lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20181216134913.1751b640@archlinux>
Date:   Sun, 16 Dec 2018 13:49:13 +0000
From:   Jonathan Cameron <jic23@...nel.org>
To:     Stefan Popa <stefan.popa@...log.com>
Cc:     <robh+dt@...nel.org>, <mark.rutland@....com>,
        <Michael.Hennerich@...log.com>, <knaack.h@....de>,
        <lars@...afoo.de>, <pmeerw@...erw.net>,
        <gregkh@...uxfoundation.org>, <linux-kernel@...r.kernel.org>,
        <linux-iio@...r.kernel.org>, <devel@...verdev.osuosl.org>,
        <stefan.popa@...log.co>
Subject: Re: [PATCH 05/11] staging: iio: adc: ad7606: Add support for
 threaded irq

On Thu, 13 Dec 2018 14:46:17 +0200
Stefan Popa <stefan.popa@...log.com> wrote:

> This patch replaces the use of a polling ring buffer with a threaded
> interrupt.
> 
> Enabling the buffer sets the CONVST signal to high. When the rising edge
> of the CONVST is applied, BUSY signal goes logic high and transitions low
> at the end of the entire conversion process. The falling edge of the BUSY
> signal triggers the interrupt.
> 
> ad7606_trigger_handler() is used as bottom half of the poll function.
> It reads data from the device and stores it in the internal buffer.
> 
> Signed-off-by: Stefan Popa <stefan.popa@...log.com>
I'd like a little more info as comments in this one. See below.
Which is another way of saying I have no idea what is going on :)

Thanks,

Jonathan.

> ---
>  drivers/staging/iio/adc/ad7606.c | 116 +++++++++++++++++++++++++++++----------
>  drivers/staging/iio/adc/ad7606.h |   6 +-
>  2 files changed, 89 insertions(+), 33 deletions(-)
> 
> diff --git a/drivers/staging/iio/adc/ad7606.c b/drivers/staging/iio/adc/ad7606.c
> index 7191d51..13aeeec 100644
> --- a/drivers/staging/iio/adc/ad7606.c
> +++ b/drivers/staging/iio/adc/ad7606.c
> @@ -21,6 +21,7 @@
>  #include <linux/iio/sysfs.h>
>  #include <linux/iio/buffer.h>
>  #include <linux/iio/trigger_consumer.h>
> +#include <linux/iio/trigger.h>
>  #include <linux/iio/triggered_buffer.h>
>  
>  #include "ad7606.h"
> @@ -81,36 +82,24 @@ static int ad7606_read_samples(struct ad7606_state *st)
>  static irqreturn_t ad7606_trigger_handler(int irq, void *p)
>  {
>  	struct iio_poll_func *pf = p;
> -	struct ad7606_state *st = iio_priv(pf->indio_dev);
> -
> -	gpiod_set_value(st->gpio_convst, 1);
> -
> -	return IRQ_HANDLED;
> -}
> -
> -/**
> - * ad7606_poll_bh_to_ring() bh of trigger launched polling to ring buffer
> - * @work_s:	the work struct through which this was scheduled
> - *
> - * Currently there is no option in this driver to disable the saving of
> - * timestamps within the ring.
> - * I think the one copy of this at a time was to avoid problems if the
> - * trigger was set far too high and the reads then locked up the computer.
> - **/
> -static void ad7606_poll_bh_to_ring(struct work_struct *work_s)
> -{
> -	struct ad7606_state *st = container_of(work_s, struct ad7606_state,
> -						poll_work);
> -	struct iio_dev *indio_dev = iio_priv_to_dev(st);
> +	struct iio_dev *indio_dev = pf->indio_dev;
> +	struct ad7606_state *st = iio_priv(indio_dev);
>  	int ret;
>  
> +	mutex_lock(&st->lock);
> +
>  	ret = ad7606_read_samples(st);
>  	if (ret == 0)
>  		iio_push_to_buffers_with_timestamp(indio_dev, st->data,
>  						   iio_get_time_ns(indio_dev));
>  
> -	gpiod_set_value(st->gpio_convst, 0);
>  	iio_trigger_notify_done(indio_dev->trig);
> +	/* The rising edge of the CONVST signal starts a new conversion. */
> +	gpiod_set_value(st->gpio_convst, 1);
> +
> +	mutex_unlock(&st->lock);
> +
> +	return IRQ_HANDLED;
>  }
>  
>  static int ad7606_scan_direct(struct iio_dev *indio_dev, unsigned int ch)
> @@ -378,8 +367,11 @@ static int ad7606_request_gpios(struct ad7606_state *st)
>  	return PTR_ERR_OR_ZERO(st->gpio_os);
>  }
>  
> -/**
> - *  Interrupt handler
> +/*
> + * The BUSY signal indicates when conversions are in progress, so when a rising
> + * edge of CONVST is applied, BUSY goes logic high and transitions low at the
> + * end of the entire conversion process. The falling edge of the BUSY signal
> + * triggers this interrupt.
>   */
>  static irqreturn_t ad7606_interrupt(int irq, void *dev_id)
>  {
> @@ -387,7 +379,8 @@ static irqreturn_t ad7606_interrupt(int irq, void *dev_id)
>  	struct ad7606_state *st = iio_priv(indio_dev);
>  
>  	if (iio_buffer_enabled(indio_dev)) {
> -		schedule_work(&st->poll_work);
> +		gpiod_set_value(st->gpio_convst, 0);
> +		iio_trigger_poll_chained(st->trig);
>  	} else {
>  		complete(&st->completion);
>  	}
> @@ -395,26 +388,74 @@ static irqreturn_t ad7606_interrupt(int irq, void *dev_id)
>  	return IRQ_HANDLED;
>  };
>  
> +static int ad7606_validate_trigger(struct iio_dev *indio_dev,
> +				   struct iio_trigger *trig)
> +{
> +	struct ad7606_state *st = iio_priv(indio_dev);
> +
> +	if (st->trig != trig)
> +		return -EINVAL;
> +
> +	return 0;
> +}
> +
> +static int ad7606_buffer_postenable(struct iio_dev *indio_dev)
> +{
> +	struct ad7606_state *st = iio_priv(indio_dev);
> +
> +	iio_triggered_buffer_postenable(indio_dev);
> +	gpiod_set_value(st->gpio_convst, 1);
> +
> +	return 0;
> +}
> +
> +static int ad7606_buffer_predisable(struct iio_dev *indio_dev)
> +{
> +	struct ad7606_state *st = iio_priv(indio_dev);
> +	int ret;
> +
> +	reinit_completion(&st->completion);
> +	gpiod_set_value(st->gpio_convst, 1);
> +	ret = wait_for_completion_timeout(&st->completion,
> +					  msecs_to_jiffies(1000));

Along with Dan's comment. I'd like to see a comment here on what
is actually going on.  Not immediately obvious a conversion should
be triggered to disable the buffer...

I'll guess there is a race against the normal handler that we
are closing with this dance, but that race needs explaining.

> +	gpiod_set_value(st->gpio_convst, 0);
> +
> +	return iio_triggered_buffer_predisable(indio_dev);
> +}
> +
> +static const struct iio_buffer_setup_ops ad7606_buffer_ops = {
> +	.postenable = &ad7606_buffer_postenable,
> +	.predisable = &ad7606_buffer_predisable,
> +};
> +
>  static const struct iio_info ad7606_info_no_os_or_range = {
>  	.read_raw = &ad7606_read_raw,
> +	.validate_trigger = &ad7606_validate_trigger,
>  };
>  
>  static const struct iio_info ad7606_info_os_and_range = {
>  	.read_raw = &ad7606_read_raw,
>  	.write_raw = &ad7606_write_raw,
>  	.attrs = &ad7606_attribute_group_os_and_range,
> +	.validate_trigger = &ad7606_validate_trigger,
>  };
>  
>  static const struct iio_info ad7606_info_os = {
>  	.read_raw = &ad7606_read_raw,
>  	.write_raw = &ad7606_write_raw,
>  	.attrs = &ad7606_attribute_group_os,
> +	.validate_trigger = &ad7606_validate_trigger,
>  };
>  
>  static const struct iio_info ad7606_info_range = {
>  	.read_raw = &ad7606_read_raw,
>  	.write_raw = &ad7606_write_raw,
>  	.attrs = &ad7606_attribute_group_range,
> +	.validate_trigger = &ad7606_validate_trigger,
> +};
> +
> +static const struct iio_trigger_ops ad7606_trigger_ops = {
> +	.validate_device = iio_trigger_validate_own_device,
>  };
>  
>  static void ad7606_regulator_disable(void *data)
> @@ -446,7 +487,6 @@ int ad7606_probe(struct device *dev, int irq, void __iomem *base_address,
>  	/* tied to logic low, analog input range is +/- 5V */
>  	st->range = 0;
>  	st->oversampling = 1;
> -	INIT_WORK(&st->poll_work, &ad7606_poll_bh_to_ring);
>  
>  	st->reg = devm_regulator_get(dev, "avcc");
>  	if (IS_ERR(st->reg))
> @@ -491,14 +531,32 @@ int ad7606_probe(struct device *dev, int irq, void __iomem *base_address,
>  	if (ret)
>  		dev_warn(st->dev, "failed to RESET: no RESET GPIO specified\n");
>  
> -	ret = devm_request_irq(dev, irq, ad7606_interrupt, IRQF_TRIGGER_FALLING,
> -			       name, indio_dev);
> +	st->trig = devm_iio_trigger_alloc(dev, "%s-dev%d",
> +					  indio_dev->name, indio_dev->id);
> +	if (!st->trig)
> +		return -ENOMEM;
> +
> +	st->trig->ops = &ad7606_trigger_ops;
> +	st->trig->dev.parent = dev;
> +	iio_trigger_set_drvdata(st->trig, indio_dev);
> +	ret = devm_iio_trigger_register(dev, st->trig);
> +	if (ret)
> +		return ret;
> +
> +	indio_dev->trig = iio_trigger_get(st->trig);
> +
> +	ret = devm_request_threaded_irq(dev, irq,
> +					NULL,
> +					&ad7606_interrupt,
> +					IRQF_TRIGGER_FALLING | IRQF_ONESHOT,
> +					name, indio_dev);
>  	if (ret)
>  		return ret;
>  
>  	ret = devm_iio_triggered_buffer_setup(dev, indio_dev,
> +					      &iio_pollfunc_store_time,
>  					      &ad7606_trigger_handler,
> -					      NULL, NULL);
> +					      &ad7606_buffer_ops);
>  	if (ret)
>  		return ret;
>  
> diff --git a/drivers/staging/iio/adc/ad7606.h b/drivers/staging/iio/adc/ad7606.h
> index 70486ef..b238e9b 100644
> --- a/drivers/staging/iio/adc/ad7606.h
> +++ b/drivers/staging/iio/adc/ad7606.h
> @@ -26,8 +26,6 @@ struct ad7606_chip_info {
>   * @dev		pointer to kernel device
>   * @chip_info		entry in the table of chips that describes this device
>   * @reg		regulator info for the the power supply of the device
> - * @poll_work		work struct for continuously reading data from the device
> - *			into an IIO triggered buffer
>   * @bops		bus operations (SPI or parallel)
>   * @range		voltage range selection, selects which scale to apply
>   * @oversampling	oversampling selection
> @@ -42,14 +40,13 @@ struct ad7606_chip_info {
>   *			is being read on the first channel
>   * @gpio_os		GPIO descriptors to control oversampling on the device
>   * @complete		completion to indicate end of conversion
> + * @trig		The IIO trigger associated with the device.
>   * @data		buffer for reading data from the device
>   */
> -
>  struct ad7606_state {
>  	struct device			*dev;
>  	const struct ad7606_chip_info	*chip_info;
>  	struct regulator		*reg;
> -	struct work_struct		poll_work;
>  	const struct ad7606_bus_ops	*bops;
>  	unsigned int			range;
>  	unsigned int			oversampling;
> @@ -62,6 +59,7 @@ struct ad7606_state {
>  	struct gpio_desc		*gpio_standby;
>  	struct gpio_desc		*gpio_frstdata;
>  	struct gpio_descs		*gpio_os;
> +	struct iio_trigger		*trig;
>  	struct completion		completion;
>  
>  	/*

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ