All of lore.kernel.org
 help / color / mirror / Atom feed
From: Michael Hennerich <michael.hennerich@analog.com>
To: Lars-Peter Clausen <lars@metafoo.de>, <jic23@kernel.org>,
	<knaack.h@gmx.de>, <pmeerw@pmeerw.net>, <robh+dt@kernel.org>,
	<mark.rutland@arm.com>
Cc: <linux-iio@vger.kernel.org>, <devicetree@vger.kernel.org>,
	<linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] iio/adc/ltc2497: Driver for Linear Technology LTC2497 ADC
Date: Wed, 29 Mar 2017 10:51:57 +0200	[thread overview]
Message-ID: <025f8461-c823-1f07-5742-e488c3e34f9f@analog.com> (raw)
In-Reply-To: <6774c073-7fe4-e259-306e-3cf4621f7c69@metafoo.de>

On 23.03.2017 12:05, Lars-Peter Clausen wrote:

Sorry - I missed some of this review feedback ...

>> +
>> +static int ltc2497_wait_conv(struct ltc2497_st *st)
>> +{
>> +	s64 time_elapsed;
>> +
>> +	time_elapsed = ktime_ms_delta(ktime_get(), st->time_prev);
>> +
>> +	if (time_elapsed < LTC2497_CONVERSION_TIME_MS) {
>> +		/* delay if conversion time not passed
>> +		 * since last read or write
>> +		 */
>> +		msleep(LTC2497_CONVERSION_TIME_MS - time_elapsed);
>
> Considering how long this sleeps msleep_interruptible() might be the better
> choice.

Wondering what should be the outcome of this?
We can't simply replace it. Actually I've seen cases here in drivers/iio 
where the delay is essential, but the return value of 
msleep_interruptible() is not being checked.
Thus causing a malicious access, in case a signal is received.

We must delay here. If we switch to msleep_interruptible() the only 
reason for this would be to cancel the read and return -EINTR to the user.

Also there is another msleep below which would also need this kind of 
handling.

>
>> +		return 0;
>> +	}
>> +
>> +	if (time_elapsed - LTC2497_CONVERSION_TIME_MS <= 0) {
>> +		/* We're in automatic mode -
>> +		 * so the last reading is stil not outdated
>> +		 */
>> +		return 0;
>> +	}
>> +
>> +	return -ETIMEDOUT;
>> +}
>> +
>> +static int ltc2497_read(struct ltc2497_st *st, u8 address, int *val)
>> +{
>> +	struct i2c_client *client = st->client;
>> +	__be32 buf = 0;
>
> transfer buffers must not be on the stack to avoid issues if the controller
> should use DMA.
>
>> +	int ret;
>> +
>> +	ret = ltc2497_wait_conv(st);
>> +	if (ret < 0 || st->addr_prev != address) {
>> +		ret = i2c_smbus_write_byte(st->client, 0xA0 | address);
>> +		if (ret < 0)
>> +			return ret;
>> +		st->addr_prev = address;
>> +		msleep(LTC2497_CONVERSION_TIME_MS);
>> +	}
>> +	ret = i2c_master_recv(client, (char *)&buf, 3);
>> +	if (ret < 0)  {
>> +		dev_err(&client->dev, "i2c_master_recv failed\n");
>> +		return ret;
>> +	}
>> +	st->time_prev = ktime_get();
>> +	*val = (be32_to_cpu(buf) >> 14) - (1 << 17);
>> +
>> +	return ret;
>> +}
> [...]
>


-- 
Greetings,
Michael

--
Analog Devices GmbH      Otl-Aicher Strasse 60-64      80807 München
Sitz der Gesellschaft München, Registergericht München HRB 40368,
Geschäftsführer: Peter Kolberg, Ali Raza Husain, Eileen Wynne

WARNING: multiple messages have this Message-ID (diff)
From: Michael Hennerich <michael.hennerich-OyLXuOCK7orQT0dZR+AlfA@public.gmane.org>
To: Lars-Peter Clausen <lars-Qo5EllUWu/uELgA04lAiVw@public.gmane.org>,
	jic23-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org,
	knaack.h-Mmb7MZpHnFY@public.gmane.org,
	pmeerw-jW+XmwGofnusTnJN9+BGXg@public.gmane.org,
	robh+dt-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org,
	mark.rutland-5wv7dgnIgG8@public.gmane.org
Cc: linux-iio-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	devicetree-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Subject: Re: [PATCH] iio/adc/ltc2497: Driver for Linear Technology LTC2497 ADC
Date: Wed, 29 Mar 2017 10:51:57 +0200	[thread overview]
Message-ID: <025f8461-c823-1f07-5742-e488c3e34f9f@analog.com> (raw)
In-Reply-To: <6774c073-7fe4-e259-306e-3cf4621f7c69-Qo5EllUWu/uELgA04lAiVw@public.gmane.org>

On 23.03.2017 12:05, Lars-Peter Clausen wrote:

Sorry - I missed some of this review feedback ...

>> +
>> +static int ltc2497_wait_conv(struct ltc2497_st *st)
>> +{
>> +	s64 time_elapsed;
>> +
>> +	time_elapsed = ktime_ms_delta(ktime_get(), st->time_prev);
>> +
>> +	if (time_elapsed < LTC2497_CONVERSION_TIME_MS) {
>> +		/* delay if conversion time not passed
>> +		 * since last read or write
>> +		 */
>> +		msleep(LTC2497_CONVERSION_TIME_MS - time_elapsed);
>
> Considering how long this sleeps msleep_interruptible() might be the better
> choice.

Wondering what should be the outcome of this?
We can't simply replace it. Actually I've seen cases here in drivers/iio 
where the delay is essential, but the return value of 
msleep_interruptible() is not being checked.
Thus causing a malicious access, in case a signal is received.

We must delay here. If we switch to msleep_interruptible() the only 
reason for this would be to cancel the read and return -EINTR to the user.

Also there is another msleep below which would also need this kind of 
handling.

>
>> +		return 0;
>> +	}
>> +
>> +	if (time_elapsed - LTC2497_CONVERSION_TIME_MS <= 0) {
>> +		/* We're in automatic mode -
>> +		 * so the last reading is stil not outdated
>> +		 */
>> +		return 0;
>> +	}
>> +
>> +	return -ETIMEDOUT;
>> +}
>> +
>> +static int ltc2497_read(struct ltc2497_st *st, u8 address, int *val)
>> +{
>> +	struct i2c_client *client = st->client;
>> +	__be32 buf = 0;
>
> transfer buffers must not be on the stack to avoid issues if the controller
> should use DMA.
>
>> +	int ret;
>> +
>> +	ret = ltc2497_wait_conv(st);
>> +	if (ret < 0 || st->addr_prev != address) {
>> +		ret = i2c_smbus_write_byte(st->client, 0xA0 | address);
>> +		if (ret < 0)
>> +			return ret;
>> +		st->addr_prev = address;
>> +		msleep(LTC2497_CONVERSION_TIME_MS);
>> +	}
>> +	ret = i2c_master_recv(client, (char *)&buf, 3);
>> +	if (ret < 0)  {
>> +		dev_err(&client->dev, "i2c_master_recv failed\n");
>> +		return ret;
>> +	}
>> +	st->time_prev = ktime_get();
>> +	*val = (be32_to_cpu(buf) >> 14) - (1 << 17);
>> +
>> +	return ret;
>> +}
> [...]
>


-- 
Greetings,
Michael

--
Analog Devices GmbH      Otl-Aicher Strasse 60-64      80807 München
Sitz der Gesellschaft München, Registergericht München HRB 40368,
Geschäftsführer: Peter Kolberg, Ali Raza Husain, Eileen Wynne

  reply	other threads:[~2017-03-29  8:51 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-03-23 10:35 [PATCH] iio/adc/ltc2497: Driver for Linear Technology LTC2497 ADC michael.hennerich
2017-03-23 10:35 ` michael.hennerich
2017-03-23 11:05 ` Lars-Peter Clausen
2017-03-23 11:05   ` Lars-Peter Clausen
2017-03-29  8:51   ` Michael Hennerich [this message]
2017-03-29  8:51     ` Michael Hennerich
2017-03-23 11:28 ` Peter Meerwald-Stadler
2017-03-23 11:28   ` Peter Meerwald-Stadler
2017-03-23 14:45   ` Michael Hennerich
2017-03-23 14:45     ` Michael Hennerich

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=025f8461-c823-1f07-5742-e488c3e34f9f@analog.com \
    --to=michael.hennerich@analog.com \
    --cc=devicetree@vger.kernel.org \
    --cc=jic23@kernel.org \
    --cc=knaack.h@gmx.de \
    --cc=lars@metafoo.de \
    --cc=linux-iio@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mark.rutland@arm.com \
    --cc=pmeerw@pmeerw.net \
    --cc=robh+dt@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.