[PATCH v8 0/2] Add iop driver for Sunplus SP7021

[PATCH v8 0/2] Add iop driver for Sunplus SP7021

[Tony Huang]

Add iop driver for Sunplus SP7021 SOC

This is a patch series for iop driver for Sunplus SP7021 SOC.

Sunplus SP7021 is an ARM Cortex A7 (4 cores) based SoC. It integrates
many peripherals (ex: UART, I2C, SPI, SDIO, eMMC, USB, SD card and
etc.) into a single chip. It is designed for industrial control.

Refer to:
https://sunplus-tibbo.atlassian.net/wiki/spaces/doc/overview
https://tibbo.com/store/plus1.html

Tony Huang (2):
  dt-binding: misc: Add iop yaml file for Sunplus SP7021
  misc: Add iop driver for Sunplus SP7021

 Documentation/ABI/testing/sysfs-platform-soc@B     |  28 ++
 .../devicetree/bindings/misc/sunplus-iop.yaml      |  76 ++++
 MAINTAINERS                                        |   7 +
 drivers/misc/Kconfig                               |  20 +
 drivers/misc/Makefile                              |   1 +
 drivers/misc/sunplus_iop.c                         | 463 +++++++++++++++++++++
 6 files changed, 595 insertions(+)
 create mode 100644 Documentation/ABI/testing/sysfs-platform-soc@B
 create mode 100644 Documentation/devicetree/bindings/misc/sunplus-iop.yaml
 create mode 100644 drivers/misc/sunplus_iop.c

-- 
2.7.4

[PATCH v8 1/2] dt-binding: misc: Add iop yaml file for Sunplus SP7021

[Tony Huang]

Add iop yaml file for Sunplus SP7021

Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Tony Huang <tonyhuang.sunplus@gmail.com>
---
Changes in v8:
 -No change

 .../devicetree/bindings/misc/sunplus-iop.yaml      | 76 ++++++++++++++++++++++
 MAINTAINERS                                        |  5 ++
 2 files changed, 81 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/misc/sunplus-iop.yaml

diff --git a/Documentation/devicetree/bindings/misc/sunplus-iop.yaml b/Documentation/devicetree/bindings/misc/sunplus-iop.yaml
new file mode 100644
index 0000000..b37e697
--- /dev/null
+++ b/Documentation/devicetree/bindings/misc/sunplus-iop.yaml
@@ -0,0 +1,76 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+# Copyright (C) Sunplus Ltd. Co. 2021
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/misc/sunplus-iop.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Sunplus IOP(8051) controller
+
+maintainers:
+  - Tony Huang <tonyhuang.sunplus@gmail.com>
+
+description: |
+  Processor for I/O control, RTC wake-up procedure management,
+  and cooperation with CPU&PMC in power management.
+
+properties:
+  compatible:
+    enum:
+      - sunplus,sp7021-iop
+
+  reg:
+    items:
+      - description: IOP registers regions
+      - description: PMC registers regions
+      - description: MOON0 registers regions
+
+  reg-names:
+    items:
+      - const: iop
+      - const: iop_pmc
+      - const: moon0
+
+  interrupts:
+    items:
+      - description: IOP_INT0. IOP to system Interrupt 0.
+                     Sent by IOP to system RISC.
+      - description: IOP_INT1. IOP to System Interrupt 1.
+                     Sent by IOP to system RISC.
+
+  memory-region:
+    maxItems: 1
+
+  wakeup-gpios:
+    description: When the linux kernel system is powered off.
+      8051 is always powered. 8051 cand receive external signals
+      according to this gpio pin. 8051 receives external signal
+      through gpio pin. 8051 can power on linux kernel system.
+
+required:
+  - compatible
+  - reg
+  - reg-names
+  - interrupts
+  - memory-region
+  - wakeup-gpios
+
+additionalProperties: false
+
+examples:
+  - |
+    #include <dt-bindings/interrupt-controller/irq.h>
+    #include <dt-bindings/interrupt-controller/arm-gic.h>
+    #include <dt-bindings/gpio/gpio.h>
+    iop: iop@9c000400 {
+        compatible = "sunplus,sp7021-iop";
+        reg = <0x9c000400 0x80>, <0x9c003100 0x80>, <0x9c000000 0x80>;
+        reg-names = "iop", "iop_pmc", "moon0";
+        interrupt-parent = <&intc>;
+        interrupts = <41 IRQ_TYPE_LEVEL_HIGH>, <42 IRQ_TYPE_LEVEL_HIGH>;
+        memory-region = <&iop_reserve>;
+        pinctrl-names = "default";
+        pinctrl-0 = <&iop_pins>;
+        wakeup-gpios = <&pctl 1 GPIO_ACTIVE_HIGH>;
+    };
+...
diff --git a/MAINTAINERS b/MAINTAINERS
index fb18ce7..6f336c9 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -18242,6 +18242,11 @@ L:	netdev@vger.kernel.org
 S:	Maintained
 F:	drivers/net/ethernet/dlink/sundance.c
 
+SUNPLUS IOP DRIVER
+M:	Tony Huang <tonyhuang.sunplus@gmail.com>
+S:	Maintained
+F:	Documentation/devicetree/bindings/misc/sunplu-iop.yaml
+
 SUPERH
 M:	Yoshinori Sato <ysato@users.sourceforge.jp>
 M:	Rich Felker <dalias@libc.org>
-- 
2.7.4

[PATCH v8 2/2] misc: Add iop driver for Sunplus SP7021

[Tony Huang]

IOP(8051) embedded inside SP7021 which is used as
Processor for I/O control, monitor RTC interrupt and
cooperation with CPU & PMC in power management purpose.
The IOP core is DQ8051, so also named IOP8051,
it supports dedicated JTAG debug pins which share with SP7021.
In standby mode operation, the power spec reach 400uA.

Signed-off-by: Tony Huang <tonyhuang.sunplus@gmail.com>
---
Changes in v8:
 - Addressed comments from Greg KH.

 Documentation/ABI/testing/sysfs-platform-soc@B |  28 ++
 MAINTAINERS                                    |   2 +
 drivers/misc/Kconfig                           |  20 ++
 drivers/misc/Makefile                          |   1 +
 drivers/misc/sunplus_iop.c                     | 463 +++++++++++++++++++++++++
 5 files changed, 514 insertions(+)
 create mode 100644 Documentation/ABI/testing/sysfs-platform-soc@B
 create mode 100644 drivers/misc/sunplus_iop.c

diff --git a/Documentation/ABI/testing/sysfs-platform-soc@B b/Documentation/ABI/testing/sysfs-platform-soc@B
new file mode 100644
index 0000000..d26d6f5
--- /dev/null
+++ b/Documentation/ABI/testing/sysfs-platform-soc@B
@@ -0,0 +1,28 @@
+What:		/sys/devices/platform/soc@B/9c000400.iop/sp_iop_mailbox
+Date:		January 2022
+KernelVersion:	5.16
+Contact:	Tony Huang <tonyhuang.sunplus@gmail.com>
+Description:
+		Show IOP's mailbox0 register data.
+		Format: %x
+
+What:		/sys/devices/platform/soc@B/9c000400.iop/sp_iop_mode
+Date:		January 2022
+KernelVersion:	5.16
+Contact:	Tony Huang <tonyhuang.sunplus@gmail.com>
+Description:
+		Read-Write.
+
+		Write this file.
+		Operation mode of IOP is switched to standby mode by writing
+		"1" to sysfs.
+		Operation mode of IOP is switched to normal mode by writing
+		"0" to sysfs.
+		Writing of other values is invalid.
+
+		Read this file.
+		Show operation mode of IOP. "0" is normal mode. "1" is standby
+		mode.
+		Format: %x
+
+
diff --git a/MAINTAINERS b/MAINTAINERS
index 6f336c9..cbc8dff 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -18245,7 +18245,9 @@ F:	drivers/net/ethernet/dlink/sundance.c
 SUNPLUS IOP DRIVER
 M:	Tony Huang <tonyhuang.sunplus@gmail.com>
 S:	Maintained
+F:	Documentation/ABI/testing/sysfs-platform-soc@B
 F:	Documentation/devicetree/bindings/misc/sunplu-iop.yaml
+F:	drivers/misc/sunplus_iop.c
 
 SUPERH
 M:	Yoshinori Sato <ysato@users.sourceforge.jp>
diff --git a/drivers/misc/Kconfig b/drivers/misc/Kconfig
index 0f5a49f..c1d790e 100644
--- a/drivers/misc/Kconfig
+++ b/drivers/misc/Kconfig
@@ -470,6 +470,26 @@ config HISI_HIKEY_USB
 	  switching between the dual-role USB-C port and the USB-A host ports
 	  using only one USB controller.
 
+config SUNPLUS_IOP
+	tristate "Sunplus IOP support"
+	default ARCH_SUNPLUS
+	help
+	  Sunplus I/O processor (8051) driver.
+	  Processor for I/O control, monitor RTC interrupt
+	  and cooperation with CPU&PMC in power management.
+
+	  The IOP core is DQ8051, so also named IOP8051.
+	  Need Install DQ8051, The DQ8051 bin file generated by keil C.
+	  We will provide users with 8051 normal and standby source code.
+	  Path: https://sunplus.atlassian.net/wiki/spaces/doc/pages/610172933/
+	  How+to+use+I+O+processor+8051+of+SP7021#5.-Write-C-or-assembly-source-files-for-IOP
+	  Users can follow the operation steps to generate normal.bin and standby.bin.
+	  Path: https://sunplus.atlassian.net/wiki/spaces/doc/pages/466190338/26.+IOP8051
+	  26.5?How To Create 8051 bin file
+
+	  This driver can also be built as a module.  If so, the module
+	  will be called sunplus_iop.
+
 source "drivers/misc/c2port/Kconfig"
 source "drivers/misc/eeprom/Kconfig"
 source "drivers/misc/cb710/Kconfig"
diff --git a/drivers/misc/Makefile b/drivers/misc/Makefile
index a086197..eafeab6 100644
--- a/drivers/misc/Makefile
+++ b/drivers/misc/Makefile
@@ -52,6 +52,7 @@ obj-$(CONFIG_DW_XDATA_PCIE)	+= dw-xdata-pcie.o
 obj-$(CONFIG_PCI_ENDPOINT_TEST)	+= pci_endpoint_test.o
 obj-$(CONFIG_OCXL)		+= ocxl/
 obj-$(CONFIG_BCM_VK)		+= bcm-vk/
+obj-$(CONFIG_SUNPLUS_IOP)	+= sunplus_iop.o
 obj-y				+= cardreader/
 obj-$(CONFIG_PVPANIC)   	+= pvpanic/
 obj-$(CONFIG_HABANA_AI)		+= habanalabs/
diff --git a/drivers/misc/sunplus_iop.c b/drivers/misc/sunplus_iop.c
new file mode 100644
index 0000000..6ef2de7
--- /dev/null
+++ b/drivers/misc/sunplus_iop.c
@@ -0,0 +1,463 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * The IOP driver for Sunplus SP7021
+ *
+ * Copyright (C) 2021 Sunplus Technology Inc.
+ *
+ * All Rights Reserved.
+ */
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/firmware.h>
+#include <linux/iopoll.h>
+#include <linux/module.h>
+#include <linux/miscdevice.h>
+#include <linux/of_platform.h>
+#include <linux/of_address.h>
+#include <linux/of_gpio.h>
+
+enum IOP_Status_e {
+	IOP_SUCCESS,		/* successful */
+	IOP_ERR_IOP_BUSY,	/* IOP is busy */
+};
+
+/* moon0 register offset */
+#define IOP_CLKEN0	0x04
+#define IOP_RESET0	0x54
+
+/* IOP register offset */
+#define IOP_CONTROL	0x00
+#define IOP_DATA0	0x20
+#define IOP_DATA1	0x24
+#define IOP_DATA2	0x28
+#define IOP_DATA3	0x2c
+#define IOP_DATA4	0x30
+#define IOP_DATA5	0x34
+#define IOP_DATA6	0x38
+#define IOP_DATA7	0x3c
+#define IOP_DATA8	0x40
+#define IOP_DATA9	0x44
+#define IOP_DATA10	0x48
+#define IOP_DATA11	0x4c
+#define IOP_BASE_ADR_L	0x50
+#define IOP_BASE_ADR_H	0x54
+
+/* PMC register offset */
+#define IOP_PMC_TIMER		0x00
+#define IOP_PMC_CTRL		0x04
+#define IOP_XTAL27M_PASSWORD_I	0x08
+#define IOP_XTAL27M_PASSWORD_II	0x0c
+#define IOP_XTAL32K_PASSWORD_I	0x10
+#define IOP_XTAL32K_PASSWORD_II	0x14
+#define IOP_CLK27M_PASSWORD_I	0x18
+#define IOP_CLK27M_PASSWORD_II	0x1c
+#define IOP_PMC_TIMER2		0x20
+
+#define NORMAL_CODE_MAX_SIZE 0X1000	/* Max size of normal.bin that can be received */
+#define STANDBY_CODE_MAX_SIZE 0x4000	/* Max size of standby.bin that can be received */
+struct sp_iop {
+	struct miscdevice dev;
+	struct mutex write_lock;	/* avoid parallel access */
+	void __iomem *iop_regs;
+	void __iomem *pmc_regs;
+	void __iomem *moon0_regs;
+	int irq;
+	int gpio_wakeup;
+	unsigned char iop_normal_code[NORMAL_CODE_MAX_SIZE];
+	unsigned char iop_standby_code[STANDBY_CODE_MAX_SIZE];
+	resource_size_t iop_mem_start;
+	resource_size_t iop_mem_size;
+	unsigned char bin_code_mode;
+};
+
+static void sp_iop_normal_mode(struct sp_iop *iop)
+{
+	void __iomem *iop_kernel_base;
+	unsigned int reg;
+
+	iop_kernel_base = ioremap(iop->iop_mem_start, NORMAL_CODE_MAX_SIZE);
+	memset(iop_kernel_base, 0, NORMAL_CODE_MAX_SIZE);
+	memcpy(iop_kernel_base, iop->iop_normal_code, NORMAL_CODE_MAX_SIZE);
+
+	writel(0x00100010, iop->moon0_regs + IOP_CLKEN0);
+
+	reg = readl(iop->iop_regs + IOP_CONTROL);
+	reg |= 0x01;
+	writel(reg, iop->iop_regs + IOP_CONTROL);
+
+	reg = readl(iop->iop_regs + IOP_CONTROL);
+	reg &= ~(0x8000);
+	writel(reg, iop->iop_regs + IOP_CONTROL);
+
+	reg = readl(iop->iop_regs + IOP_CONTROL);
+	reg |= 0x0200;// disable watchdog event reset IOP
+	writel(reg, iop->iop_regs + IOP_CONTROL);
+
+	reg = (iop->iop_mem_start & 0xFFFF);
+	writel(reg, iop->iop_regs + IOP_BASE_ADR_L);
+	reg	= (iop->iop_mem_start >> 16);
+	writel(reg, iop->iop_regs + IOP_BASE_ADR_H);
+
+	reg = readl(iop->iop_regs + IOP_CONTROL);
+	reg &= ~(0x01);
+	writel(reg, iop->iop_regs + IOP_CONTROL);
+	iop->bin_code_mode = 0;
+}
+
+static void sp_iop_standby_mode(struct sp_iop *iop)
+{
+	void __iomem *iop_kernel_base;
+	unsigned long reg;
+
+	iop_kernel_base = ioremap(iop->iop_mem_start, STANDBY_CODE_MAX_SIZE);
+	memset(iop_kernel_base, 0, STANDBY_CODE_MAX_SIZE);
+	memcpy(iop_kernel_base, iop->iop_standby_code, STANDBY_CODE_MAX_SIZE);
+
+	writel(0x00100010, iop->moon0_regs + IOP_CLKEN0);
+
+	reg = readl(iop->iop_regs + IOP_CONTROL);
+	reg |= 0x01;
+	writel(reg, iop->iop_regs + IOP_CONTROL);
+
+	reg = readl(iop->iop_regs + IOP_CONTROL);
+	reg &= ~(0x8000);
+	writel(reg, iop->iop_regs + IOP_CONTROL);
+
+	reg = readl(iop->iop_regs + IOP_CONTROL);
+	reg |= 0x0200;// disable watchdog event reset IOP
+	writel(reg, iop->iop_regs + IOP_CONTROL);
+
+	reg = (iop->iop_mem_start & 0xFFFF);
+	writel(reg, iop->iop_regs + IOP_BASE_ADR_L);
+	reg = (iop->iop_mem_start >> 16);
+	writel(reg, iop->iop_regs + IOP_BASE_ADR_H);
+
+	reg = readl(iop->iop_regs + IOP_CONTROL);
+	reg &= ~(0x01);
+	writel(reg, iop->iop_regs + IOP_CONTROL);
+	iop->bin_code_mode = 1;
+}
+
+/* 8051 informs linux kerenl. 8051 has been switched to standby.bin code. */
+#define IOP_READY	0x0004
+#define RISC_READY	0x0008
+
+/* System linux kernel tells 8051 which  gpio pin to wake-up through. */
+#define WAKEUP_PIN	0xFE02
+
+/* System linux kernel tells 8051 to execute S1 or S3 mode. */
+#define S1	0x5331
+#define S3	0x5333
+
+static int sp_iop_s3mode(struct device *dev, struct sp_iop *iop)
+{
+	unsigned int reg;
+	int ret, value;
+
+	/* PMC set */
+	writel(0x00010001, iop->pmc_regs + IOP_PMC_TIMER);
+	reg = readl(iop->pmc_regs + IOP_PMC_CTRL);
+	/* disable system reset PMC, enalbe power down IOP Domain, enable gating clock 27Mhz */
+	reg |= 0x23;
+	writel(reg, iop->pmc_regs + IOP_PMC_CTRL);
+
+	writel(0x55aa00ff, iop->pmc_regs + IOP_XTAL27M_PASSWORD_I);
+	writel(0x00ff55aa, iop->pmc_regs + IOP_XTAL27M_PASSWORD_II);
+	writel(0xaa00ff55, iop->pmc_regs + IOP_XTAL32K_PASSWORD_I);
+	writel(0xff55aa00, iop->pmc_regs + IOP_XTAL32K_PASSWORD_II);
+	writel(0xaaff0055, iop->pmc_regs + IOP_CLK27M_PASSWORD_I);
+	writel(0x5500aaff, iop->pmc_regs + IOP_CLK27M_PASSWORD_II);
+	writel(0x01000100, iop->pmc_regs + IOP_PMC_TIMER2);
+
+	/* IOP Hardware IP reset */
+	reg = readl(iop->moon0_regs + IOP_RESET0);
+	reg |= 0x10;
+	writel(reg, (iop->moon0_regs + IOP_RESET0));
+	reg &= ~(0x10);
+	writel(reg, (iop->moon0_regs + IOP_RESET0));
+
+	writel(0x00ff0085, (iop->moon0_regs + 32 * 4 * 1 + 4 * 1));
+
+	reg = readl(iop->moon0_regs + 32 * 4 * 1 + 4 * 2);
+	reg |= 0x08000800;
+	writel(reg, (iop->moon0_regs + 32 * 4 * 1 + 4 * 2));
+
+	writel(WAKEUP_PIN, iop->iop_regs + IOP_DATA0);
+	writel(iop->gpio_wakeup, iop->iop_regs + IOP_DATA1);
+
+	ret = readl_poll_timeout(iop->iop_regs + IOP_DATA2, value,
+				 (value & IOP_READY) == IOP_READY, 1000, 10000);
+	if (ret) {
+		dev_err(dev, "timed out\n");
+		return ret;
+	}
+
+	reg = RISC_READY;
+	writel(reg, iop->iop_regs + IOP_DATA2);
+	reg = 0x0000;
+	writel(reg, iop->iop_regs + IOP_DATA5);
+	reg = 0x0060;
+	writel(reg, iop->iop_regs + IOP_DATA6);
+
+	ret = readl_poll_timeout(iop->iop_regs + IOP_DATA7, value,
+				 value == 0xaaaa, 1000, 10000);
+	if (ret) {
+		dev_err(dev, "timed out\n");
+		return ret;
+	}
+
+	/* 8051 bin file call Ultra low function. */
+	writel(0xdd, iop->iop_regs + IOP_DATA1);
+	/*
+	 * When the execution is here, the system linux kernel
+	 * is about to be powered off
+	 * The purpose of mdelay(10): Do not let the system linux
+	 * kernel continue to run other programs.
+	 */
+	mdelay(10);
+	return 0;
+}
+
+static int sp_iop_s1mode(struct device *dev, struct sp_iop *iop)
+{
+	int ret, value;
+
+	ret = readl_poll_timeout(iop->iop_regs + IOP_DATA2, value,
+				 (value & IOP_READY) == IOP_READY, 1000, 10000);
+	if (ret) {
+		dev_err(dev, "timed out\n");
+		return ret;
+	}
+
+	writel(RISC_READY, iop->iop_regs + IOP_DATA2);
+	writel(0x0000, iop->iop_regs + IOP_DATA5);
+	writel(0x0060, iop->iop_regs + IOP_DATA6);
+
+	ret = readl_poll_timeout(iop->iop_regs + IOP_DATA7, value,
+				 value == 0xaaaa, 1000, 10000);
+	if (ret) {
+		dev_err(dev, "timed out\n");
+		return ret;
+	}
+
+	/* 8051 bin file call S1_mode function. */
+	writel(0xee, iop->iop_regs + IOP_DATA1);
+	/*
+	 * When the execution is here, the system linux kernel
+	 * is about to be powered off
+	 * The purpose of mdelay(10): Do not let the system linux
+	 * kernel continue to run other programs.
+	 */
+	mdelay(10);
+	return 0;
+}
+
+static ssize_t mailbox_show(struct device *dev, struct device_attribute *attr, char *buf)
+{
+	struct sp_iop *iop = dev_get_drvdata(dev);
+	unsigned int mailbox;
+
+	mailbox = readl(iop->iop_regs + IOP_DATA0);
+	return sysfs_emit(buf, "%x\n", mailbox);
+}
+
+static DEVICE_ATTR_RO(mailbox);
+
+static ssize_t mode_show(struct device *dev, struct device_attribute *attr, char *buf)
+{
+	struct sp_iop *iop = dev_get_drvdata(dev);
+
+	return sysfs_emit(buf, "%x\n", iop->bin_code_mode);
+}
+
+static ssize_t mode_store(struct device *dev, struct device_attribute *attr,
+			  const char *buf, size_t count)
+{
+	struct sp_iop *iop = dev_get_drvdata(dev);
+
+	if (sysfs_streq(buf, "0"))
+		sp_iop_normal_mode(iop);
+	else if (sysfs_streq(buf, "1"))
+		sp_iop_standby_mode(iop);
+	else
+		return -EINVAL;
+	return count;
+}
+
+static DEVICE_ATTR_RW(mode);
+
+static struct attribute *sp_iop_attrs[] = {
+	&dev_attr_mailbox.attr,
+	&dev_attr_mode.attr,
+	NULL,
+};
+
+ATTRIBUTE_GROUPS(sp_iop);
+
+static int  sp_iop_get_normal_code(struct device *dev, struct sp_iop *iop)
+{
+	const struct firmware *fw;
+	static const char file[] = "normal.bin";
+	unsigned int err, i;
+
+	err = request_firmware(&fw, file, dev);
+	if (err) {
+		dev_err(dev, "get bin file error\n");
+		return err;
+	}
+
+	for (i = 0; i < NORMAL_CODE_MAX_SIZE; i++) {
+		char temp;
+
+		temp = fw->data[i];
+		iop->iop_normal_code[i] = temp;
+	}
+	release_firmware(fw);
+	return err;
+}
+
+static int  sp_iop_get_standby_code(struct device *dev, struct sp_iop *iop)
+{
+	const struct firmware *fw;
+	static const char file[] = "standby.bin";
+	unsigned int err, i;
+
+	err = request_firmware(&fw, file, dev);
+	if (err) {
+		dev_err(dev, "get bin file error\n");
+		return err;
+	}
+
+	for (i = 0; i < STANDBY_CODE_MAX_SIZE; i++) {
+		char temp;
+
+		temp = fw->data[i];
+		iop->iop_standby_code[i] = temp;
+	}
+	release_firmware(fw);
+	return err;
+}
+
+static int sp_iop_get_resources(struct platform_device *pdev, struct sp_iop *p_sp_iop_info)
+{
+	struct resource *r;
+
+	r = platform_get_resource_byname(pdev, IORESOURCE_MEM, "iop");
+	p_sp_iop_info->iop_regs = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(p_sp_iop_info->iop_regs)) {
+		dev_err(&pdev->dev, "ioremap fail\n");
+		return PTR_ERR(p_sp_iop_info->iop_regs);
+	}
+
+	r = platform_get_resource_byname(pdev, IORESOURCE_MEM, "iop_pmc");
+	p_sp_iop_info->pmc_regs = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(p_sp_iop_info->pmc_regs)) {
+		dev_err(&pdev->dev, "ioremap fail\n");
+		return PTR_ERR(p_sp_iop_info->pmc_regs);
+	}
+
+	r = platform_get_resource_byname(pdev, IORESOURCE_MEM, "moon0");
+	p_sp_iop_info->moon0_regs = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(p_sp_iop_info->moon0_regs)) {
+		dev_err(&pdev->dev, "ioremap fail\n");
+		return PTR_ERR(p_sp_iop_info->moon0_regs);
+	}
+	return IOP_SUCCESS;
+}
+
+static int sp_iop_platform_driver_probe(struct platform_device *pdev)
+{
+	int ret = -ENXIO;
+	int rc;
+	struct sp_iop *iop;
+	struct device_node *memnp;
+	struct resource mem_res;
+
+	iop = devm_kzalloc(&pdev->dev, sizeof(struct sp_iop), GFP_KERNEL);
+	if (!iop) {
+		ret	= -ENOMEM;
+		goto fail_kmalloc;
+	}
+	/* init */
+	mutex_init(&iop->write_lock);
+	ret = sp_iop_get_resources(pdev, iop);
+
+	// Get reserve address
+	memnp = of_parse_phandle(pdev->dev.of_node, "memory-region", 0);
+	if (!memnp) {
+		dev_err(&pdev->dev, "no memory-region node\n");
+		return -EINVAL;
+	}
+
+	rc = of_address_to_resource(memnp, 0, &mem_res);
+	of_node_put(memnp);
+	if (rc) {
+		dev_err(&pdev->dev, "failed to translate memory-region to a resource\n");
+		return -EINVAL;
+	}
+
+	iop->iop_mem_start = mem_res.start;
+	iop->iop_mem_size = resource_size(&mem_res);
+
+	ret = sp_iop_get_normal_code(&pdev->dev, iop);
+	if (ret != 0) {
+		dev_err(&pdev->dev, "get normal code err=%d\n", ret);
+		return ret;
+	}
+
+	ret = sp_iop_get_standby_code(&pdev->dev, iop);
+	if (ret != 0) {
+		dev_err(&pdev->dev, "get standby code err=%d\n", ret);
+		return ret;
+	}
+
+	sp_iop_normal_mode(iop);
+	platform_set_drvdata(pdev, iop);
+	iop->gpio_wakeup = of_get_named_gpio(pdev->dev.of_node, "iop-wakeup", 0);
+	return 0;
+
+fail_kmalloc:
+	return ret;
+}
+
+static void sp_iop_platform_driver_shutdown(struct platform_device *pdev)
+{
+	struct sp_iop *iop = platform_get_drvdata(pdev);
+	unsigned int ret, value;
+
+	value = readl(iop->iop_regs + IOP_DATA11);
+	sp_iop_standby_mode(iop);
+
+	ret = readl_poll_timeout(iop->iop_regs + IOP_DATA0, value,
+				 value == 0x2222, 1000, 100000);
+	if (ret)
+		dev_warn(&pdev->dev, "timed out\n");
+
+	if (value == S1)
+		sp_iop_s1mode(&pdev->dev, iop);
+	else
+		sp_iop_s3mode(&pdev->dev, iop);
+}
+
+static const struct of_device_id sp_iop_of_match[] = {
+	{ .compatible = "sunplus,sp7021-iop" },
+	{ /* sentinel */ },
+};
+
+MODULE_DEVICE_TABLE(of, sp_iop_of_match);
+
+static struct platform_driver sp_iop_platform_driver = {
+	.probe		= sp_iop_platform_driver_probe,
+	.shutdown	= sp_iop_platform_driver_shutdown,
+	.driver = {
+		.name	= "sunplus,sp7021-iop",
+		.of_match_table = sp_iop_of_match,
+		.dev_groups = sp_iop_groups,
+	}
+};
+
+module_platform_driver(sp_iop_platform_driver);
+
+MODULE_AUTHOR("Tony Huang <tonyhuang.sunplus@gmail.com>");
+MODULE_DESCRIPTION("Sunplus IOP Driver");
+MODULE_LICENSE("GPL v2");
-- 
2.7.4

Re: [PATCH v8 2/2] misc: Add iop driver for Sunplus SP7021

[Arnd Bergmann]

On Mon, Feb 7, 2022 at 7:30 AM Tony Huang <tonyhuang.sunplus@gmail.com> wrote:
>
> IOP(8051) embedded inside SP7021 which is used as
> Processor for I/O control, monitor RTC interrupt and
> cooperation with CPU & PMC in power management purpose.
> The IOP core is DQ8051, so also named IOP8051,
> it supports dedicated JTAG debug pins which share with SP7021.
> In standby mode operation, the power spec reach 400uA.
>
> Signed-off-by: Tony Huang <tonyhuang.sunplus@gmail.com>
> ---
> Changes in v8:
>  - Addressed comments from Greg KH.
>
>  Documentation/ABI/testing/sysfs-platform-soc@B |  28 ++
>  MAINTAINERS                                    |   2 +
>  drivers/misc/Kconfig                           |  20 ++
>  drivers/misc/Makefile                          |   1 +
>  drivers/misc/sunplus_iop.c                     | 463 +++++++++++++++++++++++++
>  5 files changed, 514 insertions(+)
>  create mode 100644 Documentation/ABI/testing/sysfs-platform-soc@B
>  create mode 100644 drivers/misc/sunplus_iop.c
>
> diff --git a/Documentation/ABI/testing/sysfs-platform-soc@B b/Documentation/ABI/testing/sysfs-platform-soc@B
> new file mode 100644
> index 0000000..d26d6f5
> --- /dev/null
> +++ b/Documentation/ABI/testing/sysfs-platform-soc@B
> @@ -0,0 +1,28 @@
> +What:          /sys/devices/platform/soc@B/9c000400.iop/sp_iop_mailbox
> +Date:          January 2022
> +KernelVersion: 5.16
> +Contact:       Tony Huang <tonyhuang.sunplus@gmail.com>
> +Description:
> +               Show IOP's mailbox0 register data.
> +               Format: %x
> +
> +What:          /sys/devices/platform/soc@B/9c000400.iop/sp_iop_mode
> +Date:          January 2022
> +KernelVersion: 5.16
> +Contact:       Tony Huang <tonyhuang.sunplus@gmail.com>
> +Description:
> +               Read-Write.
> +
> +               Write this file.
> +               Operation mode of IOP is switched to standby mode by writing
> +               "1" to sysfs.
> +               Operation mode of IOP is switched to normal mode by writing
> +               "0" to sysfs.
> +               Writing of other values is invalid.
> +
> +               Read this file.
> +               Show operation mode of IOP. "0" is normal mode. "1" is standby
> +               mode.
> +               Format: %x

As discussed before, I would suggest leaving out all custom attributes for now,
and first hooking up the driver to all the in-kernel subsystems.

The mailbox0 register data definitely feels like an implementation detail,
not something that should be exposed to user space as an
interface.

For standby mode, this would normally be handled by the
power management subsystem in the kernel. not a custom
interface. From your earlier description, I assume this interface
puts the main CPU into standby mode, not the IOP, right?

CPU standby is handled by the cpuidle subsystem, so you
need a driver in drivers/cpuidle/ to replace your sysfs attribute.
If you plan to hook up the driver to multiple subsystems, keeping
a generic driver file is ok, so  you'll end up with two driver
modules, with one of them calling into the other, using
EXPORT_SYMBOL() to link between them.

         Arnd

RE: [PATCH v8 2/2] misc: Add iop driver for Sunplus SP7021

[Tony Huang 黃懷厚]

Dear Arnd:

> -----Original Message-----
> From: Arnd Bergmann <arnd@arndb.de>
> Sent: Monday, February 7, 2022 3:48 PM
> To: Tony Huang <tonyhuang.sunplus@gmail.com>
> Cc: Rob Herring <robh+dt@kernel.org>; DTML <devicetree@vger.kernel.org>;
> Linux Kernel Mailing List <linux-kernel@vger.kernel.org>; Derek Kiernan
> <derek.kiernan@xilinx.com>; Dragan Cvetic <dragan.cvetic@xilinx.com>; Arnd
> Bergmann <arnd@arndb.de>; gregkh <gregkh@linuxfoundation.org>; Tony
> Huang 黃懷厚 <tony.huang@sunplus.com>; Wells Lu 呂芳騰
> <wells.lu@sunplus.com>
> Subject: Re: [PATCH v8 2/2] misc: Add iop driver for Sunplus SP7021
> 
> On Mon, Feb 7, 2022 at 7:30 AM Tony Huang
> <tonyhuang.sunplus@gmail.com> wrote:
> >
> > IOP(8051) embedded inside SP7021 which is used as Processor for I/O
> > control, monitor RTC interrupt and cooperation with CPU & PMC in power
> > management purpose.
> > The IOP core is DQ8051, so also named IOP8051, it supports dedicated
> > JTAG debug pins which share with SP7021.
> > In standby mode operation, the power spec reach 400uA.
> >
> > Signed-off-by: Tony Huang <tonyhuang.sunplus@gmail.com>
> > ---
> > Changes in v8:
> >  - Addressed comments from Greg KH.
> >
> >  Documentation/ABI/testing/sysfs-platform-soc@B |  28 ++
> >  MAINTAINERS                                    |   2 +
> >  drivers/misc/Kconfig                           |  20 ++
> >  drivers/misc/Makefile                          |   1 +
> >  drivers/misc/sunplus_iop.c                     | 463
> +++++++++++++++++++++++++
> >  5 files changed, 514 insertions(+)
> >  create mode 100644 Documentation/ABI/testing/sysfs-platform-soc@B
> >  create mode 100644 drivers/misc/sunplus_iop.c
> >
> > diff --git a/Documentation/ABI/testing/sysfs-platform-soc@B
> > b/Documentation/ABI/testing/sysfs-platform-soc@B
> > new file mode 100644
> > index 0000000..d26d6f5
> > --- /dev/null
> > +++ b/Documentation/ABI/testing/sysfs-platform-soc@B
> > @@ -0,0 +1,28 @@
> > +What:
> /sys/devices/platform/soc@B/9c000400.iop/sp_iop_mailbox
> > +Date:          January 2022
> > +KernelVersion: 5.16
> > +Contact:       Tony Huang <tonyhuang.sunplus@gmail.com>
> > +Description:
> > +               Show IOP's mailbox0 register data.
> > +               Format: %x
> > +
> > +What:
> /sys/devices/platform/soc@B/9c000400.iop/sp_iop_mode
> > +Date:          January 2022
> > +KernelVersion: 5.16
> > +Contact:       Tony Huang <tonyhuang.sunplus@gmail.com>
> > +Description:
> > +               Read-Write.
> > +
> > +               Write this file.
> > +               Operation mode of IOP is switched to standby mode by
> writing
> > +               "1" to sysfs.
> > +               Operation mode of IOP is switched to normal mode by
> writing
> > +               "0" to sysfs.
> > +               Writing of other values is invalid.
> > +
> > +               Read this file.
> > +               Show operation mode of IOP. "0" is normal mode. "1" is
> standby
> > +               mode.
> > +               Format: %x
> 
> As discussed before, I would suggest leaving out all custom attributes for now,
> and first hooking up the driver to all the in-kernel subsystems.
> 
> The mailbox0 register data definitely feels like an implementation detail, not
> something that should be exposed to user space as an interface.
> 
> For standby mode, this would normally be handled by the power management
> subsystem in the kernel. not a custom interface. From your earlier description,
> I assume this interface puts the main CPU into standby mode, not the IOP,
> right?
> 
> CPU standby is handled by the cpuidle subsystem, so you need a driver in
> drivers/cpuidle/ to replace your sysfs attribute.
> If you plan to hook up the driver to multiple subsystems, keeping a generic
> driver file is ok, so  you'll end up with two driver modules, with one of them
> calling into the other, using
> EXPORT_SYMBOL() to link between them.
> 

The purpose of adding sysfs is only for users to debug.
So this is not needed?

Thanks

Re: [PATCH v8 2/2] misc: Add iop driver for Sunplus SP7021

[gregkh]

On Mon, Feb 07, 2022 at 08:29:40AM +0000, Tony Huang 黃懷厚 wrote:
> Dear Arnd:
> 
> > -----Original Message-----
> > From: Arnd Bergmann <arnd@arndb.de>
> > Sent: Monday, February 7, 2022 3:48 PM
> > To: Tony Huang <tonyhuang.sunplus@gmail.com>
> > Cc: Rob Herring <robh+dt@kernel.org>; DTML <devicetree@vger.kernel.org>;
> > Linux Kernel Mailing List <linux-kernel@vger.kernel.org>; Derek Kiernan
> > <derek.kiernan@xilinx.com>; Dragan Cvetic <dragan.cvetic@xilinx.com>; Arnd
> > Bergmann <arnd@arndb.de>; gregkh <gregkh@linuxfoundation.org>; Tony
> > Huang 黃懷厚 <tony.huang@sunplus.com>; Wells Lu 呂芳騰
> > <wells.lu@sunplus.com>
> > Subject: Re: [PATCH v8 2/2] misc: Add iop driver for Sunplus SP7021
> > 
> > On Mon, Feb 7, 2022 at 7:30 AM Tony Huang
> > <tonyhuang.sunplus@gmail.com> wrote:
> > >
> > > IOP(8051) embedded inside SP7021 which is used as Processor for I/O
> > > control, monitor RTC interrupt and cooperation with CPU & PMC in power
> > > management purpose.
> > > The IOP core is DQ8051, so also named IOP8051, it supports dedicated
> > > JTAG debug pins which share with SP7021.
> > > In standby mode operation, the power spec reach 400uA.
> > >
> > > Signed-off-by: Tony Huang <tonyhuang.sunplus@gmail.com>
> > > ---
> > > Changes in v8:
> > >  - Addressed comments from Greg KH.
> > >
> > >  Documentation/ABI/testing/sysfs-platform-soc@B |  28 ++
> > >  MAINTAINERS                                    |   2 +
> > >  drivers/misc/Kconfig                           |  20 ++
> > >  drivers/misc/Makefile                          |   1 +
> > >  drivers/misc/sunplus_iop.c                     | 463
> > +++++++++++++++++++++++++
> > >  5 files changed, 514 insertions(+)
> > >  create mode 100644 Documentation/ABI/testing/sysfs-platform-soc@B
> > >  create mode 100644 drivers/misc/sunplus_iop.c
> > >
> > > diff --git a/Documentation/ABI/testing/sysfs-platform-soc@B
> > > b/Documentation/ABI/testing/sysfs-platform-soc@B
> > > new file mode 100644
> > > index 0000000..d26d6f5
> > > --- /dev/null
> > > +++ b/Documentation/ABI/testing/sysfs-platform-soc@B
> > > @@ -0,0 +1,28 @@
> > > +What:
> > /sys/devices/platform/soc@B/9c000400.iop/sp_iop_mailbox
> > > +Date:          January 2022
> > > +KernelVersion: 5.16
> > > +Contact:       Tony Huang <tonyhuang.sunplus@gmail.com>
> > > +Description:
> > > +               Show IOP's mailbox0 register data.
> > > +               Format: %x
> > > +
> > > +What:
> > /sys/devices/platform/soc@B/9c000400.iop/sp_iop_mode
> > > +Date:          January 2022
> > > +KernelVersion: 5.16
> > > +Contact:       Tony Huang <tonyhuang.sunplus@gmail.com>
> > > +Description:
> > > +               Read-Write.
> > > +
> > > +               Write this file.
> > > +               Operation mode of IOP is switched to standby mode by
> > writing
> > > +               "1" to sysfs.
> > > +               Operation mode of IOP is switched to normal mode by
> > writing
> > > +               "0" to sysfs.
> > > +               Writing of other values is invalid.
> > > +
> > > +               Read this file.
> > > +               Show operation mode of IOP. "0" is normal mode. "1" is
> > standby
> > > +               mode.
> > > +               Format: %x
> > 
> > As discussed before, I would suggest leaving out all custom attributes for now,
> > and first hooking up the driver to all the in-kernel subsystems.
> > 
> > The mailbox0 register data definitely feels like an implementation detail, not
> > something that should be exposed to user space as an interface.
> > 
> > For standby mode, this would normally be handled by the power management
> > subsystem in the kernel. not a custom interface. From your earlier description,
> > I assume this interface puts the main CPU into standby mode, not the IOP,
> > right?
> > 
> > CPU standby is handled by the cpuidle subsystem, so you need a driver in
> > drivers/cpuidle/ to replace your sysfs attribute.
> > If you plan to hook up the driver to multiple subsystems, keeping a generic
> > driver file is ok, so  you'll end up with two driver modules, with one of them
> > calling into the other, using
> > EXPORT_SYMBOL() to link between them.
> > 
> 
> The purpose of adding sysfs is only for users to debug.
> So this is not needed?

If this is only for debugging, please put it in debugfs and not sysfs.

thanks,

greg k-h

Re: [PATCH v8 2/2] misc: Add iop driver for Sunplus SP7021

[Arnd Bergmann]

On Mon, Feb 7, 2022 at 9:39 AM gregkh <gregkh@linuxfoundation.org> wrote:
> On Mon, Feb 07, 2022 at 08:29:40AM +0000, Tony Huang 黃懷厚 wrote:
> > >
> > > As discussed before, I would suggest leaving out all custom attributes for now,
> > > and first hooking up the driver to all the in-kernel subsystems.
> > >
> > > The mailbox0 register data definitely feels like an implementation detail, not
> > > something that should be exposed to user space as an interface.
> > >
> > > For standby mode, this would normally be handled by the power management
> > > subsystem in the kernel. not a custom interface. From your earlier description,
> > > I assume this interface puts the main CPU into standby mode, not the IOP,
> > > right?
> > >
> > > CPU standby is handled by the cpuidle subsystem, so you need a driver in
> > > drivers/cpuidle/ to replace your sysfs attribute.
> > > If you plan to hook up the driver to multiple subsystems, keeping a generic
> > > driver file is ok, so  you'll end up with two driver modules, with one of them
> > > calling into the other, using
> > > EXPORT_SYMBOL() to link between them.
> > >
> >
> > The purpose of adding sysfs is only for users to debug.
> > So this is not needed?
>
> If this is only for debugging, please put it in debugfs and not sysfs.

I don't think that works for the idle mode interface, as this is the only
thing the driver really does at the moment.

In a previous review round, I already asked for the driver to implement
at least two in-kernel interfaces before any custom user interface
is added.

         Arnd