Re: [PATCH v2 5/5] iio: dac: ad5686: Add PWM as a trigger source
On Tue, 23 Feb 2021 17:37:40 +0100 Lars-Peter Clausen wrote: > On 2/18/21 3:05 PM, Jonathan Cameron wrote: > > On Wed, 17 Feb 2021 10:34:38 +0200 > > Alexandru Ardelean wrote: > > > >> From: Mircea Caprioru > >> > >> A PWM signal will be used as a trigger source to have a deterministic > >> sampling frequency since this family of DAC has no hardware interrupt > >> source. > >> > >> This feature is made optional however, as there are some board setups where > >> this isn't used. > >> > > So this is taking a very generic setup, but then implementing it > > as a bit of a hack within the driver. > > > > It's effectively a PWM connected up to an instance > > of iio/triggers/iio-trig-interrupt.c > > > > Now, I've not looked at that trigger driver for a while, so you may well > > need to figure out how to add a binding to instantiate it. > > (looks like no one has used it since board file days, or via instantiation > > from another driver). > > > > It's a slightly odd corner case as what it reflects is that we have > > an interrupt available that is intended to drive some sort of data > > capture or output (it's a trigger signal) - but exactly what is done > > is a runtime configurable. In this particular case that interrupt > > is hooked up to a PWM and we also want to represent that. > > > > The fact it's being driven via a PWM is interesting but we should be > > able to extend that trigger driver to optionally accept a pwm provider > > and if it has one provide frequency control. > > > > Binding might look something like the following.. > > > > interrupt-trigger { > > interrupts = <>; > > pwms = <&pwm 0 4000 PWM_POLARITY_INVERTED>; > > }; > > > > @Rob, what do you think of this odd beast? > > > > So all in all, this generic facility needs a generic implementation, not > > one buried in a driver. > > > > Another open question here is whether you really can't just use an hrtimer > > to get similar precision? Way back at the dawn of time in IIO we had > > code to use the RTC periodic ticks as a trigger with the theory that they > > would give very precise and even timing. In the end it turned out that > > hrtimers worked just as well (and RTCs drivers emulated the periodic > > ticks via hrtimers, dropping their use of the hardware periodic timers). > > > The way this DAC works is that it has a "latch" pin and some shadow > registers. The way this is supposed to be used is that you update the > shadow registers and then when the there is a rising edge on the latch > pin all the shadow register values are transferred to DAC output registers. > > This means if you hook up a periodic signal like a PWM or clock to the > latch pin you can generate very precise waveforms that have much lower > jitter than when using a hrtimer since there is no variable interrupt > latency for the update step itself. This is useful when generating > periodic signals. > > But you could for example also use a GPIO to update multiple discrete > DACs at the same time. > > This is not specific to this particular chip. There are quite a few ADI > (and probably from other vendors) precision DACs that have this > functionality. I agree that this should be a some sort of generic > trigger helper module. > > Now for the implementation since there is a direct connection between > the PWM and the DAC I think it makes sense to describe this connection > in the DT. After all if there is no connection this will not work. Thanks for the detailed description. That makes a lot more sense. This is some sort of hybrid of the hardware internal triggers we have for some SoC ADCs and wiring up a gpio pin to trigger the latch signal. PWM is one valid way of wiring it up (possibly most sensible one), but not necessarily the only one. I guess the one behind element is also a bit non intuitive (data is put in place on previous interrupt / edge but latched on the next one) Hmm. If we makes sure the binding is cleanly defined, we could do a driver specific implementation for now, with the option to figure something else out later. Exactly how to do this needs some thought... + lifting this description of hot it works into the patch description would help :) Jonathan > > As for the interrupt, most PWM controllers do have the ability to > generate an IRQ by themselves once per period. There should be not need > for a hardware loopback. Unfortunately the PWM framework does not have a > mechanism yet to expose those IRQs and register a callback. > > A similar feature btw exists for many of the ADCs and we did have this > special Blackfin PWM trigger[1] back in the day to support this. The > bfin PWM trigger driver essentially implements what I'm describing > above, but without using the PWM framework. > > - Lars > > [1] > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/staging/iio/trigger/iio-trig-bfin-timer.c?h=v3.15 >
Re: [PATCH v2 5/5] iio: dac: ad5686: Add PWM as a trigger source
On 2/18/21 3:05 PM, Jonathan Cameron wrote: On Wed, 17 Feb 2021 10:34:38 +0200 Alexandru Ardelean wrote: From: Mircea Caprioru A PWM signal will be used as a trigger source to have a deterministic sampling frequency since this family of DAC has no hardware interrupt source. This feature is made optional however, as there are some board setups where this isn't used. So this is taking a very generic setup, but then implementing it as a bit of a hack within the driver. It's effectively a PWM connected up to an instance of iio/triggers/iio-trig-interrupt.c Now, I've not looked at that trigger driver for a while, so you may well need to figure out how to add a binding to instantiate it. (looks like no one has used it since board file days, or via instantiation from another driver). It's a slightly odd corner case as what it reflects is that we have an interrupt available that is intended to drive some sort of data capture or output (it's a trigger signal) - but exactly what is done is a runtime configurable. In this particular case that interrupt is hooked up to a PWM and we also want to represent that. The fact it's being driven via a PWM is interesting but we should be able to extend that trigger driver to optionally accept a pwm provider and if it has one provide frequency control. Binding might look something like the following.. interrupt-trigger { interrupts = <>; pwms = <&pwm 0 4000 PWM_POLARITY_INVERTED>; }; @Rob, what do you think of this odd beast? So all in all, this generic facility needs a generic implementation, not one buried in a driver. Another open question here is whether you really can't just use an hrtimer to get similar precision? Way back at the dawn of time in IIO we had code to use the RTC periodic ticks as a trigger with the theory that they would give very precise and even timing. In the end it turned out that hrtimers worked just as well (and RTCs drivers emulated the periodic ticks via hrtimers, dropping their use of the hardware periodic timers). The way this DAC works is that it has a "latch" pin and some shadow registers. The way this is supposed to be used is that you update the shadow registers and then when the there is a rising edge on the latch pin all the shadow register values are transferred to DAC output registers. This means if you hook up a periodic signal like a PWM or clock to the latch pin you can generate very precise waveforms that have much lower jitter than when using a hrtimer since there is no variable interrupt latency for the update step itself. This is useful when generating periodic signals. But you could for example also use a GPIO to update multiple discrete DACs at the same time. This is not specific to this particular chip. There are quite a few ADI (and probably from other vendors) precision DACs that have this functionality. I agree that this should be a some sort of generic trigger helper module. Now for the implementation since there is a direct connection between the PWM and the DAC I think it makes sense to describe this connection in the DT. After all if there is no connection this will not work. As for the interrupt, most PWM controllers do have the ability to generate an IRQ by themselves once per period. There should be not need for a hardware loopback. Unfortunately the PWM framework does not have a mechanism yet to expose those IRQs and register a callback. A similar feature btw exists for many of the ADCs and we did have this special Blackfin PWM trigger[1] back in the day to support this. The bfin PWM trigger driver essentially implements what I'm describing above, but without using the PWM framework. - Lars [1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/staging/iio/trigger/iio-trig-bfin-timer.c?h=v3.15
Re: [PATCH v2 5/5] iio: dac: ad5686: Add PWM as a trigger source
On Thu, Feb 18, 2021 at 5:31 PM Rob Herring wrote: > > On Thu, Feb 18, 2021 at 8:05 AM Jonathan Cameron wrote: > > > > On Wed, 17 Feb 2021 10:34:38 +0200 > > Alexandru Ardelean wrote: > > > > > From: Mircea Caprioru > > > > > > A PWM signal will be used as a trigger source to have a deterministic > > > sampling frequency since this family of DAC has no hardware interrupt > > > source. > > > > > > This feature is made optional however, as there are some board setups > > > where > > > this isn't used. > > > > > > > So this is taking a very generic setup, but then implementing it > > as a bit of a hack within the driver. > > > > It's effectively a PWM connected up to an instance > > of iio/triggers/iio-trig-interrupt.c > > > > Now, I've not looked at that trigger driver for a while, so you may well > > need to figure out how to add a binding to instantiate it. > > (looks like no one has used it since board file days, or via instantiation > > from another driver). > > > > It's a slightly odd corner case as what it reflects is that we have > > an interrupt available that is intended to drive some sort of data > > capture or output (it's a trigger signal) - but exactly what is done > > is a runtime configurable. In this particular case that interrupt > > is hooked up to a PWM and we also want to represent that. > > > > The fact it's being driven via a PWM is interesting but we should be > > able to extend that trigger driver to optionally accept a pwm provider > > and if it has one provide frequency control. > > So, the main intent here was to provide a user for this new output kfifo. I don't think I have time to re-spin this into a proper solution. Someone else may come about and do it. I'll drop this from the series [for now]. > > Binding might look something like the following.. > > > > interrupt-trigger { > >interrupts = <>; > >pwms = <&pwm 0 4000 PWM_POLARITY_INVERTED>; > > }; > > > > @Rob, what do you think of this odd beast? > > So a PWM routed back to a GPIO interrupt? It needs a compatible, but > otherwise I wouldn't object to the binding if that's what the h/w > looks like. But from an OS perspective, I don't think you need it. > > > So all in all, this generic facility needs a generic implementation, not > > one buried in a driver. > > > > Another open question here is whether you really can't just use an hrtimer > > to get similar precision? Way back at the dawn of time in IIO we had > > code to use the RTC periodic ticks as a trigger with the theory that they > > would give very precise and even timing. In the end it turned out that > > hrtimers worked just as well (and RTCs drivers emulated the periodic > > ticks via hrtimers, dropping their use of the hardware periodic timers). > > +100 > > A hrtimer is likely going to be more precise. IIRC, timers are > serviced first. Either way, you're going to have some amount of > interrupt service latency, so any precision you think you are gaining > by 'doing it in h/w' isn't really there. > > Rob
Re: [PATCH v2 5/5] iio: dac: ad5686: Add PWM as a trigger source
On Thu, Feb 18, 2021 at 8:05 AM Jonathan Cameron wrote: > > On Wed, 17 Feb 2021 10:34:38 +0200 > Alexandru Ardelean wrote: > > > From: Mircea Caprioru > > > > A PWM signal will be used as a trigger source to have a deterministic > > sampling frequency since this family of DAC has no hardware interrupt > > source. > > > > This feature is made optional however, as there are some board setups where > > this isn't used. > > > > So this is taking a very generic setup, but then implementing it > as a bit of a hack within the driver. > > It's effectively a PWM connected up to an instance > of iio/triggers/iio-trig-interrupt.c > > Now, I've not looked at that trigger driver for a while, so you may well > need to figure out how to add a binding to instantiate it. > (looks like no one has used it since board file days, or via instantiation > from another driver). > > It's a slightly odd corner case as what it reflects is that we have > an interrupt available that is intended to drive some sort of data > capture or output (it's a trigger signal) - but exactly what is done > is a runtime configurable. In this particular case that interrupt > is hooked up to a PWM and we also want to represent that. > > The fact it's being driven via a PWM is interesting but we should be > able to extend that trigger driver to optionally accept a pwm provider > and if it has one provide frequency control. > > Binding might look something like the following.. > > interrupt-trigger { >interrupts = <>; >pwms = <&pwm 0 4000 PWM_POLARITY_INVERTED>; > }; > > @Rob, what do you think of this odd beast? So a PWM routed back to a GPIO interrupt? It needs a compatible, but otherwise I wouldn't object to the binding if that's what the h/w looks like. But from an OS perspective, I don't think you need it. > So all in all, this generic facility needs a generic implementation, not > one buried in a driver. > > Another open question here is whether you really can't just use an hrtimer > to get similar precision? Way back at the dawn of time in IIO we had > code to use the RTC periodic ticks as a trigger with the theory that they > would give very precise and even timing. In the end it turned out that > hrtimers worked just as well (and RTCs drivers emulated the periodic > ticks via hrtimers, dropping their use of the hardware periodic timers). +100 A hrtimer is likely going to be more precise. IIRC, timers are serviced first. Either way, you're going to have some amount of interrupt service latency, so any precision you think you are gaining by 'doing it in h/w' isn't really there. Rob
Re: [PATCH v2 5/5] iio: dac: ad5686: Add PWM as a trigger source
On Wed, 17 Feb 2021 10:34:38 +0200 Alexandru Ardelean wrote: > From: Mircea Caprioru > > A PWM signal will be used as a trigger source to have a deterministic > sampling frequency since this family of DAC has no hardware interrupt > source. > > This feature is made optional however, as there are some board setups where > this isn't used. > So this is taking a very generic setup, but then implementing it as a bit of a hack within the driver. It's effectively a PWM connected up to an instance of iio/triggers/iio-trig-interrupt.c Now, I've not looked at that trigger driver for a while, so you may well need to figure out how to add a binding to instantiate it. (looks like no one has used it since board file days, or via instantiation from another driver). It's a slightly odd corner case as what it reflects is that we have an interrupt available that is intended to drive some sort of data capture or output (it's a trigger signal) - but exactly what is done is a runtime configurable. In this particular case that interrupt is hooked up to a PWM and we also want to represent that. The fact it's being driven via a PWM is interesting but we should be able to extend that trigger driver to optionally accept a pwm provider and if it has one provide frequency control. Binding might look something like the following.. interrupt-trigger { interrupts = <>; pwms = <&pwm 0 4000 PWM_POLARITY_INVERTED>; }; @Rob, what do you think of this odd beast? So all in all, this generic facility needs a generic implementation, not one buried in a driver. Another open question here is whether you really can't just use an hrtimer to get similar precision? Way back at the dawn of time in IIO we had code to use the RTC periodic ticks as a trigger with the theory that they would give very precise and even timing. In the end it turned out that hrtimers worked just as well (and RTCs drivers emulated the periodic ticks via hrtimers, dropping their use of the hardware periodic timers). Jonathan > Signed-off-by: Mircea Caprioru > Signed-off-by: Mihail Chindris > Signed-off-by: Alexandru Ardelean > --- > drivers/iio/dac/ad5686-spi.c | 2 +- > drivers/iio/dac/ad5686.c | 146 ++- > drivers/iio/dac/ad5686.h | 7 +- > drivers/iio/dac/ad5696-i2c.c | 2 +- > 4 files changed, 152 insertions(+), 5 deletions(-) > > diff --git a/drivers/iio/dac/ad5686-spi.c b/drivers/iio/dac/ad5686-spi.c > index 0188ded5137c..07fadcf8e1e3 100644 > --- a/drivers/iio/dac/ad5686-spi.c > +++ b/drivers/iio/dac/ad5686-spi.c > @@ -92,7 +92,7 @@ static int ad5686_spi_probe(struct spi_device *spi) > const struct spi_device_id *id = spi_get_device_id(spi); > > return ad5686_probe(&spi->dev, id->driver_data, id->name, > - ad5686_spi_write, ad5686_spi_read); > + ad5686_spi_write, ad5686_spi_read, spi->irq); > } > > static int ad5686_spi_remove(struct spi_device *spi) > diff --git a/drivers/iio/dac/ad5686.c b/drivers/iio/dac/ad5686.c > index 7d6792ac1020..9e48559ec566 100644 > --- a/drivers/iio/dac/ad5686.c > +++ b/drivers/iio/dac/ad5686.c > @@ -16,6 +16,10 @@ > > #include > #include > +#include > +#include > +#include > +#include > > #include "ad5686.h" > > @@ -123,6 +127,7 @@ static int ad5686_read_raw(struct iio_dev *indio_dev, > long m) > { > struct ad5686_state *st = iio_priv(indio_dev); > + struct pwm_state state; > int ret; > > switch (m) { > @@ -139,6 +144,10 @@ static int ad5686_read_raw(struct iio_dev *indio_dev, > *val = st->vref_mv; > *val2 = chan->scan_type.realbits; > return IIO_VAL_FRACTIONAL_LOG2; > + case IIO_CHAN_INFO_SAMP_FREQ: > + pwm_get_state(st->pwm, &state); > + *val = DIV_ROUND_CLOSEST_ULL(10ULL, state.period); > + return IIO_VAL_INT; > } > return -EINVAL; > } > @@ -150,6 +159,7 @@ static int ad5686_write_raw(struct iio_dev *indio_dev, > long mask) > { > struct ad5686_state *st = iio_priv(indio_dev); > + struct pwm_state state; > int ret; > > switch (mask) { > @@ -164,6 +174,14 @@ static int ad5686_write_raw(struct iio_dev *indio_dev, > val << chan->scan_type.shift); > mutex_unlock(&st->lock); > break; > + case IIO_CHAN_INFO_SAMP_FREQ: > + pwm_get_state(st->pwm, &state); > + > + state.period = DIV_ROUND_CLOSEST_ULL(10ULL, val); > + pwm_set_relative_duty_cycle(&state, 50, 100); > + > + ret = pwm_apply_state(st->pwm, &state); > + break; > default: > ret = -EINVAL; > } > @@ -171,7 +189,37 @@ static int ad5686_write_raw(struct iio_dev *indio_dev, > return ret; > } > > +static int ad5686_trig_set_state(struct iio_trigger *trig, > +