* [PATCH v2 0/3] Add read-only SpacemiT K3 eFuse support @ 2026-11-16 7:54 ` Jiayu Du 0 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-16 7:54 UTC (permalink / raw) To: srini Cc: robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel, Jiayu Du This series adds read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse controller. These banks contain temperature calibration parameters and other configuration data. Tested on a SpacemiT K3 COM260 Kit: shadow reload completed successfully, and reads of both banks matched v1. Repeated reads and reads crossing bank/keepout boundaries returned the expected data. Changes in v2: - Trigger a shadow reload before polling FUSE_READY; fail probe on timeout. - Enable the clock only while accessing hardware and leave the shared reset deasserted on teardown. - Explain why the root key and SSK registers are excluded. - Use an aligned __le32 cache instead of put_unaligned_le32(). - Drop unnecessary Kconfig dependencies and the bootloader initialization requirement. - Use devm_platform_ioremap_resource() and align the register definitions. - Enable the eFuse node by default in k3.dtsi. - Add Conor's Acked-by to the binding patch. v1: https://lore.kernel.org/all/20261106060525.346873-1-duge1256@gmail.com/ Jiayu Du (3): dt-bindings: nvmem: Add SpacemiT K3 eFuse controller nvmem: Add read-only SpacemiT K3 eFuse support riscv: dts: spacemit: Add K3 eFuse and temperature trim cells .../bindings/nvmem/spacemit,k3-efuse.yaml | 63 ++++++ MAINTAINERS | 6 + arch/riscv/boot/dts/spacemit/k3.dtsi | 23 ++ drivers/nvmem/Kconfig | 11 + drivers/nvmem/Makefile | 1 + drivers/nvmem/spacemit-k3-efuse.c | 199 ++++++++++++++++++ 6 files changed, 303 insertions(+) create mode 100644 Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml create mode 100644 drivers/nvmem/spacemit-k3-efuse.c base-commit: d6bcfc94eb4da6e089c5473970cd96727292fd97 -- 2.43.0 ^ permalink raw reply [flat|nested] 18+ messages in thread
* [PATCH v2 0/3] Add read-only SpacemiT K3 eFuse support @ 2026-11-16 7:54 ` Jiayu Du 0 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-16 7:54 UTC (permalink / raw) To: srini Cc: robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel, Jiayu Du This series adds read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse controller. These banks contain temperature calibration parameters and other configuration data. Tested on a SpacemiT K3 COM260 Kit: shadow reload completed successfully, and reads of both banks matched v1. Repeated reads and reads crossing bank/keepout boundaries returned the expected data. Changes in v2: - Trigger a shadow reload before polling FUSE_READY; fail probe on timeout. - Enable the clock only while accessing hardware and leave the shared reset deasserted on teardown. - Explain why the root key and SSK registers are excluded. - Use an aligned __le32 cache instead of put_unaligned_le32(). - Drop unnecessary Kconfig dependencies and the bootloader initialization requirement. - Use devm_platform_ioremap_resource() and align the register definitions. - Enable the eFuse node by default in k3.dtsi. - Add Conor's Acked-by to the binding patch. v1: https://lore.kernel.org/all/20261106060525.346873-1-duge1256@gmail.com/ Jiayu Du (3): dt-bindings: nvmem: Add SpacemiT K3 eFuse controller nvmem: Add read-only SpacemiT K3 eFuse support riscv: dts: spacemit: Add K3 eFuse and temperature trim cells .../bindings/nvmem/spacemit,k3-efuse.yaml | 63 ++++++ MAINTAINERS | 6 + arch/riscv/boot/dts/spacemit/k3.dtsi | 23 ++ drivers/nvmem/Kconfig | 11 + drivers/nvmem/Makefile | 1 + drivers/nvmem/spacemit-k3-efuse.c | 199 ++++++++++++++++++ 6 files changed, 303 insertions(+) create mode 100644 Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml create mode 100644 drivers/nvmem/spacemit-k3-efuse.c base-commit: d6bcfc94eb4da6e089c5473970cd96727292fd97 -- 2.43.0 _______________________________________________ linux-riscv mailing list linux-riscv@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-riscv ^ permalink raw reply [flat|nested] 18+ messages in thread
* [PATCH v2 1/3] dt-bindings: nvmem: Add SpacemiT K3 eFuse controller 2026-11-16 7:54 ` Jiayu Du @ 2026-11-16 7:54 ` Jiayu Du -1 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-16 7:54 UTC (permalink / raw) To: srini Cc: robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel, Jiayu Du, Conor Dooley Describe the SpacemiT K3 eFuse controller and its readable bank 0 and bank 7 shadow windows, which contain temperature calibration parameters and other configuration data. Define the AES_WTM clock and reset, little-endian data and controller-relative NVMEM offsets. Use fixed-layout cells for the temperature calibration parameters. Acked-by: Conor Dooley <conor.dooley@microchip.com> Signed-off-by: Jiayu Du <duge1256@gmail.com> --- .../bindings/nvmem/spacemit,k3-efuse.yaml | 63 +++++++++++++++++++ MAINTAINERS | 5 ++ 2 files changed, 68 insertions(+) create mode 100644 Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml diff --git a/Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml b/Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml new file mode 100644 index 000000000000..93f4301c8289 --- /dev/null +++ b/Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml @@ -0,0 +1,63 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/nvmem/spacemit,k3-efuse.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: SpacemiT K3 eFuse controller + +maintainers: + - Jiayu Du <duge1256@gmail.com> + +description: + The K3 eFuse controller provides configuration and calibration data through + shadow registers. NVMEM uses little-endian data and controller-relative + offsets. Read access is limited to bank 0 (0x104-0x123) and bank 7 (0x190-0x1af). + +allOf: + - $ref: nvmem.yaml# + +properties: + compatible: + const: spacemit,k3-efuse + + reg: + maxItems: 1 + description: The 0x400-byte eFuse controller register block. + + clocks: + maxItems: 1 + + resets: + maxItems: 1 + +required: + - compatible + - reg + - clocks + - resets + +unevaluatedProperties: false + +examples: + - | + #include <dt-bindings/clock/spacemit,k3-clocks.h> + #include <dt-bindings/reset/spacemit,k3-resets.h> + + efuse@f0702800 { + compatible = "spacemit,k3-efuse"; + reg = <0xf0702800 0x400>; + clocks = <&syscon_apmu CLK_APMU_AES_WTM>; + resets = <&syscon_apmu RESET_APMU_AES_WTM>; + + nvmem-layout { + compatible = "fixed-layout"; + #address-cells = <1>; + #size-cells = <1>; + + temperature-trim@19a { + reg = <0x19a 0x2>; + bits = <4 8>; + }; + }; + }; diff --git a/MAINTAINERS b/MAINTAINERS index 46c416cadacd..0a9eb764ffee 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -26525,6 +26525,11 @@ S: Maintained F: Documentation/devicetree/bindings/sound/spacemit,k1-i2s.yaml F: sound/soc/spacemit/k1_i2s.c +SPACEMIT K3 EFUSE DRIVER +M: Jiayu Du <duge1256@gmail.com> +S: Maintained +F: Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml + SPACEMIT RESET CONTROLLER DRIVER M: Guodong Xu <docular.xu@gmail.com> L: linux-riscv@lists.infradead.org -- 2.43.0 ^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH v2 1/3] dt-bindings: nvmem: Add SpacemiT K3 eFuse controller @ 2026-11-16 7:54 ` Jiayu Du 0 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-16 7:54 UTC (permalink / raw) To: srini Cc: robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel, Jiayu Du, Conor Dooley Describe the SpacemiT K3 eFuse controller and its readable bank 0 and bank 7 shadow windows, which contain temperature calibration parameters and other configuration data. Define the AES_WTM clock and reset, little-endian data and controller-relative NVMEM offsets. Use fixed-layout cells for the temperature calibration parameters. Acked-by: Conor Dooley <conor.dooley@microchip.com> Signed-off-by: Jiayu Du <duge1256@gmail.com> --- .../bindings/nvmem/spacemit,k3-efuse.yaml | 63 +++++++++++++++++++ MAINTAINERS | 5 ++ 2 files changed, 68 insertions(+) create mode 100644 Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml diff --git a/Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml b/Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml new file mode 100644 index 000000000000..93f4301c8289 --- /dev/null +++ b/Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml @@ -0,0 +1,63 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/nvmem/spacemit,k3-efuse.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: SpacemiT K3 eFuse controller + +maintainers: + - Jiayu Du <duge1256@gmail.com> + +description: + The K3 eFuse controller provides configuration and calibration data through + shadow registers. NVMEM uses little-endian data and controller-relative + offsets. Read access is limited to bank 0 (0x104-0x123) and bank 7 (0x190-0x1af). + +allOf: + - $ref: nvmem.yaml# + +properties: + compatible: + const: spacemit,k3-efuse + + reg: + maxItems: 1 + description: The 0x400-byte eFuse controller register block. + + clocks: + maxItems: 1 + + resets: + maxItems: 1 + +required: + - compatible + - reg + - clocks + - resets + +unevaluatedProperties: false + +examples: + - | + #include <dt-bindings/clock/spacemit,k3-clocks.h> + #include <dt-bindings/reset/spacemit,k3-resets.h> + + efuse@f0702800 { + compatible = "spacemit,k3-efuse"; + reg = <0xf0702800 0x400>; + clocks = <&syscon_apmu CLK_APMU_AES_WTM>; + resets = <&syscon_apmu RESET_APMU_AES_WTM>; + + nvmem-layout { + compatible = "fixed-layout"; + #address-cells = <1>; + #size-cells = <1>; + + temperature-trim@19a { + reg = <0x19a 0x2>; + bits = <4 8>; + }; + }; + }; diff --git a/MAINTAINERS b/MAINTAINERS index 46c416cadacd..0a9eb764ffee 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -26525,6 +26525,11 @@ S: Maintained F: Documentation/devicetree/bindings/sound/spacemit,k1-i2s.yaml F: sound/soc/spacemit/k1_i2s.c +SPACEMIT K3 EFUSE DRIVER +M: Jiayu Du <duge1256@gmail.com> +S: Maintained +F: Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml + SPACEMIT RESET CONTROLLER DRIVER M: Guodong Xu <docular.xu@gmail.com> L: linux-riscv@lists.infradead.org -- 2.43.0 _______________________________________________ linux-riscv mailing list linux-riscv@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-riscv ^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH v2 2/3] nvmem: Add read-only SpacemiT K3 eFuse support 2026-11-16 7:54 ` Jiayu Du @ 2026-11-16 7:54 ` Jiayu Du -1 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-16 7:54 UTC (permalink / raw) To: srini Cc: robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel, Jiayu Du Add read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse controller. These banks contain temperature calibration parameters and other configuration data. Keep the root key and SSK registers outside the exposed bank ranges. FUSE_READY reports completion of a software-triggered reload. Trigger a reload through GEU_CONFIG bit 22 with a 1 us reset pulse, then poll FUSE_READY at a 1 us interval with a 10 ms timeout before taking the snapshot. Fail probe if the reload times out. Snapshot the shadow registers with aligned 32-bit reads and expose the data in little-endian order at controller-relative offsets. Serve reads from the snapshot and use keepouts for the remaining register ranges. Enable the AES_WTM clock only while accessing the hardware. Leave its shared reset deasserted on teardown because WTM_RST resets the entire GEU. Signed-off-by: Jiayu Du <duge1256@gmail.com> --- MAINTAINERS | 1 + drivers/nvmem/Kconfig | 11 ++ drivers/nvmem/Makefile | 1 + drivers/nvmem/spacemit-k3-efuse.c | 199 ++++++++++++++++++++++++++++++ 4 files changed, 212 insertions(+) create mode 100644 drivers/nvmem/spacemit-k3-efuse.c diff --git a/MAINTAINERS b/MAINTAINERS index 0a9eb764ffee..e0af83ecee37 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -26529,6 +26529,7 @@ SPACEMIT K3 EFUSE DRIVER M: Jiayu Du <duge1256@gmail.com> S: Maintained F: Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml +F: drivers/nvmem/spacemit-k3-efuse.c SPACEMIT RESET CONTROLLER DRIVER M: Guodong Xu <docular.xu@gmail.com> diff --git a/drivers/nvmem/Kconfig b/drivers/nvmem/Kconfig index d2f5bb236663..4a8f975b35c7 100644 --- a/drivers/nvmem/Kconfig +++ b/drivers/nvmem/Kconfig @@ -476,6 +476,17 @@ config NVMEM_SNVS_LPGPR This driver can also be built as a module. If so, the module will be called nvmem-snvs-lpgpr. +config NVMEM_SPACEMIT_K3_EFUSE + tristate "SpacemiT K3 eFuse support" + depends on ARCH_SPACEMIT || COMPILE_TEST + depends on OF && HAS_IOMEM + help + Say Y here to read configuration and calibration data from the + SpacemiT K3 eFuse shadow banks. Programming the fuses is not supported. + + This driver can also be built as a module. If so, the module will + be called spacemit-k3-efuse. + config NVMEM_SPMI_SDAM tristate "SPMI SDAM Support" depends on SPMI diff --git a/drivers/nvmem/Makefile b/drivers/nvmem/Makefile index 5e70bad9f0f9..d29716957faf 100644 --- a/drivers/nvmem/Makefile +++ b/drivers/nvmem/Makefile @@ -81,6 +81,7 @@ obj-$(CONFIG_NVMEM_SC27XX_EFUSE) += nvmem-sc27xx-efuse.o nvmem-sc27xx-efuse-y := sc27xx-efuse.o obj-$(CONFIG_NVMEM_SNVS_LPGPR) += nvmem_snvs_lpgpr.o nvmem_snvs_lpgpr-y := snvs_lpgpr.o +obj-$(CONFIG_NVMEM_SPACEMIT_K3_EFUSE) += spacemit-k3-efuse.o obj-$(CONFIG_NVMEM_SPMI_SDAM) += nvmem_qcom-spmi-sdam.o nvmem_qcom-spmi-sdam-y += qcom-spmi-sdam.o obj-$(CONFIG_NVMEM_SPRD_EFUSE) += nvmem_sprd_efuse.o diff --git a/drivers/nvmem/spacemit-k3-efuse.c b/drivers/nvmem/spacemit-k3-efuse.c new file mode 100644 index 000000000000..e126902130d4 --- /dev/null +++ b/drivers/nvmem/spacemit-k3-efuse.c @@ -0,0 +1,199 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * SpacemiT K3 read-only eFuse driver + * + * Copyright (c) 2026 Spacemit Co. Ltd. + */ + +#include <linux/array_size.h> +#include <linux/bits.h> +#include <linux/clk.h> +#include <linux/delay.h> +#include <linux/device.h> +#include <linux/err.h> +#include <linux/errno.h> +#include <linux/io.h> +#include <linux/iopoll.h> +#include <linux/module.h> +#include <linux/nvmem-provider.h> +#include <linux/of.h> +#include <linux/platform_device.h> +#include <linux/reset.h> +#include <linux/slab.h> +#include <linux/string.h> + +#include <asm/byteorder.h> + +#define K3_EFUSE_REG_SIZE 0x400 +#define K3_EFUSE_BANK_SIZE 0x20 +#define K3_EFUSE_CONFIG 0x004 +#define K3_EFUSE_SOFTWARE_RESET BIT(22) +#define K3_EFUSE_BANK0_OFFSET 0x104 +#define K3_EFUSE_STATUS 0x184 +#define K3_EFUSE_READY BIT(1) +#define K3_EFUSE_BANK7_OFFSET 0x190 +#define K3_EFUSE_RESET_DELAY_US 1 +#define K3_EFUSE_POLL_INTERVAL_US 1 +#define K3_EFUSE_POLL_TIMEOUT_US 10000 + +static const unsigned int k3_efuse_bank_offsets[] = { + K3_EFUSE_BANK0_OFFSET, + K3_EFUSE_BANK7_OFFSET, +}; + +/* Prevent NVMEM reads from exposing the root key and SSK registers. */ +static const struct nvmem_keepout k3_efuse_keepouts[] = { + { + .start = 0, + .end = K3_EFUSE_BANK0_OFFSET, + }, { + .start = K3_EFUSE_BANK0_OFFSET + K3_EFUSE_BANK_SIZE, + .end = K3_EFUSE_BANK7_OFFSET, + }, { + .start = K3_EFUSE_BANK7_OFFSET + K3_EFUSE_BANK_SIZE, + .end = K3_EFUSE_REG_SIZE, + }, +}; + +struct k3_efuse { + __le32 data[ARRAY_SIZE(k3_efuse_bank_offsets)][K3_EFUSE_BANK_SIZE / sizeof(u32)]; +}; + +static int k3_efuse_read(void *context, unsigned int offset, void *val, + size_t bytes) +{ + struct k3_efuse *efuse = context; + unsigned int bank, index; + + for (bank = 0; bank < ARRAY_SIZE(k3_efuse_bank_offsets); bank++) { + if (offset < k3_efuse_bank_offsets[bank]) + continue; + + index = offset - k3_efuse_bank_offsets[bank]; + if (index >= K3_EFUSE_BANK_SIZE || bytes > K3_EFUSE_BANK_SIZE - index) + continue; + + memcpy(val, (u8 *)efuse->data[bank] + index, bytes); + return 0; + } + + return -EINVAL; +} + +static int k3_efuse_reload(void __iomem *base) +{ + u32 val; + + val = readl(base + K3_EFUSE_CONFIG); + writel(val | K3_EFUSE_SOFTWARE_RESET, base + K3_EFUSE_CONFIG); + readl(base + K3_EFUSE_CONFIG); + udelay(K3_EFUSE_RESET_DELAY_US); + val = readl(base + K3_EFUSE_CONFIG); + writel(val & ~K3_EFUSE_SOFTWARE_RESET, base + K3_EFUSE_CONFIG); + + return readl_poll_timeout(base + K3_EFUSE_STATUS, val, + val & K3_EFUSE_READY, + K3_EFUSE_POLL_INTERVAL_US, + K3_EFUSE_POLL_TIMEOUT_US); +} + +static int k3_efuse_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct nvmem_config config = { + .dev = dev, + .name = dev_name(dev), + .id = NVMEM_DEVID_NONE, + .owner = THIS_MODULE, + .type = NVMEM_TYPE_OTP, + .read_only = true, + .root_only = true, + .reg_read = k3_efuse_read, + .word_size = 1, + .stride = 1, + .size = K3_EFUSE_REG_SIZE, + .keepout = k3_efuse_keepouts, + .nkeepout = ARRAY_SIZE(k3_efuse_keepouts), + }; + struct reset_control *reset; + struct nvmem_device *nvmem; + struct k3_efuse *efuse; + void __iomem *base; + struct clk *clk; + unsigned int bank, i; + int ret; + + base = devm_platform_ioremap_resource(pdev, 0); + if (IS_ERR(base)) + return PTR_ERR(base); + + efuse = devm_kzalloc(dev, sizeof(*efuse), GFP_KERNEL); + if (!efuse) + return -ENOMEM; + + clk = devm_clk_get(dev, NULL); + if (IS_ERR(clk)) + return dev_err_probe(dev, PTR_ERR(clk), "Failed to get clock\n"); + + reset = devm_reset_control_get_shared(dev, NULL); + if (IS_ERR(reset)) + return dev_err_probe(dev, PTR_ERR(reset), "Failed to get reset\n"); + + ret = clk_prepare_enable(clk); + if (ret) + return dev_err_probe(dev, ret, "Failed to enable clock\n"); + + /* WTM_RST resets the entire GEU; leave it deasserted on teardown. */ + ret = reset_control_deassert(reset); + if (ret) { + dev_err_probe(dev, ret, "Failed to deassert reset\n"); + goto disable_clk; + } + + ret = k3_efuse_reload(base); + if (ret) { + dev_err_probe(dev, ret, "Timed out waiting for shadow reload\n"); + goto disable_clk; + } + + for (bank = 0; bank < ARRAY_SIZE(k3_efuse_bank_offsets); bank++) { + for (i = 0; i < ARRAY_SIZE(efuse->data[bank]); i++) { + u32 val = readl(base + k3_efuse_bank_offsets[bank] + i * sizeof(u32)); + + efuse->data[bank][i] = cpu_to_le32(val); + } + } + + clk_disable_unprepare(clk); + + config.priv = efuse; + + nvmem = devm_nvmem_register(dev, &config); + if (IS_ERR(nvmem)) + return dev_err_probe(dev, PTR_ERR(nvmem), "Failed to register NVMEM\n"); + + return 0; + +disable_clk: + clk_disable_unprepare(clk); + return ret; +} + +static const struct of_device_id k3_efuse_of_match[] = { + { .compatible = "spacemit,k3-efuse" }, + { } +}; +MODULE_DEVICE_TABLE(of, k3_efuse_of_match); + +static struct platform_driver k3_efuse_driver = { + .probe = k3_efuse_probe, + .driver = { + .name = "spacemit-k3-efuse", + .of_match_table = k3_efuse_of_match, + }, +}; +module_platform_driver(k3_efuse_driver); + +MODULE_AUTHOR("Jiayu Du <duge1256@gmail.com>"); +MODULE_DESCRIPTION("SpacemiT K3 read-only eFuse driver"); +MODULE_LICENSE("GPL"); -- 2.43.0 ^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH v2 2/3] nvmem: Add read-only SpacemiT K3 eFuse support @ 2026-11-16 7:54 ` Jiayu Du 0 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-16 7:54 UTC (permalink / raw) To: srini Cc: robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel, Jiayu Du Add read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse controller. These banks contain temperature calibration parameters and other configuration data. Keep the root key and SSK registers outside the exposed bank ranges. FUSE_READY reports completion of a software-triggered reload. Trigger a reload through GEU_CONFIG bit 22 with a 1 us reset pulse, then poll FUSE_READY at a 1 us interval with a 10 ms timeout before taking the snapshot. Fail probe if the reload times out. Snapshot the shadow registers with aligned 32-bit reads and expose the data in little-endian order at controller-relative offsets. Serve reads from the snapshot and use keepouts for the remaining register ranges. Enable the AES_WTM clock only while accessing the hardware. Leave its shared reset deasserted on teardown because WTM_RST resets the entire GEU. Signed-off-by: Jiayu Du <duge1256@gmail.com> --- MAINTAINERS | 1 + drivers/nvmem/Kconfig | 11 ++ drivers/nvmem/Makefile | 1 + drivers/nvmem/spacemit-k3-efuse.c | 199 ++++++++++++++++++++++++++++++ 4 files changed, 212 insertions(+) create mode 100644 drivers/nvmem/spacemit-k3-efuse.c diff --git a/MAINTAINERS b/MAINTAINERS index 0a9eb764ffee..e0af83ecee37 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -26529,6 +26529,7 @@ SPACEMIT K3 EFUSE DRIVER M: Jiayu Du <duge1256@gmail.com> S: Maintained F: Documentation/devicetree/bindings/nvmem/spacemit,k3-efuse.yaml +F: drivers/nvmem/spacemit-k3-efuse.c SPACEMIT RESET CONTROLLER DRIVER M: Guodong Xu <docular.xu@gmail.com> diff --git a/drivers/nvmem/Kconfig b/drivers/nvmem/Kconfig index d2f5bb236663..4a8f975b35c7 100644 --- a/drivers/nvmem/Kconfig +++ b/drivers/nvmem/Kconfig @@ -476,6 +476,17 @@ config NVMEM_SNVS_LPGPR This driver can also be built as a module. If so, the module will be called nvmem-snvs-lpgpr. +config NVMEM_SPACEMIT_K3_EFUSE + tristate "SpacemiT K3 eFuse support" + depends on ARCH_SPACEMIT || COMPILE_TEST + depends on OF && HAS_IOMEM + help + Say Y here to read configuration and calibration data from the + SpacemiT K3 eFuse shadow banks. Programming the fuses is not supported. + + This driver can also be built as a module. If so, the module will + be called spacemit-k3-efuse. + config NVMEM_SPMI_SDAM tristate "SPMI SDAM Support" depends on SPMI diff --git a/drivers/nvmem/Makefile b/drivers/nvmem/Makefile index 5e70bad9f0f9..d29716957faf 100644 --- a/drivers/nvmem/Makefile +++ b/drivers/nvmem/Makefile @@ -81,6 +81,7 @@ obj-$(CONFIG_NVMEM_SC27XX_EFUSE) += nvmem-sc27xx-efuse.o nvmem-sc27xx-efuse-y := sc27xx-efuse.o obj-$(CONFIG_NVMEM_SNVS_LPGPR) += nvmem_snvs_lpgpr.o nvmem_snvs_lpgpr-y := snvs_lpgpr.o +obj-$(CONFIG_NVMEM_SPACEMIT_K3_EFUSE) += spacemit-k3-efuse.o obj-$(CONFIG_NVMEM_SPMI_SDAM) += nvmem_qcom-spmi-sdam.o nvmem_qcom-spmi-sdam-y += qcom-spmi-sdam.o obj-$(CONFIG_NVMEM_SPRD_EFUSE) += nvmem_sprd_efuse.o diff --git a/drivers/nvmem/spacemit-k3-efuse.c b/drivers/nvmem/spacemit-k3-efuse.c new file mode 100644 index 000000000000..e126902130d4 --- /dev/null +++ b/drivers/nvmem/spacemit-k3-efuse.c @@ -0,0 +1,199 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * SpacemiT K3 read-only eFuse driver + * + * Copyright (c) 2026 Spacemit Co. Ltd. + */ + +#include <linux/array_size.h> +#include <linux/bits.h> +#include <linux/clk.h> +#include <linux/delay.h> +#include <linux/device.h> +#include <linux/err.h> +#include <linux/errno.h> +#include <linux/io.h> +#include <linux/iopoll.h> +#include <linux/module.h> +#include <linux/nvmem-provider.h> +#include <linux/of.h> +#include <linux/platform_device.h> +#include <linux/reset.h> +#include <linux/slab.h> +#include <linux/string.h> + +#include <asm/byteorder.h> + +#define K3_EFUSE_REG_SIZE 0x400 +#define K3_EFUSE_BANK_SIZE 0x20 +#define K3_EFUSE_CONFIG 0x004 +#define K3_EFUSE_SOFTWARE_RESET BIT(22) +#define K3_EFUSE_BANK0_OFFSET 0x104 +#define K3_EFUSE_STATUS 0x184 +#define K3_EFUSE_READY BIT(1) +#define K3_EFUSE_BANK7_OFFSET 0x190 +#define K3_EFUSE_RESET_DELAY_US 1 +#define K3_EFUSE_POLL_INTERVAL_US 1 +#define K3_EFUSE_POLL_TIMEOUT_US 10000 + +static const unsigned int k3_efuse_bank_offsets[] = { + K3_EFUSE_BANK0_OFFSET, + K3_EFUSE_BANK7_OFFSET, +}; + +/* Prevent NVMEM reads from exposing the root key and SSK registers. */ +static const struct nvmem_keepout k3_efuse_keepouts[] = { + { + .start = 0, + .end = K3_EFUSE_BANK0_OFFSET, + }, { + .start = K3_EFUSE_BANK0_OFFSET + K3_EFUSE_BANK_SIZE, + .end = K3_EFUSE_BANK7_OFFSET, + }, { + .start = K3_EFUSE_BANK7_OFFSET + K3_EFUSE_BANK_SIZE, + .end = K3_EFUSE_REG_SIZE, + }, +}; + +struct k3_efuse { + __le32 data[ARRAY_SIZE(k3_efuse_bank_offsets)][K3_EFUSE_BANK_SIZE / sizeof(u32)]; +}; + +static int k3_efuse_read(void *context, unsigned int offset, void *val, + size_t bytes) +{ + struct k3_efuse *efuse = context; + unsigned int bank, index; + + for (bank = 0; bank < ARRAY_SIZE(k3_efuse_bank_offsets); bank++) { + if (offset < k3_efuse_bank_offsets[bank]) + continue; + + index = offset - k3_efuse_bank_offsets[bank]; + if (index >= K3_EFUSE_BANK_SIZE || bytes > K3_EFUSE_BANK_SIZE - index) + continue; + + memcpy(val, (u8 *)efuse->data[bank] + index, bytes); + return 0; + } + + return -EINVAL; +} + +static int k3_efuse_reload(void __iomem *base) +{ + u32 val; + + val = readl(base + K3_EFUSE_CONFIG); + writel(val | K3_EFUSE_SOFTWARE_RESET, base + K3_EFUSE_CONFIG); + readl(base + K3_EFUSE_CONFIG); + udelay(K3_EFUSE_RESET_DELAY_US); + val = readl(base + K3_EFUSE_CONFIG); + writel(val & ~K3_EFUSE_SOFTWARE_RESET, base + K3_EFUSE_CONFIG); + + return readl_poll_timeout(base + K3_EFUSE_STATUS, val, + val & K3_EFUSE_READY, + K3_EFUSE_POLL_INTERVAL_US, + K3_EFUSE_POLL_TIMEOUT_US); +} + +static int k3_efuse_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct nvmem_config config = { + .dev = dev, + .name = dev_name(dev), + .id = NVMEM_DEVID_NONE, + .owner = THIS_MODULE, + .type = NVMEM_TYPE_OTP, + .read_only = true, + .root_only = true, + .reg_read = k3_efuse_read, + .word_size = 1, + .stride = 1, + .size = K3_EFUSE_REG_SIZE, + .keepout = k3_efuse_keepouts, + .nkeepout = ARRAY_SIZE(k3_efuse_keepouts), + }; + struct reset_control *reset; + struct nvmem_device *nvmem; + struct k3_efuse *efuse; + void __iomem *base; + struct clk *clk; + unsigned int bank, i; + int ret; + + base = devm_platform_ioremap_resource(pdev, 0); + if (IS_ERR(base)) + return PTR_ERR(base); + + efuse = devm_kzalloc(dev, sizeof(*efuse), GFP_KERNEL); + if (!efuse) + return -ENOMEM; + + clk = devm_clk_get(dev, NULL); + if (IS_ERR(clk)) + return dev_err_probe(dev, PTR_ERR(clk), "Failed to get clock\n"); + + reset = devm_reset_control_get_shared(dev, NULL); + if (IS_ERR(reset)) + return dev_err_probe(dev, PTR_ERR(reset), "Failed to get reset\n"); + + ret = clk_prepare_enable(clk); + if (ret) + return dev_err_probe(dev, ret, "Failed to enable clock\n"); + + /* WTM_RST resets the entire GEU; leave it deasserted on teardown. */ + ret = reset_control_deassert(reset); + if (ret) { + dev_err_probe(dev, ret, "Failed to deassert reset\n"); + goto disable_clk; + } + + ret = k3_efuse_reload(base); + if (ret) { + dev_err_probe(dev, ret, "Timed out waiting for shadow reload\n"); + goto disable_clk; + } + + for (bank = 0; bank < ARRAY_SIZE(k3_efuse_bank_offsets); bank++) { + for (i = 0; i < ARRAY_SIZE(efuse->data[bank]); i++) { + u32 val = readl(base + k3_efuse_bank_offsets[bank] + i * sizeof(u32)); + + efuse->data[bank][i] = cpu_to_le32(val); + } + } + + clk_disable_unprepare(clk); + + config.priv = efuse; + + nvmem = devm_nvmem_register(dev, &config); + if (IS_ERR(nvmem)) + return dev_err_probe(dev, PTR_ERR(nvmem), "Failed to register NVMEM\n"); + + return 0; + +disable_clk: + clk_disable_unprepare(clk); + return ret; +} + +static const struct of_device_id k3_efuse_of_match[] = { + { .compatible = "spacemit,k3-efuse" }, + { } +}; +MODULE_DEVICE_TABLE(of, k3_efuse_of_match); + +static struct platform_driver k3_efuse_driver = { + .probe = k3_efuse_probe, + .driver = { + .name = "spacemit-k3-efuse", + .of_match_table = k3_efuse_of_match, + }, +}; +module_platform_driver(k3_efuse_driver); + +MODULE_AUTHOR("Jiayu Du <duge1256@gmail.com>"); +MODULE_DESCRIPTION("SpacemiT K3 read-only eFuse driver"); +MODULE_LICENSE("GPL"); -- 2.43.0 _______________________________________________ linux-riscv mailing list linux-riscv@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-riscv ^ permalink raw reply related [flat|nested] 18+ messages in thread
* Re: [PATCH v2 2/3] nvmem: Add read-only SpacemiT K3 eFuse support 2026-11-16 7:54 ` Jiayu Du @ 2026-11-16 11:54 ` Philipp Zabel -1 siblings, 0 replies; 18+ messages in thread From: Philipp Zabel @ 2026-11-16 11:54 UTC (permalink / raw) To: Jiayu Du, srini Cc: robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, devicetree, linux-riscv, spacemit, linux-kernel On Mo, 2026-11-16 at 15:54 +0800, Jiayu Du wrote: > Add read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse > controller. These banks contain temperature calibration parameters and > other configuration data. > > Keep the root key and SSK registers outside the exposed bank ranges. > > FUSE_READY reports completion of a software-triggered reload. Trigger a > reload through GEU_CONFIG bit 22 with a 1 us reset pulse, then poll > FUSE_READY at a 1 us interval with a 10 ms timeout before taking the > snapshot. Fail probe if the reload times out. > > Snapshot the shadow registers with aligned 32-bit reads and expose the > data in little-endian order at controller-relative offsets. Serve reads > from the snapshot and use keepouts for the remaining register ranges. > > Enable the AES_WTM clock only while accessing the hardware. Leave its > shared reset deasserted on teardown because WTM_RST resets the entire GEU. I don't understand this, possibly because I don't know what GEU is. What is this reset shared with, and why exactly does the reset need to stay deasserted? regards Philipp ^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH v2 2/3] nvmem: Add read-only SpacemiT K3 eFuse support @ 2026-11-16 11:54 ` Philipp Zabel 0 siblings, 0 replies; 18+ messages in thread From: Philipp Zabel @ 2026-11-16 11:54 UTC (permalink / raw) To: Jiayu Du, srini Cc: robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, devicetree, linux-riscv, spacemit, linux-kernel On Mo, 2026-11-16 at 15:54 +0800, Jiayu Du wrote: > Add read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse > controller. These banks contain temperature calibration parameters and > other configuration data. > > Keep the root key and SSK registers outside the exposed bank ranges. > > FUSE_READY reports completion of a software-triggered reload. Trigger a > reload through GEU_CONFIG bit 22 with a 1 us reset pulse, then poll > FUSE_READY at a 1 us interval with a 10 ms timeout before taking the > snapshot. Fail probe if the reload times out. > > Snapshot the shadow registers with aligned 32-bit reads and expose the > data in little-endian order at controller-relative offsets. Serve reads > from the snapshot and use keepouts for the remaining register ranges. > > Enable the AES_WTM clock only while accessing the hardware. Leave its > shared reset deasserted on teardown because WTM_RST resets the entire GEU. I don't understand this, possibly because I don't know what GEU is. What is this reset shared with, and why exactly does the reset need to stay deasserted? regards Philipp _______________________________________________ linux-riscv mailing list linux-riscv@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-riscv ^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH v2 2/3] nvmem: Add read-only SpacemiT K3 eFuse support 2026-11-16 11:54 ` Philipp Zabel @ 2026-11-17 12:22 ` Jiayu Du -1 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-17 12:22 UTC (permalink / raw) To: Philipp Zabel Cc: srini, robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, devicetree, linux-riscv, spacemit, linux-kernel On Mon, Nov 16, 2026 at 12:54:42PM +0100, Philipp Zabel wrote: > On Mo, 2026-11-16 at 15:54 +0800, Jiayu Du wrote: > > Add read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse > > controller. These banks contain temperature calibration parameters and > > other configuration data. > > > > Keep the root key and SSK registers outside the exposed bank ranges. > > > > FUSE_READY reports completion of a software-triggered reload. Trigger a > > reload through GEU_CONFIG bit 22 with a 1 us reset pulse, then poll > > FUSE_READY at a 1 us interval with a 10 ms timeout before taking the > > snapshot. Fail probe if the reload times out. > > > > Snapshot the shadow registers with aligned 32-bit reads and expose the > > data in little-endian order at controller-relative offsets. Serve reads > > from the snapshot and use keepouts for the remaining register ranges. > > > > Enable the AES_WTM clock only while accessing the hardware. Leave its > > shared reset deasserted on teardown because WTM_RST resets the entire GEU. > > I don't understand this, possibly because I don't know what GEU is. > > What is this reset shared with, and why exactly does the reset need to > stay deasserted? GEU stands for Generic Encryption and Fuse Unit. This reset is not shared with another consumer and only resets the controller state. My explanation about needing to keep it deasserted was incorrect. For v3, I've switched to devm_reset_control_get_exclusive_deasserted() and removed that explanation. The reset is now asserted on driver detach or a subsequent probe failure. Best regards Jiayu Du > > regards > Philipp ^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH v2 2/3] nvmem: Add read-only SpacemiT K3 eFuse support @ 2026-11-17 12:22 ` Jiayu Du 0 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-17 12:22 UTC (permalink / raw) To: Philipp Zabel Cc: srini, robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, devicetree, linux-riscv, spacemit, linux-kernel On Mon, Nov 16, 2026 at 12:54:42PM +0100, Philipp Zabel wrote: > On Mo, 2026-11-16 at 15:54 +0800, Jiayu Du wrote: > > Add read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse > > controller. These banks contain temperature calibration parameters and > > other configuration data. > > > > Keep the root key and SSK registers outside the exposed bank ranges. > > > > FUSE_READY reports completion of a software-triggered reload. Trigger a > > reload through GEU_CONFIG bit 22 with a 1 us reset pulse, then poll > > FUSE_READY at a 1 us interval with a 10 ms timeout before taking the > > snapshot. Fail probe if the reload times out. > > > > Snapshot the shadow registers with aligned 32-bit reads and expose the > > data in little-endian order at controller-relative offsets. Serve reads > > from the snapshot and use keepouts for the remaining register ranges. > > > > Enable the AES_WTM clock only while accessing the hardware. Leave its > > shared reset deasserted on teardown because WTM_RST resets the entire GEU. > > I don't understand this, possibly because I don't know what GEU is. > > What is this reset shared with, and why exactly does the reset need to > stay deasserted? GEU stands for Generic Encryption and Fuse Unit. This reset is not shared with another consumer and only resets the controller state. My explanation about needing to keep it deasserted was incorrect. For v3, I've switched to devm_reset_control_get_exclusive_deasserted() and removed that explanation. The reset is now asserted on driver detach or a subsequent probe failure. Best regards Jiayu Du > > regards > Philipp _______________________________________________ linux-riscv mailing list linux-riscv@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-riscv ^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH v2 2/3] nvmem: Add read-only SpacemiT K3 eFuse support 2026-11-16 7:54 ` Jiayu Du @ 2026-11-17 21:09 ` Aurelien Jarno -1 siblings, 0 replies; 18+ messages in thread From: Aurelien Jarno @ 2026-11-17 21:09 UTC (permalink / raw) To: Jiayu Du Cc: srini, robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel Hi, On 2026-11-16 15:54, Jiayu Du wrote: > Add read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse > controller. These banks contain temperature calibration parameters and > other configuration data. > > Keep the root key and SSK registers outside the exposed bank ranges. > > FUSE_READY reports completion of a software-triggered reload. Trigger a > reload through GEU_CONFIG bit 22 with a 1 us reset pulse, then poll > FUSE_READY at a 1 us interval with a 10 ms timeout before taking the > snapshot. Fail probe if the reload times out. > > Snapshot the shadow registers with aligned 32-bit reads and expose the > data in little-endian order at controller-relative offsets. Serve reads > from the snapshot and use keepouts for the remaining register ranges. > > Enable the AES_WTM clock only while accessing the hardware. Leave its > shared reset deasserted on teardown because WTM_RST resets the entire GEU. > > Signed-off-by: Jiayu Du <duge1256@gmail.com> > --- > MAINTAINERS | 1 + > drivers/nvmem/Kconfig | 11 ++ > drivers/nvmem/Makefile | 1 + > drivers/nvmem/spacemit-k3-efuse.c | 199 ++++++++++++++++++++++++++++++ > 4 files changed, 212 insertions(+) > create mode 100644 drivers/nvmem/spacemit-k3-efuse.c I have been looking at the code with more details. Is there any reason to "expose" the whole 1024-byte efuse address space in the nvmem layout (of course with some masking) as opposed to the 16 banks x 32 bytes = 4096 bits only (still with masking)? I find strange that user space has to know the address of bank 0 or bank 7 instead of just looking respectively at addresses 0x000 and 0x0e0. As the layout exposed to user space also includes status or control registers (although masked), this might makes difficult to support a newer version of the efuse controller which have a similar data layout, but slightly different registers colliding with the bank data. Regards Aurelien -- Aurelien Jarno GPG: 4096R/1DDD8C9B aurelien@aurel32.net http://aurel32.net ^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH v2 2/3] nvmem: Add read-only SpacemiT K3 eFuse support @ 2026-11-17 21:09 ` Aurelien Jarno 0 siblings, 0 replies; 18+ messages in thread From: Aurelien Jarno @ 2026-11-17 21:09 UTC (permalink / raw) To: Jiayu Du Cc: srini, robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel Hi, On 2026-11-16 15:54, Jiayu Du wrote: > Add read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse > controller. These banks contain temperature calibration parameters and > other configuration data. > > Keep the root key and SSK registers outside the exposed bank ranges. > > FUSE_READY reports completion of a software-triggered reload. Trigger a > reload through GEU_CONFIG bit 22 with a 1 us reset pulse, then poll > FUSE_READY at a 1 us interval with a 10 ms timeout before taking the > snapshot. Fail probe if the reload times out. > > Snapshot the shadow registers with aligned 32-bit reads and expose the > data in little-endian order at controller-relative offsets. Serve reads > from the snapshot and use keepouts for the remaining register ranges. > > Enable the AES_WTM clock only while accessing the hardware. Leave its > shared reset deasserted on teardown because WTM_RST resets the entire GEU. > > Signed-off-by: Jiayu Du <duge1256@gmail.com> > --- > MAINTAINERS | 1 + > drivers/nvmem/Kconfig | 11 ++ > drivers/nvmem/Makefile | 1 + > drivers/nvmem/spacemit-k3-efuse.c | 199 ++++++++++++++++++++++++++++++ > 4 files changed, 212 insertions(+) > create mode 100644 drivers/nvmem/spacemit-k3-efuse.c I have been looking at the code with more details. Is there any reason to "expose" the whole 1024-byte efuse address space in the nvmem layout (of course with some masking) as opposed to the 16 banks x 32 bytes = 4096 bits only (still with masking)? I find strange that user space has to know the address of bank 0 or bank 7 instead of just looking respectively at addresses 0x000 and 0x0e0. As the layout exposed to user space also includes status or control registers (although masked), this might makes difficult to support a newer version of the efuse controller which have a similar data layout, but slightly different registers colliding with the bank data. Regards Aurelien -- Aurelien Jarno GPG: 4096R/1DDD8C9B aurelien@aurel32.net http://aurel32.net _______________________________________________ linux-riscv mailing list linux-riscv@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-riscv ^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH v2 2/3] nvmem: Add read-only SpacemiT K3 eFuse support 2026-11-17 21:09 ` Aurelien Jarno @ 2026-11-18 6:15 ` Jiayu Du -1 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-18 6:15 UTC (permalink / raw) To: Aurelien Jarno Cc: srini, robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel On Tue, Nov 17, 2026 at 10:09:45PM +0100, Aurelien Jarno wrote: > Hi, > > On 2026-11-16 15:54, Jiayu Du wrote: > > Add read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse > > controller. These banks contain temperature calibration parameters and > > other configuration data. > > > > Keep the root key and SSK registers outside the exposed bank ranges. > > > > FUSE_READY reports completion of a software-triggered reload. Trigger a > > reload through GEU_CONFIG bit 22 with a 1 us reset pulse, then poll > > FUSE_READY at a 1 us interval with a 10 ms timeout before taking the > > snapshot. Fail probe if the reload times out. > > > > Snapshot the shadow registers with aligned 32-bit reads and expose the > > data in little-endian order at controller-relative offsets. Serve reads > > from the snapshot and use keepouts for the remaining register ranges. > > > > Enable the AES_WTM clock only while accessing the hardware. Leave its > > shared reset deasserted on teardown because WTM_RST resets the entire GEU. > > > > Signed-off-by: Jiayu Du <duge1256@gmail.com> > > --- > > MAINTAINERS | 1 + > > drivers/nvmem/Kconfig | 11 ++ > > drivers/nvmem/Makefile | 1 + > > drivers/nvmem/spacemit-k3-efuse.c | 199 ++++++++++++++++++++++++++++++ > > 4 files changed, 212 insertions(+) > > create mode 100644 drivers/nvmem/spacemit-k3-efuse.c > > I have been looking at the code with more details. Is there any reason > to "expose" the whole 1024-byte efuse address space in the nvmem layout > (of course with some masking) as opposed to the 16 banks x 32 bytes = > 4096 bits only (still with masking)? > > I find strange that user space has to know the address of bank 0 or bank > 7 instead of just looking respectively at addresses 0x000 and 0x0e0. > > As the layout exposed to user space also includes status or control > registers (although masked), this might makes difficult to support a > newer version of the efuse controller which have a similar data layout, > but slightly different registers colliding with the bank data. > Hi Aurelien, I will use a 512-byte NVMEM space, with bank 0 starting at 0x000 and bank 7 at 0x0e0. The other banks will remain covered by keepouts. I will also update the temperature calibration cell offsets in the device tree to 0x1c and 0xea. The hardware accesses will stay unchanged. Best regards, Jiayu Du ^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH v2 2/3] nvmem: Add read-only SpacemiT K3 eFuse support @ 2026-11-18 6:15 ` Jiayu Du 0 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-18 6:15 UTC (permalink / raw) To: Aurelien Jarno Cc: srini, robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel On Tue, Nov 17, 2026 at 10:09:45PM +0100, Aurelien Jarno wrote: > Hi, > > On 2026-11-16 15:54, Jiayu Du wrote: > > Add read-only NVMEM access to bank 0 and bank 7 of the SpacemiT K3 eFuse > > controller. These banks contain temperature calibration parameters and > > other configuration data. > > > > Keep the root key and SSK registers outside the exposed bank ranges. > > > > FUSE_READY reports completion of a software-triggered reload. Trigger a > > reload through GEU_CONFIG bit 22 with a 1 us reset pulse, then poll > > FUSE_READY at a 1 us interval with a 10 ms timeout before taking the > > snapshot. Fail probe if the reload times out. > > > > Snapshot the shadow registers with aligned 32-bit reads and expose the > > data in little-endian order at controller-relative offsets. Serve reads > > from the snapshot and use keepouts for the remaining register ranges. > > > > Enable the AES_WTM clock only while accessing the hardware. Leave its > > shared reset deasserted on teardown because WTM_RST resets the entire GEU. > > > > Signed-off-by: Jiayu Du <duge1256@gmail.com> > > --- > > MAINTAINERS | 1 + > > drivers/nvmem/Kconfig | 11 ++ > > drivers/nvmem/Makefile | 1 + > > drivers/nvmem/spacemit-k3-efuse.c | 199 ++++++++++++++++++++++++++++++ > > 4 files changed, 212 insertions(+) > > create mode 100644 drivers/nvmem/spacemit-k3-efuse.c > > I have been looking at the code with more details. Is there any reason > to "expose" the whole 1024-byte efuse address space in the nvmem layout > (of course with some masking) as opposed to the 16 banks x 32 bytes = > 4096 bits only (still with masking)? > > I find strange that user space has to know the address of bank 0 or bank > 7 instead of just looking respectively at addresses 0x000 and 0x0e0. > > As the layout exposed to user space also includes status or control > registers (although masked), this might makes difficult to support a > newer version of the efuse controller which have a similar data layout, > but slightly different registers colliding with the bank data. > Hi Aurelien, I will use a 512-byte NVMEM space, with bank 0 starting at 0x000 and bank 7 at 0x0e0. The other banks will remain covered by keepouts. I will also update the temperature calibration cell offsets in the device tree to 0x1c and 0xea. The hardware accesses will stay unchanged. Best regards, Jiayu Du _______________________________________________ linux-riscv mailing list linux-riscv@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-riscv ^ permalink raw reply [flat|nested] 18+ messages in thread
* [PATCH v2 3/3] riscv: dts: spacemit: Add K3 eFuse and temperature trim cells 2026-11-16 7:54 ` Jiayu Du @ 2026-11-16 7:54 ` Jiayu Du -1 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-16 7:54 UTC (permalink / raw) To: srini Cc: robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel, Jiayu Du Add the K3 eFuse controller with its AES_WTM clock and reset. Describe fixed-layout cells for the temperature calibration parameters stored in bank 0 and bank 7, using controller-relative offsets. Signed-off-by: Jiayu Du <duge1256@gmail.com> --- arch/riscv/boot/dts/spacemit/k3.dtsi | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/arch/riscv/boot/dts/spacemit/k3.dtsi b/arch/riscv/boot/dts/spacemit/k3.dtsi index e89d2cc5ca8c..b548f5029d59 100644 --- a/arch/riscv/boot/dts/spacemit/k3.dtsi +++ b/arch/riscv/boot/dts/spacemit/k3.dtsi @@ -1439,6 +1439,29 @@ saplic: interrupt-controller@e0804000 { /* sec_i2c3: 0xf0614000, not available from Linux */ + efuse: efuse@f0702800 { + compatible = "spacemit,k3-efuse"; + reg = <0x0 0xf0702800 0x0 0x400>; + clocks = <&syscon_apmu CLK_APMU_AES_WTM>; + resets = <&syscon_apmu RESET_APMU_AES_WTM>; + + nvmem-layout { + compatible = "fixed-layout"; + #address-cells = <1>; + #size-cells = <1>; + + soc_rtemp_trim1: temperature-trim@120 { + reg = <0x120 0x1>; + bits = <0 8>; + }; + + soc_rtemp_trim: temperature-trim@19a { + reg = <0x19a 0x2>; + bits = <4 8>; + }; + }; + }; + mimsic: interrupt-controller@f1000000 { compatible = "spacemit,k3-imsics", "riscv,imsics"; reg = <0x0 0xf1000000 0x0 0x10000>; -- 2.43.0 ^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH v2 3/3] riscv: dts: spacemit: Add K3 eFuse and temperature trim cells @ 2026-11-16 7:54 ` Jiayu Du 0 siblings, 0 replies; 18+ messages in thread From: Jiayu Du @ 2026-11-16 7:54 UTC (permalink / raw) To: srini Cc: robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, aurelien, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel, Jiayu Du Add the K3 eFuse controller with its AES_WTM clock and reset. Describe fixed-layout cells for the temperature calibration parameters stored in bank 0 and bank 7, using controller-relative offsets. Signed-off-by: Jiayu Du <duge1256@gmail.com> --- arch/riscv/boot/dts/spacemit/k3.dtsi | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/arch/riscv/boot/dts/spacemit/k3.dtsi b/arch/riscv/boot/dts/spacemit/k3.dtsi index e89d2cc5ca8c..b548f5029d59 100644 --- a/arch/riscv/boot/dts/spacemit/k3.dtsi +++ b/arch/riscv/boot/dts/spacemit/k3.dtsi @@ -1439,6 +1439,29 @@ saplic: interrupt-controller@e0804000 { /* sec_i2c3: 0xf0614000, not available from Linux */ + efuse: efuse@f0702800 { + compatible = "spacemit,k3-efuse"; + reg = <0x0 0xf0702800 0x0 0x400>; + clocks = <&syscon_apmu CLK_APMU_AES_WTM>; + resets = <&syscon_apmu RESET_APMU_AES_WTM>; + + nvmem-layout { + compatible = "fixed-layout"; + #address-cells = <1>; + #size-cells = <1>; + + soc_rtemp_trim1: temperature-trim@120 { + reg = <0x120 0x1>; + bits = <0 8>; + }; + + soc_rtemp_trim: temperature-trim@19a { + reg = <0x19a 0x2>; + bits = <4 8>; + }; + }; + }; + mimsic: interrupt-controller@f1000000 { compatible = "spacemit,k3-imsics", "riscv,imsics"; reg = <0x0 0xf1000000 0x0 0x10000>; -- 2.43.0 _______________________________________________ linux-riscv mailing list linux-riscv@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-riscv ^ permalink raw reply related [flat|nested] 18+ messages in thread
* Re: [PATCH v2 3/3] riscv: dts: spacemit: Add K3 eFuse and temperature trim cells 2026-11-16 7:54 ` Jiayu Du @ 2026-11-17 21:10 ` Aurelien Jarno -1 siblings, 0 replies; 18+ messages in thread From: Aurelien Jarno @ 2026-11-17 21:10 UTC (permalink / raw) To: Jiayu Du Cc: srini, robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel Hi, On 2026-11-16 15:54, Jiayu Du wrote: > Add the K3 eFuse controller with its AES_WTM clock and reset. Describe > fixed-layout cells for the temperature calibration parameters stored in > bank 0 and bank 7, using controller-relative offsets. > > Signed-off-by: Jiayu Du <duge1256@gmail.com> > --- > arch/riscv/boot/dts/spacemit/k3.dtsi | 23 +++++++++++++++++++++++ > 1 file changed, 23 insertions(+) > > diff --git a/arch/riscv/boot/dts/spacemit/k3.dtsi b/arch/riscv/boot/dts/spacemit/k3.dtsi > index e89d2cc5ca8c..b548f5029d59 100644 > --- a/arch/riscv/boot/dts/spacemit/k3.dtsi > +++ b/arch/riscv/boot/dts/spacemit/k3.dtsi > @@ -1439,6 +1439,29 @@ saplic: interrupt-controller@e0804000 { > > /* sec_i2c3: 0xf0614000, not available from Linux */ > > + efuse: efuse@f0702800 { > + compatible = "spacemit,k3-efuse"; > + reg = <0x0 0xf0702800 0x0 0x400>; > + clocks = <&syscon_apmu CLK_APMU_AES_WTM>; > + resets = <&syscon_apmu RESET_APMU_AES_WTM>; > + > + nvmem-layout { > + compatible = "fixed-layout"; > + #address-cells = <1>; > + #size-cells = <1>; > + > + soc_rtemp_trim1: temperature-trim@120 { > + reg = <0x120 0x1>; > + bits = <0 8>; > + }; > + > + soc_rtemp_trim: temperature-trim@19a { > + reg = <0x19a 0x2>; > + bits = <4 8>; > + }; > + }; > + }; > + > mimsic: interrupt-controller@f1000000 { > compatible = "spacemit,k3-imsics", "riscv,imsics"; > reg = <0x0 0xf1000000 0x0 0x10000>; Thanks for this new version which is now enabled directly at the K3 level instead of the board level. This looks good to me, except the offsets values, given my remark on the other patch. Regards Aurelien -- Aurelien Jarno GPG: 4096R/1DDD8C9B aurelien@aurel32.net http://aurel32.net _______________________________________________ linux-riscv mailing list linux-riscv@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-riscv ^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH v2 3/3] riscv: dts: spacemit: Add K3 eFuse and temperature trim cells @ 2026-11-17 21:10 ` Aurelien Jarno 0 siblings, 0 replies; 18+ messages in thread From: Aurelien Jarno @ 2026-11-17 21:10 UTC (permalink / raw) To: Jiayu Du Cc: srini, robh, krzk+dt, conor+dt, dlan, troy.mitchell, me, pjw, palmer, aou, alex, p.zabel, devicetree, linux-riscv, spacemit, linux-kernel Hi, On 2026-11-16 15:54, Jiayu Du wrote: > Add the K3 eFuse controller with its AES_WTM clock and reset. Describe > fixed-layout cells for the temperature calibration parameters stored in > bank 0 and bank 7, using controller-relative offsets. > > Signed-off-by: Jiayu Du <duge1256@gmail.com> > --- > arch/riscv/boot/dts/spacemit/k3.dtsi | 23 +++++++++++++++++++++++ > 1 file changed, 23 insertions(+) > > diff --git a/arch/riscv/boot/dts/spacemit/k3.dtsi b/arch/riscv/boot/dts/spacemit/k3.dtsi > index e89d2cc5ca8c..b548f5029d59 100644 > --- a/arch/riscv/boot/dts/spacemit/k3.dtsi > +++ b/arch/riscv/boot/dts/spacemit/k3.dtsi > @@ -1439,6 +1439,29 @@ saplic: interrupt-controller@e0804000 { > > /* sec_i2c3: 0xf0614000, not available from Linux */ > > + efuse: efuse@f0702800 { > + compatible = "spacemit,k3-efuse"; > + reg = <0x0 0xf0702800 0x0 0x400>; > + clocks = <&syscon_apmu CLK_APMU_AES_WTM>; > + resets = <&syscon_apmu RESET_APMU_AES_WTM>; > + > + nvmem-layout { > + compatible = "fixed-layout"; > + #address-cells = <1>; > + #size-cells = <1>; > + > + soc_rtemp_trim1: temperature-trim@120 { > + reg = <0x120 0x1>; > + bits = <0 8>; > + }; > + > + soc_rtemp_trim: temperature-trim@19a { > + reg = <0x19a 0x2>; > + bits = <4 8>; > + }; > + }; > + }; > + > mimsic: interrupt-controller@f1000000 { > compatible = "spacemit,k3-imsics", "riscv,imsics"; > reg = <0x0 0xf1000000 0x0 0x10000>; Thanks for this new version which is now enabled directly at the K3 level instead of the board level. This looks good to me, except the offsets values, given my remark on the other patch. Regards Aurelien -- Aurelien Jarno GPG: 4096R/1DDD8C9B aurelien@aurel32.net http://aurel32.net ^ permalink raw reply [flat|nested] 18+ messages in thread
end of thread, other threads:[~2026-11-18 6:15 UTC | newest] Thread overview: 18+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2026-11-16 7:54 [PATCH v2 0/3] Add read-only SpacemiT K3 eFuse support Jiayu Du 2026-11-16 7:54 ` Jiayu Du 2026-11-16 7:54 ` [PATCH v2 1/3] dt-bindings: nvmem: Add SpacemiT K3 eFuse controller Jiayu Du 2026-11-16 7:54 ` Jiayu Du 2026-11-16 7:54 ` [PATCH v2 2/3] nvmem: Add read-only SpacemiT K3 eFuse support Jiayu Du 2026-11-16 7:54 ` Jiayu Du 2026-11-16 11:54 ` Philipp Zabel 2026-11-16 11:54 ` Philipp Zabel 2026-11-17 12:22 ` Jiayu Du 2026-11-17 12:22 ` Jiayu Du 2026-11-17 21:09 ` Aurelien Jarno 2026-11-17 21:09 ` Aurelien Jarno 2026-11-18 6:15 ` Jiayu Du 2026-11-18 6:15 ` Jiayu Du 2026-11-16 7:54 ` [PATCH v2 3/3] riscv: dts: spacemit: Add K3 eFuse and temperature trim cells Jiayu Du 2026-11-16 7:54 ` Jiayu Du 2026-11-17 21:10 ` Aurelien Jarno 2026-11-17 21:10 ` Aurelien Jarno
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.