Preliminary Platform Driver Support for Lemote Yeeloong Laptops
Subsystems (3)
v2 cover Preliminary Platform Driver Support for Lemote Yeeloong Laptops
v2:
- Fix the missing io.h header for the MFD driver.
- Use SIMPLE_DEV_PM_OPS instead of the inappropriate .suspend/resume in the SCI driver.
Lemote Yeeloong is a laptop powered by Loongson 2F MIPS processor, primarily a demo platform for hobbyists and developers. It uses an ENE KB3310B Embedded Controller with customized firmware to implement hardware and power management.
Due to the lack of separation of responsibilities and functions between different subsystems, the original platform driver submitted to the mailing list was a piece of monolithic module that implements numerous drivers, as a result, it was never accepted.
This patch series was carefully reorganized and partially reimplemented to avoid those pitfalls in the original driver. This is the first part of the submission, which modifies the core MIPS code to introduce preliminary preparation for other subdrivers. Notably, it leverages a notify chain and MFD cells to avoid introducing unnecessary code into the core MIPS subsystem.
It features the following...
* Create a MFD driver for KB3310B controller, and move the original KB3310B controller code from mips/loongson64 to our new MFD driver. This needs to be reviewed by the MFD subsystem maintainer before the following can proceed.
* Add yeeloong_sci driver support of System Control Interrupt to arch/mips/loongson64/lemote-2f. This piece of code is crucial and has to be included in loongson64/lemote-2f because it reprograms CS5536 southbridge GPIO to handle the underlying interrupts.
* Originally, the SCI code was mixed into the hotkey driver. The new yeeloong_sci driver was written to avoid introducing unrelated subdriver code that is unsuitable for Linux/MIPS. It only handle a minimum set of core operations directly related to the hardware platform, and pass the events to other subdrivers via a notifier chain. As a result, the SCI logic and the specific hotkey handling logic have been decoupled.
* Subdrivers - hwmon, battery, backlight, lcd and hotkey are registered as MFD cells in the MFD driver. It means onlf the MFD driver is resposible to register the upcoming subdrivers, the core board files mips/loongson64/ will not contain unrelated code.
I've done everything I can think of to factor unrelated code out of the MIPS subsystem. I hope the refactored code would be acceptable now. Please review my changeset to see whether it's appropriate.
Thanks,
Yifeng Li (7):
mfd: yeeloong_kb3310b: support KB3310B EC for Lemote Yeeloong laptops.
mips: loongson64: select MFD_YEELOONG_KB3310B for LEMOTE_MACH2F.
mips: loongson64: remove ec_kb3310b.c, use MFD driver.
mips: loongson64: remove yeeloong_report_lid_status from pm.c
mips: loongson64: register per-board platform drivers for lemote-2f
mips: loongson64: Support System Control Interrupts for Lemote
Yeeloong.
MAINTAINERS: add myself as a maintainer of MIPS/Loongson2 platform
code.
MAINTAINERS | 16 + .../include/asm/mach-loongson64/loongson.h | 3 + arch/mips/loongson64/Kconfig | 1 + arch/mips/loongson64/common/platform.c | 15 + arch/mips/loongson64/lemote-2f/Makefile | 2 +- arch/mips/loongson64/lemote-2f/ec_kb3310b.c | 129 ------ arch/mips/loongson64/lemote-2f/ec_kb3310b.h | 188 --------- arch/mips/loongson64/lemote-2f/platform.c | 47 +++ arch/mips/loongson64/lemote-2f/pm.c | 38 +- arch/mips/loongson64/lemote-2f/reset.c | 4 +- arch/mips/loongson64/lemote-2f/sci.c | 394 ++++++++++++++++++ drivers/mfd/Kconfig | 10 + drivers/mfd/Makefile | 1 + drivers/mfd/yeeloong_kb3310b.c | 207 +++++++++ include/linux/mfd/yeeloong_kb3310b.h | 211 ++++++++++ 15 files changed, 916 insertions(+), 350 deletions(-) delete mode 100644 arch/mips/loongson64/lemote-2f/ec_kb3310b.c delete mode 100644 arch/mips/loongson64/lemote-2f/ec_kb3310b.h create mode 100644 arch/mips/loongson64/lemote-2f/platform.c create mode 100644 arch/mips/loongson64/lemote-2f/sci.c create mode 100644 drivers/mfd/yeeloong_kb3310b.c create mode 100644 include/linux/mfd/yeeloong_kb3310b.h
Signature · 1 line
2.20.1
PATCH v2 1/7 mfd: yeeloong_kb3310b: support KB3310B EC for Lemote Yeeloong laptops.
Lemote Yeeloong is a laptop powered by Loongson 2F MIPS processor, primarily a demo platform for hobbyists and developers. It uses an ENE KB3310B Embedded Controller with customized firmware to implement hardware and power management.
A monolithic platform driver code for those functionality has existed out-of-tree for many years. This commit creates a MFD driver for the EC chip on Yeeloong laptop to isolate EC-related code from core MIPS code, and serves as the foundation of various subdrivers.
My original attempt was to create a regmap for subdrivers to access the EC, unfortunately, the board files in Linux/MIPS still needs to access the EC directly for power management. Unless we find a better home for those code, we simply export the EC-related functions.
--- MAINTAINERS | 7 + drivers/mfd/Kconfig | 10 ++ drivers/mfd/Makefile | 1 + drivers/mfd/yeeloong_kb3310b.c | 207 ++++++++++++++++++++++++++ include/linux/mfd/yeeloong_kb3310b.h | 211 +++++++++++++++++++++++++++ 5 files changed, 436 insertions(+) create mode 100644 drivers/mfd/yeeloong_kb3310b.c create mode 100644 include/linux/mfd/yeeloong_kb3310b.h
MAINTAINERS +7 -0
@@ -16839,6 +16839,13 @@ S: Maintained F: Documentation/input/devices/yealink.rst F: drivers/input/misc/yealink.* +YEELOONG ENE KB3310B MFD DRIVER +M: Tom Li <[email protected]> +L: [email protected] +S: Maintained +F: drivers/mfd/yeeloong_kb3310b.c +F: include/linux/mfd/yeeloong_kb3310b.h + Z8530 DRIVER FOR AX.25 M: Joerg Reuter <[email protected]> W: http://yaina.de/jreuter/
drivers/mfd/Kconfig +10 -0
@@ -1872,6 +1872,16 @@ config MFD_STM32_TIMERS for PWM and IIO Timer. This driver allow to share the registers between the others drivers. +config MFD_YEELOONG_KB3310B + bool "ENE KB3310B Embedded Controller on Lemote Yeeloong laptops" + depends on (MIPS && LEMOTE_MACH2F) || COMPILE_TEST + select MFD_CORE + help + Select this option to enable ENE KB3310B Embedded Controller + driver used on Lemote Yeeloong laptops, providing power, battery + and backlight services. This is a mandatory dependency for + Lemote 2F systems. + menu "Multimedia Capabilities Port drivers" depends on ARCH_SA1100
drivers/mfd/Makefile +1 -0
@@ -224,6 +224,7 @@ obj-$(CONFIG_MFD_HI655X_PMIC) += hi655x-pmic.o obj-$(CONFIG_MFD_DLN2) += dln2.o obj-$(CONFIG_MFD_RT5033) += rt5033.o obj-$(CONFIG_MFD_SKY81452) += sky81452.o +obj-$(CONFIG_MFD_YEELOONG_KB3310B) += yeeloong_kb3310b.o intel-soc-pmic-objs := intel_soc_pmic_core.o intel_soc_pmic_crc.o obj-$(CONFIG_INTEL_SOC_PMIC) += intel-soc-pmic.o
drivers/mfd/yeeloong_kb3310b.c new file +207 -0
207 lines not on this page · raw message
include/linux/mfd/yeeloong_kb3310b.h new file +211 -0
211 lines not on this page · raw message
Signature · 1 line
2.20.1
Re: [PATCH v2 1/7] mfd: yeeloong_kb3310b: support KB3310B EC for Lemote Yeeloong laptops.
Hi Tom,
Overall I think this is much better than the older out of tree platform driver - thanks for cleaning it up.
One general comment - it would be good to run the patches through ./scripts/checkpatch.pl & fix up any warnings it gives unless you have a good reason to disagree with them.
Quoted 10 lines · Yifeng Li
Lemote Yeeloong is a laptop powered by Loongson 2F MIPS processor, primarily a demo platform for hobbyists and developers. It uses an ENE KB3310B Embedded Controller with customized firmware to implement hardware and power management.
A monolithic platform driver code for those functionality has existed out-of-tree for many years. This commit creates a MFD driver for the EC chip on Yeeloong laptop to isolate EC-related code from core MIPS code, and serves as the foundation of various subdrivers.
Quoted 4 lines
My original attempt was to create a regmap for subdrivers to access the EC, unfortunately, the board files in Linux/MIPS still needs to access the EC directly for power management. Unless we find a better home for those code, we simply export the EC-related functions.
I'm not sure I follow why the power management code prevents use of regmap?
Are you talking about the wakeup_loongson() function? Perhaps it would make sense for the suspend code to be part of one of the possible subdrivers you mention. The lemote-2f seems to be the only system that provides an implementation of wakeup_loongson() so perhaps a driver could instead just register its own struct platform_suspend_ops & avoid the need for code in arch/mips to care about the EC.
Quoted 28 lines
drivers/mfd/yeeloong_kb3310b.c new file +22
-0@@ -0,0 +1,207 @@ +// SPDX-License-Identifier: GPL-2.0-or-later + +/* +* MFD driver for ENE KB3310B embedded controller on Lemote Yeeloong laptops +* +* Copyright (C) 2008 Lemote Inc. +* Author: liujl <[email protected]>, 2008-04-20 +* +* Copyright (C) 2018 Yifeng Li +* Author: Yifeng Li <[email protected]> +* +* This is a MFD driver for the ENE KB3310B Embedded Controller for Lemote +* Yeeloong laptops to provide utility functions to access the chip from +* subdrivers, and handle events and interrupts in board files. This is a +* special-purpose driver, and it's only used on Lemote Yeeloong laptops, +* and is a mandatory dependency. +* +* My original attempt was to create a regmap for subdrivers to access the +* EC, unfortunately, the board files in Linux/MIPS still needs to access +* the EC directly for power management. Unless we find a better home for +* those code, we simply export the EC-related functions. +*/
drivers/mfd/yeeloong_kb3310b.c line 32
Quoted 10 lines
drivers/mfd/yeeloong_kb3310b.c new file +10
-0+ +#include <linux/export.h> +#include <linux/platform_device.h> +#include <linux/mfd/core.h> +#include <linux/io.h> +#include <linux/delay.h> + +#include <linux/mfd/yeeloong_kb3310b.h> + +#define DRV_NAME "yeeloong_kb3310b: "
Defining pr_fmt() would be cleaner - you wouldn't need to manually include DRV_NAME in your messages later.
drivers/mfd/yeeloong_kb3310b.c line 38
Quoted 6 lines
drivers/mfd/yeeloong_kb3310b.c new file +6
-0+ +/***************************************************************************** +* Most drivers, such as battery or backlight drivers, uses the I/O ports to +* access the Index Registers to obtain hardware status and information from +* EC chip. +****************************************************************************/
Nit: the lines of asterisks aren't part of the kernel's general comment style & I think it would looks cleaner to remove them.
Quoted 11 lines
drivers/mfd/yeeloong_kb3310b.c new file +11
-0+static struct kb3310b_chip *kb3310b_fwinfo; + +static const struct mfd_cell kb3310b_cells[] = { + { + .name ="yeeloong_sci" + }, + { + .name ="yeeloong_hwmon" + }, + { + .name ="yeeloong_battery"
drivers/mfd/yeeloong_kb3310b.c line 59
Quoted 10 lines
drivers/mfd/yeeloong_kb3310b.c new file +10
-0+ }, + { + .name = "yeeloong_backlight" + }, + { + .name ="yeeloong_lcd" + }, + { + .name ="yeeloong_hotkey" + },
Nit: I think it'd look cleaner if you remove the newlines within each array entry, eg:
{ .name = "yeeloong_sci" },
{ .name = "yeelong_hwmon" },
...Quoted 25 lines
drivers/mfd/yeeloong_kb3310b.c new file +25
-0+}; + +static DEFINE_SPINLOCK (kb3310b_index_lock); + +u8 kb3310b_read (u16 reg) +{ +unsigned long flags; +u8 val; + +spin_lock_irqsave (&kb3310b_index_lock, flags); + +outb ((reg &0xff00 ) >>8 ,KB3310B_IO_PORT_HIGH ); +outb ((reg &0x00ff ),KB3310B_IO_PORT_LOW ); + val =inb (KB3310B_IO_PORT_DATA ); + +spin_unlock_irqrestore (&kb3310b_index_lock, flags); + +return val; +} +EXPORT_SYMBOL_GPL (kb3310b_read); + +void kb3310b_write (u16 reg,u8 val) +{ +unsigned long flags; +
drivers/mfd/yeeloong_kb3310b.c line 94
Quoted 10 lines
drivers/mfd/yeeloong_kb3310b.c new file +10
-0+spin_lock_irqsave (&kb3310b_index_lock, flags); + +outb ((reg &0xff00 ) >>8 ,KB3310B_IO_PORT_HIGH ); +outb ((reg &0x00ff ),KB3310B_IO_PORT_LOW ); +outb (val,KB3310B_IO_PORT_DATA ); +inb (KB3310B_IO_PORT_DATA );/* flush pending writes */ + +spin_unlock_irqrestore (&kb3310b_index_lock, flags); +} +EXPORT_SYMBOL_GPL (kb3310b_write);
If you do switch to using regmap then the above could be neatly represented as a bus - see struct regmap_bus, and various users of it throughout drivers/. regmap would even handle the locking for you.
Quoted 20 lines
drivers/mfd/yeeloong_kb3310b.c new file +20
-0+ +bool kb3310b_fw_earlier (char *version) +{ +return (strncasecmp (kb3310b_fwinfo->version, + version,KB3310B_VERSION_LEN ) <0 ); +} +EXPORT_SYMBOL_GPL (kb3310b_fw_earlier); + +static int kb3310b_probe (struct platform_device *pdev) +{ + kb3310b_fwinfo =dev_get_platdata (&pdev->dev); +pr_info (DRV_NAME "firmware version %s" , kb3310b_fwinfo->version); + +return devm_mfd_add_devices (&pdev->dev,-1 , kb3310b_cells, +ARRAY_SIZE (kb3310b_cells),NULL ,0 ,NULL ); +} + +static struct platform_driver kb3310b_driver = { + .driver = { + .name ="yeeloong_kb3310b" ,
drivers/mfd/yeeloong_kb3310b.c line 124
Quoted 10 lines
drivers/mfd/yeeloong_kb3310b.c new file +10
-0+ }, + .probe = kb3310b_probe, +}; +builtin_platform_driver (kb3310b_driver); + +/***************************************************************************** +* For interrupt handling and power management, the EC chip is also needed to +* be queried from the board file at arch/mips/loongson64, through a separate +* command port. +*****************************************************************************/
As before, I think it'd be cleaner without the lines of asterisks. Probably without the empty line before the variable too.
drivers/mfd/yeeloong_kb3310b.c line 126
Quoted 2 lines
drivers/mfd/yeeloong_kb3310b.c new file +2
-0+ +static DEFINE_SPINLOCK (kb3310b_command_lock);
Since this is only used in kb3310b_query_seq() could you just declare it (still static) inside that function?
Thanks,
Paul
Quoted 301 lines
drivers/mfd/yeeloong_kb3310b.c new file +81
-0+ +/* +* This function is used for EC command writes and corresponding status queries. +*/ +int kb3310b_query_seq (unsigned char cmd) +{ +int timeout; +unsigned char status; +unsigned long flags; + +spin_lock_irqsave (&kb3310b_command_lock, flags); + +/* make chip goto reset mode */ +udelay (KB3310B_REG_UDELAY ); +outb (cmd,KB3310B_CMD_PORT ); +udelay (KB3310B_REG_UDELAY ); + +/* check if the command is received by EC */ + timeout =KB3310B_CMD_TIMEOUT ; + status =inb (KB3310B_STS_PORT ); +while (timeout-- && (status & (1 <<1 ))) { + status =inb (KB3310B_STS_PORT ); +udelay (KB3310B_REG_UDELAY ); + } + +spin_unlock_irqrestore (&kb3310b_command_lock, flags); + +if (timeout <=0 ) { +pr_err (DRV_NAME +"(%x/NA) failed to issue command %d, no response!\n" , + timeout, cmd); +return -EINVAL ; + } + +pr_info (DRV_NAME +"(%x/%x) issued command %d, status: 0x%x\n" , + timeout,KB3310B_CMD_TIMEOUT - timeout, + cmd, status); + +return 0 ; +} +EXPORT_SYMBOL_GPL (kb3310b_query_seq); + +/* +* Send query command to EC to get the proper event number. +*/ +int kb3310b_query_event_num (void ) +{ +return kb3310b_query_seq (KB3310B_CMD_GET_EVENT_NUM ); +} +EXPORT_SYMBOL_GPL (kb3310b_query_event_num); + +/* +* Get event number from EC. +* +* NOTE: This routine must follow the query_event_num function in the +* interrupt. +*/ +int kb3310b_get_event_num (void ) +{ +int timeout =100 ; +unsigned char value; +unsigned char status; + +udelay (KB3310B_REG_UDELAY ); + status =inb (KB3310B_STS_PORT ); +udelay (KB3310B_REG_UDELAY ); +while (timeout-- && !(status & (1 <<0 ))) { + status =inb (KB3310B_STS_PORT ); +udelay (KB3310B_REG_UDELAY ); + } +if (timeout <=0 ) { +pr_info ("%s: get event number timeout.\n" , __func__); +return -EINVAL ; + } + value =inb (KB3310B_DAT_PORT ); +udelay (KB3310B_REG_UDELAY ); + +return value; +} +EXPORT_SYMBOL_GPL (kb3310b_get_event_num);include/linux/mfd/yeeloong_kb3310b.h new file +211
-0@@ -0,0 +1,211 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ + +/* +* MFD driver for ENE KB3310B embedded controller on Lemote Yeeloong laptops +* +* Copyright (C) 2008 Lemote Inc. +* Author: liujl <[email protected]>, 2008-04-20 +* +* Copyright (C) 2018 Yifeng Li +* Author: Yifeng Li <[email protected]> +*/ + +#ifndef __LINUX_MFD_YEELOONG_KB3310B_H +#define __LINUX_MFD_YEELOONG_KB3310B_H + +extern u8 kb3310b_read (u16 reg); +extern void kb3310b_write (u16 reg,u8 val); +extern bool kb3310b_fw_earlier (char *version); +extern int kb3310b_query_seq (unsigned char cmd); +extern int kb3310b_query_event_num (void ); +extern int kb3310b_get_event_num (void ); + +typedef int (*sci_handler ) (int status); + +extern int yeeloong_sci_register_notify (struct notifier_block *nb); +extern int yeeloong_sci_unregister_notify (struct notifier_block *nb); + +#define KB3310B_VERSION_LEN 8 + +struct kb3310b_chip { +char version[KB3310B_VERSION_LEN ]; +}; + +#define KB3310B_SCI_IRQ_NUM 0x0A + +/* +* The following registers are determined by the EC index configuration. +* 1, fill the PORT_HIGH as EC register high part. +* 2, fill the PORT_LOW as EC register low part. +* 3, fill the PORT_DATA as EC register write data or get the data from it. +*/ +#define KB3310B_IO_PORT_HIGH 0x0381 +#define KB3310B_IO_PORT_LOW 0x0382 +#define KB3310B_IO_PORT_DATA 0x0383 + +/* +* EC delay time is 500us for register and status access +*/ +#define KB3310B_REG_UDELAY 500 +#define KB3310B_CMD_TIMEOUT 0x1000 + +/* +* EC access port for SCI communication +*/ +#define KB3310B_CMD_PORT 0x66 +#define KB3310B_STS_PORT 0x66 +#define KB3310B_DAT_PORT 0x62 +#define KB3310B_CMD_INIT_IDLE_MODE 0xdd +#define KB3310B_CMD_EXIT_IDLE_MODE 0xdf +#define KB3310B_CMD_INIT_RESET_MODE 0xd8 +#define KB3310B_CMD_REBOOT_SYSTEM 0x8c +#define KB3310B_CMD_GET_EVENT_NUM 0x84 +#define KB3310B_CMD_PROGRAM_PIECE 0xda + +/* temperature & fan registers */ +#define KB3310B_REG_TEMPERATURE_VALUE 0xF458 +#define KB3310B_REG_FAN_AUTO_MAN_SWITCH 0xF459 +#define KB3310B_BIT_FAN_AUTO 0 +#define KB3310B_BIT_FAN_MANUAL 1 +#define KB3310B_REG_FAN_CONTROL 0xF4D2 +#define KB3310B_BIT_FAN_CONTROL_ON (1 << 0) +#define KB3310B_BIT_FAN_CONTROL_OFF (0 << 0) +#define KB3310B_REG_FAN_STATUS 0xF4DA +#define KB3310B_BIT_FAN_STATUS_ON (1 << 0) +#define KB3310B_BIT_FAN_STATUS_OFF (0 << 0) +#define KB3310B_REG_FAN_SPEED_HIGH 0xFE22 +#define KB3310B_REG_FAN_SPEED_LOW 0xFE23 +#define KB3310B_REG_FAN_SPEED_LEVEL 0xF4CC + +/* fan speed divider */ +#define KB3310B_FAN_SPEED_DIVIDER 480000/* (60*1000*1000/62.5/2)*/ + +/* battery registers */ +#define KB3310B_REG_BAT_DESIGN_CAP_HIGH 0xF77D +#define KB3310B_REG_BAT_DESIGN_CAP_LOW 0xF77E +#define KB3310B_REG_BAT_FULLCHG_CAP_HIGH 0xF780 +#define KB3310B_REG_BAT_FULLCHG_CAP_LOW 0xF781 +#define KB3310B_REG_BAT_DESIGN_VOL_HIGH 0xF782 +#define KB3310B_REG_BAT_DESIGN_VOL_LOW 0xF783 +#define KB3310B_REG_BAT_CURRENT_HIGH 0xF784 +#define KB3310B_REG_BAT_CURRENT_LOW 0xF785 +#define KB3310B_REG_BAT_VOLTAGE_HIGH 0xF786 +#define KB3310B_REG_BAT_VOLTAGE_LOW 0xF787 +#define KB3310B_REG_BAT_TEMPERATURE_HIGH 0xF788 +#define KB3310B_REG_BAT_TEMPERATURE_LOW 0xF789 +#define KB3310B_REG_BAT_RELATIVE_CAP_HIGH 0xF492 +#define KB3310B_REG_BAT_RELATIVE_CAP_LOW 0xF493 +#define KB3310B_REG_BAT_VENDOR 0xF4C4 +#define KB3310B_FLAG_BAT_VENDOR_SANYO 0x01 +#define KB3310B_FLAG_BAT_VENDOR_SIMPLO 0x02 +#define KB3310B_REG_BAT_CELL_COUNT 0xF4C6 +#define KB3310B_FLAG_BAT_CELL_3S1P 0x03 +#define KB3310B_FLAG_BAT_CELL_3S2P 0x06 +#define KB3310B_REG_BAT_CHARGE 0xF4A2 +#define KB3310B_FLAG_BAT_CHARGE_DISCHARGE 0x01 +#define KB3310B_FLAG_BAT_CHARGE_CHARGE 0x02 +#define KB3310B_FLAG_BAT_CHARGE_ACPOWER 0x00 +#define KB3310B_REG_BAT_STATUS 0xF4B0 +#define KB3310B_BIT_BAT_STATUS_LOW (1 << 5) +#define KB3310B_BIT_BAT_STATUS_DESTROY (1 << 2) +#define KB3310B_BIT_BAT_STATUS_FULL (1 << 1) +#define KB3310B_BIT_BAT_STATUS_IN (1 << 0) +#define KB3310B_REG_BAT_CHARGE_STATUS 0xF4B1 +#define KB3310B_BIT_BAT_CHARGE_STATUS_OVERTEMP (1 << 2) +#define KB3310B_BIT_BAT_CHARGE_STATUS_PRECHG (1 << 1) +#define KB3310B_REG_BAT_STATE 0xF482 +#define KB3310B_BIT_BAT_STATE_CHARGING (1 << 1) +#define KB3310B_BIT_BAT_STATE_DISCHARGING (1 << 0) +#define KB3310B_REG_BAT_POWER 0xF440 +#define KB3310B_BIT_BAT_POWER_S3 (1 << 2) +#define KB3310B_BIT_BAT_POWER_ON (1 << 1) +#define KB3310B_BIT_BAT_POWER_ACIN (1 << 0) + +/* other registers */ + +/* Audio: rd/wr */ +#define KB3310B_REG_AUDIO_VOLUME 0xF46C +#define KB3310B_REG_AUDIO_MUTE 0xF4E7 +#define KB3310B_REG_AUDIO_BEEP 0xF4D0 + +/* USB port power or not: rd/wr */ +#define KB3310B_REG_USB0_FLAG 0xF461 +#define KB3310B_REG_USB1_FLAG 0xF462 +#define KB3310B_REG_USB2_FLAG 0xF463 +#define KB3310B_BIT_USB_FLAG_ON 1 +#define KB3310B_BIT_USB_FLAG_OFF 0 + +/* LID */ +#define KB3310B_REG_LID_DETECT 0xF4BD +#define KB3310B_BIT_LID_DETECT_ON 1 +#define KB3310B_BIT_LID_DETECT_OFF 0 + +/* CRT */ +#define KB3310B_REG_CRT_DETECT 0xF4AD +#define KB3310B_BIT_CRT_DETECT_PLUG 1 +#define KB3310B_BIT_CRT_DETECT_UNPLUG 0 + +/* LCD backlight brightness adjust: 9 levels */ +#define KB3310B_REG_DISPLAY_BRIGHTNESS 0xF4F5 + +/* Black screen status */ +#define KB3310B_REG_DISPLAY_LCD 0xF79F +#define KB3310B_BIT_DISPLAY_LCD_ON 1 +#define KB3310B_BIT_DISPLAY_LCD_OFF 0 + +/* LCD backlight control: off/restore */ +#define KB3310B_REG_BACKLIGHT_CTRL 0xF7BD +#define KB3310B_BIT_BACKLIGHT_ON 1 +#define KB3310B_BIT_BACKLIGHT_OFF 0 + +/* Reset the machine auto-clear: rd/wr */ +#define KB3310B_REG_RESET 0xF4EC +#define KB3310B_BIT_RESET_ON 1 + +/* Light the LED: rd/wr */ +#define KB3310B_REG_LED 0xF4C8 +#define KB3310B_BIT_LED_RED_POWER (1 << 0) +#define KB3310B_BIT_LED_ORANGE_POWER (1 << 1) +#define KB3310B_BIT_LED_GREEN_CHARGE (1 << 2) +#define KB3310B_BIT_LED_RED_CHARGE (1 << 3) +#define KB3310B_BIT_LED_NUMLOCK (1 << 4) + +/* Test LED mode, all LED on/off */ +#define KB3310B_REG_LED_TEST 0xF4C2 +#define KB3310B_BIT_LED_TEST_IN 1 +#define KB3310B_BIT_LED_TEST_OUT 0 + +/* Camera on/off */ +#define KB3310B_REG_CAMERA_STATUS 0xF46A +#define KB3310B_BIT_CAMERA_STATUS_ON 1 +#define KB3310B_BIT_CAMERA_STATUS_OFF 0 +#define KB3310B_REG_CAMERA_CONTROL 0xF7B7 +#define KB3310B_BIT_CAMERA_CONTROL_OFF 0 +#define KB3310B_BIT_CAMERA_CONTROL_ON 1 + +/* WLAN Status */ +#define KB3310B_REG_WLAN 0xF4FA +#define KB3310B_BIT_WLAN_ON 1 +#define KB3310B_BIT_WLAN_OFF 0 + +/* SCI Event Number from EC */ +enum { +KB3310B_EVENT_START =0x22 , +KB3310B_EVENT_LID =0x23 ,/* LID open/close */ +KB3310B_EVENT_DISPLAY_TOGGLE ,/* Fn+F3 for display switch */ +KB3310B_EVENT_SLEEP ,/* Fn+F1 for entering sleep mode */ +KB3310B_EVENT_OVERTEMP ,/* Over-temperature occurred */ +KB3310B_EVENT_CRT_DETECT ,/* CRT is connected */ +KB3310B_EVENT_CAMERA ,/* Camera on/off */ +KB3310B_EVENT_USB_OC2 ,/* USB2 Overcurrent occurred */ +KB3310B_EVENT_USB_OC0 ,/* USB0 Overcurrent occurred */ +KB3310B_EVENT_BLACK_SCREEN ,/* Turn on/off backlight */ +KB3310B_EVENT_AUDIO_MUTE ,/* Mute on/off */ +KB3310B_EVENT_DISPLAY_BRIGHTNESS ,/* LCD backlight brightness adjust */ +KB3310B_EVENT_AC_BAT ,/* AC & Battery relative issue */ +KB3310B_EVENT_AUDIO_VOLUME ,/* Volume adjust */ +KB3310B_EVENT_WLAN ,/* WLAN on/off */ +KB3310B_EVENT_END +}; + +#endif /* !__LINUX_MFD_YEELOONG_KB3310B_H */ Signature · 1 line
2.20.1
Re: [PATCH v2 1/7] mfd: yeeloong_kb3310b: support KB3310B EC for Lemote Yeeloong laptops.
Quoted 4 lines · Paul Burton
Hi Tom,
Overall I think this is much better than the older out of tree platform driver - thanks for cleaning it up.
I'm glad to hear that.
Quoted 3 lines
One general comment - it would be good to run the patches through
./scripts/checkpatch.pl& fix up any warnings it gives unless you have a good reason to disagree with them.
The entire patchset has already been checked by checkpatch.pl, and all generated warnings have been corrected before submission. So I don't think the script can help me improving the style beyond this point. Thanks for pointing out all the additional style issues.
Quoted 2 lines
Nit: the lines of asterisks aren't part of the kernel's general comment style & I think it would looks cleaner to remove them.
Sure, I will remove these asterisks banners.
In fact I learned to use these asterisks from other kernel drivers to separate logical sections... and as a said, checkpatch.pl does not generate a warning about it, so I suspect that, historically, a clear policy about asterisks banners was never enforced. After finish working on these patches, perhaps it would a good time to patch checkpatch.pl and add a check for it?
Quoted 9 lines
I'm not sure I follow why the power management code prevents use of regmap?
Are you talking about the
wakeup_loongson()function? Perhaps it would make sense for the suspend code to be part of one of the possible subdrivers you mention. The lemote-2f seems to be the only system that provides an implementation ofwakeup_loongson()so perhaps a driver could instead just register its own structplatform_suspend_ops& avoid the need for code inarch/mipsto care about the EC.
I'll reword my misleading commit message about using regmap.
Let me clarify it - nothing prevents the use of regmap. I originally thought it would be a good idea since the manual locking can be eliminated.
But later I realized it would mean more changes. In particular, not only pm.c needs to be a new subdriver, the shutdown code in ml2f_reboot() also calls ec_write(), and it needs to be converted to a subdriver in drivers/power/reset/ too. Yes, it would be good to refactor these code and move them to a more suitable place, but currently, the primary goal is to add platform drivers, not to touch functions I've just fixed and possibly introduce new regressions...
As you've seen, there is a TODO item in [PATCH 6/7] in this series, it says the CS5536 code also needs some major refactoring: the GPIO needs to be changed, the clocksource driver needs to be merged with the clockevent driver under drivers/clocksource, etc. I think we can, as well, save the pm.c/reset.c for later. Currently, I think it's better to focus on the platform drivers first.
If you still have objections on it, please let me know.
Quoted 4 lines
Quoted 1 line
+#define
DRV_NAME"yeeloong_kb3310b: "Defining
pr_fmt()would be cleaner - you wouldn't need to manually includeDRV_NAMEin your messages later.
Noted.
Quoted 25 lines
Quoted 11 lines
drivers/mfd/yeeloong_kb3310b.c new file +11
-0+static struct kb3310b_chip *kb3310b_fwinfo; + +static const struct mfd_cell kb3310b_cells[] = { + { + .name ="yeeloong_sci" + }, + { + .name ="yeeloong_hwmon" + }, + { + .name ="yeeloong_battery" drivers/mfd/yeeloong_kb3310b.c line 59
Quoted 10 lines
drivers/mfd/yeeloong_kb3310b.c new file +10
-0+ }, + { + .name = "yeeloong_backlight" + }, + { + .name ="yeeloong_lcd" + }, + { + .name ="yeeloong_hotkey" + },Nit: I think it'd look cleaner if you remove the newlines within each array entry, eg:
Quoted 3 lines
{ .name = "yeeloong_sci" },
{ .name = "yeelong_hwmon" },
...Yes.
Quoted 5 lines
Review
Quoted 2 lines
(quoted diff) +2
-0+ +static DEFINE_SPINLOCK(kb3310b_command_lock);Since this is only used in
kb3310b_query_seq()could you just declare it (still static) inside that function?
No problem.
Thanks for your review, I'll correct all the issues above and send v3 today, if there's no addition problem in v3, I will start sending the actual platform drivers (battery/hwmon/etc) for the next round of review.
P.S: This time, I hope Lee Jones, as the MFD subsystem maintainer, has received my mail and my patch, including this one. Unfortunately, all signs indicated he hasn't received it. Jones, if you have received this mail but currently too busy to review, please reply to confirm, thanks!
Sincerely yours,
Tom Li
PATCH v2 2/7 mips: loongson64: select MFD_YEELOONG_KB3310B for LEMOTE_MACH2F.
In reply to Yifeng LiTo ease the support of platform drivers on Lemote Yeeloong laptop, code for accessing the embedded controller has been separated from arch/mips, as a MFD driver. Since the board files here still need to access the EC directly to handle reboot/shutdown and interrupts, we make MFD_YEELOONG_KB3310B as a mandatory dependency.
--- arch/mips/loongson64/Kconfig | 1 + 1 file changed, 1 insertion(+)
arch/mips/loongson64/Kconfig +1 -0
@@ -56,6 +56,7 @@ config LEMOTE_MACH2F select SYS_SUPPORTS_HIGHMEM select SYS_SUPPORTS_LITTLE_ENDIAN select LOONGSON_MC146818 + select MFD_YEELOONG_KB3310B help Lemote Loongson 2F family machines utilize the 2F revision of Loongson processor and the AMD CS5536 south bridge.
Signature · 1 line
2.20.1
PATCH v2 3/7 mips: loongson64: remove ec_kb3310b.c, use MFD driver.
In reply to Yifeng LiWe have already converted the supporting code for ENE KB3310B embedded controller as a separate MFD driver, and select it as a dependency of LEMOTE_MACH2F.
This commit removes the original implementation of ec_kb3310b.c, and converts all EC operations to use the utility function provided by the yeeloong_kb3310b MFD driver instead.
--- arch/mips/loongson64/lemote-2f/Makefile | 2 +- arch/mips/loongson64/lemote-2f/ec_kb3310b.c | 129 -------------- arch/mips/loongson64/lemote-2f/ec_kb3310b.h | 188 -------------------- arch/mips/loongson64/lemote-2f/pm.c | 18 +- arch/mips/loongson64/lemote-2f/reset.c | 4 +- 5 files changed, 12 insertions(+), 329 deletions(-) delete mode 100644 arch/mips/loongson64/lemote-2f/ec_kb3310b.c delete mode 100644 arch/mips/loongson64/lemote-2f/ec_kb3310b.h
arch/mips/loongson64/lemote-2f/Makefile +1 -1
@@ -2,7 +2,7 @@ # Makefile for lemote loongson2f family machines # -obj-y += clock.o machtype.o irq.o reset.o dma.o ec_kb3310b.o +obj-y += clock.o machtype.o irq.o reset.o dma.o # # Suspend Support
arch/mips/loongson64/lemote-2f/ec_kb3310b.c deleted +0 -129
129 lines not on this page · raw message
arch/mips/loongson64/lemote-2f/ec_kb3310b.h deleted +0 -188
188 lines not on this page · raw message
arch/mips/loongson64/lemote-2f/pm.c +9 -9
@@ -23,7 +23,7 @@#include <loongson.h> #include <cs5536/cs5536_mfgpt.h> -#include "ec_kb3310b.h" +#include <linux/mfd/yeeloong_kb3310b.h> #define I8042_KBD_IRQ 1#define I8042_CTR_KBDINT 0x01
@@ -70,7 +70,7 @@ void setup_wakeup_events(void)/* Wakeup CPU via SCI lid open event */ outb (irq_mask & ~(1 <<PIC_CASCADE_IR ),PIC_MASTER_IMR );inb (PIC_MASTER_IMR ); -outb (0xff & ~(1 << (SCI_IRQ_NUM -8 )),PIC_SLAVE_IMR ); +outb (0xff & ~(1 << (KB3310B_SCI_IRQ_NUM -8 )),PIC_SLAVE_IMR );inb (PIC_SLAVE_IMR );break ;
@@ -88,7 +88,7 @@ EXPORT_SYMBOL(yeeloong_report_lid_status);static void yeeloong_lid_update_task (struct work_struct *work) {if (yeeloong_report_lid_status) -yeeloong_report_lid_status (BIT_LID_DETECT_ON ); +yeeloong_report_lid_status (KB3310B_BIT_LID_DETECT_ON ); }int wakeup_loongson (void )
@@ -104,21 +104,21 @@ int wakeup_loongson(void)if (irq ==I8042_KBD_IRQ )return 1 ; -else if (irq ==SCI_IRQ_NUM ) { +else if (irq ==KB3310B_SCI_IRQ_NUM ) {int ret, sci_event;/* query the event number */ - ret =ec_query_seq (CMD_GET_EVENT_NUM ); + ret =kb3310b_query_seq (KB3310B_CMD_GET_EVENT_NUM );if (ret <0 )return 0 ; - sci_event =ec_get_event_num (); + sci_event =kb3310b_get_event_num ();if (sci_event <0 )return 0 ; -if (sci_event ==EVENT_LID ) { +if (sci_event ==KB3310B_EVENT_LID ) {int lid_status;/* check the LID status */ - lid_status =ec_read (REG_LID_DETECT ); + lid_status =kb3310b_read (KB3310B_REG_LID_DETECT );/* wakeup cpu when people open the LID */ -if (lid_status ==BIT_LID_DETECT_ON ) { +if (lid_status ==KB3310B_BIT_LID_DETECT_ON ) {/* If we call it directly here, the WARNING * will be sent out by getnstimeofday * via "WARN_ON(timekeeping_suspended);"
arch/mips/loongson64/lemote-2f/reset.c +2 -2
@@ -20,7 +20,7 @@#include <loongson.h> #include <cs5536/cs5536.h> -#include "ec_kb3310b.h" +#include <linux/mfd/yeeloong_kb3310b.h> static void reset_cpu (void ) {
@@ -81,7 +81,7 @@ static void ml2f_reboot(void)reset_cpu ();/* sending an reset signal to EC(embedded controller) */ -ec_write (REG_RESET ,BIT_RESET_ON ); +kb3310b_write (KB3310B_REG_RESET ,KB3310B_BIT_RESET_ON ); }#define yl2f89_reboot ml2f_reboot
Signature · 1 line
2.20.1
PATCH v2 4/7 mips: loongson64: remove yeeloong_report_lid_status from pm.c
In reply to Yifeng LiThere is some complicated logic in lemote-2f/pm.c. During wakeup, it creates a delayed_work to execute a callback to the function yeeloong_report_lid_status(). It's only purpose is to report the current status of the laptop lid switch, and this callback function wan not implemented in the mainline kernel.
This level of overenginnering hardly makes sense. All we need is to report the laptop lid switch unconditionally upon wakeup in the future PM code, which is being worked on.
--- arch/mips/loongson64/lemote-2f/pm.c | 22 ---------------------- 1 file changed, 22 deletions(-)
arch/mips/loongson64/lemote-2f/pm.c +0 -22
@@ -80,17 +80,6 @@ void setup_wakeup_events(void) } } -static struct delayed_work lid_task; -static int initialized; -/* yeeloong_report_lid_status will be implemented in yeeloong_laptop.c */ -sci_handler yeeloong_report_lid_status; -EXPORT_SYMBOL (yeeloong_report_lid_status); -static void yeeloong_lid_update_task (struct work_struct *work) -{ -if (yeeloong_report_lid_status) -yeeloong_report_lid_status (KB3310B_BIT_LID_DETECT_ON ); -} -int wakeup_loongson (void ) {int irq;
@@ -119,17 +108,6 @@ int wakeup_loongson(void) lid_status =kb3310b_read (KB3310B_REG_LID_DETECT );/* wakeup cpu when people open the LID */ if (lid_status ==KB3310B_BIT_LID_DETECT_ON ) { -/* If we call it directly here, the WARNING -* will be sent out by getnstimeofday -* via "WARN_ON(timekeeping_suspended);" -* because we can not schedule in suspend mode. -*/ -if (initialized ==0 ) { -INIT_DELAYED_WORK (&lid_task, - yeeloong_lid_update_task); - initialized =1 ; - } -schedule_delayed_work (&lid_task,1 );return 1 ; } }
Signature · 1 line
2.20.1
PATCH v2 5/7 mips: loongson64: register per-board platform drivers for lemote-2f
In reply to Yifeng LiCurrently, common/platform.c registers the Loongson 2F cpufreq driver during boot time for all boards. To support platform drivers for Lemote Yeeloong laptops, we need to register more drivers.
First, we add support for per-board platform drivers. Just like how IRQ, DMA, or reset logic is implemented for each board, we introduces a call of mach_platform_init() in common/platform.c, to allow each board to have its own platform.c to register platform drivers.
Then, we implement lemote-2f/platform.c to register the MFD driver for Yeeloong laptops.
So far, only one board, lemote-f2, is using this facility, so we hardcode the call of mach_platform_init() in common/platform.c as a ifdef for now.
--- .../include/asm/mach-loongson64/loongson.h | 3 ++ arch/mips/loongson64/common/platform.c | 15 ++++++ arch/mips/loongson64/lemote-2f/Makefile | 2 +- arch/mips/loongson64/lemote-2f/platform.c | 47 +++++++++++++++++++ 4 files changed, 66 insertions(+), 1 deletion(-) create mode 100644 arch/mips/loongson64/lemote-2f/platform.c
arch/mips/include/asm/mach-loongson64/loongson.h +3 -0
@@ -23,6 +23,9 @@ extern void bonito_irq_init(void);extern void mach_prepare_reboot (void );extern void mach_prepare_shutdown (void ); +/* machine-specific platform driver registration */ +extern int mach_platform_init (void )__init ; +/* environment arguments from bootloader */ extern u32 cpu_clock_freq;extern u32 memsize, highmemsize;
arch/mips/loongson64/common/platform.c +15 -0
@@ -11,6 +11,7 @@#include <linux/err.h> #include <linux/smp.h> #include <linux/platform_device.h> +#include <loongson.h> static struct platform_device loongson2_cpufreq_device = { .name ="loongson2_cpufreq" ,
@@ -29,3 +30,17 @@ static int __init loongson2_cpufreq_init(void) }arch_initcall (loongson2_cpufreq_init); + +/* +* Currently, only LEMOTE_MACH2F implements mach_platform_init(); +* Fuloong-2E or Loongson3 does not have platform drivers to register +* at here yet. +*/ +#ifdef CONFIG_LEMOTE_MACH2F +static int __init loongson2_platform_init (void ) +{ +return mach_platform_init (); +} + +device_initcall (loongson2_platform_init); +#endif
arch/mips/loongson64/lemote-2f/Makefile +1 -1
@@ -2,7 +2,7 @@ # Makefile for lemote loongson2f family machines # -obj-y += clock.o machtype.o irq.o reset.o dma.o +obj-y += clock.o machtype.o irq.o reset.o dma.o platform.o # # Suspend Support
arch/mips/loongson64/lemote-2f/platform.c new file +47 -0
47 lines not on this page · raw message
Signature · 1 line
2.20.1
PATCH v2 6/7 mips: loongson64: Support System Control Interrupts for Lemote Yeeloong.
In reply to Yifeng LiThe hardware design of Yeeloong laptops is similar to OLPC: low-level hardware events are processed by the ENE KB3310B Embedded Controller, which is connected to the AMD CS5536 southbridge through a GPIO port.
When a hardware event occurs, such as a short-circuit on the USB port, removing the power supply, plugging in a VGA adapter, opening the lid of the laptop, or pressing a hotkey on the keyboard, the EC sends a pulse to CS5536, which then fires a System Control Interrupt to notify the kernel.
In the previous attempted submission, the logic for handling SCI and hotkeys was tightly-coupled, and the driver was called the "hotkey" driver, which was misleading. In this implementation of sci.c, we only handle the minimum number of things that deal with the underlying platform hardware directly, such as setting up the CS5536 GPIO and IRQ, handling the the power supply of USB and camera. The vast majority of events are passed to the subdrivers via a notification chain, thus, the SCI logic and the specific hotkey handling logic have been decoupled.
--- arch/mips/loongson64/lemote-2f/Makefile | 2 +- arch/mips/loongson64/lemote-2f/sci.c | 394 ++++++++++++++++++++++++ 2 files changed, 395 insertions(+), 1 deletion(-) create mode 100644 arch/mips/loongson64/lemote-2f/sci.c
arch/mips/loongson64/lemote-2f/Makefile +1 -1
@@ -2,7 +2,7 @@ # Makefile for lemote loongson2f family machines # -obj-y += clock.o machtype.o irq.o reset.o dma.o platform.o +obj-y += clock.o machtype.o irq.o reset.o dma.o platform.o sci.o # # Suspend Support
arch/mips/loongson64/lemote-2f/sci.c new file +394 -0
394 lines not on this page · raw message
Signature · 1 line
2.20.1
PATCH v2 7/7 MAINTAINERS: add myself as a maintainer of MIPS/Loongson2 platform code.
In reply to Yifeng LiI've introduced platform code for Lemote Yeeloong computers and modified power management-related files. Add myself as a maintainer of these code.
--- MAINTAINERS | 9 +++++++++ 1 file changed, 9 insertions(+)
MAINTAINERS +9 -0
@@ -10134,6 +10134,15 @@ F: arch/mips/include/asm/mach-loongson64/ F: drivers/*/*loongson2* F: drivers/*/*/*loongson2* +MIPS/LOONGSON2 LEMOTE PLATFORM AND POWER MANAGEMENT DRIVER +M: Tom Li <[email protected]> +L: [email protected] +S: Maintained +F: arch/mips/loongson64/common/platform.c +F: arch/mips/loongson64/lemote-2f/pm.c +F: arch/mips/loongson64/lemote-2f/sci.c +F: arch/mips/loongson64/lemote-2f/platform.c + MIPS/LOONGSON3 ARCHITECTURE M: Huacai Chen <[email protected]> L: [email protected]
Signature · 1 line
2.20.1