[v5,3/3] w1: add UART w1 bus driver

From: Christoph Winklhofer <cj.winklhofer@gmail.com>

Add a UART 1-Wire bus driver. The driver utilizes the UART interface via
the Serial Device Bus to create the 1-Wire timing patterns. The driver
was tested on a "Raspberry Pi 3B" with a DS18B20 and on a "Variscite
DART-6UL" with a DS18S20 temperature sensor.

The 1-Wire timing pattern and the corresponding UART baud-rate with the
interpretation of the transferred bytes are described in the document:

Link: https://www.analog.com/en/technical-articles/using-a-uart-to-implement-a-1wire-bus-master.html

In short, the UART peripheral must support full-duplex and operate in
open-drain mode. The timing patterns are generated by a specific
combination of baud-rate and transmitted byte, which corresponds to a
1-Wire read bit, write bit or reset.

Signed-off-by: Christoph Winklhofer <cj.winklhofer@gmail.com>
---
 Documentation/w1/masters/index.rst   |   1 +
 Documentation/w1/masters/w1-uart.rst |  54 +++++
 drivers/w1/masters/Kconfig           |  10 +
 drivers/w1/masters/Makefile          |   1 +
 drivers/w1/masters/w1-uart.c         | 402 +++++++++++++++++++++++++++++++++++
 5 files changed, 468 insertions(+)

On 26/01/2024 16:42, Christoph Winklhofer via B4 Relay wrote:
> From: Christoph Winklhofer <cj.winklhofer@gmail.com>
> 
> Add a UART 1-Wire bus driver. The driver utilizes the UART interface via
> the Serial Device Bus to create the 1-Wire timing patterns. The driver
> was tested on a "Raspberry Pi 3B" with a DS18B20 and on a "Variscite
> DART-6UL" with a DS18S20 temperature sensor.
> 

...

> + * struct w1_uart_config - configuration for 1-Wire operation
> + *
> + * @baudrate: baud-rate returned from serdev
> + * @delay_us: delay to complete a 1-Wire cycle (in us)
> + * @tx_byte: byte to generate 1-Wire timing pattern
> + */
> +struct w1_uart_config {
> +	unsigned int baudrate;
> +	unsigned int delay_us;
> +	u8 tx_byte;
> +};
> +
> +struct w1_uart_device {
> +	struct serdev_device *serdev;
> +	struct w1_bus_master bus;
> +
> +	struct w1_uart_config cfg_reset;
> +	struct w1_uart_config cfg_touch_0;
> +	struct w1_uart_config cfg_touch_1;
> +
> +	struct completion rx_byte_received;
> +	int rx_err;
> +	u8 rx_byte;
> +

Missing documentation of mutex scope. What does it protect?

> +	struct mutex mutex;
> +};
> +

...

> +/*
> + * Send one byte (tx_byte) and read one byte (rx_byte) via serdev.
> + */
> +static int w1_uart_serdev_tx_rx(struct w1_uart_device *w1dev,
> +				const struct w1_uart_config *w1cfg, u8 *rx_byte)
> +{
> +	struct serdev_device *serdev = w1dev->serdev;
> +	int ret;
> +
> +	serdev_device_write_flush(serdev);
> +	serdev_device_set_baudrate(serdev, w1cfg->baudrate);
> +
> +	/* write and immediately read one byte */
> +	reinit_completion(&w1dev->rx_byte_received);
> +	ret = serdev_device_write_buf(serdev, &w1cfg->tx_byte, 1);
> +	if (ret != 1)
> +		return -EIO;
> +	ret = wait_for_completion_interruptible_timeout(
> +		&w1dev->rx_byte_received, W1_UART_TIMEOUT);
> +	if (ret <= 0)
> +		return -EIO;
> +
> +	/* locking could fail during driver remove or when serdev is

It's not netdev, so:
/*
 *

> +	 * unexpectedly in the receive callback.
> +	 */
> +	if (!mutex_trylock(&w1dev->mutex))
> +		return -EIO;
> +
> +	ret = w1dev->rx_err;
> +	if (ret == 0)
> +		*rx_byte = w1dev->rx_byte;
> +
> +	if (w1cfg->delay_us > 0)
> +		fsleep(w1cfg->delay_us);
> +
> +	mutex_unlock(&w1dev->mutex);
> +
> +	return ret;
> +}
> +
> +static ssize_t w1_uart_serdev_receive_buf(struct serdev_device *serdev,
> +					  const u8 *buf, size_t count)
> +{
> +	struct w1_uart_device *w1dev = serdev_device_get_drvdata(serdev);
> +
> +	mutex_lock(&w1dev->mutex);
> +
> +	/* sent a single byte and receive one single byte */
> +	if (count == 1) {
> +		w1dev->rx_byte = buf[0];
> +		w1dev->rx_err = 0;
> +	} else {
> +		w1dev->rx_err = -EIO;
> +	}
> +
> +	mutex_unlock(&w1dev->mutex);
> +	complete(&w1dev->rx_byte_received);
> +
> +	return count;
> +}
> +
> +static const struct serdev_device_ops w1_uart_serdev_ops = {
> +	.receive_buf = w1_uart_serdev_receive_buf,
> +	.write_wakeup = serdev_device_write_wakeup,
> +};
> +
> +/*
> + * 1-wire reset and presence detect: A present slave will manipulate
> + * the received byte by pulling the 1-Wire low.
> + */
> +static u8 w1_uart_reset_bus(void *data)
> +{
> +	struct w1_uart_device *w1dev = data;
> +	const struct w1_uart_config *w1cfg = &w1dev->cfg_reset;
> +	int ret;
> +	u8 val;
> +
> +	ret = w1_uart_serdev_tx_rx(w1dev, w1cfg, &val);
> +	if (ret < 0)
> +		return -1;
> +
> +	/* Device present (0) or no device (1) */
> +	return val != w1cfg->tx_byte ? 0 : 1;
> +}
> +
> +/*
> + * 1-Wire read and write cycle: Only the read-0 manipulates the
> + * received byte, all others left the line untouched.
> + */
> +static u8 w1_uart_touch_bit(void *data, u8 bit)
> +{
> +	struct w1_uart_device *w1dev = data;
> +	const struct w1_uart_config *w1cfg = bit ? &w1dev->cfg_touch_1 :
> +						   &w1dev->cfg_touch_0;
> +	int ret;
> +	u8 val;
> +
> +	ret = w1_uart_serdev_tx_rx(w1dev, w1cfg, &val);
> +
> +	/* return inactive bus state on error */
> +	if (ret < 0)
> +		return 1;
> +
> +	return val == w1cfg->tx_byte ? 1 : 0;
> +}
> +
> +static int w1_uart_probe(struct serdev_device *serdev)
> +{
> +	struct device *dev = &serdev->dev;
> +	struct w1_uart_device *w1dev;
> +	int ret;
> +
> +	w1dev = devm_kzalloc(dev, sizeof(*w1dev), GFP_KERNEL);
> +	if (!w1dev)
> +		return -ENOMEM;
> +	w1dev->bus.data = w1dev;
> +	w1dev->bus.reset_bus = w1_uart_reset_bus;
> +	w1dev->bus.touch_bit = w1_uart_touch_bit;
> +	w1dev->serdev = serdev;
> +
> +	init_completion(&w1dev->rx_byte_received);
> +	mutex_init(&w1dev->mutex);
> +
> +	ret = w1_uart_serdev_open(w1dev);
> +	if (ret < 0)
> +		return ret;
> +	serdev_device_set_drvdata(serdev, w1dev);
> +	serdev_device_set_client_ops(serdev, &w1_uart_serdev_ops);
> +
> +	return w1_add_master_device(&w1dev->bus);
> +}
> +
> +static void w1_uart_remove(struct serdev_device *serdev)
> +{
> +	struct w1_uart_device *w1dev = serdev_device_get_drvdata(serdev);
> +
> +	mutex_lock(&w1dev->mutex);
> +
> +	w1_remove_master_device(&w1dev->bus);
> +
> +	mutex_unlock(&w1dev->mutex);

This is still suspicious. You do not have serdev_device_close and you
want to protect from concurrent access but it looks insufficient.

This code assumes that:

w1_uart_remove()
  <-- here concurrent read/write might start
  mutex_lock()
  w1_remove_master_device()
  mutex_unlock()
  <-- now w1_uart_serdev_tx_rx() or w1_uart_serdev_receive_buf() can be
executed, but device is removed. So what's the point of the mutex here?

What exactly is protected by the mutex? So far it looks like only some
contents of w1dev, but it does not matter, because it that memory is
still valid at this point.

After describing what is protected we can think whether it is really
protected...


> 

Best regards,
Krzysztof

On Wed, Jan 31, 2024 at 02:12:34PM +0100, Krzysztof Kozlowski wrote:
> On 26/01/2024 16:42, Christoph Winklhofer via B4 Relay wrote:
> > From: Christoph Winklhofer <cj.winklhofer@gmail.com>
> > 
> > Add a UART 1-Wire bus driver. The driver utilizes the UART interface via
> > the Serial Device Bus to create the 1-Wire timing patterns. The driver
> > was tested on a "Raspberry Pi 3B" with a DS18B20 and on a "Variscite
> > DART-6UL" with a DS18S20 temperature sensor.
> > 
> 
> ...
> 
> > + * struct w1_uart_config - configuration for 1-Wire operation
> > + *
> > + * @baudrate: baud-rate returned from serdev
> > + * @delay_us: delay to complete a 1-Wire cycle (in us)
> > + * @tx_byte: byte to generate 1-Wire timing pattern
> > + */
> > +struct w1_uart_config {
> > +	unsigned int baudrate;
> > +	unsigned int delay_us;
> > +	u8 tx_byte;
> > +};
> > +
> > +struct w1_uart_device {
> > +	struct serdev_device *serdev;
> > +	struct w1_bus_master bus;
> > +
> > +	struct w1_uart_config cfg_reset;
> > +	struct w1_uart_config cfg_touch_0;
> > +	struct w1_uart_config cfg_touch_1;
> > +
> > +	struct completion rx_byte_received;
> > +	int rx_err;
> > +	u8 rx_byte;
> > +
> 
> Missing documentation of mutex scope. What does it protect?
> 

The mutex should protect concurrent access to rx_err and rx_byte. It
would be not be required in the good case: a write is initiated solely
by the w1-callbacks in 'w1_uart_serdev_tx_rx' and a completion is used
to wait for the result of serdev-receive.

However, in case the UART is not configured as a loop, a serdev-receive
may occur when w1_uart_serdev_tx_rx evaluates rx_err and rx_byte in
w1_uart_serdev_tx_rx, so it is protected - however, I will try to find
a better way to detect such an error.

In addition, the w1-callbacks should also return during a 'remove' (with
the mutex_try_lock) - see comment on that below.

> > +	struct mutex mutex;
> > +};
> > +
> 
> ...
> 
> > +/*
> > + * Send one byte (tx_byte) and read one byte (rx_byte) via serdev.
> > + */
> > +static int w1_uart_serdev_tx_rx(struct w1_uart_device *w1dev,
> > +				const struct w1_uart_config *w1cfg, u8 *rx_byte)
> > +{
> > +	struct serdev_device *serdev = w1dev->serdev;
> > +	int ret;
> > +
> > +	serdev_device_write_flush(serdev);
> > +	serdev_device_set_baudrate(serdev, w1cfg->baudrate);
> > +
> > +	/* write and immediately read one byte */
> > +	reinit_completion(&w1dev->rx_byte_received);
> > +	ret = serdev_device_write_buf(serdev, &w1cfg->tx_byte, 1);
> > +	if (ret != 1)
> > +		return -EIO;
> > +	ret = wait_for_completion_interruptible_timeout(
> > +		&w1dev->rx_byte_received, W1_UART_TIMEOUT);
> > +	if (ret <= 0)
> > +		return -EIO;
> > +
> > +	/* locking could fail during driver remove or when serdev is
> 
> It's not netdev, so:
> /*
>  *
> 

Ok.

> > +	 * unexpectedly in the receive callback.
> > +	 */
> > +	if (!mutex_trylock(&w1dev->mutex))
> > +		return -EIO;
> > +
> > +	ret = w1dev->rx_err;
> > +	if (ret == 0)
> > +		*rx_byte = w1dev->rx_byte;
> > +
> > +	if (w1cfg->delay_us > 0)
> > +		fsleep(w1cfg->delay_us);
> > +
> > +	mutex_unlock(&w1dev->mutex);
> > +
> > +	return ret;
> > +}
> > +
> > +static ssize_t w1_uart_serdev_receive_buf(struct serdev_device *serdev,
> > +					  const u8 *buf, size_t count)
> > +{
> > +	struct w1_uart_device *w1dev = serdev_device_get_drvdata(serdev);
> > +
> > +	mutex_lock(&w1dev->mutex);
> > +
> > +	/* sent a single byte and receive one single byte */
> > +	if (count == 1) {
> > +		w1dev->rx_byte = buf[0];
> > +		w1dev->rx_err = 0;
> > +	} else {
> > +		w1dev->rx_err = -EIO;
> > +	}
> > +
> > +	mutex_unlock(&w1dev->mutex);
> > +	complete(&w1dev->rx_byte_received);
> > +
> > +	return count;
> > +}
> > +
> > +static const struct serdev_device_ops w1_uart_serdev_ops = {
> > +	.receive_buf = w1_uart_serdev_receive_buf,
> > +	.write_wakeup = serdev_device_write_wakeup,
> > +};
> > +
> > +/*
> > + * 1-wire reset and presence detect: A present slave will manipulate
> > + * the received byte by pulling the 1-Wire low.
> > + */
> > +static u8 w1_uart_reset_bus(void *data)
> > +{
> > +	struct w1_uart_device *w1dev = data;
> > +	const struct w1_uart_config *w1cfg = &w1dev->cfg_reset;
> > +	int ret;
> > +	u8 val;
> > +
> > +	ret = w1_uart_serdev_tx_rx(w1dev, w1cfg, &val);
> > +	if (ret < 0)
> > +		return -1;
> > +
> > +	/* Device present (0) or no device (1) */
> > +	return val != w1cfg->tx_byte ? 0 : 1;
> > +}
> > +
> > +/*
> > + * 1-Wire read and write cycle: Only the read-0 manipulates the
> > + * received byte, all others left the line untouched.
> > + */
> > +static u8 w1_uart_touch_bit(void *data, u8 bit)
> > +{
> > +	struct w1_uart_device *w1dev = data;
> > +	const struct w1_uart_config *w1cfg = bit ? &w1dev->cfg_touch_1 :
> > +						   &w1dev->cfg_touch_0;
> > +	int ret;
> > +	u8 val;
> > +
> > +	ret = w1_uart_serdev_tx_rx(w1dev, w1cfg, &val);
> > +
> > +	/* return inactive bus state on error */
> > +	if (ret < 0)
> > +		return 1;
> > +
> > +	return val == w1cfg->tx_byte ? 1 : 0;
> > +}
> > +
> > +static int w1_uart_probe(struct serdev_device *serdev)
> > +{
> > +	struct device *dev = &serdev->dev;
> > +	struct w1_uart_device *w1dev;
> > +	int ret;
> > +
> > +	w1dev = devm_kzalloc(dev, sizeof(*w1dev), GFP_KERNEL);
> > +	if (!w1dev)
> > +		return -ENOMEM;
> > +	w1dev->bus.data = w1dev;
> > +	w1dev->bus.reset_bus = w1_uart_reset_bus;
> > +	w1dev->bus.touch_bit = w1_uart_touch_bit;
> > +	w1dev->serdev = serdev;
> > +
> > +	init_completion(&w1dev->rx_byte_received);
> > +	mutex_init(&w1dev->mutex);
> > +
> > +	ret = w1_uart_serdev_open(w1dev);
> > +	if (ret < 0)
> > +		return ret;
> > +	serdev_device_set_drvdata(serdev, w1dev);
> > +	serdev_device_set_client_ops(serdev, &w1_uart_serdev_ops);
> > +
> > +	return w1_add_master_device(&w1dev->bus);
> > +}
> > +
> > +static void w1_uart_remove(struct serdev_device *serdev)
> > +{
> > +	struct w1_uart_device *w1dev = serdev_device_get_drvdata(serdev);
> > +
> > +	mutex_lock(&w1dev->mutex);
> > +
> > +	w1_remove_master_device(&w1dev->bus);
> > +
> > +	mutex_unlock(&w1dev->mutex);
> 
> This is still suspicious. You do not have serdev_device_close and you
> want to protect from concurrent access but it looks insufficient.
> 
> This code assumes that:
> 
> w1_uart_remove()
>   <-- here concurrent read/write might start
>   mutex_lock()
>   w1_remove_master_device()
>   mutex_unlock()
>   <-- now w1_uart_serdev_tx_rx() or w1_uart_serdev_receive_buf() can be
> executed, but device is removed. So what's the point of the mutex here?
> 
> What exactly is protected by the mutex? So far it looks like only some
> contents of w1dev, but it does not matter, because it that memory is
> still valid at this point.
> 
> After describing what is protected we can think whether it is really
> protected...
> 
> 
> > 
> 
> Best regards,
> Krzysztof
> 

Yes, it is still suspicious, sorry..

After w1_uart_remove, serdev is closed and w1dev is released. Therefore
the w1-callback (w1_uart_serdev_tx_rx) must be finished before returning
from w1_uart_remove. That was the intention with the lock and trylock.

I thought that after w1_remove_master_device, the w1-callback
(w1_uart_serdev_tx_rx) is finished which is not the case. I will check
the working of w1_remove_master_device, probably it requires a lock to
a mutex from w1-bus.

Many thanks for your review and pointing the things out!

Kind regards,
Christoph

On 01/02/2024 08:29, Christoph Winklhofer wrote:
>>> +
>>> +static void w1_uart_remove(struct serdev_device *serdev)
>>> +{
>>> +	struct w1_uart_device *w1dev = serdev_device_get_drvdata(serdev);
>>> +
>>> +	mutex_lock(&w1dev->mutex);
>>> +
>>> +	w1_remove_master_device(&w1dev->bus);
>>> +
>>> +	mutex_unlock(&w1dev->mutex);
>>
>> This is still suspicious. You do not have serdev_device_close and you
>> want to protect from concurrent access but it looks insufficient.
>>
>> This code assumes that:
>>
>> w1_uart_remove()
>>   <-- here concurrent read/write might start
>>   mutex_lock()
>>   w1_remove_master_device()
>>   mutex_unlock()
>>   <-- now w1_uart_serdev_tx_rx() or w1_uart_serdev_receive_buf() can be
>> executed, but device is removed. So what's the point of the mutex here?
>>
>> What exactly is protected by the mutex? So far it looks like only some
>> contents of w1dev, but it does not matter, because it that memory is
>> still valid at this point.
>>
>> After describing what is protected we can think whether it is really
>> protected...
>>
>>
>>>
>>
>> Best regards,
>> Krzysztof
>>
> 
> Yes, it is still suspicious, sorry..
> 
> After w1_uart_remove, serdev is closed and w1dev is released. Therefore
> the w1-callback (w1_uart_serdev_tx_rx) must be finished before returning

I did not even go to end of w1_uart_remove(). In my code above, that
thread is still in w1_uart_remove(), just after unlocking mutex.

> from w1_uart_remove. That was the intention with the lock and trylock.

Then it does not look really protected. To be honest, w1-gpio and other
drivers also might have a race here. You can test it by adding long
sleeps in read/write paths and then trying to unbind device. Maybe this
should be fixed everywhere, but this mutex here brings more questions.


Best regards,
Krzysztof

On Thu, Feb 01, 2024 at 10:35:32AM +0100, Krzysztof Kozlowski wrote:
> On 01/02/2024 08:29, Christoph Winklhofer wrote:
> >>> +
> >>> +static void w1_uart_remove(struct serdev_device *serdev)
> >>> +{
> >>> +	struct w1_uart_device *w1dev = serdev_device_get_drvdata(serdev);
> >>> +
> >>> +	mutex_lock(&w1dev->mutex);
> >>> +
> >>> +	w1_remove_master_device(&w1dev->bus);
> >>> +
> >>> +	mutex_unlock(&w1dev->mutex);
> >>
> >> This is still suspicious. You do not have serdev_device_close and you
> >> want to protect from concurrent access but it looks insufficient.
> >>
> >> This code assumes that:
> >>
> >> w1_uart_remove()
> >>   <-- here concurrent read/write might start
> >>   mutex_lock()
> >>   w1_remove_master_device()
> >>   mutex_unlock()
> >>   <-- now w1_uart_serdev_tx_rx() or w1_uart_serdev_receive_buf() can be
> >> executed, but device is removed. So what's the point of the mutex here?
> >>
> >> What exactly is protected by the mutex? So far it looks like only some
> >> contents of w1dev, but it does not matter, because it that memory is
> >> still valid at this point.
> >>
> >> After describing what is protected we can think whether it is really
> >> protected...
> >>
> >>
> >>>
> >>
> >> Best regards,
> >> Krzysztof
> >>
> > 
> > Yes, it is still suspicious, sorry..
> > 
> > After w1_uart_remove, serdev is closed and w1dev is released. Therefore
> > the w1-callback (w1_uart_serdev_tx_rx) must be finished before returning
> 
> I did not even go to end of w1_uart_remove(). In my code above, that
> thread is still in w1_uart_remove(), just after unlocking mutex.
> 

Ok, I looked more closely at the underlying w1-code and think it is
sufficient to only call w1_remove_master_device() in w1_uart_remove.
This function waits until all slaves have finished their work, so
w1_uart_serdev_tx_rx finished too. The lock is not required.

Hence, I will use the mutex only to protected concurrent access of
serdev and w1-callbacks (for rx_byte and rx_err).

> > from w1_uart_remove. That was the intention with the lock and trylock.
> 
> Then it does not look really protected. To be honest, w1-gpio and other
> drivers also might have a race here. You can test it by adding long
> sleeps in read/write paths and then trying to unbind device. Maybe this
> should be fixed everywhere, but this mutex here brings more questions.
> 

I added a delay, unbind takes longer but works without problems when its
done during a temperature read.

IMO also the other drivers should be fine. From w1_int.c: The w1_master
is ref-counted and released when it is unused (2). In w1_slave_detach
(1), the slaves decrement this count, perform specific clean up (in
remove_slave) and remove sysfs entries.

w1_int.c:

void __w1_remove_master_device(struct w1_master *dev)
...
	list_for_each_entry_safe(sl, sln, &dev->slist, w1_slave_entry) {
		mutex_unlock(&dev->list_mutex);
		w1_slave_detach(sl); 		(1)
...
	while (atomic_read(&dev->refcnt)) { 	(2)
...	
	}
...
	w1_free_dev(dev);

For example w1_therm waits in remove_slave until its w1-operations are
done. The other slave-drivers (e.g. w1_ds2405.c) use w1-operations in
their sysfs-callback and I suppose that sysfs-removal waits until the
sysfs-read is finished.

Kind regards,
Christoph