
 	struct vmxnet3_hw *hw = dev->data->dev_private;
Signature · 1 line
2.7.4
Acked-by Acked-by on v6 1/1
#12 lore raw

-----Original Message-----
From: Eduard Serra Miralles <[email protected]>
Date: Wednesday, March 25, 2020 at 12:19 PM
To: "[email protected]" <[email protected]>
Cc: Yong Wang <[email protected]>
Subject: [PATCH v6] net/vmxnet3: Added mtu_set() function to allow setting MTU.

(Picked up from @Charles Myers patch
https://patchwork.dpdk.org/patch/57771/)

When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500.

This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS.

---

     drivers/net/vmxnet3/vmxnet3_ethdev.c | 16 ++++++++++++++++
     1 file changed, 16 insertions(+)
    diff --git a/drivers/net/vmxnet3/vmxnet3_ethdev.c b/drivers/net/vmxnet3/vmxnet3_ethdev.c
    index 705e976..c6e11ad 100644
    --- a/drivers/net/vmxnet3/vmxnet3_ethdev.c
    +++ b/drivers/net/vmxnet3/vmxnet3_ethdev.c
    @@ -87,6 +87,7 @@ static int vmxnet3_dev_info_get(struct rte_eth_dev *dev,
     				struct rte_eth_dev_info *dev_info);
     static const uint32_t *
     vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev);
    +static int vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu);
     static int vmxnet3_dev_vlan_filter_set(struct rte_eth_dev *dev,
     				       uint16_t vid, int on);
     static int vmxnet3_dev_vlan_offload_set(struct rte_eth_dev *dev, int mask);
    @@ -124,6 +125,7 @@ static const struct eth_dev_ops vmxnet3_eth_dev_ops = {
     	.mac_addr_set         = vmxnet3_mac_addr_set,
     	.dev_infos_get        = vmxnet3_dev_info_get,
     	.dev_supported_ptypes_get = vmxnet3_dev_supported_ptypes_get,
    +	.mtu_set              = vmxnet3_dev_mtu_set,
     	.vlan_filter_set      = vmxnet3_dev_vlan_filter_set,
     	.vlan_offload_set     = vmxnet3_dev_vlan_offload_set,
     	.rx_queue_setup       = vmxnet3_dev_rx_queue_setup,
    @@ -1166,6 +1168,8 @@ vmxnet3_dev_info_get(struct rte_eth_dev *dev,
     	dev_info->max_tx_queues = VMXNET3_MAX_TX_QUEUES;
     	dev_info->min_rx_bufsize = 1518 + RTE_PKTMBUF_HEADROOM;
     	dev_info->max_rx_pktlen = 16384; /* includes CRC, cf MAXFRS register */
    +	dev_info->min_mtu = VMXNET3_MIN_MTU;
    +	dev_info->max_mtu = VMXNET3_MAX_MTU;
     	dev_info->speed_capa = ETH_LINK_SPEED_10G;
     	dev_info->max_mac_addrs = VMXNET3_MAX_MAC_ADDRS;
     
    @@ -1212,6 +1216,18 @@ vmxnet3_dev_supported_ptypes_get(struct rte_eth_dev *dev)
     }
     
     static int
    +vmxnet3_dev_mtu_set(struct rte_eth_dev *dev, __rte_unused uint16_t mtu)
    +{
    +	if (dev->data->dev_started) {
    +		PMD_DRV_LOG(ERR, "Port %d must be stopped to configure MTU",
    +			    dev->data->port_id);
    +		return -EBUSY;
    +	}
    +
    +	return 0;
    +}
    +
    +static int
     vmxnet3_mac_addr_set(struct rte_eth_dev *dev, struct rte_ether_addr *mac_addr)
     {
     	struct vmxnet3_hw *hw = dev->data->dev_private;
    -- 
    2.7.4
Applied
#13 lore raw
Quoted 10 lines · Yong Wang

-----Original Message-----
From: Eduard Serra Miralles <[email protected]>
Date: Wednesday, March 25, 2020 at 12:19 PM
To: "[email protected]" <[email protected]>
Cc: Yong Wang <[email protected]>
Subject: [PATCH v6] net/vmxnet3: Added mtu_set() function to allow setting MTU.

(Picked up from @Charles Myers patch
https://patchwork.dpdk.org/patch/57771/)

Thanks for resurrecting the patch.

Quoted 6 lines

When the mtu_set() function is not implemented, rte_eth_dev_set_mtu() fails with -ENOTSUP and mtu is not stored in the mtu field in the rte_eth_dev_data. This causes the mtu in Vmxnet3_MiscConf which is shared with hypervisor to always be set to 1500.

Quoted 9 lines

This may cause issues receiving jumbo frames on Enhanced Data Path N-VDS.

---

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