[2/3] hwmon: xgene: Adds hwmon driver

Guenter Roeck linux at roeck-us.net
Sun May 29 22:25:07 PDT 2016


On Mon, May 16, 2016 at 09:17:26AM -0700, hotran wrote:
> This patch adds hardware temperature and power reading support for
> APM X-Gene SoC's using the mailbox communication interface.
> 
Please drop the "'".

> 
> Signed-off-by: Hoan Tran <hotran at apm.com>
> ---
>  Documentation/hwmon/xgene-hwmon |  32 ++
>  drivers/hwmon/Kconfig           |   7 +
>  drivers/hwmon/Makefile          |   1 +
>  drivers/hwmon/xgene-hwmon.c     | 801 ++++++++++++++++++++++++++++++++++++++++
>  4 files changed, 841 insertions(+)
>  create mode 100644 Documentation/hwmon/xgene-hwmon
>  create mode 100644 drivers/hwmon/xgene-hwmon.c
> 
> diff --git a/Documentation/hwmon/xgene-hwmon b/Documentation/hwmon/xgene-hwmon
> new file mode 100644
> index 0000000..040a7f2
> --- /dev/null
> +++ b/Documentation/hwmon/xgene-hwmon
> @@ -0,0 +1,32 @@
> +Kernel driver xgene-hwmon
> +========================
> +
> +Supported chips:
> + * APM X-Gene SoC
> +
> +Description
> +-----------
> +
> +This driver adds hardware temperature and power reading support for
> +APM X-Gene SoC's using the mailbox communication interface.
> 
No "'".
> 
> +For device tree, it is the standard DT mailbox.
> +For ACPI, it is the PCC mailbox.
> +
> +The following sensors are supported
> +
> +  * Temperature
> +    - SoC on-die temperature in milli-degree
> +    - Alarm when high/over temperature occurs
> +  * Power
> +    - CPU power in uW
> +    - IO power in uW
> +    - CPU and IO power in uW
> +
> +sysfs-Interface
> +---------------
> +
> +temp1_input - SoC on-die temperature (milli-degree C)
> +temp1_critical_alarm - An 1 would indicates on-die temperature exceeded threshold
> +power1_input - CPU power in (uW)
> +power2_input - IO power in (uW)
> +power3_input - CPU and IO power in (uW)
> diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
> index 5c2d13a..91c3056 100644
> --- a/drivers/hwmon/Kconfig
> +++ b/drivers/hwmon/Kconfig
> @@ -1776,6 +1776,13 @@ config SENSORS_ULTRA45
>  	  This driver provides support for the Ultra45 workstation environmental
>  	  sensors.
>  
> +config SENSORS_XGENE
> +	tristate "APM X-Gene SoC hardware monitoring driver"
> +	depends on XGENE_SLIMPRO_MBOX || PCC
> +	help
> +	  If you say yes here you get support for the temperature
> +	  and power sensors for APM X-Gene SoC.
> +
>  if ACPI
>  
>  comment "ACPI drivers"
> diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
> index 58cc3ac..668d0f1 100644
> --- a/drivers/hwmon/Makefile
> +++ b/drivers/hwmon/Makefile
> @@ -161,6 +161,7 @@ obj-$(CONFIG_SENSORS_W83L785TS)	+= w83l785ts.o
>  obj-$(CONFIG_SENSORS_W83L786NG)	+= w83l786ng.o
>  obj-$(CONFIG_SENSORS_WM831X)	+= wm831x-hwmon.o
>  obj-$(CONFIG_SENSORS_WM8350)	+= wm8350-hwmon.o
> +obj-$(CONFIG_SENSORS_XGENE)	+= xgene-hwmon.o
>  
>  obj-$(CONFIG_PMBUS)		+= pmbus/
>  
> diff --git a/drivers/hwmon/xgene-hwmon.c b/drivers/hwmon/xgene-hwmon.c
> new file mode 100644
> index 0000000..bc4ad29
> --- /dev/null
> +++ b/drivers/hwmon/xgene-hwmon.c
> @@ -0,0 +1,801 @@
> +/*
> + * APM X-Gene SoC Hardware Monitoring Driver
> + *
> + * Copyright (c) 2016, Applied Micro Circuits Corporation
> + * Author: Loc Ho <lho at apm.com>
> + *         Hoan Tran <hotran at apm.com>
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License as
> + * published by the Free Software Foundation; either version 2 of
> + * the License, or (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, see <http://www.gnu.org/licenses/>.
> + *
> + * This driver provides the following features:
> + *  - Retrieve CPU's total power (uW)
> + *  - Retrieve IO's total power (uW)
 
Please drop the "'s".

> + *  - Retrieve SoC total power (uW)
> + *  - Retrieve SoC temperature (milli-degree C) and alarm
> + */
> +#include <linux/acpi.h>
> +#include <linux/dma-mapping.h>
> +#include <linux/hwmon.h>
> +#include <linux/hwmon-sysfs.h>
> +#include <linux/kfifo.h>
> +#include <linux/interrupt.h>
> +#include <linux/mailbox_client.h>
> +#include <linux/module.h>
> +#include <linux/of.h>
> +#include <linux/platform_device.h>
> +#include <acpi/cppc_acpi.h>
> +
> +/* SLIMpro message defines */
> +#define SLIMPRO_MSG_TYPE_DBG_ID		0
> +#define SLIMPRO_MSG_TYPE_ERR_ID		7
> +#define SLIMPRO_MSG_TYPE_PWRMGMT_ID	9
> +
> +#define SLIMPRO_MSG_TYPE(v)		(((v) & 0xF0000000) >> 28)
> +#define SLIMPRO_MSG_TYPE_SET(v)		(((v) << 28) & 0xF0000000)
> +#define SLIMPRO_MSG_SUBTYPE(v)		(((v) & 0x0F000000) >> 24)
> +#define SLIMPRO_MSG_SUBTYPE_SET(v)	(((v) << 24) & 0x0F000000)
> +
> +#define SLIMPRO_DBG_SUBTYPE_SENSOR_READ	4
> +#define SLIMPRO_SENSOR_READ_MSG		0x04FFE902
> +#define SLIMPRO_SENSOR_READ_ENCODE_ADDR(a) \
> +	((a) & 0x000FFFFF)
> +#define PMD_PWR_MW_REG			0x26
> +#define SOC_PWR_REG			0x21
> +#define SOC_TEMP_REG			0x10
> +
> +#define SLIMPRO_PWRMGMT_SUBTYPE_TPC	1
> +#define SLIMPRO_TPC_ALARM		2
> +#define SLIMPRO_TPC_GET_ALARM		3
> +#define SLIMPRO_TPC_CMD(v)		(((v) & 0x00FF0000) >> 16)
> +#define SLIMPRO_TPC_CMD_SET(v)		(((v) << 16) & 0x00FF0000)
> +#define SLIMPRO_TPC_ENCODE_MSG(hndl, cmd, type) \
> +	(SLIMPRO_MSG_TYPE_SET(SLIMPRO_MSG_TYPE_PWRMGMT_ID) | \
> +	SLIMPRO_MSG_SUBTYPE_SET(hndl) | \
> +	SLIMPRO_TPC_CMD_SET(cmd) | \
> +	type)
> +

Do you _have_ to use such long defines ? This causes lots of "line too long"
checkpatch warnings, which would really be easy to avoid with shorter defines.

> +/* PCC defines */
> +#define SLIMPRO_MSG_PCC_SUBSPACE	7
> +#define PCC_SIGNATURE_MASK		0x50424300
> +#define PCCC_GENERATE_DB_INT		BIT(15)
> +#define PCCS_CMD_COMPLETE		BIT(0)
> +#define PCCS_SCI_DOORBEL		BIT(1)
> +#define PCCS_PLATFORM_NOTIFICATION	BIT(3)
> +/*
> + * Arbitrary retries in case the remote processor is slow to respond
> + * to PCC commands
> + */
> +#define PCC_NUM_RETRIES			500
> +
> +#define ASYNC_MSG_FIFO_SIZE		16
> +#define MBOX_HWMON_INDEX		0
> +#define MBOX_OP_TIMEOUTMS		1000
> +
> +#define SOC_TEMP			0
> +#define CPU_POWER			0
> +#define IO_POWER			1
> +#define SOC_POWER			2
> +

Are those random defines, to be used for index values ?
If so, how about enums ?

> +#define WATT_TO_mWATT(x)		((x) * 1000)
> +#define mWATT_TO_uWATT(x)		((x) * 1000)
> +#define WATT_TO_uWATT(x)		((x) * 1000000)
> +#define CELSIUS_TO_mCELSIUS(x)		((x) * 1000)
> +
> +#define to_xgene_hwmon_dev(cl)		\
> +	container_of(cl, struct xgene_hwmon_dev, mbox_client)
> +
> +struct slimpro_resp_msg {
> +	u32 msg;
> +	u32 param1;
> +	u32 param2;
> +} __packed;
> +
> +struct xgene_hwmon_dev {
> +	struct device		*dev;
> +	struct mbox_chan	*mbox_chan;
> +	struct mbox_client	mbox_client;
> +
> +	spinlock_t		lock;
> +	struct completion	rd_complete;
> +	int			resp_pending;
> +	struct slimpro_resp_msg sync_msg;
> +
> +	struct work_struct	workq;
> +	struct kfifo_rec_ptr_1	async_msg_fifo;
> +
> +	struct device		*hwmon_dev;
> +	bool			temp_critical_alarm;
> +
> +	phys_addr_t		comm_base_addr;
> +	void			*pcc_comm_addr;
> +	u64			usecs_lat;
> +};
> +
> +/*
> + * This function tests and clears a bitmask then returns its old value
> + */
> +static u16 xgene_word_tst_and_clr(u16 *addr, u16 mask)
> +{
> +	u16 ret, val;
> +
> +	val = readw_relaxed(addr);
> +	ret = val & mask;
> +	val &= ~mask;
> +	writew_relaxed(val, addr);
> +
> +	return ret;
> +}
> +
> +static int xgene_hwmon_pcc_rd(struct xgene_hwmon_dev *ctx, u32 *msg)
> +{
> +	struct acpi_pcct_shared_memory *generic_comm_base = ctx->pcc_comm_addr;
> +	void *ptr = generic_comm_base + 1;
> +	unsigned long flags;
> +	u16 val;
> +	int rc;
> +
> +	spin_lock_irqsave(&ctx->lock, flags);
> +	if (ctx->resp_pending) {
> +		spin_unlock_irqrestore(&ctx->lock, flags);
> +		return -EAGAIN;
> +	}
> +
> +	init_completion(&ctx->rd_complete);
> +	ctx->resp_pending = true;
> +
> +	/* Write signature for subspace */
> +	writel_relaxed(PCC_SIGNATURE_MASK | SLIMPRO_MSG_PCC_SUBSPACE,
> +		       &generic_comm_base->signature);
> +
> +	/* Write to the shared command region */
> +	writew_relaxed(SLIMPRO_MSG_TYPE(msg[0]) | PCCC_GENERATE_DB_INT,
> +		       &generic_comm_base->command);
> +
> +	/* Flip CMD COMPLETE bit */
> +	val = readw_relaxed(&generic_comm_base->status);
> +	val &= ~PCCS_CMD_COMPLETE;
> +	writew_relaxed(val, &generic_comm_base->status);
> +
> +	/* Copy the message to the PCC comm space */
> +	memcpy(ptr, msg, sizeof(struct slimpro_resp_msg));
> +
> +	/* Ring the doorbell */
> +	rc = mbox_send_message(ctx->mbox_chan, msg);
> +	if (rc < 0) {
> +		dev_err(ctx->dev, "Mailbox send error %d\n", rc);
> +		goto err;
> +	}
> +	spin_unlock_irqrestore(&ctx->lock, flags);
> +	if (!wait_for_completion_timeout(&ctx->rd_complete,
> +					 usecs_to_jiffies(ctx->usecs_lat))) {
> +		spin_lock_irqsave(&ctx->lock, flags);
> +		dev_err(ctx->dev, "Mailbox operation timed out\n");
> +		rc = -EIO;

	-ETIMEDOUT ?

> +		goto err;
> +	}
> +	spin_lock_irqsave(&ctx->lock, flags);
> +
> +	/* Check for invalid data or no device */
> +	if (SLIMPRO_MSG_TYPE(ctx->sync_msg.msg) == SLIMPRO_MSG_TYPE_ERR_ID ||
> +	    ctx->sync_msg.msg == 0xffffffff) {
> +		rc = -ENODEV;

Two different errors ?

> +		goto err;
> +	}
> +
> +	msg[0] = ctx->sync_msg.msg;
> +	msg[1] = ctx->sync_msg.param1;
> +	msg[2] = ctx->sync_msg.param2;
> +
> +err:
> +	mbox_chan_txdone(ctx->mbox_chan, 0);
> +	ctx->resp_pending = false;
> +	spin_unlock_irqrestore(&ctx->lock, flags);
> +	return rc;
> +}
> +
> +static int xgene_hwmon_rd(struct xgene_hwmon_dev *ctx, u32 *msg)
> +{
> +	unsigned long flags;
> +	int rc;
> +
> +	spin_lock_irqsave(&ctx->lock, flags);
> +	if (ctx->resp_pending) {
> +		spin_unlock_irqrestore(&ctx->lock, flags);
> +		return -EAGAIN;
> +	}
> +
> +	init_completion(&ctx->rd_complete);
> +	ctx->resp_pending = true;
> +	rc = mbox_send_message(ctx->mbox_chan, msg);
> +	if (rc < 0) {
> +		dev_err(ctx->dev, "Mailbox send error %d\n", rc);
> +		goto err;
> +	}
> +	spin_unlock_irqrestore(&ctx->lock, flags);
> +
> +	if (!wait_for_completion_timeout(&ctx->rd_complete,
> +					 msecs_to_jiffies(MBOX_OP_TIMEOUTMS))) {
> +		spin_lock_irqsave(&ctx->lock, flags);
> +		dev_err(ctx->dev, "Mailbox operation timed out\n");
> +		rc = -EIO;

	-ETIMEDOUT ?

> +		goto err;
> +	}
> +	spin_lock_irqsave(&ctx->lock, flags);
> +
> +	/* Check for invalid data or no device */
> +	if (SLIMPRO_MSG_TYPE(ctx->sync_msg.msg) == SLIMPRO_MSG_TYPE_ERR_ID ||
> +	    ctx->sync_msg.msg == 0xffffffff) {
> +		rc = -ENODEV;

Isn't that two different errors ?

> +		goto err;
> +	}
> +
> +	msg[0] = ctx->sync_msg.msg;
> +	msg[1] = ctx->sync_msg.param1;
> +	msg[2] = ctx->sync_msg.param2;
> +
> +err:
> +	ctx->resp_pending = false;
> +	spin_unlock_irqrestore(&ctx->lock, flags);
> +	return rc;
> +}
> +
> +static int xgene_hwmon_reg_map_rd(struct xgene_hwmon_dev *ctx, u32 addr,
> +				  u32 *data)
> +{
> +	u32 msg[3];
> +	int rc;
> +
> +	msg[0] = SLIMPRO_SENSOR_READ_MSG;
> +	msg[1] = SLIMPRO_SENSOR_READ_ENCODE_ADDR(addr);
> +	msg[2] = 0;
> +
Does the msg endianness always match the CPU endianness ?
Probably yes, just want to make sure.

> +	if (ACPI_COMPANION(ctx->dev))
> +		rc = xgene_hwmon_pcc_rd(ctx, msg);
> +	else
> +		rc = xgene_hwmon_rd(ctx, msg);
> +	if (rc < 0) {
> +		/* To support compatibility with firmware, return 0 */
> +		dev_err(ctx->dev, "SLIMpro register 0x%02X read error %d\n",
> +			addr, rc);

Is this error mesage necessary ?

> +		*data = 0;
> +		return rc;
> +	}
> +	*data = msg[1];
> +
> +	return rc;
> +}
> +
> +static int xgene_hwmon_get_notification_msg(struct xgene_hwmon_dev *ctx,
> +					    u32 *amsg)
> +{
> +	u32 msg[3];
> +	int rc;
> +
> +	msg[0] = SLIMPRO_TPC_ENCODE_MSG(SLIMPRO_PWRMGMT_SUBTYPE_TPC,
> +					SLIMPRO_TPC_GET_ALARM, 0);
> +	msg[1] = 0;
> +	msg[2] = 0;
> +
> +	rc = xgene_hwmon_pcc_rd(ctx, msg);
> +	if (rc < 0) {
> +		dev_err(ctx->dev, "PCC Alarm read error %d\n", rc);

Is this error message necessary ?

> +		return rc;
> +	}
> +	amsg[0] = msg[0];
> +	amsg[1] = msg[1];
> +	amsg[2] = msg[2];
> +
> +	return rc;
> +}
> +
> +static int xgene_hwmon_get_cpu_pwr(struct xgene_hwmon_dev *ctx, u32 *val)
> +{
> +	return xgene_hwmon_reg_map_rd(ctx, PMD_PWR_MW_REG, val);
> +}
> +
> +static int xgene_hwmon_get_io_pwr(struct xgene_hwmon_dev *ctx, u32 *val)
> +{
> +	return xgene_hwmon_reg_map_rd(ctx, SOC_PWR_REG, val);
> +}
> +
> +static int xgene_hwmon_get_soc_power(struct xgene_hwmon_dev *ctx, u32 *val)
> +{
> +	u32 pmd_vrm_power;
> +	u32 io_vrm_power;
> +	int rc;
> +
> +	rc = xgene_hwmon_get_cpu_pwr(ctx, &pmd_vrm_power);
> +	if (rc < 0)
> +		return rc;
> +
> +	rc = xgene_hwmon_get_io_pwr(ctx, &io_vrm_power);
> +	if (rc < 0)
> +		return rc;
> +
> +	*val = pmd_vrm_power + WATT_TO_mWATT(io_vrm_power);
> +

So the SoC power is just adding CPU and IO power, ie it reports the sum
of both ?  Please drop. Such calculations should be done in user space
if wanted.

> +	return 0;
> +}
> +
> +static int xgene_hwmon_get_temp(struct xgene_hwmon_dev *ctx, u32 *val)
> +{
> +	return xgene_hwmon_reg_map_rd(ctx, SOC_TEMP_REG, val);
> +}
> +
> +/*
> + * Sensor temperature/power functions
> + */
> +static const char * const sensor_temp_input_names[] = {
> +	[SOC_TEMP] = "SoC Temperature",
> +};
> +
> +static const char * const sensor_pwr_input_names[] = {
> +	[CPU_POWER] = "CPU's power",
> +	[IO_POWER] = "IO's power",

Please drop the "'s".

> +	[SOC_POWER] = "SoC power"
> +};
> +
> +static ssize_t xgene_hwmon_show_name(struct device *dev,
> +				     struct device_attribute *attr, char *buf)
> +{
> +	return snprintf(buf, PAGE_SIZE, "APM X-Gene\n");
> +}
> +
> +static ssize_t xgene_hwmon_show_temp(struct device *dev,
> +				     struct device_attribute *attr,
> +				     char *buf)
> +{
> +	struct xgene_hwmon_dev *ctx = dev_get_drvdata(dev);
> +	u32 val;
> +	int rc;
> +
> +	rc = xgene_hwmon_get_temp(ctx, &val);
> +	if (rc < 0)
> +		return rc;
> +
> +	val = CELSIUS_TO_mCELSIUS(val);
> +
> +	return snprintf(buf, PAGE_SIZE, "%d\n", val);

val is u32 (is it reported by the chip as unsigned ?)

> +}
> +
> +static ssize_t xgene_hwmon_show_temp_label(struct device *dev,
> +					   struct device_attribute *attr,
> +					   char *buf)
> +{
> +	int channel = to_sensor_dev_attr(attr)->index - 1;
> +
> +	return snprintf(buf, PAGE_SIZE, "%s\n",
> +			sensor_temp_input_names[channel]);
> +}
> +
> +static ssize_t xgene_hwmon_show_temp_critical_alarm(struct device *dev,
> +						    struct device_attribute *devattr,
> +						    char *buf)
> +{
> +	struct xgene_hwmon_dev *ctx = dev_get_drvdata(dev);
> +
> +	return snprintf(buf, PAGE_SIZE, "%d\n", ctx->temp_critical_alarm);
> +}
> +
> +static ssize_t xgene_hwmon_show_pwr_label(struct device *dev,
> +					  struct device_attribute *attr,
> +					  char *buf)
> +{
> +	int channel = to_sensor_dev_attr(attr)->index - 1;
> +
> +	return snprintf(buf, PAGE_SIZE, "%s\n",
> +			sensor_pwr_input_names[channel]);
> +}
> +
> +static ssize_t xgene_hwmon_show_pwr(struct device *dev,
> +				    struct device_attribute *attr,
> +				    char *buf)
> +{
> +	struct xgene_hwmon_dev *ctx = dev_get_drvdata(dev);
> +	int channel = to_sensor_dev_attr(attr)->index - 1;
> +	u32 val;
> +	int rc;
> +
> +	switch (channel) {
> +	case CPU_POWER:
> +		rc = xgene_hwmon_get_cpu_pwr(ctx, &val);
> +		break;
> +	case IO_POWER:
> +		rc = xgene_hwmon_get_io_pwr(ctx, &val);
> +		break;
> +	case SOC_POWER:
> +		rc = xgene_hwmon_get_soc_power(ctx, &val);
> +		break;

Those are not the index values specified. Please use those constants
in the attribute declaration.

Overall, it would probably be easier to just have two separate
functions for CPU an IO power.

> +	default:
> +		rc = -EINVAL;
> +		break;
> +	}
> +	if (rc < 0)
> +		return rc;
> +
> +	if (channel == IO_POWER)
> +		val = WATT_TO_uWATT(val);

What is the maximum possible reported IO power ? I understand that
the maximum supported value with the current code is 4,294 kW, but
if the hardware can report more than that it would be prudent to
use an unsigned long for the value.

> +	else
> +		val = mWATT_TO_uWATT(val);
> +
> +	return snprintf(buf, PAGE_SIZE, "%d\n", val);

val is u32.

> +}
> +
> +/* Chip name, required by hwmon */
> +static DEVICE_ATTR(name, S_IRUGO, xgene_hwmon_show_name, NULL);
> +

Unnecesary - see below.

> +#define SENSOR_ATTR_TEMP(index) \
> +	static SENSOR_DEVICE_ATTR(temp##index##_label, S_IRUGO, \
> +				  xgene_hwmon_show_temp_label, NULL, index); \
> +	static SENSOR_DEVICE_ATTR(temp##index##_input, S_IRUGO, \
> +				  xgene_hwmon_show_temp, NULL, index); \
> +	static SENSOR_DEVICE_ATTR(temp##index##_critical_alarm, S_IRUGO, \
> +				  xgene_hwmon_show_temp_critical_alarm, NULL, \
> +				  index);
> +#define SENSOR_ATTR_PWR(index) \
> +	static SENSOR_DEVICE_ATTR(power##index##_label, S_IRUGO, \
> +				  xgene_hwmon_show_pwr_label, NULL, index); \
> +	static SENSOR_DEVICE_ATTR(power##index##_input, S_IRUGO, \
> +				  xgene_hwmon_show_pwr, NULL, index);
> +
> +#define APM_TEMP_SENSOR_ATTR(index) \
> +	&sensor_dev_attr_temp##index##_input.dev_attr.attr, \
> +	&sensor_dev_attr_temp##index##_label.dev_attr.attr, \
> +	&sensor_dev_attr_temp##index##_critical_alarm.dev_attr.attr
> +#define APM_PWR_SENSOR_ATTR(index) \
> +	&sensor_dev_attr_power##index##_input.dev_attr.attr, \
> +	&sensor_dev_attr_power##index##_label.dev_attr.attr
> +
> +SENSOR_ATTR_TEMP(1); /* SoC temperature */
> +SENSOR_ATTR_PWR(1);  /* CPU power */
> +SENSOR_ATTR_PWR(2);  /* IO power */
> +SENSOR_ATTR_PWR(3);  /* SoC power */
> +
> +static struct attribute *xgene_hwmon_attributes[] = {
> +	&dev_attr_name.attr,
> +	APM_TEMP_SENSOR_ATTR(1),
> +	APM_PWR_SENSOR_ATTR(1),
> +	APM_PWR_SENSOR_ATTR(2),
> +	APM_PWR_SENSOR_ATTR(3),

Please drop the macros and write explicit code.

'index' values are always accessed with '- 1'. Please start the index with 0
and drop the '- 1'.

> +	NULL,
> +};
> +
> +static const struct attribute_group xgene_hwmon_attr_group = {
> +	.attrs	= xgene_hwmon_attributes,
> +};
> +
> +static int xgene_hwmon_tpc_alarm(struct xgene_hwmon_dev *ctx,
> +				 struct slimpro_resp_msg *amsg)
> +{
> +	char name[40];
> +
> +	ctx->temp_critical_alarm = amsg->param2 ? true : false;
> +	snprintf(name, sizeof(name), "temp%d_critical_alarm", SOC_TEMP + 1);
> +	sysfs_notify(&ctx->dev->kobj, NULL, name);
> +	dev_alert(ctx->dev, "SoC temperature alarm at %d degree\n",
> +		  amsg->param1);

Please drop this message. That is what the notification is for.
User space can then decide if it wants to log or not.

> +	return 0;
> +}
> +
> +static void xgene_hwmon_process_pwrmsg(struct xgene_hwmon_dev *ctx,
> +				       struct slimpro_resp_msg *amsg)
> +{
> +	switch (SLIMPRO_MSG_SUBTYPE(amsg->msg)) {
> +	case SLIMPRO_PWRMGMT_SUBTYPE_TPC:
> +		switch (SLIMPRO_TPC_CMD(amsg->msg)) {
> +		case SLIMPRO_TPC_ALARM:
> +			xgene_hwmon_tpc_alarm(ctx, amsg);
> +			break;
> +		default:
> +			dev_warn(ctx->dev,
> +				 "Un-supported TPC message received 0x%08X\n",
> +				 amsg->msg);
> +			break;
> +		}
> +		break;
> +	default:
> +		dev_warn(ctx->dev, "Un-supported message received 0x%08X\n",
> +			 amsg->msg);

Are those messages necessary ? Can they clog the log ?

> +		break;
> +	}
> +}
> +
> +/*
> + * This function is called to process async work queue
> + */
> +static void xgene_hwmon_evt_work(struct work_struct *work)
> +{
> +	struct slimpro_resp_msg amsg;
> +	struct xgene_hwmon_dev *ctx;
> +	int ret;
> +
> +	ctx = container_of(work, struct xgene_hwmon_dev, workq);
> +	while (kfifo_out_spinlocked(&ctx->async_msg_fifo, &amsg,
> +				    sizeof(struct slimpro_resp_msg),
> +				    &ctx->lock)) {
> +		/*
> +		 * If PCC, send a consumer command to Platform to get info
> +		 * If Slimpro Mailbox, get message from specific FIFO
> +		 */
> +		if (ACPI_COMPANION(ctx->dev)) {
> +			ret = xgene_hwmon_get_notification_msg(ctx,
> +							       (u32 *)&amsg);
> +			if (ret < 0)
> +				continue;
> +		}
> +		switch (SLIMPRO_MSG_TYPE(amsg.msg)) {
> +		case SLIMPRO_MSG_TYPE_PWRMGMT_ID:
> +			xgene_hwmon_process_pwrmsg(ctx, &amsg);
> +			break;
> +		default:
> +			dev_warn(ctx->dev,
> +				 "Invalid mailbox msg received 0x%08X 0x%08X 0x%08X\n",
> +				 amsg.msg, amsg.param1, amsg.param2);

Is this message necessary ? I am concerned about the logging noise it might
create.

> +			break;
> +		}
> +	}
> +}
> +
> +/*
> + * This function is called when the SLIMpro/PCC Mailbox received a message
> + */
> +static void xgene_hwmon_rx_cb(struct mbox_client *cl, void *msg)
> +{
> +	struct xgene_hwmon_dev *ctx = to_xgene_hwmon_dev(cl);
> +	struct acpi_pcct_shared_memory *generic_comm_base = ctx->pcc_comm_addr;
> +	struct slimpro_resp_msg amsg;
> +
> +	/* If PCC mailbox controller, get msg from shared memory */
> +	if (ACPI_COMPANION(ctx->dev)) {
> +		msg = generic_comm_base + 1;
> +		/* Check if platform sends interrupt */
> +		if (!xgene_word_tst_and_clr(&generic_comm_base->status,
> +					    PCCS_SCI_DOORBEL))
> +			return;
> +	}
> +
> +	/*
> +	 * Response message format:
> +	 * msg[0] is the return code of the operation
> +	 * msg[1] is the first parameter word
> +	 * msg[2] is the second parameter word
> +	 *
> +	 * As message only supports dword size, just assign it.
> +	 */
> +
> +	/* Check for sync query */
> +	if (ctx->resp_pending &&
> +	    ((SLIMPRO_MSG_TYPE(((u32 *) msg)[0]) == SLIMPRO_MSG_TYPE_ERR_ID) ||
> +	     (SLIMPRO_MSG_TYPE(((u32 *) msg)[0]) == SLIMPRO_MSG_TYPE_DBG_ID &&
> +	      SLIMPRO_MSG_SUBTYPE(((u32 *) msg)[0]) == SLIMPRO_DBG_SUBTYPE_SENSOR_READ) ||
> +	     (SLIMPRO_MSG_TYPE(((u32 *) msg)[0]) == SLIMPRO_MSG_TYPE_PWRMGMT_ID &&
> +	      SLIMPRO_MSG_SUBTYPE(((u32 *) msg)[0]) == SLIMPRO_PWRMGMT_SUBTYPE_TPC &&
> +	      SLIMPRO_TPC_CMD(((u32 *) msg)[0]) == SLIMPRO_TPC_ALARM))) {
> +		if (ACPI_COMPANION(ctx->dev)) {
> +			/* Check if platform completes command */
> +			if (!xgene_word_tst_and_clr(&generic_comm_base->status,
> +						    PCCS_CMD_COMPLETE))
> +				goto notify;

How about reversing the logic here and dropping the goto ?

Overall, I have to say that the code is quite complex due to the repeated
checks for ACPI. I am close to suggest having two separate drivers,
one for ACPI and one for non-ACPI. Any chance to separate ACPI vs. non-ACPI
code a bit better ?

> +		}
> +		ctx->sync_msg.msg = ((u32 *) msg)[0];
> +		ctx->sync_msg.param1 = ((u32 *) msg)[1];
> +		ctx->sync_msg.param2 = ((u32 *) msg)[2];
> +
> +		/* Operation waiting for response */
> +		complete(&ctx->rd_complete);
> +
> +		return;
> +	}
> +
> +	/*
> +	 * If PCC, platform notifies interrupt to OSPM.
> +	 * OPSM schedules a consumer command to get this information
> +	 * in a workqueue. Platform must wait until OSPM has issued
> +	 * a consumer command that serves this notification.
> +	 */
> +notify:
> +	if (ACPI_COMPANION(ctx->dev)) {
> +		/* Check and clear Platform Notification bit */
> +		if (!xgene_word_tst_and_clr(&generic_comm_base->status,
> +					    PCCS_PLATFORM_NOTIFICATION))
> +			return;
> +	} else {
> +		amsg.msg   = ((u32 *) msg)[0];
> +		amsg.param1 = ((u32 *) msg)[1];
> +		amsg.param2 = ((u32 *) msg)[2];
> +	}
> +
> +	/* Enqueue to the FIFO */
> +	kfifo_in_spinlocked(&ctx->async_msg_fifo, &amsg,
> +			    sizeof(struct slimpro_resp_msg), &ctx->lock);
> +	/* Schedule the bottom handler */
> +	schedule_work(&ctx->workq);
> +}
> +
> +static void xgene_hwmon_tx_done(struct mbox_client *cl, void *msg, int ret)
> +{
> +	if (ret) {
> +		dev_dbg(cl->dev, "TX did not complete: CMD sent:%x, ret:%d\n",
> +			*(u16 *) msg, ret);
> +	} else {
> +		dev_dbg(cl->dev, "TX completed. CMD sent:%x, ret:%d\n",
> +			*(u16 *) msg, ret);
> +	}
> +}
> +
> +static int __init xgene_hwmon_probe(struct platform_device *pdev)
> +{
> +	struct xgene_hwmon_dev *ctx;
> +	struct mbox_client *cl;
> +	int rc;
> +
> +	ctx = devm_kzalloc(&pdev->dev, sizeof(*ctx), GFP_KERNEL);
> +	if (!ctx)
> +		return -ENOMEM;
> +
> +	ctx->dev = &pdev->dev;
> +	platform_set_drvdata(pdev, ctx);
> +	cl = &ctx->mbox_client;
> +
> +	/* Request mailbox channel */
> +	cl->dev = &pdev->dev;
> +	cl->rx_callback = xgene_hwmon_rx_cb;
> +	cl->tx_done = xgene_hwmon_tx_done;
> +	cl->tx_block = false;
> +	cl->tx_tout = MBOX_OP_TIMEOUTMS;
> +	cl->knows_txdone = false;
> +	if (!ACPI_COMPANION(cl->dev)) {
> +		ctx->mbox_chan = mbox_request_channel(cl, MBOX_HWMON_INDEX);
> +		if (IS_ERR(ctx->mbox_chan)) {
> +			dev_err(&pdev->dev,
> +				"SLIMpro mailbox channel request failed\n");
> +			return PTR_ERR(ctx->mbox_chan);
> +		}
> +	} else {
> +		struct acpi_pcct_hw_reduced *cppc_ss;
> +
> +		ctx->mbox_chan =
> +			pcc_mbox_request_channel(cl, SLIMPRO_MSG_PCC_SUBSPACE);
> +		if (IS_ERR(ctx->mbox_chan)) {
> +			dev_err(&pdev->dev,
> +				"PPC channel request failed\n");
> +			return PTR_ERR(ctx->mbox_chan);
> +		}
> +
> +		/*
> +		 * The PCC mailbox controller driver should
> +		 * have parsed the PCCT (global table of all
> +		 * PCC channels) and stored pointers to the
> +		 * subspace communication region in con_priv.
> +		 */
> +		cppc_ss = ctx->mbox_chan->con_priv;
> +		if (!cppc_ss) {
> +			dev_err(&pdev->dev, "PPC subspace not found\n");
> +			rc = -ENODEV;
> +			goto out_mbox_free;
> +		}
> +
> +		if (!ctx->mbox_chan->mbox->txdone_irq) {
> +			dev_err(&pdev->dev, "PCC IRQ not supported\n");
> +			rc = -ENODEV;
> +			goto out_mbox_free;
> +		}
> +
> +		/*
> +		 * This is the shared communication region
> +		 * for the OS and Platform to communicate over.
> +		 */
> +		ctx->comm_base_addr = cppc_ss->base_address;
> +		if (ctx->comm_base_addr) {
> +			ctx->pcc_comm_addr =
> +					acpi_os_ioremap(ctx->comm_base_addr,
> +							cppc_ss->length);
> +		} else {
> +			dev_err(&pdev->dev, "Failed to get PCC comm region\n");
> +			rc = -ENODEV;
> +			goto out_mbox_free;
> +		}
> +
> +		if (!ctx->pcc_comm_addr) {
> +			dev_err(&pdev->dev,
> +				"Failed to ioremap PCC comm region\n");
> +			rc = -ENOMEM;
> +			goto out_mbox_free;
> +		}
> +
> +		/*
> +		 * cppc_ss->latency is just a Nominal value. In reality
> +		 * the remote processor could be much slower to reply.
> +		 * So add an arbitrary amount of wait on top of Nominal.
> +		 */
> +		ctx->usecs_lat = PCC_NUM_RETRIES * cppc_ss->latency;
> +	}
> +
> +	spin_lock_init(&ctx->lock);
> +
> +	rc = kfifo_alloc(&ctx->async_msg_fifo,
> +			 sizeof(struct slimpro_resp_msg) * ASYNC_MSG_FIFO_SIZE,
> +			 GFP_KERNEL);
> +	if (rc)
> +		goto out_mbox_free;
> +
> +	INIT_WORK(&ctx->workq, xgene_hwmon_evt_work);
> +
> +	/* Hook up sysfs for sensor monitor */
> +	rc = sysfs_create_group(&pdev->dev.kobj, &xgene_hwmon_attr_group);
> +	if (rc) {
> +		dev_err(&pdev->dev, "Fail to create sysfs\n");
> +		goto out_kfifo_free;
> +	}
> +
> +	ctx->hwmon_dev = hwmon_device_register(ctx->dev);
> +	if (IS_ERR(ctx->hwmon_dev)) {
> +		dev_err(&pdev->dev, "Failed to register HW monitor device\n");
> +		rc = PTR_ERR(ctx->hwmon_dev);
> +		goto out;
> +	}

Please use hwmon_device_register_with_groups().

> +
> +	dev_info(&pdev->dev, "APM X-Gene SoC HW monitor driver registered\n");
> +
> +	return rc;
> +
> +out:
> +	sysfs_remove_group(&pdev->dev.kobj, &xgene_hwmon_attr_group);
> +out_kfifo_free:
> +	kfifo_free(&ctx->async_msg_fifo);
> +out_mbox_free:
> +	mbox_free_channel(ctx->mbox_chan);
> +
> +	return rc;
> +}
> +
> +static int xgene_hwmon_remove(struct platform_device *pdev)
> +{
> +	struct xgene_hwmon_dev *ctx = platform_get_drvdata(pdev);
> +
> +	hwmon_device_unregister(ctx->hwmon_dev);
> +	sysfs_remove_group(&pdev->dev.kobj, &xgene_hwmon_attr_group);
> +	kfifo_free(&ctx->async_msg_fifo);
> +	mbox_free_channel(ctx->mbox_chan);
> +
> +	return 0;
> +}
> +
> +#ifdef CONFIG_ACPI
> +static const struct acpi_device_id xgene_hwmon_acpi_match[] = {
> +	{"APMC0D29", 0},
> +	{},
> +};
> +MODULE_DEVICE_TABLE(acpi, xgene_hwmon_acpi_match);
> +#endif
> +
> +static const struct of_device_id xgene_hwmon_of_match[] = {
> +	{.compatible = "apm,xgene-slimpro-hwmon"},
> +	{}
> +};
> +MODULE_DEVICE_TABLE(of, xgene_hwmon_of_match);
> +
> +static struct platform_driver xgene_hwmon_driver __refdata = {
> +	.probe = xgene_hwmon_probe,
> +	.remove = xgene_hwmon_remove,
> +	.driver = {
> +		.name = "xgene-slimpro-hwmon",
> +		.of_match_table = xgene_hwmon_of_match,
> +		.acpi_match_table = ACPI_PTR(xgene_hwmon_acpi_match),
> +	},
> +};
> +module_platform_driver(xgene_hwmon_driver);
> +
> +MODULE_DESCRIPTION("APM X-Gene SoC hardware monitor");
> +MODULE_LICENSE("GPL");



More information about the linux-arm-kernel mailing list