[PATCH 2/2] net: thunderbolt: Update ring indices only after all frames are queued

[Mika Westerberg <mika.westerberg@linux.intel.com>]

Take advantage of the new API and update the ring producer/consumer
indices only after the frames have been queued. This reduces the number
of doorbell register writes from one per frame to one per batch.

Measured using iperf3 over a Thunderbolt/USB4 net connection between
Arrow Lake and Panther Lake systems (average of three runs):

  before         after
  ~14.3 Gbit/s   ~15.3 Gbit/s

Assisted-by: LLM
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
---
 drivers/net/thunderbolt/main.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/net/thunderbolt/main.c b/drivers/net/thunderbolt/main.c
index d9fb587a62c5..cf51b9c39f4e 100644
--- a/drivers/net/thunderbolt/main.c
+++ b/drivers/net/thunderbolt/main.c
@@ -540,11 +540,12 @@ static int tbnet_alloc_rx_buffers(struct tbnet *net, unsigned int nbuffers)
 		trace_tbnet_alloc_rx_frame(index, tf->page, dma_addr,
 					   DMA_FROM_DEVICE);
 
-		tb_ring_rx(ring->ring, &tf->frame);
+		tb_ring_rx_more(ring->ring, &tf->frame);
 
 		ring->prod++;
 	}
 
+	tb_ring_notify(ring->ring);
 	return 0;
 
 err_free:
@@ -1243,7 +1244,8 @@ static netdev_tx_t tbnet_start_xmit(struct sk_buff *skb,
 		goto err_drop;
 
 	for (i = 0; i < frame_index + 1; i++)
-		tb_ring_tx(net->tx_ring.ring, &frames[i]->frame);
+		tb_ring_tx_more(net->tx_ring.ring, &frames[i]->frame);
+	tb_ring_notify(net->tx_ring.ring);
 
 	if (net->svc->prtcstns & TBNET_MATCH_FRAGS_ID)
 		atomic_inc(&net->frame_id);
-- 
2.50.1