Re: [PATCH v3 1/2] hwmon: Add driver for Texas Instruments TMP512/513 sensor chips
On Thu, Oct 03, 2019 at 06:34:17PM +, Tremblay, Eric wrote: > hwmon: Add driver for Texas Instruments TMP512/513 sensor chips. > > TI's TMP512/513 are I2C/SMBus system monitor chips. These chips > monitor the supply voltage, supply current, power consumption > and provide one local and up to three (TMP513) remote temperature sensors. > > It has been tested using a TI TMP513 development kit (TMP513EVM) > > Signed-off-by: Eric Tremblay > --- > Documentation/hwmon/tmp513.rst | 102 +++ > MAINTAINERS | 7 + > drivers/hwmon/Kconfig| 10 + > drivers/hwmon/Makefile | 1 + > drivers/hwmon/tmp513.c | 901 +++ > include/linux/platform_data/tmp513.h | 28 + > 6 files changed, 1049 insertions(+) > create mode 100644 Documentation/hwmon/tmp513.rst > create mode 100644 drivers/hwmon/tmp513.c > create mode 100644 include/linux/platform_data/tmp513.h > > diff --git a/Documentation/hwmon/tmp513.rst b/Documentation/hwmon/tmp513.rst > new file mode 100644 > index ..b3837748ca7e > --- /dev/null > +++ b/Documentation/hwmon/tmp513.rst > @@ -0,0 +1,102 @@ > +.. SPDX-License-Identifier: GPL-2.0 > +Kernel driver tmp513 > + > + > +Supported chips: > + > + * Texas Instruments TMP512 > + > +Prefix: 'tmp512' > + > +Datasheet: http://www.ti.com/lit/ds/symlink/tmp512.pdf > + > + * Texas Instruments TMP513 > + > +Prefix: 'tmp513' > + > +Datasheet: http://www.ti.com/lit/ds/symlink/tmp513.pdf > + > +Authors: > + > + Eric Tremblay > + > +Description > +--- > + > +This driver implements support for Texas Instruments TMP512, and TMP513 > +temperature and power supply sensor chips. These chips implement one > +local and up to three (TMP513) temperature sensors. The chips also implement > +power supply monitoring such as shunt voltage, shunt current, bus voltage > +and power consumption. > + > +The temperatures are measured in degrees Celsius with a range of > +-40 to + 125 degrees with a resolution of 0.0625 degree C. > + > +For hysteresis value, only the first channel is writable. Writing to it > +will affect all other values since each channels are sharing the same > +hysteresis value. The hysteresis is in degrees Celsius with a range of > +0 to 127.5 degrees with a resolution of 0.5 degree. > + > +The driver exports the temperature values via the following sysfs files: > + > +**temp[1-4]_input** > + > +**temp[1-4]_crit** > + > +**temp[1-4]_crit_alarm** > + > +**temp[1-4]_crit_hyst** > + > +The shunt voltage is measured in micro volts. There are four configurable > +range: 320mV, 160mV, 80mV and 40mV all of them with a resolution of 10 uV. > +The range depends on the pga gain specified in the device tree > +with "pga-gain" (default to 8). The range will be equal to pga gain > +multiply by 40mV. > + > +The driver exports the shunt voltage values via the following sysFs files: > + > +**in0_input** > + > +**in0_lcrit** > + > +**in0_lcrit_alarm** > + > +**in0_crit** > + > +**in0_crit_alarm** > + > +The bus voltage is measured in milli volts. There are two configuable > +range: 32V and 16V both with a resolution of 4mV. It can be configured in the > +device tree with "bus-voltage-range" (default to 32V); > + > +The driver exports the bus voltage values via the following sysFs files: > + > +**in0_input** > + > +**in0_lcrit** > + > +**in0_lcrit_alarm** > + > +**in0_crit** > + > +**in0_crit_alarm** > + > +The power and the currents range and resolution depends on the calibration > +register value. Those values are calculate by the hardware using those > +formula: > + > +Current = (ShuntVoltage * CalibrationRegister) / 4096 > +Power = (Current * BusVoltage) / 5000 > + > +The driver exports the current and power values via the following sysFs > files: > + > +**curr0_input** > + > +**power1_input** > + > +**power1_crit** > + > +**power1_crit_alarm** > + > +The calibration process follow the procedure of the datasheet (without > overflow) > +and depend on the shunt resistor value and the pga_gain value. > diff --git a/MAINTAINERS b/MAINTAINERS > index 296de2b51c83..1d13e5be6661 100644 > --- a/MAINTAINERS > +++ b/MAINTAINERS > @@ -16369,6 +16369,13 @@ S: Maintained > F: Documentation/hwmon/tmp401.rst > F: drivers/hwmon/tmp401.c > > +TMP513 HARDWARE MONITOR DRIVER > +M: Eric Tremblay > +L: linux-hwmon@vger.kernel.org > +S: Maintained > +F: Documentation/hwmon/tmp513.rst > +F: drivers/hwmon/tmp513.c > + > TMPFS (SHMEM FILESYSTEM) > M: Hugh Dickins > L: linux...@kvack.org > diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig > index 13a6b4afb4b3..926e3c98dbdf 100644 > --- a/drivers/hwmon/Kconfig > +++ b/drivers/hwmon/Kconfig > @@ -1709,6 +1709,16 @@ config SENSORS_TMP421 > This driver can also be built as a module. If so, the module > will be called tmp421. > > +config SENSORS_TMP513 > + tristate "Texas Instruments
Re: [PATCH 09/10] hwmon: (lm70) Avoid undefined reference to match table
On Fri, Oct 04, 2019 at 02:43:33PM -0700, Stephen Boyd wrote: > We're going to remove of_match_ptr() from the definition of > of_match_device() when CONFIG_OF=n. This way we can always be certain > that of_match_device() acts the same when CONFIG_OF is set and when it > isn't. Add of_match_ptr() here so that this doesn't break when that > change is made to the of_match_device() API. > > Cc: Arnd Bergmann > Cc: Geert Uytterhoeven > Cc: Jean Delvare > Cc: Guenter Roeck > Cc: Rob Herring > Cc: Frank Rowand > Cc: > Signed-off-by: Stephen Boyd Acked-by: Guenter Roeck > --- > > Please ack or pick for immediate merge so the last patch can be merged. > > drivers/hwmon/lm70.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/hwmon/lm70.c b/drivers/hwmon/lm70.c > index 4122e59f0bb4..57480dada955 100644 > --- a/drivers/hwmon/lm70.c > +++ b/drivers/hwmon/lm70.c > @@ -155,7 +155,7 @@ static int lm70_probe(struct spi_device *spi) > struct lm70 *p_lm70; > int chip; > > - match = of_match_device(lm70_of_ids, >dev); > + match = of_match_device(of_match_ptr(lm70_of_ids), >dev); > if (match) > chip = (int)(uintptr_t)match->data; > else > -- > Sent by a computer through tubes >
[PATCH 09/10] hwmon: (lm70) Avoid undefined reference to match table
We're going to remove of_match_ptr() from the definition of of_match_device() when CONFIG_OF=n. This way we can always be certain that of_match_device() acts the same when CONFIG_OF is set and when it isn't. Add of_match_ptr() here so that this doesn't break when that change is made to the of_match_device() API. Cc: Arnd Bergmann Cc: Geert Uytterhoeven Cc: Jean Delvare Cc: Guenter Roeck Cc: Rob Herring Cc: Frank Rowand Cc: Signed-off-by: Stephen Boyd --- Please ack or pick for immediate merge so the last patch can be merged. drivers/hwmon/lm70.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/hwmon/lm70.c b/drivers/hwmon/lm70.c index 4122e59f0bb4..57480dada955 100644 --- a/drivers/hwmon/lm70.c +++ b/drivers/hwmon/lm70.c @@ -155,7 +155,7 @@ static int lm70_probe(struct spi_device *spi) struct lm70 *p_lm70; int chip; - match = of_match_device(lm70_of_ids, >dev); + match = of_match_device(of_match_ptr(lm70_of_ids), >dev); if (match) chip = (int)(uintptr_t)match->data; else -- Sent by a computer through tubes
[PATCH 00/10] Stop NULLifying match pointer in of_match_device()
of_match_device() uses of_match_ptr() to make the match table argument NULL via the pre-processor when CONFIG_OF=n. This makes life harder for compilers who think that match tables are never used and warn about unused variables when CONFIG_OF=n. This series changes various callers to use of_device_get_match_data() instead, which doesn't have this problem, and removes the of_match_ptr() usage from of_match_device() so that the compiler can stop complaining about unused variables. It will do dead code elimination instead and remove the match table if it isn't actually used. Huge Cc list! Cc: Alessandro Zummo Cc: Alexandre Belloni Cc: Alexandre Torgue Cc: Cc: Andrew Lunn Cc: Arnd Bergmann Cc: Dan Murphy Cc: "David S. Miller" Cc: Frank Rowand Cc: Geert Uytterhoeven Cc: Greg Kroah-Hartman Cc: Gregory Clement Cc: Grygorii Strashko Cc: Guenter Roeck Cc: Jacek Anaszewski Cc: Jacopo Mondi Cc: Jaroslav Kysela Cc: Jason Cooper Cc: Jean Delvare Cc: Jiri Slaby Cc: Liam Girdwood Cc: Cc: Cc: Cc: Cc: Cc: Cc: Cc: Cc: Cc: Mark Brown Cc: Mauro Carvalho Chehab Cc: Maxime Coquelin Cc: Paul Cercueil Cc: Pavel Machek Cc: Richard Leitner Cc: Riku Voipio Cc: Rob Herring Cc: Sebastian Hesselbarth Cc: Takashi Iwai Stephen Boyd (10): leds: pca953x: Use of_device_get_match_data() media: renesas-ceu: Use of_device_get_match_data() rtc: armada38x: Use of_device_get_match_data() drivers: net: davinci_mdio: Use of_device_get_match_data() serial: stm32: Use of_device_get_match_data() usb: usb251xb: Use of_device_get_match_data() ASoC: jz4740: Use of_device_get_match_data() spi: gpio: Look for a device node instead of match hwmon: (lm70) Avoid undefined reference to match table of/device: Don't NULLify match table in of_match_device() with CONFIG_OF=n drivers/hwmon/lm70.c | 2 +- drivers/leds/leds-pca9532.c| 14 + drivers/media/platform/renesas-ceu.c | 2 +- drivers/net/ethernet/ti/davinci_mdio.c | 12 ++--- drivers/rtc/rtc-armada38x.c| 10 ++-- drivers/spi/spi-gpio.c | 5 +- drivers/tty/serial/stm32-usart.c | 71 -- drivers/tty/serial/stm32-usart.h | 2 +- drivers/usb/misc/usb251xb.c| 12 ++--- include/linux/of_device.h | 4 +- sound/soc/jz4740/jz4740-i2s.c | 5 +- 11 files changed, 55 insertions(+), 84 deletions(-) base-commit: 54ecb8f7028c5eb3d740bb82b0f1d90f2df63c5c -- Sent by a computer through tubes
Re: [PATCH v3 2/2] hwmon: Add driver for Texas Instruments TMP512/513 sensor chips
On Thu, Oct 03, 2019 at 06:34:41PM +, Tremblay, Eric wrote:
> hwmon: Add DT bindings for TMP513 driver
>
$subject is misleading.
> Signed-off-by: Eric Tremblay
> ---
> .../devicetree/bindings/hwmon/tmp513.txt | 33 +++
> 1 file changed, 33 insertions(+)
> create mode 100644 Documentation/devicetree/bindings/hwmon/tmp513.txt
>
> diff --git a/Documentation/devicetree/bindings/hwmon/tmp513.txt
> b/Documentation/devicetree/bindings/hwmon/tmp513.txt
> new file mode 100644
> index ..341c764e9173
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/hwmon/tmp513.txt
> @@ -0,0 +1,33 @@
> +TMP513 system monitor sensor
> +-
> +
> +Require node properties:
> +- compatible: one of
> + "ti,tmp512"
> + "ti,tmp513"
> +- reg : the I2C address of the device. This is 0x5c, 0x5d, 0x5e, or 0x5f.
> +
> +Optional properties:
> +- shunt-resistor-uohm : The shunt resistor value in uOhm. If 0, the
> calibration process
> + will be skiped and the current and power measurement engine
> will not work.
> + Temperature and voltage measurement will continue to work. The
> shunt value also
> + need to respect : rshunt <= pga-gain * 40 * 1000 * 1000. If
> not, it's not possible
> + to compute a valid calibration value.
If this is not mandatory, I would suggest to declare a sensible default,
such as 1 mOhm / 1000 uOhm.
> +- bus-voltage-range: 32V or 16V, default to 32V.
Question for Rob: Does this need "-volt" as part of the name ?
> +- pga-gain: The gain value for the PGA function. This is 8, 4, 2 or 1.
> Default to 8.
It might be useful to explain "pga".
> +- temp_2_nfactor: nFactor for remote channel 1. Value according to Table 11
> in datasheet.
> +- temp_3_nfactor: nFactor for remote channel 2. Value according to Table 11
> in datasheet.
> +- temp_4_nfactor: nFactor for remote channel 3. Value according to Table 11
> in datasheet,
Maybe an array and temp-nfactor, with 0x0 meaning "use default" ?
> + TMP513 only.
> +
> +Example:
> + tmp513@5c {
> + compatible = "ti,tmp513";
> + reg = <0x5C>;
> + shunt-resistor-uohm = <33>;
> + bus-voltage-range = <32>;
> + pga-gain = <8>;
> + temp_2_nfactor = <0x01>;
> + temp_3_nfactor = <0x02>;
> + temp_4_nfactor = <0x03>;
> + };
> --
> 2.17.1
>
RE: [PATCH v3 0/2] hwmon: Add driver for Texas Instruments TMP512/513 sensor chips
> On Thu, Oct 03, 2019 at 06:35:51PM +, Tremblay, Eric wrote: > > > Version three of the driver for Texas Instruments TMP512/513 sensors. > > > > > > Thanks again for the review, I think we are getting close. I'm not > > > in an easy position for now to test the driver without device tree. > > > I can put the required effort to > > > > Sorry, I mean that I'm not in a good position to test it, but I can put the > > required effort to if needed. > > All you would need to do is to use device_ API functions instead of > of_device_ API functions, so I can not entirely follow your logic. I was clearly missing some piece of knowledge about the device API... I did some research and I was able to instantiate and test the driver without using the of_device API. The version 3 work as expected with or without platform_data provided. Eric > > Guenter > > > > > > > Main changes from version 2: > > > - Remove the neccessity to provide a shunt-resistor value > > > - Remove max-expected-current-ma configuration and always > > > use max possible current for calibration > > > - Make sure calculation during calibration can't overflow > > > - Add value clamping for DT value and runtime parameters > > > - Support non DT system > > > - Move hysteresis from DT to standard attribute > > > > > > Main changes from version 1: > > > - Use the with_info API instead of sysfs attributes. > > > - Remove non-standard attributes and raw value. > > > - Move settings that were non-standard attributes to > > > device tree, update documentation as well. > > > - Fix coding style issues > > > > > > Eric Tremblay (2): > > > Add driver for Texas Instruments TMP512/513 sensor chips. > > > Add DT bindings for TMP513 driver > > > > > > .../devicetree/bindings/hwmon/tmp513.txt | 33 + > > > Documentation/hwmon/tmp513.rst| 102 ++ > > > MAINTAINERS | 7 + > > > drivers/hwmon/Kconfig | 10 + > > > drivers/hwmon/Makefile| 1 + > > > drivers/hwmon/tmp513.c| 901 ++ > > > include/linux/platform_data/tmp513.h | 28 + > > > 7 files changed, 1082 insertions(+) create mode 100644 > > > Documentation/devicetree/bindings/hwmon/tmp513.txt > > > create mode 100644 Documentation/hwmon/tmp513.rst create mode > > > 100644 drivers/hwmon/tmp513.c create mode 100644 > > > include/linux/platform_data/tmp513.h > > > > > > -- > > > 2.17.1 > >
