net: dsa: microchip: Add KSZ8463 switch support
Subsystems (5)
v4 cover net: dsa: microchip: Add KSZ8463 switch support
From: Tristram Ha <[email protected]>
This series of patches is to add KSZ8463 switch support to the KSZ DSA driver.
v4
- Fix a typo in ksz8_reg.h
- Fix logic in ksz8463_r_phy()
v3
- Replace cpu_to_be16() with swab16() to avoid compiler warning
- Disable PTP function in a separate patch
v2
- Break the KSZ8463 driver code into several patches for easy review
- Replace ntohs with cpu_to_be16
Tristram Ha (7):
dt-bindings: net: dsa: microchip: Add KSZ8463 switch support
net: dsa: microchip: Add KSZ8463 switch support to KSZ DSA driver
net: dsa: microchip: Transform register for use with KSZ8463
net: dsa: microchip: Use different registers for KSZ8463
net: dsa: microchip: Write switch MAC address differently for KSZ8463
net: dsa: microchip: Setup fiber ports for KSZ8463
net: dsa: microchip: Disable PTP function of KSZ8463
.../bindings/net/dsa/microchip,ksz.yaml | 1 + drivers/net/dsa/microchip/ksz8.c | 195 +++++++++++++++--- drivers/net/dsa/microchip/ksz8.h | 4 + drivers/net/dsa/microchip/ksz8_reg.h | 49 +++++ drivers/net/dsa/microchip/ksz_common.c | 168 ++++++++++++++- drivers/net/dsa/microchip/ksz_common.h | 104 ++++++++-- drivers/net/dsa/microchip/ksz_dcb.c | 10 +- drivers/net/dsa/microchip/ksz_spi.c | 14 ++ include/linux/platform_data/microchip-ksz.h | 1 + 9 files changed, 497 insertions(+), 49 deletions(-)
Signature · 1 line
2.34.1
PATCH v4 1/7 dt-bindings: net: dsa: microchip: Add KSZ8463 switch support
From: Tristram Ha <[email protected]>
KSZ8463 switch is a 3-port switch based from KSZ8863. Its register access is significantly different from the other KSZ SPI switches.
--- Documentation/devicetree/bindings/net/dsa/microchip,ksz.yaml | 1 + 1 file changed, 1 insertion(+)
Documentation/devicetree/bindings/net/dsa/microchip,ksz.yaml +1 -0
@@ -18,6 +18,7 @@ properties: # required and optional properties. compatible: enum: + - microchip,ksz8463 - microchip,ksz8765 - microchip,ksz8794 - microchip,ksz8795
Signature · 1 line
2.34.1
PATCH v4 2/7 net: dsa: microchip: Add KSZ8463 switch support to KSZ DSA driver
In reply to [email protected]From: Tristram Ha <[email protected]>
KSZ8463 switch is a 3-port switch based from KSZ8863. Its major difference from other KSZ SPI switches is its register access is not a simple continual 8-bit transfer with automatic address increase but uses a byte-enable mechanism specifying 8-bit, 16-bit, or 32-bit access. Its registers are also defined in 16-bit format because it shares a design with a MAC controller using 16-bit access. As a result some common register accesses need to be re-arranged. The 64-bit access used by other switches needs to be broken into 2 32-bit accesses.
This patch adds the basic structure for using KSZ8463. It cannot use the same regmap table for other KSZ switches as it interprets the 16-bit value as little-endian and its SPI commands are different.
KSZ8463's internal PHYs use standard PHY register definitions so there is no need to remap things. However, the hardware has a bug that the high word and low word of the PHY id are swapped. In addition the port registers are arranged differently so KSZ8463 has its own mapping for port registers and PHY registers.
---
v4
- Fix a typo in ksz8_reg.h
- Fix logic in ksz8463_r_phy()
drivers/net/dsa/microchip/ksz8.c | 80 ++++++++++++++ drivers/net/dsa/microchip/ksz8.h | 4 + drivers/net/dsa/microchip/ksz8_reg.h | 49 +++++++++ drivers/net/dsa/microchip/ksz_common.c | 114 ++++++++++++++++++++ drivers/net/dsa/microchip/ksz_common.h | 33 ++++++ drivers/net/dsa/microchip/ksz_spi.c | 14 +++ include/linux/platform_data/microchip-ksz.h | 1 + 7 files changed, 295 insertions(+)
drivers/net/dsa/microchip/ksz8.c +80 -0
@@ -3,6 +3,7 @@* Microchip KSZ8XXX series switch driver * * It supports the following switches: +* - KSZ8463 * - KSZ8863, KSZ8873 aka KSZ88X3 * - KSZ8895, KSZ8864 aka KSZ8895 family * - KSZ8794, KSZ8795, KSZ8765 aka KSZ87XX
@@ -194,6 +195,7 @@ int ksz8_change_mtu(struct ksz_device *dev, int port, int mtu)case KSZ8794_CHIP_ID :case KSZ8765_CHIP_ID :return ksz8795_change_mtu (dev, frame_size); +case KSZ8463_CHIP_ID :case KSZ88X3_CHIP_ID :case KSZ8864_CHIP_ID :case KSZ8895_CHIP_ID :
@@ -1947,6 +1949,84 @@ u32 ksz8_get_port_addr(int port, int offset)return PORT_CTRL_ADDR (port, offset); } +u32 ksz8463_get_port_addr (int port,int offset) +{ +return offset +0x18 * port; +} + +static inline u16 ksz8463_get_phy_addr (u16 phy,u16 reg,u16 offset) +{ +return offset + reg *2 + phy * (P2MBCR -P1MBCR ); +} + +int ksz8463_r_phy (struct ksz_device *dev,u16 phy,u16 reg,u16 *val) +{ +u16 sw_reg =0 ; +u16 data =0 ; +int ret; + +if (phy >1 ) +return -ENOSPC ; +switch (reg) { +case MII_PHYSID1 : + sw_reg =ksz8463_get_phy_addr (phy,0 ,PHY1IHR ); +break ; +case MII_PHYSID2 : + sw_reg =ksz8463_get_phy_addr (phy,0 ,PHY1ILR ); +break ; +case MII_BMCR : +case MII_BMSR : +case MII_ADVERTISE : +case MII_LPA : + sw_reg =ksz8463_get_phy_addr (phy, reg,P1MBCR ); +break ; +case MII_TPISTATUS : +/* This register holds the PHY interrupt status for simulated +* Micrel KSZ PHY. +*/ + data =0x0505 ; +break ; +default : +break ; + } +if (sw_reg) { + ret =ksz_read16 (dev, sw_reg, &data); +if (ret) +return ret; + } + *val = data; + +return 0 ; +} + +int ksz8463_w_phy (struct ksz_device *dev,u16 phy,u16 reg,u16 val) +{ +u16 sw_reg =0 ; +int ret; + +if (phy >1 ) +return -ENOSPC ; + +/* No write to fiber port. */ +if (dev->ports[phy].fiber) +return 0 ; +switch (reg) { +case MII_BMCR : +case MII_ADVERTISE : + sw_reg =ksz8463_get_phy_addr (phy, reg,P1MBCR ); +break ; +default : +break ; + } +if (sw_reg) { + ret =ksz_write16 (dev, sw_reg, val); +if (ret) +return ret; + } + +return 0 ; +} +int ksz8_switch_init (struct ksz_device *dev) { dev->cpu_port =fls (dev->info->cpu_ports) -1 ;
drivers/net/dsa/microchip/ksz8.h +4 -0
@@ -63,4 +63,8 @@ void ksz8_phylink_mac_link_up(struct phylink_config *config, bool tx_pause, bool rx_pause);int ksz8_all_queues_split (struct ksz_device *dev,int queues); +u32 ksz8463_get_port_addr (int port,int offset); +int ksz8463_r_phy (struct ksz_device *dev,u16 phy,u16 reg,u16 *val); +int ksz8463_w_phy (struct ksz_device *dev,u16 phy,u16 reg,u16 val); +#endif
drivers/net/dsa/microchip/ksz8_reg.h +49 -0
55 lines not on this page · raw message
drivers/net/dsa/microchip/ksz_common.c +114 -0
150 lines not on this page · raw message
drivers/net/dsa/microchip/ksz_common.h +33 -0
@@ -222,6 +222,7 @@ struct ksz_device {/* List of supported models */ enum ksz_model { +KSZ8463 ,KSZ8563 ,KSZ8567 ,KSZ8795 ,
@@ -761,6 +762,7 @@ static inline bool ksz_is_sgmii_port(struct ksz_device *dev, int port)#define REG_CHIP_ID0 0x00#define SW_FAMILY_ID_M GENMASK(15, 8) +#define KSZ84_FAMILY_ID 0x84#define KSZ87_FAMILY_ID 0x87#define KSZ88_FAMILY_ID 0x88#define KSZ8895_FAMILY_ID 0x95
@@ -906,6 +908,9 @@ static inline bool ksz_is_sgmii_port(struct ksz_device *dev, int port)#define KSZ_SPI_OP_RD 3#define KSZ_SPI_OP_WR 2 +#define KSZ8463_SPI_OP_RD 0 +#define KSZ8463_SPI_OP_WR 1 +#define swabnot_used (x) 0#define KSZ_SPI_OP_FLAG_MASK (opcode, swp, regbits, regpad) \
@@ -939,4 +944,32 @@ static inline bool ksz_is_sgmii_port(struct ksz_device *dev, int port) [KSZ_REGMAP_32] = KSZ_REGMAP_ENTRY(32, swp, (regbits), (regpad), (regalign)), \ } +#define KSZ8463_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ + { \ + .name = #width, \ + .val_bits = (width), \ + .reg_stride = 1, \ + .reg_bits = (regbits) + (regalign), \ + .pad_bits = (regpad), \ + .max_register = BIT(regbits) - 1, \ + .cache_type = REGCACHE_NONE, \ + .read_flag_mask = \ + KSZ_SPI_OP_FLAG_MASK(KSZ8463_SPI_OP_RD, swp, \ + regbits, regpad), \ + .write_flag_mask = \ + KSZ_SPI_OP_FLAG_MASK(KSZ8463_SPI_OP_WR, swp, \ + regbits, regpad), \ + .lock = ksz_regmap_lock, \ + .unlock = ksz_regmap_unlock, \ + .reg_format_endian = REGMAP_ENDIAN_BIG, \ + .val_format_endian = REGMAP_ENDIAN_LITTLE \ + } + +#define KSZ8463_REGMAP_TABLE (ksz, swp, regbits, regpad, regalign) \ + static const struct regmap_config ksz##_regmap_config[] = { \ + [KSZ_REGMAP_8] = KSZ8463_REGMAP_ENTRY(8, swp, (regbits), (regpad), (regalign)), \ + [KSZ_REGMAP_16] = KSZ8463_REGMAP_ENTRY(16, swp, (regbits), (regpad), (regalign)), \ + [KSZ_REGMAP_32] = KSZ8463_REGMAP_ENTRY(32, swp, (regbits), (regpad), (regalign)), \ + } +#endif
drivers/net/dsa/microchip/ksz_spi.c +14 -0
@@ -16,6 +16,10 @@#include "ksz_common.h" +#define KSZ8463_SPI_ADDR_SHIFT 13 +#define KSZ8463_SPI_ADDR_ALIGN 1 +#define KSZ8463_SPI_TURNAROUND_SHIFT 2 +#define KSZ8795_SPI_ADDR_SHIFT 12#define KSZ8795_SPI_ADDR_ALIGN 3#define KSZ8795_SPI_TURNAROUND_SHIFT 1
@@ -28,6 +32,9 @@#define KSZ9477_SPI_ADDR_ALIGN 3#define KSZ9477_SPI_TURNAROUND_SHIFT 5 +KSZ8463_REGMAP_TABLE (ksz8463,16 ,KSZ8463_SPI_ADDR_SHIFT , +KSZ8463_SPI_TURNAROUND_SHIFT ,KSZ8463_SPI_ADDR_ALIGN ); +KSZ_REGMAP_TABLE (ksz8795,16 ,KSZ8795_SPI_ADDR_SHIFT ,KSZ8795_SPI_TURNAROUND_SHIFT ,KSZ8795_SPI_ADDR_ALIGN );
@@ -58,6 +65,8 @@ static int ksz_spi_probe(struct spi_device *spi) dev->chip_id = chip->chip_id;if (chip->chip_id ==KSZ88X3_CHIP_ID ) regmap_config = ksz8863_regmap_config; +else if (chip->chip_id ==KSZ8463_CHIP_ID ) + regmap_config = ksz8463_regmap_config;else if (chip->chip_id ==KSZ8795_CHIP_ID || chip->chip_id ==KSZ8794_CHIP_ID || chip->chip_id ==KSZ8765_CHIP_ID )
@@ -125,6 +134,10 @@ static void ksz_spi_shutdown(struct spi_device *spi) }static const struct of_device_id ksz_dt_ids[] = { + { + .compatible ="microchip,ksz8463" , + .data = &ksz_switch_chips[KSZ8463 ] + }, { .compatible ="microchip,ksz8765" , .data = &ksz_switch_chips[KSZ8765 ]
@@ -214,6 +227,7 @@ static const struct of_device_id ksz_dt_ids[] = {MODULE_DEVICE_TABLE (of, ksz_dt_ids);static const struct spi_device_id ksz_spi_ids[] = { + {"ksz8463" }, {"ksz8765" }, {"ksz8794" }, {"ksz8795" },
include/linux/platform_data/microchip-ksz.h +1 -0
@@ -23,6 +23,7 @@#include <linux/platform_data/dsa.h> enum ksz_chip_id { +KSZ8463_CHIP_ID =0x8463 ,KSZ8563_CHIP_ID =0x8563 ,KSZ8795_CHIP_ID =0x8795 ,KSZ8794_CHIP_ID =0x8794 ,
Signature · 1 line
2.34.1
Simon Horman
reviewerRe: [PATCH net-next v4 2/7] net: dsa: microchip: Add KSZ8463 switch support to KSZ DSA driver
Quoted 22 lines · On Fri, Jul 18, 2025 at 06:21:01PM -0700, [email protected] wrote:
From: Tristram Ha <[email protected]>
KSZ8463 switch is a 3-port switch based from KSZ8863. Its major difference from other KSZ SPI switches is its register access is not a simple continual 8-bit transfer with automatic address increase but uses a byte-enable mechanism specifying 8-bit, 16-bit, or 32-bit access. Its registers are also defined in 16-bit format because it shares a design with a MAC controller using 16-bit access. As a result some common register accesses need to be re-arranged. The 64-bit access used by other switches needs to be broken into 2 32-bit accesses.
This patch adds the basic structure for using KSZ8463. It cannot use the same regmap table for other KSZ switches as it interprets the 16-bit value as little-endian and its SPI commands are different.
KSZ8463's internal PHYs use standard PHY register definitions so there is no need to remap things. However, the hardware has a bug that the high word and low word of the PHY id are swapped. In addition the port registers are arranged differently so KSZ8463 has its own mapping for port registers and PHY registers.
Signed-off-by Tristram Ha <[email protected]>
...
Quoted 1 line
drivers/net/dsa/microchip/ksz8.c +0
-0
...
drivers/net/dsa/microchip/ksz8.c u32 ksz8_get_port_addr(int port, int offset) · line 1960
Quoted 4 lines
drivers/net/dsa/microchip/ksz8.c +4
-0+static inline u16 ksz8463_get_phy_addr (u16 phy,u16 reg,u16 offset) +{ +return offset + reg *2 + phy * (P2MBCR -P1MBCR ); +}
nit: Please do not use the inline keyword in .c files,
unless there is a demonstrable (usually performance) reason to do so. Rather, let the compiler inline (or not) code as it sees fit.
...
PATCH v4 3/7 net: dsa: microchip: Transform register for use with KSZ8463
In reply to [email protected]From: Tristram Ha <[email protected]>
KSZ8463 uses a byte-enable mechanism to specify 8-bit, 16-bit, and 32-bit access. The register is first shifted right by 2 then left by 4. Extra 4 bits are added. If the access is 8-bit one of the 4 bits is set. If the access is 16-bit two of the 4 bits are set. If the access is 32-bit all 4 bits are set. The SPI command for read or write is then added.
As there are no automatic address increase and continuous SPI transfer the 64-bit access needs to broken into 2 32-bit accesses.
All common access functions in ksz_common.h are updated to include register transformation call so that the correct register access is used for KSZ8463.
KSZ8463/KSZ8795/KSZ8863/KSZ8895 common code for switch access is updated to include the register transformation function if common access functions in ksz_common.h are not used. In addition PORT_CTRL_ADDR is replaced with the get_port_addr helper function as KSZ8463 has different port register arrangement.
--- drivers/net/dsa/microchip/ksz8.c | 10 ++-- drivers/net/dsa/microchip/ksz_common.c | 5 +- drivers/net/dsa/microchip/ksz_common.h | 71 +++++++++++++++++++++----- 3 files changed, 67 insertions(+), 19 deletions(-)
drivers/net/dsa/microchip/ksz8.c +6 -4
@@ -36,13 +36,15 @@static void ksz_cfg (struct ksz_device *dev,u32 addr,u8 bits,bool set) { -regmap_update_bits (ksz_regmap_8 (dev), addr, bits, set ? bits :0 ); +regmap_update_bits (ksz_regmap_8 (dev),reg8 (dev, addr), bits, + set ? bits :0 ); }static void ksz_port_cfg (struct ksz_device *dev,int port,int offset,u8 bits,bool set) { -regmap_update_bits (ksz_regmap_8 (dev),PORT_CTRL_ADDR (port, offset), +regmap_update_bits (ksz_regmap_8 (dev), +reg8 (dev, dev->dev_ops->get_port_addr (port, offset)), bits, set ? bits :0 ); }
@@ -1888,14 +1890,14 @@ int ksz8_setup(struct dsa_switch *ds)ksz_cfg (dev,S_LINK_AGING_CTRL ,SW_LINK_AUTO_AGING , true);/* Enable aggressive back off algorithm in half duplex mode. */ -regmap_update_bits (ksz_regmap_8 (dev),REG_SW_CTRL_1 , +regmap_update_bits (ksz_regmap_8 (dev),reg8 (dev,REG_SW_CTRL_1 ),SW_AGGR_BACKOFF ,SW_AGGR_BACKOFF );/* * Make sure unicast VLAN boundary is set as default and * enable no excessive collision drop. */ -regmap_update_bits (ksz_regmap_8 (dev),REG_SW_CTRL_2 , +regmap_update_bits (ksz_regmap_8 (dev),reg8 (dev,REG_SW_CTRL_2 ),UNICAST_VLAN_BOUNDARY |NO_EXC_COLLISION_DROP ,UNICAST_VLAN_BOUNDARY |NO_EXC_COLLISION_DROP );
drivers/net/dsa/microchip/ksz_common.c +3 -2
@@ -2991,7 +2991,8 @@ static int ksz_setup(struct dsa_switch *ds) ds->num_tx_queues = dev->info->num_tx_queues; -regmap_update_bits (ksz_regmap_8 (dev), regs[S_MULTICAST_CTRL ], +regmap_update_bits (ksz_regmap_8 (dev), +reg8 (dev, regs[S_MULTICAST_CTRL ]),MULTICAST_STORM_DISABLE ,MULTICAST_STORM_DISABLE );ksz_init_mib_timer (dev);
@@ -3051,7 +3052,7 @@ static int ksz_setup(struct dsa_switch *ds) goto out_ptp_clock_unregister;/* start switch */ -regmap_update_bits (ksz_regmap_8 (dev), regs[S_START_CTRL ], +regmap_update_bits (ksz_regmap_8 (dev),reg8 (dev, regs[S_START_CTRL ]),SW_START ,SW_START );return 0 ;
drivers/net/dsa/microchip/ksz_common.h +58 -13
150 lines not on this page · raw message
Signature · 1 line
2.34.1
Andrew Lunn
maintainerRe: [PATCH net-next v4 3/7] net: dsa: microchip: Transform register for use with KSZ8463
Quoted 21 lines
drivers/net/dsa/microchip/ksz_common.h +21
-0+static inline bool ksz_is_ksz8463 (struct ksz_device *dev) +{ +return dev->chip_id ==KSZ8463_CHIP_ID ; +} + +static inline u32 reg8 (struct ksz_device *dev,u32 reg) +{ +if (ksz_is_ksz8463 (dev)) +return ((reg >>2 ) <<4 ) | (1 << (reg &3 )); +return reg; +} + +static inline u32 reg16 (struct ksz_device *dev,u32 reg) +{ +if (ksz_is_ksz8463 (dev)) +return ((reg >>2 ) <<4 ) | (reg &2 ?0x0c :0x03 ); +return reg; +} + +static inline u32 reg32 (struct ksz_device *dev,u32 reg) +{
drivers/net/dsa/microchip/ksz_common.h static inline struct regmap *ksz_regmap_32(struct ksz_device *dev) · line 517
Quoted 10 lines
drivers/net/dsa/microchip/ksz_common.h +6
-1+if (ksz_is_ksz8463 (dev)) +return ((reg >>2 ) <<4 ) |0xf ; +return reg; +} +static inline int ksz_read8 (struct ksz_device *dev,u32 reg,u8 *val) {unsigned int value; -int ret =regmap_read (ksz_regmap_8 (dev), reg, &value); +int ret =regmap_read (ksz_regmap_8 (dev),reg8 (dev, reg), &value);
I'm wondering if there is a less intrusive way to do this. When you create a regmap, you can optionally pass it methods to use for read/write/update etc.
struct regmap_config {
...
int (*reg_read)(void *context, unsigned int reg, unsigned int *val); int (*reg_write)(void *context, unsigned int reg, unsigned int val); int (*reg_update_bits)(void *context, unsigned int reg, unsigned int mask, unsigned int val);
Could you provide your own methods for the ksz8463 which perform the register modification, and then call the normal regmap SPI function to do the operation?
If you cannot get direct access to the regmap SPI functions, you can stack one regmap on top of another regmap. Have the top regmap do the register modifications, and then call a normal SPI regmap to do the read/write.
What i don't like about the current code is that developers adding new code could miss they need to add reg8().. to all regmap calls. So ideally you want to hide that away so they don't need to care, it just works.
Andrew
PATCH v4 4/7 net: dsa: microchip: Use different registers for KSZ8463
In reply to [email protected]From: Tristram Ha <[email protected]>
KSZ8463 does not use same set of registers as KSZ8863 so it is necessary to change some registers when using KSZ8463.
---
v3
- Replace cpu_to_be16() with swab16() to avoid compiler warning
drivers/net/dsa/microchip/ksz8.c | 78 +++++++++++++++++++------- drivers/net/dsa/microchip/ksz_common.c | 35 +++++++++--- drivers/net/dsa/microchip/ksz_dcb.c | 10 +++- 3 files changed, 94 insertions(+), 29 deletions(-)
drivers/net/dsa/microchip/ksz8.c +57 -21
@@ -143,6 +143,11 @@ int ksz8_reset_switch(struct ksz_device *dev) KSZ8863_GLOBAL_SOFTWARE_RESET | KSZ8863_PCS_RESET, true);ksz_cfg (dev,KSZ8863_REG_SW_RESET ,KSZ8863_GLOBAL_SOFTWARE_RESET |KSZ8863_PCS_RESET , false); + }else if (ksz_is_ksz8463 (dev)) { +ksz_cfg (dev,KSZ8463_REG_SW_RESET , +KSZ8463_GLOBAL_SOFTWARE_RESET , true); +ksz_cfg (dev,KSZ8463_REG_SW_RESET , +KSZ8463_GLOBAL_SOFTWARE_RESET , false); }else {/* reset switch */ ksz_write8(dev, REG_POWER_MANAGEMENT_1,
@@ -231,6 +236,11 @@ static int ksz8_port_queue_split(struct ksz_device *dev, int port, int queues) WEIGHTED_FAIR_QUEUE_ENABLE);if (ret)return ret; + }else if (ksz_is_ksz8463 (dev)) { + mask_4q =KSZ8873_PORT_4QUEUE_SPLIT_EN ; + mask_2q =KSZ8873_PORT_2QUEUE_SPLIT_EN ; + reg_4q =P1CR1 ; + reg_2q =P1CR1 +1 ; }else { mask_4q =KSZ8795_PORT_4QUEUE_SPLIT_EN ; mask_2q =KSZ8795_PORT_2QUEUE_SPLIT_EN ;
@@ -1269,12 +1279,15 @@ int ksz8_w_phy(struct ksz_device *dev, u16 phy, u16 reg, u16 val)void ksz8_cfg_port_member (struct ksz_device *dev,int port,u8 member) { +int offset =P_MIRROR_CTRL ;u8 data; -ksz_pread8 (dev, port,P_MIRROR_CTRL , &data); - data &= ~PORT_VLAN_MEMBERSHIP ; +if (ksz_is_ksz8463 (dev)) + offset =P1CR2 ; +ksz_pread8 (dev, port, offset, &data); + data &= ~dev->port_mask; data |= (member & dev->port_mask); -ksz_pwrite8 (dev, port,P_MIRROR_CTRL , data); +ksz_pwrite8 (dev, port, offset, data); }void ksz8_flush_dyn_mac_table (struct ksz_device *dev,int port)
@@ -1282,6 +1295,8 @@ void ksz8_flush_dyn_mac_table(struct ksz_device *dev, int port)u8 learn[DSA_MAX_PORTS ];int first, index, cnt;const u16 *regs; +int reg =S_FLUSH_TABLE_CTRL ; +int mask =SW_FLUSH_DYN_MAC_TABLE ; regs = dev->info->regs;
@@ -1299,7 +1314,11 @@ void ksz8_flush_dyn_mac_table(struct ksz_device *dev, int port)ksz_pwrite8 (dev, index, regs[P_STP_CTRL ], learn[index] |PORT_LEARN_DISABLE ); } -ksz_cfg (dev,S_FLUSH_TABLE_CTRL ,SW_FLUSH_DYN_MAC_TABLE , true); +if (ksz_is_ksz8463 (dev)) { + reg =KSZ8463_FLUSH_TABLE_CTRL ; + mask =KSZ8463_FLUSH_DYN_MAC_TABLE ; + } +ksz_cfg (dev, reg, mask, true);for (index = first; index < cnt; index++) {if (!(learn[index] &PORT_LEARN_DISABLE ))ksz_pwrite8 (dev, index, regs[P_STP_CTRL ], learn[index]);
@@ -1438,7 +1457,7 @@ int ksz8_fdb_del(struct ksz_device *dev, int port, const unsigned char *addr,int ksz8_port_vlan_filtering (struct ksz_device *dev,int port,bool flag,struct netlink_ext_ack *extack) { -if (ksz_is_ksz88x3 (dev)) +if (ksz_is_ksz88x3 (dev) ||ksz_is_ksz8463 (dev))return -ENOTSUPP ;/* Discard packets with VID not enabled on the switch */
@@ -1454,9 +1473,12 @@ int ksz8_port_vlan_filtering(struct ksz_device *dev, int port, bool flag,static void ksz8_port_enable_pvid (struct ksz_device *dev,int port,bool state) { -if (ksz_is_ksz88x3 (dev)) { -ksz_cfg (dev,REG_SW_INSERT_SRC_PVID , -0x03 << (4 -2 * port), state); +if (ksz_is_ksz88x3 (dev) ||ksz_is_ksz8463 (dev)) { +int reg =REG_SW_INSERT_SRC_PVID ; + +if (ksz_is_ksz8463 (dev)) + reg =KSZ8463_REG_SW_CTRL_9 ; +ksz_cfg (dev, reg,0x03 << (4 -2 * port), state); }else {ksz_pwrite8 (dev, port,REG_PORT_CTRL_12 , state ?0x0f :0x00 ); }
@@ -1471,7 +1493,7 @@ int ksz8_port_vlan_add(struct ksz_device *dev, int port,u16 data, new_pvid =0 ;u8 fid, member, valid; -if (ksz_is_ksz88x3 (dev)) +if (ksz_is_ksz88x3 (dev) ||ksz_is_ksz8463 (dev))return -ENOTSUPP ;/* If a VLAN is added with untagged flag different from the
@@ -1540,7 +1562,7 @@ int ksz8_port_vlan_del(struct ksz_device *dev, int port,u16 data, pvid;u8 fid, member, valid; -if (ksz_is_ksz88x3 (dev)) +if (ksz_is_ksz88x3 (dev) ||ksz_is_ksz8463 (dev))return -ENOTSUPP ;ksz_pread16 (dev, port,REG_PORT_CTRL_VID , &pvid);
@@ -1570,19 +1592,23 @@ int ksz8_port_mirror_add(struct ksz_device *dev, int port,struct dsa_mall_mirror_tc_entry *mirror, bool ingress, struct netlink_ext_ack *extack) { + int offset = P_MIRROR_CTRL; + +if (ksz_is_ksz8463 (dev)) + offset =P1CR2 ;if (ingress) { -ksz_port_cfg (dev ,port ,P_MIRROR_CTRL ,PORT_MIRROR_RX ,true ); +ksz_port_cfg (dev, port, offset,PORT_MIRROR_RX , true); dev->mirror_rx |=BIT (port); }else { -ksz_port_cfg (dev, port,P_MIRROR_CTRL ,PORT_MIRROR_TX , true); +ksz_port_cfg (dev, port, offset,PORT_MIRROR_TX , true); dev->mirror_tx |=BIT (port); } -ksz_port_cfg (dev, port,P_MIRROR_CTRL ,PORT_MIRROR_SNIFFER , false); +ksz_port_cfg (dev, port, offset,PORT_MIRROR_SNIFFER , false);/* configure mirror port */ if (dev->mirror_rx || dev->mirror_tx) -ksz_port_cfg (dev, mirror->to_local_port,P_MIRROR_CTRL , +ksz_port_cfg (dev, mirror->to_local_port, offset,PORT_MIRROR_SNIFFER , true);return 0 ;
@@ -1591,20 +1617,23 @@ int ksz8_port_mirror_add(struct ksz_device *dev, int port,void ksz8_port_mirror_del (struct ksz_device *dev,int port,struct dsa_mall_mirror_tc_entry *mirror) { +int offset =P_MIRROR_CTRL ;u8 data; +if (ksz_is_ksz8463 (dev)) + offset =P1CR2 ;if (mirror->ingress) { -ksz_port_cfg (dev, port,P_MIRROR_CTRL ,PORT_MIRROR_RX , false); +ksz_port_cfg (dev, port, offset,PORT_MIRROR_RX , false); dev->mirror_rx &= ~BIT (port); }else { -ksz_port_cfg (dev, port,P_MIRROR_CTRL ,PORT_MIRROR_TX , false); +ksz_port_cfg (dev, port, offset,PORT_MIRROR_TX , false); dev->mirror_tx &= ~BIT (port); } -ksz_pread8 (dev, port,P_MIRROR_CTRL , &data); +ksz_pread8 (dev, port, offset, &data);if (!dev->mirror_rx && !dev->mirror_tx) -ksz_port_cfg (dev, mirror->to_local_port,P_MIRROR_CTRL , +ksz_port_cfg (dev, mirror->to_local_port, offset,PORT_MIRROR_SNIFFER , false); }
@@ -1629,17 +1658,24 @@ void ksz8_port_setup(struct ksz_device *dev, int port, bool cpu_port)const u16 *regs = dev->info->regs;struct dsa_switch *ds = dev->ds;const u32 *masks; +int offset;u8 member; masks = dev->info->masks;/* enable broadcast storm limit */ -ksz_port_cfg (dev, port,P_BCAST_STORM_CTRL ,PORT_BROADCAST_STORM , true); + offset =P_BCAST_STORM_CTRL ; +if (ksz_is_ksz8463 (dev)) + offset =P1CR1 ; +ksz_port_cfg (dev, port, offset,PORT_BROADCAST_STORM , true);ksz8_port_queue_split (dev, port, dev->info->num_tx_queues);/* replace priority */ -ksz_port_cfg (dev, port,P_802_1P_CTRL , + offset =P_802_1P_CTRL ; +if (ksz_is_ksz8463 (dev)) + offset =P1CR2 ; +ksz_port_cfg (dev, port, offset, masks[PORT_802_1P_REMAPPING ], false);if (cpu_port)
@@ -1905,7 +1941,7 @@ int ksz8_setup(struct dsa_switch *ds)ksz_cfg (dev,S_MIRROR_CTRL ,SW_MIRROR_RX_TX , false); -if (!ksz_is_ksz88x3 (dev)) +if (!ksz_is_ksz88x3 (dev) && !ksz_is_ksz8463 (dev))ksz_cfg (dev,REG_SW_CTRL_19 ,SW_INS_TAG_ENABLE , true);for (i =0 ; i < (dev->info->num_vlans /4 ); i++)
drivers/net/dsa/microchip/ksz_common.c +28 -7
@@ -2951,6 +2951,7 @@ static int ksz_parse_drive_strength(struct ksz_device *dev);static int ksz_setup (struct dsa_switch *ds) {struct ksz_device *dev = ds->priv; +u16 storm_mask, storm_rate;struct dsa_port *dp;struct ksz_port *p;const u16 *regs;
@@ -2980,10 +2981,15 @@ static int ksz_setup(struct dsa_switch *ds) }/* set broadcast storm protection 10% rate */ -regmap_update_bits (ksz_regmap_16 (dev), regs[S_BROADCAST_CTRL ], -BROADCAST_STORM_RATE , - (BROADCAST_STORM_VALUE * -BROADCAST_STORM_PROT_RATE ) /100 ); + storm_mask =BROADCAST_STORM_RATE ; + storm_rate = (BROADCAST_STORM_VALUE *BROADCAST_STORM_PROT_RATE ) /100 ; +if (ksz_is_ksz8463 (dev)) { + storm_mask =swab16 (storm_mask); + storm_rate =swab16 (storm_rate); + } +regmap_update_bits (ksz_regmap_16 (dev), +reg16 (dev, regs[S_BROADCAST_CTRL ]), + storm_mask, storm_rate); dev->dev_ops->config_cpu_port (ds);
@@ -4222,6 +4228,17 @@ static int ksz_ets_band_to_queue(struct tc_ets_qopt_offload_replace_params *p,return p->bands -1 - band; } +static u8 ksz8463_tc_ctrl (int port,int queue) +{ +u8 reg; + + reg =0xC8 + port *4 ; + reg += ((3 - queue) /2 ) *2 ; + reg++; + reg -= (queue &1 ); +return reg; +} +/** * ksz88x3_tc_ets_add - Configure ETS (Enhanced Transmission Selection) * for a port on KSZ88x3 switch
@@ -4257,6 +4274,8 @@ static int ksz88x3_tc_ets_add(struct ksz_device *dev, int port,* port/queue */ reg =KSZ8873_TXQ_SPLIT_CTRL_REG (port, queue); +if (ksz_is_ksz8463 (dev)) + reg =ksz8463_tc_ctrl (port, queue);/* Clear WFQ enable bit to select strict priority scheduling */ ret =ksz_rmw8 (dev, reg,KSZ8873_TXQ_WFQ_ENABLE ,0 );
@@ -4292,6 +4311,8 @@ static int ksz88x3_tc_ets_del(struct ksz_device *dev, int port)* port/queue */ reg =KSZ8873_TXQ_SPLIT_CTRL_REG (port, queue); +if (ksz_is_ksz8463 (dev)) + reg =ksz8463_tc_ctrl (port, queue);/* Set WFQ enable bit to revert back to default scheduling * mode
@@ -4439,7 +4460,7 @@ static int ksz_tc_setup_qdisc_ets(struct dsa_switch *ds, int port,struct ksz_device *dev = ds->priv;int ret; -if (is_ksz8 (dev) && !ksz_is_ksz88x3 (dev)) +if (is_ksz8 (dev) && !(ksz_is_ksz88x3 (dev) ||ksz_is_ksz8463 (dev)))return -EOPNOTSUPP ;if (qopt->parent !=TC_H_ROOT ) {
@@ -4453,13 +4474,13 @@ static int ksz_tc_setup_qdisc_ets(struct dsa_switch *ds, int port,if (ret)return ret; -if (ksz_is_ksz88x3 (dev)) +if (ksz_is_ksz88x3 (dev) ||ksz_is_ksz8463 (dev))return ksz88x3_tc_ets_add (dev, port, &qopt->replace_params);else return ksz_tc_ets_add (dev, port, &qopt->replace_params);case TC_ETS_DESTROY : -if (ksz_is_ksz88x3 (dev)) +if (ksz_is_ksz88x3 (dev) ||ksz_is_ksz8463 (dev))return ksz88x3_tc_ets_del (dev, port);else return ksz_tc_ets_del (dev, port);
drivers/net/dsa/microchip/ksz_dcb.c +9 -1
@@ -16,10 +16,12 @@* Therefore, we define the base offset as 0x00 here to align with that logic. */ #define KSZ8_REG_PORT_1_CTRL_0 0x00 +#define KSZ8463_REG_PORT_1_CTRL_0 0x6C#define KSZ8_PORT_DIFFSERV_ENABLE BIT(6)#define KSZ8_PORT_802_1P_ENABLE BIT(5)#define KSZ8_PORT_BASED_PRIO_M GENMASK(4, 3) +#define KSZ8463_REG_TOS_DSCP_CTRL 0x16#define KSZ88X3_REG_TOS_DSCP_CTRL 0x60#define KSZ8765_REG_TOS_DSCP_CTRL 0x90
@@ -98,6 +100,8 @@ static void ksz_get_default_port_prio_reg(struct ksz_device *dev, int *reg, *reg =KSZ8_REG_PORT_1_CTRL_0 ; *mask =KSZ8_PORT_BASED_PRIO_M ; *shift =__bf_shf (KSZ8_PORT_BASED_PRIO_M ); +if (ksz_is_ksz8463 (dev)) + *reg =KSZ8463_REG_PORT_1_CTRL_0 ; }else { *reg =KSZ9477_REG_PORT_MRI_MAC_CTRL ; *mask =KSZ9477_PORT_BASED_PRIO_M ;
@@ -122,10 +126,12 @@ static void ksz_get_dscp_prio_reg(struct ksz_device *dev, int *reg, *reg =KSZ8765_REG_TOS_DSCP_CTRL ; *per_reg =4 ; *mask =GENMASK (1 ,0 ); - }else if (ksz_is_ksz88x3 (dev)) { + }else if (ksz_is_ksz88x3 (dev) ||ksz_is_ksz8463 (dev)) { *reg =KSZ88X3_REG_TOS_DSCP_CTRL ; *per_reg =4 ; *mask =GENMASK (1 ,0 ); +if (ksz_is_ksz8463 (dev)) + *reg =KSZ8463_REG_TOS_DSCP_CTRL ; }else { *reg =KSZ9477_REG_DIFFSERV_PRIO_MAP ; *per_reg =2 ;
@@ -151,6 +157,8 @@ static void ksz_get_apptrust_map_and_reg(struct ksz_device *dev, *map = ksz8_apptrust_map_to_bit; *reg =KSZ8_REG_PORT_1_CTRL_0 ; *mask =KSZ8_PORT_DIFFSERV_ENABLE |KSZ8_PORT_802_1P_ENABLE ; +if (ksz_is_ksz8463 (dev)) + *reg =KSZ8463_REG_PORT_1_CTRL_0 ; }else { *map = ksz9477_apptrust_map_to_bit; *reg =KSZ9477_REG_PORT_MRI_PRIO_CTRL ;
Signature · 1 line
2.34.1
Simon Horman
reviewerRe: [PATCH net-next v4 4/7] net: dsa: microchip: Use different registers for KSZ8463
Quoted 9 lines · On Fri, Jul 18, 2025 at 06:21:03PM -0700, [email protected] wrote:
From: Tristram Ha <[email protected]>
KSZ8463 does not use same set of registers as KSZ8863 so it is necessary to change some registers when using KSZ8463.
Signed-off-by Tristram Ha <[email protected]>---
v3
- Replacecpu_to_be16()withswab16()to avoid compiler warning
...
Quoted 1 line
drivers/net/dsa/microchip/ksz_common.c +0
-0
...
Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c +0
-3@@ -2980,10 +2981,15 @@ static int ksz_setup(struct dsa_switch *ds) }/* set broadcast storm protection 10% rate */ - regmap_update_bits(ksz_regmap_16(dev), regs[S_BROADCAST_CTRL], - BROADCAST_STORM_RATE, - (BROADCAST_STORM_VALUE *
drivers/net/dsa/microchip/ksz_common.c static int ksz_setup(struct dsa_switch *ds) · line 2992
Quoted 10 lines
drivers/net/dsa/microchip/ksz_common.c +9
-1- BROADCAST_STORM_PROT_RATE) / 100); + storm_mask =BROADCAST_STORM_RATE ; + storm_rate = (BROADCAST_STORM_VALUE *BROADCAST_STORM_PROT_RATE ) /100 ; +if (ksz_is_ksz8463 (dev)) { + storm_mask =swab16 (storm_mask); + storm_rate =swab16 (storm_rate); + } +regmap_update_bits (ksz_regmap_16 (dev), +reg16 (dev, regs[S_BROADCAST_CTRL ]), + storm_mask, storm_rate);
Hi Tristram,
I am confused by the use of swab16() here.
Let us say that we are running on a little endian host (likely). Then the effect of this is to pass big endian values to regmap_update_bits().
But if we are running on a big endian host, the opposite will be true:
little endian values will be passed to regmap_update_bits().
Looking at KSZ_REGMAP_ENTRY() I see:
#define KSZ_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \
{ \
...
.reg_format_endian = REGMAP_ENDIAN_BIG, \
.val_format_endian = REGMAP_ENDIAN_BIG \
}Which based on a skimming the regmap code implies to me that regmap_update_bits() should be passed host byte order values which regmap will convert to big endian when writing out these values.
It is unclear to me why changing the byte order of storm_mask and storm_rate is needed here. But it does seem clear that it will lead to inconsistent results on big endian and little endian hosts.
...
Simon Horman
reviewerRe: [PATCH net-next v4 4/7] net: dsa: microchip: Use different registers for KSZ8463
Quoted 46 lines · Simon Horman
Quoted 9 lines · On Fri, Jul 18, 2025 at 06:21:03PM -0700, [email protected] wrote:
From: Tristram Ha <[email protected]>
KSZ8463 does not use same set of registers as KSZ8863 so it is necessary to change some registers when using KSZ8463.
Signed-off-by Tristram Ha <[email protected]>---
v3
- Replacecpu_to_be16()withswab16()to avoid compiler warning...
Quoted 1 line
drivers/net/dsa/microchip/ksz_common.c +0
-0...
Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c +0
-3@@ -2980,10 +2981,15 @@ static int ksz_setup(struct dsa_switch *ds) }/* set broadcast storm protection 10% rate */ - regmap_update_bits(ksz_regmap_16(dev), regs[S_BROADCAST_CTRL], - BROADCAST_STORM_RATE, - (BROADCAST_STORM_VALUE *drivers/net/dsa/microchip/ksz_common.c static int ksz_setup(struct dsa_switch *ds) · line 2992
Quoted 10 lines
drivers/net/dsa/microchip/ksz_common.c +9
-1- BROADCAST_STORM_PROT_RATE) / 100); + storm_mask =BROADCAST_STORM_RATE ; + storm_rate = (BROADCAST_STORM_VALUE *BROADCAST_STORM_PROT_RATE ) /100 ; +if (ksz_is_ksz8463 (dev)) { + storm_mask =swab16 (storm_mask); + storm_rate =swab16 (storm_rate); + } +regmap_update_bits (ksz_regmap_16 (dev), +reg16 (dev, regs[S_BROADCAST_CTRL ]), + storm_mask, storm_rate);Hi Tristram,
I am confused by the use of
swab16()here.Let us say that we are running on a little endian host (likely). Then the effect of this is to pass big endian values to
regmap_update_bits().But if we are running on a big endian host, the opposite will be true:
little endian values will be passed toregmap_update_bits().
Quoted 8 lines
Looking at
KSZ_REGMAP_ENTRY()I see:#define KSZ_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ { \ ... .reg_format_endian = REGMAP_ENDIAN_BIG, \ .val_format_endian = REGMAP_ENDIAN_BIG \ }
Update; I now see this in another patch of the series:
+#define KSZ8463_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \
+ { \
...
+ .reg_format_endian = REGMAP_ENDIAN_BIG, \
+ .val_format_endian = REGMAP_ENDIAN_LITTLE \+ }
Which I understand to mean that the hardware is expecting little endian values. But still, my concerns raised in my previous email of this thread remain.
And I have a question: does this chip use little endian register values whereas other chips used big endian register values?
Quoted 13 lines
Which based on a skimming the regmap code implies to me that
regmap_update_bits()should be passed host byte order values which regmap will convert to big endian when writing out these values.It is unclear to me why changing the byte order of
storm_maskandstorm_rateis needed here. But it does seem clear that it will lead to inconsistent results on big endian and little endian hosts....
RE: [PATCH net-next v4 4/7] net: dsa: microchip: Use different registers for KSZ8463
Quoted 67 lines · Simon Horman
Quoted 6 lines · On Fri, Jul 18, 2025 at 06:21:03PM -0700, [email protected] wrote:
Quoted 5 lines
From: Tristram Ha <[email protected]>
KSZ8463 does not use same set of registers as KSZ8863 so it is necessary to change some registers when using KSZ8463.
Quoted 8 lines
Quoted 4 lines
Signed-off-by Tristram Ha <[email protected]>---
v3
- Replacecpu_to_be16()withswab16()to avoid compiler warning...
Review
Quoted 1 line
(quoted diff) +0
-0
b/drivers/net/dsa/microchip/ksz_common.cQuoted 6 lines
...
Quoted 3 lines
(quoted diff) +0
-0@@ -2980,10 +2981,15 @@ static int ksz_setup(struct dsa_switch *ds) }Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c static int ksz_setup(struct dsa_switch *ds)
Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c +2
-4/* set broadcast storm protection 10% rate */ -regmap_update_bits (ksz_regmap_16 (dev), regs[S_BROADCAST_CTRL ], -BROADCAST_STORM_RATE , - (BROADCAST_STORM_VALUE * -BROADCAST_STORM_PROT_RATE ) /100 ); + storm_mask =BROADCAST_STORM_RATE ; + storm_rate = (BROADCAST_STORM_VALUE *
BROADCAST_STORM_PROT_RATE) / 100;Quoted 19 lines
drivers/net/dsa/microchip/ksz_common.c static int ksz_setup(struct dsa_switch *ds) · line 2992
Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c +7
-0+if (ksz_is_ksz8463 (dev)) { + storm_mask =swab16 (storm_mask); + storm_rate =swab16 (storm_rate); + } +regmap_update_bits (ksz_regmap_16 (dev), +reg16 (dev, regs[S_BROADCAST_CTRL ]), + storm_mask, storm_rate);Hi Tristram,
I am confused by the use of
swab16()here.Let us say that we are running on a little endian host (likely). Then the effect of this is to pass big endian values to
regmap_update_bits().But if we are running on a big endian host, the opposite will be true:
little endian values will be passed toregmap_update_bits().Quoted 8 lines
Looking at
KSZ_REGMAP_ENTRY()I see:#define KSZ_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ { \ ... .reg_format_endian = REGMAP_ENDIAN_BIG, \ .val_format_endian = REGMAP_ENDIAN_BIG \ }Update; I now see this in another patch of the series:
(quoted diff) +5
-0+#define KSZ8463_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ + { \ + .reg_format_endian = REGMAP_ENDIAN_BIG, \ + .val_format_endian = REGMAP_ENDIAN_LITTLE \ + }
Quoted 17 lines
Which I understand to mean that the hardware is expecting little endian values. But still, my concerns raised in my previous email of this thread remain.
And I have a question: does this chip use little endian register values whereas other chips used big endian register values?
Quoted 10 lines
Which based on a skimming the regmap code implies to me that
regmap_update_bits()should be passed host byte order values which regmap will convert to big endian when writing out these values.It is unclear to me why changing the byte order of
storm_maskandstorm_rateis needed here. But it does seem clear that it will lead to inconsistent results on big endian and little endian hosts.
The broadcast storm value 0x7ff is stored in registers 6 and 7 in KSZ8863 where register 6 holds the 0x7 part while register 7 holds the 0xff part. In KSZ8463 register 6 is defined as 16-bit where the 0x7 part is held in lower byte and the 0xff part is held in higher byte. It is necessary to swap the bytes when the value is passed to the 16-bit write function.
All other KSZ switches use 8-bit access with automatic address increase so a write to register 0 with value 0x12345678 means 0=0x12, 1=0x34, 2=0x56, and 3=0x78.
Simon Horman
reviewerRe: [PATCH net-next v4 4/7] net: dsa: microchip: Use different registers for KSZ8463
Quoted 85 lines · On Wed, Jul 23, 2025 at 02:25:27AM +0000, [email protected] wrote:
Quoted 67 lines · Simon Horman
Quoted 6 lines · On Fri, Jul 18, 2025 at 06:21:03PM -0700, [email protected] wrote:
Quoted 5 lines
From: Tristram Ha <[email protected]>
KSZ8463 does not use same set of registers as KSZ8863 so it is necessary to change some registers when using KSZ8463.
Quoted 8 lines
Quoted 4 lines
Signed-off-by Tristram Ha <[email protected]>---
v3
- Replacecpu_to_be16()withswab16()to avoid compiler warning...
Review
Quoted 1 line
(quoted diff) +0
-0
b/drivers/net/dsa/microchip/ksz_common.cQuoted 6 lines
...
Quoted 3 lines
(quoted diff) +0
-0@@ -2980,10 +2981,15 @@ static int ksz_setup(struct dsa_switch *ds) }Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c static int ksz_setup(struct dsa_switch *ds)
Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c +2
-4/* set broadcast storm protection 10% rate */ -regmap_update_bits (ksz_regmap_16 (dev), regs[S_BROADCAST_CTRL ], -BROADCAST_STORM_RATE , - (BROADCAST_STORM_VALUE * -BROADCAST_STORM_PROT_RATE ) /100 ); + storm_mask =BROADCAST_STORM_RATE ; + storm_rate = (BROADCAST_STORM_VALUE *
BROADCAST_STORM_PROT_RATE) / 100;Quoted 19 lines
drivers/net/dsa/microchip/ksz_common.c static int ksz_setup(struct dsa_switch *ds) · line 2992
Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c +7
-0+if (ksz_is_ksz8463 (dev)) { + storm_mask =swab16 (storm_mask); + storm_rate =swab16 (storm_rate); + } +regmap_update_bits (ksz_regmap_16 (dev), +reg16 (dev, regs[S_BROADCAST_CTRL ]), + storm_mask, storm_rate);Hi Tristram,
I am confused by the use of
swab16()here.Let us say that we are running on a little endian host (likely). Then the effect of this is to pass big endian values to
regmap_update_bits().But if we are running on a big endian host, the opposite will be true:
little endian values will be passed toregmap_update_bits().Quoted 8 lines
Looking at
KSZ_REGMAP_ENTRY()I see:#define KSZ_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ { \ ... .reg_format_endian = REGMAP_ENDIAN_BIG, \ .val_format_endian = REGMAP_ENDIAN_BIG \ }Update; I now see this in another patch of the series:
(quoted diff) +5
-0+#define KSZ8463_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ + { \ + .reg_format_endian = REGMAP_ENDIAN_BIG, \ + .val_format_endian = REGMAP_ENDIAN_LITTLE \ + }Quoted 17 lines
Which I understand to mean that the hardware is expecting little endian values. But still, my concerns raised in my previous email of this thread remain.
And I have a question: does this chip use little endian register values whereas other chips used big endian register values?
Quoted 10 lines
Which based on a skimming the regmap code implies to me that
regmap_update_bits()should be passed host byte order values which regmap will convert to big endian when writing out these values.It is unclear to me why changing the byte order of
storm_maskandstorm_rateis needed here. But it does seem clear that it will lead to inconsistent results on big endian and little endian hosts.
Quoted 5 lines
The broadcast storm value
0x7ffis stored in registers 6 and 7 in KSZ8863 where register 6 holds the0x7part while register 7 holds the0xffpart. In KSZ8463 register 6 is defined as 16-bit where the0x7part is held in lower byte and the0xffpart is held in higher byte. It is necessary to swap the bytes when the value is passed to the 16-bit write function.
Perhaps naively, I would have expected
.val_format_endian = REGMAP_ENDIAN_LITTLE
to handle writing the 16-bit value 0x7ff such that 0x7 is in the lower byte, while 0xff is in the upper byte. Is that not the case?
If not, do you get the desired result by removing the swab16() calls and using
.val_format_endian = REGMAP_ENDIAN_BIG
But perhaps I misunderstand how .val_format_endian works.
Quoted 5 lines
All other KSZ switches use 8-bit access with automatic address increase so a write to register 0 with value
0x12345678means 0=0x12, 1=0x34, 2=0x56, and 3=0x78.
RE: [PATCH net-next v4 4/7] net: dsa: microchip: Use different registers for KSZ8463
Quoted 99 lines · On Wed, Jul 23, 2025 at 02:25:27AM +0000, [email protected] wrote:
Quoted 85 lines
Quoted 67 lines · Simon Horman
Quoted 6 lines · On Fri, Jul 18, 2025 at 06:21:03PM -0700, [email protected] wrote:
Quoted 5 lines
From: Tristram Ha <[email protected]>
KSZ8463 does not use same set of registers as KSZ8863 so it is necessary to change some registers when using KSZ8463.
Quoted 8 lines
Quoted 4 lines
Signed-off-by Tristram Ha <[email protected]>---
v3
- Replacecpu_to_be16()withswab16()to avoid compiler warning...
Review
Quoted 1 line
(quoted diff) +0
-0
b/drivers/net/dsa/microchip/ksz_common.cQuoted 6 lines
...
Quoted 3 lines
(quoted diff) +0
-0@@ -2980,10 +2981,15 @@ static int ksz_setup(struct dsa_switch *ds) }Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c static int ksz_setup(struct dsa_switch *ds)
Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c +2
-4/* set broadcast storm protection 10% rate */ -regmap_update_bits (ksz_regmap_16 (dev), regs[S_BROADCAST_CTRL ], -BROADCAST_STORM_RATE , - (BROADCAST_STORM_VALUE * -BROADCAST_STORM_PROT_RATE ) /100 ); + storm_mask =BROADCAST_STORM_RATE ; + storm_rate = (BROADCAST_STORM_VALUE *
BROADCAST_STORM_PROT_RATE) / 100;Quoted 19 lines
drivers/net/dsa/microchip/ksz_common.c static int ksz_setup(struct dsa_switch *ds) · line 2992
Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c +7
-0+if (ksz_is_ksz8463 (dev)) { + storm_mask =swab16 (storm_mask); + storm_rate =swab16 (storm_rate); + } +regmap_update_bits (ksz_regmap_16 (dev), +reg16 (dev, regs[S_BROADCAST_CTRL ]), + storm_mask, storm_rate);Hi Tristram,
I am confused by the use of
swab16()here.Let us say that we are running on a little endian host (likely). Then the effect of this is to pass big endian values to
regmap_update_bits().But if we are running on a big endian host, the opposite will be true:
little endian values will be passed toregmap_update_bits().Quoted 8 lines
Looking at
KSZ_REGMAP_ENTRY()I see:#define KSZ_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ { \ ... .reg_format_endian = REGMAP_ENDIAN_BIG, \ .val_format_endian = REGMAP_ENDIAN_BIG \ }Update; I now see this in another patch of the series:
(quoted diff) +5
-0+#define KSZ8463_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ + { \ + .reg_format_endian = REGMAP_ENDIAN_BIG, \ + .val_format_endian = REGMAP_ENDIAN_LITTLE \ + }Quoted 17 lines
Which I understand to mean that the hardware is expecting little endian values. But still, my concerns raised in my previous email of this thread remain.
And I have a question: does this chip use little endian register values whereas other chips used big endian register values?
Quoted 10 lines
Which based on a skimming the regmap code implies to me that
regmap_update_bits()should be passed host byte order values which regmap will convert to big endian when writing out these values.It is unclear to me why changing the byte order of
storm_maskandstorm_rateis needed here. But it does seem clear that it will lead to inconsistent results on big endian and little endian hosts.Quoted 5 lines
The broadcast storm value
0x7ffis stored in registers 6 and 7 in KSZ8863 where register 6 holds the0x7part while register 7 holds the0xffpart. In KSZ8463 register 6 is defined as 16-bit where the0x7part is held in lower byte and the0xffpart is held in higher byte. It is necessary to swap the bytes when the value is passed to the 16-bit write function.Perhaps naively, I would have expected
.val_format_endian =
REGMAP_ENDIAN_LITTLEto handle writing the 16-bit value
0x7ffsuch that0x7is in the lower byte, while0xffis in the upper byte. Is that not the case?
Quoted 11 lines
If not, do you get the desired result by removing the
swab16()calls and using.val_format_endian =
REGMAP_ENDIAN_BIGBut perhaps I misunderstand how .val_format_endian works.
Quoted 4 lines
All other KSZ switches use 8-bit access with automatic address increase so a write to register 0 with value
0x12345678means 0=0x12, 1=0x34, 2=0x56, and 3=0x78.
It is not about big-endian or little-endian. It is just the presentation of this register is different between KSZ8863 and KSZ8463. KSZ8863 uses big-endian for register value as the access is 8-bit and the address is automatically increased by 1. Writing a value 0x03ff to register 6 means 6=0x03 and 7=0xff. The actual SPI transfer commands are "02 06 03 ff." KSZ8463 uses little-endian for register value as the access is fixed at 8-bit, 16-bit, or 32-bit. Writing 0x03ff results in the actual SPI transfer commands "80 70 ff 03" where the correct commands are "80 70 03 ff."
Simon Horman
reviewerRe: [PATCH net-next v4 4/7] net: dsa: microchip: Use different registers for KSZ8463
Quoted 111 lines · On Thu, Jul 24, 2025 at 02:28:56AM +0000, [email protected] wrote:
Quoted 99 lines · On Wed, Jul 23, 2025 at 02:25:27AM +0000, [email protected] wrote:
Quoted 85 lines
Quoted 67 lines · Simon Horman
Quoted 6 lines · On Fri, Jul 18, 2025 at 06:21:03PM -0700, [email protected] wrote:
Quoted 5 lines
From: Tristram Ha <[email protected]>
KSZ8463 does not use same set of registers as KSZ8863 so it is necessary to change some registers when using KSZ8463.
Quoted 8 lines
Quoted 4 lines
Signed-off-by Tristram Ha <[email protected]>---
v3
- Replacecpu_to_be16()withswab16()to avoid compiler warning...
Review
Quoted 1 line
(quoted diff) +0
-0
b/drivers/net/dsa/microchip/ksz_common.cQuoted 6 lines
...
Quoted 3 lines
(quoted diff) +0
-0@@ -2980,10 +2981,15 @@ static int ksz_setup(struct dsa_switch *ds) }Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c static int ksz_setup(struct dsa_switch *ds)
Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c +2
-4/* set broadcast storm protection 10% rate */ -regmap_update_bits (ksz_regmap_16 (dev), regs[S_BROADCAST_CTRL ], -BROADCAST_STORM_RATE , - (BROADCAST_STORM_VALUE * -BROADCAST_STORM_PROT_RATE ) /100 ); + storm_mask =BROADCAST_STORM_RATE ; + storm_rate = (BROADCAST_STORM_VALUE *
BROADCAST_STORM_PROT_RATE) / 100;Quoted 19 lines
drivers/net/dsa/microchip/ksz_common.c static int ksz_setup(struct dsa_switch *ds) · line 2992
Quoted 7 lines
drivers/net/dsa/microchip/ksz_common.c +7
-0+if (ksz_is_ksz8463 (dev)) { + storm_mask =swab16 (storm_mask); + storm_rate =swab16 (storm_rate); + } +regmap_update_bits (ksz_regmap_16 (dev), +reg16 (dev, regs[S_BROADCAST_CTRL ]), + storm_mask, storm_rate);Hi Tristram,
I am confused by the use of
swab16()here.Let us say that we are running on a little endian host (likely). Then the effect of this is to pass big endian values to
regmap_update_bits().But if we are running on a big endian host, the opposite will be true:
little endian values will be passed toregmap_update_bits().Quoted 8 lines
Looking at
KSZ_REGMAP_ENTRY()I see:#define KSZ_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ { \ ... .reg_format_endian = REGMAP_ENDIAN_BIG, \ .val_format_endian = REGMAP_ENDIAN_BIG \ }Update; I now see this in another patch of the series:
(quoted diff) +5
-0+#define KSZ8463_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ + { \ + .reg_format_endian = REGMAP_ENDIAN_BIG, \ + .val_format_endian = REGMAP_ENDIAN_LITTLE \ + }Quoted 17 lines
Which I understand to mean that the hardware is expecting little endian values. But still, my concerns raised in my previous email of this thread remain.
And I have a question: does this chip use little endian register values whereas other chips used big endian register values?
Quoted 10 lines
Which based on a skimming the regmap code implies to me that
regmap_update_bits()should be passed host byte order values which regmap will convert to big endian when writing out these values.It is unclear to me why changing the byte order of
storm_maskandstorm_rateis needed here. But it does seem clear that it will lead to inconsistent results on big endian and little endian hosts.Quoted 5 lines
The broadcast storm value
0x7ffis stored in registers 6 and 7 in KSZ8863 where register 6 holds the0x7part while register 7 holds the0xffpart. In KSZ8463 register 6 is defined as 16-bit where the0x7part is held in lower byte and the0xffpart is held in higher byte. It is necessary to swap the bytes when the value is passed to the 16-bit write function.Perhaps naively, I would have expected
.val_format_endian =
REGMAP_ENDIAN_LITTLEto handle writing the 16-bit value
0x7ffsuch that0x7is in the lower byte, while0xffis in the upper byte. Is that not the case?Quoted 11 lines
If not, do you get the desired result by removing the
swab16()calls and using.val_format_endian =
REGMAP_ENDIAN_BIGBut perhaps I misunderstand how .val_format_endian works.
Quoted 4 lines
All other KSZ switches use 8-bit access with automatic address increase so a write to register 0 with value
0x12345678means 0=0x12, 1=0x34, 2=0x56, and 3=0x78.
Quoted 9 lines
It is not about big-endian or little-endian. It is just the presentation of this register is different between KSZ8863 and KSZ8463. KSZ8863 uses big-endian for register value as the access is 8-bit and the address is automatically increased by 1. Writing a value
0x03ffto register 6 means 6=0x03 and 7=0xff. The actual SPI transfer commands are "02 06 03 ff." KSZ8463 uses little-endian for register value as the access is fixed at 8-bit, 16-bit, or 32-bit. Writing0x03ffresults in the actual SPI transfer commands "80 70 ff 03" where the correct commands are "80 70 03 ff."
The difference between expressing a 16-bit value as "ff 03" and "03 ff" sounds a lot like endianness to me.
"ff 03" is the little endian representation of 0x3ff. "03 ff" is the big endian representation of 0x3ff.
I am very confused as to why you say "KSZ8463 uses little-endian for register value". And then go on to say that the correct transfer command is "02 06 03 ff", where the value in that command is "03 ff." That looks like a big endian value to me.
In my reading of your code, it takes a host byte order value, and swapping it's byte order. It is then passing it to an API that expects a host byte order value. I think it would be much better to avoid doing that. This is my main point.
Let us consider the (likely) case that the host is little endian. The value (and mask) are byte swapped, becoming big endian. Thisbig endian value (and mask) is passed to regmap_update_bits().
Now let us assume that, because REGMAP_ENDIAN_LITTLE is used, they then pass through something like cpu_to_le16().
That's a noop on a little endian system. So the value remains big endian.
Next, let us consider a big endian host.
The value (and mask) are byte swapped, becoming little endian. This little endian value (and mask) is passed to regmap_update_bits().
Then, let us assume that, because REGMAP_ENDIAN_LITTLE is used, they then pass through something like cpu_to_le16(). This is a byte-swap on big endian hosts.
So the value (and mask) become big endian.
The result turns out to be the same for little endian and big endian hosts, which is nice. But now let us assume that instead of passing byte-swapped values to APIs that expect host byte order values, we instead pass host byte order values and use REGMAP_ENDIAN_BIG.
In this case the host byte order values are passed to regmap_update_bits(). Then, as per our earlier assumptions, because REGMAP_ENDIAN_BIG is used, the value (and mask) pass through cpu_to_be16 or similar. After which they are big endian. The same result as above. But without passing byte-swapped values to APIs that expect host byte order values.
Is my understanding of the effect of REGMAP_ENDIAN_LITTLE and REGMAP_ENDIAN_BIG incorrect? Is some other part of my reasoning faulty?
I feel that we are talking past each other.
Let's try to find a common understanding.
RE: [PATCH net-next v4 4/7] net: dsa: microchip: Use different registers for KSZ8463
Quoted 168 lines · On Thu, Jul 24, 2025 at 02:28:56AM +0000, [email protected] wrote:
Quoted 3 lines
Quoted 3 lines · On Wed, Jul 23, 2025 at 02:25:27AM +0000, [email protected] wrote:
Quoted 2 lines
Quoted 2 lines · Simon Horman
Quoted 1 line
On Fri, Jul 18, 2025 at 06:21:03PM -0700, [email protected]
Quoted 108 lines · wrote:
Quoted 96 lines
Quoted 83 lines
Quoted 65 lines
Quoted 5 lines
Quoted 5 lines
From: Tristram Ha <[email protected]>
KSZ8463 does not use same set of registers as KSZ8863 so it is necessary to change some registers when using KSZ8463.
Quoted 8 lines
Quoted 4 lines
Signed-off-by Tristram Ha <[email protected]>---
v3
- Replacecpu_to_be16()withswab16()to avoid compiler warning...
Review
Quoted 1 line
(quoted diff) +0
-0
b/drivers/net/dsa/microchip/ksz_common.cQuoted 6 lines
...
Quoted 3 lines
(quoted diff) +0
-0@@ -2980,10 +2981,15 @@ static int ksz_setup(struct dsa_switch *ds) }Quoted 7 lines
Review static int ksz_setup(struct dsa_switch *ds) · line 2985
Quoted 7 lines
(quoted diff) +2
-4/* set broadcast storm protection 10% rate */ - regmap_update_bits(ksz_regmap_16(dev), regs[S_BROADCAST_CTRL], - BROADCAST_STORM_RATE, - (BROADCAST_STORM_VALUE * - BROADCAST_STORM_PROT_RATE) / 100); + storm_mask = BROADCAST_STORM_RATE; + storm_rate = (BROADCAST_STORM_VALUE *
BROADCAST_STORM_PROT_RATE) / 100;Quoted 19 lines
Review static int ksz_setup(struct dsa_switch *ds) · line 2992
Quoted 7 lines
(quoted diff) +7
-0+ if (ksz_is_ksz8463(dev)) { + storm_mask = swab16(storm_mask); + storm_rate = swab16(storm_rate); + } + regmap_update_bits(ksz_regmap_16(dev), + reg16(dev, regs[S_BROADCAST_CTRL]), + storm_mask, storm_rate);Hi Tristram,
I am confused by the use of
swab16()here.Let us say that we are running on a little endian host (likely). Then the effect of this is to pass big endian values to
regmap_update_bits().But if we are running on a big endian host, the opposite will be true:
little endian values will be passed toregmap_update_bits().Quoted 8 lines
Looking at
KSZ_REGMAP_ENTRY()I see:#define KSZ_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ { \ ... .reg_format_endian = REGMAP_ENDIAN_BIG, \ .val_format_endian = REGMAP_ENDIAN_BIG \ }Update; I now see this in another patch of the series:
(quoted diff) +5
-0+#define KSZ8463_REGMAP_ENTRY(width, swp, regbits, regpad, regalign) \ + { \ + .reg_format_endian = REGMAP_ENDIAN_BIG, \ + .val_format_endian = REGMAP_ENDIAN_LITTLE \ + }Quoted 17 lines
Which I understand to mean that the hardware is expecting little endian values. But still, my concerns raised in my previous email of this thread remain.
And I have a question: does this chip use little endian register values whereas other chips used big endian register values?
Quoted 10 lines
Which based on a skimming the regmap code implies to me that
regmap_update_bits()should be passed host byte order values which regmap will convert to big endian when writing out these values.It is unclear to me why changing the byte order of
storm_maskandstorm_rateis needed here. But it does seem clear that it will lead to inconsistent results on big endian and little endian hosts.Quoted 5 lines
The broadcast storm value
0x7ffis stored in registers 6 and 7 in KSZ8863 where register 6 holds the0x7part while register 7 holds the0xffpart. In KSZ8463 register 6 is defined as 16-bit where the0x7part is held in lower byte and the0xffpart is held in higher byte. It is necessary to swap the bytes when the value is passed to the 16-bit write function.Perhaps naively, I would have expected
.val_format_endian =
REGMAP_ENDIAN_LITTLEto handle writing the 16-bit value
0x7ffsuch that0x7is in the lower byte, while0xffis in the upper byte. Is that not the case?Quoted 11 lines
If not, do you get the desired result by removing the
swab16()calls and using.val_format_endian =
REGMAP_ENDIAN_BIGBut perhaps I misunderstand how .val_format_endian works.
Quoted 4 lines
All other KSZ switches use 8-bit access with automatic address increase so a write to register 0 with value
0x12345678means 0=0x12, 1=0x34, 2=0x56, and 3=0x78.Quoted 9 lines
It is not about big-endian or little-endian. It is just the presentation of this register is different between KSZ8863 and KSZ8463. KSZ8863 uses big-endian for register value as the access is 8-bit and the address is automatically increased by 1. Writing a value
0x03ffto register 6 means 6=0x03 and 7=0xff. The actual SPI transfer commands are "02 06 03 ff." KSZ8463 uses little-endian for register value as the access is fixed at 8-bit, 16-bit, or 32-bit. Writing0x03ffresults in the actual SPI transfer commands "80 70 ff 03" where the correct commands are "80 70 03 ff."The difference between expressing a 16-bit value as "ff 03" and "03 ff" sounds a lot like endianness to me.
"ff 03" is the little endian representation of
0x3ff. "03 ff" is the big endian representation of0x3ff.I am very confused as to why you say "KSZ8463 uses little-endian for register value". And then go on to say that the correct transfer command is "02 06 03 ff", where the value in that command is "03 ff." That looks like a big endian value to me.
In my reading of your code, it takes a host byte order value, and swapping it's byte order. It is then passing it to an API that expects a host byte order value. I think it would be much better to avoid doing that. This is my main point.
Let us consider the (likely) case that the host is little endian. The value (and mask) are byte swapped, becoming big endian. Thisbig endian value (and mask) is passed to
regmap_update_bits().Now let us assume that, because
REGMAP_ENDIAN_LITTLEis used, they then pass through something likecpu_to_le16().
That's a noop on a little endian system. So the value remains big endian.Next, let us consider a big endian host.
The value (and mask) are byte swapped, becoming little endian. This little endian value (and mask) is passed toregmap_update_bits().Then, let us assume that, because
REGMAP_ENDIAN_LITTLEis used, they then pass through something likecpu_to_le16(). This is a byte-swap on big endian hosts.
So the value (and mask) become big endian.The result turns out to be the same for little endian and big endian hosts, which is nice. But now let us assume that instead of passing byte-swapped values to APIs that expect host byte order values, we instead pass host byte order values and use
REGMAP_ENDIAN_BIG.In this case the host byte order values are passed to
regmap_update_bits(). Then, as per our earlier assumptions, becauseREGMAP_ENDIAN_BIGis used, the value (and mask) pass throughcpu_to_be16or similar. After which they are big endian. The same result as above. But without passing byte-swapped values to APIs that expect host byte order values.
Quoted 6 lines
Is my understanding of the effect of
REGMAP_ENDIAN_LITTLEandREGMAP_ENDIAN_BIGincorrect? Is some other part of my reasoning faulty?I feel that we are talking past each other.
Let's try to find a common understanding.
It is really about the register definition of this specific register. In KSZ8863 when presenting in 16-bit the value is 0x07ff, but in KSZ8463 it is 0xff07. It is the fault of the hardware to define such value.
Note that in the new patch KSZ8463 SPI driver implements its own access functions, so native mode is used instead and there is no automatic swapping depending on the big-endian or little-endian format. Still this code is needed to program the register correctly.
Simon Horman
reviewerRe: [PATCH net-next v4 4/7] net: dsa: microchip: Use different registers for KSZ8463
Quoted 5 lines · On Fri, Jul 25, 2025 at 12:17:26AM +0000, [email protected] wrote:
Quoted 5 lines · On Thu, Jul 24, 2025 at 02:28:56AM +0000, [email protected] wrote:
Quoted 3 lines
Quoted 3 lines · On Wed, Jul 23, 2025 at 02:25:27AM +0000, [email protected] wrote:
Quoted 2 lines
Quoted 2 lines · Simon Horman
Quoted 1 line
On Fri, Jul 18, 2025 at 06:21:03PM -0700, [email protected]
wrote:
...
Quoted 6 lines
Quoted 2 lines
I feel that we are talking past each other.
Let's try to find a common understanding.It is really about the register definition of this specific register. In KSZ8863 when presenting in 16-bit the value is
0x07ff, but in KSZ8463 it is0xff07. It is the fault of the hardware to define such value.
If the behaviour of this register is different to others, then I guess a special case is a reasonable approach. Although I would have thought there is a better way.
I would suggest adding a comment to the code explaining that this is a special case. And I would suggest revisiting this if more special cases are needed.
Quoted 4 lines
Note that in the new patch KSZ8463 SPI driver implements its own access functions, so native mode is used instead and there is no automatic swapping depending on the big-endian or little-endian format. Still this code is needed to program the register correctly.
Thanks for taking time to respond to my questions. I think we should let the matter rest here.
PATCH v4 5/7 net: dsa: microchip: Write switch MAC address differently for KSZ8463
In reply to [email protected]From: Tristram Ha <[email protected]>
KSZ8463 uses 16-bit register definitions so it writes differently for 8-bit switch MAC address.
--- drivers/net/dsa/microchip/ksz_common.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-)
drivers/net/dsa/microchip/ksz_common.c +10 -1
@@ -4823,7 +4823,16 @@ int ksz_switch_macaddr_get(struct dsa_switch *ds, int port,/* Program the switch MAC address to hardware */ for (i =0 ; i <ETH_ALEN ; i++) { - ret =ksz_write8 (dev, regs[REG_SW_MAC_ADDR ] + i, addr[i]); +if (ksz_is_ksz8463 (dev)) { +u16 addr16 = ((u16 )addr[i] <<8 ) | addr[i +1 ]; + + ret =ksz_write16 (dev, regs[REG_SW_MAC_ADDR ] + i, + addr16); + i++; + }else { + ret =ksz_write8 (dev, regs[REG_SW_MAC_ADDR ] + i, + addr[i]); + }if (ret) goto macaddr_drop; }
Signature · 1 line
2.34.1
PATCH v4 6/7 net: dsa: microchip: Setup fiber ports for KSZ8463
In reply to [email protected]From: Tristram Ha <[email protected]>
The fiber ports in KSZ8463 cannot be detected internally, so it requires specifying that condition in the device tree. Like the one used in Micrel PHY the port link can only be read and there is no write to the PHY. The driver programs registers to operate fiber ports correctly.
---
v3
- Disable PTP function in a separate patch
drivers/net/dsa/microchip/ksz8.c | 16 ++++++++++++++++ drivers/net/dsa/microchip/ksz_common.c | 3 +++ 2 files changed, 19 insertions(+)
drivers/net/dsa/microchip/ksz8.c +16 -0
@@ -1715,6 +1715,7 @@ void ksz8_config_cpu_port(struct dsa_switch *ds)const u32 *masks;const u16 *regs;u8 remote; +u8 fiber_ports =0 ;int i; masks = dev->info->masks;
@@ -1745,6 +1746,21 @@ void ksz8_config_cpu_port(struct dsa_switch *ds)else ksz_port_cfg (dev, i, regs[P_STP_CTRL ],PORT_FORCE_FLOW_CTRL , false); +if (p->fiber) + fiber_ports |= (1 << i); + } +if (ksz_is_ksz8463 (dev)) { +/* Setup fiber ports. */ +if (fiber_ports) { + fiber_ports &=3 ; +regmap_update_bits (ksz_regmap_16 (dev), +reg16 (dev,KSZ8463_REG_CFG_CTRL ), + fiber_ports <<PORT_COPPER_MODE_S , +0 ); +regmap_update_bits (ksz_regmap_16 (dev), +reg16 (dev,KSZ8463_REG_DSP_CTRL_6 ), +COPPER_RECEIVE_ADJUSTMENT ,0 ); + } } }
drivers/net/dsa/microchip/ksz_common.c +3 -0
@@ -5441,6 +5441,9 @@ int ksz_switch_register(struct ksz_device *dev) &dev->ports[port_num].interface);ksz_parse_rgmii_delay (dev, port_num, port); + dev->ports[port_num].fiber = +of_property_read_bool (port, +"micrel,fiber-mode" ); }of_node_put (ports); }
Signature · 1 line
2.34.1
PATCH v4 7/7 net: dsa: microchip: Disable PTP function of KSZ8463
In reply to [email protected]From: Tristram Ha <[email protected]>
The PTP function of KSZ8463 is on by default. However, its proprietary way of storing timestamp directly in a reserved field inside the PTP message header is not suitable for use with the current Linux PTP stack implementation. It is necessary to disable the PTP function to not interfere the normal operation of the MAC.
Note the PTP driver for KSZ switches does not work for KSZ8463 and is not activated for it.
--- drivers/net/dsa/microchip/ksz8.c | 11 +++++++++++ 1 file changed, 11 insertions(+)
drivers/net/dsa/microchip/ksz8.c +11 -0
@@ -1761,6 +1761,17 @@ void ksz8_config_cpu_port(struct dsa_switch *ds)reg16 (dev,KSZ8463_REG_DSP_CTRL_6 ), COPPER_RECEIVE_ADJUSTMENT, 0); } + +/* Turn off PTP function as the switch's proprietary way of +* handling timestamp is not supported in current Linux PTP +* stack implementation. +*/ +regmap_update_bits (ksz_regmap_16 (dev), +reg16 (dev,KSZ8463_PTP_MSG_CONF1 ), +PTP_ENABLE ,0 ); +regmap_update_bits (ksz_regmap_16 (dev), +reg16 (dev,KSZ8463_PTP_CLK_CTRL ), +PTP_CLK_ENABLE ,0 ); } }
Signature · 1 line
2.34.1