From: Jonathan Cameron <jic23@kernel.org>
To: Lorenzo Bianconi <lorenzo.bianconi83@gmail.com>
Cc: Bobby Jones <rjones@gateworks.com>,
Lorenzo Bianconi <lorenzo@kernel.org>,
Martin Kepplinger <martin.kepplinger@puri.sm>,
linux-iio@vger.kernel.org
Subject: Re: LSM9DS1 testing with st_lsm6dsx driver
Date: Sat, 5 Oct 2019 12:22:07 +0100 [thread overview]
Message-ID: <20191005122207.5979efd2@archlinux> (raw)
In-Reply-To: <CAA2SeN+QR5sd_WrjOxiPx0JY7fK3W0wenP_RzxfrP9isk9wCXQ@mail.gmail.com>
On Tue, 24 Sep 2019 20:55:09 +0200
Lorenzo Bianconi <lorenzo.bianconi83@gmail.com> wrote:
> >
> > > > > > > > > > > LSM9DS1 does not support hw FIFO for the moment. Are you trying to enable
> > > > > > > > > > > buffered reading?
> > > > > > > > > >
> > > > > > > > > > I am not manually interacting with the device at all when this
> > > > > > > > > > exception occurs. This happens during the driver probe.
> > > > > > > > > >
> > > > > > > > > > > Could you please try if the following patch helps? (just compiled)
> > > > > > > > > >
> > > > > > > > > > I no longer receive the exception with this patch and it makes sense, thanks.
> > > > > > > > >
> > > > > > > > > Hi Bobby,
> > > > > > > > >
> > > > > > > > > thx a lot for testing. Could you please try to drop the previous patch and
> > > > > > > > > apply the following one? Does it fix the issue as well?
> > > > > > > >
> > > > > > > > No problem, happy to help. I just tested, and unfortunately the issue
> > > > > > > > is still present with this patch.
> > > > > > >
> > > > > > > re-looking at the code this patch will actually does not anything since in the
> > > > > > > current implementation st_lsm6dsx_handler_irq will return IRQ_NONE and the
> > > > > > > thread handler will not be run. Anyway since st_lsm6dsx_handler_irq will be
> > > > > > > removed soon I will respin this patch ontop of Sean series.
> > > > > > >
> > > > > > > > I gave the datasheet and the hardware reference manual of the
> > > > > > > > connected CPU a closer look and suspected a problem with my pin
> > > > > > > > config. However even after various combinations of pull ups/downs and
> > > > > > > > IRQ_TYPE_LEVEL_LOW/IRQ_TYPE_LEVEL_HIGH the same exception occurs, so I
> > > > > > > > don't think that there's an issue with pin config stopping the
> > > > > > > > interrupt line from being deasserted.
> > > > > > >
> > > > > > > are you able to monitor the line activity through an oscilloscope?
> > > > > > > The issue is the irq line is never asserted and the kernel complains about
> > > > > > > lot of interrupts not managed
> > > > > > >
> > > > > > > Looking at the datasheet, LSM9DS1 does not support IRQ_TYPE_LEVEL_LOW.
> > > > > > > ST_LSM6DSX_REG_HLACTIVE_ADDR and ST_LSM6DSX_REG_PP_OD_ADDR registers are
> > > > > > > actually not defined for LSM9DS1 and I will move them in hw_settings map.
> > > > > >
> > > > > > The datasheet I have for the LSM9DS1 shows a HLACTIVE bit in the
> > > > > > CTRL_REG8 (0x22) register, I thought that indicated support for both
> > > > > > high and low IRQ types.
> > > > >
> > > > > Yes, right. I will update the series. Looking at the datasheet even
> > > > > the BDU register definition seems wrong. I will fix it as well.
> > > > >
> > > > > > Either way, after applying your recent (Sep 22nd) patches to this
> > > > > > driver and changing my device tree node I have no kernel exception and
> > > > > > the iio device enumerates as normal. For anyone curious my device tree
> > > > > > node is now:
> > > > > >
> > > > >
> > > > > What happen if you set the irq line active high? Does the issue occur?
> > > >
> > > > No issue with IRQ_TYPE_LEVEL_HIGH either. Although all I'm doing at
> > > > this point to test is checking for the exception in the kernel prints
> > > > and reading the /sys/bus/iio/devices/iio:device*/in*raw of the device.
> > >
> > > Ok, thx a lot for testing. Could you please try to understand which
> > > patch of the series fix the issue?
> > > In particular, could you please try to apply just '[PATCH 3/3] iio:
> > > imu: st_lsm6dsx: add sanity check for read_fifo pointer' and double
> > > check if the issue occurs?
> >
> > No problem. It looks like that was the one, only having that patch
> > applied works fine.
>
> Ok, cool. Thx a lot for testing.
>
> @Jonathan: how do you prefer to proceed? Push just this patch to
> 'togreg' branch and rebase the other fixes ontop of Sean's series or
> merge all the fixes now? (they will conflict with Sean's series)
Assuming I don't find any issues in Sean's latest, I'll apply that
first. Then we put this on top but also look at a backport fix
to get this into stable after the next merge window.
It's always tricky to deal with these cases when a driver is
in a state of fairly rapid change.
I'll take a proper look at patch 3 if that is the only one
'necessary'. It looks small so we might just push the fix
along side Sean's patch and let everyone concerned know that
a merge conflict is expected by trivial.
Thanks,
Jonathan
>
> Regards,
> Lorenzo
>
> >
> > Thanks,
> > Bobby Jones
> >
> > >
> > > Regards,
> > > Lorenzo
> > >
> > > >
> > > > Thanks,
> > > > Bobby Jones
> > > >
> > > > >
> > > > > Regards,
> > > > > Lorenzo
> > > > >
> > > > > > lsm9ds1_ag@6a {
> > > > > > compatible = "st,lsm9ds1-imu";
> > > > > > reg = <0x6a>;
> > > > > > st,drdy-int-pin = <1>;
> > > > > > pinctrl-names = "default";
> > > > > > pinctrl-0 = <&pinctrl_acc_gyro>;
> > > > > > interrupt-parent = <&gpio7>;
> > > > > > interrupts = <13 IRQ_TYPE_EDGE_RISING>;
> > > > > > };
> > > > > >
> > > > > > >
> > > > > > > @Jonathan: do you prefer this patch to be ontop of Sean's series?
> > > > > > >
> > > > > > > Regards,
> > > > > > > Lorenzo
> > > > > > >
> > > > > > > >
> > > > > > > > >
> > > > > > > > > iio: imu: st_lsm6dsx: check read_fifo pointer in st_lsm6dsx_handler_thread
> > > > > > > > >
> > > > > > > > > Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
> > > > > > > > > ---
> > > > > > > > > drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c | 3 +++
> > > > > > > > > 1 file changed, 3 insertions(+)
> > > > > > > > >
> > > > > > > > > diff --git a/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c b/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c
> > > > > > > > > index b0f3da1976e4..f4fd4842bd79 100644
> > > > > > > > > --- a/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c
> > > > > > > > > +++ b/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c
> > > > > > > > > @@ -666,6 +666,9 @@ static irqreturn_t st_lsm6dsx_handler_thread(int irq, void *private)
> > > > > > > > > struct st_lsm6dsx_hw *hw = private;
> > > > > > > > > int count;
> > > > > > > > >
> > > > > > > > > + if (!hw->settings->fifo_ops.read_fifo)
> > > > > > > > > + return IRQ_NONE;
> > > > > > > > > +
> > > > > > > > > mutex_lock(&hw->fifo_lock);
> > > > > > > > > count = hw->settings->fifo_ops.read_fifo(hw);
> > > > > > > > > mutex_unlock(&hw->fifo_lock);
> > > > > > > > > --
> > > > > > > > > 2.21.0
> > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > For context I'm working with a board that has every data ready and
> > > > > > > > > > interrupt signal connected to the LSM9DS1. Could you clarify what the
> > > > > > > > > > proper usage of the "st,drdy-int-pin" would be in this case and
> > > > > > > > > > whether or not I need more than one interrupt called out in my device
> > > > > > > > > > tree node?
> > > > > > > > > > I'm not really understanding how they're currently being utilized for
> > > > > > > > > > this device in the driver.
> > > > > > > > >
> > > > > > > > > For the moment irq line in lsm9ds1 (acc/gyro) is not used at all,
> > > > > > > > > so you can omit the "st,drdy-int-pin" property
> > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > Also, I know support for this device was added recently and the combo
> > > > > > > > > > device hardware FIFO is complex, but is support for this something
> > > > > > > > > > that's currently being worked on?
> > > > > > > > >
> > > > > > > > > It is actually in my ToDo list but I have no this device at the moment, so
> > > > > > > > > patches are welcome :)
> > > > > > > > >
> > > > > > > > > Regards,
> > > > > > > > > Lorenzo
> > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > Thanks,
> > > > > > > > > > Bobby Jones
> > > > > > > > > >
> > > > > > > > > > > Regards,
> > > > > > > > > > > Lorenzo
> > > > > > > > > > >
> > > > > > > > > > > iio: imu: st_lsm6dsx: do not configure the fifo if not supported
> > > > > > > > > > >
> > > > > > > > > > > Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
> > > > > > > > > > > ---
> > > > > > > > > > > drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_core.c | 2 +-
> > > > > > > > > > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > > > > > > > > > >
> > > > > > > > > > > diff --git a/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_core.c b/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_core.c
> > > > > > > > > > > index b65a6ca775e0..90a0e5ce44e5 100644
> > > > > > > > > > > --- a/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_core.c
> > > > > > > > > > > +++ b/drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_core.c
> > > > > > > > > > > @@ -1572,7 +1572,7 @@ int st_lsm6dsx_probe(struct device *dev, int irq, int hw_id,
> > > > > > > > > > > return err;
> > > > > > > > > > > }
> > > > > > > > > > >
> > > > > > > > > > > - if (hw->irq > 0) {
> > > > > > > > > > > + if (hw->irq > 0 && hw->settings->fifo_ops.update_fifo) {
> > > > > > > > > > > err = st_lsm6dsx_fifo_setup(hw);
> > > > > > > > > > > if (err < 0)
> > > > > > > > > > > return err;
> > > > > > > > > > > --
> > > > > > > > > > > 2.21.0
> > > > > > > > > > >
> > > > > > > > > > >
>
>
>
next prev parent reply other threads:[~2019-10-05 11:22 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-09-20 0:53 LSM9DS1 testing with st_lsm6dsx driver Bobby Jones
2019-09-20 6:42 ` Lorenzo Bianconi
2019-09-20 18:39 ` Bobby Jones
2019-09-20 21:55 ` Lorenzo Bianconi
2019-09-20 23:21 ` Bobby Jones
2019-09-21 8:06 ` Lorenzo Bianconi
2019-09-23 23:23 ` Bobby Jones
2019-09-24 6:18 ` Lorenzo Bianconi
2019-09-24 17:30 ` Bobby Jones
2019-09-24 17:40 ` Lorenzo Bianconi
2019-09-24 18:21 ` Bobby Jones
2019-09-24 18:55 ` Lorenzo Bianconi
2019-10-05 11:22 ` Jonathan Cameron [this message]
2019-10-05 12:31 ` 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=20191005122207.5979efd2@archlinux \
--to=jic23@kernel.org \
--cc=linux-iio@vger.kernel.org \
--cc=lorenzo.bianconi83@gmail.com \
--cc=lorenzo@kernel.org \
--cc=martin.kepplinger@puri.sm \
--cc=rjones@gateworks.com \
/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).