From: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
To: Jonathan Cameron <jic23@kernel.org>
Cc: Lorenzo Bianconi <lorenzo@kernel.org>,
linux-iio@vger.kernel.org,
Linus Walleij <linus.walleij@linaro.org>,
Denis CIOCCA <denis.ciocca@st.com>
Subject: Re: [PATCH v3] iio: common: st_sensors: fix possible infinite loop in st_sensors_irq_thread
Date: Sun, 13 Dec 2020 20:23:35 +0100 [thread overview]
Message-ID: <CAJ0CqmXmFvEEnt_fQa+H9Lrsu9d-kj+zTWgVXakBF8z2KqEQYA@mail.gmail.com> (raw)
In-Reply-To: <20201213150447.119eec7c@archlinux>
>
> On Tue, 8 Dec 2020 15:36:40 +0100
> Lorenzo Bianconi <lorenzo@kernel.org> wrote:
>
> > Return a boolean value in st_sensors_new_samples_available routine in
> > order to avoid an infinite loop in st_sensors_irq_thread if
> > stat_drdy.addr is not defined or stat_drdy read fails
> >
> > Fixes: 90efe05562921 ("iio: st_sensors: harden interrupt handling")
> > Reported-by: Jonathan Cameron <jic23@kernel.org>
> > Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
> > Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
>
> One trivial inline. Note I'm looking for an Ack form Dennis on this one
> as it may result in a some breakage if any devices are using
> edge interrupts.
Looking at the current code I am wondering if it is possible since we
would have already triggered the infinite loop in this case. I think
nobody is currently using edge interrupts if the status register is
not available. What do you think?
Regards,
Lorenzo
>
> It's possible we would be better falling back to interrupt free support
> in that case rather than failing to probe at all.
> I think that would be best done by moving the check out to the
> various per type drivers so we fail in the same fashion as no irq
> provided + a warning.
>
> thanks,
>
> Jonathan
>
>
> > ---
> > Changes since v2:
> > - return -EOPNOTSUPP if the drv requests edge IRQ and the sensor does not support
> > status register
> >
> > Changes since v1:
> > - return true if the sensor does not have stat_drdy register
> > ---
> > .../common/st_sensors/st_sensors_trigger.c | 31 ++++++++++---------
> > 1 file changed, 17 insertions(+), 14 deletions(-)
> >
> > diff --git a/drivers/iio/common/st_sensors/st_sensors_trigger.c b/drivers/iio/common/st_sensors/st_sensors_trigger.c
> > index 0507283bd4c1..2dbd2646e44e 100644
> > --- a/drivers/iio/common/st_sensors/st_sensors_trigger.c
> > +++ b/drivers/iio/common/st_sensors/st_sensors_trigger.c
> > @@ -23,35 +23,31 @@
> > * @sdata: Sensor data.
> > *
> > * returns:
> > - * 0 - no new samples available
> > - * 1 - new samples available
> > - * negative - error or unknown
> > + * false - no new samples available or read error
> > + * true - new samples available
> > */
> > -static int st_sensors_new_samples_available(struct iio_dev *indio_dev,
> > - struct st_sensor_data *sdata)
> > +static bool st_sensors_new_samples_available(struct iio_dev *indio_dev,
> > + struct st_sensor_data *sdata)
> > {
> > int ret, status;
> >
> > /* How would I know if I can't check it? */
> > if (!sdata->sensor_settings->drdy_irq.stat_drdy.addr)
> > - return -EINVAL;
> > + return true;
> >
> > /* No scan mask, no interrupt */
> > if (!indio_dev->active_scan_mask)
> > - return 0;
> > + return false;
> >
> > ret = regmap_read(sdata->regmap,
> > sdata->sensor_settings->drdy_irq.stat_drdy.addr,
> > &status);
> > if (ret < 0) {
> > dev_err(sdata->dev, "error checking samples available\n");
> > - return ret;
> > + return false;
> > }
> >
> > - if (status & sdata->sensor_settings->drdy_irq.stat_drdy.mask)
> > - return 1;
> > -
> > - return 0;
> > + return !!(status & sdata->sensor_settings->drdy_irq.stat_drdy.mask);
>
> No need for the !! as you can rely on type conversion to a boolean.
>
> > }
> >
> > /**
> > @@ -180,9 +176,15 @@ int st_sensors_allocate_trigger(struct iio_dev *indio_dev,
> >
> > /* Tell the interrupt handler that we're dealing with edges */
> > if (irq_trig == IRQF_TRIGGER_FALLING ||
> > - irq_trig == IRQF_TRIGGER_RISING)
> > + irq_trig == IRQF_TRIGGER_RISING) {
> > + if (!sdata->sensor_settings->drdy_irq.stat_drdy.addr) {
> > + dev_err(&indio_dev->dev,
> > + "edge IRQ not supported w/o stat register.\n");
> > + err = -EOPNOTSUPP;
> > + goto iio_trigger_free;
> > + }
> > sdata->edge_irq = true;
> > - else
> > + } else {
> > /*
> > * If we're not using edges (i.e. level interrupts) we
> > * just mask off the IRQ, handle one interrupt, then
> > @@ -190,6 +192,7 @@ int st_sensors_allocate_trigger(struct iio_dev *indio_dev,
> > * interrupt handler top half again and start over.
> > */
> > irq_trig |= IRQF_ONESHOT;
> > + }
> >
> > /*
> > * If the interrupt pin is Open Drain, by definition this
>
next prev parent reply other threads:[~2020-12-13 19:25 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-12-08 14:36 [PATCH v3] iio: common: st_sensors: fix possible infinite loop in st_sensors_irq_thread Lorenzo Bianconi
2020-12-09 14:41 ` Andy Shevchenko
2020-12-09 14:42 ` Andy Shevchenko
2020-12-13 15:04 ` Jonathan Cameron
2020-12-13 19:23 ` Lorenzo Bianconi [this message]
2020-12-22 10:53 ` Lorenzo Bianconi
2020-12-30 12:08 ` Jonathan Cameron
2021-01-11 8:21 ` Lorenzo Bianconi
2021-01-11 13:10 ` Jonathan Cameron
2021-01-11 13:26 ` Lorenzo Bianconi
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=CAJ0CqmXmFvEEnt_fQa+H9Lrsu9d-kj+zTWgVXakBF8z2KqEQYA@mail.gmail.com \
--to=lorenzo.bianconi@redhat.com \
--cc=denis.ciocca@st.com \
--cc=jic23@kernel.org \
--cc=linus.walleij@linaro.org \
--cc=linux-iio@vger.kernel.org \
--cc=lorenzo@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).