From: Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
To: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>,
barebox@lists.infradead.org
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Subject: Re: [PATCH 4/5] net: Add driver for Armada 370/XP 10/100/1000 Mbps network controller
Date: Tue, 26 Aug 2014 16:30:24 +0200 [thread overview]
Message-ID: <53FC9A00.9090608@gmail.com> (raw)
In-Reply-To: <1408834420-899-5-git-send-email-ezequiel.garcia@free-electrons.com>
On 08/24/2014 12:53 AM, Ezequiel Garcia wrote:
> This patch introduces the mvneta driver to support the network controller
> found in Armada 370/XP SoCs.
>
> Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
> ---
> drivers/net/Kconfig | 6 +
> drivers/net/Makefile | 1 +
> drivers/net/mvneta.c | 758 +++++++++++++++++++++++++++++++++++++++++++++++++++
> 3 files changed, 765 insertions(+)
> create mode 100644 drivers/net/mvneta.c
>
> diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
> index c99fcc8..f6584d4 100644
> --- a/drivers/net/Kconfig
> +++ b/drivers/net/Kconfig
> @@ -137,6 +137,12 @@ config DRIVER_NET_MPC5200
> depends on ARCH_MPC5200
> select PHYLIB
>
> +config DRIVER_NET_MVNETA
> + bool "Marvell NETA"
> + depends on ARCH_MVEBU
nit: depends on ARCH_ARMADA_370 || ARCH_ARMADA_XP ?
I'll also give it a try later.
Sebastian
> + select PHYLIB
> + select MDIO_MVEBU
> +
> config DRIVER_NET_NETX
> bool "Hilscher Netx ethernet driver"
> depends on HAS_NETX_ETHER
> diff --git a/drivers/net/Makefile b/drivers/net/Makefile
> index 1b85778..b124220 100644
> --- a/drivers/net/Makefile
> +++ b/drivers/net/Makefile
> @@ -19,6 +19,7 @@ obj-$(CONFIG_DRIVER_NET_KS8851_MLL) += ks8851_mll.o
> obj-$(CONFIG_DRIVER_NET_MACB) += macb.o
> obj-$(CONFIG_DRIVER_NET_MICREL) += ksz8864rmn.o
> obj-$(CONFIG_DRIVER_NET_MPC5200) += fec_mpc5200.o
> +obj-$(CONFIG_DRIVER_NET_MVNETA) += mvneta.o
> obj-$(CONFIG_DRIVER_NET_NETX) += netx_eth.o
> obj-$(CONFIG_DRIVER_NET_ORION) += orion-gbe.o
> obj-$(CONFIG_DRIVER_NET_RTL8139) += rtl8139.o
> diff --git a/drivers/net/mvneta.c b/drivers/net/mvneta.c
> new file mode 100644
> index 0000000..8768b0d
> --- /dev/null
> +++ b/drivers/net/mvneta.c
> @@ -0,0 +1,758 @@
> +/*
> + * (C) Copyright 2014 - Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
> + *
> + * based on mvneta driver from linux
> + * (C) Copyright 2012 Marvell
> + * Rami Rosen <rosenr@marvell.com>
> + * Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> + *
> + * based on orion-gbe driver from barebox
> + * (C) Copyright 2014
> + * Pengutronix, Michael Grzeschik <mgr@pengutronix.de>
> + * Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
> +
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License as
> + * published by the Free Software Foundation; either version 2 of
> + * the License, or (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + * GNU General Public License for more details.
> + *
> + */
> +
> +#include <common.h>
> +#include <init.h>
> +#include <io.h>
> +#include <net.h>
> +#include <of_net.h>
> +#include <sizes.h>
> +#include <asm/mmu.h>
> +#include <linux/clk.h>
> +#include <linux/err.h>
> +#include <linux/mbus.h>
> +
> +/* Registers */
> +
> +/* Rx queue */
> +#define MVNETA_RXQ_CONFIG_REG(q) (0x1400 + ((q) << 2))
> +#define MVNETA_RXQ_PKT_OFFSET_ALL_MASK (0xf << 8)
> +#define MVNETA_RXQ_PKT_OFFSET_MASK(offs) ((offs) << 8)
> +#define MVNETA_RXQ_THRESHOLD_REG(q) (0x14c0 + ((q) << 2))
> +#define MVNETA_RXQ_NON_OCCUPIED(v) ((v) << 16)
> +#define MVNETA_RXQ_BASE_ADDR_REG(q) (0x1480 + ((q) << 2))
> +#define MVNETA_RXQ_SIZE_REG(q) (0x14a0 + ((q) << 2))
> +#define MVNETA_RXQ_BUF_SIZE_SHIFT 19
> +#define MVNETA_RXQ_BUF_SIZE_MASK (0x1fff << 19)
> +#define MVNETA_RXQ_STATUS_REG(q) (0x14e0 + ((q) << 2))
> +#define MVNETA_RXQ_OCCUPIED_ALL_MASK 0x3fff
> +#define MVNETA_RXQ_STATUS_UPDATE_REG(q) (0x1500 + ((q) << 2))
> +#define MVNETA_RXQ_ADD_NON_OCCUPIED_SHIFT 16
> +#define MVNETA_RXQ_ADD_NON_OCCUPIED_MAX 255
> +
> +#define MVNETA_PORT_RX_RESET 0x1cc0
> +#define MVNETA_MBUS_RETRY 0x2010
> +#define MVNETA_UNIT_INTR_CAUSE 0x2080
> +#define MVNETA_UNIT_CONTROL 0x20B0
> +#define MVNETA_WIN_BASE(w) (0x2200 + ((w) << 3))
> +#define MVNETA_WIN_SIZE(w) (0x2204 + ((w) << 3))
> +#define MVNETA_WIN_REMAP(w) (0x2280 + ((w) << 2))
> +#define MVNETA_BASE_ADDR_ENABLE 0x2290
> +#define MVNETA_PORT_CONFIG 0x2400
> +#define MVNETA_DEF_RXQ(q) ((q) << 1)
> +#define MVNETA_DEF_RXQ_ARP(q) ((q) << 4)
> +#define MVNETA_TX_UNSET_ERR_SUM BIT(12)
> +#define MVNETA_DEF_RXQ_TCP(q) ((q) << 16)
> +#define MVNETA_DEF_RXQ_UDP(q) ((q) << 19)
> +#define MVNETA_DEF_RXQ_BPDU(q) ((q) << 22)
> +#define MVNETA_RX_CSUM_WITH_PSEUDO_HDR BIT(25)
> +#define MVNETA_PORT_CONFIG_DEFL_VALUE(q) (MVNETA_DEF_RXQ(q) | \
> + MVNETA_DEF_RXQ_ARP(q) | \
> + MVNETA_DEF_RXQ_TCP(q) | \
> + MVNETA_DEF_RXQ_UDP(q) | \
> + MVNETA_DEF_RXQ_BPDU(q) | \
> + MVNETA_TX_UNSET_ERR_SUM | \
> + MVNETA_RX_CSUM_WITH_PSEUDO_HDR)
> +#define MVNETA_PORT_CONFIG_EXTEND 0x2404
> +#define MVNETA_MAC_ADDR_LOW 0x2414
> +#define MVNETA_MAC_ADDR_HIGH 0x2418
> +#define MVNETA_SDMA_CONFIG 0x241c
> +#define MVNETA_SDMA_BRST_SIZE_16 4
> +#define MVNETA_RX_BRST_SZ_MASK(burst) ((burst) << 1)
> +#define MVNETA_RX_NO_DATA_SWAP BIT(4)
> +#define MVNETA_TX_NO_DATA_SWAP BIT(5)
> +#define MVNETA_DESC_SWAP BIT(6)
> +#define MVNETA_TX_BRST_SZ_MASK(burst) ((burst) << 22)
> +#define MVNETA_RX_MIN_FRAME_SIZE 0x247c
> +#define MVNETA_SERDES_CFG 0x24a0
> +#define MVNETA_SGMII_SERDES 0x0cc7
> +#define MVNETA_QSGMII_SERDES 0x0667
> +#define MVNETA_TYPE_PRIO 0x24bc
> +#define MVNETA_TXQ_CMD_1 0x24e4
> +#define MVNETA_TXQ_CMD 0x2448
> +#define MVNETA_TXQ_DISABLE_SHIFT 8
> +#define MVNETA_TXQ_ENABLE_MASK 0x000000ff
> +#define MVNETA_ACC_MODE 0x2500
> +#define MVNETA_CPU_MAP(cpu) (0x2540 + ((cpu) << 2))
> +
> +#define MVNETA_INTR_NEW_CAUSE 0x25a0
> +#define MVNETA_INTR_NEW_MASK 0x25a4
> +#define MVNETA_INTR_OLD_CAUSE 0x25a8
> +#define MVNETA_INTR_OLD_MASK 0x25ac
> +#define MVNETA_INTR_MISC_CAUSE 0x25b0
> +#define MVNETA_INTR_MISC_MASK 0x25b4
> +#define MVNETA_INTR_ENABLE 0x25b8
> +
> +#define MVNETA_RXQ_CMD 0x2680
> +#define MVNETA_RXQ_DISABLE_SHIFT 8
> +#define MVNETA_RXQ_ENABLE_MASK 0x000000ff
> +#define MVETH_TXQ_TOKEN_COUNT_REG(q) (0x2700 + ((q) << 4))
> +#define MVETH_TXQ_TOKEN_CFG_REG(q) (0x2704 + ((q) << 4))
> +#define MVNETA_GMAC_CTRL_0 0x2c00
> +#define MVNETA_GMAC_CTRL_2 0x2c08
> +#define MVNETA_GMAC2_PCS_ENABLE BIT(3)
> +#define MVNETA_GMAC2_PORT_RGMII BIT(4)
> +#define MVNETA_GMAC2_PORT_RESET BIT(6)
> +#define MVNETA_GMAC_AUTONEG_CONFIG 0x2c0c
> +#define MVNETA_GMAC_FORCE_LINK_DOWN BIT(0)
> +#define MVNETA_GMAC_FORCE_LINK_PASS BIT(1)
> +#define MVNETA_GMAC_CONFIG_MII_SPEED BIT(5)
> +#define MVNETA_GMAC_CONFIG_GMII_SPEED BIT(6)
> +#define MVNETA_GMAC_AN_SPEED_EN BIT(7)
> +#define MVNETA_GMAC_CONFIG_FULL_DUPLEX BIT(12)
> +#define MVNETA_GMAC_AN_DUPLEX_EN BIT(13)
> +#define MVNETA_MIB_COUNTERS_BASE 0x3080
> +#define MVNETA_MIB_LATE_COLLISION 0x7c
> +#define MVNETA_DA_FILT_SPEC_MCAST 0x3400
> +#define MVNETA_DA_FILT_OTH_MCAST 0x3500
> +#define MVNETA_DA_FILT_UCAST_BASE 0x3600
> +#define MVNETA_TXQ_BASE_ADDR_REG(q) (0x3c00 + ((q) << 2))
> +#define MVNETA_TXQ_SIZE_REG(q) (0x3c20 + ((q) << 2))
> +#define MVNETA_TXQ_UPDATE_REG(q) (0x3c60 + ((q) << 2))
> +#define MVNETA_TXQ_DEC_SENT_SHIFT 16
> +#define MVNETA_TXQ_STATUS_REG(q) (0x3c40 + ((q) << 2))
> +#define MVNETA_PORT_TX_RESET 0x3cf0
> +#define MVNETA_TX_MTU 0x3e0c
> +#define MVNETA_TX_TOKEN_SIZE 0x3e14
> +#define MVNETA_TXQ_TOKEN_SIZE_REG(q) (0x3e40 + ((q) << 2))
> +
> +/* The mvneta_tx_desc and mvneta_rx_desc structures describe the
> + * layout of the transmit and reception DMA descriptors, and their
> + * layout is therefore defined by the hardware design
> + */
> +
> +#define MVNETA_TX_L3_OFF_SHIFT 0
> +#define MVNETA_TX_IP_HLEN_SHIFT 8
> +#define MVNETA_TX_L4_UDP BIT(16)
> +#define MVNETA_TX_L3_IP6 BIT(17)
> +#define MVNETA_TXD_IP_CSUM BIT(18)
> +#define MVNETA_TXD_Z_PAD BIT(19)
> +#define MVNETA_TXD_L_DESC BIT(20)
> +#define MVNETA_TXD_F_DESC BIT(21)
> +#define MVNETA_TXD_FLZ_DESC (MVNETA_TXD_Z_PAD | \
> + MVNETA_TXD_L_DESC | \
> + MVNETA_TXD_F_DESC)
> +#define MVNETA_TX_L4_CSUM_FULL BIT(30)
> +#define MVNETA_TX_L4_CSUM_NOT BIT(31)
> +
> +#define MVNETA_TXD_ERROR BIT(0)
> +#define TXD_ERROR_MASK 0x6
> +#define TXD_ERROR_SHIFT 1
> +
> +#define MVNETA_RXD_ERR_CRC 0x0
> +#define MVNETA_RXD_ERR_SUMMARY BIT(16)
> +#define MVNETA_RXD_ERR_OVERRUN BIT(17)
> +#define MVNETA_RXD_ERR_LEN BIT(18)
> +#define MVNETA_RXD_ERR_RESOURCE (BIT(17) | BIT(18))
> +#define MVNETA_RXD_ERR_CODE_MASK (BIT(17) | BIT(18))
> +#define MVNETA_RXD_L3_IP4 BIT(25)
> +#define MVNETA_RXD_FIRST_LAST_DESC (BIT(26) | BIT(27))
> +#define MVNETA_RXD_L4_CSUM_OK BIT(30)
> +
> +#define MVNETA_MH_SIZE 2
> +#define TXQ_NUM 8
> +#define RX_RING_SIZE 4
> +#define TRANSFER_TIMEOUT (10 * MSECOND)
> +
> +struct rxdesc {
> + u32 cmd_sts; /* Info about received packet */
> + u16 reserved1; /* pnc_info - (for future use, PnC) */
> + u16 data_size; /* Size of received packet in bytes */
> +
> + u32 buf_phys_addr; /* Physical address of the buffer */
> + u32 reserved2; /* pnc_flow_id (for future use, PnC) */
> +
> + u32 buf_cookie; /* cookie for access to RX buffer in rx path */
> + u16 reserved3; /* prefetch_cmd, for future use */
> + u16 reserved4; /* csum_l4 - (for future use, PnC) */
> +
> + u32 reserved5; /* pnc_extra PnC (for future use, PnC) */
> + u32 reserved6; /* hw_cmd (for future use, PnC and HWF) */
> +};
> +
> +struct txdesc {
> + u32 cmd_sts; /* Options used by HW for packet transmitting.*/
> + u16 reserverd1; /* csum_l4 (for future use) */
> + u16 byte_cnt; /* Data size of transmitted packet in bytes */
> + u32 buf_ptr; /* Physical addr of transmitted buffer */
> + u8 error; /* */
> + u8 reserved2; /* Reserved - (for future use) */
> + u16 reserved3; /* Reserved - (for future use) */
> + u32 reserved4[4]; /* Reserved - (for future use) */
> +};
> +
> +struct mvneta_port {
> + void __iomem *reg;
> + struct device_d dev;
> + struct eth_device edev;
> + struct clk *clk;
> +
> + struct txdesc *txdesc;
> + struct rxdesc *rxdesc;
> + int curr_rxdesc;
> + u8 *rxbuf;
> + phy_interface_t intf;
> +};
> +
> +static void mvneta_conf_mbus_windows(struct mvneta_port *priv)
> +{
> + const struct mbus_dram_target_info *dram = mvebu_mbus_dram_info();
> + u32 win_enable, win_protect;
> + int i;
> +
> + for (i = 0; i < 6; i++) {
> + writel(0, priv->reg + MVNETA_WIN_BASE(i));
> + writel(0, priv->reg + MVNETA_WIN_SIZE(i));
> +
> + if (i < 4)
> + writel(0, priv->reg + MVNETA_WIN_REMAP(i));
> + }
> +
> + win_enable = 0x3f;
> + win_protect = 0;
> +
> + for (i = 0; i < dram->num_cs; i++) {
> + const struct mbus_dram_window *cs = dram->cs + i;
> +
> + writel((cs->base & 0xffff0000) |
> + (cs->mbus_attr << 8) |
> + dram->mbus_dram_target_id,
> + priv->reg + MVNETA_WIN_BASE(i));
> +
> + writel((cs->size - 1) & 0xffff0000,
> + priv->reg + MVNETA_WIN_SIZE(i));
> +
> + win_enable &= ~(1 << i);
> + win_protect |= 3 << (2 * i);
> + }
> +
> + writel(win_enable, priv->reg + MVNETA_BASE_ADDR_ENABLE);
> +}
> +
> +static void mvneta_clear_mcast_table(struct mvneta_port *priv)
> +{
> + int offset;
> +
> + for (offset = 0; offset <= 0xfc; offset += 4) {
> + writel(0, priv->reg + MVNETA_DA_FILT_UCAST_BASE + offset);
> + writel(0, priv->reg + MVNETA_DA_FILT_SPEC_MCAST + offset);
> + writel(0, priv->reg + MVNETA_DA_FILT_OTH_MCAST + offset);
> + }
> +}
> +
> +/* Set unicast address */
> +static void mvneta_set_ucast_addr(struct mvneta_port *priv, u8 last_nibble)
> +{
> + unsigned int tbl_offset, reg_offset;
> + int queue = 0;
> + u32 val;
> +
> + /* Locate the Unicast table entry */
> + last_nibble = (0xf & last_nibble);
> +
> + /* offset from unicast tbl base */
> + tbl_offset = (last_nibble / 4) * 4;
> +
> + /* offset within the above reg */
> + reg_offset = last_nibble % 4;
> +
> + val = readl(priv->reg + MVNETA_DA_FILT_UCAST_BASE + tbl_offset);
> +
> + val &= ~(0xff << (8 * reg_offset));
> + val |= ((0x01 | (queue << 1)) << (8 * reg_offset));
> +
> + writel(val, priv->reg + MVNETA_DA_FILT_UCAST_BASE + tbl_offset);
> +}
> +
> +static void mvneta_clear_ucast_addr(struct mvneta_port *priv, u8 last_nibble)
> +{
> + unsigned int tbl_offset, reg_offset;
> + u32 val;
> +
> + /* Locate the Unicast table entry */
> + last_nibble = (0xf & last_nibble);
> +
> + /* offset from unicast tbl base */
> + tbl_offset = (last_nibble / 4) * 4;
> +
> + /* offset within the above reg */
> + reg_offset = last_nibble % 4;
> +
> + val = readl(priv->reg + MVNETA_DA_FILT_UCAST_BASE + tbl_offset);
> +
> + /* Clear accepts frame bit at specified unicast DA tbl entry */
> + val &= ~(0xff << (8 * reg_offset));
> +
> + writel(val, priv->reg + MVNETA_DA_FILT_UCAST_BASE + tbl_offset);
> +}
> +
> +static void mvneta_rx_unicast_promisc_clear(struct mvneta_port *priv)
> +{
> + u32 portcfg, val;
> +
> + portcfg = readl(priv->reg + MVNETA_PORT_CONFIG);
> + val = readl(priv->reg + MVNETA_TYPE_PRIO);
> +
> + /* Reject all Unicast addresses */
> + writel(portcfg & ~BIT(0), priv->reg + MVNETA_PORT_CONFIG);
> + writel(val & ~BIT(21), priv->reg + MVNETA_TYPE_PRIO);
> +}
> +
> +/* Clear all MIB counters */
> +static void mvneta_mib_counters_clear(struct mvneta_port *priv)
> +{
> + int i;
> +
> + /* Perform dummy reads from MIB counters */
> + for (i = 0; i < MVNETA_MIB_LATE_COLLISION; i += 4)
> + readl(priv->reg + MVNETA_MIB_COUNTERS_BASE + i);
> +}
> +
> +static int mvneta_pending_tx(struct mvneta_port *priv)
> +{
> + u32 val = readl(priv->reg + MVNETA_TXQ_STATUS_REG(0));
> +
> + return val & 0x3fff;
> +}
> +
> +static int mvneta_pending_rx(struct mvneta_port *priv)
> +{
> + u32 val = readl(priv->reg + MVNETA_RXQ_STATUS_REG(0));
> +
> + return val & 0x3fff;
> +}
> +
> +static void mvneta_port_stop(struct mvneta_port *priv)
> +{
> + u32 val;
> +
> + /* Stop all queues */
> + writel(0xff << MVNETA_RXQ_DISABLE_SHIFT, priv->reg + MVNETA_RXQ_CMD);
> + writel(0xff << MVNETA_TXQ_DISABLE_SHIFT, priv->reg + MVNETA_TXQ_CMD);
> +
> + /* Reset Tx */
> + writel(BIT(0), priv->reg + MVNETA_PORT_TX_RESET);
> + writel(0, priv->reg + MVNETA_PORT_TX_RESET);
> +
> + /* Reset Rx */
> + writel(BIT(0), priv->reg + MVNETA_PORT_RX_RESET);
> + writel(0, priv->reg + MVNETA_PORT_RX_RESET);
> +
> + /* Disable port 0 */
> + val = readl(priv->reg + MVNETA_GMAC_CTRL_0);
> + writel(val & ~BIT(0), priv->reg + MVNETA_GMAC_CTRL_0);
> + udelay(200);
> +
> + /* Clear all Cause registers */
> + writel(0, priv->reg + MVNETA_INTR_NEW_CAUSE);
> + writel(0, priv->reg + MVNETA_INTR_OLD_CAUSE);
> + writel(0, priv->reg + MVNETA_INTR_MISC_CAUSE);
> + writel(0, priv->reg + MVNETA_UNIT_INTR_CAUSE);
> +
> + /* Mask all interrupts */
> + writel(0, priv->reg + MVNETA_INTR_NEW_MASK);
> + writel(0, priv->reg + MVNETA_INTR_OLD_MASK);
> + writel(0, priv->reg + MVNETA_INTR_MISC_MASK);
> + writel(0, priv->reg + MVNETA_INTR_ENABLE);
> +}
> +
> +static void mvneta_halt(struct eth_device *edev)
> +{
> + mvneta_port_stop((struct mvneta_port *)edev->priv);
> +}
> +
> +static int mvneta_send(struct eth_device *edev, void *data, int len)
> +{
> + struct mvneta_port *priv = edev->priv;
> + struct txdesc *txdesc = priv->txdesc;
> + int ret, error, last_desc;
> +
> + /* Flush transmit data */
> + dma_flush_range((unsigned long)data, (unsigned long)data+len);
> +
> + /* Fill the Tx descriptor */
> + txdesc->cmd_sts |= MVNETA_TX_L4_CSUM_NOT | MVNETA_TXD_FLZ_DESC;
> + txdesc->buf_ptr = (u32)data;
> + txdesc->byte_cnt = len;
> +
> + /* Increase the number of prepared descriptors (one), by writing
> + * to the 'NoOfWrittenDescriptors' field in the PTXSU register.
> + */
> + writel(1, priv->reg + MVNETA_TXQ_UPDATE_REG(0));
> +
> + /* The controller updates the number of transmitted descriptors in
> + * the Tx port status register (PTXS).
> + */
> + ret = wait_on_timeout(TRANSFER_TIMEOUT, !mvneta_pending_tx(priv));
> + if (ret) {
> + dev_err(&edev->dev, "transmit timeout\n");
> + return ret;
> + }
> +
> + last_desc = readl(&txdesc->cmd_sts) & MVNETA_TXD_L_DESC;
> + error = readl(&txdesc->error);
> + if (last_desc && error & MVNETA_TXD_ERROR) {
> + dev_err(&edev->dev, "transmit error %d\n",
> + (error & TXD_ERROR_MASK) >> TXD_ERROR_SHIFT);
> + return -EIO;
> + }
> +
> + /* Release the transmitted descriptor by writing to the
> + * 'NoOfReleasedBuffers' field in the PTXSU register.
> + */
> + writel(1 << MVNETA_TXQ_DEC_SENT_SHIFT,
> + priv->reg + MVNETA_TXQ_UPDATE_REG(0));
> +
> + return 0;
> +}
> +
> +static int mvneta_recv(struct eth_device *edev)
> +{
> + struct mvneta_port *priv = edev->priv;
> + struct rxdesc *rxdesc = &priv->rxdesc[priv->curr_rxdesc];
> + int ret, pending;
> + u32 cmd_sts;
> +
> + /* wait for received packet */
> + pending = mvneta_pending_rx(priv);
> + if (!pending)
> + return 0;
> +
> + /* drop malicious packets */
> + cmd_sts = readl(&rxdesc->cmd_sts);
> + if ((cmd_sts & MVNETA_RXD_FIRST_LAST_DESC) !=
> + MVNETA_RXD_FIRST_LAST_DESC) {
> + dev_err(&edev->dev, "rx packet spread on multiple descriptors\n");
> + ret = -EIO;
> + goto recv_err;
> + }
> +
> + if (cmd_sts & MVNETA_RXD_ERR_SUMMARY) {
> + dev_err(&edev->dev, "receive error\n");
> + ret = -EIO;
> + goto recv_err;
> + }
> +
> + /* invalidate current receive buffer */
> + dma_inv_range((unsigned long)rxdesc->buf_phys_addr,
> + (unsigned long)rxdesc->buf_phys_addr +
> + ALIGN(PKTSIZE, 8));
> +
> + /* received packet is padded with two null bytes (Marvell header) */
> + net_receive(edev, (void *)(rxdesc->buf_phys_addr + MVNETA_MH_SIZE),
> + rxdesc->data_size - MVNETA_MH_SIZE);
> + ret = 0;
> +
> +recv_err:
> + /* reset this and get next rx descriptor*/
> + rxdesc->data_size = 0;
> + rxdesc->cmd_sts = 0;
> +
> + priv->curr_rxdesc++;
> + if (priv->curr_rxdesc == RX_RING_SIZE)
> + priv->curr_rxdesc = 0;
> +
> + /* Descriptor processed and refilled */
> + writel(1 | 1 << MVNETA_RXQ_ADD_NON_OCCUPIED_SHIFT,
> + priv->reg + MVNETA_RXQ_STATUS_UPDATE_REG(0));
> + return ret;
> +}
> +
> +static int mvneta_set_ethaddr(struct eth_device *edev, unsigned char *mac)
> +{
> + struct mvneta_port *priv = edev->priv;
> + u32 mac_h = (mac[0] << 24) | (mac[1] << 16) | (mac[2] << 8) | mac[3];
> + u32 mac_l = (mac[4] << 8) | mac[5];
> +
> + mvneta_clear_ucast_addr(priv, mac[5]);
> +
> + writel(mac_l, priv->reg + MVNETA_MAC_ADDR_LOW);
> + writel(mac_h, priv->reg + MVNETA_MAC_ADDR_HIGH);
> +
> + /* accept frames for this address */
> + mvneta_set_ucast_addr(priv, mac[5]);
> +
> + return 0;
> +}
> +
> +static int mvneta_get_ethaddr(struct eth_device *edev, unsigned char *mac)
> +{
> + struct mvneta_port *priv = edev->priv;
> + u32 mac_l = readl(priv->reg + MVNETA_MAC_ADDR_LOW);
> + u32 mac_h = readl(priv->reg + MVNETA_MAC_ADDR_HIGH);
> +
> + mac[0] = (mac_h >> 24) & 0xff;
> + mac[1] = (mac_h >> 16) & 0xff;
> + mac[2] = (mac_h >> 8) & 0xff;
> + mac[3] = mac_h & 0xff;
> + mac[4] = (mac_l >> 8) & 0xff;
> + mac[5] = mac_l & 0xff;
> +
> + return 0;
> +}
> +
> +static void mvneta_adjust_link(struct eth_device *edev)
> +{
> + struct mvneta_port *priv = edev->priv;
> + struct phy_device *phy = edev->phydev;
> + u32 val;
> +
> + if (!phy->link)
> + return;
> +
> + val = readl(priv->reg + MVNETA_GMAC_AUTONEG_CONFIG);
> + val &= ~(MVNETA_GMAC_CONFIG_MII_SPEED |
> + MVNETA_GMAC_CONFIG_GMII_SPEED |
> + MVNETA_GMAC_CONFIG_FULL_DUPLEX |
> + MVNETA_GMAC_AN_SPEED_EN |
> + MVNETA_GMAC_AN_DUPLEX_EN);
> +
> + if (phy->speed == SPEED_1000)
> + val |= MVNETA_GMAC_CONFIG_GMII_SPEED;
> + else if (phy->speed == SPEED_100)
> + val |= MVNETA_GMAC_CONFIG_MII_SPEED;
> +
> + if (phy->duplex)
> + val |= MVNETA_GMAC_CONFIG_FULL_DUPLEX;
> +
> + val |= MVNETA_GMAC_FORCE_LINK_PASS | MVNETA_GMAC_FORCE_LINK_DOWN;
> +
> + writel(val, priv->reg + MVNETA_GMAC_AUTONEG_CONFIG);
> +
> + mvneta_mib_counters_clear(priv);
> +
> + /* Enable first Tx and first Rx queues */
> + writel(BIT(0), priv->reg + MVNETA_TXQ_CMD);
> + writel(BIT(0), priv->reg + MVNETA_RXQ_CMD);
> +}
> +
> +static int mvneta_open(struct eth_device *edev)
> +{
> + struct mvneta_port *priv = edev->priv;
> + int ret;
> + u32 val;
> +
> + ret = phy_device_connect(&priv->edev, NULL, -1,
> + mvneta_adjust_link, 0, priv->intf);
> + if (ret)
> + return ret;
> +
> + /* Enable the first port */
> + val = readl(priv->reg + MVNETA_GMAC_CTRL_0);
> + writel(val | BIT(0), priv->reg + MVNETA_GMAC_CTRL_0);
> +
> + return 0;
> +}
> +
> +static void mvneta_init_rx_ring(struct mvneta_port *priv)
> +{
> + int i;
> +
> + for (i = 0; i < RX_RING_SIZE; i++) {
> + struct rxdesc *desc = &priv->rxdesc[i];
> +
> + desc->buf_phys_addr = (u32)priv->rxbuf + i * ALIGN(PKTSIZE, 8);
> + }
> +
> + priv->curr_rxdesc = 0;
> +}
> +
> +void mvneta_setup_tx_rx(struct mvneta_port *priv)
> +{
> + u32 val;
> +
> + /* Allocate descriptors and buffers */
> + priv->txdesc = dma_alloc_coherent(ALIGN(sizeof(*priv->txdesc), 32));
> + priv->rxdesc = dma_alloc_coherent(RX_RING_SIZE *
> + ALIGN(sizeof(*priv->rxdesc), 32));
> + priv->rxbuf = dma_alloc(RX_RING_SIZE * ALIGN(PKTSIZE, 8));
> +
> + mvneta_init_rx_ring(priv);
> +
> + /* Configure the Rx queue */
> + val = readl(priv->reg + MVNETA_RXQ_CONFIG_REG(0));
> + val &= ~MVNETA_RXQ_PKT_OFFSET_ALL_MASK;
> + writel(val, priv->reg + MVNETA_RXQ_CONFIG_REG(0));
> +
> + /* Configure the Tx descriptor */
> + writel(1, priv->reg + MVNETA_TXQ_SIZE_REG(0));
> + writel((u32)priv->txdesc, priv->reg + MVNETA_TXQ_BASE_ADDR_REG(0));
> +
> + /* Configure the Rx descriptor. Packet size is in 8-byte units. */
> + val = RX_RING_SIZE;
> + val |= ((PKTSIZE >> 3) << MVNETA_RXQ_BUF_SIZE_SHIFT);
> + writel(val , priv->reg + MVNETA_RXQ_SIZE_REG(0));
> + writel((u32)priv->rxdesc, priv->reg + MVNETA_RXQ_BASE_ADDR_REG(0));
> +
> + /* Set the number of available Rx descriptors */
> + writel(RX_RING_SIZE << MVNETA_RXQ_ADD_NON_OCCUPIED_SHIFT,
> + priv->reg + MVNETA_RXQ_STATUS_UPDATE_REG(0));
> +}
> +
> +static int mvneta_port_config(struct mvneta_port *priv)
> +{
> + int queue;
> + u32 val;
> +
> + /* Enable MBUS Retry bit16 */
> + writel(0x20, priv->reg + MVNETA_MBUS_RETRY);
> +
> + /* Map the first Tx queue and first Rx queue to CPU0 */
> + writel(BIT(0) | (BIT(0) << 8), priv->reg + MVNETA_CPU_MAP(0));
> +
> + /* Reset Tx/Rx DMA */
> + writel(BIT(0), priv->reg + MVNETA_PORT_TX_RESET);
> + writel(BIT(0), priv->reg + MVNETA_PORT_RX_RESET);
> +
> + /* Disable Legacy WRR, Disable EJP, Release from reset */
> + writel(0, priv->reg + MVNETA_TXQ_CMD_1);
> +
> + /* Set maximum bandwidth for the first TX queue */
> + writel(0x3ffffff, priv->reg + MVETH_TXQ_TOKEN_CFG_REG(0));
> + writel(0x3ffffff, priv->reg + MVETH_TXQ_TOKEN_COUNT_REG(0));
> +
> + /* Minimum bandwidth on the rest of them */
> + for (queue = 1; queue < TXQ_NUM; queue++) {
> + writel(0, priv->reg + MVETH_TXQ_TOKEN_COUNT_REG(queue));
> + writel(0, priv->reg + MVETH_TXQ_TOKEN_CFG_REG(queue));
> + }
> +
> + writel(0, priv->reg + MVNETA_PORT_RX_RESET);
> + writel(0, priv->reg + MVNETA_PORT_TX_RESET);
> +
> + /* Disable hardware PHY polling */
> + val = readl(priv->reg + MVNETA_UNIT_CONTROL);
> + writel(val & ~BIT(1), priv->reg + MVNETA_UNIT_CONTROL);
> +
> + /* Port Acceleration Mode */
> + writel(0x1, priv->reg + MVNETA_ACC_MODE);
> +
> + /* Port default configuration for the first Rx queue */
> + val = MVNETA_PORT_CONFIG_DEFL_VALUE(0);
> + writel(val, priv->reg + MVNETA_PORT_CONFIG);
> + writel(0, priv->reg + MVNETA_PORT_CONFIG_EXTEND);
> + writel(64, priv->reg + MVNETA_RX_MIN_FRAME_SIZE);
> +
> + /* Default burst size */
> + val = 0;
> + val |= MVNETA_TX_BRST_SZ_MASK(MVNETA_SDMA_BRST_SIZE_16);
> + val |= MVNETA_RX_BRST_SZ_MASK(MVNETA_SDMA_BRST_SIZE_16);
> + val |= MVNETA_RX_NO_DATA_SWAP | MVNETA_TX_NO_DATA_SWAP;
> + writel(val, priv->reg + MVNETA_SDMA_CONFIG);
> +
> + mvneta_clear_mcast_table(priv);
> + mvneta_rx_unicast_promisc_clear(priv);
> +
> + /* Configure maximum MTU and token size */
> + writel(0x0003ffff, priv->reg + MVNETA_TX_MTU);
> + writel(0xffffffff, priv->reg + MVNETA_TX_TOKEN_SIZE);
> + writel(0x7fffffff, priv->reg + MVNETA_TXQ_TOKEN_SIZE_REG(0));
> +
> + val = readl(priv->reg + MVNETA_GMAC_CTRL_2);
> +
> + /* Even though it might look weird, when we're configured in
> + * SGMII or QSGMII mode, the RGMII bit needs to be set.
> + */
> + switch (priv->intf) {
> + case PHY_INTERFACE_MODE_QSGMII:
> + writel(MVNETA_QSGMII_SERDES, priv->reg + MVNETA_SERDES_CFG);
> + val |= MVNETA_GMAC2_PCS_ENABLE | MVNETA_GMAC2_PORT_RGMII;
> + break;
> + case PHY_INTERFACE_MODE_SGMII:
> + writel(MVNETA_SGMII_SERDES, priv->reg + MVNETA_SERDES_CFG);
> + val |= MVNETA_GMAC2_PCS_ENABLE | MVNETA_GMAC2_PORT_RGMII;
> + break;
> + case PHY_INTERFACE_MODE_RGMII:
> + case PHY_INTERFACE_MODE_RGMII_ID:
> + val |= MVNETA_GMAC2_PORT_RGMII;
> + break;
> + default:
> + return -EINVAL;
> + }
> +
> + /* Cancel Port Reset */
> + val &= ~MVNETA_GMAC2_PORT_RESET;
> + writel(val, priv->reg + MVNETA_GMAC_CTRL_2);
> + while (readl(priv->reg + MVNETA_GMAC_CTRL_2) & MVNETA_GMAC2_PORT_RESET)
> + continue;
> +
> + return 0;
> +}
> +
> +static int mvneta_probe(struct device_d *dev)
> +{
> + struct mvneta_port *priv;
> + int ret;
> +
> + priv = xzalloc(sizeof(*priv));
> +
> + priv->reg = dev_get_mem_region(dev, 0);
> +
> + priv->clk = clk_get(dev, 0);
> + if (IS_ERR(priv->clk))
> + return PTR_ERR(priv->clk);
> + clk_enable(priv->clk);
> +
> + ret = of_get_phy_mode(dev->device_node);
> + if (ret < 0)
> + return ret;
> + priv->intf = ret;
> +
> + mvneta_port_stop(priv);
> +
> + ret = mvneta_port_config(priv);
> + if (ret)
> + return ret;
> +
> + mvneta_conf_mbus_windows(priv);
> + mvneta_setup_tx_rx(priv);
> +
> + /* register eth device */
> + priv->edev.priv = priv;
> + priv->edev.open = mvneta_open;
> + priv->edev.send = mvneta_send;
> + priv->edev.recv = mvneta_recv;
> + priv->edev.halt = mvneta_halt;
> + priv->edev.set_ethaddr = mvneta_set_ethaddr;
> + priv->edev.get_ethaddr = mvneta_get_ethaddr;
> + priv->edev.parent = dev;
> +
> + ret = eth_register(&priv->edev);
> + if (ret)
> + return ret;
> + return 0;
> +}
> +
> +static struct of_device_id mvneta_dt_ids[] = {
> + { .compatible = "marvell,armada-370-neta", },
> + { }
> +};
> +
> +static struct driver_d mvneta_driver = {
> + .name = "mvneta",
> + .probe = mvneta_probe,
> + .of_compatible = DRV_OF_COMPAT(mvneta_dt_ids),
> +};
> +device_platform_driver(mvneta_driver);
>
_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox
next prev parent reply other threads:[~2014-08-26 14:30 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-08-23 22:53 [PATCH 0/5] mvebu: Add network support for Armada 370/XP Ezequiel Garcia
2014-08-23 22:53 ` [PATCH 1/5] ARM: mvebu: Enable PUP register Ezequiel Garcia
2014-08-26 14:28 ` Sebastian Hesselbarth
2014-08-23 22:53 ` [PATCH 2/5] net: phy: Support Marvell 88EE1545 PHY Ezequiel Garcia
2014-08-23 22:53 ` [PATCH 3/5] net: phy: Support Marvell 88EE1543 PHY Ezequiel Garcia
2014-08-26 20:13 ` [PATCH] net: phy: Support Marvell 88E1318S and 88E1510 PHYs Sebastian Hesselbarth
2014-08-23 22:53 ` [PATCH 4/5] net: Add driver for Armada 370/XP 10/100/1000 Mbps network controller Ezequiel Garcia
2014-08-26 14:30 ` Sebastian Hesselbarth [this message]
2014-08-26 16:20 ` Ezequiel Garcia
2014-08-26 20:06 ` Sebastian Hesselbarth
2014-08-23 22:53 ` [PATCH 5/5] configs: Add network options to Armada 370/XP boards Ezequiel Garcia
2014-09-01 8:55 ` [PATCH 0/5] mvebu: Add network support for Armada 370/XP Sascha Hauer
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=53FC9A00.9090608@gmail.com \
--to=sebastian.hesselbarth@gmail.com \
--cc=barebox@lists.infradead.org \
--cc=ezequiel.garcia@free-electrons.com \
--cc=thomas.petazzoni@free-electrons.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox