This series includes adding realtek automotive ethernet driver
and adding rtase ethernet driver entry in MAINTAINERS file.
This ethernet device driver for the PCIe interface of
Realtek Automotive Ethernet Switch,applicable to
RTL9054, RTL9068, RTL9072, RTL9075, RTL9068, RTL9071.
v1 -> v2:
- Remove redundent debug message.
- Modify coding rule.
- Remove other function codes not related to netdev.
v2 -> v3:
- Remove SR-IOV function - We will add the SR-IOV function together when
uploading the vf driver in the future.
- Remove other unnecessary code and macro.
v3 -> v4:
- Remove function prototype - Our driver does not use recursion, so we
have reordered the code and removed the function prototypes.
- Define macro precisely - Improve macro code readability to make the
source code cleaner.
v4 -> v5:
- Modify ethtool function - Remove some unnecessary code.
- Don't use inline function - Let the compiler decide.
v5 -> v6:
- Some old macro definitions have been removed and replaced with the
lastest usage.
- Replace s32 with int to ensure consistency.
- Clearly point out the objects of the service and remove unnecessary
struct.
v6 -> v7:
- Split this driver into multiple patches.
- Reorganize this driver code and remove redundant code to make this
driver more concise.
v7 -> v8:
- Add the function to calculate time mitigation and the function to
calculate packet number mitigation. Users can use these two functions
to calculate the reg value that needs to be set for the mitigation value
they want to set.
- This device is usually used in automotive embedded systems. The page
pool api will use more memory in receiving packets and requires more
verification, so we currently do not plan to use it in this patch.
v8 -> v9:
- Declare functions that are not extern as static functions and increase
the size of the character array named name in the rtase_int_vector struct
to correct the build warning noticed by the kernel test robot.
v9 -> v10:
- Currently we change to use the page pool api. However, when we allocate
more than one page to an rx buffer, it will cause system errors
in some cases. Therefore, we set the rx buffer to fixed size with 3776
(PAGE_SIZE - SKB_DATA_ALIGN(sizeof(skb_shared_info) )), and the maximum
value of mtu is set to 3754(rx buffer size - VLAN_ETH_HLEN - ETH_FCS_LEN).
- When ndo_tx_timeout is called, it will dump some device information,
which can be used for debugging.
- When the mtu is greater than 1500, the device supports checksums
but not TSO.
- Fix compiler warnning.
v10 -> v11:
- Added error handling of rtase_init_ring().
- Modify the error related to asymmetric pause in rtase_get_settings.
- Fix compiler error.
v11 -> v12:
- Use pm_sleep_ptr and related macros.
- Remove multicast filter limit.
- Remove VLAN support and CBS offload functions.
- Remove redundent code.
- Fix compiler warnning.
v12 -> v13:
- Fixed the compiler warning of unuse rtase_suspend() and rtase_resume()
when there is no define CONFIG_PM_SLEEP.
v13 -> v14:
- Remove unuse include.
- call eth_hw_addr_random() to generate random MAC and set device flag
- use pci_enable_msix_exact() instead of pci_enable_msix_range()
- If dev->dma_mask is non-NULL, dma_set_mask_and_coherent with a 64-bit
mask will never fail, so remove the part that determines the 32-bit mask.
- set dev->pcpu_stat_type before register_netdev() and core will allocate
stats
- call NAPI instance at the right location
v14 -> v15:
- In rtase_open, when the request interrupt fails, all request interrupts
are freed.
- When calling netif_device_detach, there is no need to call
netif_stop_queue.
- Call netif_tx_disable() instead of stop_queue(), it takes the tx lock so
there is no need to worry about the packets being transmitted.
- In rtase_tx_handler, napi budget is no longer used, but a customized
tx budget is used.
- Use the start / stop macros from include/net/netdev_queues.h.
- Remove redundent code.
v15 -> v16:
- Re-upload v15 patch set
v16 -> v17:
- Prefix the names of some rtase-specific macros, structs, and enums.
- Fix the abnormal problem when returning page_pool resources.
v17 -> v18:
- Limit the width of each line to 80 colums.
- Use reverse xmas tree order.
- Modify the error handling of rtase_alloc_msix and rtase_alloc_interrupt.
v18 -> v19:
- Use dma_wmb() instead of wmb() to ensure the order of access
instructions for a memory shared by DMA and CPU.
- Add error message when allocate dma memory fails.
- Add .get_eth_mac_stats function to report hardware information.
- Remove .get_ethtool_stats function.
- In rtase_tx_csum, when the packet is not ipv6 or ipv4, a warning will
no longer be issued.
v19 -> v20:
- Modify the description of switch architecture.
Justin Lai (13):
rtase: Add pci table supported in this module
rtase: Implement the .ndo_open function
rtase: Implement the rtase_down function
rtase: Implement the interrupt routine and rtase_poll
rtase: Implement hardware configuration function
rtase: Implement .ndo_start_xmit function
rtase: Implement a function to receive packets
rtase: Implement net_device_ops
rtase: Implement pci_driver suspend and resume function
rtase: Implement ethtool function
rtase: Add a Makefile in the rtase folder
realtek: Update the Makefile and Kconfig in the realtek folder
MAINTAINERS: Add the rtase ethernet driver entry
MAINTAINERS | 7 +
drivers/net/ethernet/realtek/Kconfig | 19 +
drivers/net/ethernet/realtek/Makefile | 1 +
drivers/net/ethernet/realtek/rtase/Makefile | 10 +
drivers/net/ethernet/realtek/rtase/rtase.h | 327 +++
.../net/ethernet/realtek/rtase/rtase_main.c | 2366 +++++++++++++++++
6 files changed, 2730 insertions(+)
create mode 100644 drivers/net/ethernet/realtek/rtase/Makefile
create mode 100644 drivers/net/ethernet/realtek/rtase/rtase.h
create mode 100644 drivers/net/ethernet/realtek/rtase/rtase_main.c
--
2.34.1
Add pci table supported in this module, and implement pci_driver function
to initialize this driver, remove this driver, or shutdown this driver.
Signed-off-by: Justin Lai <[email protected]>
---
drivers/net/ethernet/realtek/rtase/rtase.h | 327 +++++++++
.../net/ethernet/realtek/rtase/rtase_main.c | 640 ++++++++++++++++++
2 files changed, 967 insertions(+)
create mode 100644 drivers/net/ethernet/realtek/rtase/rtase.h
create mode 100644 drivers/net/ethernet/realtek/rtase/rtase_main.c
diff --git a/drivers/net/ethernet/realtek/rtase/rtase.h b/drivers/net/ethernet/realtek/rtase/rtase.h
new file mode 100644
index 000000000000..d9d4871216cd
--- /dev/null
+++ b/drivers/net/ethernet/realtek/rtase/rtase.h
@@ -0,0 +1,327 @@
+/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */
+/*
+ * rtase is the Linux device driver released for Realtek Automotive Switch
+ * controllers with PCI-Express interface.
+ *
+ * Copyright(c) 2024 Realtek Semiconductor Corp.
+ */
+
+#ifndef _RTASE_H_
+#define _RTASE_H_
+
+#define RTASE_HW_VER_MASK 0x7C800000
+
+#define RTASE_RX_DMA_BURST_256 4
+#define RTASE_TX_DMA_BURST_UNLIMITED 7
+
+#define RTASE_RX_BUF_SIZE (PAGE_SIZE - \
+ SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))
+#define RTASE_MAX_JUMBO_SIZE (RTASE_RX_BUF_SIZE - VLAN_ETH_HLEN - ETH_FCS_LEN)
+
+/* 3 means InterFrameGap = the shortest one */
+#define RTASE_INTERFRAMEGAP 0x03
+
+#define RTASE_REGS_SIZE 256
+#define RTASE_PCI_REGS_SIZE 0x100
+
+#define RTASE_MULTICAST_FILTER_MASK GENMASK(30, 26)
+
+#define RTASE_VLAN_FILTER_ENTRY_NUM 32
+#define RTASE_NUM_TX_QUEUE 8
+#define RTASE_NUM_RX_QUEUE 4
+
+#define RTASE_TXQ_CTRL 1
+#define RTASE_FUNC_TXQ_NUM 1
+#define RTASE_FUNC_RXQ_NUM 1
+#define RTASE_INTERRUPT_NUM 1
+
+#define RTASE_MITI_TIME_COUNT_MASK GENMASK(3, 0)
+#define RTASE_MITI_TIME_UNIT_MASK GENMASK(7, 4)
+#define RTASE_MITI_DEFAULT_TIME 128
+#define RTASE_MITI_MAX_TIME 491520
+#define RTASE_MITI_PKT_NUM_COUNT_MASK GENMASK(11, 8)
+#define RTASE_MITI_PKT_NUM_UNIT_MASK GENMASK(13, 12)
+#define RTASE_MITI_DEFAULT_PKT_NUM 64
+#define RTASE_MITI_MAX_PKT_NUM_IDX 3
+#define RTASE_MITI_MAX_PKT_NUM_UNIT 16
+#define RTASE_MITI_MAX_PKT_NUM 240
+#define RTASE_MITI_COUNT_BIT_NUM 4
+
+#define RTASE_NUM_MSIX 4
+
+#define RTASE_DWORD_MOD 16
+
+/*****************************************************************************/
+enum rtase_registers {
+ RTASE_MAC0 = 0x0000,
+ RTASE_MAC4 = 0x0004,
+ RTASE_MAR0 = 0x0008,
+ RTASE_MAR1 = 0x000C,
+ RTASE_DTCCR0 = 0x0010,
+ RTASE_DTCCR4 = 0x0014,
+#define RTASE_COUNTER_RESET BIT(0)
+#define RTASE_COUNTER_DUMP BIT(3)
+
+ RTASE_FCR = 0x0018,
+#define RTASE_FCR_RXQ_MASK GENMASK(5, 4)
+
+ RTASE_LBK_CTRL = 0x001A,
+#define RTASE_LBK_ATLD BIT(1)
+#define RTASE_LBK_CLR BIT(0)
+
+ RTASE_TX_DESC_ADDR0 = 0x0020,
+ RTASE_TX_DESC_ADDR4 = 0x0024,
+ RTASE_TX_DESC_COMMAND = 0x0028,
+#define RTASE_TX_DESC_CMD_CS BIT(15)
+#define RTASE_TX_DESC_CMD_WE BIT(14)
+
+ RTASE_BOOT_CTL = 0x6004,
+ RTASE_CLKSW_SET = 0x6018,
+
+ RTASE_CHIP_CMD = 0x0037,
+#define RTASE_STOP_REQ BIT(7)
+#define RTASE_STOP_REQ_DONE BIT(6)
+#define RTASE_RE BIT(3)
+#define RTASE_TE BIT(2)
+
+ RTASE_IMR0 = 0x0038,
+ RTASE_ISR0 = 0x003C,
+#define RTASE_TOK7 BIT(30)
+#define RTASE_TOK6 BIT(28)
+#define RTASE_TOK5 BIT(26)
+#define RTASE_TOK4 BIT(24)
+#define RTASE_FOVW BIT(6)
+#define RTASE_RDU BIT(4)
+#define RTASE_TOK BIT(2)
+#define RTASE_ROK BIT(0)
+
+ RTASE_IMR1 = 0x0800,
+ RTASE_ISR1 = 0x0802,
+#define RTASE_Q_TOK BIT(4)
+#define RTASE_Q_RDU BIT(1)
+#define RTASE_Q_ROK BIT(0)
+
+ RTASE_EPHY_ISR = 0x6014,
+ RTASE_EPHY_IMR = 0x6016,
+
+ RTASE_TX_CONFIG_0 = 0x0040,
+#define RTASE_TX_INTER_FRAME_GAP_MASK GENMASK(25, 24)
+ /* DMA burst value (0-7) is shift this many bits */
+#define RTASE_TX_DMA_MASK GENMASK(10, 8)
+
+ RTASE_RX_CONFIG_0 = 0x0044,
+#define RTASE_RX_SINGLE_FETCH BIT(14)
+#define RTASE_RX_SINGLE_TAG BIT(13)
+#define RTASE_RX_MX_DMA_MASK GENMASK(10, 8)
+#define RTASE_ACPT_FLOW BIT(7)
+#define RTASE_ACCEPT_ERR BIT(5)
+#define RTASE_ACCEPT_RUNT BIT(4)
+#define RTASE_ACCEPT_BROADCAST BIT(3)
+#define RTASE_ACCEPT_MULTICAST BIT(2)
+#define RTASE_ACCEPT_MYPHYS BIT(1)
+#define RTASE_ACCEPT_ALLPHYS BIT(0)
+#define RTASE_ACCEPT_MASK (RTASE_ACPT_FLOW | RTASE_ACCEPT_ERR | \
+ RTASE_ACCEPT_RUNT | RTASE_ACCEPT_BROADCAST | \
+ RTASE_ACCEPT_MULTICAST | RTASE_ACCEPT_MYPHYS | \
+ RTASE_ACCEPT_ALLPHYS)
+
+ RTASE_RX_CONFIG_1 = 0x0046,
+#define RTASE_RX_MAX_FETCH_DESC_MASK GENMASK(15, 11)
+#define RTASE_RX_NEW_DESC_FORMAT_EN BIT(8)
+#define RTASE_OUTER_VLAN_DETAG_EN BIT(7)
+#define RTASE_INNER_VLAN_DETAG_EN BIT(6)
+#define RTASE_PCIE_NEW_FLOW BIT(2)
+#define RTASE_PCIE_RELOAD_EN BIT(0)
+
+ RTASE_EEM = 0x0050,
+#define RTASE_EEM_UNLOCK 0xC0
+
+ RTASE_TDFNR = 0x0057,
+ RTASE_TPPOLL = 0x0090,
+ RTASE_PDR = 0x00B0,
+ RTASE_FIFOR = 0x00D3,
+#define RTASE_TX_FIFO_EMPTY BIT(5)
+#define RTASE_RX_FIFO_EMPTY BIT(4)
+
+ RTASE_RMS = 0x00DA,
+ RTASE_CPLUS_CMD = 0x00E0,
+#define RTASE_FORCE_RXFLOW_EN BIT(11)
+#define RTASE_FORCE_TXFLOW_EN BIT(10)
+#define RTASE_RX_CHKSUM BIT(5)
+
+ RTASE_Q0_RX_DESC_ADDR0 = 0x00E4,
+ RTASE_Q0_RX_DESC_ADDR4 = 0x00E8,
+ RTASE_Q1_RX_DESC_ADDR0 = 0x4000,
+ RTASE_Q1_RX_DESC_ADDR4 = 0x4004,
+ RTASE_MTPS = 0x00EC,
+#define RTASE_TAG_NUM_SEL_MASK GENMASK(10, 8)
+
+ RTASE_MISC = 0x00F2,
+#define RTASE_RX_DV_GATE_EN BIT(3)
+
+ RTASE_TFUN_CTRL = 0x0400,
+#define RTASE_TX_NEW_DESC_FORMAT_EN BIT(0)
+
+ RTASE_TX_CONFIG_1 = 0x203E,
+#define RTASE_TC_MODE_MASK GENMASK(11, 10)
+
+ RTASE_TOKSEL = 0x2046,
+ RTASE_RFIFONFULL = 0x4406,
+ RTASE_INT_MITI_TX = 0x0A00,
+ RTASE_INT_MITI_RX = 0x0A80,
+
+ RTASE_VLAN_ENTRY_0 = 0xAC80,
+};
+
+enum rtase_desc_status_bit {
+ RTASE_DESC_OWN = BIT(31), /* Descriptor is owned by NIC */
+ RTASE_RING_END = BIT(30), /* End of descriptor ring */
+};
+
+enum rtase_sw_flag_content {
+ RTASE_SWF_MSI_ENABLED = BIT(1),
+ RTASE_SWF_MSIX_ENABLED = BIT(2),
+};
+
+#define RSVD_MASK 0x3FFFC000
+
+struct rtase_tx_desc {
+ __le32 opts1;
+ __le32 opts2;
+ __le64 addr;
+ __le32 opts3;
+ __le32 reserved1;
+ __le32 reserved2;
+ __le32 reserved3;
+} __packed;
+
+/*------ offset 0 of tx descriptor ------*/
+#define RTASE_TX_FIRST_FRAG BIT(29) /* Tx First segment of a packet */
+#define RTASE_TX_LAST_FRAG BIT(28) /* Tx Final segment of a packet */
+#define RTASE_GIANT_SEND_V4 BIT(26) /* TCP Giant Send Offload V4 (GSOv4) */
+#define RTASE_GIANT_SEND_V6 BIT(25) /* TCP Giant Send Offload V6 (GSOv6) */
+#define RTASE_TX_VLAN_TAG BIT(17) /* Add VLAN tag */
+
+/*------ offset 4 of tx descriptor ------*/
+#define RTASE_TX_UDPCS_C BIT(31) /* Calculate UDP/IP checksum */
+#define RTASE_TX_TCPCS_C BIT(30) /* Calculate TCP/IP checksum */
+#define RTASE_TX_IPCS_C BIT(29) /* Calculate IP checksum */
+#define RTASE_TX_IPV6F_C BIT(28) /* Indicate it is an IPv6 packet */
+
+union rtase_rx_desc {
+ struct {
+ __le64 header_buf_addr;
+ __le32 reserved1;
+ __le32 opts_header_len;
+ __le64 addr;
+ __le32 reserved2;
+ __le32 opts1;
+ } __packed desc_cmd;
+
+ struct {
+ __le32 reserved1;
+ __le32 reserved2;
+ __le32 rss;
+ __le32 opts4;
+ __le32 reserved3;
+ __le32 opts3;
+ __le32 opts2;
+ __le32 opts1;
+ } __packed desc_status;
+} __packed;
+
+/*------ offset 28 of rx descriptor ------*/
+#define RTASE_RX_FIRST_FRAG BIT(25) /* Rx First segment of a packet */
+#define RTASE_RX_LAST_FRAG BIT(24) /* Rx Final segment of a packet */
+#define RTASE_RX_RES BIT(20)
+#define RTASE_RX_RUNT BIT(19)
+#define RTASE_RX_RWT BIT(18)
+#define RTASE_RX_CRC BIT(16)
+#define RTASE_RX_V6F BIT(31)
+#define RTASE_RX_V4F BIT(30)
+#define RTASE_RX_UDPT BIT(29)
+#define RTASE_RX_TCPT BIT(28)
+#define RTASE_RX_IPF BIT(26) /* IP checksum failed */
+#define RTASE_RX_UDPF BIT(25) /* UDP/IP checksum failed */
+#define RTASE_RX_TCPF BIT(24) /* TCP/IP checksum failed */
+#define RTASE_RX_VLAN_TAG BIT(16) /* VLAN tag available */
+
+#define RTASE_NUM_DESC 1024
+#define RTASE_TX_BUDGET_DEFAULT 256
+#define RTASE_TX_RING_DESC_SIZE (RTASE_NUM_DESC * sizeof(struct rtase_tx_desc))
+#define RTASE_RX_RING_DESC_SIZE (RTASE_NUM_DESC * sizeof(union rtase_rx_desc))
+#define RTASE_TX_STOP_THRS (MAX_SKB_FRAGS + 1)
+#define RTASE_TX_START_THRS (2 * RTASE_TX_STOP_THRS)
+#define RTASE_VLAN_TAG_MASK GENMASK(15, 0)
+#define RTASE_RX_PKT_SIZE_MASK GENMASK(13, 0)
+
+#define RTASE_IVEC_NAME_SIZE (IFNAMSIZ + 10)
+
+struct rtase_int_vector {
+ struct rtase_private *tp;
+ unsigned int irq;
+ char name[RTASE_IVEC_NAME_SIZE];
+ u16 index;
+ u16 imr_addr;
+ u16 isr_addr;
+ u32 imr;
+ struct list_head ring_list;
+ struct napi_struct napi;
+ int (*poll)(struct napi_struct *napi, int budget);
+};
+
+struct rtase_ring {
+ struct rtase_int_vector *ivec;
+ void *desc;
+ dma_addr_t phy_addr;
+ u32 cur_idx;
+ u32 dirty_idx;
+ u16 index;
+
+ struct sk_buff *skbuff[RTASE_NUM_DESC];
+ union {
+ u32 len[RTASE_NUM_DESC];
+ dma_addr_t data_phy_addr[RTASE_NUM_DESC];
+ } mis;
+
+ struct list_head ring_entry;
+ int (*ring_handler)(struct rtase_ring *ring, int budget);
+};
+
+struct rtase_private {
+ void __iomem *mmio_addr;
+ u32 sw_flag;
+
+ struct pci_dev *pdev;
+ struct net_device *dev;
+ u32 rx_buf_sz;
+
+ struct page_pool *page_pool;
+ struct rtase_ring tx_ring[RTASE_NUM_TX_QUEUE];
+ struct rtase_ring rx_ring[RTASE_NUM_RX_QUEUE];
+ struct rtase_counters *tally_vaddr;
+ dma_addr_t tally_paddr;
+
+ u32 vlan_filter_ctrl;
+ u16 vlan_filter_vid[RTASE_VLAN_FILTER_ENTRY_NUM];
+
+ struct msix_entry msix_entry[RTASE_NUM_MSIX];
+ struct rtase_int_vector int_vector[RTASE_NUM_MSIX];
+
+ u16 tx_queue_ctrl;
+ u16 func_tx_queue_num;
+ u16 func_rx_queue_num;
+ u16 int_nums;
+ u16 tx_int_mit;
+ u16 rx_int_mit;
+};
+
+#define RTASE_LSO_64K 64000
+
+#define RTASE_NIC_MAX_PHYS_BUF_COUNT_LSO2 (16 * 4)
+
+#define RTASE_TCPHO_MASK GENMASK(24, 18)
+
+#define RTASE_MSS_MASK GENMASK(28, 18)
+
+#endif /* _RTASE_H_ */
diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c b/drivers/net/ethernet/realtek/rtase/rtase_main.c
new file mode 100644
index 000000000000..b2749cc72ebe
--- /dev/null
+++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
@@ -0,0 +1,640 @@
+// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause
+/*
+ * rtase is the Linux device driver released for Realtek Automotive Switch
+ * controllers with PCI-Express interface.
+ *
+ * Copyright(c) 2024 Realtek Semiconductor Corp.
+ *
+ * Below is a simplified block diagram of the chip and its relevant interfaces.
+ *
+ * *************************
+ * * *
+ * * CPU network device *
+ * * *
+ * * +-------------+ *
+ * * | PCIE Host | *
+ * ***********++************
+ * ||
+ * PCIE
+ * ||
+ * ********************++**********************
+ * * | PCIE Endpoint | *
+ * * +---------------+ *
+ * * | GMAC | *
+ * * +--++--+ Realtek *
+ * * || RTL90xx Series *
+ * * || *
+ * * +-------------++----------------+ *
+ * * | | MAC | | *
+ * * | +-----+ | *
+ * * | | *
+ * * | Ethernet Switch Core | *
+ * * | | *
+ * * | +-----+ +-----+ | *
+ * * | | MAC |...........| MAC | | *
+ * * +---+-----+-----------+-----+---+ *
+ * * | PHY |...........| PHY | *
+ * * +--++-+ +--++-+ *
+ * *************||****************||***********
+ *
+ * The block of the Realtek RTL90xx series is our entire chip architecture,
+ * the GMAC is connected to the switch core, and there is no PHY in between.
+ * In addition, this driver is mainly used to control GMAC, but does not
+ * control the switch core, so it is not the same as DSA. Linux only plays
+ * the role of a normal leaf node in this model.
+ */
+
+#include <linux/crc32.h>
+#include <linux/dma-mapping.h>
+#include <linux/etherdevice.h>
+#include <linux/if_vlan.h>
+#include <linux/in.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/iopoll.h>
+#include <linux/ip.h>
+#include <linux/ipv6.h>
+#include <linux/mdio.h>
+#include <linux/module.h>
+#include <linux/netdevice.h>
+#include <linux/pci.h>
+#include <linux/pm_runtime.h>
+#include <linux/prefetch.h>
+#include <linux/rtnetlink.h>
+#include <linux/tcp.h>
+#include <asm/irq.h>
+#include <net/ip6_checksum.h>
+#include <net/netdev_queues.h>
+#include <net/page_pool/helpers.h>
+#include <net/pkt_cls.h>
+
+#include "rtase.h"
+
+#define RTK_OPTS1_DEBUG_VALUE 0x0BADBEEF
+#define RTK_MAGIC_NUMBER 0x0BADBADBADBADBAD
+
+static const struct pci_device_id rtase_pci_tbl[] = {
+ {PCI_VDEVICE(REALTEK, 0x906A)},
+ {}
+};
+
+MODULE_DEVICE_TABLE(pci, rtase_pci_tbl);
+
+MODULE_AUTHOR("Realtek ARD Software Team");
+MODULE_DESCRIPTION("Network Driver for the PCIe interface of Realtek Automotive Ethernet Switch");
+MODULE_LICENSE("Dual BSD/GPL");
+
+struct rtase_counters {
+ __le64 tx_packets;
+ __le64 rx_packets;
+ __le64 tx_errors;
+ __le32 rx_errors;
+ __le16 rx_missed;
+ __le16 align_errors;
+ __le32 tx_one_collision;
+ __le32 tx_multi_collision;
+ __le64 rx_unicast;
+ __le64 rx_broadcast;
+ __le32 rx_multicast;
+ __le16 tx_aborted;
+ __le16 tx_underun;
+} __packed;
+
+static void rtase_w8(const struct rtase_private *tp, u16 reg, u8 val8)
+{
+ writeb(val8, tp->mmio_addr + reg);
+}
+
+static void rtase_w16(const struct rtase_private *tp, u16 reg, u16 val16)
+{
+ writew(val16, tp->mmio_addr + reg);
+}
+
+static void rtase_w32(const struct rtase_private *tp, u16 reg, u32 val32)
+{
+ writel(val32, tp->mmio_addr + reg);
+}
+
+static u8 rtase_r8(const struct rtase_private *tp, u16 reg)
+{
+ return readb(tp->mmio_addr + reg);
+}
+
+static u16 rtase_r16(const struct rtase_private *tp, u16 reg)
+{
+ return readw(tp->mmio_addr + reg);
+}
+
+static u32 rtase_r32(const struct rtase_private *tp, u16 reg)
+{
+ return readl(tp->mmio_addr + reg);
+}
+
+static void rtase_tally_counter_clear(const struct rtase_private *tp)
+{
+ u32 cmd = lower_32_bits(tp->tally_paddr);
+
+ rtase_w32(tp, RTASE_DTCCR4, upper_32_bits(tp->tally_paddr));
+ rtase_w32(tp, RTASE_DTCCR0, cmd | RTASE_COUNTER_RESET);
+}
+
+static void rtase_enable_eem_write(const struct rtase_private *tp)
+{
+ u8 val;
+
+ val = rtase_r8(tp, RTASE_EEM);
+ rtase_w8(tp, RTASE_EEM, val | RTASE_EEM_UNLOCK);
+}
+
+static void rtase_disable_eem_write(const struct rtase_private *tp)
+{
+ u8 val;
+
+ val = rtase_r8(tp, RTASE_EEM);
+ rtase_w8(tp, RTASE_EEM, val & ~RTASE_EEM_UNLOCK);
+}
+
+static void rtase_rar_set(const struct rtase_private *tp, const u8 *addr)
+{
+ u32 rar_low, rar_high;
+
+ rar_low = (u32)addr[0] | ((u32)addr[1] << 8) |
+ ((u32)addr[2] << 16) | ((u32)addr[3] << 24);
+
+ rar_high = (u32)addr[4] | ((u32)addr[5] << 8);
+
+ rtase_enable_eem_write(tp);
+ rtase_w32(tp, RTASE_MAC0, rar_low);
+ rtase_w32(tp, RTASE_MAC4, rar_high);
+ rtase_disable_eem_write(tp);
+ rtase_w16(tp, RTASE_LBK_CTRL, RTASE_LBK_ATLD | RTASE_LBK_CLR);
+}
+
+static void rtase_get_mac_address(struct net_device *dev)
+{
+ struct rtase_private *tp = netdev_priv(dev);
+ u8 mac_addr[ETH_ALEN] __aligned(2) = {};
+ u32 i;
+
+ for (i = 0; i < ETH_ALEN; i++)
+ mac_addr[i] = rtase_r8(tp, RTASE_MAC0 + i);
+
+ if (!is_valid_ether_addr(mac_addr)) {
+ eth_hw_addr_random(dev);
+ netdev_warn(dev, "Random ether addr %pM\n", dev->dev_addr);
+ } else {
+ eth_hw_addr_set(dev, mac_addr);
+ ether_addr_copy(dev->perm_addr, dev->dev_addr);
+ }
+
+ rtase_rar_set(tp, dev->dev_addr);
+}
+
+static void rtase_reset_interrupt(struct pci_dev *pdev,
+ const struct rtase_private *tp)
+{
+ if (tp->sw_flag & RTASE_SWF_MSIX_ENABLED)
+ pci_disable_msix(pdev);
+ else
+ pci_disable_msi(pdev);
+}
+
+static int rtase_alloc_msix(struct pci_dev *pdev, struct rtase_private *tp)
+{
+ int ret, irq;
+ u16 i;
+
+ memset(tp->msix_entry, 0x0, RTASE_NUM_MSIX *
+ sizeof(struct msix_entry));
+
+ for (i = 0; i < RTASE_NUM_MSIX; i++)
+ tp->msix_entry[i].entry = i;
+
+ ret = pci_enable_msix_exact(pdev, tp->msix_entry, tp->int_nums);
+
+ if (ret)
+ return ret;
+
+ for (i = 0; i < tp->int_nums; i++) {
+ irq = pci_irq_vector(pdev, i);
+ if (!irq) {
+ pci_disable_msix(pdev);
+ return irq;
+ }
+
+ tp->int_vector[i].irq = irq;
+ }
+
+ return 0;
+}
+
+static int rtase_alloc_interrupt(struct pci_dev *pdev,
+ struct rtase_private *tp)
+{
+ int ret;
+
+ ret = rtase_alloc_msix(pdev, tp);
+ if (ret) {
+ ret = pci_enable_msi(pdev);
+ if (ret) {
+ dev_err(&pdev->dev,
+ "unable to alloc interrupt.(MSI)\n");
+ return ret;
+ }
+
+ tp->sw_flag |= RTASE_SWF_MSI_ENABLED;
+ } else {
+ tp->sw_flag |= RTASE_SWF_MSIX_ENABLED;
+ }
+
+ return 0;
+}
+
+static void rtase_init_hardware(const struct rtase_private *tp)
+{
+ u16 i;
+
+ for (i = 0; i < RTASE_VLAN_FILTER_ENTRY_NUM; i++)
+ rtase_w32(tp, RTASE_VLAN_ENTRY_0 + i * 4, 0);
+}
+
+static void rtase_init_int_vector(struct rtase_private *tp)
+{
+ u16 i;
+
+ /* interrupt vector 0 */
+ tp->int_vector[0].tp = tp;
+ tp->int_vector[0].index = 0;
+ tp->int_vector[0].imr_addr = RTASE_IMR0;
+ tp->int_vector[0].isr_addr = RTASE_ISR0;
+ tp->int_vector[0].imr = RTASE_ROK | RTASE_RDU | RTASE_TOK |
+ RTASE_TOK4 | RTASE_TOK5 | RTASE_TOK6 |
+ RTASE_TOK7;
+ tp->int_vector[0].poll = rtase_poll;
+
+ memset(tp->int_vector[0].name, 0x0, sizeof(tp->int_vector[0].name));
+ INIT_LIST_HEAD(&tp->int_vector[0].ring_list);
+
+ netif_napi_add(tp->dev, &tp->int_vector[0].napi,
+ tp->int_vector[0].poll);
+
+ /* interrupt vector 1 ~ 3 */
+ for (i = 1; i < tp->int_nums; i++) {
+ tp->int_vector[i].tp = tp;
+ tp->int_vector[i].index = i;
+ tp->int_vector[i].imr_addr = RTASE_IMR1 + (i - 1) * 4;
+ tp->int_vector[i].isr_addr = RTASE_ISR1 + (i - 1) * 4;
+ tp->int_vector[i].imr = RTASE_Q_ROK | RTASE_Q_RDU |
+ RTASE_Q_TOK;
+ tp->int_vector[i].poll = rtase_poll;
+
+ memset(tp->int_vector[i].name, 0x0,
+ sizeof(tp->int_vector[0].name));
+ INIT_LIST_HEAD(&tp->int_vector[i].ring_list);
+
+ netif_napi_add(tp->dev, &tp->int_vector[i].napi,
+ tp->int_vector[i].poll);
+ }
+}
+
+static u16 rtase_calc_time_mitigation(u32 time_us)
+{
+ u8 msb, time_count, time_unit;
+ u16 int_miti;
+
+ time_us = min_t(int, time_us, RTASE_MITI_MAX_TIME);
+
+ msb = fls(time_us);
+ if (msb >= RTASE_MITI_COUNT_BIT_NUM) {
+ time_unit = msb - RTASE_MITI_COUNT_BIT_NUM;
+ time_count = time_us >> (msb - RTASE_MITI_COUNT_BIT_NUM);
+ } else {
+ time_unit = 0;
+ time_count = time_us;
+ }
+
+ int_miti = u16_encode_bits(time_count, RTASE_MITI_TIME_COUNT_MASK) |
+ u16_encode_bits(time_unit, RTASE_MITI_TIME_UNIT_MASK);
+
+ return int_miti;
+}
+
+static u16 rtase_calc_packet_num_mitigation(u16 pkt_num)
+{
+ u8 msb, pkt_num_count, pkt_num_unit;
+ u16 int_miti;
+
+ pkt_num = min_t(int, pkt_num, RTASE_MITI_MAX_PKT_NUM);
+
+ if (pkt_num > 60) {
+ pkt_num_unit = RTASE_MITI_MAX_PKT_NUM_IDX;
+ pkt_num_count = pkt_num / RTASE_MITI_MAX_PKT_NUM_UNIT;
+ } else {
+ msb = fls(pkt_num);
+ if (msb >= RTASE_MITI_COUNT_BIT_NUM) {
+ pkt_num_unit = msb - RTASE_MITI_COUNT_BIT_NUM;
+ pkt_num_count = pkt_num >> (msb -
+ RTASE_MITI_COUNT_BIT_NUM);
+ } else {
+ pkt_num_unit = 0;
+ pkt_num_count = pkt_num;
+ }
+ }
+
+ int_miti = u16_encode_bits(pkt_num_count,
+ RTASE_MITI_PKT_NUM_COUNT_MASK) |
+ u16_encode_bits(pkt_num_unit,
+ RTASE_MITI_PKT_NUM_UNIT_MASK);
+
+ return int_miti;
+}
+
+static void rtase_init_software_variable(struct pci_dev *pdev,
+ struct rtase_private *tp)
+{
+ u16 int_miti;
+
+ tp->tx_queue_ctrl = RTASE_TXQ_CTRL;
+ tp->func_tx_queue_num = RTASE_FUNC_TXQ_NUM;
+ tp->func_rx_queue_num = RTASE_FUNC_RXQ_NUM;
+ tp->int_nums = RTASE_INTERRUPT_NUM;
+
+ int_miti = rtase_calc_time_mitigation(RTASE_MITI_DEFAULT_TIME) |
+ rtase_calc_packet_num_mitigation(RTASE_MITI_DEFAULT_PKT_NUM);
+ tp->tx_int_mit = int_miti;
+ tp->rx_int_mit = int_miti;
+
+ tp->sw_flag = 0;
+
+ rtase_init_int_vector(tp);
+
+ /* MTU range: 60 - hw-specific max */
+ tp->dev->min_mtu = ETH_ZLEN;
+ tp->dev->max_mtu = RTASE_MAX_JUMBO_SIZE;
+}
+
+static bool rtase_check_mac_version_valid(struct rtase_private *tp)
+{
+ u32 hw_ver = rtase_r32(tp, RTASE_TX_CONFIG_0) & RTASE_HW_VER_MASK;
+ bool known_ver = false;
+
+ switch (hw_ver) {
+ case 0x00800000:
+ case 0x04000000:
+ case 0x04800000:
+ known_ver = true;
+ break;
+ }
+
+ return known_ver;
+}
+
+static int rtase_init_board(struct pci_dev *pdev, struct net_device **dev_out,
+ void __iomem **ioaddr_out)
+{
+ struct net_device *dev;
+ void __iomem *ioaddr;
+ int ret = -ENOMEM;
+
+ /* dev zeroed in alloc_etherdev */
+ dev = alloc_etherdev_mq(sizeof(struct rtase_private),
+ RTASE_FUNC_TXQ_NUM);
+ if (!dev)
+ goto err_out;
+
+ SET_NETDEV_DEV(dev, &pdev->dev);
+
+ ret = pci_enable_device(pdev);
+ if (ret < 0)
+ goto err_out_free_dev;
+
+ /* make sure PCI base addr 1 is MMIO */
+ if (!(pci_resource_flags(pdev, 2) & IORESOURCE_MEM)) {
+ ret = -ENODEV;
+ goto err_out_disable;
+ }
+
+ /* check for weird/broken PCI region reporting */
+ if (pci_resource_len(pdev, 2) < RTASE_REGS_SIZE) {
+ ret = -ENODEV;
+ goto err_out_disable;
+ }
+
+ ret = pci_request_regions(pdev, KBUILD_MODNAME);
+ if (ret < 0)
+ goto err_out_disable;
+
+ if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64))) {
+ dev_err(&pdev->dev, "no usable dma addressing method\n");
+ goto err_out_free_res;
+ }
+
+ pci_set_master(pdev);
+
+ /* ioremap MMIO region */
+ ioaddr = ioremap(pci_resource_start(pdev, 2),
+ pci_resource_len(pdev, 2));
+ if (!ioaddr) {
+ ret = -EIO;
+ goto err_out_free_res;
+ }
+
+ *ioaddr_out = ioaddr;
+ *dev_out = dev;
+
+ return ret;
+
+err_out_free_res:
+ pci_release_regions(pdev);
+
+err_out_disable:
+ pci_disable_device(pdev);
+
+err_out_free_dev:
+ free_netdev(dev);
+
+err_out:
+ *ioaddr_out = NULL;
+ *dev_out = NULL;
+
+ return ret;
+}
+
+static void rtase_release_board(struct pci_dev *pdev, struct net_device *dev,
+ void __iomem *ioaddr)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+
+ rtase_rar_set(tp, tp->dev->perm_addr);
+ iounmap(ioaddr);
+
+ if ((tp->sw_flag & RTASE_SWF_MSIX_ENABLED))
+ pci_disable_msix(pdev);
+ else
+ pci_disable_msi(pdev);
+
+ pci_release_regions(pdev);
+ pci_disable_device(pdev);
+ free_netdev(dev);
+}
+
+static int rtase_init_one(struct pci_dev *pdev,
+ const struct pci_device_id *ent)
+{
+ struct net_device *dev = NULL;
+ struct rtase_int_vector *ivec;
+ void __iomem *ioaddr = NULL;
+ struct rtase_private *tp;
+ int ret, i;
+
+ if (!pdev->is_physfn && pdev->is_virtfn) {
+ dev_err(&pdev->dev,
+ "This module does not support a virtual function.");
+ return -EINVAL;
+ }
+
+ dev_dbg(&pdev->dev, "Automotive Switch Ethernet driver loaded\n");
+
+ ret = rtase_init_board(pdev, &dev, &ioaddr);
+ if (ret != 0)
+ return ret;
+
+ tp = netdev_priv(dev);
+ tp->mmio_addr = ioaddr;
+ tp->dev = dev;
+ tp->pdev = pdev;
+
+ /* identify chip attached to board */
+ if (!rtase_check_mac_version_valid(tp)) {
+ return dev_err_probe(&pdev->dev, -ENODEV,
+ "unknown chip version, contact rtase "
+ "maintainers (see MAINTAINERS file)\n");
+ }
+
+ rtase_init_software_variable(pdev, tp);
+ rtase_init_hardware(tp);
+
+ ret = rtase_alloc_interrupt(pdev, tp);
+ if (ret < 0) {
+ dev_err(&pdev->dev, "unable to alloc MSIX/MSI\n");
+ goto err_out_1;
+ }
+
+ rtase_init_netdev_ops(dev);
+
+ dev->pcpu_stat_type = NETDEV_PCPU_STAT_TSTATS;
+
+ dev->features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_CTAG_RX;
+
+ dev->features |= NETIF_F_IP_CSUM;
+ dev->features |= NETIF_F_HIGHDMA;
+ dev->features |= NETIF_F_RXCSUM | NETIF_F_SG | NETIF_F_TSO;
+ dev->features |= NETIF_F_IPV6_CSUM | NETIF_F_TSO6;
+ dev->hw_features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO |
+ NETIF_F_RXCSUM | NETIF_F_HW_VLAN_CTAG_TX |
+ NETIF_F_HW_VLAN_CTAG_RX;
+ dev->hw_features |= NETIF_F_RXALL;
+ dev->hw_features |= NETIF_F_RXFCS;
+ dev->hw_features |= NETIF_F_IPV6_CSUM | NETIF_F_TSO6;
+ dev->vlan_features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO |
+ NETIF_F_HIGHDMA;
+ dev->priv_flags |= IFF_LIVE_ADDR_CHANGE;
+ netif_set_tso_max_size(dev, RTASE_LSO_64K);
+ netif_set_tso_max_segs(dev, RTASE_NIC_MAX_PHYS_BUF_COUNT_LSO2);
+
+ rtase_get_mac_address(dev);
+
+ tp->tally_vaddr = dma_alloc_coherent(&pdev->dev,
+ sizeof(*tp->tally_vaddr),
+ &tp->tally_paddr,
+ GFP_KERNEL);
+ if (!tp->tally_vaddr) {
+ ret = -ENOMEM;
+ goto err_out;
+ }
+
+ rtase_tally_counter_clear(tp);
+
+ pci_set_drvdata(pdev, dev);
+
+ netif_carrier_off(dev);
+
+ ret = register_netdev(dev);
+ if (ret != 0)
+ goto err_out;
+
+ netdev_dbg(dev, "%pM, IRQ %d\n", dev->dev_addr, dev->irq);
+
+ return 0;
+
+err_out:
+ if (tp->tally_vaddr) {
+ dma_free_coherent(&pdev->dev,
+ sizeof(*tp->tally_vaddr),
+ tp->tally_vaddr,
+ tp->tally_paddr);
+
+ tp->tally_vaddr = NULL;
+ }
+
+err_out_1:
+ for (i = 0; i < tp->int_nums; i++) {
+ ivec = &tp->int_vector[i];
+ netif_napi_del(&ivec->napi);
+ }
+
+ rtase_release_board(pdev, dev, ioaddr);
+
+ return ret;
+}
+
+static void rtase_remove_one(struct pci_dev *pdev)
+{
+ struct net_device *dev = pci_get_drvdata(pdev);
+ struct rtase_private *tp = netdev_priv(dev);
+ struct rtase_int_vector *ivec;
+ u32 i;
+
+ unregister_netdev(dev);
+
+ for (i = 0; i < tp->int_nums; i++) {
+ ivec = &tp->int_vector[i];
+ netif_napi_del(&ivec->napi);
+ }
+
+ rtase_reset_interrupt(pdev, tp);
+ if (tp->tally_vaddr) {
+ dma_free_coherent(&pdev->dev,
+ sizeof(*tp->tally_vaddr),
+ tp->tally_vaddr,
+ tp->tally_paddr);
+ tp->tally_vaddr = NULL;
+ }
+
+ rtase_release_board(pdev, dev, tp->mmio_addr);
+ pci_set_drvdata(pdev, NULL);
+}
+
+static void rtase_shutdown(struct pci_dev *pdev)
+{
+ struct net_device *dev = pci_get_drvdata(pdev);
+ const struct rtase_private *tp;
+
+ tp = netdev_priv(dev);
+
+ if (netif_running(dev))
+ rtase_close(dev);
+
+ rtase_reset_interrupt(pdev, tp);
+}
+
+static struct pci_driver rtase_pci_driver = {
+ .name = KBUILD_MODNAME,
+ .id_table = rtase_pci_tbl,
+ .probe = rtase_init_one,
+ .remove = rtase_remove_one,
+ .shutdown = rtase_shutdown,
+};
+
+module_pci_driver(rtase_pci_driver);
--
2.34.1
Implement the .ndo_open function to set default hardware settings
and initialize the descriptor ring and interrupts. Among them,
when requesting irq, because the first group of interrupts needs to
process more events, the overall structure will be different from
other groups of interrupts, so it needs to be processed separately.
Signed-off-by: Justin Lai <[email protected]>
---
.../net/ethernet/realtek/rtase/rtase_main.c | 425 ++++++++++++++++++
1 file changed, 425 insertions(+)
diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c b/drivers/net/ethernet/realtek/rtase/rtase_main.c
index b2749cc72ebe..80ee963971c0 100644
--- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
+++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
@@ -131,6 +131,297 @@ static u32 rtase_r32(const struct rtase_private *tp, u16 reg)
return readl(tp->mmio_addr + reg);
}
+static int rtase_alloc_desc(struct rtase_private *tp)
+{
+ struct pci_dev *pdev = tp->pdev;
+ u32 i;
+
+ /* rx and tx descriptors needs 256 bytes alignment.
+ * dma_alloc_coherent provides more.
+ */
+ for (i = 0; i < tp->func_tx_queue_num; i++) {
+ tp->tx_ring[i].desc =
+ dma_alloc_coherent(&pdev->dev,
+ RTASE_TX_RING_DESC_SIZE,
+ &tp->tx_ring[i].phy_addr,
+ GFP_KERNEL);
+ if (!tp->tx_ring[i].desc) {
+ netdev_err(tp->dev, "Failed to allocate dma memory of "
+ "tx descriptor.\n");
+ return -ENOMEM;
+ }
+ }
+
+ for (i = 0; i < tp->func_rx_queue_num; i++) {
+ tp->rx_ring[i].desc =
+ dma_alloc_coherent(&pdev->dev,
+ RTASE_RX_RING_DESC_SIZE,
+ &tp->rx_ring[i].phy_addr,
+ GFP_KERNEL);
+ if (!tp->rx_ring[i].desc) {
+ netdev_err(tp->dev, "Failed to allocate dma memory of "
+ "rx descriptor.\n");
+ return -ENOMEM;
+ }
+ }
+
+ return 0;
+}
+
+static void rtase_free_desc(struct rtase_private *tp)
+{
+ struct pci_dev *pdev = tp->pdev;
+ u32 i;
+
+ for (i = 0; i < tp->func_tx_queue_num; i++) {
+ if (!tp->tx_ring[i].desc)
+ continue;
+
+ dma_free_coherent(&pdev->dev, RTASE_TX_RING_DESC_SIZE,
+ tp->tx_ring[i].desc,
+ tp->tx_ring[i].phy_addr);
+ tp->tx_ring[i].desc = NULL;
+ }
+
+ for (i = 0; i < tp->func_rx_queue_num; i++) {
+ if (!tp->rx_ring[i].desc)
+ continue;
+
+ dma_free_coherent(&pdev->dev, RTASE_RX_RING_DESC_SIZE,
+ tp->rx_ring[i].desc,
+ tp->rx_ring[i].phy_addr);
+ tp->rx_ring[i].desc = NULL;
+ }
+}
+
+static void rtase_mark_to_asic(union rtase_rx_desc *desc, u32 rx_buf_sz)
+{
+ u32 eor = le32_to_cpu(desc->desc_cmd.opts1) & RTASE_RING_END;
+
+ desc->desc_status.opts2 = 0;
+ /* force memory writes to complete before releasing descriptor */
+ dma_wmb();
+ WRITE_ONCE(desc->desc_cmd.opts1,
+ cpu_to_le32(RTASE_DESC_OWN | eor | rx_buf_sz));
+}
+
+static void rtase_tx_desc_init(struct rtase_private *tp, u16 idx)
+{
+ struct rtase_ring *ring = &tp->tx_ring[idx];
+ struct rtase_tx_desc *desc;
+ u32 i;
+
+ memset(ring->desc, 0x0, RTASE_TX_RING_DESC_SIZE);
+ memset(ring->skbuff, 0x0, sizeof(ring->skbuff));
+ ring->cur_idx = 0;
+ ring->dirty_idx = 0;
+ ring->index = idx;
+
+ for (i = 0; i < RTASE_NUM_DESC; i++) {
+ ring->mis.len[i] = 0;
+ if ((RTASE_NUM_DESC - 1) == i) {
+ desc = ring->desc + sizeof(struct rtase_tx_desc) * i;
+ desc->opts1 = cpu_to_le32(RTASE_RING_END);
+ }
+ }
+
+ ring->ring_handler = tx_handler;
+ if (idx < 4) {
+ ring->ivec = &tp->int_vector[idx];
+ list_add_tail(&ring->ring_entry,
+ &tp->int_vector[idx].ring_list);
+ } else {
+ ring->ivec = &tp->int_vector[0];
+ list_add_tail(&ring->ring_entry, &tp->int_vector[0].ring_list);
+ }
+}
+
+static void rtase_map_to_asic(union rtase_rx_desc *desc, dma_addr_t mapping,
+ u32 rx_buf_sz)
+{
+ desc->desc_cmd.addr = cpu_to_le64(mapping);
+
+ rtase_mark_to_asic(desc, rx_buf_sz);
+}
+
+static void rtase_make_unusable_by_asic(union rtase_rx_desc *desc)
+{
+ desc->desc_cmd.addr = cpu_to_le64(RTK_MAGIC_NUMBER);
+ desc->desc_cmd.opts1 &= ~cpu_to_le32(RTASE_DESC_OWN | RSVD_MASK);
+}
+
+static int rtase_alloc_rx_skb(const struct rtase_ring *ring,
+ struct sk_buff **p_sk_buff,
+ union rtase_rx_desc *desc,
+ dma_addr_t *rx_phy_addr, u8 in_intr)
+{
+ struct rtase_int_vector *ivec = ring->ivec;
+ const struct rtase_private *tp = ivec->tp;
+ struct sk_buff *skb = NULL;
+ dma_addr_t mapping;
+ struct page *page;
+ void *buf_addr;
+ int ret = 0;
+
+ page = page_pool_dev_alloc_pages(tp->page_pool);
+ if (!page) {
+ netdev_err(tp->dev, "failed to alloc page\n");
+ goto err_out;
+ }
+
+ buf_addr = page_address(page);
+ mapping = page_pool_get_dma_addr(page);
+
+ skb = build_skb(buf_addr, PAGE_SIZE);
+ if (!skb) {
+ page_pool_put_full_page(tp->page_pool, page, true);
+ netdev_err(tp->dev, "failed to build skb\n");
+ goto err_out;
+ }
+
+ *p_sk_buff = skb;
+ *rx_phy_addr = mapping;
+ rtase_map_to_asic(desc, mapping, tp->rx_buf_sz);
+
+ return ret;
+
+err_out:
+ if (skb)
+ dev_kfree_skb(skb);
+
+ ret = -ENOMEM;
+ rtase_make_unusable_by_asic(desc);
+
+ return ret;
+}
+
+static u32 rtase_rx_ring_fill(struct rtase_ring *ring, u32 ring_start,
+ u32 ring_end, u8 in_intr)
+{
+ union rtase_rx_desc *desc_base = ring->desc;
+ u32 cur;
+
+ for (cur = ring_start; ring_end - cur > 0; cur++) {
+ u32 i = cur % RTASE_NUM_DESC;
+ union rtase_rx_desc *desc = desc_base + i;
+ int ret;
+
+ if (ring->skbuff[i])
+ continue;
+
+ ret = rtase_alloc_rx_skb(ring, &ring->skbuff[i], desc,
+ &ring->mis.data_phy_addr[i],
+ in_intr);
+ if (ret)
+ break;
+ }
+
+ return cur - ring_start;
+}
+
+static void rtase_mark_as_last_descriptor(union rtase_rx_desc *desc)
+{
+ desc->desc_cmd.opts1 |= cpu_to_le32(RTASE_RING_END);
+}
+
+static void rtase_rx_ring_clear(struct rtase_ring *ring)
+{
+ union rtase_rx_desc *desc;
+ u32 i;
+
+ for (i = 0; i < RTASE_NUM_DESC; i++) {
+ desc = ring->desc + sizeof(union rtase_rx_desc) * i;
+
+ if (!ring->skbuff[i])
+ continue;
+
+ skb_mark_for_recycle(ring->skbuff[i]);
+
+ dev_kfree_skb(ring->skbuff[i]);
+
+ ring->skbuff[i] = NULL;
+
+ rtase_make_unusable_by_asic(desc);
+ }
+}
+
+static void rtase_rx_desc_init(struct rtase_private *tp, u16 idx)
+{
+ struct rtase_ring *ring = &tp->rx_ring[idx];
+ u16 i;
+
+ memset(ring->desc, 0x0, RTASE_RX_RING_DESC_SIZE);
+ memset(ring->skbuff, 0x0, sizeof(ring->skbuff));
+ ring->cur_idx = 0;
+ ring->dirty_idx = 0;
+ ring->index = idx;
+
+ for (i = 0; i < RTASE_NUM_DESC; i++)
+ ring->mis.data_phy_addr[i] = 0;
+
+ ring->ring_handler = rx_handler;
+ ring->ivec = &tp->int_vector[idx];
+ list_add_tail(&ring->ring_entry, &tp->int_vector[idx].ring_list);
+}
+
+static void rtase_rx_clear(struct rtase_private *tp)
+{
+ u32 i;
+
+ for (i = 0; i < tp->func_rx_queue_num; i++)
+ rtase_rx_ring_clear(&tp->rx_ring[i]);
+
+ page_pool_destroy(tp->page_pool);
+ tp->page_pool = NULL;
+}
+
+static int rtase_init_ring(const struct net_device *dev)
+{
+ struct rtase_private *tp = netdev_priv(dev);
+ struct page_pool_params pp_params = { 0 };
+ struct page_pool *page_pool;
+ u32 num;
+ u16 i;
+
+ pp_params.flags = PP_FLAG_DMA_MAP | PP_FLAG_DMA_SYNC_DEV;
+ pp_params.order = 0;
+ pp_params.pool_size = RTASE_NUM_DESC * tp->func_rx_queue_num;
+ pp_params.nid = dev_to_node(&tp->pdev->dev);
+ pp_params.dev = &tp->pdev->dev;
+ pp_params.dma_dir = DMA_FROM_DEVICE;
+ pp_params.max_len = PAGE_SIZE;
+ pp_params.offset = 0;
+
+ page_pool = page_pool_create(&pp_params);
+ if (IS_ERR(page_pool)) {
+ netdev_err(tp->dev, "failed to create page pool\n");
+ return -ENOMEM;
+ }
+
+ tp->page_pool = page_pool;
+
+ for (i = 0; i < tp->func_tx_queue_num; i++)
+ rtase_tx_desc_init(tp, i);
+
+ for (i = 0; i < tp->func_rx_queue_num; i++) {
+ rtase_rx_desc_init(tp, i);
+ num = rtase_rx_ring_fill(&tp->rx_ring[i], 0,
+ RTASE_NUM_DESC, 0);
+ if (num != RTASE_NUM_DESC)
+ goto err_out;
+
+ rtase_mark_as_last_descriptor(tp->rx_ring[i].desc +
+ sizeof(union rtase_rx_desc) *
+ (RTASE_NUM_DESC - 1));
+ }
+
+ return 0;
+
+err_out:
+ rtase_rx_clear(tp);
+ return -ENOMEM;
+}
+
static void rtase_tally_counter_clear(const struct rtase_private *tp)
{
u32 cmd = lower_32_bits(tp->tally_paddr);
@@ -139,6 +430,130 @@ static void rtase_tally_counter_clear(const struct rtase_private *tp)
rtase_w32(tp, RTASE_DTCCR0, cmd | RTASE_COUNTER_RESET);
}
+static void rtase_nic_enable(const struct net_device *dev)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+ u16 rcr = rtase_r16(tp, RTASE_RX_CONFIG_1);
+ u8 val;
+
+ rtase_w16(tp, RTASE_RX_CONFIG_1, rcr & ~RTASE_PCIE_RELOAD_EN);
+ rtase_w16(tp, RTASE_RX_CONFIG_1, rcr | RTASE_PCIE_RELOAD_EN);
+
+ val = rtase_r8(tp, RTASE_CHIP_CMD);
+ rtase_w8(tp, RTASE_CHIP_CMD, val | RTASE_TE | RTASE_RE);
+
+ val = rtase_r8(tp, RTASE_MISC);
+ rtase_w8(tp, RTASE_MISC, val & ~RTASE_RX_DV_GATE_EN);
+}
+
+static void rtase_enable_hw_interrupt(const struct rtase_private *tp)
+{
+ const struct rtase_int_vector *ivec = &tp->int_vector[0];
+ u32 i;
+
+ rtase_w32(tp, ivec->imr_addr, ivec->imr);
+
+ for (i = 1; i < tp->int_nums; i++) {
+ ivec = &tp->int_vector[i];
+ rtase_w16(tp, ivec->imr_addr, ivec->imr);
+ }
+}
+
+static void rtase_hw_start(const struct net_device *dev)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+
+ rtase_nic_enable(dev);
+ rtase_enable_hw_interrupt(tp);
+}
+
+static int rtase_open(struct net_device *dev)
+{
+ struct rtase_private *tp = netdev_priv(dev);
+ const struct pci_dev *pdev = tp->pdev;
+ struct rtase_int_vector *ivec;
+ u16 i = 0, j;
+ int ret;
+
+ ivec = &tp->int_vector[0];
+ tp->rx_buf_sz = RTASE_RX_BUF_SIZE;
+
+ ret = rtase_alloc_desc(tp);
+ if (ret)
+ goto err_free_all_allocated_mem;
+
+ ret = rtase_init_ring(dev);
+ if (ret)
+ goto err_free_all_allocated_mem;
+
+ rtase_hw_config(dev);
+
+ if (tp->sw_flag & RTASE_SWF_MSIX_ENABLED) {
+ ret = request_irq(ivec->irq, rtase_interrupt, 0,
+ dev->name, ivec);
+ if (ret)
+ goto err_free_all_allocated_irq;
+
+ /* request other interrupts to handle multiqueue */
+ for (i = 1; i < tp->int_nums; i++) {
+ ivec = &tp->int_vector[i];
+ snprintf(ivec->name, sizeof(ivec->name), "%s_int%i",
+ tp->dev->name, i);
+ ret = request_irq(ivec->irq, rtase_q_interrupt, 0,
+ ivec->name, ivec);
+ if (ret)
+ goto err_free_all_allocated_irq;
+ }
+ } else {
+ ret = request_irq(pdev->irq, rtase_interrupt, 0, dev->name,
+ ivec);
+ if (ret)
+ goto err_free_all_allocated_mem;
+ }
+
+ rtase_hw_start(dev);
+
+ for (i = 0; i < tp->int_nums; i++) {
+ ivec = &tp->int_vector[i];
+ napi_enable(&ivec->napi);
+ }
+
+ netif_carrier_on(dev);
+ netif_wake_queue(dev);
+
+ return 0;
+
+err_free_all_allocated_irq:
+ for (j = 0; j < i; j++)
+ free_irq(tp->int_vector[j].irq, &tp->int_vector[j]);
+
+err_free_all_allocated_mem:
+ rtase_free_desc(tp);
+
+ return ret;
+}
+
+static int rtase_close(struct net_device *dev)
+{
+ struct rtase_private *tp = netdev_priv(dev);
+ const struct pci_dev *pdev = tp->pdev;
+ u32 i;
+
+ rtase_down(dev);
+
+ if (tp->sw_flag & RTASE_SWF_MSIX_ENABLED) {
+ for (i = 0; i < tp->int_nums; i++)
+ free_irq(tp->int_vector[i].irq, &tp->int_vector[i]);
+
+ } else {
+ free_irq(pdev->irq, &tp->int_vector[0]);
+ }
+
+ rtase_free_desc(tp);
+
+ return 0;
+}
+
static void rtase_enable_eem_write(const struct rtase_private *tp)
{
u8 val;
@@ -171,6 +586,11 @@ static void rtase_rar_set(const struct rtase_private *tp, const u8 *addr)
rtase_w16(tp, RTASE_LBK_CTRL, RTASE_LBK_ATLD | RTASE_LBK_CLR);
}
+static const struct net_device_ops rtase_netdev_ops = {
+ .ndo_open = rtase_open,
+ .ndo_stop = rtase_close,
+};
+
static void rtase_get_mac_address(struct net_device *dev)
{
struct rtase_private *tp = netdev_priv(dev);
@@ -191,6 +611,11 @@ static void rtase_get_mac_address(struct net_device *dev)
rtase_rar_set(tp, dev->dev_addr);
}
+static void rtase_init_netdev_ops(struct net_device *dev)
+{
+ dev->netdev_ops = &rtase_netdev_ops;
+}
+
static void rtase_reset_interrupt(struct pci_dev *pdev,
const struct rtase_private *tp)
{
--
2.34.1
Implement the rtase_down function to disable hardware setting
and interrupt and clear descriptor ring.
Signed-off-by: Justin Lai <[email protected]>
---
.../net/ethernet/realtek/rtase/rtase_main.c | 152 ++++++++++++++++++
1 file changed, 152 insertions(+)
diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c b/drivers/net/ethernet/realtek/rtase/rtase_main.c
index 80ee963971c0..caff2cfe78c7 100644
--- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
+++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
@@ -194,6 +194,57 @@ static void rtase_free_desc(struct rtase_private *tp)
}
}
+static void rtase_unmap_tx_skb(struct pci_dev *pdev, u32 len,
+ struct rtase_tx_desc *desc)
+{
+ dma_unmap_single(&pdev->dev, le64_to_cpu(desc->addr), len,
+ DMA_TO_DEVICE);
+ desc->opts1 = cpu_to_le32(RTK_OPTS1_DEBUG_VALUE);
+ desc->opts2 = 0x00;
+ desc->addr = cpu_to_le64(RTK_MAGIC_NUMBER);
+}
+
+static void rtase_tx_clear_range(struct rtase_ring *ring, u32 start, u32 n)
+{
+ const struct rtase_private *tp = ring->ivec->tp;
+ struct rtase_tx_desc *desc_base = ring->desc;
+ struct net_device *dev = tp->dev;
+ u32 i;
+
+ for (i = 0; i < n; i++) {
+ u32 entry = (start + i) % RTASE_NUM_DESC;
+ struct rtase_tx_desc *desc = desc_base + entry;
+ u32 len = ring->mis.len[entry];
+ struct sk_buff *skb;
+
+ if (len == 0)
+ continue;
+
+ rtase_unmap_tx_skb(tp->pdev, len, desc);
+ ring->mis.len[entry] = 0;
+ skb = ring->skbuff[entry];
+ if (!skb)
+ continue;
+
+ dev->stats.tx_dropped++;
+ dev_kfree_skb_any(skb);
+ ring->skbuff[entry] = NULL;
+ }
+}
+
+static void rtase_tx_clear(struct rtase_private *tp)
+{
+ struct rtase_ring *ring;
+ u16 i;
+
+ for (i = 0; i < tp->func_tx_queue_num; i++) {
+ ring = &tp->tx_ring[i];
+ rtase_tx_clear_range(ring, ring->dirty_idx, RTASE_NUM_DESC);
+ ring->cur_idx = 0;
+ ring->dirty_idx = 0;
+ }
+}
+
static void rtase_mark_to_asic(union rtase_rx_desc *desc, u32 rx_buf_sz)
{
u32 eor = le32_to_cpu(desc->desc_cmd.opts1) & RTASE_RING_END;
@@ -430,6 +481,80 @@ static void rtase_tally_counter_clear(const struct rtase_private *tp)
rtase_w32(tp, RTASE_DTCCR0, cmd | RTASE_COUNTER_RESET);
}
+static void rtase_irq_dis_and_clear(const struct rtase_private *tp)
+{
+ const struct rtase_int_vector *ivec = &tp->int_vector[0];
+ u32 val1;
+ u16 val2;
+ u8 i;
+
+ rtase_w32(tp, ivec->imr_addr, 0);
+ val1 = rtase_r32(tp, ivec->isr_addr);
+ rtase_w32(tp, ivec->isr_addr, val1);
+
+ for (i = 1; i < tp->int_nums; i++) {
+ ivec = &tp->int_vector[i];
+ rtase_w16(tp, ivec->imr_addr, 0);
+ val2 = rtase_r16(tp, ivec->isr_addr);
+ rtase_w16(tp, ivec->isr_addr, val2);
+ }
+}
+
+static void rtase_poll_timeout(const struct rtase_private *tp, u32 cond,
+ u32 sleep_us, u64 timeout_us, u16 reg)
+{
+ int err;
+ u8 val;
+
+ err = read_poll_timeout(rtase_r8, val, val & cond, sleep_us,
+ timeout_us, false, tp, reg);
+
+ if (err == -ETIMEDOUT)
+ netdev_err(tp->dev, "poll reg 0x00%x timeout\n", reg);
+}
+
+static void rtase_nic_reset(const struct net_device *dev)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+ u16 rx_config;
+ u8 val;
+
+ rx_config = rtase_r16(tp, RTASE_RX_CONFIG_0);
+ rtase_w16(tp, RTASE_RX_CONFIG_0, rx_config & ~RTASE_ACCEPT_MASK);
+
+ val = rtase_r8(tp, RTASE_MISC);
+ rtase_w8(tp, RTASE_MISC, val | RTASE_RX_DV_GATE_EN);
+
+ val = rtase_r8(tp, RTASE_CHIP_CMD);
+ rtase_w8(tp, RTASE_CHIP_CMD, val | RTASE_STOP_REQ);
+ mdelay(2);
+
+ rtase_poll_timeout(tp, RTASE_STOP_REQ_DONE, 100, 150000,
+ RTASE_CHIP_CMD);
+
+ rtase_poll_timeout(tp, RTASE_TX_FIFO_EMPTY, 100, 100000,
+ RTASE_FIFOR);
+
+ rtase_poll_timeout(tp, RTASE_RX_FIFO_EMPTY, 100, 100000,
+ RTASE_FIFOR);
+
+ val = rtase_r8(tp, RTASE_CHIP_CMD);
+ rtase_w8(tp, RTASE_CHIP_CMD, val & ~(RTASE_TE | RTASE_RE));
+ val = rtase_r8(tp, RTASE_CHIP_CMD);
+ rtase_w8(tp, RTASE_CHIP_CMD, val & ~RTASE_STOP_REQ);
+
+ rtase_w16(tp, RTASE_RX_CONFIG_0, rx_config);
+}
+
+static void rtase_hw_reset(const struct net_device *dev)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+
+ rtase_irq_dis_and_clear(tp);
+
+ rtase_nic_reset(dev);
+}
+
static void rtase_nic_enable(const struct net_device *dev)
{
const struct rtase_private *tp = netdev_priv(dev);
@@ -533,6 +658,33 @@ static int rtase_open(struct net_device *dev)
return ret;
}
+static void rtase_down(struct net_device *dev)
+{
+ struct rtase_private *tp = netdev_priv(dev);
+ struct rtase_int_vector *ivec;
+ struct rtase_ring *ring, *tmp;
+ u32 i;
+
+ for (i = 0; i < tp->int_nums; i++) {
+ ivec = &tp->int_vector[i];
+ napi_disable(&ivec->napi);
+ list_for_each_entry_safe(ring, tmp, &ivec->ring_list,
+ ring_entry) {
+ list_del(&ring->ring_entry);
+ }
+ }
+
+ netif_tx_disable(dev);
+
+ netif_carrier_off(dev);
+
+ rtase_hw_reset(dev);
+
+ rtase_tx_clear(tp);
+
+ rtase_rx_clear(tp);
+}
+
static int rtase_close(struct net_device *dev)
{
struct rtase_private *tp = netdev_priv(dev);
--
2.34.1
Implement rtase_hw_config to set default hardware settings, including
setting interrupt mitigation, tx/rx DMA burst, interframe gap time,
rx packet filter, near fifo threshold and fill descriptor ring and
tally counter address, and enable flow control. When filling the
rx descriptor ring, the first group of queues needs to be processed
separately because the positions of the first group of queues are not
regular with other subsequent groups. The other queues are all newly
added features, but we want to retain the original design. So they were
not put together.
Signed-off-by: Justin Lai <[email protected]>
---
.../net/ethernet/realtek/rtase/rtase_main.c | 238 ++++++++++++++++++
1 file changed, 238 insertions(+)
diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c b/drivers/net/ethernet/realtek/rtase/rtase_main.c
index f971110027ee..23406c195cff 100644
--- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
+++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
@@ -473,6 +473,23 @@ static int rtase_init_ring(const struct net_device *dev)
return -ENOMEM;
}
+static void rtase_interrupt_mitigation(const struct rtase_private *tp)
+{
+ u32 i;
+
+ for (i = 0; i < tp->func_tx_queue_num; i++)
+ rtase_w16(tp, RTASE_INT_MITI_TX + i * 2, tp->tx_int_mit);
+
+ for (i = 0; i < tp->func_rx_queue_num; i++)
+ rtase_w16(tp, RTASE_INT_MITI_RX + i * 2, tp->rx_int_mit);
+}
+
+static void rtase_tally_counter_addr_fill(const struct rtase_private *tp)
+{
+ rtase_w32(tp, RTASE_DTCCR4, upper_32_bits(tp->tally_paddr));
+ rtase_w32(tp, RTASE_DTCCR0, lower_32_bits(tp->tally_paddr));
+}
+
static void rtase_tally_counter_clear(const struct rtase_private *tp)
{
u32 cmd = lower_32_bits(tp->tally_paddr);
@@ -481,6 +498,119 @@ static void rtase_tally_counter_clear(const struct rtase_private *tp)
rtase_w32(tp, RTASE_DTCCR0, cmd | RTASE_COUNTER_RESET);
}
+static void rtase_desc_addr_fill(const struct rtase_private *tp)
+{
+ const struct rtase_ring *ring;
+ u16 i, cmd, val;
+ int err;
+
+ for (i = 0; i < tp->func_tx_queue_num; i++) {
+ ring = &tp->tx_ring[i];
+
+ rtase_w32(tp, RTASE_TX_DESC_ADDR0,
+ lower_32_bits(ring->phy_addr));
+ rtase_w32(tp, RTASE_TX_DESC_ADDR4,
+ upper_32_bits(ring->phy_addr));
+
+ cmd = i | RTASE_TX_DESC_CMD_WE | RTASE_TX_DESC_CMD_CS;
+ rtase_w16(tp, RTASE_TX_DESC_COMMAND, cmd);
+
+ err = read_poll_timeout(rtase_r16, val,
+ !(val & RTASE_TX_DESC_CMD_CS), 10,
+ 1000, false, tp,
+ RTASE_TX_DESC_COMMAND);
+
+ if (err == -ETIMEDOUT)
+ netdev_err(tp->dev,
+ "error occurred in fill tx descriptor\n");
+ }
+
+ for (i = 0; i < tp->func_rx_queue_num; i++) {
+ ring = &tp->rx_ring[i];
+
+ if (i == 0) {
+ rtase_w32(tp, RTASE_Q0_RX_DESC_ADDR0,
+ lower_32_bits(ring->phy_addr));
+ rtase_w32(tp, RTASE_Q0_RX_DESC_ADDR4,
+ upper_32_bits(ring->phy_addr));
+ } else {
+ rtase_w32(tp, (RTASE_Q1_RX_DESC_ADDR0 + ((i - 1) * 8)),
+ lower_32_bits(ring->phy_addr));
+ rtase_w32(tp, (RTASE_Q1_RX_DESC_ADDR4 + ((i - 1) * 8)),
+ upper_32_bits(ring->phy_addr));
+ }
+ }
+}
+
+static void rtase_hw_set_features(const struct net_device *dev,
+ netdev_features_t features)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+ u16 rx_config, val;
+
+ rx_config = rtase_r16(tp, RTASE_RX_CONFIG_0);
+ if (features & NETIF_F_RXALL)
+ rx_config |= (RTASE_ACCEPT_ERR | RTASE_ACCEPT_RUNT);
+ else
+ rx_config &= ~(RTASE_ACCEPT_ERR | RTASE_ACCEPT_RUNT);
+
+ rtase_w16(tp, RTASE_RX_CONFIG_0, rx_config);
+
+ val = rtase_r16(tp, RTASE_CPLUS_CMD);
+ if (features & NETIF_F_RXCSUM)
+ rtase_w16(tp, RTASE_CPLUS_CMD, val | RTASE_RX_CHKSUM);
+ else
+ rtase_w16(tp, RTASE_CPLUS_CMD, val & ~RTASE_RX_CHKSUM);
+
+ rx_config = rtase_r16(tp, RTASE_RX_CONFIG_1);
+ if (dev->features & NETIF_F_HW_VLAN_CTAG_RX)
+ rx_config |= (RTASE_INNER_VLAN_DETAG_EN |
+ RTASE_OUTER_VLAN_DETAG_EN);
+ else
+ rx_config &= ~(RTASE_INNER_VLAN_DETAG_EN |
+ RTASE_OUTER_VLAN_DETAG_EN);
+
+ rtase_w16(tp, RTASE_RX_CONFIG_1, rx_config);
+}
+
+static void rtase_hw_set_rx_packet_filter(struct net_device *dev)
+{
+ u32 mc_filter[2] = { 0xFFFFFFFF, 0xFFFFFFFF };
+ struct rtase_private *tp = netdev_priv(dev);
+ u16 rx_mode;
+
+ rx_mode = rtase_r16(tp, RTASE_RX_CONFIG_0) & ~RTASE_ACCEPT_MASK;
+ rx_mode |= RTASE_ACCEPT_BROADCAST | RTASE_ACCEPT_MYPHYS;
+
+ if (dev->flags & IFF_PROMISC) {
+ rx_mode |= RTASE_ACCEPT_MULTICAST | RTASE_ACCEPT_ALLPHYS;
+ } else if (dev->flags & IFF_ALLMULTI) {
+ rx_mode |= RTASE_ACCEPT_MULTICAST;
+ } else {
+ struct netdev_hw_addr *hw_addr;
+
+ mc_filter[0] = 0;
+ mc_filter[1] = 0;
+
+ netdev_for_each_mc_addr(hw_addr, dev) {
+ u32 bit_nr = eth_hw_addr_crc(hw_addr);
+ u32 idx = u32_get_bits(bit_nr, BIT(31));
+ u32 bit = u32_get_bits(bit_nr,
+ RTASE_MULTICAST_FILTER_MASK);
+
+ mc_filter[idx] |= BIT(bit);
+ rx_mode |= RTASE_ACCEPT_MULTICAST;
+ }
+ }
+
+ if (dev->features & NETIF_F_RXALL)
+ rx_mode |= RTASE_ACCEPT_ERR | RTASE_ACCEPT_RUNT;
+
+ rtase_w32(tp, RTASE_MAR0, swab32(mc_filter[1]));
+ rtase_w32(tp, RTASE_MAR1, swab32(mc_filter[0]));
+ rtase_w16(tp, RTASE_RX_CONFIG_0, rx_mode);
+}
+
static void rtase_irq_dis_and_clear(const struct rtase_private *tp)
{
const struct rtase_int_vector *ivec = &tp->int_vector[0];
@@ -555,6 +685,114 @@ static void rtase_hw_reset(const struct net_device *dev)
rtase_nic_reset(dev);
}
+static void rtase_set_rx_queue(const struct rtase_private *tp)
+{
+ u16 reg_data;
+
+ reg_data = rtase_r16(tp, RTASE_FCR);
+ switch (tp->func_rx_queue_num) {
+ case 1:
+ u16p_replace_bits(®_data, 0x1, RTASE_FCR_RXQ_MASK);
+ break;
+ case 2:
+ u16p_replace_bits(®_data, 0x2, RTASE_FCR_RXQ_MASK);
+ break;
+ case 4:
+ u16p_replace_bits(®_data, 0x3, RTASE_FCR_RXQ_MASK);
+ break;
+ }
+ rtase_w16(tp, RTASE_FCR, reg_data);
+}
+
+static void rtase_set_tx_queue(const struct rtase_private *tp)
+{
+ u16 reg_data;
+
+ reg_data = rtase_r16(tp, RTASE_TX_CONFIG_1);
+ switch (tp->tx_queue_ctrl) {
+ case 1:
+ u16p_replace_bits(®_data, 0x0, RTASE_TC_MODE_MASK);
+ break;
+ case 2:
+ u16p_replace_bits(®_data, 0x1, RTASE_TC_MODE_MASK);
+ break;
+ case 3:
+ case 4:
+ u16p_replace_bits(®_data, 0x2, RTASE_TC_MODE_MASK);
+ break;
+ default:
+ u16p_replace_bits(®_data, 0x3, RTASE_TC_MODE_MASK);
+ break;
+ }
+ rtase_w16(tp, RTASE_TX_CONFIG_1, reg_data);
+}
+
+static void rtase_hw_config(struct net_device *dev)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+ u32 reg_data32;
+ u16 reg_data16;
+
+ rtase_hw_reset(dev);
+
+ /* set rx dma burst */
+ reg_data16 = rtase_r16(tp, RTASE_RX_CONFIG_0);
+ reg_data16 &= ~(RTASE_RX_SINGLE_TAG | RTASE_RX_SINGLE_FETCH);
+ u16p_replace_bits(®_data16, RTASE_RX_DMA_BURST_256,
+ RTASE_RX_MX_DMA_MASK);
+ rtase_w16(tp, RTASE_RX_CONFIG_0, reg_data16);
+
+ /* new rx descritpor */
+ reg_data16 = rtase_r16(tp, RTASE_RX_CONFIG_1);
+ reg_data16 |= RTASE_RX_NEW_DESC_FORMAT_EN | RTASE_PCIE_NEW_FLOW;
+ u16p_replace_bits(®_data16, 0xF, RTASE_RX_MAX_FETCH_DESC_MASK);
+ rtase_w16(tp, RTASE_RX_CONFIG_1, reg_data16);
+
+ rtase_set_rx_queue(tp);
+
+ rtase_interrupt_mitigation(tp);
+
+ /* set tx dma burst size and interframe gap time */
+ reg_data32 = rtase_r32(tp, RTASE_TX_CONFIG_0);
+ u32p_replace_bits(®_data32, RTASE_TX_DMA_BURST_UNLIMITED,
+ RTASE_TX_DMA_MASK);
+ u32p_replace_bits(®_data32, RTASE_INTERFRAMEGAP,
+ RTASE_TX_INTER_FRAME_GAP_MASK);
+ rtase_w32(tp, RTASE_TX_CONFIG_0, reg_data32);
+
+ /* new tx descriptor */
+ reg_data16 = rtase_r16(tp, RTASE_TFUN_CTRL);
+ rtase_w16(tp, RTASE_TFUN_CTRL, reg_data16 |
+ RTASE_TX_NEW_DESC_FORMAT_EN);
+
+ /* tx fetch desc number */
+ rtase_w8(tp, RTASE_TDFNR, 0x10);
+
+ /* tag num select */
+ reg_data16 = rtase_r16(tp, RTASE_MTPS);
+ u16p_replace_bits(®_data16, 0x4, RTASE_TAG_NUM_SEL_MASK);
+ rtase_w16(tp, RTASE_MTPS, reg_data16);
+
+ rtase_set_tx_queue(tp);
+
+ rtase_w16(tp, RTASE_TOKSEL, 0x5555);
+
+ rtase_tally_counter_addr_fill(tp);
+ rtase_desc_addr_fill(tp);
+ rtase_hw_set_features(dev, dev->features);
+
+ /* enable flow control */
+ reg_data16 = rtase_r16(tp, RTASE_CPLUS_CMD);
+ reg_data16 |= (RTASE_FORCE_TXFLOW_EN | RTASE_FORCE_RXFLOW_EN);
+ rtase_w16(tp, RTASE_CPLUS_CMD, reg_data16);
+ /* set near fifo threshold - rx missed issue. */
+ rtase_w16(tp, RTASE_RFIFONFULL, 0x190);
+
+ rtase_w16(tp, RTASE_RMS, tp->rx_buf_sz);
+
+ rtase_hw_set_rx_packet_filter(dev);
+}
+
static void rtase_nic_enable(const struct net_device *dev)
{
const struct rtase_private *tp = netdev_priv(dev);
--
2.34.1
1. Implement rtase_interrupt to handle txQ0/rxQ0, txQ4~txQ7 interrupts,
and implement rtase_q_interrupt to handle txQ1/rxQ1, txQ2/rxQ2 and
txQ3/rxQ3 interrupts.
2. Implement rtase_poll to call ring_handler to process the tx or
rx packet of each ring. If the returned value is budget,it means that
there is still work of a certain ring that has not yet been completed.
Signed-off-by: Justin Lai <[email protected]>
---
.../net/ethernet/realtek/rtase/rtase_main.c | 89 +++++++++++++++++++
1 file changed, 89 insertions(+)
diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c b/drivers/net/ethernet/realtek/rtase/rtase_main.c
index caff2cfe78c7..f971110027ee 100644
--- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
+++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
@@ -592,6 +592,76 @@ static void rtase_hw_start(const struct net_device *dev)
rtase_enable_hw_interrupt(tp);
}
+/* the interrupt handler does RXQ0 and TXQ0, TXQ4~7 interrutp status
+ */
+static irqreturn_t rtase_interrupt(int irq, void *dev_instance)
+{
+ const struct rtase_private *tp;
+ struct rtase_int_vector *ivec;
+ u32 status;
+
+ ivec = dev_instance;
+ tp = ivec->tp;
+ status = rtase_r32(tp, ivec->isr_addr);
+
+ rtase_w32(tp, ivec->imr_addr, 0x0);
+ rtase_w32(tp, ivec->isr_addr, status & ~RTASE_FOVW);
+
+ if (napi_schedule_prep(&ivec->napi))
+ __napi_schedule(&ivec->napi);
+
+ return IRQ_HANDLED;
+}
+
+/* the interrupt handler does RXQ1&TXQ1 or RXQ2&TXQ2 or RXQ3&TXQ3 interrupt
+ * status according to interrupt vector
+ */
+static irqreturn_t rtase_q_interrupt(int irq, void *dev_instance)
+{
+ const struct rtase_private *tp;
+ struct rtase_int_vector *ivec;
+ u16 status;
+
+ ivec = dev_instance;
+ tp = ivec->tp;
+ status = rtase_r16(tp, ivec->isr_addr);
+
+ rtase_w16(tp, ivec->imr_addr, 0x0);
+ rtase_w16(tp, ivec->isr_addr, status);
+
+ if (napi_schedule_prep(&ivec->napi))
+ __napi_schedule(&ivec->napi);
+
+ return IRQ_HANDLED;
+}
+
+static int rtase_poll(struct napi_struct *napi, int budget)
+{
+ const struct rtase_int_vector *ivec;
+ const struct rtase_private *tp;
+ struct rtase_ring *ring;
+ int total_workdone = 0;
+
+ ivec = container_of(napi, struct rtase_int_vector, napi);
+ tp = ivec->tp;
+
+ list_for_each_entry(ring, &ivec->ring_list, ring_entry) {
+ total_workdone += ring->ring_handler(ring, budget);
+ }
+
+ if (total_workdone >= budget)
+ return budget;
+
+ if (napi_complete_done(napi, total_workdone)) {
+ if (!ivec->index)
+ rtase_w32(tp, ivec->imr_addr, ivec->imr);
+ else
+ rtase_w16(tp, ivec->imr_addr, ivec->imr);
+ }
+
+ return total_workdone;
+}
+
static int rtase_open(struct net_device *dev)
{
struct rtase_private *tp = netdev_priv(dev);
@@ -738,9 +808,28 @@ static void rtase_rar_set(const struct rtase_private *tp, const u8 *addr)
rtase_w16(tp, RTASE_LBK_CTRL, RTASE_LBK_ATLD | RTASE_LBK_CLR);
}
+#ifdef CONFIG_NET_POLL_CONTROLLER
+/* Polling 'interrupt' - used by things like netconsole to send skbs
+ * without having to re-enable interrupts. It's not called while
+ * the interrupt routine is executing.
+ */
+static void rtase_netpoll(struct net_device *dev)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+ const struct pci_dev *pdev = tp->pdev;
+
+ disable_irq(pdev->irq);
+ rtase_interrupt(pdev->irq, dev);
+ enable_irq(pdev->irq);
+}
+#endif
+
static const struct net_device_ops rtase_netdev_ops = {
.ndo_open = rtase_open,
.ndo_stop = rtase_close,
+#ifdef CONFIG_NET_POLL_CONTROLLER
+ .ndo_poll_controller = rtase_netpoll,
+#endif
};
static void rtase_get_mac_address(struct net_device *dev)
--
2.34.1
Implement rx_handler to read the information of the rx descriptor,
thereby checking the packet accordingly and storing the packet
in the socket buffer to complete the reception of the packet.
Signed-off-by: Justin Lai <[email protected]>
---
.../net/ethernet/realtek/rtase/rtase_main.c | 146 ++++++++++++++++++
1 file changed, 146 insertions(+)
diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c b/drivers/net/ethernet/realtek/rtase/rtase_main.c
index 6bdb4edbfbc1..9ee9f30444f3 100644
--- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
+++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
@@ -458,6 +458,152 @@ static void rtase_rx_ring_clear(struct rtase_ring *ring)
}
}
+static int rtase_fragmented_frame(u32 status)
+{
+ return (status & (RTASE_RX_FIRST_FRAG | RTASE_RX_LAST_FRAG)) !=
+ (RTASE_RX_FIRST_FRAG | RTASE_RX_LAST_FRAG);
+}
+
+static void rtase_rx_csum(const struct rtase_private *tp, struct sk_buff *skb,
+ const union rtase_rx_desc *desc)
+{
+ u32 opts2 = le32_to_cpu(desc->desc_status.opts2);
+
+ /* rx csum offload */
+ if (((opts2 & RTASE_RX_V4F) && !(opts2 & RTASE_RX_IPF)) ||
+ (opts2 & RTASE_RX_V6F)) {
+ if (((opts2 & RTASE_RX_TCPT) && !(opts2 & RTASE_RX_TCPF)) ||
+ ((opts2 & RTASE_RX_UDPT) && !(opts2 & RTASE_RX_UDPF))) {
+ skb->ip_summed = CHECKSUM_UNNECESSARY;
+ } else {
+ skb->ip_summed = CHECKSUM_NONE;
+ }
+ } else {
+ skb->ip_summed = CHECKSUM_NONE;
+ }
+}
+
+static void rtase_rx_vlan_skb(union rtase_rx_desc *desc, struct sk_buff *skb)
+{
+ u32 opts2 = le32_to_cpu(desc->desc_status.opts2);
+
+ if (!(opts2 & RTASE_RX_VLAN_TAG))
+ return;
+
+ __vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q),
+ swab16(opts2 & RTASE_VLAN_TAG_MASK));
+}
+
+static void rtase_rx_skb(const struct rtase_ring *ring, struct sk_buff *skb)
+{
+ struct rtase_int_vector *ivec = ring->ivec;
+
+ napi_gro_receive(&ivec->napi, skb);
+}
+
+static int rx_handler(struct rtase_ring *ring, int budget)
+{
+ const struct rtase_private *tp = ring->ivec->tp;
+ union rtase_rx_desc *desc_base = ring->desc;
+ u32 pkt_size, cur_rx, delta, entry, status;
+ struct net_device *dev = tp->dev;
+ union rtase_rx_desc *desc;
+ struct sk_buff *skb;
+ int workdone = 0;
+
+ cur_rx = ring->cur_idx;
+ entry = cur_rx % RTASE_NUM_DESC;
+ desc = &desc_base[entry];
+
+ do {
+ /* make sure discriptor has been updated */
+ rmb();
+ status = le32_to_cpu(desc->desc_status.opts1);
+
+ if (status & RTASE_DESC_OWN)
+ break;
+
+ if (unlikely(status & RTASE_RX_RES)) {
+ if (net_ratelimit())
+ netdev_warn(dev, "Rx ERROR. status = %08x\n",
+ status);
+
+ dev->stats.rx_errors++;
+
+ if (status & (RTASE_RX_RWT | RTASE_RX_RUNT))
+ dev->stats.rx_length_errors++;
+
+ if (status & RTASE_RX_CRC)
+ dev->stats.rx_crc_errors++;
+
+ if (dev->features & NETIF_F_RXALL)
+ goto process_pkt;
+
+ rtase_mark_to_asic(desc, tp->rx_buf_sz);
+ goto skip_process_pkt;
+ }
+
+process_pkt:
+ pkt_size = status & RTASE_RX_PKT_SIZE_MASK;
+ if (likely(!(dev->features & NETIF_F_RXFCS)))
+ pkt_size -= ETH_FCS_LEN;
+
+ /* the driver does not support incoming fragmented
+ * frames. they are seen as a symptom of over-mtu
+ * sized frames
+ */
+ if (unlikely(rtase_fragmented_frame(status))) {
+ dev->stats.rx_dropped++;
+ dev->stats.rx_length_errors++;
+ rtase_mark_to_asic(desc, tp->rx_buf_sz);
+ continue;
+ }
+
+ skb = ring->skbuff[entry];
+ dma_sync_single_for_cpu(&tp->pdev->dev,
+ ring->mis.data_phy_addr[entry],
+ tp->rx_buf_sz, DMA_FROM_DEVICE);
+
+ ring->skbuff[entry] = NULL;
+
+ if (dev->features & NETIF_F_RXCSUM)
+ rtase_rx_csum(tp, skb, desc);
+
+ skb->dev = dev;
+ skb_put(skb, pkt_size);
+ skb_mark_for_recycle(skb);
+ skb->protocol = eth_type_trans(skb, dev);
+
+ if (skb->pkt_type == PACKET_MULTICAST)
+ dev->stats.multicast++;
+
+ rtase_rx_vlan_skb(desc, skb);
+ rtase_rx_skb(ring, skb);
+
+ dev_sw_netstats_rx_add(dev, pkt_size);
+
+skip_process_pkt:
+ workdone++;
+ cur_rx++;
+ entry = cur_rx % RTASE_NUM_DESC;
+ desc = ring->desc + sizeof(union rtase_rx_desc) * entry;
+ prefetch(desc);
+ } while (workdone != budget);
+
+ ring->cur_idx = cur_rx;
+ delta = rtase_rx_ring_fill(ring, ring->dirty_idx, ring->cur_idx, 1);
+
+ if (!delta && workdone)
+ netdev_info(dev, "no Rx buffer allocated\n");
+
+ ring->dirty_idx += delta;
+
+ if ((ring->dirty_idx + RTASE_NUM_DESC) == ring->cur_idx)
+ netdev_emerg(dev, "Rx buffers exhausted\n");
+
+ return workdone;
+}
+
static void rtase_rx_desc_init(struct rtase_private *tp, u16 idx)
{
struct rtase_ring *ring = &tp->rx_ring[idx];
--
2.34.1
1. Implement .ndo_set_rx_mode so that the device can change address
list filtering.
2. Implement .ndo_set_mac_address so that mac address can be changed.
3. Implement .ndo_change_mtu so that mtu can be changed.
4. Implement .ndo_tx_timeout to perform related processing when the
transmitter does not make any progress.
5. Implement .ndo_get_stats64 to provide statistics that are called
when the user wants to get network device usage.
6. Implement .ndo_vlan_rx_add_vid to register VLAN ID when the device
supports VLAN filtering.
7. Implement .ndo_vlan_rx_kill_vid to unregister VLAN ID when the device
supports VLAN filtering.
8. Implement the .ndo_setup_tc to enable setting any "tc" scheduler,
classifier or action on dev.
9. Implement .ndo_fix_features enables adjusting requested feature flags
based on device-specific constraints.
10. Implement .ndo_set_features enables updating device configuration to
new features.
Signed-off-by: Justin Lai <[email protected]>
---
.../net/ethernet/realtek/rtase/rtase_main.c | 237 ++++++++++++++++++
1 file changed, 237 insertions(+)
diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c b/drivers/net/ethernet/realtek/rtase/rtase_main.c
index 9ee9f30444f3..38faff333a93 100644
--- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
+++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
@@ -1444,6 +1444,11 @@ static netdev_tx_t rtase_start_xmit(struct sk_buff *skb,
return NETDEV_TX_BUSY;
}
+static void rtase_set_rx_mode(struct net_device *dev)
+{
+ rtase_hw_set_rx_packet_filter(dev);
+}
+
static void rtase_enable_eem_write(const struct rtase_private *tp)
{
u8 val;
@@ -1476,6 +1481,205 @@ static void rtase_rar_set(const struct rtase_private *tp, const u8 *addr)
rtase_w16(tp, RTASE_LBK_CTRL, RTASE_LBK_ATLD | RTASE_LBK_CLR);
}
+static int rtase_set_mac_address(struct net_device *dev, void *p)
+{
+ struct rtase_private *tp = netdev_priv(dev);
+ int ret;
+
+ ret = eth_mac_addr(dev, p);
+ if (ret)
+ return ret;
+
+ rtase_rar_set(tp, dev->dev_addr);
+
+ return 0;
+}
+
+static int rtase_change_mtu(struct net_device *dev, int new_mtu)
+{
+ dev->mtu = new_mtu;
+
+ netdev_update_features(dev);
+
+ return 0;
+}
+
+static void rtase_wait_for_quiescence(const struct net_device *dev)
+{
+ struct rtase_private *tp = netdev_priv(dev);
+ struct rtase_int_vector *ivec;
+ u32 i;
+
+ for (i = 0; i < tp->int_nums; i++) {
+ ivec = &tp->int_vector[i];
+ synchronize_irq(ivec->irq);
+ /* wait for any pending NAPI task to complete */
+ napi_disable(&ivec->napi);
+ }
+
+ rtase_irq_dis_and_clear(tp);
+
+ for (i = 0; i < tp->int_nums; i++) {
+ ivec = &tp->int_vector[i];
+ napi_enable(&ivec->napi);
+ }
+}
+
+static void rtase_sw_reset(struct net_device *dev)
+{
+ struct rtase_private *tp = netdev_priv(dev);
+ int ret;
+
+ netif_stop_queue(dev);
+ netif_carrier_off(dev);
+ rtase_hw_reset(dev);
+
+ /* let's wait a bit while any (async) irq lands on */
+ rtase_wait_for_quiescence(dev);
+ rtase_tx_clear(tp);
+ rtase_rx_clear(tp);
+
+ ret = rtase_init_ring(dev);
+ if (ret) {
+ netdev_err(dev, "unable to init ring\n");
+ rtase_free_desc(tp);
+ return;
+ }
+
+ rtase_hw_config(dev);
+ /* always link, so start to transmit & receive */
+ rtase_hw_start(dev);
+
+ netif_carrier_on(dev);
+ netif_wake_queue(dev);
+}
+
+static void rtase_dump_tally_counter(const struct rtase_private *tp)
+{
+ dma_addr_t paddr = tp->tally_paddr;
+ u32 cmd = lower_32_bits(paddr);
+ u32 val;
+ int err;
+
+ rtase_w32(tp, RTASE_DTCCR4, upper_32_bits(paddr));
+ rtase_w32(tp, RTASE_DTCCR0, cmd);
+ rtase_w32(tp, RTASE_DTCCR0, cmd | RTASE_COUNTER_DUMP);
+
+ err = read_poll_timeout(rtase_r32, val, !(val & RTASE_COUNTER_DUMP),
+ 10, 250, false, tp, RTASE_DTCCR0);
+
+ if (err == -ETIMEDOUT)
+ netdev_err(tp->dev, "error occurred in dump tally counter\n");
+}
+
+static void rtase_dump_state(const struct net_device *dev)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+ int max_reg_size = RTASE_PCI_REGS_SIZE;
+ const struct rtase_counters *counters;
+ const struct rtase_ring *ring;
+ u32 dword_rd;
+ int n = 0;
+
+ ring = &tp->tx_ring[0];
+ netdev_err(dev, "Tx descriptor info:\n");
+ netdev_err(dev, "Tx curIdx = 0x%x\n", ring->cur_idx);
+ netdev_err(dev, "Tx dirtyIdx = 0x%x\n", ring->dirty_idx);
+ netdev_err(dev, "Tx phyAddr = %pad\n", &ring->phy_addr);
+
+ ring = &tp->rx_ring[0];
+ netdev_err(dev, "Rx descriptor info:\n");
+ netdev_err(dev, "Rx curIdx = 0x%x\n", ring->cur_idx);
+ netdev_err(dev, "Rx dirtyIdx = 0x%x\n", ring->dirty_idx);
+ netdev_err(dev, "Rx phyAddr = %pad\n", &ring->phy_addr);
+
+ netdev_err(dev, "Device Registers:\n");
+ netdev_err(dev, "Chip Command = 0x%02x\n",
+ rtase_r8(tp, RTASE_CHIP_CMD));
+ netdev_err(dev, "IMR = %08x\n", rtase_r32(tp, RTASE_IMR0));
+ netdev_err(dev, "ISR = %08x\n", rtase_r32(tp, RTASE_ISR0));
+ netdev_err(dev, "Boot Ctrl Reg(0xE004) = %04x\n",
+ rtase_r16(tp, RTASE_BOOT_CTL));
+ netdev_err(dev, "EPHY ISR(0xE014) = %04x\n",
+ rtase_r16(tp, RTASE_EPHY_ISR));
+ netdev_err(dev, "EPHY IMR(0xE016) = %04x\n",
+ rtase_r16(tp, RTASE_EPHY_IMR));
+ netdev_err(dev, "CLKSW SET REG(0xE018) = %04x\n",
+ rtase_r16(tp, RTASE_CLKSW_SET));
+
+ netdev_err(dev, "Dump PCI Registers:\n");
+
+ while (n < max_reg_size) {
+ if ((n % RTASE_DWORD_MOD) == 0)
+ netdev_err(tp->dev, "0x%03x:\n", n);
+
+ pci_read_config_dword(tp->pdev, n, &dword_rd);
+ netdev_err(tp->dev, "%08x\n", dword_rd);
+ n += 4;
+ }
+
+ netdev_err(dev, "Dump tally counter:\n");
+ counters = tp->tally_vaddr;
+ rtase_dump_tally_counter(tp);
+
+ netdev_err(dev, "tx_packets %lld\n",
+ le64_to_cpu(counters->tx_packets));
+ netdev_err(dev, "rx_packets %lld\n",
+ le64_to_cpu(counters->rx_packets));
+ netdev_err(dev, "tx_errors %lld\n",
+ le64_to_cpu(counters->tx_errors));
+ netdev_err(dev, "rx_errors %d\n",
+ le32_to_cpu(counters->rx_errors));
+ netdev_err(dev, "rx_missed %d\n",
+ le16_to_cpu(counters->rx_missed));
+ netdev_err(dev, "align_errors %d\n",
+ le16_to_cpu(counters->align_errors));
+ netdev_err(dev, "tx_one_collision %d\n",
+ le32_to_cpu(counters->tx_one_collision));
+ netdev_err(dev, "tx_multi_collision %d\n",
+ le32_to_cpu(counters->tx_multi_collision));
+ netdev_err(dev, "rx_unicast %lld\n",
+ le64_to_cpu(counters->rx_unicast));
+ netdev_err(dev, "rx_broadcast %lld\n",
+ le64_to_cpu(counters->rx_broadcast));
+ netdev_err(dev, "rx_multicast %d\n",
+ le32_to_cpu(counters->rx_multicast));
+ netdev_err(dev, "tx_aborted %d\n",
+ le16_to_cpu(counters->tx_aborted));
+ netdev_err(dev, "tx_underun %d\n",
+ le16_to_cpu(counters->tx_underun));
+}
+
+static void rtase_tx_timeout(struct net_device *dev, unsigned int txqueue)
+{
+ rtase_dump_state(dev);
+ rtase_sw_reset(dev);
+}
+
+static void rtase_get_stats64(struct net_device *dev,
+ struct rtnl_link_stats64 *stats)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+ const struct rtase_counters *counters;
+
+ counters = tp->tally_vaddr;
+
+ if (!counters)
+ return;
+
+ netdev_stats_to_stats64(stats, &dev->stats);
+ dev_fetch_sw_netstats(stats, dev->tstats);
+
+ /* fetch additional counter values missing in stats collected by driver
+ * from tally counter
+ */
+ rtase_dump_tally_counter(tp);
+ stats->tx_errors = le64_to_cpu(counters->tx_errors);
+ stats->rx_missed_errors = le16_to_cpu(counters->rx_missed);
+ stats->collisions = le32_to_cpu(counters->tx_multi_collision);
+ stats->tx_aborted_errors = le16_to_cpu(counters->tx_aborted);
+}
+
#ifdef CONFIG_NET_POLL_CONTROLLER
/* Polling 'interrupt' - used by things like netconsole to send skbs
* without having to re-enable interrupts. It's not called while
@@ -1492,13 +1696,46 @@ static void rtase_netpoll(struct net_device *dev)
}
#endif
+static netdev_features_t rtase_fix_features(struct net_device *dev,
+ netdev_features_t features)
+{
+ netdev_features_t features_fix = features;
+
+ /* not support TSO for jumbo frames */
+ if (dev->mtu > ETH_DATA_LEN)
+ features_fix &= ~NETIF_F_ALL_TSO;
+
+ return features_fix;
+}
+
+static int rtase_set_features(struct net_device *dev,
+ netdev_features_t features)
+{
+ netdev_features_t features_set = features;
+
+ features_set &= NETIF_F_RXALL | NETIF_F_RXCSUM |
+ NETIF_F_HW_VLAN_CTAG_RX;
+
+ if (features_set ^ dev->features)
+ rtase_hw_set_features(dev, features_set);
+
+ return 0;
+}
+
static const struct net_device_ops rtase_netdev_ops = {
.ndo_open = rtase_open,
.ndo_stop = rtase_close,
.ndo_start_xmit = rtase_start_xmit,
+ .ndo_set_rx_mode = rtase_set_rx_mode,
+ .ndo_set_mac_address = rtase_set_mac_address,
+ .ndo_change_mtu = rtase_change_mtu,
+ .ndo_tx_timeout = rtase_tx_timeout,
+ .ndo_get_stats64 = rtase_get_stats64,
#ifdef CONFIG_NET_POLL_CONTROLLER
.ndo_poll_controller = rtase_netpoll,
#endif
+ .ndo_fix_features = rtase_fix_features,
+ .ndo_set_features = rtase_set_features,
};
static void rtase_get_mac_address(struct net_device *dev)
--
2.34.1
Implement the pci_driver suspend function to enable the device
to sleep, and implement the resume function to enable the device
to resume operation.
Signed-off-by: Justin Lai <[email protected]>
---
.../net/ethernet/realtek/rtase/rtase_main.c | 51 +++++++++++++++++++
1 file changed, 51 insertions(+)
diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c b/drivers/net/ethernet/realtek/rtase/rtase_main.c
index 38faff333a93..e9bbafc212c5 100644
--- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
+++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
@@ -2201,12 +2201,63 @@ static void rtase_shutdown(struct pci_dev *pdev)
rtase_reset_interrupt(pdev, tp);
}
+static int rtase_suspend(struct device *device)
+{
+ struct net_device *dev = dev_get_drvdata(device);
+
+ if (netif_running(dev)) {
+ netif_device_detach(dev);
+ rtase_hw_reset(dev);
+ }
+
+ return 0;
+}
+
+static int rtase_resume(struct device *device)
+{
+ struct net_device *dev = dev_get_drvdata(device);
+ struct rtase_private *tp = netdev_priv(dev);
+ int ret;
+
+ /* restore last modified mac address */
+ rtase_rar_set(tp, dev->dev_addr);
+
+ if (!netif_running(dev))
+ goto out;
+
+ rtase_wait_for_quiescence(dev);
+
+ rtase_tx_clear(tp);
+ rtase_rx_clear(tp);
+
+ ret = rtase_init_ring(dev);
+ if (ret) {
+ netdev_err(dev, "unable to init ring\n");
+ rtase_free_desc(tp);
+ return -ENOMEM;
+ }
+
+ rtase_hw_config(dev);
+ /* always link, so start to transmit & receive */
+ rtase_hw_start(dev);
+
+ netif_device_attach(dev);
+out:
+
+ return 0;
+}
+
+static const struct dev_pm_ops rtase_pm_ops = {
+ SYSTEM_SLEEP_PM_OPS(rtase_suspend, rtase_resume)
+};
+
static struct pci_driver rtase_pci_driver = {
.name = KBUILD_MODNAME,
.id_table = rtase_pci_tbl,
.probe = rtase_init_one,
.remove = rtase_remove_one,
.shutdown = rtase_shutdown,
+ .driver.pm = pm_ptr(&rtase_pm_ops),
};
module_pci_driver(rtase_pci_driver);
--
2.34.1
1. Add the RTASE entry in the Kconfig.
2. Add the CONFIG_RTASE entry in the Makefile.
Signed-off-by: Justin Lai <[email protected]>
---
drivers/net/ethernet/realtek/Kconfig | 19 +++++++++++++++++++
drivers/net/ethernet/realtek/Makefile | 1 +
2 files changed, 20 insertions(+)
diff --git a/drivers/net/ethernet/realtek/Kconfig b/drivers/net/ethernet/realtek/Kconfig
index 03015b665f4e..8a8ea51c639e 100644
--- a/drivers/net/ethernet/realtek/Kconfig
+++ b/drivers/net/ethernet/realtek/Kconfig
@@ -120,4 +120,23 @@ config R8169_LEDS
Optional support for controlling the NIC LED's with the netdev
LED trigger.
+config RTASE
+ tristate "Realtek Automotive Switch 9054/9068/9072/9075/9068/9071 PCIe Interface support"
+ depends on PCI
+ select CRC32
+ select PAGE_POOL
+ help
+ Say Y here and it will be compiled and linked with the kernel
+ if you have a Realtek Ethernet adapter belonging to the
+ following families:
+ RTL9054 5GBit Ethernet
+ RTL9068 5GBit Ethernet
+ RTL9072 5GBit Ethernet
+ RTL9075 5GBit Ethernet
+ RTL9068 5GBit Ethernet
+ RTL9071 5GBit Ethernet
+
+ To compile this driver as a module, choose M here: the module
+ will be called rtase. This is recommended.
+
endif # NET_VENDOR_REALTEK
diff --git a/drivers/net/ethernet/realtek/Makefile b/drivers/net/ethernet/realtek/Makefile
index 635491d8826e..046adf503ff4 100644
--- a/drivers/net/ethernet/realtek/Makefile
+++ b/drivers/net/ethernet/realtek/Makefile
@@ -9,3 +9,4 @@ obj-$(CONFIG_ATP) += atp.o
r8169-y += r8169_main.o r8169_firmware.o r8169_phy_config.o
r8169-$(CONFIG_R8169_LEDS) += r8169_leds.o
obj-$(CONFIG_R8169) += r8169.o
+obj-$(CONFIG_RTASE) += rtase/
--
2.34.1
Add a Makefile in the rtase folder to build rtase driver.
Signed-off-by: Justin Lai <[email protected]>
---
drivers/net/ethernet/realtek/rtase/Makefile | 10 ++++++++++
1 file changed, 10 insertions(+)
create mode 100644 drivers/net/ethernet/realtek/rtase/Makefile
diff --git a/drivers/net/ethernet/realtek/rtase/Makefile b/drivers/net/ethernet/realtek/rtase/Makefile
new file mode 100644
index 000000000000..ba3d8550f9e6
--- /dev/null
+++ b/drivers/net/ethernet/realtek/rtase/Makefile
@@ -0,0 +1,10 @@
+# SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause
+# Copyright(c) 2024 Realtek Semiconductor Corp. All rights reserved.
+
+#
+# Makefile for the Realtek PCIe driver
+#
+
+obj-$(CONFIG_RTASE) += rtase.o
+
+rtase-objs := rtase_main.o
--
2.34.1
Add myself and Larry Chiu as the maintainer for the rtase ethernet driver.
Signed-off-by: Justin Lai <[email protected]>
---
MAINTAINERS | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 94fddfcec2fb..5ee95a083591 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -19296,6 +19296,13 @@ L: [email protected]
S: Maintained
F: drivers/tty/rpmsg_tty.c
+RTASE ETHERNET DRIVER
+M: Justin Lai <[email protected]>
+M: Larry Chiu <[email protected]>
+L: [email protected]
+S: Maintained
+F: drivers/net/ethernet/realtek/rtase/
+
RTL2830 MEDIA DRIVER
L: [email protected]
S: Orphan
--
2.34.1
Implement the ethtool function to support users to obtain network card
information, including obtaining various device settings, Report whether
physical link is up, Report pause parameters, Set pause parameters,
Return a set of strings that describe the requested objects, Get number
of strings that @get_strings will write, Return extended statistics
about the device.
Signed-off-by: Justin Lai <[email protected]>
---
.../net/ethernet/realtek/rtase/rtase_main.c | 103 ++++++++++++++++++
1 file changed, 103 insertions(+)
diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c b/drivers/net/ethernet/realtek/rtase/rtase_main.c
index e9bbafc212c5..da2627ea99fe 100644
--- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
+++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
@@ -1758,9 +1758,112 @@ static void rtase_get_mac_address(struct net_device *dev)
rtase_rar_set(tp, dev->dev_addr);
}
+static void rtase_get_drvinfo(struct net_device *dev,
+ struct ethtool_drvinfo *drvinfo)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+
+ strscpy(drvinfo->driver, KBUILD_MODNAME, 32);
+ strscpy(drvinfo->bus_info, pci_name(tp->pdev), 32);
+}
+
+static int rtase_get_settings(struct net_device *dev,
+ struct ethtool_link_ksettings *cmd)
+{
+ u32 supported = SUPPORTED_MII | SUPPORTED_Pause | SUPPORTED_Asym_Pause;
+
+ ethtool_convert_legacy_u32_to_link_mode(cmd->link_modes.supported,
+ supported);
+ cmd->base.speed = SPEED_5000;
+ cmd->base.duplex = DUPLEX_FULL;
+ cmd->base.port = PORT_MII;
+ cmd->base.autoneg = AUTONEG_DISABLE;
+
+ return 0;
+}
+
+static void rtase_get_pauseparam(struct net_device *dev,
+ struct ethtool_pauseparam *pause)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+ u16 value = rtase_r16(tp, RTASE_CPLUS_CMD);
+
+ pause->autoneg = AUTONEG_DISABLE;
+
+ if ((value & (RTASE_FORCE_TXFLOW_EN | RTASE_FORCE_RXFLOW_EN)) ==
+ (RTASE_FORCE_TXFLOW_EN | RTASE_FORCE_RXFLOW_EN)) {
+ pause->rx_pause = 1;
+ pause->tx_pause = 1;
+ } else if ((value & RTASE_FORCE_TXFLOW_EN)) {
+ pause->tx_pause = 1;
+ } else if ((value & RTASE_FORCE_RXFLOW_EN)) {
+ pause->rx_pause = 1;
+ }
+}
+
+static int rtase_set_pauseparam(struct net_device *dev,
+ struct ethtool_pauseparam *pause)
+{
+ const struct rtase_private *tp = netdev_priv(dev);
+ u16 value = rtase_r16(tp, RTASE_CPLUS_CMD);
+
+ if (pause->autoneg)
+ return -EOPNOTSUPP;
+
+ value &= ~(RTASE_FORCE_TXFLOW_EN | RTASE_FORCE_RXFLOW_EN);
+
+ if (pause->tx_pause)
+ value |= RTASE_FORCE_TXFLOW_EN;
+
+ if (pause->rx_pause)
+ value |= RTASE_FORCE_RXFLOW_EN;
+
+ rtase_w16(tp, RTASE_CPLUS_CMD, value);
+ return 0;
+}
+
+static void rtase_get_eth_mac_stats(struct net_device *dev,
+ struct ethtool_eth_mac_stats *stats)
+{
+ struct rtase_private *tp = netdev_priv(dev);
+ const struct rtase_counters *counters;
+
+ counters = tp->tally_vaddr;
+ if (!counters)
+ return;
+
+ rtase_dump_tally_counter(tp);
+
+ stats->FramesTransmittedOK = le64_to_cpu(counters->tx_packets);
+ stats->SingleCollisionFrames = le32_to_cpu(counters->tx_one_collision);
+ stats->MultipleCollisionFrames =
+ le32_to_cpu(counters->tx_multi_collision);
+ stats->FramesReceivedOK = le64_to_cpu(counters->rx_packets);
+ stats->FrameCheckSequenceErrors = le32_to_cpu(counters->rx_errors);
+ stats->AlignmentErrors = le16_to_cpu(counters->align_errors);
+ stats->FramesAbortedDueToXSColls = le16_to_cpu(counters->tx_aborted);
+ stats->FramesLostDueToIntMACXmitError =
+ le64_to_cpu(counters->tx_errors);
+ stats->FramesLostDueToIntMACRcvError =
+ le16_to_cpu(counters->rx_missed);
+ stats->MulticastFramesReceivedOK = le32_to_cpu(counters->rx_multicast);
+ stats->BroadcastFramesReceivedOK = le64_to_cpu(counters->rx_broadcast);
+}
+
+static const struct ethtool_ops rtase_ethtool_ops = {
+ .get_drvinfo = rtase_get_drvinfo,
+ .get_link = ethtool_op_get_link,
+ .get_link_ksettings = rtase_get_settings,
+ .get_pauseparam = rtase_get_pauseparam,
+ .set_pauseparam = rtase_set_pauseparam,
+ .get_eth_mac_stats = rtase_get_eth_mac_stats,
+ .get_ts_info = ethtool_op_get_ts_info,
+};
+
static void rtase_init_netdev_ops(struct net_device *dev)
{
dev->netdev_ops = &rtase_netdev_ops;
+ dev->ethtool_ops = &rtase_ethtool_ops;
}
static void rtase_reset_interrupt(struct pci_dev *pdev,
--
2.34.1
Implement .ndo_start_xmit function to fill the information of the packet
to be transmitted into the tx descriptor, and then the hardware will
transmit the packet using the information in the tx descriptor.
In addition, we also implemented the tx_handler function to enable the
tx descriptor to be reused.
Signed-off-by: Justin Lai <[email protected]>
---
.../net/ethernet/realtek/rtase/rtase_main.c | 285 ++++++++++++++++++
1 file changed, 285 insertions(+)
diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c b/drivers/net/ethernet/realtek/rtase/rtase_main.c
index 23406c195cff..6bdb4edbfbc1 100644
--- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
+++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
@@ -256,6 +256,68 @@ static void rtase_mark_to_asic(union rtase_rx_desc *desc, u32 rx_buf_sz)
cpu_to_le32(RTASE_DESC_OWN | eor | rx_buf_sz));
}
+static u32 rtase_tx_avail(struct rtase_ring *ring)
+{
+ return READ_ONCE(ring->dirty_idx) + RTASE_NUM_DESC -
+ READ_ONCE(ring->cur_idx);
+}
+
+static int tx_handler(struct rtase_ring *ring, int budget)
+{
+ const struct rtase_private *tp = ring->ivec->tp;
+ struct net_device *dev = tp->dev;
+ u32 dirty_tx, tx_left;
+ u32 bytes_compl = 0;
+ u32 pkts_compl = 0;
+ int workdone = 0;
+
+ dirty_tx = ring->dirty_idx;
+ tx_left = READ_ONCE(ring->cur_idx) - dirty_tx;
+
+ while (tx_left > 0) {
+ u32 entry = dirty_tx % RTASE_NUM_DESC;
+ struct rtase_tx_desc *desc = ring->desc +
+ sizeof(struct rtase_tx_desc) * entry;
+ u32 status;
+
+ status = le32_to_cpu(desc->opts1);
+
+ if (status & RTASE_DESC_OWN)
+ break;
+
+ rtase_unmap_tx_skb(tp->pdev, ring->mis.len[entry], desc);
+ ring->mis.len[entry] = 0;
+ if (ring->skbuff[entry]) {
+ pkts_compl++;
+ bytes_compl += ring->skbuff[entry]->len;
+ napi_consume_skb(ring->skbuff[entry], budget);
+ ring->skbuff[entry] = NULL;
+ }
+
+ dirty_tx++;
+ tx_left--;
+ workdone++;
+
+ if (workdone == RTASE_TX_BUDGET_DEFAULT)
+ break;
+ }
+
+ if (ring->dirty_idx != dirty_tx) {
+ dev_sw_netstats_tx_add(dev, pkts_compl, bytes_compl);
+ WRITE_ONCE(ring->dirty_idx, dirty_tx);
+
+ netif_subqueue_completed_wake(dev, ring->index, pkts_compl,
+ bytes_compl,
+ rtase_tx_avail(ring),
+ RTASE_TX_START_THRS);
+
+ if (ring->cur_idx != dirty_tx)
+ rtase_w8(tp, RTASE_TPPOLL, BIT(ring->index));
+ }
+
+ return 0;
+}
+
static void rtase_tx_desc_init(struct rtase_private *tp, u16 idx)
{
struct rtase_ring *ring = &tp->tx_ring[idx];
@@ -1014,6 +1076,228 @@ static int rtase_close(struct net_device *dev)
return 0;
}
+static u32 rtase_tx_vlan_tag(const struct rtase_private *tp,
+ const struct sk_buff *skb)
+{
+ return (skb_vlan_tag_present(skb)) ?
+ (RTASE_TX_VLAN_TAG | swab16(skb_vlan_tag_get(skb))) : 0x00;
+}
+
+static u32 rtase_tx_csum(struct sk_buff *skb, const struct net_device *dev)
+{
+ u32 csum_cmd = 0;
+ u8 ip_protocol;
+
+ switch (vlan_get_protocol(skb)) {
+ case htons(ETH_P_IP):
+ csum_cmd = RTASE_TX_IPCS_C;
+ ip_protocol = ip_hdr(skb)->protocol;
+ break;
+
+ case htons(ETH_P_IPV6):
+ csum_cmd = RTASE_TX_IPV6F_C;
+ ip_protocol = ipv6_hdr(skb)->nexthdr;
+ break;
+
+ default:
+ ip_protocol = IPPROTO_RAW;
+ break;
+ }
+
+ if (ip_protocol == IPPROTO_TCP)
+ csum_cmd |= RTASE_TX_TCPCS_C;
+ else if (ip_protocol == IPPROTO_UDP)
+ csum_cmd |= RTASE_TX_UDPCS_C;
+
+ csum_cmd |= u32_encode_bits(skb_transport_offset(skb),
+ RTASE_TCPHO_MASK);
+
+ return csum_cmd;
+}
+
+static int rtase_xmit_frags(struct rtase_ring *ring, struct sk_buff *skb,
+ u32 opts1, u32 opts2)
+{
+ const struct skb_shared_info *info = skb_shinfo(skb);
+ const struct rtase_private *tp = ring->ivec->tp;
+ const u8 nr_frags = info->nr_frags;
+ struct rtase_tx_desc *txd = NULL;
+ u32 cur_frag, entry;
+
+ entry = ring->cur_idx;
+ for (cur_frag = 0; cur_frag < nr_frags; cur_frag++) {
+ const skb_frag_t *frag = &info->frags[cur_frag];
+ dma_addr_t mapping;
+ u32 status, len;
+ void *addr;
+
+ entry = (entry + 1) % RTASE_NUM_DESC;
+
+ txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
+ len = skb_frag_size(frag);
+ addr = skb_frag_address(frag);
+ mapping = dma_map_single(&tp->pdev->dev, addr, len,
+ DMA_TO_DEVICE);
+
+ if (unlikely(dma_mapping_error(&tp->pdev->dev, mapping))) {
+ if (unlikely(net_ratelimit()))
+ netdev_err(tp->dev,
+ "Failed to map TX fragments DMA!\n");
+
+ goto err_out;
+ }
+
+ if (((entry + 1) % RTASE_NUM_DESC) == 0)
+ status = (opts1 | len | RTASE_RING_END);
+ else
+ status = opts1 | len;
+
+ if (cur_frag == (nr_frags - 1)) {
+ ring->skbuff[entry] = skb;
+ status |= RTASE_TX_LAST_FRAG;
+ }
+
+ ring->mis.len[entry] = len;
+ txd->addr = cpu_to_le64(mapping);
+ txd->opts2 = cpu_to_le32(opts2);
+
+ /* make sure the operating fields have been updated */
+ dma_wmb();
+ txd->opts1 = cpu_to_le32(status);
+ }
+
+ return cur_frag;
+
+err_out:
+ rtase_tx_clear_range(ring, ring->cur_idx + 1, cur_frag);
+ return -EIO;
+}
+
+static netdev_tx_t rtase_start_xmit(struct sk_buff *skb,
+ struct net_device *dev)
+{
+ struct skb_shared_info *shinfo = skb_shinfo(skb);
+ struct rtase_private *tp = netdev_priv(dev);
+ u32 q_idx, entry, len, opts1, opts2;
+ struct netdev_queue *tx_queue;
+ bool stop_queue, door_bell;
+ u32 mss = shinfo->gso_size;
+ struct rtase_tx_desc *txd;
+ struct rtase_ring *ring;
+ dma_addr_t mapping;
+ int frags;
+
+ /* multiqueues */
+ q_idx = skb_get_queue_mapping(skb);
+ ring = &tp->tx_ring[q_idx];
+ tx_queue = netdev_get_tx_queue(dev, q_idx);
+
+ if (unlikely(!rtase_tx_avail(ring))) {
+ if (net_ratelimit())
+ netdev_err(dev, "BUG! Tx Ring full when queue awake!\n");
+ goto err_stop;
+ }
+
+ entry = ring->cur_idx % RTASE_NUM_DESC;
+ txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
+
+ opts1 = RTASE_DESC_OWN;
+ opts2 = rtase_tx_vlan_tag(tp, skb);
+
+ /* tcp segmentation offload (or tcp large send) */
+ if (mss) {
+ if (shinfo->gso_type & SKB_GSO_TCPV4) {
+ opts1 |= RTASE_GIANT_SEND_V4;
+ } else if (shinfo->gso_type & SKB_GSO_TCPV6) {
+ if (skb_cow_head(skb, 0))
+ goto err_dma_0;
+
+ tcp_v6_gso_csum_prep(skb);
+ opts1 |= RTASE_GIANT_SEND_V6;
+ } else {
+ WARN_ON_ONCE(1);
+ }
+
+ opts1 |= u32_encode_bits(skb_transport_offset(skb),
+ RTASE_TCPHO_MASK);
+ opts2 |= u32_encode_bits(mss, RTASE_MSS_MASK);
+ } else if (skb->ip_summed == CHECKSUM_PARTIAL) {
+ opts2 |= rtase_tx_csum(skb, dev);
+ }
+
+ frags = rtase_xmit_frags(ring, skb, opts1, opts2);
+ if (unlikely(frags < 0))
+ goto err_dma_0;
+
+ if (frags) {
+ len = skb_headlen(skb);
+ opts1 |= RTASE_TX_FIRST_FRAG;
+ } else {
+ len = skb->len;
+ ring->skbuff[entry] = skb;
+ opts1 |= RTASE_TX_FIRST_FRAG | RTASE_TX_LAST_FRAG;
+ }
+
+ if (((entry + 1) % RTASE_NUM_DESC) == 0)
+ opts1 |= (len | RTASE_RING_END);
+ else
+ opts1 |= len;
+
+ mapping = dma_map_single(&tp->pdev->dev, skb->data, len,
+ DMA_TO_DEVICE);
+
+ if (unlikely(dma_mapping_error(&tp->pdev->dev, mapping))) {
+ if (unlikely(net_ratelimit()))
+ netdev_err(dev, "Failed to map TX DMA!\n");
+
+ goto err_dma_1;
+ }
+
+ ring->mis.len[entry] = len;
+ txd->addr = cpu_to_le64(mapping);
+ txd->opts2 = cpu_to_le32(opts2);
+ txd->opts1 = cpu_to_le32(opts1 & ~RTASE_DESC_OWN);
+
+ /* make sure the operating fields have been updated */
+ dma_wmb();
+
+ door_bell = __netdev_tx_sent_queue(tx_queue, skb->len,
+ netdev_xmit_more());
+
+ txd->opts1 = cpu_to_le32(opts1);
+
+ skb_tx_timestamp(skb);
+
+ /* tx needs to see descriptor changes before updated cur_idx */
+ smp_wmb();
+
+ WRITE_ONCE(ring->cur_idx, ring->cur_idx + frags + 1);
+
+ stop_queue = !netif_subqueue_maybe_stop(dev, ring->index,
+ rtase_tx_avail(ring),
+ RTASE_TX_STOP_THRS,
+ RTASE_TX_START_THRS);
+
+ if (door_bell || stop_queue)
+ rtase_w8(tp, RTASE_TPPOLL, BIT(ring->index));
+
+ return NETDEV_TX_OK;
+
+err_dma_1:
+ ring->skbuff[entry] = NULL;
+ rtase_tx_clear_range(ring, ring->cur_idx + 1, frags);
+
+err_dma_0:
+ dev->stats.tx_dropped++;
+ dev_kfree_skb_any(skb);
+ return NETDEV_TX_OK;
+
+err_stop:
+ netif_stop_queue(dev);
+ dev->stats.tx_dropped++;
+ return NETDEV_TX_BUSY;
+}
+
static void rtase_enable_eem_write(const struct rtase_private *tp)
{
u8 val;
@@ -1065,6 +1349,7 @@ static void rtase_netpoll(struct net_device *dev)
static const struct net_device_ops rtase_netdev_ops = {
.ndo_open = rtase_open,
.ndo_stop = rtase_close,
+ .ndo_start_xmit = rtase_start_xmit,
#ifdef CONFIG_NET_POLL_CONTROLLER
.ndo_poll_controller = rtase_netpoll,
#endif
--
2.34.1
> Implement .ndo_start_xmit function to fill the information of the packet to be
> transmitted into the tx descriptor, and then the hardware will transmit the
> packet using the information in the tx descriptor.
> In addition, we also implemented the tx_handler function to enable the tx
> descriptor to be reused.
>
> Signed-off-by: Justin Lai <[email protected]>
> ---
> .../net/ethernet/realtek/rtase/rtase_main.c | 285 ++++++++++++++++++
> 1 file changed, 285 insertions(+)
>
> diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> index 23406c195cff..6bdb4edbfbc1 100644
> --- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> +++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> @@ -256,6 +256,68 @@ static void rtase_mark_to_asic(union rtase_rx_desc
> *desc, u32 rx_buf_sz)
> cpu_to_le32(RTASE_DESC_OWN | eor | rx_buf_sz)); }
>
> +static u32 rtase_tx_avail(struct rtase_ring *ring) {
> + return READ_ONCE(ring->dirty_idx) + RTASE_NUM_DESC -
> + READ_ONCE(ring->cur_idx);
> +}
> +
> +static int tx_handler(struct rtase_ring *ring, int budget) {
> + const struct rtase_private *tp = ring->ivec->tp;
> + struct net_device *dev = tp->dev;
> + u32 dirty_tx, tx_left;
> + u32 bytes_compl = 0;
> + u32 pkts_compl = 0;
> + int workdone = 0;
> +
> + dirty_tx = ring->dirty_idx;
> + tx_left = READ_ONCE(ring->cur_idx) - dirty_tx;
> +
> + while (tx_left > 0) {
> + u32 entry = dirty_tx % RTASE_NUM_DESC;
> + struct rtase_tx_desc *desc = ring->desc +
> + sizeof(struct rtase_tx_desc) * entry;
> + u32 status;
> +
> + status = le32_to_cpu(desc->opts1);
> +
> + if (status & RTASE_DESC_OWN)
> + break;
> +
> + rtase_unmap_tx_skb(tp->pdev, ring->mis.len[entry], desc);
> + ring->mis.len[entry] = 0;
> + if (ring->skbuff[entry]) {
> + pkts_compl++;
> + bytes_compl += ring->skbuff[entry]->len;
> + napi_consume_skb(ring->skbuff[entry], budget);
> + ring->skbuff[entry] = NULL;
> + }
> +
> + dirty_tx++;
> + tx_left--;
> + workdone++;
> +
> + if (workdone == RTASE_TX_BUDGET_DEFAULT)
> + break;
> + }
> +
> + if (ring->dirty_idx != dirty_tx) {
> + dev_sw_netstats_tx_add(dev, pkts_compl, bytes_compl);
> + WRITE_ONCE(ring->dirty_idx, dirty_tx);
> +
> + netif_subqueue_completed_wake(dev, ring->index,
> pkts_compl,
> + bytes_compl,
> + rtase_tx_avail(ring),
> + RTASE_TX_START_THRS);
> +
> + if (ring->cur_idx != dirty_tx)
> + rtase_w8(tp, RTASE_TPPOLL, BIT(ring->index));
> + }
> +
> + return 0;
> +}
> +
> static void rtase_tx_desc_init(struct rtase_private *tp, u16 idx) {
> struct rtase_ring *ring = &tp->tx_ring[idx]; @@ -1014,6 +1076,228
> @@ static int rtase_close(struct net_device *dev)
> return 0;
> }
>
> +static u32 rtase_tx_vlan_tag(const struct rtase_private *tp,
> + const struct sk_buff *skb)
> +{
> + return (skb_vlan_tag_present(skb)) ?
> + (RTASE_TX_VLAN_TAG | swab16(skb_vlan_tag_get(skb))) :
> 0x00; }
> +
Vlan protocol can be either 0x8100 or 0x88A8, how does hardware know which header to insert?
Thanks,
Hariprasad k
> +static u32 rtase_tx_csum(struct sk_buff *skb, const struct net_device
> +*dev) {
> + u32 csum_cmd = 0;
> + u8 ip_protocol;
> +
> + switch (vlan_get_protocol(skb)) {
> + case htons(ETH_P_IP):
> + csum_cmd = RTASE_TX_IPCS_C;
> + ip_protocol = ip_hdr(skb)->protocol;
> + break;
> +
> + case htons(ETH_P_IPV6):
> + csum_cmd = RTASE_TX_IPV6F_C;
> + ip_protocol = ipv6_hdr(skb)->nexthdr;
> + break;
> +
> + default:
> + ip_protocol = IPPROTO_RAW;
> + break;
> + }
> +
> + if (ip_protocol == IPPROTO_TCP)
> + csum_cmd |= RTASE_TX_TCPCS_C;
> + else if (ip_protocol == IPPROTO_UDP)
> + csum_cmd |= RTASE_TX_UDPCS_C;
> +
> + csum_cmd |= u32_encode_bits(skb_transport_offset(skb),
> + RTASE_TCPHO_MASK);
> +
> + return csum_cmd;
> +}
> +
> +static int rtase_xmit_frags(struct rtase_ring *ring, struct sk_buff *skb,
> + u32 opts1, u32 opts2)
> +{
> + const struct skb_shared_info *info = skb_shinfo(skb);
> + const struct rtase_private *tp = ring->ivec->tp;
> + const u8 nr_frags = info->nr_frags;
> + struct rtase_tx_desc *txd = NULL;
> + u32 cur_frag, entry;
> +
> + entry = ring->cur_idx;
> + for (cur_frag = 0; cur_frag < nr_frags; cur_frag++) {
> + const skb_frag_t *frag = &info->frags[cur_frag];
> + dma_addr_t mapping;
> + u32 status, len;
> + void *addr;
> +
> + entry = (entry + 1) % RTASE_NUM_DESC;
> +
> + txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
> + len = skb_frag_size(frag);
> + addr = skb_frag_address(frag);
> + mapping = dma_map_single(&tp->pdev->dev, addr, len,
> + DMA_TO_DEVICE);
> +
> + if (unlikely(dma_mapping_error(&tp->pdev->dev, mapping)))
> {
> + if (unlikely(net_ratelimit()))
> + netdev_err(tp->dev,
> + "Failed to map TX fragments
> DMA!\n");
> +
> + goto err_out;
> + }
> +
> + if (((entry + 1) % RTASE_NUM_DESC) == 0)
> + status = (opts1 | len | RTASE_RING_END);
> + else
> + status = opts1 | len;
> +
> + if (cur_frag == (nr_frags - 1)) {
> + ring->skbuff[entry] = skb;
> + status |= RTASE_TX_LAST_FRAG;
> + }
> +
> + ring->mis.len[entry] = len;
> + txd->addr = cpu_to_le64(mapping);
> + txd->opts2 = cpu_to_le32(opts2);
> +
> + /* make sure the operating fields have been updated */
> + dma_wmb();
> + txd->opts1 = cpu_to_le32(status);
> + }
> +
> + return cur_frag;
> +
> +err_out:
> + rtase_tx_clear_range(ring, ring->cur_idx + 1, cur_frag);
> + return -EIO;
> +}
> +
> +static netdev_tx_t rtase_start_xmit(struct sk_buff *skb,
> + struct net_device *dev)
> +{
> + struct skb_shared_info *shinfo = skb_shinfo(skb);
> + struct rtase_private *tp = netdev_priv(dev);
> + u32 q_idx, entry, len, opts1, opts2;
> + struct netdev_queue *tx_queue;
> + bool stop_queue, door_bell;
> + u32 mss = shinfo->gso_size;
> + struct rtase_tx_desc *txd;
> + struct rtase_ring *ring;
> + dma_addr_t mapping;
> + int frags;
> +
> + /* multiqueues */
> + q_idx = skb_get_queue_mapping(skb);
> + ring = &tp->tx_ring[q_idx];
> + tx_queue = netdev_get_tx_queue(dev, q_idx);
> +
> + if (unlikely(!rtase_tx_avail(ring))) {
> + if (net_ratelimit())
> + netdev_err(dev, "BUG! Tx Ring full when queue
> awake!\n");
> + goto err_stop;
> + }
> +
> + entry = ring->cur_idx % RTASE_NUM_DESC;
> + txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
> +
> + opts1 = RTASE_DESC_OWN;
> + opts2 = rtase_tx_vlan_tag(tp, skb);
> +
> + /* tcp segmentation offload (or tcp large send) */
> + if (mss) {
> + if (shinfo->gso_type & SKB_GSO_TCPV4) {
> + opts1 |= RTASE_GIANT_SEND_V4;
> + } else if (shinfo->gso_type & SKB_GSO_TCPV6) {
> + if (skb_cow_head(skb, 0))
> + goto err_dma_0;
> +
> + tcp_v6_gso_csum_prep(skb);
> + opts1 |= RTASE_GIANT_SEND_V6;
> + } else {
> + WARN_ON_ONCE(1);
> + }
> +
> + opts1 |= u32_encode_bits(skb_transport_offset(skb),
> + RTASE_TCPHO_MASK);
> + opts2 |= u32_encode_bits(mss, RTASE_MSS_MASK);
> + } else if (skb->ip_summed == CHECKSUM_PARTIAL) {
> + opts2 |= rtase_tx_csum(skb, dev);
> + }
> +
> + frags = rtase_xmit_frags(ring, skb, opts1, opts2);
> + if (unlikely(frags < 0))
> + goto err_dma_0;
> +
> + if (frags) {
> + len = skb_headlen(skb);
> + opts1 |= RTASE_TX_FIRST_FRAG;
> + } else {
> + len = skb->len;
> + ring->skbuff[entry] = skb;
> + opts1 |= RTASE_TX_FIRST_FRAG | RTASE_TX_LAST_FRAG;
> + }
> +
> + if (((entry + 1) % RTASE_NUM_DESC) == 0)
> + opts1 |= (len | RTASE_RING_END);
> + else
> + opts1 |= len;
> +
> + mapping = dma_map_single(&tp->pdev->dev, skb->data, len,
> + DMA_TO_DEVICE);
> +
> + if (unlikely(dma_mapping_error(&tp->pdev->dev, mapping))) {
> + if (unlikely(net_ratelimit()))
> + netdev_err(dev, "Failed to map TX DMA!\n");
> +
> + goto err_dma_1;
> + }
> +
> + ring->mis.len[entry] = len;
> + txd->addr = cpu_to_le64(mapping);
> + txd->opts2 = cpu_to_le32(opts2);
> + txd->opts1 = cpu_to_le32(opts1 & ~RTASE_DESC_OWN);
> +
> + /* make sure the operating fields have been updated */
> + dma_wmb();
> +
> + door_bell = __netdev_tx_sent_queue(tx_queue, skb->len,
> + netdev_xmit_more());
> +
> + txd->opts1 = cpu_to_le32(opts1);
> +
> + skb_tx_timestamp(skb);
> +
> + /* tx needs to see descriptor changes before updated cur_idx */
> + smp_wmb();
> +
> + WRITE_ONCE(ring->cur_idx, ring->cur_idx + frags + 1);
> +
> + stop_queue = !netif_subqueue_maybe_stop(dev, ring->index,
> + rtase_tx_avail(ring),
> + RTASE_TX_STOP_THRS,
> + RTASE_TX_START_THRS);
> +
> + if (door_bell || stop_queue)
> + rtase_w8(tp, RTASE_TPPOLL, BIT(ring->index));
> +
> + return NETDEV_TX_OK;
> +
> +err_dma_1:
> + ring->skbuff[entry] = NULL;
> + rtase_tx_clear_range(ring, ring->cur_idx + 1, frags);
> +
> +err_dma_0:
> + dev->stats.tx_dropped++;
> + dev_kfree_skb_any(skb);
> + return NETDEV_TX_OK;
> +
> +err_stop:
> + netif_stop_queue(dev);
> + dev->stats.tx_dropped++;
> + return NETDEV_TX_BUSY;
> +}
> +
> static void rtase_enable_eem_write(const struct rtase_private *tp) {
> u8 val;
> @@ -1065,6 +1349,7 @@ static void rtase_netpoll(struct net_device *dev)
> static const struct net_device_ops rtase_netdev_ops = {
> .ndo_open = rtase_open,
> .ndo_stop = rtase_close,
> + .ndo_start_xmit = rtase_start_xmit,
> #ifdef CONFIG_NET_POLL_CONTROLLER
> .ndo_poll_controller = rtase_netpoll,
> #endif
> --
> 2.34.1
>
On 2024-06-07 at 14:13:14, Justin Lai ([email protected]) wrote:
> Implement .ndo_start_xmit function to fill the information of the packet
> to be transmitted into the tx descriptor, and then the hardware will
> transmit the packet using the information in the tx descriptor.
> In addition, we also implemented the tx_handler function to enable the
> tx descriptor to be reused.
>
> Signed-off-by: Justin Lai <[email protected]>
> ---
> .../net/ethernet/realtek/rtase/rtase_main.c | 285 ++++++++++++++++++
> 1 file changed, 285 insertions(+)
>
> diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> index 23406c195cff..6bdb4edbfbc1 100644
> --- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> +++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> @@ -256,6 +256,68 @@ static void rtase_mark_to_asic(union rtase_rx_desc *desc, u32 rx_buf_sz)
> cpu_to_le32(RTASE_DESC_OWN | eor | rx_buf_sz));
> }
>
> +static u32 rtase_tx_avail(struct rtase_ring *ring)
> +{
> + return READ_ONCE(ring->dirty_idx) + RTASE_NUM_DESC -
> + READ_ONCE(ring->cur_idx);
> +}
dirty_idx and cur_idx wont wrap ? its 32bit in size.
>
> On 2024-06-07 at 14:13:14, Justin Lai ([email protected]) wrote:
> > Implement .ndo_start_xmit function to fill the information of the
> > packet to be transmitted into the tx descriptor, and then the hardware
> > will transmit the packet using the information in the tx descriptor.
> > In addition, we also implemented the tx_handler function to enable the
> > tx descriptor to be reused.
> >
> > Signed-off-by: Justin Lai <[email protected]>
> > ---
> > .../net/ethernet/realtek/rtase/rtase_main.c | 285 ++++++++++++++++++
> > 1 file changed, 285 insertions(+)
> >
> > diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > index 23406c195cff..6bdb4edbfbc1 100644
> > --- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > +++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > @@ -256,6 +256,68 @@ static void rtase_mark_to_asic(union rtase_rx_desc
> *desc, u32 rx_buf_sz)
> > cpu_to_le32(RTASE_DESC_OWN | eor | rx_buf_sz)); }
> >
> > +static u32 rtase_tx_avail(struct rtase_ring *ring) {
> > + return READ_ONCE(ring->dirty_idx) + RTASE_NUM_DESC -
> > + READ_ONCE(ring->cur_idx); }
> dirty_idx and cur_idx wont wrap ? its 32bit in size.
>
> >
cur_idx and dirty_idx may wrap, but all we want is the difference between
them, so this won't have any effect. In addition, the difference between
the two will not exceed RTASE_NUM_DESC, and dirty_idx will not exceed
cur_idx, so the calculation won't go wrong.
> > Implement .ndo_start_xmit function to fill the information of the
> > packet to be transmitted into the tx descriptor, and then the hardware
> > will transmit the packet using the information in the tx descriptor.
> > In addition, we also implemented the tx_handler function to enable the
> > tx descriptor to be reused.
> >
> > Signed-off-by: Justin Lai <[email protected]>
> > ---
> > .../net/ethernet/realtek/rtase/rtase_main.c | 285 ++++++++++++++++++
> > 1 file changed, 285 insertions(+)
> >
> > diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > index 23406c195cff..6bdb4edbfbc1 100644
> > --- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > +++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > @@ -256,6 +256,68 @@ static void rtase_mark_to_asic(union
> > rtase_rx_desc *desc, u32 rx_buf_sz)
> > cpu_to_le32(RTASE_DESC_OWN | eor | rx_buf_sz)); }
> >
> > +static u32 rtase_tx_avail(struct rtase_ring *ring) {
> > + return READ_ONCE(ring->dirty_idx) + RTASE_NUM_DESC -
> > + READ_ONCE(ring->cur_idx); }
> > +
> > +static int tx_handler(struct rtase_ring *ring, int budget) {
> > + const struct rtase_private *tp = ring->ivec->tp;
> > + struct net_device *dev = tp->dev;
> > + u32 dirty_tx, tx_left;
> > + u32 bytes_compl = 0;
> > + u32 pkts_compl = 0;
> > + int workdone = 0;
> > +
> > + dirty_tx = ring->dirty_idx;
> > + tx_left = READ_ONCE(ring->cur_idx) - dirty_tx;
> > +
> > + while (tx_left > 0) {
> > + u32 entry = dirty_tx % RTASE_NUM_DESC;
> > + struct rtase_tx_desc *desc = ring->desc +
> > + sizeof(struct rtase_tx_desc) *
> entry;
> > + u32 status;
> > +
> > + status = le32_to_cpu(desc->opts1);
> > +
> > + if (status & RTASE_DESC_OWN)
> > + break;
> > +
> > + rtase_unmap_tx_skb(tp->pdev, ring->mis.len[entry], desc);
> > + ring->mis.len[entry] = 0;
> > + if (ring->skbuff[entry]) {
> > + pkts_compl++;
> > + bytes_compl += ring->skbuff[entry]->len;
> > + napi_consume_skb(ring->skbuff[entry], budget);
> > + ring->skbuff[entry] = NULL;
> > + }
> > +
> > + dirty_tx++;
> > + tx_left--;
> > + workdone++;
> > +
> > + if (workdone == RTASE_TX_BUDGET_DEFAULT)
> > + break;
> > + }
> > +
> > + if (ring->dirty_idx != dirty_tx) {
> > + dev_sw_netstats_tx_add(dev, pkts_compl, bytes_compl);
> > + WRITE_ONCE(ring->dirty_idx, dirty_tx);
> > +
> > + netif_subqueue_completed_wake(dev, ring->index,
> > pkts_compl,
> > + bytes_compl,
> > + rtase_tx_avail(ring),
> > +
> RTASE_TX_START_THRS);
> > +
> > + if (ring->cur_idx != dirty_tx)
> > + rtase_w8(tp, RTASE_TPPOLL, BIT(ring->index));
> > + }
> > +
> > + return 0;
> > +}
> > +
> > static void rtase_tx_desc_init(struct rtase_private *tp, u16 idx) {
> > struct rtase_ring *ring = &tp->tx_ring[idx]; @@ -1014,6
> > +1076,228 @@ static int rtase_close(struct net_device *dev)
> > return 0;
> > }
> >
> > +static u32 rtase_tx_vlan_tag(const struct rtase_private *tp,
> > + const struct sk_buff *skb) {
> > + return (skb_vlan_tag_present(skb)) ?
> > + (RTASE_TX_VLAN_TAG | swab16(skb_vlan_tag_get(skb))) :
> > 0x00; }
> > +
> Vlan protocol can be either 0x8100 or 0x88A8, how does
> hardware know which header to insert?
> Thanks,
> Hariprasad k
We only allow the hardware to add 0x8100, the VLAN must at least have
0x8100 to potentially have 0x88a8, skb_vlan_tag_present indicates that
VLAN exists, hence at least the 0x8100 VLAN would exist.
>
> > +static u32 rtase_tx_csum(struct sk_buff *skb, const struct net_device
> > +*dev) {
> > + u32 csum_cmd = 0;
> > + u8 ip_protocol;
> > +
> > + switch (vlan_get_protocol(skb)) {
> > + case htons(ETH_P_IP):
> > + csum_cmd = RTASE_TX_IPCS_C;
> > + ip_protocol = ip_hdr(skb)->protocol;
> > + break;
> > +
> > + case htons(ETH_P_IPV6):
> > + csum_cmd = RTASE_TX_IPV6F_C;
> > + ip_protocol = ipv6_hdr(skb)->nexthdr;
> > + break;
> > +
> > + default:
> > + ip_protocol = IPPROTO_RAW;
> > + break;
> > + }
> > +
> > + if (ip_protocol == IPPROTO_TCP)
> > + csum_cmd |= RTASE_TX_TCPCS_C;
> > + else if (ip_protocol == IPPROTO_UDP)
> > + csum_cmd |= RTASE_TX_UDPCS_C;
> > +
> > + csum_cmd |= u32_encode_bits(skb_transport_offset(skb),
> > + RTASE_TCPHO_MASK);
> > +
> > + return csum_cmd;
> > +}
> > +
> > +static int rtase_xmit_frags(struct rtase_ring *ring, struct sk_buff *skb,
> > + u32 opts1, u32 opts2) {
> > + const struct skb_shared_info *info = skb_shinfo(skb);
> > + const struct rtase_private *tp = ring->ivec->tp;
> > + const u8 nr_frags = info->nr_frags;
> > + struct rtase_tx_desc *txd = NULL;
> > + u32 cur_frag, entry;
> > +
> > + entry = ring->cur_idx;
> > + for (cur_frag = 0; cur_frag < nr_frags; cur_frag++) {
> > + const skb_frag_t *frag = &info->frags[cur_frag];
> > + dma_addr_t mapping;
> > + u32 status, len;
> > + void *addr;
> > +
> > + entry = (entry + 1) % RTASE_NUM_DESC;
> > +
> > + txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
> > + len = skb_frag_size(frag);
> > + addr = skb_frag_address(frag);
> > + mapping = dma_map_single(&tp->pdev->dev, addr, len,
> > + DMA_TO_DEVICE);
> > +
> > + if (unlikely(dma_mapping_error(&tp->pdev->dev,
> > + mapping)))
> > {
> > + if (unlikely(net_ratelimit()))
> > + netdev_err(tp->dev,
> > + "Failed to map TX
> fragments
> > DMA!\n");
> > +
> > + goto err_out;
> > + }
> > +
> > + if (((entry + 1) % RTASE_NUM_DESC) == 0)
> > + status = (opts1 | len | RTASE_RING_END);
> > + else
> > + status = opts1 | len;
> > +
> > + if (cur_frag == (nr_frags - 1)) {
> > + ring->skbuff[entry] = skb;
> > + status |= RTASE_TX_LAST_FRAG;
> > + }
> > +
> > + ring->mis.len[entry] = len;
> > + txd->addr = cpu_to_le64(mapping);
> > + txd->opts2 = cpu_to_le32(opts2);
> > +
> > + /* make sure the operating fields have been updated */
> > + dma_wmb();
> > + txd->opts1 = cpu_to_le32(status);
> > + }
> > +
> > + return cur_frag;
> > +
> > +err_out:
> > + rtase_tx_clear_range(ring, ring->cur_idx + 1, cur_frag);
> > + return -EIO;
> > +}
> > +
> > +static netdev_tx_t rtase_start_xmit(struct sk_buff *skb,
> > + struct net_device *dev) {
> > + struct skb_shared_info *shinfo = skb_shinfo(skb);
> > + struct rtase_private *tp = netdev_priv(dev);
> > + u32 q_idx, entry, len, opts1, opts2;
> > + struct netdev_queue *tx_queue;
> > + bool stop_queue, door_bell;
> > + u32 mss = shinfo->gso_size;
> > + struct rtase_tx_desc *txd;
> > + struct rtase_ring *ring;
> > + dma_addr_t mapping;
> > + int frags;
> > +
> > + /* multiqueues */
> > + q_idx = skb_get_queue_mapping(skb);
> > + ring = &tp->tx_ring[q_idx];
> > + tx_queue = netdev_get_tx_queue(dev, q_idx);
> > +
> > + if (unlikely(!rtase_tx_avail(ring))) {
> > + if (net_ratelimit())
> > + netdev_err(dev, "BUG! Tx Ring full when queue
> > awake!\n");
> > + goto err_stop;
> > + }
> > +
> > + entry = ring->cur_idx % RTASE_NUM_DESC;
> > + txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
> > +
> > + opts1 = RTASE_DESC_OWN;
> > + opts2 = rtase_tx_vlan_tag(tp, skb);
> > +
> > + /* tcp segmentation offload (or tcp large send) */
> > + if (mss) {
> > + if (shinfo->gso_type & SKB_GSO_TCPV4) {
> > + opts1 |= RTASE_GIANT_SEND_V4;
> > + } else if (shinfo->gso_type & SKB_GSO_TCPV6) {
> > + if (skb_cow_head(skb, 0))
> > + goto err_dma_0;
> > +
> > + tcp_v6_gso_csum_prep(skb);
> > + opts1 |= RTASE_GIANT_SEND_V6;
> > + } else {
> > + WARN_ON_ONCE(1);
> > + }
> > +
> > + opts1 |= u32_encode_bits(skb_transport_offset(skb),
> > + RTASE_TCPHO_MASK);
> > + opts2 |= u32_encode_bits(mss, RTASE_MSS_MASK);
> > + } else if (skb->ip_summed == CHECKSUM_PARTIAL) {
> > + opts2 |= rtase_tx_csum(skb, dev);
> > + }
> > +
> > + frags = rtase_xmit_frags(ring, skb, opts1, opts2);
> > + if (unlikely(frags < 0))
> > + goto err_dma_0;
> > +
> > + if (frags) {
> > + len = skb_headlen(skb);
> > + opts1 |= RTASE_TX_FIRST_FRAG;
> > + } else {
> > + len = skb->len;
> > + ring->skbuff[entry] = skb;
> > + opts1 |= RTASE_TX_FIRST_FRAG | RTASE_TX_LAST_FRAG;
> > + }
> > +
> > + if (((entry + 1) % RTASE_NUM_DESC) == 0)
> > + opts1 |= (len | RTASE_RING_END);
> > + else
> > + opts1 |= len;
> > +
> > + mapping = dma_map_single(&tp->pdev->dev, skb->data, len,
> > + DMA_TO_DEVICE);
> > +
> > + if (unlikely(dma_mapping_error(&tp->pdev->dev, mapping))) {
> > + if (unlikely(net_ratelimit()))
> > + netdev_err(dev, "Failed to map TX DMA!\n");
> > +
> > + goto err_dma_1;
> > + }
> > +
> > + ring->mis.len[entry] = len;
> > + txd->addr = cpu_to_le64(mapping);
> > + txd->opts2 = cpu_to_le32(opts2);
> > + txd->opts1 = cpu_to_le32(opts1 & ~RTASE_DESC_OWN);
> > +
> > + /* make sure the operating fields have been updated */
> > + dma_wmb();
> > +
> > + door_bell = __netdev_tx_sent_queue(tx_queue, skb->len,
> > + netdev_xmit_more());
> > +
> > + txd->opts1 = cpu_to_le32(opts1);
> > +
> > + skb_tx_timestamp(skb);
> > +
> > + /* tx needs to see descriptor changes before updated cur_idx */
> > + smp_wmb();
> > +
> > + WRITE_ONCE(ring->cur_idx, ring->cur_idx + frags + 1);
> > +
> > + stop_queue = !netif_subqueue_maybe_stop(dev, ring->index,
> > + rtase_tx_avail(ring),
> > +
> RTASE_TX_STOP_THRS,
> > +
> RTASE_TX_START_THRS);
> > +
> > + if (door_bell || stop_queue)
> > + rtase_w8(tp, RTASE_TPPOLL, BIT(ring->index));
> > +
> > + return NETDEV_TX_OK;
> > +
> > +err_dma_1:
> > + ring->skbuff[entry] = NULL;
> > + rtase_tx_clear_range(ring, ring->cur_idx + 1, frags);
> > +
> > +err_dma_0:
> > + dev->stats.tx_dropped++;
> > + dev_kfree_skb_any(skb);
> > + return NETDEV_TX_OK;
> > +
> > +err_stop:
> > + netif_stop_queue(dev);
> > + dev->stats.tx_dropped++;
> > + return NETDEV_TX_BUSY;
> > +}
> > +
> > static void rtase_enable_eem_write(const struct rtase_private *tp) {
> > u8 val;
> > @@ -1065,6 +1349,7 @@ static void rtase_netpoll(struct net_device
> > *dev) static const struct net_device_ops rtase_netdev_ops = {
> > .ndo_open = rtase_open,
> > .ndo_stop = rtase_close,
> > + .ndo_start_xmit = rtase_start_xmit,
> > #ifdef CONFIG_NET_POLL_CONTROLLER
> > .ndo_poll_controller = rtase_netpoll, #endif
> > --
> > 2.34.1
> >
> > > Implement .ndo_start_xmit function to fill the information of the
> > > packet to be transmitted into the tx descriptor, and then the
> > > hardware will transmit the packet using the information in the tx
> descriptor.
> > > In addition, we also implemented the tx_handler function to enable
> > > the tx descriptor to be reused.
> > >
> > > Signed-off-by: Justin Lai <[email protected]>
> > > ---
> > > .../net/ethernet/realtek/rtase/rtase_main.c | 285 ++++++++++++++++++
> > > 1 file changed, 285 insertions(+)
> > >
> > > diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > index 23406c195cff..6bdb4edbfbc1 100644
> > > --- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > +++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > @@ -256,6 +256,68 @@ static void rtase_mark_to_asic(union
> > > rtase_rx_desc *desc, u32 rx_buf_sz)
> > > cpu_to_le32(RTASE_DESC_OWN | eor | rx_buf_sz)); }
> > >
> > > +static u32 rtase_tx_avail(struct rtase_ring *ring) {
> > > + return READ_ONCE(ring->dirty_idx) + RTASE_NUM_DESC -
> > > + READ_ONCE(ring->cur_idx); }
> > > +
> > > +static int tx_handler(struct rtase_ring *ring, int budget) {
> > > + const struct rtase_private *tp = ring->ivec->tp;
> > > + struct net_device *dev = tp->dev;
> > > + u32 dirty_tx, tx_left;
> > > + u32 bytes_compl = 0;
> > > + u32 pkts_compl = 0;
> > > + int workdone = 0;
> > > +
> > > + dirty_tx = ring->dirty_idx;
> > > + tx_left = READ_ONCE(ring->cur_idx) - dirty_tx;
> > > +
> > > + while (tx_left > 0) {
> > > + u32 entry = dirty_tx % RTASE_NUM_DESC;
> > > + struct rtase_tx_desc *desc = ring->desc +
> > > + sizeof(struct rtase_tx_desc) *
> > entry;
> > > + u32 status;
> > > +
> > > + status = le32_to_cpu(desc->opts1);
> > > +
> > > + if (status & RTASE_DESC_OWN)
> > > + break;
> > > +
> > > + rtase_unmap_tx_skb(tp->pdev, ring->mis.len[entry], desc);
> > > + ring->mis.len[entry] = 0;
> > > + if (ring->skbuff[entry]) {
> > > + pkts_compl++;
> > > + bytes_compl += ring->skbuff[entry]->len;
> > > + napi_consume_skb(ring->skbuff[entry], budget);
> > > + ring->skbuff[entry] = NULL;
> > > + }
> > > +
> > > + dirty_tx++;
> > > + tx_left--;
> > > + workdone++;
> > > +
> > > + if (workdone == RTASE_TX_BUDGET_DEFAULT)
> > > + break;
> > > + }
> > > +
> > > + if (ring->dirty_idx != dirty_tx) {
> > > + dev_sw_netstats_tx_add(dev, pkts_compl, bytes_compl);
> > > + WRITE_ONCE(ring->dirty_idx, dirty_tx);
> > > +
> > > + netif_subqueue_completed_wake(dev, ring->index,
> > > pkts_compl,
> > > + bytes_compl,
> > > + rtase_tx_avail(ring),
> > > +
> > RTASE_TX_START_THRS);
> > > +
> > > + if (ring->cur_idx != dirty_tx)
> > > + rtase_w8(tp, RTASE_TPPOLL, BIT(ring->index));
> > > + }
> > > +
> > > + return 0;
> > > +}
> > > +
> > > static void rtase_tx_desc_init(struct rtase_private *tp, u16 idx) {
> > > struct rtase_ring *ring = &tp->tx_ring[idx]; @@ -1014,6
> > > +1076,228 @@ static int rtase_close(struct net_device *dev)
> > > return 0;
> > > }
> > >
> > > +static u32 rtase_tx_vlan_tag(const struct rtase_private *tp,
> > > + const struct sk_buff *skb) {
> > > + return (skb_vlan_tag_present(skb)) ?
> > > + (RTASE_TX_VLAN_TAG | swab16(skb_vlan_tag_get(skb))) :
> > > 0x00; }
> > > +
> > Vlan protocol can be either 0x8100 or 0x88A8, how does
> > hardware know which header to insert?
> > Thanks,
> > Hariprasad k
>
> We only allow the hardware to add 0x8100, the VLAN must at least have
> 0x8100 to potentially have 0x88a8, skb_vlan_tag_present indicates that VLAN
> exists, hence at least the 0x8100 VLAN would exist.
> >
Thanks for the explanation, but one question which bothers me is that "how hardware knows offset with in the packet"
For example
Case 1: DMAC + SMAC + 8100 VLAN_ID + IP
Here offset is right after the SMAC.
Case 2: DMAC + SMAC + 88A8 VLAN_ID + 8100 VLAN_ID + IP
Here offset is right after first vlan tag.
Thanks,
Hariprasad k
> > > +static u32 rtase_tx_csum(struct sk_buff *skb, const struct
> > > +net_device
> > > +*dev) {
> > > + u32 csum_cmd = 0;
> > > + u8 ip_protocol;
> > > +
> > > + switch (vlan_get_protocol(skb)) {
> > > + case htons(ETH_P_IP):
> > > + csum_cmd = RTASE_TX_IPCS_C;
> > > + ip_protocol = ip_hdr(skb)->protocol;
> > > + break;
> > > +
> > > + case htons(ETH_P_IPV6):
> > > + csum_cmd = RTASE_TX_IPV6F_C;
> > > + ip_protocol = ipv6_hdr(skb)->nexthdr;
> > > + break;
> > > +
> > > + default:
> > > + ip_protocol = IPPROTO_RAW;
> > > + break;
> > > + }
> > > +
> > > + if (ip_protocol == IPPROTO_TCP)
> > > + csum_cmd |= RTASE_TX_TCPCS_C;
> > > + else if (ip_protocol == IPPROTO_UDP)
> > > + csum_cmd |= RTASE_TX_UDPCS_C;
> > > +
> > > + csum_cmd |= u32_encode_bits(skb_transport_offset(skb),
> > > + RTASE_TCPHO_MASK);
> > > +
> > > + return csum_cmd;
> > > +}
> > > +
> > > +static int rtase_xmit_frags(struct rtase_ring *ring, struct sk_buff *skb,
> > > + u32 opts1, u32 opts2) {
> > > + const struct skb_shared_info *info = skb_shinfo(skb);
> > > + const struct rtase_private *tp = ring->ivec->tp;
> > > + const u8 nr_frags = info->nr_frags;
> > > + struct rtase_tx_desc *txd = NULL;
> > > + u32 cur_frag, entry;
> > > +
> > > + entry = ring->cur_idx;
> > > + for (cur_frag = 0; cur_frag < nr_frags; cur_frag++) {
> > > + const skb_frag_t *frag = &info->frags[cur_frag];
> > > + dma_addr_t mapping;
> > > + u32 status, len;
> > > + void *addr;
> > > +
> > > + entry = (entry + 1) % RTASE_NUM_DESC;
> > > +
> > > + txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
> > > + len = skb_frag_size(frag);
> > > + addr = skb_frag_address(frag);
> > > + mapping = dma_map_single(&tp->pdev->dev, addr, len,
> > > + DMA_TO_DEVICE);
> > > +
> > > + if (unlikely(dma_mapping_error(&tp->pdev->dev,
> > > + mapping)))
> > > {
> > > + if (unlikely(net_ratelimit()))
> > > + netdev_err(tp->dev,
> > > + "Failed to map TX
> > fragments
> > > DMA!\n");
> > > +
> > > + goto err_out;
> > > + }
> > > +
> > > + if (((entry + 1) % RTASE_NUM_DESC) == 0)
> > > + status = (opts1 | len | RTASE_RING_END);
> > > + else
> > > + status = opts1 | len;
> > > +
> > > + if (cur_frag == (nr_frags - 1)) {
> > > + ring->skbuff[entry] = skb;
> > > + status |= RTASE_TX_LAST_FRAG;
> > > + }
> > > +
> > > + ring->mis.len[entry] = len;
> > > + txd->addr = cpu_to_le64(mapping);
> > > + txd->opts2 = cpu_to_le32(opts2);
> > > +
> > > + /* make sure the operating fields have been updated */
> > > + dma_wmb();
> > > + txd->opts1 = cpu_to_le32(status);
> > > + }
> > > +
> > > + return cur_frag;
> > > +
> > > +err_out:
> > > + rtase_tx_clear_range(ring, ring->cur_idx + 1, cur_frag);
> > > + return -EIO;
> > > +}
> > > +
> > > +static netdev_tx_t rtase_start_xmit(struct sk_buff *skb,
> > > + struct net_device *dev) {
> > > + struct skb_shared_info *shinfo = skb_shinfo(skb);
> > > + struct rtase_private *tp = netdev_priv(dev);
> > > + u32 q_idx, entry, len, opts1, opts2;
> > > + struct netdev_queue *tx_queue;
> > > + bool stop_queue, door_bell;
> > > + u32 mss = shinfo->gso_size;
> > > + struct rtase_tx_desc *txd;
> > > + struct rtase_ring *ring;
> > > + dma_addr_t mapping;
> > > + int frags;
> > > +
> > > + /* multiqueues */
> > > + q_idx = skb_get_queue_mapping(skb);
> > > + ring = &tp->tx_ring[q_idx];
> > > + tx_queue = netdev_get_tx_queue(dev, q_idx);
> > > +
> > > + if (unlikely(!rtase_tx_avail(ring))) {
> > > + if (net_ratelimit())
> > > + netdev_err(dev, "BUG! Tx Ring full when queue
> > > awake!\n");
> > > + goto err_stop;
> > > + }
> > > +
> > > + entry = ring->cur_idx % RTASE_NUM_DESC;
> > > + txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
> > > +
> > > + opts1 = RTASE_DESC_OWN;
> > > + opts2 = rtase_tx_vlan_tag(tp, skb);
> > > +
> > > + /* tcp segmentation offload (or tcp large send) */
> > > + if (mss) {
> > > + if (shinfo->gso_type & SKB_GSO_TCPV4) {
> > > + opts1 |= RTASE_GIANT_SEND_V4;
> > > + } else if (shinfo->gso_type & SKB_GSO_TCPV6) {
> > > + if (skb_cow_head(skb, 0))
> > > + goto err_dma_0;
> > > +
> > > + tcp_v6_gso_csum_prep(skb);
> > > + opts1 |= RTASE_GIANT_SEND_V6;
> > > + } else {
> > > + WARN_ON_ONCE(1);
> > > + }
> > > +
> > > + opts1 |= u32_encode_bits(skb_transport_offset(skb),
> > > + RTASE_TCPHO_MASK);
> > > + opts2 |= u32_encode_bits(mss, RTASE_MSS_MASK);
> > > + } else if (skb->ip_summed == CHECKSUM_PARTIAL) {
> > > + opts2 |= rtase_tx_csum(skb, dev);
> > > + }
> > > +
> > > + frags = rtase_xmit_frags(ring, skb, opts1, opts2);
> > > + if (unlikely(frags < 0))
> > > + goto err_dma_0;
> > > +
> > > + if (frags) {
> > > + len = skb_headlen(skb);
> > > + opts1 |= RTASE_TX_FIRST_FRAG;
> > > + } else {
> > > + len = skb->len;
> > > + ring->skbuff[entry] = skb;
> > > + opts1 |= RTASE_TX_FIRST_FRAG | RTASE_TX_LAST_FRAG;
> > > + }
> > > +
> > > + if (((entry + 1) % RTASE_NUM_DESC) == 0)
> > > + opts1 |= (len | RTASE_RING_END);
> > > + else
> > > + opts1 |= len;
> > > +
> > > + mapping = dma_map_single(&tp->pdev->dev, skb->data, len,
> > > + DMA_TO_DEVICE);
> > > +
> > > + if (unlikely(dma_mapping_error(&tp->pdev->dev, mapping))) {
> > > + if (unlikely(net_ratelimit()))
> > > + netdev_err(dev, "Failed to map TX DMA!\n");
> > > +
> > > + goto err_dma_1;
> > > + }
> > > +
> > > + ring->mis.len[entry] = len;
> > > + txd->addr = cpu_to_le64(mapping);
> > > + txd->opts2 = cpu_to_le32(opts2);
> > > + txd->opts1 = cpu_to_le32(opts1 & ~RTASE_DESC_OWN);
> > > +
> > > + /* make sure the operating fields have been updated */
> > > + dma_wmb();
> > > +
> > > + door_bell = __netdev_tx_sent_queue(tx_queue, skb->len,
> > > + netdev_xmit_more());
> > > +
> > > + txd->opts1 = cpu_to_le32(opts1);
> > > +
> > > + skb_tx_timestamp(skb);
> > > +
> > > + /* tx needs to see descriptor changes before updated cur_idx */
> > > + smp_wmb();
> > > +
> > > + WRITE_ONCE(ring->cur_idx, ring->cur_idx + frags + 1);
> > > +
> > > + stop_queue = !netif_subqueue_maybe_stop(dev, ring->index,
> > > + rtase_tx_avail(ring),
> > > +
> > RTASE_TX_STOP_THRS,
> > > +
> > RTASE_TX_START_THRS);
> > > +
> > > + if (door_bell || stop_queue)
> > > + rtase_w8(tp, RTASE_TPPOLL, BIT(ring->index));
> > > +
> > > + return NETDEV_TX_OK;
> > > +
> > > +err_dma_1:
> > > + ring->skbuff[entry] = NULL;
> > > + rtase_tx_clear_range(ring, ring->cur_idx + 1, frags);
> > > +
> > > +err_dma_0:
> > > + dev->stats.tx_dropped++;
> > > + dev_kfree_skb_any(skb);
> > > + return NETDEV_TX_OK;
> > > +
> > > +err_stop:
> > > + netif_stop_queue(dev);
> > > + dev->stats.tx_dropped++;
> > > + return NETDEV_TX_BUSY;
> > > +}
> > > +
> > > static void rtase_enable_eem_write(const struct rtase_private *tp) {
> > > u8 val;
> > > @@ -1065,6 +1349,7 @@ static void rtase_netpoll(struct net_device
> > > *dev) static const struct net_device_ops rtase_netdev_ops = {
> > > .ndo_open = rtase_open,
> > > .ndo_stop = rtase_close,
> > > + .ndo_start_xmit = rtase_start_xmit,
> > > #ifdef CONFIG_NET_POLL_CONTROLLER
> > > .ndo_poll_controller = rtase_netpoll, #endif
> > > --
> > > 2.34.1
> > >
On Fri, 7 Jun 2024 16:43:18 +0800 Justin Lai wrote:
> Implement the ethtool function to support users to obtain network card
> information, including obtaining various device settings, Report whether
> physical link is up, Report pause parameters, Set pause parameters,
> Return a set of strings that describe the requested objects, Get number
> of strings that @get_strings will write, Return extended statistics
> about the device.
You don't implement get_strings any more.
> +static void rtase_get_drvinfo(struct net_device *dev,
> + struct ethtool_drvinfo *drvinfo)
> +{
> + const struct rtase_private *tp = netdev_priv(dev);
> +
> + strscpy(drvinfo->driver, KBUILD_MODNAME, 32);
sizeof(drvinfo->driver) instead of the literal 32?
> + strscpy(drvinfo->bus_info, pci_name(tp->pdev), 32);
Can you double check that overwriting these fields is actually needed?
I think core will fill this in for you in ethtool_get_drvinfo()
> + if ((value & (RTASE_FORCE_TXFLOW_EN | RTASE_FORCE_RXFLOW_EN)) ==
> + (RTASE_FORCE_TXFLOW_EN | RTASE_FORCE_RXFLOW_EN)) {
> + pause->rx_pause = 1;
> + pause->tx_pause = 1;
> + } else if ((value & RTASE_FORCE_TXFLOW_EN)) {
unnecessary parenthesis
> + pause->tx_pause = 1;
> + } else if ((value & RTASE_FORCE_RXFLOW_EN)) {
same here
> + pause->rx_pause = 1;
> + }
> +}
> +
> +static int rtase_set_pauseparam(struct net_device *dev,
> + struct ethtool_pauseparam *pause)
> +{
> + const struct rtase_private *tp = netdev_priv(dev);
> + u16 value = rtase_r16(tp, RTASE_CPLUS_CMD);
> +
> + if (pause->autoneg)
> + return -EOPNOTSUPP;
> +
> + value &= ~(RTASE_FORCE_TXFLOW_EN | RTASE_FORCE_RXFLOW_EN);
> +
> + if (pause->tx_pause)
> + value |= RTASE_FORCE_TXFLOW_EN;
> +
> + if (pause->rx_pause)
> + value |= RTASE_FORCE_RXFLOW_EN;
> +
> + rtase_w16(tp, RTASE_CPLUS_CMD, value);
> + return 0;
> +}
> +
> +static void rtase_get_eth_mac_stats(struct net_device *dev,
> + struct ethtool_eth_mac_stats *stats)
> +{
> + struct rtase_private *tp = netdev_priv(dev);
> + const struct rtase_counters *counters;
> +
> + counters = tp->tally_vaddr;
> + if (!counters)
you fail probe if this is NULL, why check if here?
> + return;
> +
> + rtase_dump_tally_counter(tp);
> +
> + stats->FramesTransmittedOK = le64_to_cpu(counters->tx_packets);
> + stats->SingleCollisionFrames = le32_to_cpu(counters->tx_one_collision);
> + stats->MultipleCollisionFrames =
> + le32_to_cpu(counters->tx_multi_collision);
> + stats->FramesReceivedOK = le64_to_cpu(counters->rx_packets);
> + stats->FrameCheckSequenceErrors = le32_to_cpu(counters->rx_errors);
You dont report this in rtase_get_stats64() as crc errors, are these
really CRC / FCS errors or other errors?
> + stats->AlignmentErrors = le16_to_cpu(counters->align_errors);
> + stats->FramesAbortedDueToXSColls = le16_to_cpu(counters->tx_aborted);
> + stats->FramesLostDueToIntMACXmitError =
> + le64_to_cpu(counters->tx_errors);
> + stats->FramesLostDueToIntMACRcvError =
> + le16_to_cpu(counters->rx_missed);
Are you sure this is the correct statistic to report as?
What's the definition of rx_missed in the datasheet?
Also is 16 bits enough for a packet counter at 5Gbps?
Don't you have to periodically accumulate this counter so that it
doesn't wrap around?
> + stats->MulticastFramesReceivedOK = le32_to_cpu(counters->rx_multicast);
> + stats->BroadcastFramesReceivedOK = le64_to_cpu(counters->rx_broadcast);
> +}
> +
> +static const struct ethtool_ops rtase_ethtool_ops = {
> + .get_drvinfo = rtase_get_drvinfo,
> + .get_link = ethtool_op_get_link,
> + .get_link_ksettings = rtase_get_settings,
> + .get_pauseparam = rtase_get_pauseparam,
> + .set_pauseparam = rtase_set_pauseparam,
> + .get_eth_mac_stats = rtase_get_eth_mac_stats,
> + .get_ts_info = ethtool_op_get_ts_info,
> +};
> +
> static void rtase_init_netdev_ops(struct net_device *dev)
> {
> dev->netdev_ops = &rtase_netdev_ops;
> + dev->ethtool_ops = &rtase_ethtool_ops;
> }
>
> static void rtase_reset_interrupt(struct pci_dev *pdev,
On Fri, 7 Jun 2024 16:43:16 +0800 Justin Lai wrote:
> +static void rtase_get_stats64(struct net_device *dev,
> + struct rtnl_link_stats64 *stats)
> +{
> + const struct rtase_private *tp = netdev_priv(dev);
> + const struct rtase_counters *counters;
> +
> + counters = tp->tally_vaddr;
> +
> + if (!counters)
> + return;
Same question about how this can be null as in the ethtool patch..
> + netdev_stats_to_stats64(stats, &dev->stats);
Please dont use this field, there is a comment on it:
struct net_device_stats stats; /* not used by modern drivers */
You can store the fields you need for counters in struct rtase_private
> + dev_fetch_sw_netstats(stats, dev->tstats);
On Fri, 7 Jun 2024 16:43:15 +0800 Justin Lai wrote:
> +static int rx_handler(struct rtase_ring *ring, int budget)
> +{
> + const struct rtase_private *tp = ring->ivec->tp;
> + union rtase_rx_desc *desc_base = ring->desc;
> + u32 pkt_size, cur_rx, delta, entry, status;
> + struct net_device *dev = tp->dev;
> + union rtase_rx_desc *desc;
> + struct sk_buff *skb;
> + int workdone = 0;
> +
> + cur_rx = ring->cur_idx;
> + entry = cur_rx % RTASE_NUM_DESC;
> + desc = &desc_base[entry];
> +
> + do {
> + /* make sure discriptor has been updated */
> + rmb();
Barriers are between things. What is this barrier between?
> + status = le32_to_cpu(desc->desc_status.opts1);
> +
> + if (status & RTASE_DESC_OWN)
> + break;
> On Fri, 7 Jun 2024 16:43:16 +0800 Justin Lai wrote:
> > +static void rtase_get_stats64(struct net_device *dev,
> > + struct rtnl_link_stats64 *stats) {
> > + const struct rtase_private *tp = netdev_priv(dev);
> > + const struct rtase_counters *counters;
> > +
> > + counters = tp->tally_vaddr;
> > +
> > + if (!counters)
> > + return;
>
> Same question about how this can be null as in the ethtool patch..
This check seems unnecessary, I will remove it.
>
> > + netdev_stats_to_stats64(stats, &dev->stats);
>
> Please dont use this field, there is a comment on it:
>
> struct net_device_stats stats; /* not used by modern drivers */
>
> You can store the fields you need for counters in struct rtase_private
Ok, I will remove all parts that use struct net_device_stats stats.
>
> > + dev_fetch_sw_netstats(stats, dev->tstats);
> > > > Implement .ndo_start_xmit function to fill the information of the
> > > > packet to be transmitted into the tx descriptor, and then the
> > > > hardware will transmit the packet using the information in the tx
> > descriptor.
> > > > In addition, we also implemented the tx_handler function to enable
> > > > the tx descriptor to be reused.
> > > >
> > > > Signed-off-by: Justin Lai <[email protected]>
> > > > ---
> > > > .../net/ethernet/realtek/rtase/rtase_main.c | 285
> ++++++++++++++++++
> > > > 1 file changed, 285 insertions(+)
> > > >
> > > > diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > > b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > > index 23406c195cff..6bdb4edbfbc1 100644
> > > > --- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > > +++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > > @@ -256,6 +256,68 @@ static void rtase_mark_to_asic(union
> > > > rtase_rx_desc *desc, u32 rx_buf_sz)
> > > > cpu_to_le32(RTASE_DESC_OWN | eor | rx_buf_sz));
> > > > }
> > > >
> > > > +static u32 rtase_tx_avail(struct rtase_ring *ring) {
> > > > + return READ_ONCE(ring->dirty_idx) + RTASE_NUM_DESC -
> > > > + READ_ONCE(ring->cur_idx); }
> > > > +
> > > > +static int tx_handler(struct rtase_ring *ring, int budget) {
> > > > + const struct rtase_private *tp = ring->ivec->tp;
> > > > + struct net_device *dev = tp->dev;
> > > > + u32 dirty_tx, tx_left;
> > > > + u32 bytes_compl = 0;
> > > > + u32 pkts_compl = 0;
> > > > + int workdone = 0;
> > > > +
> > > > + dirty_tx = ring->dirty_idx;
> > > > + tx_left = READ_ONCE(ring->cur_idx) - dirty_tx;
> > > > +
> > > > + while (tx_left > 0) {
> > > > + u32 entry = dirty_tx % RTASE_NUM_DESC;
> > > > + struct rtase_tx_desc *desc = ring->desc +
> > > > + sizeof(struct rtase_tx_desc)
> > > > + *
> > > entry;
> > > > + u32 status;
> > > > +
> > > > + status = le32_to_cpu(desc->opts1);
> > > > +
> > > > + if (status & RTASE_DESC_OWN)
> > > > + break;
> > > > +
> > > > + rtase_unmap_tx_skb(tp->pdev, ring->mis.len[entry],
> desc);
> > > > + ring->mis.len[entry] = 0;
> > > > + if (ring->skbuff[entry]) {
> > > > + pkts_compl++;
> > > > + bytes_compl += ring->skbuff[entry]->len;
> > > > + napi_consume_skb(ring->skbuff[entry],
> budget);
> > > > + ring->skbuff[entry] = NULL;
> > > > + }
> > > > +
> > > > + dirty_tx++;
> > > > + tx_left--;
> > > > + workdone++;
> > > > +
> > > > + if (workdone == RTASE_TX_BUDGET_DEFAULT)
> > > > + break;
> > > > + }
> > > > +
> > > > + if (ring->dirty_idx != dirty_tx) {
> > > > + dev_sw_netstats_tx_add(dev, pkts_compl,
> bytes_compl);
> > > > + WRITE_ONCE(ring->dirty_idx, dirty_tx);
> > > > +
> > > > + netif_subqueue_completed_wake(dev, ring->index,
> > > > pkts_compl,
> > > > + bytes_compl,
> > > > +
> rtase_tx_avail(ring),
> > > > +
> > > RTASE_TX_START_THRS);
> > > > +
> > > > + if (ring->cur_idx != dirty_tx)
> > > > + rtase_w8(tp, RTASE_TPPOLL,
> BIT(ring->index));
> > > > + }
> > > > +
> > > > + return 0;
> > > > +}
> > > > +
> > > > static void rtase_tx_desc_init(struct rtase_private *tp, u16 idx) {
> > > > struct rtase_ring *ring = &tp->tx_ring[idx]; @@ -1014,6
> > > > +1076,228 @@ static int rtase_close(struct net_device *dev)
> > > > return 0;
> > > > }
> > > >
> > > > +static u32 rtase_tx_vlan_tag(const struct rtase_private *tp,
> > > > + const struct sk_buff *skb) {
> > > > + return (skb_vlan_tag_present(skb)) ?
> > > > + (RTASE_TX_VLAN_TAG |
> swab16(skb_vlan_tag_get(skb))) :
> > > > 0x00; }
> > > > +
> > > Vlan protocol can be either 0x8100 or 0x88A8, how
> > > does hardware know which header to insert?
> > > Thanks,
> > > Hariprasad k
> >
> > We only allow the hardware to add 0x8100, the VLAN must at least have
> > 0x8100 to potentially have 0x88a8, skb_vlan_tag_present indicates that
> > VLAN exists, hence at least the 0x8100 VLAN would exist.
> > >
> Thanks for the explanation, but one question which bothers me is that "how
> hardware knows offset with in the packet"
>
> For example
> Case 1: DMAC + SMAC + 8100 VLAN_ID + IP
> Here offset is right after the SMAC.
> Case 2: DMAC + SMAC + 88A8 VLAN_ID + 8100 VLAN_ID + IP
> Here offset is right after first vlan tag.
>
> Thanks,
> Hariprasad k
This driver only enables NETIF_F_HW_VLAN_CTAG_TX, and we only support case 1.
>
> > > > +static u32 rtase_tx_csum(struct sk_buff *skb, const struct
> > > > +net_device
> > > > +*dev) {
> > > > + u32 csum_cmd = 0;
> > > > + u8 ip_protocol;
> > > > +
> > > > + switch (vlan_get_protocol(skb)) {
> > > > + case htons(ETH_P_IP):
> > > > + csum_cmd = RTASE_TX_IPCS_C;
> > > > + ip_protocol = ip_hdr(skb)->protocol;
> > > > + break;
> > > > +
> > > > + case htons(ETH_P_IPV6):
> > > > + csum_cmd = RTASE_TX_IPV6F_C;
> > > > + ip_protocol = ipv6_hdr(skb)->nexthdr;
> > > > + break;
> > > > +
> > > > + default:
> > > > + ip_protocol = IPPROTO_RAW;
> > > > + break;
> > > > + }
> > > > +
> > > > + if (ip_protocol == IPPROTO_TCP)
> > > > + csum_cmd |= RTASE_TX_TCPCS_C;
> > > > + else if (ip_protocol == IPPROTO_UDP)
> > > > + csum_cmd |= RTASE_TX_UDPCS_C;
> > > > +
> > > > + csum_cmd |= u32_encode_bits(skb_transport_offset(skb),
> > > > + RTASE_TCPHO_MASK);
> > > > +
> > > > + return csum_cmd;
> > > > +}
> > > > +
> > > > +static int rtase_xmit_frags(struct rtase_ring *ring, struct sk_buff *skb,
> > > > + u32 opts1, u32 opts2) {
> > > > + const struct skb_shared_info *info = skb_shinfo(skb);
> > > > + const struct rtase_private *tp = ring->ivec->tp;
> > > > + const u8 nr_frags = info->nr_frags;
> > > > + struct rtase_tx_desc *txd = NULL;
> > > > + u32 cur_frag, entry;
> > > > +
> > > > + entry = ring->cur_idx;
> > > > + for (cur_frag = 0; cur_frag < nr_frags; cur_frag++) {
> > > > + const skb_frag_t *frag = &info->frags[cur_frag];
> > > > + dma_addr_t mapping;
> > > > + u32 status, len;
> > > > + void *addr;
> > > > +
> > > > + entry = (entry + 1) % RTASE_NUM_DESC;
> > > > +
> > > > + txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
> > > > + len = skb_frag_size(frag);
> > > > + addr = skb_frag_address(frag);
> > > > + mapping = dma_map_single(&tp->pdev->dev, addr, len,
> > > > + DMA_TO_DEVICE);
> > > > +
> > > > + if (unlikely(dma_mapping_error(&tp->pdev->dev,
> > > > + mapping)))
> > > > {
> > > > + if (unlikely(net_ratelimit()))
> > > > + netdev_err(tp->dev,
> > > > + "Failed to map TX
> > > fragments
> > > > DMA!\n");
> > > > +
> > > > + goto err_out;
> > > > + }
> > > > +
> > > > + if (((entry + 1) % RTASE_NUM_DESC) == 0)
> > > > + status = (opts1 | len | RTASE_RING_END);
> > > > + else
> > > > + status = opts1 | len;
> > > > +
> > > > + if (cur_frag == (nr_frags - 1)) {
> > > > + ring->skbuff[entry] = skb;
> > > > + status |= RTASE_TX_LAST_FRAG;
> > > > + }
> > > > +
> > > > + ring->mis.len[entry] = len;
> > > > + txd->addr = cpu_to_le64(mapping);
> > > > + txd->opts2 = cpu_to_le32(opts2);
> > > > +
> > > > + /* make sure the operating fields have been updated */
> > > > + dma_wmb();
> > > > + txd->opts1 = cpu_to_le32(status);
> > > > + }
> > > > +
> > > > + return cur_frag;
> > > > +
> > > > +err_out:
> > > > + rtase_tx_clear_range(ring, ring->cur_idx + 1, cur_frag);
> > > > + return -EIO;
> > > > +}
> > > > +
> > > > +static netdev_tx_t rtase_start_xmit(struct sk_buff *skb,
> > > > + struct net_device *dev) {
> > > > + struct skb_shared_info *shinfo = skb_shinfo(skb);
> > > > + struct rtase_private *tp = netdev_priv(dev);
> > > > + u32 q_idx, entry, len, opts1, opts2;
> > > > + struct netdev_queue *tx_queue;
> > > > + bool stop_queue, door_bell;
> > > > + u32 mss = shinfo->gso_size;
> > > > + struct rtase_tx_desc *txd;
> > > > + struct rtase_ring *ring;
> > > > + dma_addr_t mapping;
> > > > + int frags;
> > > > +
> > > > + /* multiqueues */
> > > > + q_idx = skb_get_queue_mapping(skb);
> > > > + ring = &tp->tx_ring[q_idx];
> > > > + tx_queue = netdev_get_tx_queue(dev, q_idx);
> > > > +
> > > > + if (unlikely(!rtase_tx_avail(ring))) {
> > > > + if (net_ratelimit())
> > > > + netdev_err(dev, "BUG! Tx Ring full when
> > > > + queue
> > > > awake!\n");
> > > > + goto err_stop;
> > > > + }
> > > > +
> > > > + entry = ring->cur_idx % RTASE_NUM_DESC;
> > > > + txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
> > > > +
> > > > + opts1 = RTASE_DESC_OWN;
> > > > + opts2 = rtase_tx_vlan_tag(tp, skb);
> > > > +
> > > > + /* tcp segmentation offload (or tcp large send) */
> > > > + if (mss) {
> > > > + if (shinfo->gso_type & SKB_GSO_TCPV4) {
> > > > + opts1 |= RTASE_GIANT_SEND_V4;
> > > > + } else if (shinfo->gso_type & SKB_GSO_TCPV6) {
> > > > + if (skb_cow_head(skb, 0))
> > > > + goto err_dma_0;
> > > > +
> > > > + tcp_v6_gso_csum_prep(skb);
> > > > + opts1 |= RTASE_GIANT_SEND_V6;
> > > > + } else {
> > > > + WARN_ON_ONCE(1);
> > > > + }
> > > > +
> > > > + opts1 |= u32_encode_bits(skb_transport_offset(skb),
> > > > + RTASE_TCPHO_MASK);
> > > > + opts2 |= u32_encode_bits(mss, RTASE_MSS_MASK);
> > > > + } else if (skb->ip_summed == CHECKSUM_PARTIAL) {
> > > > + opts2 |= rtase_tx_csum(skb, dev);
> > > > + }
> > > > +
> > > > + frags = rtase_xmit_frags(ring, skb, opts1, opts2);
> > > > + if (unlikely(frags < 0))
> > > > + goto err_dma_0;
> > > > +
> > > > + if (frags) {
> > > > + len = skb_headlen(skb);
> > > > + opts1 |= RTASE_TX_FIRST_FRAG;
> > > > + } else {
> > > > + len = skb->len;
> > > > + ring->skbuff[entry] = skb;
> > > > + opts1 |= RTASE_TX_FIRST_FRAG |
> RTASE_TX_LAST_FRAG;
> > > > + }
> > > > +
> > > > + if (((entry + 1) % RTASE_NUM_DESC) == 0)
> > > > + opts1 |= (len | RTASE_RING_END);
> > > > + else
> > > > + opts1 |= len;
> > > > +
> > > > + mapping = dma_map_single(&tp->pdev->dev, skb->data, len,
> > > > + DMA_TO_DEVICE);
> > > > +
> > > > + if (unlikely(dma_mapping_error(&tp->pdev->dev, mapping))) {
> > > > + if (unlikely(net_ratelimit()))
> > > > + netdev_err(dev, "Failed to map TX DMA!\n");
> > > > +
> > > > + goto err_dma_1;
> > > > + }
> > > > +
> > > > + ring->mis.len[entry] = len;
> > > > + txd->addr = cpu_to_le64(mapping);
> > > > + txd->opts2 = cpu_to_le32(opts2);
> > > > + txd->opts1 = cpu_to_le32(opts1 & ~RTASE_DESC_OWN);
> > > > +
> > > > + /* make sure the operating fields have been updated */
> > > > + dma_wmb();
> > > > +
> > > > + door_bell = __netdev_tx_sent_queue(tx_queue, skb->len,
> > > > + netdev_xmit_more());
> > > > +
> > > > + txd->opts1 = cpu_to_le32(opts1);
> > > > +
> > > > + skb_tx_timestamp(skb);
> > > > +
> > > > + /* tx needs to see descriptor changes before updated cur_idx */
> > > > + smp_wmb();
> > > > +
> > > > + WRITE_ONCE(ring->cur_idx, ring->cur_idx + frags + 1);
> > > > +
> > > > + stop_queue = !netif_subqueue_maybe_stop(dev, ring->index,
> > > > +
> > > > + rtase_tx_avail(ring),
> > > > +
> > > RTASE_TX_STOP_THRS,
> > > > +
> > > RTASE_TX_START_THRS);
> > > > +
> > > > + if (door_bell || stop_queue)
> > > > + rtase_w8(tp, RTASE_TPPOLL, BIT(ring->index));
> > > > +
> > > > + return NETDEV_TX_OK;
> > > > +
> > > > +err_dma_1:
> > > > + ring->skbuff[entry] = NULL;
> > > > + rtase_tx_clear_range(ring, ring->cur_idx + 1, frags);
> > > > +
> > > > +err_dma_0:
> > > > + dev->stats.tx_dropped++;
> > > > + dev_kfree_skb_any(skb);
> > > > + return NETDEV_TX_OK;
> > > > +
> > > > +err_stop:
> > > > + netif_stop_queue(dev);
> > > > + dev->stats.tx_dropped++;
> > > > + return NETDEV_TX_BUSY;
> > > > +}
> > > > +
> > > > static void rtase_enable_eem_write(const struct rtase_private *tp) {
> > > > u8 val;
> > > > @@ -1065,6 +1349,7 @@ static void rtase_netpoll(struct net_device
> > > > *dev) static const struct net_device_ops rtase_netdev_ops = {
> > > > .ndo_open = rtase_open,
> > > > .ndo_stop = rtase_close,
> > > > + .ndo_start_xmit = rtase_start_xmit,
> > > > #ifdef CONFIG_NET_POLL_CONTROLLER
> > > > .ndo_poll_controller = rtase_netpoll, #endif
> > > > --
> > > > 2.34.1
> > > >
> > > > > Implement .ndo_start_xmit function to fill the information of
> > > > > the packet to be transmitted into the tx descriptor, and then
> > > > > the hardware will transmit the packet using the information in
> > > > > the tx
> > > descriptor.
> > > > > In addition, we also implemented the tx_handler function to
> > > > > enable the tx descriptor to be reused.
> > > > >
> > > > > Signed-off-by: Justin Lai <[email protected]>
> > > > > ---
> > > > > .../net/ethernet/realtek/rtase/rtase_main.c | 285
> > ++++++++++++++++++
> > > > > 1 file changed, 285 insertions(+)
> > > > >
> > > > > diff --git a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > > > b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > > > index 23406c195cff..6bdb4edbfbc1 100644
> > > > > --- a/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > > > +++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> > > > > @@ -256,6 +256,68 @@ static void rtase_mark_to_asic(union
> > > > > rtase_rx_desc *desc, u32 rx_buf_sz)
> > > > > cpu_to_le32(RTASE_DESC_OWN | eor | rx_buf_sz));
> > > > > }
> > > > >
> > > > > +static u32 rtase_tx_avail(struct rtase_ring *ring) {
> > > > > + return READ_ONCE(ring->dirty_idx) + RTASE_NUM_DESC -
> > > > > + READ_ONCE(ring->cur_idx); }
> > > > > +
> > > > > +static int tx_handler(struct rtase_ring *ring, int budget) {
> > > > > + const struct rtase_private *tp = ring->ivec->tp;
> > > > > + struct net_device *dev = tp->dev;
> > > > > + u32 dirty_tx, tx_left;
> > > > > + u32 bytes_compl = 0;
> > > > > + u32 pkts_compl = 0;
> > > > > + int workdone = 0;
> > > > > +
> > > > > + dirty_tx = ring->dirty_idx;
> > > > > + tx_left = READ_ONCE(ring->cur_idx) - dirty_tx;
> > > > > +
> > > > > + while (tx_left > 0) {
> > > > > + u32 entry = dirty_tx % RTASE_NUM_DESC;
> > > > > + struct rtase_tx_desc *desc = ring->desc +
> > > > > + sizeof(struct
> > > > > + rtase_tx_desc)
> > > > > + *
> > > > entry;
> > > > > + u32 status;
> > > > > +
> > > > > + status = le32_to_cpu(desc->opts1);
> > > > > +
> > > > > + if (status & RTASE_DESC_OWN)
> > > > > + break;
> > > > > +
> > > > > + rtase_unmap_tx_skb(tp->pdev, ring->mis.len[entry],
> > desc);
> > > > > + ring->mis.len[entry] = 0;
> > > > > + if (ring->skbuff[entry]) {
> > > > > + pkts_compl++;
> > > > > + bytes_compl += ring->skbuff[entry]->len;
> > > > > + napi_consume_skb(ring->skbuff[entry],
> > budget);
> > > > > + ring->skbuff[entry] = NULL;
> > > > > + }
> > > > > +
> > > > > + dirty_tx++;
> > > > > + tx_left--;
> > > > > + workdone++;
> > > > > +
> > > > > + if (workdone == RTASE_TX_BUDGET_DEFAULT)
> > > > > + break;
> > > > > + }
> > > > > +
> > > > > + if (ring->dirty_idx != dirty_tx) {
> > > > > + dev_sw_netstats_tx_add(dev, pkts_compl,
> > bytes_compl);
> > > > > + WRITE_ONCE(ring->dirty_idx, dirty_tx);
> > > > > +
> > > > > + netif_subqueue_completed_wake(dev, ring->index,
> > > > > pkts_compl,
> > > > > + bytes_compl,
> > > > > +
> > rtase_tx_avail(ring),
> > > > > +
> > > > RTASE_TX_START_THRS);
> > > > > +
> > > > > + if (ring->cur_idx != dirty_tx)
> > > > > + rtase_w8(tp, RTASE_TPPOLL,
> > BIT(ring->index));
> > > > > + }
> > > > > +
> > > > > + return 0;
> > > > > +}
> > > > > +
> > > > > static void rtase_tx_desc_init(struct rtase_private *tp, u16 idx) {
> > > > > struct rtase_ring *ring = &tp->tx_ring[idx]; @@ -1014,6
> > > > > +1076,228 @@ static int rtase_close(struct net_device *dev)
> > > > > return 0;
> > > > > }
> > > > >
> > > > > +static u32 rtase_tx_vlan_tag(const struct rtase_private *tp,
> > > > > + const struct sk_buff *skb) {
> > > > > + return (skb_vlan_tag_present(skb)) ?
> > > > > + (RTASE_TX_VLAN_TAG |
> > swab16(skb_vlan_tag_get(skb))) :
> > > > > 0x00; }
> > > > > +
> > > > Vlan protocol can be either 0x8100 or 0x88A8, how
> > > > does hardware know which header to insert?
> > > > Thanks,
> > > > Hariprasad k
> > >
> > > We only allow the hardware to add 0x8100, the VLAN must at least
> > > have
> > > 0x8100 to potentially have 0x88a8, skb_vlan_tag_present indicates
> > > that VLAN exists, hence at least the 0x8100 VLAN would exist.
> > > >
> > Thanks for the explanation, but one question which bothers me is that
> > "how hardware knows offset with in the packet"
> >
> > For example
> > Case 1: DMAC + SMAC + 8100 VLAN_ID + IP
> > Here offset is right after the SMAC.
> > Case 2: DMAC + SMAC + 88A8 VLAN_ID + 8100 VLAN_ID + IP
> > Here offset is right after first vlan tag.
> >
> > Thanks,
> > Hariprasad k
>
> This driver only enables NETIF_F_HW_VLAN_CTAG_TX, and we only support
> case 1.
>
Got it . So the hardware is programmed with fixed offset.
> >
> > > > > +static u32 rtase_tx_csum(struct sk_buff *skb, const struct
> > > > > +net_device
> > > > > +*dev) {
> > > > > + u32 csum_cmd = 0;
> > > > > + u8 ip_protocol;
> > > > > +
> > > > > + switch (vlan_get_protocol(skb)) {
> > > > > + case htons(ETH_P_IP):
> > > > > + csum_cmd = RTASE_TX_IPCS_C;
> > > > > + ip_protocol = ip_hdr(skb)->protocol;
> > > > > + break;
> > > > > +
> > > > > + case htons(ETH_P_IPV6):
> > > > > + csum_cmd = RTASE_TX_IPV6F_C;
> > > > > + ip_protocol = ipv6_hdr(skb)->nexthdr;
> > > > > + break;
> > > > > +
> > > > > + default:
> > > > > + ip_protocol = IPPROTO_RAW;
> > > > > + break;
> > > > > + }
> > > > > +
> > > > > + if (ip_protocol == IPPROTO_TCP)
> > > > > + csum_cmd |= RTASE_TX_TCPCS_C;
> > > > > + else if (ip_protocol == IPPROTO_UDP)
> > > > > + csum_cmd |= RTASE_TX_UDPCS_C;
> > > > > +
> > > > > + csum_cmd |= u32_encode_bits(skb_transport_offset(skb),
> > > > > + RTASE_TCPHO_MASK);
> > > > > +
> > > > > + return csum_cmd;
> > > > > +}
> > > > > +
> > > > > +static int rtase_xmit_frags(struct rtase_ring *ring, struct sk_buff *skb,
> > > > > + u32 opts1, u32 opts2) {
> > > > > + const struct skb_shared_info *info = skb_shinfo(skb);
> > > > > + const struct rtase_private *tp = ring->ivec->tp;
> > > > > + const u8 nr_frags = info->nr_frags;
> > > > > + struct rtase_tx_desc *txd = NULL;
> > > > > + u32 cur_frag, entry;
> > > > > +
> > > > > + entry = ring->cur_idx;
> > > > > + for (cur_frag = 0; cur_frag < nr_frags; cur_frag++) {
> > > > > + const skb_frag_t *frag = &info->frags[cur_frag];
> > > > > + dma_addr_t mapping;
> > > > > + u32 status, len;
> > > > > + void *addr;
> > > > > +
> > > > > + entry = (entry + 1) % RTASE_NUM_DESC;
> > > > > +
> > > > > + txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
> > > > > + len = skb_frag_size(frag);
> > > > > + addr = skb_frag_address(frag);
> > > > > + mapping = dma_map_single(&tp->pdev->dev, addr, len,
> > > > > + DMA_TO_DEVICE);
> > > > > +
> > > > > + if (unlikely(dma_mapping_error(&tp->pdev->dev,
> > > > > + mapping)))
> > > > > {
> > > > > + if (unlikely(net_ratelimit()))
> > > > > + netdev_err(tp->dev,
> > > > > + "Failed to map TX
> > > > fragments
> > > > > DMA!\n");
> > > > > +
> > > > > + goto err_out;
> > > > > + }
> > > > > +
> > > > > + if (((entry + 1) % RTASE_NUM_DESC) == 0)
> > > > > + status = (opts1 | len | RTASE_RING_END);
> > > > > + else
> > > > > + status = opts1 | len;
> > > > > +
> > > > > + if (cur_frag == (nr_frags - 1)) {
> > > > > + ring->skbuff[entry] = skb;
> > > > > + status |= RTASE_TX_LAST_FRAG;
> > > > > + }
> > > > > +
> > > > > + ring->mis.len[entry] = len;
> > > > > + txd->addr = cpu_to_le64(mapping);
> > > > > + txd->opts2 = cpu_to_le32(opts2);
> > > > > +
> > > > > + /* make sure the operating fields have been updated */
> > > > > + dma_wmb();
> > > > > + txd->opts1 = cpu_to_le32(status);
> > > > > + }
> > > > > +
> > > > > + return cur_frag;
> > > > > +
> > > > > +err_out:
> > > > > + rtase_tx_clear_range(ring, ring->cur_idx + 1, cur_frag);
> > > > > + return -EIO;
> > > > > +}
> > > > > +
> > > > > +static netdev_tx_t rtase_start_xmit(struct sk_buff *skb,
> > > > > + struct net_device *dev) {
> > > > > + struct skb_shared_info *shinfo = skb_shinfo(skb);
> > > > > + struct rtase_private *tp = netdev_priv(dev);
> > > > > + u32 q_idx, entry, len, opts1, opts2;
> > > > > + struct netdev_queue *tx_queue;
> > > > > + bool stop_queue, door_bell;
> > > > > + u32 mss = shinfo->gso_size;
> > > > > + struct rtase_tx_desc *txd;
> > > > > + struct rtase_ring *ring;
> > > > > + dma_addr_t mapping;
> > > > > + int frags;
> > > > > +
> > > > > + /* multiqueues */
> > > > > + q_idx = skb_get_queue_mapping(skb);
> > > > > + ring = &tp->tx_ring[q_idx];
> > > > > + tx_queue = netdev_get_tx_queue(dev, q_idx);
> > > > > +
> > > > > + if (unlikely(!rtase_tx_avail(ring))) {
> > > > > + if (net_ratelimit())
> > > > > + netdev_err(dev, "BUG! Tx Ring full when
> > > > > + queue
> > > > > awake!\n");
> > > > > + goto err_stop;
> > > > > + }
> > > > > +
> > > > > + entry = ring->cur_idx % RTASE_NUM_DESC;
> > > > > + txd = ring->desc + sizeof(struct rtase_tx_desc) * entry;
> > > > > +
> > > > > + opts1 = RTASE_DESC_OWN;
> > > > > + opts2 = rtase_tx_vlan_tag(tp, skb);
> > > > > +
> > > > > + /* tcp segmentation offload (or tcp large send) */
> > > > > + if (mss) {
> > > > > + if (shinfo->gso_type & SKB_GSO_TCPV4) {
> > > > > + opts1 |= RTASE_GIANT_SEND_V4;
> > > > > + } else if (shinfo->gso_type & SKB_GSO_TCPV6) {
> > > > > + if (skb_cow_head(skb, 0))
> > > > > + goto err_dma_0;
> > > > > +
> > > > > + tcp_v6_gso_csum_prep(skb);
> > > > > + opts1 |= RTASE_GIANT_SEND_V6;
> > > > > + } else {
> > > > > + WARN_ON_ONCE(1);
> > > > > + }
> > > > > +
> > > > > + opts1 |= u32_encode_bits(skb_transport_offset(skb),
> > > > > + RTASE_TCPHO_MASK);
> > > > > + opts2 |= u32_encode_bits(mss, RTASE_MSS_MASK);
> > > > > + } else if (skb->ip_summed == CHECKSUM_PARTIAL) {
> > > > > + opts2 |= rtase_tx_csum(skb, dev);
> > > > > + }
> > > > > +
> > > > > + frags = rtase_xmit_frags(ring, skb, opts1, opts2);
> > > > > + if (unlikely(frags < 0))
> > > > > + goto err_dma_0;
> > > > > +
> > > > > + if (frags) {
> > > > > + len = skb_headlen(skb);
> > > > > + opts1 |= RTASE_TX_FIRST_FRAG;
> > > > > + } else {
> > > > > + len = skb->len;
> > > > > + ring->skbuff[entry] = skb;
> > > > > + opts1 |= RTASE_TX_FIRST_FRAG |
> > RTASE_TX_LAST_FRAG;
> > > > > + }
> > > > > +
> > > > > + if (((entry + 1) % RTASE_NUM_DESC) == 0)
> > > > > + opts1 |= (len | RTASE_RING_END);
> > > > > + else
> > > > > + opts1 |= len;
> > > > > +
> > > > > + mapping = dma_map_single(&tp->pdev->dev, skb->data, len,
> > > > > + DMA_TO_DEVICE);
> > > > > +
> > > > > + if (unlikely(dma_mapping_error(&tp->pdev->dev, mapping))) {
> > > > > + if (unlikely(net_ratelimit()))
> > > > > + netdev_err(dev, "Failed to map TX
> > > > > + DMA!\n");
> > > > > +
> > > > > + goto err_dma_1;
> > > > > + }
> > > > > +
> > > > > + ring->mis.len[entry] = len;
> > > > > + txd->addr = cpu_to_le64(mapping);
> > > > > + txd->opts2 = cpu_to_le32(opts2);
> > > > > + txd->opts1 = cpu_to_le32(opts1 & ~RTASE_DESC_OWN);
> > > > > +
> > > > > + /* make sure the operating fields have been updated */
> > > > > + dma_wmb();
> > > > > +
> > > > > + door_bell = __netdev_tx_sent_queue(tx_queue, skb->len,
> > > > > + netdev_xmit_more());
> > > > > +
> > > > > + txd->opts1 = cpu_to_le32(opts1);
> > > > > +
> > > > > + skb_tx_timestamp(skb);
> > > > > +
> > > > > + /* tx needs to see descriptor changes before updated cur_idx */
> > > > > + smp_wmb();
> > > > > +
> > > > > + WRITE_ONCE(ring->cur_idx, ring->cur_idx + frags + 1);
> > > > > +
> > > > > + stop_queue = !netif_subqueue_maybe_stop(dev, ring->index,
> > > > > +
> > > > > + rtase_tx_avail(ring),
> > > > > +
> > > > RTASE_TX_STOP_THRS,
> > > > > +
> > > > RTASE_TX_START_THRS);
> > > > > +
> > > > > + if (door_bell || stop_queue)
> > > > > + rtase_w8(tp, RTASE_TPPOLL, BIT(ring->index));
> > > > > +
> > > > > + return NETDEV_TX_OK;
> > > > > +
> > > > > +err_dma_1:
> > > > > + ring->skbuff[entry] = NULL;
> > > > > + rtase_tx_clear_range(ring, ring->cur_idx + 1, frags);
> > > > > +
> > > > > +err_dma_0:
> > > > > + dev->stats.tx_dropped++;
> > > > > + dev_kfree_skb_any(skb);
> > > > > + return NETDEV_TX_OK;
> > > > > +
> > > > > +err_stop:
> > > > > + netif_stop_queue(dev);
> > > > > + dev->stats.tx_dropped++;
> > > > > + return NETDEV_TX_BUSY;
> > > > > +}
> > > > > +
> > > > > static void rtase_enable_eem_write(const struct rtase_private *tp) {
> > > > > u8 val;
> > > > > @@ -1065,6 +1349,7 @@ static void rtase_netpoll(struct
> > > > > net_device
> > > > > *dev) static const struct net_device_ops rtase_netdev_ops = {
> > > > > .ndo_open = rtase_open,
> > > > > .ndo_stop = rtase_close,
> > > > > + .ndo_start_xmit = rtase_start_xmit,
> > > > > #ifdef CONFIG_NET_POLL_CONTROLLER
> > > > > .ndo_poll_controller = rtase_netpoll, #endif
> > > > > --
> > > > > 2.34.1
> > > > >
…
> when requesting irq, because the first group of interrupts needs to
> process more events, the overall structure will be different from
> other groups of interrupts, so it needs to be processed separately.
Can such a change description become clearer anyhow?
…
> +++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
…
> +static int rtase_alloc_desc(struct rtase_private *tp)
> +{
…
> + netdev_err(tp->dev, "Failed to allocate dma memory of "
> + "tx descriptor.\n");
…
Would you like to keep the message (from such string literals) in a single line?
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/coding-style.rst?h=v6.10-rc3#n116
…
> +static int rtase_alloc_rx_skb(const struct rtase_ring *ring,
…
> +{
…
> + struct sk_buff *skb = NULL;
…
> + int ret = 0;
…
> + if (!page) {
> + netdev_err(tp->dev, "failed to alloc page\n");
> + goto err_out;
…
> + if (!skb) {
…
> + netdev_err(tp->dev, "failed to build skb\n");
> + goto err_out;
> + }
…
> + return ret;
I find the following statement more appropriate.
return 0;
> +
> +err_out:
> + if (skb)
> + dev_kfree_skb(skb);
Why would you like to repeat such a check after it can be determined
from the control flow that the used variable contains still a null pointer?
> +
> + ret = -ENOMEM;
> + rtase_make_unusable_by_asic(desc);
> +
> + return ret;
> +}
…
It seems that the following statement can be more appropriate.
return -ENOMEM;
May the local variable “ret” be omitted here?
…
> +static int rtase_open(struct net_device *dev)
> +{
…
> + int ret;
> +
> + ivec = &tp->int_vector[0];
> + tp->rx_buf_sz = RTASE_RX_BUF_SIZE;
> +
> + ret = rtase_alloc_desc(tp);
> + if (ret)
> + goto err_free_all_allocated_mem;
…
I suggest to return directly after such a resource allocation failure.
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/coding-style.rst?h=v6.10-rc3#n532
How do you think about to increase the application of scope-based resource management?
https://elixir.bootlin.com/linux/v6.10-rc3/source/include/linux/cleanup.h#L8
Regards,
Markus
> Add pci table supported in this module, and implement pci_driver function
> to initialize this driver, remove this driver, or shutdown this driver.
Can a summary phrase like “Add support for a PCI table” be a bit nicer?
…
> +++ b/drivers/net/ethernet/realtek/rtase/rtase_main.c
> @@ -0,0 +1,640 @@
…
> +static int rtase_init_one(struct pci_dev *pdev,
> + const struct pci_device_id *ent)
> +{
…
> + /* identify chip attached to board */
> + if (!rtase_check_mac_version_valid(tp)) {
> + return dev_err_probe(&pdev->dev, -ENODEV,
> + "unknown chip version, contact rtase "
> + "maintainers (see MAINTAINERS file)\n");
> + }
…
* May curly brackets be omitted here?
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/coding-style.rst?h=v6.10-rc3#n197
* Would you like to keep the message (from such string literals) in a single line?
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/coding-style.rst?h=v6.10-rc3#n116
…
> + dev->features |= NETIF_F_IP_CSUM;
> + dev->features |= NETIF_F_HIGHDMA;
…
> + dev->hw_features |= NETIF_F_RXALL;
> + dev->hw_features |= NETIF_F_RXFCS;
…
How do you think about to reduce such assignment statements
(if all desired software options would be passed at once)?
Regards,
Markus