diff mbox series

[net-next,v5,3/5] net: ti: icssg-prueth: Add support for HSR frame forward offload

Message ID 20240906111538.1259418-4-danishanwar@ti.com (mailing list archive)
State New, archived
Headers show
Series Introduce HSR offload support for ICSSG | expand

Commit Message

MD Danish Anwar Sept. 6, 2024, 11:15 a.m. UTC
Add support for offloading HSR port-to-port frame forward to hardware.
When the slave interfaces are added to the HSR interface, the PRU cores
will be stopped and ICSSG HSR firmwares will be loaded to them.

Similarly, when HSR interface is deleted, the PRU cores will be
restarted and the last used firmwares will be reloaded. PRUeth
interfaces will be back to the last used mode.

This commit also renames some APIs that are common between switch and
hsr mode with '_fw_offload' suffix.

Signed-off-by: MD Danish Anwar <danishanwar@ti.com>
---
 .../net/ethernet/ti/icssg/icssg_classifier.c  |   1 +
 drivers/net/ethernet/ti/icssg/icssg_config.c  |  18 +--
 drivers/net/ethernet/ti/icssg/icssg_prueth.c  | 132 +++++++++++++++++-
 drivers/net/ethernet/ti/icssg/icssg_prueth.h  |   6 +
 4 files changed, 145 insertions(+), 12 deletions(-)

Comments

Roger Quadros Sept. 10, 2024, 12:12 p.m. UTC | #1
Hi,

On 06/09/2024 14:15, MD Danish Anwar wrote:
> Add support for offloading HSR port-to-port frame forward to hardware.
> When the slave interfaces are added to the HSR interface, the PRU cores
> will be stopped and ICSSG HSR firmwares will be loaded to them.
> 
> Similarly, when HSR interface is deleted, the PRU cores will be
> restarted and the last used firmwares will be reloaded. PRUeth
> interfaces will be back to the last used mode.
> 
> This commit also renames some APIs that are common between switch and
> hsr mode with '_fw_offload' suffix.
> 
> Signed-off-by: MD Danish Anwar <danishanwar@ti.com>
> ---
>  .../net/ethernet/ti/icssg/icssg_classifier.c  |   1 +
>  drivers/net/ethernet/ti/icssg/icssg_config.c  |  18 +--
>  drivers/net/ethernet/ti/icssg/icssg_prueth.c  | 132 +++++++++++++++++-
>  drivers/net/ethernet/ti/icssg/icssg_prueth.h  |   6 +
>  4 files changed, 145 insertions(+), 12 deletions(-)
> 
> diff --git a/drivers/net/ethernet/ti/icssg/icssg_classifier.c b/drivers/net/ethernet/ti/icssg/icssg_classifier.c
> index 9ec504d976d6..833ca86d0b71 100644
> --- a/drivers/net/ethernet/ti/icssg/icssg_classifier.c
> +++ b/drivers/net/ethernet/ti/icssg/icssg_classifier.c

<snip>

> +static void emac_change_hsr_feature(struct net_device *ndev,
> +				    netdev_features_t features,
> +				    u64 hsr_feature)
> +{
> +	netdev_features_t changed = ndev->features ^ features;
> +
> +	if (changed & hsr_feature) {
> +		if (features & hsr_feature)
> +			ndev->features |= hsr_feature;
> +		else
> +			ndev->features &= ~hsr_feature;

You are not supposed to change ndev->features here.

From
"https://www.kernel.org/doc/Documentation/networking/netdev-features.txt"
"
 * ndo_set_features:

Hardware should be reconfigured to match passed feature set. The set
should not be altered unless some error condition happens that can't
be reliably detected in ndo_fix_features. In this case, the callback
should update netdev->features to match resulting hardware state.
Errors returned are not (and cannot be) propagated anywhere except dmesg.
(Note: successful return is zero, >0 means silent error.)"

This means only in 

> +	}
> +}
> +
> +static int emac_ndo_set_features(struct net_device *ndev,
> +				 netdev_features_t features)
> +{
> +	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_FWD);
> +	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_DUP);
> +	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_TAG_INS);
> +	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_TAG_RM);

I don't understand this part. 

As you are not changing hardware state in ndo_set_features, I'm not sure why
you even need ndo_set_features callback.

You didn't take my feedback about using ndo_fix_features().

Please read this.
https://www.kernel.org/doc/Documentation/networking/netdev-features.txt
"Part II: Controlling enabled features"
"Part III: Implementation hints"

Also look at how _netdev_update_features() works and calls ndo_fix_features()
and ndo_set_features()

https://elixir.bootlin.com/linux/v6.11-rc7/source/net/core/dev.c#L10023

> +
> +	return 0;
> +}
> +
>  static const struct net_device_ops emac_netdev_ops = {
>  	.ndo_open = emac_ndo_open,
>  	.ndo_stop = emac_ndo_stop,
> @@ -737,6 +780,7 @@ static const struct net_device_ops emac_netdev_ops = {
>  	.ndo_eth_ioctl = icssg_ndo_ioctl,
>  	.ndo_get_stats64 = icssg_ndo_get_stats64,
>  	.ndo_get_phys_port_name = icssg_ndo_get_phys_port_name,
> +	.ndo_set_features = emac_ndo_set_features,
>  };

<snip>
Anwar, Md Danish Sept. 10, 2024, 1:13 p.m. UTC | #2
Hi Roger,

On 9/10/2024 5:42 PM, Roger Quadros wrote:
> Hi,
> 
> On 06/09/2024 14:15, MD Danish Anwar wrote:
>> Add support for offloading HSR port-to-port frame forward to hardware.
>> When the slave interfaces are added to the HSR interface, the PRU cores
>> will be stopped and ICSSG HSR firmwares will be loaded to them.
>>
>> Similarly, when HSR interface is deleted, the PRU cores will be
>> restarted and the last used firmwares will be reloaded. PRUeth
>> interfaces will be back to the last used mode.
>>
>> This commit also renames some APIs that are common between switch and
>> hsr mode with '_fw_offload' suffix.
>>
>> Signed-off-by: MD Danish Anwar <danishanwar@ti.com>
>> ---
>>  .../net/ethernet/ti/icssg/icssg_classifier.c  |   1 +
>>  drivers/net/ethernet/ti/icssg/icssg_config.c  |  18 +--
>>  drivers/net/ethernet/ti/icssg/icssg_prueth.c  | 132 +++++++++++++++++-
>>  drivers/net/ethernet/ti/icssg/icssg_prueth.h  |   6 +
>>  4 files changed, 145 insertions(+), 12 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/ti/icssg/icssg_classifier.c b/drivers/net/ethernet/ti/icssg/icssg_classifier.c
>> index 9ec504d976d6..833ca86d0b71 100644
>> --- a/drivers/net/ethernet/ti/icssg/icssg_classifier.c
>> +++ b/drivers/net/ethernet/ti/icssg/icssg_classifier.c
> 
> <snip>
> 
>> +static void emac_change_hsr_feature(struct net_device *ndev,
>> +				    netdev_features_t features,
>> +				    u64 hsr_feature)
>> +{
>> +	netdev_features_t changed = ndev->features ^ features;
>> +
>> +	if (changed & hsr_feature) {
>> +		if (features & hsr_feature)
>> +			ndev->features |= hsr_feature;
>> +		else
>> +			ndev->features &= ~hsr_feature;
> 
> You are not supposed to change ndev->features here.
> 
> From
> "https://www.kernel.org/doc/Documentation/networking/netdev-features.txt"
> "
>  * ndo_set_features:
> 
> Hardware should be reconfigured to match passed feature set. The set
> should not be altered unless some error condition happens that can't
> be reliably detected in ndo_fix_features. In this case, the callback
> should update netdev->features to match resulting hardware state.
> Errors returned are not (and cannot be) propagated anywhere except dmesg.
> (Note: successful return is zero, >0 means silent error.)"
> 
> This means only in 
> 
>> +	}
>> +}
>> +
>> +static int emac_ndo_set_features(struct net_device *ndev,
>> +				 netdev_features_t features)
>> +{
>> +	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_FWD);
>> +	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_DUP);
>> +	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_TAG_INS);
>> +	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_TAG_RM);
> 
> I don't understand this part. 
> 
> As you are not changing hardware state in ndo_set_features, I'm not sure why
> you even need ndo_set_features callback.
> 

We don't need to do any hardware configuration when the feature is
changed. Instead certain APIs will do certain actions based on which HSR
related feature is set in ndev->features. I agree this makes the whole
API redundant. Previously in v3
(https://lore.kernel.org/all/20240828091901.3120935-4-danishanwar@ti.com/)
also this API was essentially just changing ndev->features when hsr
change was requested.

+static int emac_ndo_set_features(struct net_device *ndev,
+				 netdev_features_t features)
+{
...

+
+	if (hsr_change_request)
+		ndev->features = features;
+
+	return 0;
+}

I think it would be better to drop this call back in the driver as no
action is needed in the hardware based on what feature is set here.

> You didn't take my feedback about using ndo_fix_features().
> 

Roger, I have taken your feedback regarding ndo_fix_features() and
implemented it in the next patch (patch 4/5
https://lore.kernel.org/all/20240906111538.1259418-5-danishanwar@ti.com/).
Please have a look at that patch also. The features that have
dependencies that can be addressed in ndo_fix_features, are not
introduced in this patch so I have not introduced the function here. It
is getting introduced in the next patch.

> Please read this.
> https://www.kernel.org/doc/Documentation/networking/netdev-features.txt
> "Part II: Controlling enabled features"
> "Part III: Implementation hints"
> 
> Also look at how _netdev_update_features() works and calls ndo_fix_features()
> and ndo_set_features()
> 

Thanks for this. I checked that. Based on how _netdev_update_features()
is implemented, I don't think there is any need of ndo_set_features.
Whatever hardware dependencies are related to these HSR features, can be
taken care by ndo_fix_features.

Please let me know how does this look to you.

> https://elixir.bootlin.com/linux/v6.11-rc7/source/net/core/dev.c#L10023
> 
>> +
>> +	return 0;
>> +}
>> +
>>  static const struct net_device_ops emac_netdev_ops = {
>>  	.ndo_open = emac_ndo_open,
>>  	.ndo_stop = emac_ndo_stop,
>> @@ -737,6 +780,7 @@ static const struct net_device_ops emac_netdev_ops = {
>>  	.ndo_eth_ioctl = icssg_ndo_ioctl,
>>  	.ndo_get_stats64 = icssg_ndo_get_stats64,
>>  	.ndo_get_phys_port_name = icssg_ndo_get_phys_port_name,
>> +	.ndo_set_features = emac_ndo_set_features,
>>  };
> 
> <snip>
>
diff mbox series

Patch

diff --git a/drivers/net/ethernet/ti/icssg/icssg_classifier.c b/drivers/net/ethernet/ti/icssg/icssg_classifier.c
index 9ec504d976d6..833ca86d0b71 100644
--- a/drivers/net/ethernet/ti/icssg/icssg_classifier.c
+++ b/drivers/net/ethernet/ti/icssg/icssg_classifier.c
@@ -290,6 +290,7 @@  void icssg_class_set_host_mac_addr(struct regmap *miig_rt, const u8 *mac)
 		     mac[2] << 16 | mac[3] << 24));
 	regmap_write(miig_rt, MAC_INTERFACE_1, (u32)(mac[4] | mac[5] << 8));
 }
+EXPORT_SYMBOL_GPL(icssg_class_set_host_mac_addr);
 
 void icssg_class_set_mac_addr(struct regmap *miig_rt, int slice, u8 *mac)
 {
diff --git a/drivers/net/ethernet/ti/icssg/icssg_config.c b/drivers/net/ethernet/ti/icssg/icssg_config.c
index dae52a83a378..7b2e6c192ff3 100644
--- a/drivers/net/ethernet/ti/icssg/icssg_config.c
+++ b/drivers/net/ethernet/ti/icssg/icssg_config.c
@@ -107,7 +107,7 @@  static const struct map hwq_map[2][ICSSG_NUM_OTHER_QUEUES] = {
 	},
 };
 
-static void icssg_config_mii_init_switch(struct prueth_emac *emac)
+static void icssg_config_mii_init_fw_offload(struct prueth_emac *emac)
 {
 	struct prueth *prueth = emac->prueth;
 	int mii = prueth_emac_slice(emac);
@@ -278,7 +278,7 @@  static int emac_r30_is_done(struct prueth_emac *emac)
 	return 1;
 }
 
-static int prueth_switch_buffer_setup(struct prueth_emac *emac)
+static int prueth_fw_offload_buffer_setup(struct prueth_emac *emac)
 {
 	struct icssg_buffer_pool_cfg __iomem *bpool_cfg;
 	struct icssg_rxq_ctx __iomem *rxq_ctx;
@@ -424,7 +424,7 @@  static void icssg_init_emac_mode(struct prueth *prueth)
 	icssg_class_set_host_mac_addr(prueth->miig_rt, mac);
 }
 
-static void icssg_init_switch_mode(struct prueth *prueth)
+static void icssg_init_fw_offload_mode(struct prueth *prueth)
 {
 	u32 addr = prueth->shram.pa + EMAC_ICSSG_SWITCH_DEFAULT_VLAN_TABLE_OFFSET;
 	int i;
@@ -455,8 +455,8 @@  int icssg_config(struct prueth *prueth, struct prueth_emac *emac, int slice)
 	struct icssg_flow_cfg __iomem *flow_cfg;
 	int ret;
 
-	if (prueth->is_switch_mode)
-		icssg_init_switch_mode(prueth);
+	if (prueth->is_switch_mode || prueth->is_hsr_offload_mode)
+		icssg_init_fw_offload_mode(prueth);
 	else
 		icssg_init_emac_mode(prueth);
 
@@ -472,8 +472,8 @@  int icssg_config(struct prueth *prueth, struct prueth_emac *emac, int slice)
 	regmap_update_bits(prueth->miig_rt, ICSSG_CFG_OFFSET,
 			   ICSSG_CFG_DEFAULT, ICSSG_CFG_DEFAULT);
 	icssg_miig_set_interface_mode(prueth->miig_rt, slice, emac->phy_if);
-	if (prueth->is_switch_mode)
-		icssg_config_mii_init_switch(emac);
+	if (prueth->is_switch_mode || prueth->is_hsr_offload_mode)
+		icssg_config_mii_init_fw_offload(emac);
 	else
 		icssg_config_mii_init(emac);
 	icssg_config_ipg(emac);
@@ -498,8 +498,8 @@  int icssg_config(struct prueth *prueth, struct prueth_emac *emac, int slice)
 	writeb(0, config + SPL_PKT_DEFAULT_PRIORITY);
 	writeb(0, config + QUEUE_NUM_UNTAGGED);
 
-	if (prueth->is_switch_mode)
-		ret = prueth_switch_buffer_setup(emac);
+	if (prueth->is_switch_mode || prueth->is_hsr_offload_mode)
+		ret = prueth_fw_offload_buffer_setup(emac);
 	else
 		ret = prueth_emac_buffer_setup(emac);
 	if (ret)
diff --git a/drivers/net/ethernet/ti/icssg/icssg_prueth.c b/drivers/net/ethernet/ti/icssg/icssg_prueth.c
index 1dcd68eefdc3..676168d6fded 100644
--- a/drivers/net/ethernet/ti/icssg/icssg_prueth.c
+++ b/drivers/net/ethernet/ti/icssg/icssg_prueth.c
@@ -13,6 +13,7 @@ 
 #include <linux/dma/ti-cppi5.h>
 #include <linux/etherdevice.h>
 #include <linux/genalloc.h>
+#include <linux/if_hsr.h>
 #include <linux/if_vlan.h>
 #include <linux/interrupt.h>
 #include <linux/kernel.h>
@@ -40,6 +41,8 @@ 
 #define DEFAULT_PORT_MASK	1
 #define DEFAULT_UNTAG_MASK	1
 
+#define NETIF_PRUETH_HSR_OFFLOAD_FEATURES	NETIF_F_HW_HSR_FWD
+
 /* CTRLMMR_ICSSG_RGMII_CTRL register bits */
 #define ICSSG_CTRL_RGMII_ID_MODE                BIT(24)
 
@@ -118,6 +121,19 @@  static irqreturn_t prueth_tx_ts_irq(int irq, void *dev_id)
 	return IRQ_HANDLED;
 }
 
+static struct icssg_firmwares icssg_hsr_firmwares[] = {
+	{
+		.pru = "ti-pruss/am65x-sr2-pru0-pruhsr-fw.elf",
+		.rtu = "ti-pruss/am65x-sr2-rtu0-pruhsr-fw.elf",
+		.txpru = "ti-pruss/am65x-sr2-txpru0-pruhsr-fw.elf",
+	},
+	{
+		.pru = "ti-pruss/am65x-sr2-pru1-pruhsr-fw.elf",
+		.rtu = "ti-pruss/am65x-sr2-rtu1-pruhsr-fw.elf",
+		.txpru = "ti-pruss/am65x-sr2-txpru1-pruhsr-fw.elf",
+	}
+};
+
 static struct icssg_firmwares icssg_switch_firmwares[] = {
 	{
 		.pru = "ti-pruss/am65x-sr2-pru0-prusw-fw.elf",
@@ -152,6 +168,8 @@  static int prueth_emac_start(struct prueth *prueth, struct prueth_emac *emac)
 
 	if (prueth->is_switch_mode)
 		firmwares = icssg_switch_firmwares;
+	else if (prueth->is_hsr_offload_mode)
+		firmwares = icssg_hsr_firmwares;
 	else
 		firmwares = icssg_emac_firmwares;
 
@@ -726,6 +744,31 @@  static void emac_ndo_set_rx_mode(struct net_device *ndev)
 	queue_work(emac->cmd_wq, &emac->rx_mode_work);
 }
 
+static void emac_change_hsr_feature(struct net_device *ndev,
+				    netdev_features_t features,
+				    u64 hsr_feature)
+{
+	netdev_features_t changed = ndev->features ^ features;
+
+	if (changed & hsr_feature) {
+		if (features & hsr_feature)
+			ndev->features |= hsr_feature;
+		else
+			ndev->features &= ~hsr_feature;
+	}
+}
+
+static int emac_ndo_set_features(struct net_device *ndev,
+				 netdev_features_t features)
+{
+	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_FWD);
+	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_DUP);
+	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_TAG_INS);
+	emac_change_hsr_feature(ndev, features, NETIF_F_HW_HSR_TAG_RM);
+
+	return 0;
+}
+
 static const struct net_device_ops emac_netdev_ops = {
 	.ndo_open = emac_ndo_open,
 	.ndo_stop = emac_ndo_stop,
@@ -737,6 +780,7 @@  static const struct net_device_ops emac_netdev_ops = {
 	.ndo_eth_ioctl = icssg_ndo_ioctl,
 	.ndo_get_stats64 = icssg_ndo_get_stats64,
 	.ndo_get_phys_port_name = icssg_ndo_get_phys_port_name,
+	.ndo_set_features = emac_ndo_set_features,
 };
 
 static int prueth_netdev_init(struct prueth *prueth,
@@ -865,6 +909,7 @@  static int prueth_netdev_init(struct prueth *prueth,
 	ndev->ethtool_ops = &icssg_ethtool_ops;
 	ndev->hw_features = NETIF_F_SG;
 	ndev->features = ndev->hw_features;
+	ndev->hw_features |= NETIF_PRUETH_HSR_OFFLOAD_FEATURES;
 
 	netif_napi_add(ndev, &emac->napi_rx, icssg_napi_rx_poll);
 	hrtimer_init(&emac->rx_hrtimer, CLOCK_MONOTONIC,
@@ -953,7 +998,7 @@  static void prueth_emac_restart(struct prueth *prueth)
 	netif_device_attach(emac1->ndev);
 }
 
-static void icssg_enable_switch_mode(struct prueth *prueth)
+static void icssg_change_mode(struct prueth *prueth)
 {
 	struct prueth_emac *emac;
 	int mac;
@@ -973,8 +1018,13 @@  static void icssg_enable_switch_mode(struct prueth *prueth)
 					  BIT(emac->port_id) | DEFAULT_PORT_MASK,
 					  BIT(emac->port_id) | DEFAULT_UNTAG_MASK,
 					  true);
+			if (prueth->is_hsr_offload_mode)
+				icssg_vtbl_modify(emac, DEFAULT_VID,
+						  DEFAULT_PORT_MASK,
+						  DEFAULT_UNTAG_MASK, true);
 			icssg_set_pvid(prueth, emac->port_vlan, emac->port_id);
-			icssg_set_port_state(emac, ICSSG_EMAC_PORT_VLAN_AWARE_ENABLE);
+			if (prueth->is_switch_mode)
+				icssg_set_port_state(emac, ICSSG_EMAC_PORT_VLAN_AWARE_ENABLE);
 		}
 	}
 }
@@ -1012,7 +1062,7 @@  static int prueth_netdevice_port_link(struct net_device *ndev,
 			prueth->is_switch_mode = true;
 			prueth->default_vlan = 1;
 			emac->port_vlan = prueth->default_vlan;
-			icssg_enable_switch_mode(prueth);
+			icssg_change_mode(prueth);
 		}
 	}
 
@@ -1040,6 +1090,61 @@  static void prueth_netdevice_port_unlink(struct net_device *ndev)
 		prueth->hw_bridge_dev = NULL;
 }
 
+static int prueth_hsr_port_link(struct net_device *ndev)
+{
+	struct prueth_emac *emac = netdev_priv(ndev);
+	struct prueth *prueth = emac->prueth;
+	struct prueth_emac *emac0;
+	struct prueth_emac *emac1;
+
+	emac0 = prueth->emac[PRUETH_MAC0];
+	emac1 = prueth->emac[PRUETH_MAC1];
+
+	if (prueth->is_switch_mode)
+		return -EOPNOTSUPP;
+
+	prueth->hsr_members |= BIT(emac->port_id);
+	if (!prueth->is_hsr_offload_mode) {
+		if (prueth->hsr_members & BIT(PRUETH_PORT_MII0) &&
+		    prueth->hsr_members & BIT(PRUETH_PORT_MII1)) {
+			if (!(emac0->ndev->features &
+			      NETIF_PRUETH_HSR_OFFLOAD_FEATURES) &&
+			    !(emac1->ndev->features &
+			      NETIF_PRUETH_HSR_OFFLOAD_FEATURES))
+				return -EOPNOTSUPP;
+			prueth->is_hsr_offload_mode = true;
+			prueth->default_vlan = 1;
+			emac0->port_vlan = prueth->default_vlan;
+			emac1->port_vlan = prueth->default_vlan;
+			icssg_change_mode(prueth);
+			netdev_dbg(ndev, "Enabling HSR offload mode\n");
+		}
+	}
+
+	return 0;
+}
+
+static void prueth_hsr_port_unlink(struct net_device *ndev)
+{
+	struct prueth_emac *emac = netdev_priv(ndev);
+	struct prueth *prueth = emac->prueth;
+	struct prueth_emac *emac0;
+	struct prueth_emac *emac1;
+
+	emac0 = prueth->emac[PRUETH_MAC0];
+	emac1 = prueth->emac[PRUETH_MAC1];
+
+	prueth->hsr_members &= ~BIT(emac->port_id);
+	if (prueth->is_hsr_offload_mode) {
+		prueth->is_hsr_offload_mode = false;
+		emac0->port_vlan = 0;
+		emac1->port_vlan = 0;
+		prueth->hsr_dev = NULL;
+		prueth_emac_restart(prueth);
+		netdev_dbg(ndev, "Disabling HSR Offload mode\n");
+	}
+}
+
 /* netdev notifier */
 static int prueth_netdevice_event(struct notifier_block *unused,
 				  unsigned long event, void *ptr)
@@ -1047,6 +1152,8 @@  static int prueth_netdevice_event(struct notifier_block *unused,
 	struct netlink_ext_ack *extack = netdev_notifier_info_to_extack(ptr);
 	struct net_device *ndev = netdev_notifier_info_to_dev(ptr);
 	struct netdev_notifier_changeupper_info *info;
+	struct prueth_emac *emac = netdev_priv(ndev);
+	struct prueth *prueth = emac->prueth;
 	int ret = NOTIFY_DONE;
 
 	if (ndev->netdev_ops != &emac_netdev_ops)
@@ -1056,6 +1163,25 @@  static int prueth_netdevice_event(struct notifier_block *unused,
 	case NETDEV_CHANGEUPPER:
 		info = ptr;
 
+		if ((ndev->features & NETIF_PRUETH_HSR_OFFLOAD_FEATURES) &&
+		    is_hsr_master(info->upper_dev)) {
+			if (info->linking) {
+				if (!prueth->hsr_dev) {
+					prueth->hsr_dev = info->upper_dev;
+					icssg_class_set_host_mac_addr(prueth->miig_rt,
+								      prueth->hsr_dev->dev_addr);
+				} else {
+					if (prueth->hsr_dev != info->upper_dev) {
+						netdev_dbg(ndev, "Both interfaces must be linked to same upper device\n");
+						return -EOPNOTSUPP;
+					}
+				}
+				prueth_hsr_port_link(ndev);
+			} else {
+				prueth_hsr_port_unlink(ndev);
+			}
+		}
+
 		if (netif_is_bridge_master(info->upper_dev)) {
 			if (info->linking)
 				ret = prueth_netdevice_port_link(ndev, info->upper_dev, extack);
diff --git a/drivers/net/ethernet/ti/icssg/icssg_prueth.h b/drivers/net/ethernet/ti/icssg/icssg_prueth.h
index 786bd1ba34ab..a4b025fae797 100644
--- a/drivers/net/ethernet/ti/icssg/icssg_prueth.h
+++ b/drivers/net/ethernet/ti/icssg/icssg_prueth.h
@@ -243,11 +243,14 @@  struct icssg_firmwares {
  * @iep1: pointer to IEP1 device
  * @vlan_tbl: VLAN-FID table pointer
  * @hw_bridge_dev: pointer to HW bridge net device
+ * @hsr_dev: pointer to the HSR net device
  * @br_members: bitmask of bridge member ports
+ * @hsr_members: bitmask of hsr member ports
  * @prueth_netdevice_nb: netdevice notifier block
  * @prueth_switchdev_nb: switchdev notifier block
  * @prueth_switchdev_bl_nb: switchdev blocking notifier block
  * @is_switch_mode: flag to indicate if device is in Switch mode
+ * @is_hsr_offload_mode: flag to indicate if device is in hsr offload mode
  * @is_switchmode_supported: indicates platform support for switch mode
  * @switch_id: ID for mapping switch ports to bridge
  * @default_vlan: Default VLAN for host
@@ -279,11 +282,14 @@  struct prueth {
 	struct prueth_vlan_tbl *vlan_tbl;
 
 	struct net_device *hw_bridge_dev;
+	struct net_device *hsr_dev;
 	u8 br_members;
+	u8 hsr_members;
 	struct notifier_block prueth_netdevice_nb;
 	struct notifier_block prueth_switchdev_nb;
 	struct notifier_block prueth_switchdev_bl_nb;
 	bool is_switch_mode;
+	bool is_hsr_offload_mode;
 	bool is_switchmode_supported;
 	unsigned char switch_id[MAX_PHYS_ITEM_ID_LEN];
 	int default_vlan;