[U-Boot] [PATCH 00/11] cpsw: enable DM_ETH on dra74 and am437x evms
This series adds the following * Enable DM_ETH on dra74, am437x gp and am437x sk evms. * Add support to verify of_is_device_conpatible() based on linux implementation * Fix an issue in fdtdec get addr for address and size cell length
Pushed a branch for others for testing [1] and logs at [2]
[1] - git://git.ti.com/~mugunthanvnm/ti-u-boot/mugunth-ti-u-boot.git cpsw [2] - http://pastebin.ubuntu.com/15671145/
Mugunthan V N (11): drivers: core: device: add support to check dt compatible for a device/machine lib: fdtdec: fix size cell and address cell parse from DT ti_omap5_common: eth: do not define DM_ETH for spl drivers: net: cpsw: fix cpsw dp parse when num slaves as 1 ARM: omap5: add platform specific ethernet phy modes configurations drivers: net: cpsw: add support for reading mac address from efuse arm: dts: am4372: add syscon node to cpsw to read mac address arm: dts: dra7: add syscon node to cpsw to read mac address defconfig: am437x_gp_evm: enable eth driver model defconfig: am437x_sk_evm: enable eth driver model defconfig: dra74_evm: enable eth driver model
arch/arm/dts/am4372.dtsi | 1 + arch/arm/dts/dra7.dtsi | 1 + arch/arm/include/asm/arch-omap5/cpu.h | 12 ++++ configs/am437x_gp_evm_defconfig | 1 + configs/am437x_sk_evm_defconfig | 1 + configs/dra74_evm_defconfig | 1 + drivers/core/device.c | 27 ++++++++ drivers/net/Makefile | 2 +- drivers/net/cpsw-common.c | 121 ++++++++++++++++++++++++++++++++++ drivers/net/cpsw.c | 30 +++------ include/configs/ti_omap5_common.h | 1 + include/cpsw.h | 1 + include/dm/device.h | 23 +++++++ lib/fdtdec.c | 9 ++- 14 files changed, 208 insertions(+), 23 deletions(-) create mode 100644 drivers/net/cpsw-common.c
Provide an api to check whether the given device or machine is compatible with the given compat string which helps in making decisions in drivers based on device or machine compatible.
Idea taken from Linux.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com --- drivers/core/device.c | 27 +++++++++++++++++++++++++++ include/dm/device.h | 23 +++++++++++++++++++++++ 2 files changed, 50 insertions(+)
diff --git a/drivers/core/device.c b/drivers/core/device.c index cb24a61..445b22e 100644 --- a/drivers/core/device.c +++ b/drivers/core/device.c @@ -691,3 +691,30 @@ int device_set_name(struct udevice *dev, const char *name)
return 0; } + +bool of_device_is_compatible(int offset, const char *compat) +{ + const void *fdt = gd->fdt_blob; + const char *str; + int str_len; + int len = 0; + + str = fdt_getprop(fdt, 0, "compatible", &str_len); + if(!str) { + debug("compatible field not found in node(%d)\n", offset); + return false; + } + + while (len < str_len) { + if (!strcmp(compat, &str[len])) + return true; + len += strlen(&str[len]) + 1; + } + + return false; +} + +bool of_machine_is_compatible(const char *compat) +{ + return of_device_is_compatible(0, compat); +} diff --git a/include/dm/device.h b/include/dm/device.h index 1cf8150..9a1b3d0 100644 --- a/include/dm/device.h +++ b/include/dm/device.h @@ -510,6 +510,29 @@ bool device_is_last_sibling(struct udevice *dev); int device_set_name(struct udevice *dev, const char *name);
/** + * of_device_is_compatible() - check if the device is compatible with the compat + * + * This allows to check whether the device is comaptible with the compat. + * + * @offset: Offset to the device + * @compat: Compatible string which needs to verified in the given + * device offset + * @return true if OK, false if the compatible is not found + */ +bool of_device_is_compatible(int offset, const char *compat); + +/** + * of_machine_is_compatible() - check if the machine is compatible with + * the compat + * + * This allows to check whether the machine is comaptible with the compat. + * + * @compat: Compatible string which needs to verified + * @return true if OK, false if the compatible is not found + */ +bool of_machine_is_compatible(const char *compat); + +/** * device_is_on_pci_bus - Test if a device is on a PCI bus * * @dev: device to test
Hi Mugunthan,
On 7 April 2016 at 09:17, Mugunthan V N mugunthanvnm@ti.com wrote:
Provide an api to check whether the given device or machine is compatible with the given compat string which helps in making decisions in drivers based on device or machine compatible.
Idea taken from Linux.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
drivers/core/device.c | 27 +++++++++++++++++++++++++++ include/dm/device.h | 23 +++++++++++++++++++++++ 2 files changed, 50 insertions(+)
diff --git a/drivers/core/device.c b/drivers/core/device.c index cb24a61..445b22e 100644 --- a/drivers/core/device.c +++ b/drivers/core/device.c @@ -691,3 +691,30 @@ int device_set_name(struct udevice *dev, const char *name)
return 0;}
+bool of_device_is_compatible(int offset, const char *compat)
struct udevice *dev should be the first parameter. We may one day stop using offsets and use a pointer, so the less we have to change the better.
+{
const void *fdt = gd->fdt_blob;const char *str;int str_len;int len = 0;str = fdt_getprop(fdt, 0, "compatible", &str_len);if(!str) {debug("compatible field not found in node(%d)\n", offset);return false;}while (len < str_len) {
Can you use fdt_node_check_compatible() here?
if (!strcmp(compat, &str[len]))return true;len += strlen(&str[len]) + 1;}return false;+}
+bool of_machine_is_compatible(const char *compat) +{
return of_device_is_compatible(0, compat);+} diff --git a/include/dm/device.h b/include/dm/device.h index 1cf8150..9a1b3d0 100644 --- a/include/dm/device.h +++ b/include/dm/device.h @@ -510,6 +510,29 @@ bool device_is_last_sibling(struct udevice *dev); int device_set_name(struct udevice *dev, const char *name);
/**
- of_device_is_compatible() - check if the device is compatible with the compat
- This allows to check whether the device is comaptible with the compat.
- @offset: Offset to the device
- @compat: Compatible string which needs to verified in the given
device offset
- @return true if OK, false if the compatible is not found
- */
+bool of_device_is_compatible(int offset, const char *compat);
+/**
- of_machine_is_compatible() - check if the machine is compatible with
the compat
- This allows to check whether the machine is comaptible with the compat.
- @compat: Compatible string which needs to verified
- @return true if OK, false if the compatible is not found
- */
+bool of_machine_is_compatible(const char *compat);
+/**
- device_is_on_pci_bus - Test if a device is on a PCI bus
- @dev: device to test
-- 2.8.1.101.g72d917a
Regards, Simon
Size cell and address cell should be read from the parent node and should not assume with data structures as an example TI DRA7xx SoC is enabled as 64bit as there is LPAE support but the addresses specified in DT are all 32 bit sizes. So changing the code to read from parent node instead of calculations.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com --- lib/fdtdec.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/lib/fdtdec.c b/lib/fdtdec.c index 70acc29..8a5fb8c 100644 --- a/lib/fdtdec.c +++ b/lib/fdtdec.c @@ -88,15 +88,20 @@ fdt_addr_t fdtdec_get_addr_size_fixed(const void *blob, int node, const fdt32_t *prop_addr, *prop_size, *prop_after_size; int len; fdt_addr_t addr; + int parent;
debug("%s: %s: ", __func__, prop_name);
- if (na > (sizeof(fdt_addr_t) / sizeof(fdt32_t))) { + parent = fdt_parent_offset(blob, node); + + na = fdt_address_cells(blob, parent); + if (na < 1) { debug("(na too large for fdt_addr_t type)\n"); return FDT_ADDR_T_NONE; }
- if (ns > (sizeof(fdt_size_t) / sizeof(fdt32_t))) { + ns = fdt_size_cells(blob, parent); + if (ns < 0) { debug("(ns too large for fdt_size_t type)\n"); return FDT_ADDR_T_NONE; }
+Stephen
Hi Mugunthan,
On 7 April 2016 at 09:17, Mugunthan V N mugunthanvnm@ti.com wrote:
Size cell and address cell should be read from the parent node and should not assume with data structures as an example TI DRA7xx SoC is enabled as 64bit as there is LPAE support but the addresses specified in DT are all 32 bit sizes. So changing the code to read from parent node instead of calculations.
I don't understand this. Can you please reword it and shorten the sentences?
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
lib/fdtdec.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/lib/fdtdec.c b/lib/fdtdec.c index 70acc29..8a5fb8c 100644 --- a/lib/fdtdec.c +++ b/lib/fdtdec.c @@ -88,15 +88,20 @@ fdt_addr_t fdtdec_get_addr_size_fixed(const void *blob, int node, const fdt32_t *prop_addr, *prop_size, *prop_after_size; int len; fdt_addr_t addr;
int parent; debug("%s: %s: ", __func__, prop_name);
if (na > (sizeof(fdt_addr_t) / sizeof(fdt32_t))) {
parent = fdt_parent_offset(blob, node);
This is a very slow function. I hope this changes is not needed.
na = fdt_address_cells(blob, parent);if (na < 1) { debug("(na too large for fdt_addr_t type)\n"); return FDT_ADDR_T_NONE; }
if (ns > (sizeof(fdt_size_t) / sizeof(fdt32_t))) {
ns = fdt_size_cells(blob, parent);if (ns < 0) { debug("(ns too large for fdt_size_t type)\n"); return FDT_ADDR_T_NONE; }-- 2.8.1.101.g72d917a
Regards, Simon
On 04/09/2016 12:35 PM, Simon Glass wrote:
+Stephen
Hi Mugunthan,
On 7 April 2016 at 09:17, Mugunthan V N mugunthanvnm@ti.com wrote:
Size cell and address cell should be read from the parent node and should not assume with data structures as an example TI DRA7xx SoC is enabled as 64bit as there is LPAE support but the addresses specified in DT are all 32 bit sizes. So changing the code to read from parent node instead of calculations.
I don't understand this. Can you please reword it and shorten the sentences?
diff --git a/lib/fdtdec.c b/lib/fdtdec.c index 70acc29..8a5fb8c 100644 --- a/lib/fdtdec.c +++ b/lib/fdtdec.c @@ -88,15 +88,20 @@ fdt_addr_t fdtdec_get_addr_size_fixed(const void *blob, int node, const fdt32_t *prop_addr, *prop_size, *prop_after_size; int len; fdt_addr_t addr;
int parent; debug("%s: %s: ", __func__, prop_name);
if (na > (sizeof(fdt_addr_t) / sizeof(fdt32_t))) {
parent = fdt_parent_offset(blob, node);This is a very slow function. I hope this changes is not needed.
na = fdt_address_cells(blob, parent);if (na < 1) { debug("(na too large for fdt_addr_t type)\n"); return FDT_ADDR_T_NONE; }
if (ns > (sizeof(fdt_size_t) / sizeof(fdt32_t))) {
ns = fdt_size_cells(blob, parent);if (ns < 0) { debug("(ns too large for fdt_size_t type)\n"); return FDT_ADDR_T_NONE; }
The entire point of fdtdec_get_addr_size_fixed() is for use-cases where na and ns are known and fixed ahead of time, or have already been retrieved from the parent node by the caller. This patch is incorrect. I expect the correct fix is to call e.g. fdtdec_get_addr_size_auto_parent() or fdtdec_get_addr_size_auto_noparent() from somewhere, rather than calling fdtdec_get_addr_size_fixed().
Stephen
On Sunday 10 April 2016 09:11 AM, Stephen Warren wrote:
On 04/09/2016 12:35 PM, Simon Glass wrote:
+Stephen
Hi Mugunthan,
On 7 April 2016 at 09:17, Mugunthan V N mugunthanvnm@ti.com wrote:
Size cell and address cell should be read from the parent node and should not assume with data structures as an example TI DRA7xx SoC is enabled as 64bit as there is LPAE support but the addresses specified in DT are all 32 bit sizes. So changing the code to read from parent node instead of calculations.
I don't understand this. Can you please reword it and shorten the sentences?
diff --git a/lib/fdtdec.c b/lib/fdtdec.c index 70acc29..8a5fb8c 100644 --- a/lib/fdtdec.c +++ b/lib/fdtdec.c @@ -88,15 +88,20 @@ fdt_addr_t fdtdec_get_addr_size_fixed(const void *blob, int node, const fdt32_t *prop_addr, *prop_size, *prop_after_size; int len; fdt_addr_t addr;
int parent; debug("%s: %s: ", __func__, prop_name);
if (na > (sizeof(fdt_addr_t) / sizeof(fdt32_t))) {
parent = fdt_parent_offset(blob, node);This is a very slow function. I hope this changes is not needed.
na = fdt_address_cells(blob, parent);if (na < 1) { debug("(na too large for fdt_addr_t type)\n"); return FDT_ADDR_T_NONE; }
if (ns > (sizeof(fdt_size_t) / sizeof(fdt32_t))) {
ns = fdt_size_cells(blob, parent);if (ns < 0) { debug("(ns too large for fdt_size_t type)\n"); return FDT_ADDR_T_NONE; }The entire point of fdtdec_get_addr_size_fixed() is for use-cases where na and ns are known and fixed ahead of time, or have already been retrieved from the parent node by the caller. This patch is incorrect. I expect the correct fix is to call e.g. fdtdec_get_addr_size_auto_parent() or fdtdec_get_addr_size_auto_noparent() from somewhere, rather than calling fdtdec_get_addr_size_fixed().
Will fix thos in v2 by using fdtdec_get_addr_size_auto_parent()
Regards Mugunthan V N
Since omap's spl doesn't support DM currently, do not define DM_ETH for spl build.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com --- include/configs/ti_omap5_common.h | 1 + 1 file changed, 1 insertion(+)
diff --git a/include/configs/ti_omap5_common.h b/include/configs/ti_omap5_common.h index 202b18c..c592646 100644 --- a/include/configs/ti_omap5_common.h +++ b/include/configs/ti_omap5_common.h @@ -155,6 +155,7 @@ #ifdef CONFIG_SPL_BUILD #undef CONFIG_DM_MMC #undef CONFIG_TIMER +#undef CONFIG_DM_ETH #endif
#endif /* __CONFIG_TI_OMAP5_COMMON_H */
Hi Mugunthan,
On 7 April 2016 at 09:17, Mugunthan V N mugunthanvnm@ti.com wrote:
Since omap's spl doesn't support DM currently, do not define DM_ETH for spl build.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
include/configs/ti_omap5_common.h | 1 + 1 file changed, 1 insertion(+)
Reviewed-by: Simon Glass sjg@chromium.org
I suppose this is OK as a temporary convenience, although we should use defconfig for this sort of thing.
diff --git a/include/configs/ti_omap5_common.h b/include/configs/ti_omap5_common.h index 202b18c..c592646 100644 --- a/include/configs/ti_omap5_common.h +++ b/include/configs/ti_omap5_common.h @@ -155,6 +155,7 @@ #ifdef CONFIG_SPL_BUILD #undef CONFIG_DM_MMC #undef CONFIG_TIMER +#undef CONFIG_DM_ETH #endif
#endif /* __CONFIG_TI_OMAP5_COMMON_H */
2.8.1.101.g72d917a
On Sunday 10 April 2016 12:05 AM, Simon Glass wrote:
Hi Mugunthan,
On 7 April 2016 at 09:17, Mugunthan V N mugunthanvnm@ti.com wrote:
Since omap's spl doesn't support DM currently, do not define DM_ETH for spl build.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
include/configs/ti_omap5_common.h | 1 + 1 file changed, 1 insertion(+)
Reviewed-by: Simon Glass sjg@chromium.org
I suppose this is OK as a temporary convenience, although we should use defconfig for this sort of thing.
Yes, this is a temporary. When omap's spl updated to support DM, these undefs has to be removed.
Regards Mugunthan V N
On Thu, Apr 07, 2016 at 08:47:02PM +0530, Mugunthan V N wrote:
Since omap's spl doesn't support DM currently, do not define DM_ETH for spl build.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
Reviewed-by: Tom Rini trini@konsulko.com
On some boards number of slaves can be 1 when only one port ethernet is pinned out. So do not break when slave_index and num slaves check fails, instead continue to parse the next child.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com --- drivers/net/cpsw.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/net/cpsw.c b/drivers/net/cpsw.c index 7104754..971ebf0 100644 --- a/drivers/net/cpsw.c +++ b/drivers/net/cpsw.c @@ -1209,10 +1209,8 @@ static int cpsw_eth_ofdata_to_platdata(struct udevice *dev) if (!strncmp(name, "slave", 5)) { u32 phy_id[2];
- if (slave_index >= priv->data.slaves) { - printf("error: num slaves and slave nodes did not match\n"); - return -EINVAL; - } + if (slave_index >= priv->data.slaves) + continue; phy_mode = fdt_getprop(fdt, subnode, "phy-mode", NULL); if (phy_mode) priv->data.slave_data[slave_index].phy_if =
On Thu, Apr 07, 2016 at 08:47:03PM +0530, Mugunthan V N wrote:
On some boards number of slaves can be 1 when only one port ethernet is pinned out. So do not break when slave_index and num slaves check fails, instead continue to parse the next child.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
Reviewed-by: Tom Rini trini@konsulko.com
Add platforms specific phy mode configuration bits to be used to configure phy mode in control module.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com --- arch/arm/include/asm/arch-omap5/cpu.h | 12 ++++++++++++ 1 file changed, 12 insertions(+)
diff --git a/arch/arm/include/asm/arch-omap5/cpu.h b/arch/arm/include/asm/arch-omap5/cpu.h index b1513e9..683d905 100644 --- a/arch/arm/include/asm/arch-omap5/cpu.h +++ b/arch/arm/include/asm/arch-omap5/cpu.h @@ -116,4 +116,16 @@ struct watchdog { #define CPSW_BASE 0x48484000 #define CPSW_MDIO_BASE 0x48485000
+/* gmii_sel register defines */ +#define GMII1_SEL_MII 0x0 +#define GMII1_SEL_RMII 0x1 +#define GMII1_SEL_RGMII 0x2 +#define GMII2_SEL_MII (GMII1_SEL_MII << 4) +#define GMII2_SEL_RMII (GMII1_SEL_RMII << 4) +#define GMII2_SEL_RGMII (GMII1_SEL_RGMII << 4) + +#define MII_MODE_ENABLE (GMII1_SEL_MII | GMII2_SEL_MII) +#define RMII_MODE_ENABLE (GMII1_SEL_RMII | GMII2_SEL_RMII) +#define RGMII_MODE_ENABLE (GMII1_SEL_RGMII | GMII2_SEL_RGMII) + #endif /* _CPU_H */
On Thu, Apr 07, 2016 at 08:47:04PM +0530, Mugunthan V N wrote:
Add platforms specific phy mode configuration bits to be used to configure phy mode in control module.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
Reviewed-by: Tom Rini trini@konsulko.com
Different TI platforms has to read with different combination to get the mac address from efuse. So add support to read mac address based on machine/device compatibles.
The code is taken from Linux drivers/net/ethernet/ti/cpsw-common.c done by Tony Lindgren.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com --- drivers/net/Makefile | 2 +- drivers/net/cpsw-common.c | 121 ++++++++++++++++++++++++++++++++++++++++++++++ drivers/net/cpsw.c | 24 +++------ include/cpsw.h | 1 + 4 files changed, 131 insertions(+), 17 deletions(-) create mode 100644 drivers/net/cpsw-common.c
diff --git a/drivers/net/Makefile b/drivers/net/Makefile index fbedd04..d5e4a97 100644 --- a/drivers/net/Makefile +++ b/drivers/net/Makefile @@ -59,7 +59,7 @@ obj-$(CONFIG_SMC91111) += smc91111.o obj-$(CONFIG_SMC911X) += smc911x.o obj-$(CONFIG_DRIVER_TI_EMAC) += davinci_emac.o obj-$(CONFIG_TSEC_ENET) += tsec.o fsl_mdio.o -obj-$(CONFIG_DRIVER_TI_CPSW) += cpsw.o +obj-$(CONFIG_DRIVER_TI_CPSW) += cpsw.o cpsw-common.o obj-$(CONFIG_FMAN_ENET) += fsl_mdio.o obj-$(CONFIG_TSI108_ETH) += tsi108_eth.o obj-$(CONFIG_ULI526X) += uli526x.o diff --git a/drivers/net/cpsw-common.c b/drivers/net/cpsw-common.c new file mode 100644 index 0000000..086ffc9 --- /dev/null +++ b/drivers/net/cpsw-common.c @@ -0,0 +1,121 @@ +/* + * CPSW common - libs used across TI ethernet devices. + * + * Copyright (C) 2016, Texas Instruments, Incorporated + * + * SPDX-License-Identifier: GPL-2.0+ + */ + +#include <common.h> +#include <dm.h> +#include <fdt_support.h> +#include <asm/io.h> +#include <cpsw.h> + +DECLARE_GLOBAL_DATA_PTR; + +#define CTRL_MAC_REG(offset, id) ((offset) + 0x8 * (id)) + +static int davinci_emac_3517_get_macid(struct udevice *dev, u16 offset, + int slave, u8 *mac_addr) +{ + void *fdt = (void *)gd->fdt_blob; + int node = dev->of_offset; + u32 macid_lsb; + u32 macid_msb; + fdt32_t gmii = 0; + int syscon; + u32 addr; + + syscon = fdtdec_lookup_phandle(fdt, node, "syscon"); + if (syscon < 0) { + error("Syscon offset not found\n"); + return -ENOENT; + } + + addr = (u32)map_physmem(fdt_translate_address(fdt, syscon, &gmii), + sizeof(u32), MAP_NOCACHE); + if (addr == FDT_ADDR_T_NONE) { + error("Not able to get syscon address to get mac efuse address\n"); + return -ENOENT; + } + + addr += CTRL_MAC_REG(offset, slave); + + /* try reading mac address from efuse */ + macid_lsb = readl(addr); + macid_msb = readl(addr + 4); + + mac_addr[0] = (macid_msb >> 16) & 0xff; + mac_addr[1] = (macid_msb >> 8) & 0xff; + mac_addr[2] = macid_msb & 0xff; + mac_addr[3] = (macid_lsb >> 16) & 0xff; + mac_addr[4] = (macid_lsb >> 8) & 0xff; + mac_addr[5] = macid_lsb & 0xff; + + return 0; +} + +static int cpsw_am33xx_cm_get_macid(struct udevice *dev, u16 offset, int slave, + u8 *mac_addr) +{ + void *fdt = (void *)gd->fdt_blob; + int node = dev->of_offset; + u32 macid_lo; + u32 macid_hi; + fdt32_t gmii = 0; + int syscon; + u32 addr; + + syscon = fdtdec_lookup_phandle(fdt, node, "syscon"); + if (syscon < 0) { + error("Syscon offset not found\n"); + return -ENOENT; + } + + addr = (u32)map_physmem(fdt_translate_address(fdt, syscon, &gmii), + sizeof(u32), MAP_NOCACHE); + if (addr == FDT_ADDR_T_NONE) { + error("Not able to get syscon address to get mac efuse address\n"); + return -ENOENT; + } + + addr += CTRL_MAC_REG(offset, slave); + + /* try reading mac address from efuse */ + macid_lo = readl(addr); + macid_hi = readl(addr + 4); + + mac_addr[5] = (macid_lo >> 8) & 0xff; + mac_addr[4] = macid_lo & 0xff; + mac_addr[3] = (macid_hi >> 24) & 0xff; + mac_addr[2] = (macid_hi >> 16) & 0xff; + mac_addr[1] = (macid_hi >> 8) & 0xff; + mac_addr[0] = macid_hi & 0xff; + + return 0; +} + +int ti_cm_get_macid(struct udevice *dev, int slave, u8 *mac_addr) +{ + if (of_machine_is_compatible("ti,dm8148")) + return cpsw_am33xx_cm_get_macid(dev, 0x630, slave, mac_addr); + + if (of_machine_is_compatible("ti,am33xx")) + return cpsw_am33xx_cm_get_macid(dev, 0x630, slave, mac_addr); + + if (of_device_is_compatible(dev->of_offset, "ti,am3517-emac")) + return davinci_emac_3517_get_macid(dev, 0x110, slave, mac_addr); + + if (of_device_is_compatible(dev->of_offset, "ti,dm816-emac")) + return cpsw_am33xx_cm_get_macid(dev, 0x30, slave, mac_addr); + + if (of_machine_is_compatible("ti,am4372")) + return cpsw_am33xx_cm_get_macid(dev, 0x630, slave, mac_addr); + + if (of_machine_is_compatible("ti,dra7")) + return davinci_emac_3517_get_macid(dev, 0x514, slave, mac_addr); + + dev_err(dev, "incompatible machine/device type for reading mac address\n"); + return -ENOENT; +} diff --git a/drivers/net/cpsw.c b/drivers/net/cpsw.c index 971ebf0..cc63833 100644 --- a/drivers/net/cpsw.c +++ b/drivers/net/cpsw.c @@ -26,6 +26,7 @@ #include <phy.h> #include <asm/arch/cpu.h> #include <dm.h> +#include <fdt_support.h>
DECLARE_GLOBAL_DATA_PTR;
@@ -1146,9 +1147,8 @@ static int cpsw_eth_ofdata_to_platdata(struct udevice *dev) int node = dev->of_offset; int subnode; int slave_index = 0; - uint32_t mac_hi, mac_lo; - fdt32_t gmii = 0; int active_slave; + int ret;
pdata->iobase = dev_get_addr(dev); priv->data.version = CPSW_CTRL_VERSION_2; @@ -1234,20 +1234,11 @@ static int cpsw_eth_ofdata_to_platdata(struct udevice *dev) priv->data.slave_data[1].sliver_reg_ofs = CPSW_SLIVER1_OFFSET; }
- subnode = fdtdec_lookup_phandle(fdt, node, "syscon"); - priv->data.mac_id = fdt_translate_address((void *)fdt, subnode, &gmii); - priv->data.mac_id += AM335X_GMII_SEL_OFFSET; - priv->data.mac_id += active_slave * 8; - - /* try reading mac address from efuse */ - mac_lo = readl(priv->data.mac_id); - mac_hi = readl(priv->data.mac_id + 4); - pdata->enetaddr[0] = mac_hi & 0xFF; - pdata->enetaddr[1] = (mac_hi & 0xFF00) >> 8; - pdata->enetaddr[2] = (mac_hi & 0xFF0000) >> 16; - pdata->enetaddr[3] = (mac_hi & 0xFF000000) >> 24; - pdata->enetaddr[4] = mac_lo & 0xFF; - pdata->enetaddr[5] = (mac_lo & 0xFF00) >> 8; + ret = ti_cm_get_macid(dev, active_slave, pdata->enetaddr); + if (ret < 0) { + error("cpsw read efuse mac failed\n"); + return ret; + }
pdata->phy_interface = priv->data.slave_data[active_slave].phy_if; if (pdata->phy_interface == -1) { @@ -1268,6 +1259,7 @@ static int cpsw_eth_ofdata_to_platdata(struct udevice *dev) writel(RGMII_MODE_ENABLE, priv->data.gmii_sel); break; } + return 0; }
diff --git a/include/cpsw.h b/include/cpsw.h index cf1d30b..6255cd8 100644 --- a/include/cpsw.h +++ b/include/cpsw.h @@ -51,5 +51,6 @@ struct cpsw_platform_data { };
int cpsw_register(struct cpsw_platform_data *data); +int ti_cm_get_macid(struct udevice *dev, int slave, u8 *mac_addr);
#endif /* _CPSW_H_ */
On Thu, Apr 07, 2016 at 08:47:05PM +0530, Mugunthan V N wrote:
Different TI platforms has to read with different combination to get the mac address from efuse. So add support to read mac address based on machine/device compatibles.
The code is taken from Linux drivers/net/ethernet/ti/cpsw-common.c done by Tony Lindgren.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
... but we don't do anything with the half dozen copies of that code that we already have in the tree and it needs to also take care of setting ethaddr.
On Monday 11 April 2016 08:31 PM, Tom Rini wrote:
On Thu, Apr 07, 2016 at 08:47:05PM +0530, Mugunthan V N wrote:
Different TI platforms has to read with different combination to get the mac address from efuse. So add support to read mac address based on machine/device compatibles.
The code is taken from Linux drivers/net/ethernet/ti/cpsw-common.c done by Tony Lindgren.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
... but we don't do anything with the half dozen copies of that code that we already have in the tree and it needs to also take care of setting ethaddr.
The file is just a copy of the file from Kernel and modified header file inclusions and defines as per u-boot requirement. Kept the functions intact so that later patches from Kernel can be pulled to u-boot or vice-versa when new SoC is brought up.
Driver has to update the pdata->enetaddr, it will be set to env in eth uclass driver post probe function.
Regards Mugunthan V N
On Tue, Apr 12, 2016 at 09:54:24AM +0530, Mugunthan V N wrote:
On Monday 11 April 2016 08:31 PM, Tom Rini wrote:
On Thu, Apr 07, 2016 at 08:47:05PM +0530, Mugunthan V N wrote:
Different TI platforms has to read with different combination to get the mac address from efuse. So add support to read mac address based on machine/device compatibles.
The code is taken from Linux drivers/net/ethernet/ti/cpsw-common.c done by Tony Lindgren.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
... but we don't do anything with the half dozen copies of that code that we already have in the tree and it needs to also take care of setting ethaddr.
The file is just a copy of the file from Kernel and modified header file inclusions and defines as per u-boot requirement. Kept the functions intact so that later patches from Kernel can be pulled to u-boot or vice-versa when new SoC is brought up.
Driver has to update the pdata->enetaddr, it will be set to env in eth uclass driver post probe function.
Ah, OK. So we'll be able to remove the board/ versions and such after we move it all to DM, I can accept that, thanks.
Add syscon node to cpsw device node to read mac address from efuse.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com --- arch/arm/dts/am4372.dtsi | 1 + 1 file changed, 1 insertion(+)
diff --git a/arch/arm/dts/am4372.dtsi b/arch/arm/dts/am4372.dtsi index c95d1d3..3ffa8e0 100644 --- a/arch/arm/dts/am4372.dtsi +++ b/arch/arm/dts/am4372.dtsi @@ -547,6 +547,7 @@ active_slave = <0>; cpts_clock_mult = <0x80000000>; cpts_clock_shift = <29>; + syscon = <&scm_conf>; ranges;
davinci_mdio: mdio@4a101000 {
On Thu, Apr 07, 2016 at 08:47:06PM +0530, Mugunthan V N wrote:
Add syscon node to cpsw device node to read mac address from efuse.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
Reviewed-by: Tom Rini trini@konsulko.com
Add syscon node to cpsw device node to read mac address from efuse.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com --- arch/arm/dts/dra7.dtsi | 1 + 1 file changed, 1 insertion(+)
diff --git a/arch/arm/dts/dra7.dtsi b/arch/arm/dts/dra7.dtsi index e7fecf7..3059273 100644 --- a/arch/arm/dts/dra7.dtsi +++ b/arch/arm/dts/dra7.dtsi @@ -1426,6 +1426,7 @@ active_slave = <0>; cpts_clock_mult = <0x80000000>; cpts_clock_shift = <29>; + syscon = <&scm_conf>; reg = <0x48484000 0x1000 0x48485200 0x2E00>; #address-cells = <1>;
On Thu, Apr 07, 2016 at 08:47:07PM +0530, Mugunthan V N wrote:
Add syscon node to cpsw device node to read mac address from efuse.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
Reviewed-by: Tom Rini trini@konsulko.com
Enable eth driver model for am437x_gp_evm as cpsw supports driver model.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com --- configs/am437x_gp_evm_defconfig | 1 + 1 file changed, 1 insertion(+)
diff --git a/configs/am437x_gp_evm_defconfig b/configs/am437x_gp_evm_defconfig index 356f6fd..760e1f3 100644 --- a/configs/am437x_gp_evm_defconfig +++ b/configs/am437x_gp_evm_defconfig @@ -22,3 +22,4 @@ CONFIG_TIMER=y CONFIG_OMAP_TIMER=y CONFIG_USB=y CONFIG_USB_GADGET=y +CONFIG_DM_ETH=y
On Thu, Apr 07, 2016 at 08:47:08PM +0530, Mugunthan V N wrote:
Enable eth driver model for am437x_gp_evm as cpsw supports driver model.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
Reviewed-by: Tom Rini trini@konsulko.com
Enable eth driver model for am437x_sk_evm as cpsw supports driver model.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com --- configs/am437x_sk_evm_defconfig | 1 + 1 file changed, 1 insertion(+)
diff --git a/configs/am437x_sk_evm_defconfig b/configs/am437x_sk_evm_defconfig index 2e2827f..324e8ed 100644 --- a/configs/am437x_sk_evm_defconfig +++ b/configs/am437x_sk_evm_defconfig @@ -26,3 +26,4 @@ CONFIG_TIMER=y CONFIG_OMAP_TIMER=y CONFIG_USB=y CONFIG_USB_GADGET=y +CONFIG_DM_ETH=y
On Thu, Apr 07, 2016 at 08:47:09PM +0530, Mugunthan V N wrote:
Enable eth driver model for am437x_sk_evm as cpsw supports driver model.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
Reviewed-by: Tom Rini trini@konsulko.com
Enable eth driver model for dra74_evm as cpsw supports driver model.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com --- configs/dra74_evm_defconfig | 1 + 1 file changed, 1 insertion(+)
diff --git a/configs/dra74_evm_defconfig b/configs/dra74_evm_defconfig index 5f50004..682c126 100644 --- a/configs/dra74_evm_defconfig +++ b/configs/dra74_evm_defconfig @@ -25,3 +25,4 @@ CONFIG_OMAP_TIMER=y CONFIG_USB=y CONFIG_USB_GADGET=y CONFIG_DM_MMC=y +CONFIG_DM_ETH=y
On Thu, Apr 07, 2016 at 08:47:10PM +0530, Mugunthan V N wrote:
Enable eth driver model for dra74_evm as cpsw supports driver model.
Signed-off-by: Mugunthan V N mugunthanvnm@ti.com
Reviewed-by: Tom Rini trini@konsulko.com
participants (4)
-
Mugunthan V N -
Simon Glass -
Stephen Warren -
Tom Rini