[PATCH net-next 1/8] net: octeontx2: convert to use .get_rx_ring_count

From: Breno Leitao

Date: Fri Jan 09 2026 - 12:42:12 EST


Use the newly introduced .get_rx_ring_count ethtool ops callback instead
of handling ETHTOOL_GRXRINGS directly in .get_rxnfc().

Signed-off-by: Breno Leitao <leitao@xxxxxxxxxx>
---
drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c | 13 +++++++++----
1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c
index b6449f0a9e7d..8918be3ce45e 100644
--- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c
+++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c
@@ -568,6 +568,13 @@ static int otx2_set_coalesce(struct net_device *netdev,
return 0;
}

+static u32 otx2_get_rx_ring_count(struct net_device *dev)
+{
+ struct otx2_nic *pfvf = netdev_priv(dev);
+
+ return pfvf->hw.rx_queues;
+}
+
static int otx2_get_rss_hash_opts(struct net_device *dev,
struct ethtool_rxfh_fields *nfc)
{
@@ -742,10 +749,6 @@ static int otx2_get_rxnfc(struct net_device *dev,
int ret = -EOPNOTSUPP;

switch (nfc->cmd) {
- case ETHTOOL_GRXRINGS:
- nfc->data = pfvf->hw.rx_queues;
- ret = 0;
- break;
case ETHTOOL_GRXCLSRLCNT:
if (netif_running(dev) && ntuple) {
nfc->rule_cnt = pfvf->flow_cfg->nr_flows;
@@ -1344,6 +1347,7 @@ static const struct ethtool_ops otx2_ethtool_ops = {
.set_coalesce = otx2_set_coalesce,
.get_rxnfc = otx2_get_rxnfc,
.set_rxnfc = otx2_set_rxnfc,
+ .get_rx_ring_count = otx2_get_rx_ring_count,
.get_rxfh_key_size = otx2_get_rxfh_key_size,
.get_rxfh_indir_size = otx2_get_rxfh_indir_size,
.get_rxfh = otx2_get_rxfh,
@@ -1462,6 +1466,7 @@ static const struct ethtool_ops otx2vf_ethtool_ops = {
.get_channels = otx2_get_channels,
.get_rxnfc = otx2_get_rxnfc,
.set_rxnfc = otx2_set_rxnfc,
+ .get_rx_ring_count = otx2_get_rx_ring_count,
.get_rxfh_key_size = otx2_get_rxfh_key_size,
.get_rxfh_indir_size = otx2_get_rxfh_indir_size,
.get_rxfh = otx2_get_rxfh,

--
2.47.3