[PATCH v2 0/6] airoha: Add initial support AN7581

[PATCH v2 0/6] airoha: Add initial support AN7581

[Christian Marangi]

This little series adds initial support for Airoha AN7581 SoC.

With the help of some backport patch, this use OF_UPSTREAM
directly.

Posting this to have the targer and the very basic driver.

Ethernet, SNAND and eMMC support is already ready downstream
and will be posted shortly after this gets approved.

Having the first driver ready permits to separately push
dedicate series for SNAND, eMMC and Ethrnet as they all depends
on basic support of clock and reset and nothing else.

Changes v2:
- Use OF_UPSTREAM directly
- Backport some patch for OF_UPSTREAM
- Make use of u-boot.dtsi files
- Fix defconfig
- Use pattern for MAINTAINERS file
- Correctly handle SYS_BOARD/SYS_CONFIG_NAME
- Drop config for header file

Christian Marangi (6):
  airoha: Add initial support for Airoha AN7581 SoC
  clk: airoha: Add support for Airoha AN7581 SoC clock
  reset: airoha: Add driver for controlling reset line of AN7581
  arm64: dts: airoha: en7581: Add Clock Controller node
  dt-bindings: clock: drop NUM_CLOCKS define for EN7581
  dt-bindings: clock: add ID for eMMC for EN7581

 arch/arm/Kconfig                              |   9 +
 arch/arm/Makefile                             |   1 +
 arch/arm/dts/an7581-u-boot.dtsi               |  18 +
 arch/arm/dts/en7581-evb-u-boot.dtsi           |  11 +
 arch/arm/mach-airoha/Kconfig                  |  32 ++
 arch/arm/mach-airoha/Makefile                 |   5 +
 arch/arm/mach-airoha/an7581/Makefile          |   3 +
 arch/arm/mach-airoha/an7581/init.c            |  47 ++
 arch/arm/mach-airoha/cpu.c                    |  20 +
 board/airoha/an7581/MAINTAINERS               |   5 +
 board/airoha/an7581/Makefile                  |   3 +
 board/airoha/an7581/an7581_rfb.c              |  16 +
 configs/an7581_evb_defconfig                  |  78 +++
 drivers/clk/Makefile                          |   1 +
 drivers/clk/airoha/Makefile                   |   3 +
 drivers/clk/airoha/clk-airoha.c               | 454 ++++++++++++++++++
 drivers/reset/Kconfig                         |   7 +
 drivers/reset/Makefile                        |   1 +
 drivers/reset/reset-airoha.c                  | 173 +++++++
 .../include/dt-bindings/clock/en7523-clk.h    |   2 +-
 dts/upstream/src/arm64/airoha/en7581.dtsi     |   8 +
 include/configs/an7581.h                      |  19 +
 22 files changed, 915 insertions(+), 1 deletion(-)
 create mode 100644 arch/arm/dts/an7581-u-boot.dtsi
 create mode 100644 arch/arm/dts/en7581-evb-u-boot.dtsi
 create mode 100644 arch/arm/mach-airoha/Kconfig
 create mode 100644 arch/arm/mach-airoha/Makefile
 create mode 100644 arch/arm/mach-airoha/an7581/Makefile
 create mode 100644 arch/arm/mach-airoha/an7581/init.c
 create mode 100644 arch/arm/mach-airoha/cpu.c
 create mode 100644 board/airoha/an7581/MAINTAINERS
 create mode 100644 board/airoha/an7581/Makefile
 create mode 100644 board/airoha/an7581/an7581_rfb.c
 create mode 100644 configs/an7581_evb_defconfig
 create mode 100644 drivers/clk/airoha/Makefile
 create mode 100644 drivers/clk/airoha/clk-airoha.c
 create mode 100644 drivers/reset/reset-airoha.c
 create mode 100644 include/configs/an7581.h

-- 
2.48.1

[PATCH v2 1/6] airoha: Add initial support for Airoha AN7581 SoC

[Christian Marangi]

Add initial support for Airoha AN7581 SoC. This adds the initial Kconfig
and Makefile entry for the SoC, an U-Boot specific DTSI and initial config
for it. Also add the initial code for CPU and RAM initialization.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
---
 arch/arm/Kconfig                     |  9 ++++
 arch/arm/Makefile                    |  1 +
 arch/arm/dts/an7581-u-boot.dtsi      | 18 +++++++
 arch/arm/dts/en7581-evb-u-boot.dtsi  | 11 ++++
 arch/arm/mach-airoha/Kconfig         | 32 ++++++++++++
 arch/arm/mach-airoha/Makefile        |  5 ++
 arch/arm/mach-airoha/an7581/Makefile |  3 ++
 arch/arm/mach-airoha/an7581/init.c   | 47 +++++++++++++++++
 arch/arm/mach-airoha/cpu.c           | 20 +++++++
 board/airoha/an7581/MAINTAINERS      |  5 ++
 board/airoha/an7581/Makefile         |  3 ++
 board/airoha/an7581/an7581_rfb.c     | 16 ++++++
 configs/an7581_evb_defconfig         | 78 ++++++++++++++++++++++++++++
 include/configs/an7581.h             | 19 +++++++
 14 files changed, 267 insertions(+)
 create mode 100644 arch/arm/dts/an7581-u-boot.dtsi
 create mode 100644 arch/arm/dts/en7581-evb-u-boot.dtsi
 create mode 100644 arch/arm/mach-airoha/Kconfig
 create mode 100644 arch/arm/mach-airoha/Makefile
 create mode 100644 arch/arm/mach-airoha/an7581/Makefile
 create mode 100644 arch/arm/mach-airoha/an7581/init.c
 create mode 100644 arch/arm/mach-airoha/cpu.c
 create mode 100644 board/airoha/an7581/MAINTAINERS
 create mode 100644 board/airoha/an7581/Makefile
 create mode 100644 board/airoha/an7581/an7581_rfb.c
 create mode 100644 configs/an7581_evb_defconfig
 create mode 100644 include/configs/an7581.h

diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index da6f1174934..4729ac39257 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -600,6 +600,13 @@ choice
 	prompt "Target select"
 	default TARGET_HIKEY
 
+config ARCH_AIROHA
+	bool "Airoha SoCs"
+	select DM
+	select OF_CONTROL
+	help
+	  Support for the Airoha soc.
+
 config ARCH_AT91
 	bool "Atmel AT91"
 	select GPIO_EXTRA_HEADER
@@ -2251,6 +2258,8 @@ config SYS_KWD_CONFIG
 	  Path within the source directory to the kwbimage.cfg file to use
 	  when packaging the U-Boot image for use.
 
+source "arch/arm/mach-airoha/Kconfig"
+
 source "arch/arm/mach-apple/Kconfig"
 
 source "arch/arm/mach-aspeed/Kconfig"
diff --git a/arch/arm/Makefile b/arch/arm/Makefile
index 7334e79965f..5ecadb2ef1b 100644
--- a/arch/arm/Makefile
+++ b/arch/arm/Makefile
@@ -51,6 +51,7 @@ PLATFORM_CPPFLAGS += $(arch-y) $(tune-y)
 
 # Machine directory name.  This list is sorted alphanumerically
 # by CONFIG_* macro name.
+machine-$(CONFIG_ARCH_AIROHA)		+= airoha
 machine-$(CONFIG_ARCH_APPLE)		+= apple
 machine-$(CONFIG_ARCH_ASPEED)		+= aspeed
 machine-$(CONFIG_ARCH_AT91)		+= at91
diff --git a/arch/arm/dts/an7581-u-boot.dtsi b/arch/arm/dts/an7581-u-boot.dtsi
new file mode 100644
index 00000000000..0316b73f3a5
--- /dev/null
+++ b/arch/arm/dts/an7581-u-boot.dtsi
@@ -0,0 +1,18 @@
+// SPDX-License-Identifier: GPL-2.0+
+
+/ {
+	reserved-memory {
+		#address-cells = <2>;
+		#size-cells = <2>;
+		ranges;
+
+		atf-reserved-memory@80000000 {
+			no-map;
+			reg = <0x0 0x80000000 0x0 0x40000>;
+		};
+	};
+};
+
+&uart1 {
+	bootph-all;
+};
diff --git a/arch/arm/dts/en7581-evb-u-boot.dtsi b/arch/arm/dts/en7581-evb-u-boot.dtsi
new file mode 100644
index 00000000000..ebd3b8b4958
--- /dev/null
+++ b/arch/arm/dts/en7581-evb-u-boot.dtsi
@@ -0,0 +1,11 @@
+// SPDX-License-Identifier: GPL-2.0+
+
+/ {
+	/* When running as a first-stage bootloader this isn't filled in automatically */
+	memory@80000000 {
+		device_type = "memory";
+		reg = <0x0 0x80000000 0x0 0x20000000>;
+	};
+};
+
+#include "an7581-u-boot.dtsi"
diff --git a/arch/arm/mach-airoha/Kconfig b/arch/arm/mach-airoha/Kconfig
new file mode 100644
index 00000000000..be3562ae3ff
--- /dev/null
+++ b/arch/arm/mach-airoha/Kconfig
@@ -0,0 +1,32 @@
+if ARCH_AIROHA
+
+config SYS_VENDOR
+	default "airoha"
+
+choice
+	prompt "Airoha board select"
+
+config TARGET_AN7581
+	bool "Airoha AN7581 SoC"
+	select ARM64
+	help
+	  The Airoha EN7581 is a ARM-based SoC with a quad-core Cortex-A7
+	  including NEON and GPU, Mali-450 graphics, several DDR3 options,
+	  crypto engine, built-in Wi-Fi / Bluetooth combo chip, JPEG decoder,
+	  video interfaces supporting HDMI and MIPI, and video codec support.
+	  Peripherals include Gigabit Ethernet, switch, USB3.0 and OTG, PCIe,
+	  I2S, PCM, S/PDIF, UART, SPI, I2C, IR TX/RX, and PWM.
+
+endchoice
+
+config SYS_SOC
+	default "an7581" if TARGET_AN7581
+
+config SYS_BOARD
+	default "an7581" if TARGET_AN7581
+
+config SYS_CONFIG_NAME
+	default "an7581" if TARGET_AN7581
+
+endif
+
diff --git a/arch/arm/mach-airoha/Makefile b/arch/arm/mach-airoha/Makefile
new file mode 100644
index 00000000000..215a300373b
--- /dev/null
+++ b/arch/arm/mach-airoha/Makefile
@@ -0,0 +1,5 @@
+# SPDX-License-Identifier:	GPL-2.0
+
+obj-y	+= cpu.o
+
+obj-$(CONFIG_TARGET_AN7581) += an7581/
diff --git a/arch/arm/mach-airoha/an7581/Makefile b/arch/arm/mach-airoha/an7581/Makefile
new file mode 100644
index 00000000000..886ab7e4eb9
--- /dev/null
+++ b/arch/arm/mach-airoha/an7581/Makefile
@@ -0,0 +1,3 @@
+# SPDX-License-Identifier:	GPL-2.0
+
+obj-y += init.o
diff --git a/arch/arm/mach-airoha/an7581/init.c b/arch/arm/mach-airoha/an7581/init.c
new file mode 100644
index 00000000000..cefe9c6db9e
--- /dev/null
+++ b/arch/arm/mach-airoha/an7581/init.c
@@ -0,0 +1,47 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <fdtdec.h>
+#include <init.h>
+#include <asm/armv8/mmu.h>
+#include <asm/system.h>
+
+int print_cpuinfo(void)
+{
+	printf("CPU:   Airoha AN7581\n");
+	return 0;
+}
+
+int dram_init(void)
+{
+	return fdtdec_setup_mem_size_base();
+}
+
+int dram_init_banksize(void)
+{
+	return fdtdec_setup_memory_banksize();
+}
+
+void reset_cpu(ulong addr)
+{
+	psci_system_reset();
+}
+
+static struct mm_region an7581_mem_map[] = {
+	{
+		/* DDR */
+		.virt = 0x80000000UL,
+		.phys = 0x80000000UL,
+		.size = 0x80000000UL,
+		.attrs = PTE_BLOCK_MEMTYPE(MT_NORMAL) | PTE_BLOCK_OUTER_SHARE,
+	}, {
+		.virt = 0x00000000UL,
+		.phys = 0x00000000UL,
+		.size = 0x20000000UL,
+		.attrs = PTE_BLOCK_MEMTYPE(MT_DEVICE_NGNRNE) |
+			 PTE_BLOCK_NON_SHARE |
+			 PTE_BLOCK_PXN | PTE_BLOCK_UXN
+	}, {
+		0,
+	}
+};
+struct mm_region *mem_map = an7581_mem_map;
diff --git a/arch/arm/mach-airoha/cpu.c b/arch/arm/mach-airoha/cpu.c
new file mode 100644
index 00000000000..a578e964664
--- /dev/null
+++ b/arch/arm/mach-airoha/cpu.c
@@ -0,0 +1,20 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <cpu_func.h>
+#include <dm.h>
+#include <init.h>
+#include <wdt.h>
+#include <dm/uclass-internal.h>
+
+int arch_cpu_init(void)
+{
+	icache_enable();
+
+	return 0;
+}
+
+void enable_caches(void)
+{
+	/* Enable D-cache. I-cache is already enabled in start.S */
+	dcache_enable();
+}
diff --git a/board/airoha/an7581/MAINTAINERS b/board/airoha/an7581/MAINTAINERS
new file mode 100644
index 00000000000..28ec2fbf2ea
--- /dev/null
+++ b/board/airoha/an7581/MAINTAINERS
@@ -0,0 +1,5 @@
+AN7581
+M:	Christian Marangi <ansuelsmth@gmail.com>
+S:	Maintained
+N:      airoha
+N:      an7581
diff --git a/board/airoha/an7581/Makefile b/board/airoha/an7581/Makefile
new file mode 100644
index 00000000000..70f8db7bce9
--- /dev/null
+++ b/board/airoha/an7581/Makefile
@@ -0,0 +1,3 @@
+# SPDX-License-Identifier:	GPL-2.0
+
+obj-y	+= an7581_rfb.o
diff --git a/board/airoha/an7581/an7581_rfb.c b/board/airoha/an7581/an7581_rfb.c
new file mode 100644
index 00000000000..aa73679d929
--- /dev/null
+++ b/board/airoha/an7581/an7581_rfb.c
@@ -0,0 +1,16 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Author: Christian Marangi <ansuelsmth@gmail.com>
+ */
+
+#include <asm/global_data.h>
+
+DECLARE_GLOBAL_DATA_PTR;
+
+int board_init(void)
+{
+	/* address of boot parameters */
+	gd->bd->bi_boot_params = CFG_SYS_SDRAM_BASE + 0x100;
+
+	return 0;
+}
diff --git a/configs/an7581_evb_defconfig b/configs/an7581_evb_defconfig
new file mode 100644
index 00000000000..f09b5b603a2
--- /dev/null
+++ b/configs/an7581_evb_defconfig
@@ -0,0 +1,78 @@
+CONFIG_ARM=y
+CONFIG_ARCH_AIROHA=y
+CONFIG_TEXT_BASE=0x81E00000
+CONFIG_SYS_MALLOC_F_LEN=0x4000
+CONFIG_NR_DRAM_BANKS=1
+CONFIG_ENV_SIZE=0x4000
+CONFIG_ENV_OFFSET=0x7c000
+CONFIG_DM_GPIO=y
+CONFIG_DEFAULT_DEVICE_TREE="airoha/en7581-evb"
+CONFIG_DM_RESET=y
+CONFIG_SYS_LOAD_ADDR=0x81800000
+CONFIG_BUILD_TARGET="u-boot.bin"
+# CONFIG_EFI_LOADER is not set
+CONFIG_FIT=y
+CONFIG_FIT_VERBOSE=y
+CONFIG_BOOTDELAY=3
+CONFIG_DEFAULT_FDT_FILE="en7581-evb"
+CONFIG_SYS_PBSIZE=1049
+CONFIG_SYS_CONSOLE_IS_IN_ENV=y
+# CONFIG_DISPLAY_BOARDINFO is not set
+CONFIG_HUSH_PARSER=y
+CONFIG_SYS_PROMPT="U-Boot> "
+CONFIG_SYS_MAXARGS=8
+CONFIG_CMD_BOOTZ=y
+CONFIG_CMD_BOOTMENU=y
+# CONFIG_CMD_ELF is not set
+# CONFIG_CMD_XIMG is not set
+CONFIG_CMD_BIND=y
+CONFIG_CMD_GPIO=y
+CONFIG_CMD_MMC=y
+CONFIG_CMD_MTD=y
+CONFIG_CMD_SF_TEST=y
+CONFIG_CMD_SPI=y
+# CONFIG_CMD_SETEXPR is not set
+CONFIG_CMD_PING=y
+CONFIG_CMD_EXT4=y
+CONFIG_CMD_FAT=y
+CONFIG_CMD_FS_GENERIC=y
+CONFIG_CMD_MTDPARTS=y
+CONFIG_CMD_LOG=y
+CONFIG_OF_UPSTREAM=y
+CONFIG_ENV_OVERWRITE=y
+CONFIG_ENV_IS_IN_MMC=y
+CONFIG_SYS_RELOC_GD_ENV_ADDR=y
+CONFIG_ENV_VARS_UBOOT_RUNTIME_CONFIG=y
+CONFIG_NET_RANDOM_ETHADDR=y
+CONFIG_REGMAP=y
+CONFIG_SYSCON=y
+CONFIG_CLK=y
+CONFIG_DMA=y
+CONFIG_LED=y
+CONFIG_LED_GPIO=y
+CONFIG_MMC_HS200_SUPPORT=y
+CONFIG_MTD=y
+CONFIG_DM_MTD=y
+CONFIG_MTD_SPI_NAND=y
+CONFIG_DM_SPI_FLASH=y
+CONFIG_SPI_FLASH_EON=y
+CONFIG_SPI_FLASH_GIGADEVICE=y
+CONFIG_SPI_FLASH_ISSI=y
+CONFIG_SPI_FLASH_MACRONIX=y
+CONFIG_SPI_FLASH_SPANSION=y
+CONFIG_SPI_FLASH_STMICRO=y
+CONFIG_SPI_FLASH_WINBOND=y
+CONFIG_SPI_FLASH_MTD=y
+CONFIG_PHYLIB=y
+CONFIG_PHY=y
+CONFIG_PINCTRL=y
+CONFIG_PINCONF=y
+CONFIG_POWER_DOMAIN=y
+CONFIG_DM_REGULATOR=y
+CONFIG_DM_REGULATOR_FIXED=y
+CONFIG_RAM=y
+CONFIG_DM_SERIAL=y
+CONFIG_SYS_NS16550=y
+CONFIG_SPI=y
+CONFIG_DM_SPI=y
+CONFIG_SHA512=y
diff --git a/include/configs/an7581.h b/include/configs/an7581.h
new file mode 100644
index 00000000000..64f04c9d9a5
--- /dev/null
+++ b/include/configs/an7581.h
@@ -0,0 +1,19 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Configuration for Airoha AN7581
+ */
+
+#ifndef __AN7581_H
+#define __AN7581_H
+
+#include <linux/sizes.h>
+
+#define CFG_SYS_UBOOT_BASE		CONFIG_TEXT_BASE
+
+#define CFG_SYS_INIT_RAM_ADDR           CONFIG_TEXT_BASE
+#define CFG_SYS_INIT_RAM_SIZE           SZ_2M
+
+/* DRAM */
+#define CFG_SYS_SDRAM_BASE		0x80000000
+
+#endif
-- 
2.48.1

[PATCH v2 2/6] clk: airoha: Add support for Airoha AN7581 SoC clock

[Christian Marangi]

Add support for Airoha AN7581 SoC clock driver. This mainly needed for
eMMC support to correctly get the current clock applied.

Based on the Linux clk-en7523.c but majorly reworked for U-Boot that
doesn't require CCF subsystem.

Major modification, support for set_rate, realtime get_rate and split
for reset part to a different driver.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
---
 drivers/clk/Makefile            |   1 +
 drivers/clk/airoha/Makefile     |   3 +
 drivers/clk/airoha/clk-airoha.c | 438 ++++++++++++++++++++++++++++++++
 3 files changed, 442 insertions(+)
 create mode 100644 drivers/clk/airoha/Makefile
 create mode 100644 drivers/clk/airoha/clk-airoha.c

diff --git a/drivers/clk/Makefile b/drivers/clk/Makefile
index fe0e49f6112..8411205ee04 100644
--- a/drivers/clk/Makefile
+++ b/drivers/clk/Makefile
@@ -14,6 +14,7 @@ obj-$(CONFIG_$(PHASE_)CLK_GPIO) += clk-gpio.o
 obj-$(CONFIG_$(PHASE_)CLK_STUB) += clk-stub.o
 
 obj-y += adi/
+obj-y += airoha/
 obj-y += analogbits/
 obj-y += imx/
 obj-$(CONFIG_CLK_JH7110) += starfive/
diff --git a/drivers/clk/airoha/Makefile b/drivers/clk/airoha/Makefile
new file mode 100644
index 00000000000..1744c5f7236
--- /dev/null
+++ b/drivers/clk/airoha/Makefile
@@ -0,0 +1,3 @@
+# SPDX-License-Identifier: GPL-2.0
+# Core
+obj-$(CONFIG_ARCH_AIROHA) += clk-airoha.o
diff --git a/drivers/clk/airoha/clk-airoha.c b/drivers/clk/airoha/clk-airoha.c
new file mode 100644
index 00000000000..96d120feba7
--- /dev/null
+++ b/drivers/clk/airoha/clk-airoha.c
@@ -0,0 +1,438 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Based on the Linux clk-en7523.c but majorly reworked
+ * for U-Boot that doesn't require CCF subsystem.
+ *
+ * Major modification, support for set_rate, realtime
+ * get_rate and split for reset part to a different driver.
+ *
+ * Author: Lorenzo Bianconi <lorenzo@kernel.org> (original driver)
+ *	   Christian Marangi <ansuelsmth@gmail.com>
+ */
+
+#include <clk-uclass.h>
+#include <dm.h>
+#include <dm/devres.h>
+#include <dm/device_compat.h>
+#include <dm/lists.h>
+#include <regmap.h>
+#include <syscon.h>
+
+#include <dt-bindings/clock/en7523-clk.h>
+
+#define REG_GSW_CLK_DIV_SEL		0x1b4
+#define REG_EMI_CLK_DIV_SEL		0x1b8
+#define REG_BUS_CLK_DIV_SEL		0x1bc
+#define REG_SPI_CLK_DIV_SEL		0x1c4
+#define REG_SPI_CLK_FREQ_SEL		0x1c8
+#define REG_NPU_CLK_DIV_SEL		0x1fc
+
+#define REG_NP_SCU_PCIC			0x88
+#define REG_NP_SCU_SSTR			0x9c
+#define REG_PCIE_XSI0_SEL_MASK		GENMASK(14, 13)
+#define REG_PCIE_XSI1_SEL_MASK		GENMASK(12, 11)
+#define REG_CRYPTO_CLKSRC2		0x20c
+
+#define EN7581_MAX_CLKS			9
+
+struct airoha_clk_desc {
+	int id;
+	const char *name;
+	u32 base_reg;
+	u8 base_bits;
+	u8 base_shift;
+	union {
+		const unsigned int *base_values;
+		unsigned int base_value;
+	};
+	size_t n_base_values;
+
+	u16 div_reg;
+	u8 div_bits;
+	u8 div_shift;
+	u16 div_val0;
+	u8 div_step;
+	u8 div_offset;
+};
+
+struct airoha_clk_priv {
+	struct regmap *chip_scu_map;
+	struct airoha_clk_soc_data *data;
+};
+
+struct airoha_clk_soc_data {
+	u32 num_clocks;
+	const struct airoha_clk_desc *descs;
+};
+
+static const u32 gsw_base[] = { 400000000, 500000000 };
+static const u32 slic_base[] = { 100000000, 3125000 };
+
+static const u32 emi7581_base[] = { 540000000, 480000000, 400000000, 300000000 };
+static const u32 bus7581_base[] = { 600000000, 540000000 };
+static const u32 npu7581_base[] = { 800000000, 750000000, 720000000, 600000000 };
+static const u32 crypto_base[] = { 540000000, 480000000 };
+static const u32 emmc7581_base[] = { 200000000, 150000000 };
+
+static const struct airoha_clk_desc en7581_base_clks[EN7581_MAX_CLKS] = {
+	[EN7523_CLK_GSW] = {
+		.id = EN7523_CLK_GSW,
+		.name = "gsw",
+
+		.base_reg = REG_GSW_CLK_DIV_SEL,
+		.base_bits = 1,
+		.base_shift = 8,
+		.base_values = gsw_base,
+		.n_base_values = ARRAY_SIZE(gsw_base),
+
+		.div_bits = 3,
+		.div_shift = 0,
+		.div_step = 1,
+		.div_offset = 1,
+	},
+	[EN7523_CLK_EMI] = {
+		.id = EN7523_CLK_EMI,
+		.name = "emi",
+
+		.base_reg = REG_EMI_CLK_DIV_SEL,
+		.base_bits = 2,
+		.base_shift = 8,
+		.base_values = emi7581_base,
+		.n_base_values = ARRAY_SIZE(emi7581_base),
+
+		.div_bits = 3,
+		.div_shift = 0,
+		.div_step = 1,
+		.div_offset = 1,
+	},
+	[EN7523_CLK_BUS] = {
+		.id = EN7523_CLK_BUS,
+		.name = "bus",
+
+		.base_reg = REG_BUS_CLK_DIV_SEL,
+		.base_bits = 1,
+		.base_shift = 8,
+		.base_values = bus7581_base,
+		.n_base_values = ARRAY_SIZE(bus7581_base),
+
+		.div_bits = 3,
+		.div_shift = 0,
+		.div_step = 1,
+		.div_offset = 1,
+	},
+	[EN7523_CLK_SLIC] = {
+		.id = EN7523_CLK_SLIC,
+		.name = "slic",
+
+		.base_reg = REG_SPI_CLK_FREQ_SEL,
+		.base_bits = 1,
+		.base_shift = 0,
+		.base_values = slic_base,
+		.n_base_values = ARRAY_SIZE(slic_base),
+
+		.div_reg = REG_SPI_CLK_DIV_SEL,
+		.div_bits = 5,
+		.div_shift = 24,
+		.div_val0 = 20,
+		.div_step = 2,
+	},
+	[EN7523_CLK_SPI] = {
+		.id = EN7523_CLK_SPI,
+		.name = "spi",
+
+		.base_reg = REG_SPI_CLK_DIV_SEL,
+
+		.base_value = 400000000,
+
+		.div_bits = 5,
+		.div_shift = 8,
+		.div_val0 = 40,
+		.div_step = 2,
+	},
+	[EN7523_CLK_NPU] = {
+		.id = EN7523_CLK_NPU,
+		.name = "npu",
+
+		.base_reg = REG_NPU_CLK_DIV_SEL,
+		.base_bits = 2,
+		.base_shift = 8,
+		.base_values = npu7581_base,
+		.n_base_values = ARRAY_SIZE(npu7581_base),
+
+		.div_bits = 3,
+		.div_shift = 0,
+		.div_step = 1,
+		.div_offset = 1,
+	},
+	[EN7523_CLK_CRYPTO] = {
+		.id = EN7523_CLK_CRYPTO,
+		.name = "crypto",
+
+		.base_reg = REG_CRYPTO_CLKSRC2,
+		.base_bits = 1,
+		.base_shift = 0,
+		.base_values = crypto_base,
+		.n_base_values = ARRAY_SIZE(crypto_base),
+	},
+	[EN7581_CLK_EMMC] = {
+		.id = EN7581_CLK_EMMC,
+		.name = "emmc",
+
+		.base_reg = REG_CRYPTO_CLKSRC2,
+		.base_bits = 1,
+		.base_shift = 12,
+		.base_values = emmc7581_base,
+		.n_base_values = ARRAY_SIZE(emmc7581_base),
+	}
+};
+
+static u32 airoha_clk_get_base_rate(const struct airoha_clk_desc *desc, u32 val)
+{
+	if (!desc->base_bits)
+		return desc->base_value;
+
+	val >>= desc->base_shift;
+	val &= (1 << desc->base_bits) - 1;
+
+	if (val >= desc->n_base_values)
+		return 0;
+
+	return desc->base_values[val];
+}
+
+static u32 airoha_clk_get_div(const struct airoha_clk_desc *desc, u32 val)
+{
+	if (!desc->div_bits)
+		return 1;
+
+	val >>= desc->div_shift;
+	val &= (1 << desc->div_bits) - 1;
+
+	if (!val && desc->div_val0)
+		return desc->div_val0;
+
+	return (val + desc->div_offset) * desc->div_step;
+}
+
+static int airoha_clk_enable(struct clk *clk)
+{
+	struct airoha_clk_priv *priv = dev_get_priv(clk->dev);
+	struct airoha_clk_soc_data *data = priv->data;
+	int id = clk->id;
+
+	if (id > data->num_clocks)
+		return -EINVAL;
+
+	return 0;
+}
+
+static int airoha_clk_disable(struct clk *clk)
+{
+	return 0;
+}
+
+static ulong airoha_clk_get_rate(struct clk *clk)
+{
+	struct airoha_clk_priv *priv = dev_get_priv(clk->dev);
+	struct airoha_clk_soc_data *data = priv->data;
+	const struct airoha_clk_desc *desc;
+	struct regmap *map = priv->chip_scu_map;
+	int id = clk->id;
+	u32 reg, val;
+	ulong rate;
+	int ret;
+
+	if (id > data->num_clocks) {
+		dev_err(clk->dev, "Invalid clk ID %d\n", id);
+		return 0;
+	}
+
+	desc = &data->descs[id];
+
+	ret = regmap_read(map, desc->base_reg, &val);
+	if (ret) {
+		dev_err(clk->dev, "Failed to read reg for clock %s\n",
+			desc->name);
+		return 0;
+	}
+
+	rate = airoha_clk_get_base_rate(desc, val);
+
+	reg = desc->div_reg ? desc->div_reg : desc->base_reg;
+	ret = regmap_read(map, reg, &val);
+	if (ret) {
+		dev_err(clk->dev, "Failed to read reg for clock %s\n",
+			desc->name);
+		return 0;
+	}
+
+	rate /= airoha_clk_get_div(desc, val);
+
+	return rate;
+}
+
+static int airoha_clk_search_rate(const struct airoha_clk_desc *desc, int div,
+				  ulong rate)
+{
+	int i;
+
+	/* Single base rate */
+	if (!desc->base_bits) {
+		if (rate != desc->base_value / div)
+			goto err;
+
+		return 0;
+	}
+
+	/* Check every base rate with provided divisor */
+	for (i = 0; i < desc->n_base_values; i++)
+		if (rate == desc->base_values[i] / div)
+			return i;
+
+err:
+	return -EINVAL;
+}
+
+static ulong airoha_clk_set_rate(struct clk *clk, ulong rate)
+{
+	struct airoha_clk_priv *priv = dev_get_priv(clk->dev);
+	struct airoha_clk_soc_data *data = priv->data;
+	const struct airoha_clk_desc *desc;
+	struct regmap *map = priv->chip_scu_map;
+	int div_val, base_val;
+	u32 reg, val, mask;
+	int id = clk->id;
+	int div;
+	int ret;
+
+	if (id > data->num_clocks) {
+		dev_err(clk->dev, "Invalid clk ID %d\n", id);
+		return 0;
+	}
+
+	desc = &data->descs[id];
+
+	if (!desc->base_bits && !desc->div_bits) {
+		dev_err(clk->dev, "Can't set rate for fixed clock %s\n",
+			desc->name);
+		return 0;
+	}
+
+	if (!desc->div_bits) {
+		/* Divisor not supported, just search in base rate */
+		div_val = 0;
+		base_val = airoha_clk_search_rate(desc, 1, rate);
+		if (base_val < 0) {
+			dev_err(clk->dev, "Invalid rate for clock %s\n",
+				desc->name);
+			return 0;
+		}
+	} else {
+		div_val = 0;
+
+		/* Check if div0 satisfy the request */
+		if (desc->div_val0) {
+			base_val = airoha_clk_search_rate(desc, desc->div_val0,
+							  rate);
+			if (base_val >= 0) {
+				div_val = 0;
+				goto apply;
+			}
+
+			/* Skip checking first divisor val */
+			div_val = 1;
+		}
+
+		/* Simulate rate with every divisor supported */
+		for (div_val = div_val + desc->div_offset;
+		     div_val < BIT(desc->div_bits) - 1; div_val++) {
+			div = div_val * desc->div_step;
+
+			base_val = airoha_clk_search_rate(desc, div, rate);
+			if (base_val >= 0)
+				break;
+		}
+
+		if (div_val == BIT(desc->div_bits) - 1) {
+			dev_err(clk->dev, "Invalid rate for clock %s\n",
+				desc->name);
+			return 0;
+		}
+	}
+
+apply:
+	if (desc->div_bits) {
+		reg = desc->div_reg ? desc->div_reg : desc->base_reg;
+
+		mask = (BIT(desc->div_bits) - 1) << desc->div_shift;
+		val = div_val << desc->div_shift;
+
+		ret = regmap_update_bits(map, reg, mask, val);
+		if (ret) {
+			dev_err(clk->dev, "Failed to update div reg for clock %s\n",
+				desc->name);
+			return 0;
+		}
+	}
+
+	if (desc->base_bits) {
+		mask = (BIT(desc->base_bits) - 1) << desc->base_shift;
+		val = base_val << desc->base_shift;
+
+		ret = regmap_update_bits(map, desc->base_reg, mask, val);
+		if (ret) {
+			dev_err(clk->dev, "Failed to update reg for clock %s\n",
+				desc->name);
+			return 0;
+		}
+	}
+
+	return rate;
+}
+
+const struct clk_ops airoha_clk_ops = {
+	.enable = airoha_clk_enable,
+	.disable = airoha_clk_disable,
+	.get_rate = airoha_clk_get_rate,
+	.set_rate = airoha_clk_set_rate,
+};
+
+static int airoha_clk_probe(struct udevice *dev)
+{
+	struct airoha_clk_priv *priv = dev_get_priv(dev);
+	ofnode chip_scu_node;
+
+	chip_scu_node = ofnode_by_compatible(ofnode_null(),
+					     "airoha,en7581-chip-scu");
+	if (!ofnode_valid(chip_scu_node))
+		return -EINVAL;
+
+	priv->chip_scu_map = syscon_node_to_regmap(chip_scu_node);
+	if (IS_ERR(priv->chip_scu_map))
+		return PTR_ERR(priv->chip_scu_map);
+
+	priv->data = (void *)dev_get_driver_data(dev);
+
+	return 0;
+}
+
+static const struct airoha_clk_soc_data en7581_data = {
+	.num_clocks = ARRAY_SIZE(en7581_base_clks),
+	.descs = en7581_base_clks,
+};
+
+static const struct udevice_id airoha_clk_ids[] = {
+	{ .compatible = "airoha,en7581-scu",
+	  .data = (ulong)&en7581_data,
+	},
+	{ }
+};
+
+U_BOOT_DRIVER(airoha_clk) = {
+	.name = "clk-airoha",
+	.id = UCLASS_CLK,
+	.of_match = airoha_clk_ids,
+	.probe = airoha_clk_probe,
+	.priv_auto = sizeof(struct airoha_clk_priv),
+	.ops = &airoha_clk_ops,
+};
-- 
2.48.1

[PATCH v2 3/6] reset: airoha: Add driver for controlling reset line of AN7581

[Christian Marangi]

Add driver for controlling the reset lines of AN7581. This is a detached
version of the clock controller driver present in Linux only used to
control reset lines. Driver gets loaded with the bind of the clock
driver and doesn't require a compatible. This is needed as they share
the same registers.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
---
 drivers/clk/airoha/clk-airoha.c |  16 +++
 drivers/reset/Kconfig           |   7 ++
 drivers/reset/Makefile          |   1 +
 drivers/reset/reset-airoha.c    | 173 ++++++++++++++++++++++++++++++++
 4 files changed, 197 insertions(+)
 create mode 100644 drivers/reset/reset-airoha.c

diff --git a/drivers/clk/airoha/clk-airoha.c b/drivers/clk/airoha/clk-airoha.c
index 96d120feba7..1b2c4c98de5 100644
--- a/drivers/clk/airoha/clk-airoha.c
+++ b/drivers/clk/airoha/clk-airoha.c
@@ -416,6 +416,21 @@ static int airoha_clk_probe(struct udevice *dev)
 	return 0;
 }
 
+static int airoha_clk_bind(struct udevice *dev)
+{
+	struct udevice *rst_dev;
+	int ret = 0;
+
+	if (CONFIG_IS_ENABLED(RESET_AIROHA)) {
+		ret = device_bind_driver_to_node(dev, "airoha-reset", "reset",
+						 dev_ofnode(dev), &rst_dev);
+		if (ret)
+			debug("Warning: failed to bind reset controller\n");
+	}
+
+	return ret;
+}
+
 static const struct airoha_clk_soc_data en7581_data = {
 	.num_clocks = ARRAY_SIZE(en7581_base_clks),
 	.descs = en7581_base_clks,
@@ -433,6 +448,7 @@ U_BOOT_DRIVER(airoha_clk) = {
 	.id = UCLASS_CLK,
 	.of_match = airoha_clk_ids,
 	.probe = airoha_clk_probe,
+	.bind = airoha_clk_bind,
 	.priv_auto = sizeof(struct airoha_clk_priv),
 	.ops = &airoha_clk_ops,
 };
diff --git a/drivers/reset/Kconfig b/drivers/reset/Kconfig
index fe5c1214f57..ee923c19f41 100644
--- a/drivers/reset/Kconfig
+++ b/drivers/reset/Kconfig
@@ -49,6 +49,13 @@ config TEGRA186_RESET
 	  Enable support for manipulating Tegra's on-SoC reset signals via IPC
 	  requests to the BPMP (Boot and Power Management Processor).
 
+config RESET_AIROHA
+	bool "Reset controller driver for Airoha SoCs"
+	depends on DM_RESET && ARCH_AIROHA
+	default y
+	help
+	  Support for reset controller on Airoha SoCs.
+
 config RESET_TI_SCI
 	bool "TI System Control Interface (TI SCI) reset driver"
 	depends on DM_RESET && TI_SCI_PROTOCOL
diff --git a/drivers/reset/Makefile b/drivers/reset/Makefile
index d99a78c9828..524fc064417 100644
--- a/drivers/reset/Makefile
+++ b/drivers/reset/Makefile
@@ -10,6 +10,7 @@ obj-$(CONFIG_STI_RESET) += sti-reset.o
 obj-$(CONFIG_STM32_RESET) += stm32-reset.o
 obj-$(CONFIG_TEGRA_CAR_RESET) += tegra-car-reset.o
 obj-$(CONFIG_TEGRA186_RESET) += tegra186-reset.o
+obj-$(CONFIG_RESET_AIROHA) += reset-airoha.o
 obj-$(CONFIG_RESET_TI_SCI) += reset-ti-sci.o
 obj-$(CONFIG_RESET_HSDK) += reset-hsdk.o
 obj-$(CONFIG_RESET_BCM6345) += reset-bcm6345.o
diff --git a/drivers/reset/reset-airoha.c b/drivers/reset/reset-airoha.c
new file mode 100644
index 00000000000..e878af6167c
--- /dev/null
+++ b/drivers/reset/reset-airoha.c
@@ -0,0 +1,173 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Based on Linux drivers/clk/clk-en7523.c reworked
+ * and detached to a dedicated driver
+ *
+ * Author: Lorenzo Bianconi <lorenzo@kernel.org> (original driver)
+ *	   Christian Marangi <ansuelsmth@gmail.com>
+ */
+
+#include <dm.h>
+#include <linux/io.h>
+#include <reset-uclass.h>
+
+#include <dt-bindings/reset/airoha,en7581-reset.h>
+
+#define RST_NR_PER_BANK			32
+
+#define REG_RESET_CONTROL2		0x830
+#define REG_RESET_CONTROL1		0x834
+
+struct airoha_reset_priv {
+	const u16 *bank_ofs;
+	const u16 *idx_map;
+	void __iomem *base;
+};
+
+static const u16 en7581_rst_ofs[] = {
+	REG_RESET_CONTROL2,
+	REG_RESET_CONTROL1,
+};
+
+static const u16 en7581_rst_map[] = {
+	/* RST_CTRL2 */
+	[EN7581_XPON_PHY_RST]		= 0,
+	[EN7581_CPU_TIMER2_RST]		= 2,
+	[EN7581_HSUART_RST]		= 3,
+	[EN7581_UART4_RST]		= 4,
+	[EN7581_UART5_RST]		= 5,
+	[EN7581_I2C2_RST]		= 6,
+	[EN7581_XSI_MAC_RST]		= 7,
+	[EN7581_XSI_PHY_RST]		= 8,
+	[EN7581_NPU_RST]		= 9,
+	[EN7581_I2S_RST]		= 10,
+	[EN7581_TRNG_RST]		= 11,
+	[EN7581_TRNG_MSTART_RST]	= 12,
+	[EN7581_DUAL_HSI0_RST]		= 13,
+	[EN7581_DUAL_HSI1_RST]		= 14,
+	[EN7581_HSI_RST]		= 15,
+	[EN7581_DUAL_HSI0_MAC_RST]	= 16,
+	[EN7581_DUAL_HSI1_MAC_RST]	= 17,
+	[EN7581_HSI_MAC_RST]		= 18,
+	[EN7581_WDMA_RST]		= 19,
+	[EN7581_WOE0_RST]		= 20,
+	[EN7581_WOE1_RST]		= 21,
+	[EN7581_HSDMA_RST]		= 22,
+	[EN7581_TDMA_RST]		= 24,
+	[EN7581_EMMC_RST]		= 25,
+	[EN7581_SOE_RST]		= 26,
+	[EN7581_PCIE2_RST]		= 27,
+	[EN7581_XFP_MAC_RST]		= 28,
+	[EN7581_USB_HOST_P1_RST]	= 29,
+	[EN7581_USB_HOST_P1_U3_PHY_RST]	= 30,
+	/* RST_CTRL1 */
+	[EN7581_PCM1_ZSI_ISI_RST]	= RST_NR_PER_BANK + 0,
+	[EN7581_FE_PDMA_RST]		= RST_NR_PER_BANK + 1,
+	[EN7581_FE_QDMA_RST]		= RST_NR_PER_BANK + 2,
+	[EN7581_PCM_SPIWP_RST]		= RST_NR_PER_BANK + 4,
+	[EN7581_CRYPTO_RST]		= RST_NR_PER_BANK + 6,
+	[EN7581_TIMER_RST]		= RST_NR_PER_BANK + 8,
+	[EN7581_PCM1_RST]		= RST_NR_PER_BANK + 11,
+	[EN7581_UART_RST]		= RST_NR_PER_BANK + 12,
+	[EN7581_GPIO_RST]		= RST_NR_PER_BANK + 13,
+	[EN7581_GDMA_RST]		= RST_NR_PER_BANK + 14,
+	[EN7581_I2C_MASTER_RST]		= RST_NR_PER_BANK + 16,
+	[EN7581_PCM2_ZSI_ISI_RST]	= RST_NR_PER_BANK + 17,
+	[EN7581_SFC_RST]		= RST_NR_PER_BANK + 18,
+	[EN7581_UART2_RST]		= RST_NR_PER_BANK + 19,
+	[EN7581_GDMP_RST]		= RST_NR_PER_BANK + 20,
+	[EN7581_FE_RST]			= RST_NR_PER_BANK + 21,
+	[EN7581_USB_HOST_P0_RST]	= RST_NR_PER_BANK + 22,
+	[EN7581_GSW_RST]		= RST_NR_PER_BANK + 23,
+	[EN7581_SFC2_PCM_RST]		= RST_NR_PER_BANK + 25,
+	[EN7581_PCIE0_RST]		= RST_NR_PER_BANK + 26,
+	[EN7581_PCIE1_RST]		= RST_NR_PER_BANK + 27,
+	[EN7581_CPU_TIMER_RST]		= RST_NR_PER_BANK + 28,
+	[EN7581_PCIE_HB_RST]		= RST_NR_PER_BANK + 29,
+	[EN7581_XPON_MAC_RST]		= RST_NR_PER_BANK + 31,
+};
+
+static int airoha_reset_update(struct airoha_reset_priv *priv,
+			       unsigned long id, bool assert)
+{
+	void __iomem *addr = priv->base + priv->bank_ofs[id / RST_NR_PER_BANK];
+	u32 val;
+
+	val = readl(addr);
+	if (assert)
+		val |= BIT(id % RST_NR_PER_BANK);
+	else
+		val &= ~BIT(id % RST_NR_PER_BANK);
+	writel(val, addr);
+
+	return 0;
+}
+
+static int airoha_reset_assert(struct reset_ctl *reset_ctl)
+{
+	struct airoha_reset_priv *priv = dev_get_priv(reset_ctl->dev);
+	int id = reset_ctl->id;
+
+	return airoha_reset_update(priv, id, true);
+}
+
+static int airoha_reset_deassert(struct reset_ctl *reset_ctl)
+{
+	struct airoha_reset_priv *priv = dev_get_priv(reset_ctl->dev);
+	int id = reset_ctl->id;
+
+	return airoha_reset_update(priv, id, false);
+}
+
+static int airoha_reset_status(struct reset_ctl *reset_ctl)
+{
+	struct airoha_reset_priv *priv = dev_get_priv(reset_ctl->dev);
+	int id = reset_ctl->id;
+	void __iomem *addr;
+
+	addr = priv->base + priv->bank_ofs[id / RST_NR_PER_BANK];
+
+	return !!(readl(addr) & BIT(id % RST_NR_PER_BANK));
+}
+
+static int airoha_reset_xlate(struct reset_ctl *reset_ctl,
+			      struct ofnode_phandle_args *args)
+{
+	struct airoha_reset_priv *priv = dev_get_priv(reset_ctl->dev);
+
+	if (args->args[0] >= ARRAY_SIZE(en7581_rst_map))
+		return -EINVAL;
+
+	reset_ctl->id = priv->idx_map[args->args[0]];
+
+	return 0;
+}
+
+static struct reset_ops airoha_reset_ops = {
+	.of_xlate = airoha_reset_xlate,
+	.rst_assert = airoha_reset_assert,
+	.rst_deassert = airoha_reset_deassert,
+	.rst_status = airoha_reset_status,
+};
+
+static int airoha_reset_probe(struct udevice *dev)
+{
+	struct airoha_reset_priv *priv = dev_get_priv(dev);
+
+	priv->base = dev_remap_addr(dev);
+	if (!priv->base)
+		return -ENOMEM;
+
+	priv->bank_ofs = en7581_rst_ofs;
+	priv->idx_map = en7581_rst_map;
+
+	return 0;
+}
+
+U_BOOT_DRIVER(airoha_reset) = {
+	.name = "airoha-reset",
+	.id = UCLASS_RESET,
+	.probe = airoha_reset_probe,
+	.ops = &airoha_reset_ops,
+	.priv_auto = sizeof(struct airoha_reset_priv),
+};
-- 
2.48.1

[PATCH v2 4/6] arm64: dts: airoha: en7581: Add Clock Controller node

[Christian Marangi]

Add Clock Controller node for EN7581 SoC to correctly expose supported
clock for any user in the SoC.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://lore.kernel.org/r/20250105150328.15172-1-ansuelsmth@gmail.com
Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>

[ upstream commit: 7693017580e9be839fa5f27130bb6500f3597595 ]
---
 dts/upstream/src/arm64/airoha/en7581.dtsi | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/dts/upstream/src/arm64/airoha/en7581.dtsi b/dts/upstream/src/arm64/airoha/en7581.dtsi
index 55eb1762fb1..f584409e72c 100644
--- a/dts/upstream/src/arm64/airoha/en7581.dtsi
+++ b/dts/upstream/src/arm64/airoha/en7581.dtsi
@@ -2,6 +2,7 @@
 
 #include <dt-bindings/interrupt-controller/irq.h>
 #include <dt-bindings/interrupt-controller/arm-gic.h>
+#include <dt-bindings/clock/en7523-clk.h>
 
 / {
 	interrupt-parent = <&gic>;
@@ -142,6 +143,13 @@
 			interrupts = <GIC_PPI 9 IRQ_TYPE_LEVEL_LOW>;
 		};
 
+		scuclk: clock-controller@1fa20000 {
+			compatible = "airoha,en7581-scu";
+			reg = <0x0 0x1fb00000 0x0 0x970>;
+			#clock-cells = <1>;
+			#reset-cells = <1>;
+		};
+
 		uart1: serial@1fbf0000 {
 			compatible = "ns16550";
 			reg = <0x0 0x1fbf0000 0x0 0x30>;
-- 
2.48.1

[PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Christian Marangi]

Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
should not be placed here. Value is derived internally in the user
driver.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20250113231030.6735-3-ansuelsmth@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

[ upstream commit: 02d3b7557ce28c373ea1e925ae16ab5988284313 ]
---
 dts/upstream/include/dt-bindings/clock/en7523-clk.h | 2 --
 1 file changed, 2 deletions(-)

diff --git a/dts/upstream/include/dt-bindings/clock/en7523-clk.h b/dts/upstream/include/dt-bindings/clock/en7523-clk.h
index 717d23a5e5a..28e56745ccf 100644
--- a/dts/upstream/include/dt-bindings/clock/en7523-clk.h
+++ b/dts/upstream/include/dt-bindings/clock/en7523-clk.h
@@ -12,6 +12,4 @@
 #define EN7523_CLK_CRYPTO	6
 #define EN7523_CLK_PCIE		7
 
-#define EN7523_NUM_CLOCKS	8
-
 #endif /* _DT_BINDINGS_CLOCK_AIROHA_EN7523_H_ */
-- 
2.48.1

[PATCH v2 6/6] dt-bindings: clock: add ID for eMMC for EN7581

[Christian Marangi]

Add ID for eMMC for EN7581. This is to control clock selection of eMMC
between 200MHz and 150MHz.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Acked-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20250113231030.6735-4-ansuelsmth@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

[ upstream commit: 82108ad3285f58f314ad41398f44017c7dbe44de ]
---
 dts/upstream/include/dt-bindings/clock/en7523-clk.h | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/dts/upstream/include/dt-bindings/clock/en7523-clk.h b/dts/upstream/include/dt-bindings/clock/en7523-clk.h
index 28e56745ccf..edfa64045f5 100644
--- a/dts/upstream/include/dt-bindings/clock/en7523-clk.h
+++ b/dts/upstream/include/dt-bindings/clock/en7523-clk.h
@@ -12,4 +12,6 @@
 #define EN7523_CLK_CRYPTO	6
 #define EN7523_CLK_PCIE		7
 
+#define EN7581_CLK_EMMC		8
+
 #endif /* _DT_BINDINGS_CLOCK_AIROHA_EN7523_H_ */
-- 
2.48.1

Re: [PATCH v2 0/6] airoha: Add initial support AN7581

[Tom Rini]

[-- Attachment #1: Type: text/plain, Size: 907 bytes --]

On Fri, Mar 14, 2025 at 07:59:20PM +0100, Christian Marangi wrote:
`
> This little series adds initial support for Airoha AN7581 SoC.
> 
> With the help of some backport patch, this use OF_UPSTREAM
> directly.
> 
> Posting this to have the targer and the very basic driver.
> 
> Ethernet, SNAND and eMMC support is already ready downstream
> and will be posted shortly after this gets approved.
> 
> Having the first driver ready permits to separately push
> dedicate series for SNAND, eMMC and Ethrnet as they all depends
> on basic support of clock and reset and nothing else.
> 
> Changes v2:
> - Use OF_UPSTREAM directly
> - Backport some patch for OF_UPSTREAM
> - Make use of u-boot.dtsi files
> - Fix defconfig
> - Use pattern for MAINTAINERS file
> - Correctly handle SYS_BOARD/SYS_CONFIG_NAME
> - Drop config for header file

This generally looks good, thanks.

-- 
Tom

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 659 bytes --]

Re: [PATCH v2 0/6] airoha: Add initial support AN7581

[Christian Marangi]

On Fri, Mar 14, 2025 at 02:26:11PM -0600, Tom Rini wrote:
> On Fri, Mar 14, 2025 at 07:59:20PM +0100, Christian Marangi wrote:
> `
> > This little series adds initial support for Airoha AN7581 SoC.
> > 
> > With the help of some backport patch, this use OF_UPSTREAM
> > directly.
> > 
> > Posting this to have the targer and the very basic driver.
> > 
> > Ethernet, SNAND and eMMC support is already ready downstream
> > and will be posted shortly after this gets approved.
> > 
> > Having the first driver ready permits to separately push
> > dedicate series for SNAND, eMMC and Ethrnet as they all depends
> > on basic support of clock and reset and nothing else.
> > 
> > Changes v2:
> > - Use OF_UPSTREAM directly
> > - Backport some patch for OF_UPSTREAM
> > - Make use of u-boot.dtsi files
> > - Fix defconfig
> > - Use pattern for MAINTAINERS file
> > - Correctly handle SYS_BOARD/SYS_CONFIG_NAME
> > - Drop config for header file
> 
> This generally looks good, thanks.
> 

Hi Tom, may I ask status of this? I would like to continue with posting
follow-up driver like mmc nand ethernet for this. Is it ok if I say this
series is a dependency so I can start the review process for it?

-- 
	Ansuel

Re: [PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Krzysztof Kozlowski]

On 14/03/2025 19:59, Christian Marangi wrote:
> Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
> should not be placed here. Value is derived internally in the user
> driver.
> 
> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Please drop my Ack. I have never acked such patch for uboot. If I did,
it was by mistake - probably you CC-ed me for some reason.

Best regards,
Krzysztof

Re: [PATCH v2 6/6] dt-bindings: clock: add ID for eMMC for EN7581

[Krzysztof Kozlowski]

On 14/03/2025 19:59, Christian Marangi wrote:
> Add ID for eMMC for EN7581. This is to control clock selection of eMMC
> between 200MHz and 150MHz.
> 
> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> Acked-by: Conor Dooley <conor.dooley@microchip.com>
> Link: https://lore.kernel.org/r/20250113231030.6735-4-ansuelsmth@gmail.com
> Signed-off-by: Stephen Boyd <sboyd@kernel.org>
> 
> [ upstream commit: 82108ad3285f58f314ad41398f44017c7dbe44de ]
> ---
>  dts/upstream/include/dt-bindings/clock/en7523-clk.h | 2 ++


I don't think Conor ever acked above change to "dts/upstream/" folder.


Best regards,
Krzysztof

Re: [PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Christian Marangi]

On Thu, Mar 27, 2025 at 03:43:47PM +0100, Krzysztof Kozlowski wrote:
> On 14/03/2025 19:59, Christian Marangi wrote:
> > Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
> > should not be placed here. Value is derived internally in the user
> > driver.
> > 
> > Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> > Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
> Please drop my Ack. I have never acked such patch for uboot. If I did,
> it was by mistake - probably you CC-ed me for some reason.
>

Some explaination, uboot introduced the concept of upstream where they
"import" linux patch for dts and dt-bindings.

This and the other patch are the exact upstream patch with only the path
changed so I keep all the patch commit message with tags and added the

[ upstream commit ] thing.

Hope Tom can better suggest how this should be done. You were CC
probably because the git send-email included you as present in the
different tags.

-- 
	Ansuel

Re: [PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Krzysztof Kozlowski]

On 27/03/2025 15:50, Christian Marangi wrote:
> On Thu, Mar 27, 2025 at 03:43:47PM +0100, Krzysztof Kozlowski wrote:
>> On 14/03/2025 19:59, Christian Marangi wrote:
>>> Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
>>> should not be placed here. Value is derived internally in the user
>>> driver.
>>>
>>> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
>>> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
>> Please drop my Ack. I have never acked such patch for uboot. If I did,
>> it was by mistake - probably you CC-ed me for some reason.
>>
> 
> Some explaination, uboot introduced the concept of upstream where they
> "import" linux patch for dts and dt-bindings.

I expected OF_UPSTREAM to be taking the sources, not patches.

> 
> This and the other patch are the exact upstream patch with only the path
> changed so I keep all the patch commit message with tags and added the
> 
> [ upstream commit ] thing.
> 
> Hope Tom can better suggest how this should be done. You were CC
> probably because the git send-email included you as present in the
> different tags.

Well, Ack is still not valid because I did not Ack exactly that change.
It does not matter for the ack, but for Reviewed-by it would matter,
because it is a statement (of oversight...). I cannot control what you
put into patches taken out of kernel, but at least do not Cc me on that.

Best regards,
Krzysztof

Re: [PATCH v2 0/6] airoha: Add initial support AN7581

[Andreas Gnau]

On 2025-03-14 19:59, Christian Marangi wrote:
> This little series adds initial support for Airoha AN7581 SoC.

Very nice. Will you (or someone else) upstream or at least open-source 
ATF parts as well, similar to how Mediatek does it?

Re: [PATCH v2 0/6] airoha: Add initial support AN7581

[Christian Marangi]

On Tue, Apr 01, 2025 at 12:45:21PM +0200, Andreas Gnau wrote:
> On 2025-03-14 19:59, Christian Marangi wrote:
> > This little series adds initial support for Airoha AN7581 SoC.
> 
> Very nice. Will you (or someone else) upstream or at least open-source ATF
> parts as well, similar to how Mediatek does it?

It's planned. I have the source and I have permission to opesource it
(there are lots of .o file but whatever)

It needs lots of love tho as they modified ATF fork A LOT. I'm giving
priority on U-Boot currently.

-- 
	Ansuel

Re: [PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Tom Rini]

[-- Attachment #1: Type: text/plain, Size: 2013 bytes --]

On Thu, Mar 27, 2025 at 03:58:52PM +0100, Krzysztof Kozlowski wrote:
> On 27/03/2025 15:50, Christian Marangi wrote:
> > On Thu, Mar 27, 2025 at 03:43:47PM +0100, Krzysztof Kozlowski wrote:
> >> On 14/03/2025 19:59, Christian Marangi wrote:
> >>> Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
> >>> should not be placed here. Value is derived internally in the user
> >>> driver.
> >>>
> >>> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> >>> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
> >> Please drop my Ack. I have never acked such patch for uboot. If I did,
> >> it was by mistake - probably you CC-ed me for some reason.
> >>
> > 
> > Some explaination, uboot introduced the concept of upstream where they
> > "import" linux patch for dts and dt-bindings.
> 
> I expected OF_UPSTREAM to be taking the sources, not patches.
> 
> > 
> > This and the other patch are the exact upstream patch with only the path
> > changed so I keep all the patch commit message with tags and added the
> > 
> > [ upstream commit ] thing.
> > 
> > Hope Tom can better suggest how this should be done. You were CC
> > probably because the git send-email included you as present in the
> > different tags.
> 
> Well, Ack is still not valid because I did not Ack exactly that change.
> It does not matter for the ack, but for Reviewed-by it would matter,
> because it is a statement (of oversight...). I cannot control what you
> put into patches taken out of kernel, but at least do not Cc me on that.

In specifics, yes, we should update doc/develop/devicetree/control.rst
and maybe doc/develop/sending_patches.rst to use --suppress-cc=all for
dts/upstream.

But in general, what do you expect people to be doing with content from
devicetree-rebasing? We're doing some direct cherry-picks in between
merging of the tags. I think it would be weird to be dropping the tags
and un-attributing peoples work.

-- 
Tom

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 659 bytes --]

Re: [PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Conor Dooley]

[-- Attachment #1: Type: text/plain, Size: 2396 bytes --]

On Tue, Apr 01, 2025 at 08:44:20AM -0600, Tom Rini wrote:
> On Thu, Mar 27, 2025 at 03:58:52PM +0100, Krzysztof Kozlowski wrote:
> > On 27/03/2025 15:50, Christian Marangi wrote:
> > > On Thu, Mar 27, 2025 at 03:43:47PM +0100, Krzysztof Kozlowski wrote:
> > >> On 14/03/2025 19:59, Christian Marangi wrote:
> > >>> Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
> > >>> should not be placed here. Value is derived internally in the user
> > >>> driver.
> > >>>
> > >>> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> > >>> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
> > >> Please drop my Ack. I have never acked such patch for uboot. If I did,
> > >> it was by mistake - probably you CC-ed me for some reason.
> > >>
> > > 
> > > Some explaination, uboot introduced the concept of upstream where they
> > > "import" linux patch for dts and dt-bindings.
> > 
> > I expected OF_UPSTREAM to be taking the sources, not patches.
> > 
> > > 
> > > This and the other patch are the exact upstream patch with only the path
> > > changed so I keep all the patch commit message with tags and added the
> > > 
> > > [ upstream commit ] thing.
> > > 
> > > Hope Tom can better suggest how this should be done. You were CC
> > > probably because the git send-email included you as present in the
> > > different tags.
> > 
> > Well, Ack is still not valid because I did not Ack exactly that change.
> > It does not matter for the ack, but for Reviewed-by it would matter,
> > because it is a statement (of oversight...). I cannot control what you
> > put into patches taken out of kernel, but at least do not Cc me on that.
> 
> In specifics, yes, we should update doc/develop/devicetree/control.rst
> and maybe doc/develop/sending_patches.rst to use --suppress-cc=all for
> dts/upstream.

> But in general, what do you expect people to be doing with content from
> devicetree-rebasing? We're doing some direct cherry-picks in between
> merging of the tags. I think it would be weird to be dropping the tags
> and un-attributing peoples work.

I'm okay with my tags being kept for cherry-picked commits, since they
show up with the subtree imports anyway, right? Doing some sort of
[ upstream commit <sha> ] thing like linux-stable could be a good idea
to make it obvious what the source is.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

Re: [PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Krzysztof Kozlowski]

On 01/04/2025 16:44, Tom Rini wrote:
> On Thu, Mar 27, 2025 at 03:58:52PM +0100, Krzysztof Kozlowski wrote:
>> On 27/03/2025 15:50, Christian Marangi wrote:
>>> On Thu, Mar 27, 2025 at 03:43:47PM +0100, Krzysztof Kozlowski wrote:
>>>> On 14/03/2025 19:59, Christian Marangi wrote:
>>>>> Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
>>>>> should not be placed here. Value is derived internally in the user
>>>>> driver.
>>>>>
>>>>> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
>>>>> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
>>>> Please drop my Ack. I have never acked such patch for uboot. If I did,
>>>> it was by mistake - probably you CC-ed me for some reason.
>>>>
>>>
>>> Some explaination, uboot introduced the concept of upstream where they
>>> "import" linux patch for dts and dt-bindings.
>>
>> I expected OF_UPSTREAM to be taking the sources, not patches.
>>
>>>
>>> This and the other patch are the exact upstream patch with only the path
>>> changed so I keep all the patch commit message with tags and added the
>>>
>>> [ upstream commit ] thing.
>>>
>>> Hope Tom can better suggest how this should be done. You were CC
>>> probably because the git send-email included you as present in the
>>> different tags.
>>
>> Well, Ack is still not valid because I did not Ack exactly that change.
>> It does not matter for the ack, but for Reviewed-by it would matter,
>> because it is a statement (of oversight...). I cannot control what you
>> put into patches taken out of kernel, but at least do not Cc me on that.
> 
> In specifics, yes, we should update doc/develop/devicetree/control.rst
> and maybe doc/develop/sending_patches.rst to use --suppress-cc=all for
> dts/upstream.
> 
> But in general, what do you expect people to be doing with content from
> devicetree-rebasing? We're doing some direct cherry-picks in between
> merging of the tags. I think it would be weird to be dropping the tags
> and un-attributing peoples work.


I rather expected something like how kernel is importing dtc. You just
list the commits you get. If you want the full git history, then I would
expect simple git submodule. In both cases there will be no such patches
on the lists.

For the Ack it does not matter, but I would feel uncomfortable if people
were sending stripped and modified patches with my Rb tag.

Best regards,
Krzysztof

Re: [PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Tom Rini]

[-- Attachment #1: Type: text/plain, Size: 2921 bytes --]

On Tue, Apr 01, 2025 at 05:27:30PM +0200, Krzysztof Kozlowski wrote:
> On 01/04/2025 16:44, Tom Rini wrote:
> > On Thu, Mar 27, 2025 at 03:58:52PM +0100, Krzysztof Kozlowski wrote:
> >> On 27/03/2025 15:50, Christian Marangi wrote:
> >>> On Thu, Mar 27, 2025 at 03:43:47PM +0100, Krzysztof Kozlowski wrote:
> >>>> On 14/03/2025 19:59, Christian Marangi wrote:
> >>>>> Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
> >>>>> should not be placed here. Value is derived internally in the user
> >>>>> driver.
> >>>>>
> >>>>> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> >>>>> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
> >>>> Please drop my Ack. I have never acked such patch for uboot. If I did,
> >>>> it was by mistake - probably you CC-ed me for some reason.
> >>>>
> >>>
> >>> Some explaination, uboot introduced the concept of upstream where they
> >>> "import" linux patch for dts and dt-bindings.
> >>
> >> I expected OF_UPSTREAM to be taking the sources, not patches.
> >>
> >>>
> >>> This and the other patch are the exact upstream patch with only the path
> >>> changed so I keep all the patch commit message with tags and added the
> >>>
> >>> [ upstream commit ] thing.
> >>>
> >>> Hope Tom can better suggest how this should be done. You were CC
> >>> probably because the git send-email included you as present in the
> >>> different tags.
> >>
> >> Well, Ack is still not valid because I did not Ack exactly that change.
> >> It does not matter for the ack, but for Reviewed-by it would matter,
> >> because it is a statement (of oversight...). I cannot control what you
> >> put into patches taken out of kernel, but at least do not Cc me on that.
> > 
> > In specifics, yes, we should update doc/develop/devicetree/control.rst
> > and maybe doc/develop/sending_patches.rst to use --suppress-cc=all for
> > dts/upstream.
> > 
> > But in general, what do you expect people to be doing with content from
> > devicetree-rebasing? We're doing some direct cherry-picks in between
> > merging of the tags. I think it would be weird to be dropping the tags
> > and un-attributing peoples work.
> 
> 
> I rather expected something like how kernel is importing dtc. You just
> list the commits you get. If you want the full git history, then I would
> expect simple git submodule. In both cases there will be no such patches
> on the lists.
> 
> For the Ack it does not matter, but I would feel uncomfortable if people
> were sending stripped and modified patches with my Rb tag.

I guess I'm confused. Looking at
https://patchwork.ozlabs.org/project/uboot/patch/20250314185941.27834-6-ansuelsmth@gmail.com/
we're doing the normal thing of havig "[ upstream commit <sha>]" after
the imported log. When I merge the subtree and tag it indeed gives what
you're expecting too.

-- 
Tom

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 659 bytes --]

Re: [PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Krzysztof Kozlowski]

On 01/04/2025 18:40, Tom Rini wrote:
> On Tue, Apr 01, 2025 at 05:27:30PM +0200, Krzysztof Kozlowski wrote:
>> On 01/04/2025 16:44, Tom Rini wrote:
>>> On Thu, Mar 27, 2025 at 03:58:52PM +0100, Krzysztof Kozlowski wrote:
>>>> On 27/03/2025 15:50, Christian Marangi wrote:
>>>>> On Thu, Mar 27, 2025 at 03:43:47PM +0100, Krzysztof Kozlowski wrote:
>>>>>> On 14/03/2025 19:59, Christian Marangi wrote:
>>>>>>> Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
>>>>>>> should not be placed here. Value is derived internally in the user
>>>>>>> driver.
>>>>>>>
>>>>>>> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
>>>>>>> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
>>>>>> Please drop my Ack. I have never acked such patch for uboot. If I did,
>>>>>> it was by mistake - probably you CC-ed me for some reason.
>>>>>>
>>>>>
>>>>> Some explaination, uboot introduced the concept of upstream where they
>>>>> "import" linux patch for dts and dt-bindings.
>>>>
>>>> I expected OF_UPSTREAM to be taking the sources, not patches.
>>>>
>>>>>
>>>>> This and the other patch are the exact upstream patch with only the path
>>>>> changed so I keep all the patch commit message with tags and added the
>>>>>
>>>>> [ upstream commit ] thing.
>>>>>
>>>>> Hope Tom can better suggest how this should be done. You were CC
>>>>> probably because the git send-email included you as present in the
>>>>> different tags.
>>>>
>>>> Well, Ack is still not valid because I did not Ack exactly that change.
>>>> It does not matter for the ack, but for Reviewed-by it would matter,
>>>> because it is a statement (of oversight...). I cannot control what you
>>>> put into patches taken out of kernel, but at least do not Cc me on that.
>>>
>>> In specifics, yes, we should update doc/develop/devicetree/control.rst
>>> and maybe doc/develop/sending_patches.rst to use --suppress-cc=all for
>>> dts/upstream.
>>>
>>> But in general, what do you expect people to be doing with content from
>>> devicetree-rebasing? We're doing some direct cherry-picks in between
>>> merging of the tags. I think it would be weird to be dropping the tags
>>> and un-attributing peoples work.
>>
>>
>> I rather expected something like how kernel is importing dtc. You just
>> list the commits you get. If you want the full git history, then I would
>> expect simple git submodule. In both cases there will be no such patches
>> on the lists.
>>
>> For the Ack it does not matter, but I would feel uncomfortable if people
>> were sending stripped and modified patches with my Rb tag.
> 
> I guess I'm confused. Looking at
> https://patchwork.ozlabs.org/project/uboot/patch/20250314185941.27834-6-ansuelsmth@gmail.com/
> we're doing the normal thing of havig "[ upstream commit <sha>]" after
> the imported log. When I merge the subtree and tag it indeed gives what
> you're expecting too.

When you merge subtree, the patch is not modified and it lives in
separate repo. No one sends them over lists, no one modifies them.
Unlike here (even if modification did not happen, person was touching it
so how can anyone be sure? That's not a scripted process).

Best regards,
Krzysztof

Re: [PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Tom Rini]

[-- Attachment #1: Type: text/plain, Size: 3671 bytes --]

On Tue, Apr 01, 2025 at 07:28:36PM +0200, Krzysztof Kozlowski wrote:
> On 01/04/2025 18:40, Tom Rini wrote:
> > On Tue, Apr 01, 2025 at 05:27:30PM +0200, Krzysztof Kozlowski wrote:
> >> On 01/04/2025 16:44, Tom Rini wrote:
> >>> On Thu, Mar 27, 2025 at 03:58:52PM +0100, Krzysztof Kozlowski wrote:
> >>>> On 27/03/2025 15:50, Christian Marangi wrote:
> >>>>> On Thu, Mar 27, 2025 at 03:43:47PM +0100, Krzysztof Kozlowski wrote:
> >>>>>> On 14/03/2025 19:59, Christian Marangi wrote:
> >>>>>>> Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
> >>>>>>> should not be placed here. Value is derived internally in the user
> >>>>>>> driver.
> >>>>>>>
> >>>>>>> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> >>>>>>> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
> >>>>>> Please drop my Ack. I have never acked such patch for uboot. If I did,
> >>>>>> it was by mistake - probably you CC-ed me for some reason.
> >>>>>>
> >>>>>
> >>>>> Some explaination, uboot introduced the concept of upstream where they
> >>>>> "import" linux patch for dts and dt-bindings.
> >>>>
> >>>> I expected OF_UPSTREAM to be taking the sources, not patches.
> >>>>
> >>>>>
> >>>>> This and the other patch are the exact upstream patch with only the path
> >>>>> changed so I keep all the patch commit message with tags and added the
> >>>>>
> >>>>> [ upstream commit ] thing.
> >>>>>
> >>>>> Hope Tom can better suggest how this should be done. You were CC
> >>>>> probably because the git send-email included you as present in the
> >>>>> different tags.
> >>>>
> >>>> Well, Ack is still not valid because I did not Ack exactly that change.
> >>>> It does not matter for the ack, but for Reviewed-by it would matter,
> >>>> because it is a statement (of oversight...). I cannot control what you
> >>>> put into patches taken out of kernel, but at least do not Cc me on that.
> >>>
> >>> In specifics, yes, we should update doc/develop/devicetree/control.rst
> >>> and maybe doc/develop/sending_patches.rst to use --suppress-cc=all for
> >>> dts/upstream.
> >>>
> >>> But in general, what do you expect people to be doing with content from
> >>> devicetree-rebasing? We're doing some direct cherry-picks in between
> >>> merging of the tags. I think it would be weird to be dropping the tags
> >>> and un-attributing peoples work.
> >>
> >>
> >> I rather expected something like how kernel is importing dtc. You just
> >> list the commits you get. If you want the full git history, then I would
> >> expect simple git submodule. In both cases there will be no such patches
> >> on the lists.
> >>
> >> For the Ack it does not matter, but I would feel uncomfortable if people
> >> were sending stripped and modified patches with my Rb tag.
> > 
> > I guess I'm confused. Looking at
> > https://patchwork.ozlabs.org/project/uboot/patch/20250314185941.27834-6-ansuelsmth@gmail.com/
> > we're doing the normal thing of havig "[ upstream commit <sha>]" after
> > the imported log. When I merge the subtree and tag it indeed gives what
> > you're expecting too.
> 
> When you merge subtree, the patch is not modified and it lives in
> separate repo. No one sends them over lists, no one modifies them.
> Unlike here (even if modification did not happen, person was touching it
> so how can anyone be sure? That's not a scripted process).

We merge the subtree on tags, and people cherry-pick commits in between
tags when needed. This is a case of the latter, which is why it says "[
upstream commit <sha> ]" in the commit message, which is the usual case.

-- 
Tom

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 659 bytes --]

Re: [PATCH v2 0/6] airoha: Add initial support AN7581

[Tom Rini]

On Fri, 14 Mar 2025 19:59:20 +0100, Christian Marangi wrote:

> This little series adds initial support for Airoha AN7581 SoC.
> 
> With the help of some backport patch, this use OF_UPSTREAM
> directly.
> 
> Posting this to have the targer and the very basic driver.
> 
> [...]

Applied to u-boot/next, thanks!

[1/6] airoha: Add initial support for Airoha AN7581 SoC
      commit: 793e327e2ca947a293c0eb5beb5eef27744385f0
[2/6] clk: airoha: Add support for Airoha AN7581 SoC clock
      commit: d0b81afb5ec9d2d8a3f11bfc827f2f24c32d47ec
[3/6] reset: airoha: Add driver for controlling reset line of AN7581
      commit: b02b8b767624a859a2776c11822119392702bbab
[4/6] arm64: dts: airoha: en7581: Add Clock Controller node
      commit: 83972387a52712acf98e4827c0d09b72ea970088
[5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581
      commit: ea0d64adc67d270de27a092d7ff7bfce8780fefb
[6/6] dt-bindings: clock: add ID for eMMC for EN7581
      commit: e00a318d64724a8fa1ed1ad56cf54edc1c295f3d
-- 
Tom

Re: [PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Sumit Garg]

Sorry for jumping in late in this thread as it is still using my old
Linaro email ID which should be disabled by now.

On Tue, Apr 01, 2025 at 12:02:17PM -0600, Tom Rini wrote:
> On Tue, Apr 01, 2025 at 07:28:36PM +0200, Krzysztof Kozlowski wrote:
> > On 01/04/2025 18:40, Tom Rini wrote:
> > > On Tue, Apr 01, 2025 at 05:27:30PM +0200, Krzysztof Kozlowski wrote:
> > >> On 01/04/2025 16:44, Tom Rini wrote:
> > >>> On Thu, Mar 27, 2025 at 03:58:52PM +0100, Krzysztof Kozlowski wrote:
> > >>>> On 27/03/2025 15:50, Christian Marangi wrote:
> > >>>>> On Thu, Mar 27, 2025 at 03:43:47PM +0100, Krzysztof Kozlowski wrote:
> > >>>>>> On 14/03/2025 19:59, Christian Marangi wrote:
> > >>>>>>> Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
> > >>>>>>> should not be placed here. Value is derived internally in the user
> > >>>>>>> driver.
> > >>>>>>>
> > >>>>>>> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> > >>>>>>> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
> > >>>>>> Please drop my Ack. I have never acked such patch for uboot. If I did,
> > >>>>>> it was by mistake - probably you CC-ed me for some reason.
> > >>>>>>
> > >>>>>
> > >>>>> Some explaination, uboot introduced the concept of upstream where they
> > >>>>> "import" linux patch for dts and dt-bindings.
> > >>>>
> > >>>> I expected OF_UPSTREAM to be taking the sources, not patches.
> > >>>>
> > >>>>>
> > >>>>> This and the other patch are the exact upstream patch with only the path
> > >>>>> changed so I keep all the patch commit message with tags and added the
> > >>>>>
> > >>>>> [ upstream commit ] thing.
> > >>>>>
> > >>>>> Hope Tom can better suggest how this should be done. You were CC
> > >>>>> probably because the git send-email included you as present in the
> > >>>>> different tags.
> > >>>>
> > >>>> Well, Ack is still not valid because I did not Ack exactly that change.
> > >>>> It does not matter for the ack, but for Reviewed-by it would matter,
> > >>>> because it is a statement (of oversight...). I cannot control what you
> > >>>> put into patches taken out of kernel, but at least do not Cc me on that.
> > >>>
> > >>> In specifics, yes, we should update doc/develop/devicetree/control.rst
> > >>> and maybe doc/develop/sending_patches.rst to use --suppress-cc=all for
> > >>> dts/upstream.
> > >>>
> > >>> But in general, what do you expect people to be doing with content from
> > >>> devicetree-rebasing? We're doing some direct cherry-picks in between
> > >>> merging of the tags. I think it would be weird to be dropping the tags
> > >>> and un-attributing peoples work.
> > >>
> > >>
> > >> I rather expected something like how kernel is importing dtc. You just
> > >> list the commits you get. If you want the full git history, then I would
> > >> expect simple git submodule. In both cases there will be no such patches
> > >> on the lists.
> > >>
> > >> For the Ack it does not matter, but I would feel uncomfortable if people
> > >> were sending stripped and modified patches with my Rb tag.
> > > 
> > > I guess I'm confused. Looking at
> > > https://patchwork.ozlabs.org/project/uboot/patch/20250314185941.27834-6-ansuelsmth@gmail.com/
> > > we're doing the normal thing of havig "[ upstream commit <sha>]" after
> > > the imported log. When I merge the subtree and tag it indeed gives what
> > > you're expecting too.
> > 
> > When you merge subtree, the patch is not modified and it lives in
> > separate repo. No one sends them over lists, no one modifies them.
> > Unlike here (even if modification did not happen, person was touching it
> > so how can anyone be sure? That's not a scripted process).
> 
> We merge the subtree on tags, and people cherry-pick commits in between
> tags when needed. This is a case of the latter, which is why it says "[
> upstream commit <sha> ]" in the commit message, which is the usual case.

Although we have tooling to pick patches from devicetree-rebasing tree
but I can see Krzysztof's concerns here. We can't be sure if developer
has touched the cherry picked patch or not but I suppose there would be
similar concerns for the stable backports for Linux too. So IMHO, it's
really upto maintainer applying those cherry-picked patches to see if
there is any difference from upstream.

However, there is an additional process change what we can do here is
for the developer to list just the commit IDs for the patches to be
cherry picked in dts/upstream in the cover letter. This way the
maintainer can just directly use the tooling to cherry pick those
patches before applying the patch-set.

Tooling already available for cherry-picking subtree commits as:

$ ./tools/update-subtree.sh pick dts <devicetree-rebasing-commit-id>

-Sumit

Re: [PATCH v2 5/6] dt-bindings: clock: drop NUM_CLOCKS define for EN7581

[Tom Rini]

[-- Attachment #1: Type: text/plain, Size: 5392 bytes --]

On Wed, Apr 02, 2025 at 12:42:54PM +0530, Sumit Garg wrote:
> Sorry for jumping in late in this thread as it is still using my old
> Linaro email ID which should be disabled by now.
> 
> On Tue, Apr 01, 2025 at 12:02:17PM -0600, Tom Rini wrote:
> > On Tue, Apr 01, 2025 at 07:28:36PM +0200, Krzysztof Kozlowski wrote:
> > > On 01/04/2025 18:40, Tom Rini wrote:
> > > > On Tue, Apr 01, 2025 at 05:27:30PM +0200, Krzysztof Kozlowski wrote:
> > > >> On 01/04/2025 16:44, Tom Rini wrote:
> > > >>> On Thu, Mar 27, 2025 at 03:58:52PM +0100, Krzysztof Kozlowski wrote:
> > > >>>> On 27/03/2025 15:50, Christian Marangi wrote:
> > > >>>>> On Thu, Mar 27, 2025 at 03:43:47PM +0100, Krzysztof Kozlowski wrote:
> > > >>>>>> On 14/03/2025 19:59, Christian Marangi wrote:
> > > >>>>>>> Drop NUM_CLOCKS define for EN7581 dts/upstream/src/include. This is not a binding and
> > > >>>>>>> should not be placed here. Value is derived internally in the user
> > > >>>>>>> driver.
> > > >>>>>>>
> > > >>>>>>> Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
> > > >>>>>>> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
> > > >>>>>> Please drop my Ack. I have never acked such patch for uboot. If I did,
> > > >>>>>> it was by mistake - probably you CC-ed me for some reason.
> > > >>>>>>
> > > >>>>>
> > > >>>>> Some explaination, uboot introduced the concept of upstream where they
> > > >>>>> "import" linux patch for dts and dt-bindings.
> > > >>>>
> > > >>>> I expected OF_UPSTREAM to be taking the sources, not patches.
> > > >>>>
> > > >>>>>
> > > >>>>> This and the other patch are the exact upstream patch with only the path
> > > >>>>> changed so I keep all the patch commit message with tags and added the
> > > >>>>>
> > > >>>>> [ upstream commit ] thing.
> > > >>>>>
> > > >>>>> Hope Tom can better suggest how this should be done. You were CC
> > > >>>>> probably because the git send-email included you as present in the
> > > >>>>> different tags.
> > > >>>>
> > > >>>> Well, Ack is still not valid because I did not Ack exactly that change.
> > > >>>> It does not matter for the ack, but for Reviewed-by it would matter,
> > > >>>> because it is a statement (of oversight...). I cannot control what you
> > > >>>> put into patches taken out of kernel, but at least do not Cc me on that.
> > > >>>
> > > >>> In specifics, yes, we should update doc/develop/devicetree/control.rst
> > > >>> and maybe doc/develop/sending_patches.rst to use --suppress-cc=all for
> > > >>> dts/upstream.
> > > >>>
> > > >>> But in general, what do you expect people to be doing with content from
> > > >>> devicetree-rebasing? We're doing some direct cherry-picks in between
> > > >>> merging of the tags. I think it would be weird to be dropping the tags
> > > >>> and un-attributing peoples work.
> > > >>
> > > >>
> > > >> I rather expected something like how kernel is importing dtc. You just
> > > >> list the commits you get. If you want the full git history, then I would
> > > >> expect simple git submodule. In both cases there will be no such patches
> > > >> on the lists.
> > > >>
> > > >> For the Ack it does not matter, but I would feel uncomfortable if people
> > > >> were sending stripped and modified patches with my Rb tag.
> > > > 
> > > > I guess I'm confused. Looking at
> > > > https://patchwork.ozlabs.org/project/uboot/patch/20250314185941.27834-6-ansuelsmth@gmail.com/
> > > > we're doing the normal thing of havig "[ upstream commit <sha>]" after
> > > > the imported log. When I merge the subtree and tag it indeed gives what
> > > > you're expecting too.
> > > 
> > > When you merge subtree, the patch is not modified and it lives in
> > > separate repo. No one sends them over lists, no one modifies them.
> > > Unlike here (even if modification did not happen, person was touching it
> > > so how can anyone be sure? That's not a scripted process).
> > 
> > We merge the subtree on tags, and people cherry-pick commits in between
> > tags when needed. This is a case of the latter, which is why it says "[
> > upstream commit <sha> ]" in the commit message, which is the usual case.
> 
> Although we have tooling to pick patches from devicetree-rebasing tree
> but I can see Krzysztof's concerns here. We can't be sure if developer
> has touched the cherry picked patch or not but I suppose there would be
> similar concerns for the stable backports for Linux too. So IMHO, it's
> really upto maintainer applying those cherry-picked patches to see if
> there is any difference from upstream.
> 
> However, there is an additional process change what we can do here is
> for the developer to list just the commit IDs for the patches to be
> cherry picked in dts/upstream in the cover letter. This way the
> maintainer can just directly use the tooling to cherry pick those
> patches before applying the patch-set.
> 
> Tooling already available for cherry-picking subtree commits as:
> 
> $ ./tools/update-subtree.sh pick dts <devicetree-rebasing-commit-id>

For U-Boot at least, at the end of the day the cherry-picks are a
temporary measure. We do pick up and merge each release tag in turn (and
so yeah, I should grab the newest one already) and if the merge doesn't
Just Work, I drop in the correct file from upstream to resolve the merge
conflict.

-- 
Tom

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 659 bytes --]