| 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
| ||
|
Message-ID: <CACPK8XeOT6NvEe=oBZ9dUJynHougj-mTMAC2FCwDkvpzBaTKDQ@mail.gmail.com>
Date: Tue, 24 May 2022 02:12:45 +0000
From: Joel Stanley <joel@....id.au>
To: Eddie James <eajames@...ux.ibm.com>
Cc: linux-iio@...r.kernel.org,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
Lars-Peter Clausen <lars@...afoo.de>,
Jonathan Cameron <jic23@...nel.org>
Subject: Re: [PATCH v2 2/2] iio: pressure: dps310: Reset chip if MEAS_CFG is corrupt
On Wed, 18 May 2022 at 14:48, Eddie James <eajames@...ux.ibm.com> wrote:
>
> Corruption of the MEAS_CFG register has been observed soon after
> system boot. In order to recover this scenario, check MEAS_CFG if
> measurement isn't ready, and if it's incorrect, reset the DPS310
> and execute the startup procedure.
I have some suggestions below on how to rework to make the code easier
to understand. But before we got to that, I had some high level
questions:
You don't seem to be setting the en bits in the CFG register after
doing the reset. Is that required?
Are we ok to sleep for 2.5ms in the iio_info->read_raw callback?
>
> Fixes: ba6ec48e76bc ("iio: Add driver for Infineon DPS310")
> Signed-off-by: Eddie James <eajames@...ux.ibm.com>
> ---
> drivers/iio/pressure/dps310.c | 89 ++++++++++++++++++++++++++++-------
> 1 file changed, 71 insertions(+), 18 deletions(-)
>
> diff --git a/drivers/iio/pressure/dps310.c b/drivers/iio/pressure/dps310.c
> index f79b274bb38d..c6d02679ef33 100644
> --- a/drivers/iio/pressure/dps310.c
> +++ b/drivers/iio/pressure/dps310.c
> @@ -397,6 +397,39 @@ static int dps310_get_temp_k(struct dps310_data *data)
> return scale_factors[ilog2(rc)];
> }
>
> +/* Called with lock held */
Perhaps add this to your comment: Returns a negative value on error, a
positive value when the device is not ready (and may have been reset
due to corruption), and zero when the device is ready.
> +static int dps310_check_reset_meas_cfg(struct dps310_data *data, int ready_bit)
> +{
> + int en = DPS310_PRS_EN | DPS310_TEMP_EN | DPS310_BACKGROUND;
> + int meas_cfg;
> + int rc = regmap_read(data->regmap, DPS310_MEAS_CFG, &meas_cfg);
> +
> + if (rc < 0)
> + return rc;
> +
> + if (meas_cfg & ready_bit)
> + return 0;
> +
> + if ((meas_cfg & en) != en) {
> + /* DPS310 register state corrupt, better start from scratch */
> + rc = regmap_write(data->regmap, DPS310_RESET,
> + DPS310_RESET_MAGIC);
> + if (rc < 0)
> + return rc;
> +
> + /* Wait for device chip access: 2.5ms in specification */
> + usleep_range(2500, 12000);
> + rc = dps310_startup(data);
> + if (rc)
> + return rc;
> +
> + dev_info(&data->client->dev,
> + "recovered from corrupted MEAS_CFG=%02x\n", meas_cfg);
> + }
> +
> + return 1;
I'm confused about this case. We get there when the device doesn't
have ready_bit set in meas_cfg and we've done a reset, but we also get
here when the bit isn't set and we haven't done anything to resolve it
(after re-reading the code I understand now, but perhaps reworking it
as follows will make it clear):
Could we write it like this:
if (meas_cfg & ready_bit) {
/* Device ready, must be okay */
return 0;
}
if (meas_cfg & en) {
/* Device okay (but not ready), no action required */
return 1;
}
/* DPS310 register state corrupt, better start from scratch */
...
return 1;
> +}
> +
> static int dps310_read_pres_raw(struct dps310_data *data)
> {
> int rc;
> @@ -409,15 +442,25 @@ static int dps310_read_pres_raw(struct dps310_data *data)
> if (mutex_lock_interruptible(&data->lock))
> return -EINTR;
>
> - rate = dps310_get_pres_samp_freq(data);
> - timeout = DPS310_POLL_TIMEOUT_US(rate);
> -
> - /* Poll for sensor readiness; base the timeout upon the sample rate. */
> - rc = regmap_read_poll_timeout(data->regmap, DPS310_MEAS_CFG, ready,
> - ready & DPS310_PRS_RDY,
> - DPS310_POLL_SLEEP_US(timeout), timeout);
> - if (rc)
> - goto done;
> + rc = dps310_check_reset_meas_cfg(data, DPS310_PRS_RDY);
can we do this:
if (rc < 0)
goto done;
if (rc > 0) {
}
The rework I suggest makes it clearer that we've considered the '0'
case, when the device is ready before this code runs.
> + if (rc) {
> + if (rc < 0)
> + goto done;
> +
> + rate = dps310_get_pres_samp_freq(data);
> + timeout = DPS310_POLL_TIMEOUT_US(rate);
> +
> + /*
> + * Poll for sensor readiness; base the timeout upon the sample
> + * rate.
> + */
> + rc = regmap_read_poll_timeout(data->regmap, DPS310_MEAS_CFG,
> + ready, ready & DPS310_PRS_RDY,
> + DPS310_POLL_SLEEP_US(timeout),
> + timeout);
> + if (rc)
> + goto done;
> + }
>
> rc = regmap_bulk_read(data->regmap, DPS310_PRS_BASE, val, sizeof(val));
> if (rc < 0)
> @@ -458,15 +501,25 @@ static int dps310_read_temp_raw(struct dps310_data *data)
> if (mutex_lock_interruptible(&data->lock))
> return -EINTR;
>
> - rate = dps310_get_temp_samp_freq(data);
> - timeout = DPS310_POLL_TIMEOUT_US(rate);
> -
> - /* Poll for sensor readiness; base the timeout upon the sample rate. */
> - rc = regmap_read_poll_timeout(data->regmap, DPS310_MEAS_CFG, ready,
> - ready & DPS310_TMP_RDY,
> - DPS310_POLL_SLEEP_US(timeout), timeout);
> - if (rc < 0)
> - goto done;
> + rc = dps310_check_reset_meas_cfg(data, DPS310_TMP_RDY);
> + if (rc) {
> + if (rc < 0)
> + goto done;
> +
> + rate = dps310_get_temp_samp_freq(data);
> + timeout = DPS310_POLL_TIMEOUT_US(rate);
> +
> + /*
> + * Poll for sensor readiness; base the timeout upon the sample
> + * rate.
> + */
> + rc = regmap_read_poll_timeout(data->regmap, DPS310_MEAS_CFG,
> + ready, ready & DPS310_TMP_RDY,
> + DPS310_POLL_SLEEP_US(timeout),
> + timeout);
> + if (rc < 0)
> + goto done;
> + }
>
> rc = dps310_read_temp_ready(data);
>
> --
> 2.27.0
>
Powered by blists - more mailing lists