Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Esben Haabendal <esben@geanix.com>
To:     netdev@vger.kernel.org
Cc:     linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org,
        Andrew Lunn <andrew@lunn.ch>,
        "David S . Miller" <davem@davemloft.net>,
        Michal Simek <michal.simek@xilinx.com>,
        =?UTF-8?q?Petr=20=C5=A0tetiar?= <ynezz@true.cz>
Subject: [PATCH 7/8] net: ll_temac: Make RX/TX ring sizes configurable
Date:   Tue, 18 Feb 2020 09:27:29 +0100
Message-Id: <20200218082729.7626-1-esben@geanix.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Add support for setting the RX and TX ring sizes for this driver using
ethtool. Also increase the default RX ring size as the previous default
was far too low for good performance in some configurations.

Signed-off-by: Esben Haabendal <esben@geanix.com>
---
 drivers/net/ethernet/xilinx/ll_temac.h      |  2 +
 drivers/net/ethernet/xilinx/ll_temac_main.c | 96 +++++++++++++++------
 2 files changed, 72 insertions(+), 26 deletions(-)

diff --git a/drivers/net/ethernet/xilinx/ll_temac.h b/drivers/net/ethernet/xilinx/ll_temac.h
index 463ef9eaf42d..8777ec6e21c8 100644
--- a/drivers/net/ethernet/xilinx/ll_temac.h
+++ b/drivers/net/ethernet/xilinx/ll_temac.h
@@ -369,8 +369,10 @@ struct temac_local {
 	/* Buffer descriptors */
 	struct cdmac_bd *tx_bd_v;
 	dma_addr_t tx_bd_p;
+	u32 tx_bd_num;
 	struct cdmac_bd *rx_bd_v;
 	dma_addr_t rx_bd_p;
+	u32 rx_bd_num;
 	int tx_bd_ci;
 	int tx_bd_tail;
 	int rx_bd_ci;
diff --git a/drivers/net/ethernet/xilinx/ll_temac_main.c b/drivers/net/ethernet/xilinx/ll_temac_main.c
index baf05a2b7551..e3d4857334f3 100644
--- a/drivers/net/ethernet/xilinx/ll_temac_main.c
+++ b/drivers/net/ethernet/xilinx/ll_temac_main.c
@@ -58,8 +58,11 @@
 
 #include "ll_temac.h"
 
-#define TX_BD_NUM   64
-#define RX_BD_NUM   128
+/* Descriptors defines for Tx and Rx DMA */
+#define TX_BD_NUM_DEFAULT		64
+#define RX_BD_NUM_DEFAULT		1024
+#define TX_BD_NUM_MAX			4096
+#define RX_BD_NUM_MAX			4096
 
 /* ---------------------------------------------------------------------
  * Low level register access functions
@@ -301,7 +304,7 @@ static void temac_dma_bd_release(struct net_device *ndev)
 	/* Reset Local Link (DMA) */
 	lp->dma_out(lp, DMA_CONTROL_REG, DMA_CONTROL_RST);
 
-	for (i = 0; i < RX_BD_NUM; i++) {
+	for (i = 0; i < lp->rx_bd_num; i++) {
 		if (!lp->rx_skb[i])
 			break;
 		else {
@@ -312,12 +315,12 @@ static void temac_dma_bd_release(struct net_device *ndev)
 	}
 	if (lp->rx_bd_v)
 		dma_free_coherent(ndev->dev.parent,
-				sizeof(*lp->rx_bd_v) * RX_BD_NUM,
-				lp->rx_bd_v, lp->rx_bd_p);
+				  sizeof(*lp->rx_bd_v) * lp->rx_bd_num,
+				  lp->rx_bd_v, lp->rx_bd_p);
 	if (lp->tx_bd_v)
 		dma_free_coherent(ndev->dev.parent,
-				sizeof(*lp->tx_bd_v) * TX_BD_NUM,
-				lp->tx_bd_v, lp->tx_bd_p);
+				  sizeof(*lp->tx_bd_v) * lp->tx_bd_num,
+				  lp->tx_bd_v, lp->tx_bd_p);
 }
 
 /**
@@ -330,33 +333,33 @@ static int temac_dma_bd_init(struct net_device *ndev)
 	dma_addr_t skb_dma_addr;
 	int i;
 
-	lp->rx_skb = devm_kcalloc(&ndev->dev, RX_BD_NUM, sizeof(*lp->rx_skb),
-				  GFP_KERNEL);
+	lp->rx_skb = devm_kcalloc(&ndev->dev, lp->rx_bd_num,
+				  sizeof(*lp->rx_skb), GFP_KERNEL);
 	if (!lp->rx_skb)
 		goto out;
 
 	/* allocate the tx and rx ring buffer descriptors. */
 	/* returns a virtual address and a physical address. */
 	lp->tx_bd_v = dma_alloc_coherent(ndev->dev.parent,
-					 sizeof(*lp->tx_bd_v) * TX_BD_NUM,
+					 sizeof(*lp->tx_bd_v) * lp->tx_bd_num,
 					 &lp->tx_bd_p, GFP_KERNEL);
 	if (!lp->tx_bd_v)
 		goto out;
 
 	lp->rx_bd_v = dma_alloc_coherent(ndev->dev.parent,
-					 sizeof(*lp->rx_bd_v) * RX_BD_NUM,
+					 sizeof(*lp->rx_bd_v) * lp->rx_bd_num,
 					 &lp->rx_bd_p, GFP_KERNEL);
 	if (!lp->rx_bd_v)
 		goto out;
 
-	for (i = 0; i < TX_BD_NUM; i++) {
+	for (i = 0; i < lp->tx_bd_num; i++) {
 		lp->tx_bd_v[i].next = cpu_to_be32(lp->tx_bd_p
-				+ sizeof(*lp->tx_bd_v) * ((i + 1) % TX_BD_NUM));
+			+ sizeof(*lp->tx_bd_v) * ((i + 1) % lp->tx_bd_num));
 	}
 
-	for (i = 0; i < RX_BD_NUM; i++) {
+	for (i = 0; i < lp->rx_bd_num; i++) {
 		lp->rx_bd_v[i].next = cpu_to_be32(lp->rx_bd_p
-				+ sizeof(*lp->rx_bd_v) * ((i + 1) % RX_BD_NUM));
+			+ sizeof(*lp->rx_bd_v) * ((i + 1) % lp->rx_bd_num));
 
 		skb = netdev_alloc_skb_ip_align(ndev,
 						XTE_MAX_JUMBO_FRAME_SIZE);
@@ -389,7 +392,7 @@ static int temac_dma_bd_init(struct net_device *ndev)
 	lp->tx_bd_ci = 0;
 	lp->tx_bd_tail = 0;
 	lp->rx_bd_ci = 0;
-	lp->rx_bd_tail = RX_BD_NUM - 1;
+	lp->rx_bd_tail = lp->rx_bd_num - 1;
 
 	/* Enable RX DMA transfers */
 	wmb();
@@ -784,7 +787,7 @@ static void temac_start_xmit_done(struct net_device *ndev)
 		ndev->stats.tx_bytes += be32_to_cpu(cur_p->len);
 
 		lp->tx_bd_ci++;
-		if (lp->tx_bd_ci >= TX_BD_NUM)
+		if (lp->tx_bd_ci >= lp->tx_bd_num)
 			lp->tx_bd_ci = 0;
 
 		cur_p = &lp->tx_bd_v[lp->tx_bd_ci];
@@ -810,7 +813,7 @@ static inline int temac_check_tx_bd_space(struct temac_local *lp, int num_frag)
 			return NETDEV_TX_BUSY;
 
 		tail++;
-		if (tail >= TX_BD_NUM)
+		if (tail >= lp->tx_bd_num)
 			tail = 0;
 
 		cur_p = &lp->tx_bd_v[tail];
@@ -871,7 +874,7 @@ temac_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 	ptr_to_txbd((void *)skb, cur_p);
 
 	for (ii = 0; ii < num_frag; ii++) {
-		if (++lp->tx_bd_tail >= TX_BD_NUM)
+		if (++lp->tx_bd_tail >= lp->tx_bd_num)
 			lp->tx_bd_tail = 0;
 
 		cur_p = &lp->tx_bd_v[lp->tx_bd_tail];
@@ -881,7 +884,7 @@ temac_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 					      DMA_TO_DEVICE);
 		if (dma_mapping_error(ndev->dev.parent, skb_dma_addr)) {
 			if (--lp->tx_bd_tail < 0)
-				lp->tx_bd_tail = TX_BD_NUM - 1;
+				lp->tx_bd_tail = lp->tx_bd_num - 1;
 			cur_p = &lp->tx_bd_v[lp->tx_bd_tail];
 			while (--ii >= 0) {
 				--frag;
@@ -890,7 +893,7 @@ temac_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 						 skb_frag_size(frag),
 						 DMA_TO_DEVICE);
 				if (--lp->tx_bd_tail < 0)
-					lp->tx_bd_tail = TX_BD_NUM - 1;
+					lp->tx_bd_tail = lp->tx_bd_num - 1;
 				cur_p = &lp->tx_bd_v[lp->tx_bd_tail];
 			}
 			dma_unmap_single(ndev->dev.parent,
@@ -907,7 +910,7 @@ temac_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 
 	tail_p = lp->tx_bd_p + sizeof(*lp->tx_bd_v) * lp->tx_bd_tail;
 	lp->tx_bd_tail++;
-	if (lp->tx_bd_tail >= TX_BD_NUM)
+	if (lp->tx_bd_tail >= lp->tx_bd_num)
 		lp->tx_bd_tail = 0;
 
 	skb_tx_timestamp(skb);
@@ -927,7 +930,7 @@ static int ll_temac_recv_buffers_available(struct temac_local *lp)
 		return 0;
 	available = 1 + lp->rx_bd_tail - lp->rx_bd_ci;
 	if (available <= 0)
-		available += RX_BD_NUM;
+		available += lp->rx_bd_num;
 	return available;
 }
 
@@ -996,7 +999,7 @@ static void ll_temac_recv(struct net_device *ndev)
 		ndev->stats.rx_bytes += length;
 
 		rx_bd = lp->rx_bd_ci;
-		if (++lp->rx_bd_ci >= RX_BD_NUM)
+		if (++lp->rx_bd_ci >= lp->rx_bd_num)
 			lp->rx_bd_ci = 0;
 	} while (rx_bd != lp->rx_bd_tail);
 
@@ -1027,7 +1030,7 @@ static void ll_temac_recv(struct net_device *ndev)
 		dma_addr_t skb_dma_addr;
 
 		rx_bd = lp->rx_bd_tail + 1;
-		if (rx_bd >= RX_BD_NUM)
+		if (rx_bd >= lp->rx_bd_num)
 			rx_bd = 0;
 		bd = &lp->rx_bd_v[rx_bd];
 
@@ -1243,13 +1246,52 @@ static const struct attribute_group temac_attr_group = {
 	.attrs = temac_device_attrs,
 };
 
-/* ethtool support */
+/* ---------------------------------------------------------------------
+ * ethtool support
+ */
+
+static void ll_temac_ethtools_get_ringparam(struct net_device *ndev,
+					    struct ethtool_ringparam *ering)
+{
+	struct temac_local *lp = netdev_priv(ndev);
+
+	ering->rx_max_pending = RX_BD_NUM_MAX;
+	ering->rx_mini_max_pending = 0;
+	ering->rx_jumbo_max_pending = 0;
+	ering->tx_max_pending = TX_BD_NUM_MAX;
+	ering->rx_pending = lp->rx_bd_num;
+	ering->rx_mini_pending = 0;
+	ering->rx_jumbo_pending = 0;
+	ering->tx_pending = lp->tx_bd_num;
+}
+
+static int ll_temac_ethtools_set_ringparam(struct net_device *ndev,
+					   struct ethtool_ringparam *ering)
+{
+	struct temac_local *lp = netdev_priv(ndev);
+
+	if (ering->rx_pending > RX_BD_NUM_MAX ||
+	    ering->rx_mini_pending ||
+	    ering->rx_jumbo_pending ||
+	    ering->rx_pending > TX_BD_NUM_MAX)
+		return -EINVAL;
+
+	if (netif_running(ndev))
+		return -EBUSY;
+
+	lp->rx_bd_num = ering->rx_pending;
+	lp->tx_bd_num = ering->tx_pending;
+	return 0;
+}
+
 static const struct ethtool_ops temac_ethtool_ops = {
 	.nway_reset = phy_ethtool_nway_reset,
 	.get_link = ethtool_op_get_link,
 	.get_ts_info = ethtool_op_get_ts_info,
 	.get_link_ksettings = phy_ethtool_get_link_ksettings,
 	.set_link_ksettings = phy_ethtool_set_link_ksettings,
+	.get_ringparam	= ll_temac_ethtools_get_ringparam,
+	.set_ringparam	= ll_temac_ethtools_set_ringparam,
 };
 
 static int temac_probe(struct platform_device *pdev)
@@ -1293,6 +1335,8 @@ static int temac_probe(struct platform_device *pdev)
 	lp->ndev = ndev;
 	lp->dev = &pdev->dev;
 	lp->options = XTE_OPTION_DEFAULTS;
+	lp->rx_bd_num = RX_BD_NUM_DEFAULT;
+	lp->tx_bd_num = TX_BD_NUM_DEFAULT;
 	spin_lock_init(&lp->rx_lock);
 	INIT_DELAYED_WORK(&lp->restart_work, ll_temac_restart_work_func);
 
-- 
2.25.0