| Message ID | 20240731-ad7625_r1-v1-0-a1efef5a2ab9@baylibre.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | iio: adc: add new ad7625 driver | expand |
On 2024-07-31 9:48 a.m., Trevor Gamblin wrote: > This series adds a new driver for the Analog Devices Inc. AD7625, > AD7626, AD7960, and AD7961. These chips are part of a family of > LVDS-based SAR ADCs. The initial driver implementation does not support > the devices' self-clocked mode, although that can be added later. > > One aspect that is still uncertain is whether there should be a > devicetree property indicating if the DCO+/- pins are connected, so > specific feedback on that is appreciated. > > The devices make use of two offset PWM signals, one to trigger > conversions and the other as a burst signal for transferring data to the > host. These rely on the new PWM waveform functionality being > reviewed in [1]. > > This work is being done by BayLibre and on behalf of Analog Devices > Inc., hence the maintainers are @analog.com. > > Special thanks to David Lechner for his guidance and reviews. I forgot to actually include: [1] https://lore.kernel.org/linux-pwm/cover.1722261050.git.u.kleine-koenig@baylibre.com > > Signed-off-by: Trevor Gamblin <tgamblin@baylibre.com> > --- > Trevor Gamblin (3): > dt-bindings: iio: adc: add AD762x/AD796x ADCs > iio: adc: ad7625: add driver > docs: iio: new docs for ad7625 driver > > .../devicetree/bindings/iio/adc/adi,ad7625.yaml | 176 ++++++ > Documentation/iio/ad7625.rst | 91 +++ > MAINTAINERS | 11 + > drivers/iio/adc/Kconfig | 15 + > drivers/iio/adc/Makefile | 1 + > drivers/iio/adc/ad7625.c | 626 +++++++++++++++++++++ > 6 files changed, 920 insertions(+) > --- > base-commit: ac6a258892793f0a255fe7084ec2b612131c67fc > change-id: 20240730-ad7625_r1-60d17ea28958 > > Best regards,
On Wed, 31 Jul 2024 09:48:02 -0400 Trevor Gamblin <tgamblin@baylibre.com> wrote: > This series adds a new driver for the Analog Devices Inc. AD7625, > AD7626, AD7960, and AD7961. These chips are part of a family of > LVDS-based SAR ADCs. The initial driver implementation does not support > the devices' self-clocked mode, although that can be added later. > > One aspect that is still uncertain is whether there should be a > devicetree property indicating if the DCO+/- pins are connected, so > specific feedback on that is appreciated. Would be good to give more detail. What is DCO? Seems to be a delayed clock skewed so it aligns with the data being out in response to clk. Host drives clk, but samples on dco. Given the device needs to do slightly different things depending on whether that is what the host is using, I think it definitely does need to be in DT. Maybe you need to represent it as the ADC also having a PWM output that the LVDS DT node binds to if present. That binding then indicates to the ADC driver that it needs to operating in the mode that doesn't send the synchronisation 101 pattern. If you are always representing the ADC and the lvds side of things as a single node, then need a flag in here somewhere so we can tell if they are in use or not. Given this exists as a potential difference between two separate parts pf a system I'd definitely think about whether we can give them separate representations with clear 'connectivity' between them One of those cases were a bit of ascii art would probably be good to put the problem clearly for the DT reviewers. Jonathan > > The devices make use of two offset PWM signals, one to trigger > conversions and the other as a burst signal for transferring data to the > host. These rely on the new PWM waveform functionality being > reviewed in [1]. > > This work is being done by BayLibre and on behalf of Analog Devices > Inc., hence the maintainers are @analog.com. > > Special thanks to David Lechner for his guidance and reviews. > > Signed-off-by: Trevor Gamblin <tgamblin@baylibre.com> > --- > Trevor Gamblin (3): > dt-bindings: iio: adc: add AD762x/AD796x ADCs > iio: adc: ad7625: add driver > docs: iio: new docs for ad7625 driver > > .../devicetree/bindings/iio/adc/adi,ad7625.yaml | 176 ++++++ > Documentation/iio/ad7625.rst | 91 +++ > MAINTAINERS | 11 + > drivers/iio/adc/Kconfig | 15 + > drivers/iio/adc/Makefile | 1 + > drivers/iio/adc/ad7625.c | 626 +++++++++++++++++++++ > 6 files changed, 920 insertions(+) > --- > base-commit: ac6a258892793f0a255fe7084ec2b612131c67fc > change-id: 20240730-ad7625_r1-60d17ea28958 > > Best regards,
This series adds a new driver for the Analog Devices Inc. AD7625, AD7626, AD7960, and AD7961. These chips are part of a family of LVDS-based SAR ADCs. The initial driver implementation does not support the devices' self-clocked mode, although that can be added later. One aspect that is still uncertain is whether there should be a devicetree property indicating if the DCO+/- pins are connected, so specific feedback on that is appreciated. The devices make use of two offset PWM signals, one to trigger conversions and the other as a burst signal for transferring data to the host. These rely on the new PWM waveform functionality being reviewed in [1]. This work is being done by BayLibre and on behalf of Analog Devices Inc., hence the maintainers are @analog.com. Special thanks to David Lechner for his guidance and reviews. Signed-off-by: Trevor Gamblin <tgamblin@baylibre.com> --- Trevor Gamblin (3): dt-bindings: iio: adc: add AD762x/AD796x ADCs iio: adc: ad7625: add driver docs: iio: new docs for ad7625 driver .../devicetree/bindings/iio/adc/adi,ad7625.yaml | 176 ++++++ Documentation/iio/ad7625.rst | 91 +++ MAINTAINERS | 11 + drivers/iio/adc/Kconfig | 15 + drivers/iio/adc/Makefile | 1 + drivers/iio/adc/ad7625.c | 626 +++++++++++++++++++++ 6 files changed, 920 insertions(+) --- base-commit: ac6a258892793f0a255fe7084ec2b612131c67fc change-id: 20240730-ad7625_r1-60d17ea28958 Best regards,