| Message ID | 1343110735-10640-4-git-send-email-anilkumar@ti.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi Mark, On Tue, Jul 24, 2012 at 11:48:54, AnilKumar, Chimata wrote: > Add tps65910 regulator device tree data to AM335x-EVM by adding > regulator consumers with tightened constraints and regulator-name. > TPS65910 regulator handle can be obtained by using this regulator > name. > > This patch also add I2C node with I2C frequency and tps65910 PMIC > I2C slave address. > > Signed-off-by: AnilKumar Ch <anilkumar@ti.com> > --- > arch/arm/boot/dts/am335x-evm.dts | 28 ++++++++++++++++++++++++++++ > 1 file changed, 28 insertions(+) > > diff --git a/arch/arm/boot/dts/am335x-evm.dts b/arch/arm/boot/dts/am335x-evm.dts > index d6a97d9..e4e1ccb 100644 > --- a/arch/arm/boot/dts/am335x-evm.dts > +++ b/arch/arm/boot/dts/am335x-evm.dts > @@ -18,3 +18,31 @@ > reg = <0x80000000 0x10000000>; /* 256 MB */ > }; > }; > + > +&i2c1 { > + clock-frequency = <400000>; > + > + tps: tps@2D { > + reg = <0x2D>; > + }; > +}; > + > +/include/ "tps65910.dtsi" > + > +&vdd1_reg { > + /* VDD_MPU voltage limits 0.95V - 1.26V with +/-4% tolerance */ > + regulator-name = "vdd_mpu"; > + regulator-min-microvolt = <912500>; > + regulator-max-microvolt = <1312500>; > + regulator-boot-on; > + regulator-always-on; > +}; I want to understand how to add parent regulator nodes? Like input-supply node to regulators. With "regulator: tps65910: add support for input supply" patch, regulator registration fails if we are not providing the input_supply node with proper input supply name. AM335x EVM have one dedicated input supply which provide supplies to all the PMIC regulators. According to tps65910 "VDD1 regulator" input supply is from vcc1 so we have to create a dummy/parent/some node to handle this. Example below has the details. In am335x-evm.dts file vcc1_supply: fixedregulator@0 { compatible = "regulator-fixed"; regulator-name = "vcc1"; regulator-boot-on; }; /include/ "tps65910.dtsi" &tps { regulators { vcc1-supply = <&vcc1_supply>; ... }; }; Similarly we have to add 7 more fixed regulators entries to provide input supply to different PMIC regulators. Am I doing correct thing? Or is there any better way to handle this? Regards AnilKumar
On Tue, Aug 14, 2012 at 02:25:18PM +0000, AnilKumar, Chimata wrote: > AM335x EVM have one dedicated input supply which provide supplies to all > the PMIC regulators. According to tps65910 "VDD1 regulator" input supply > is from vcc1 so we have to create a dummy/parent/some node to handle this. > Example below has the details. Your device tree should accurately reflect the hardware. > Similarly we have to add 7 more fixed regulators entries to provide input > supply to different PMIC regulators. If you have all these unrelated supplies in the system then your device tree should accurately reflect that. However this would be a very unusual hardware design. Usually there would be a very small set of system supply rails, frequently only one.
Hi Mark, On Wed, Aug 15, 2012 at 18:23:39, Mark Brown wrote: > On Tue, Aug 14, 2012 at 02:25:18PM +0000, AnilKumar, Chimata wrote: > > > AM335x EVM have one dedicated input supply which provide supplies to all > > the PMIC regulators. According to tps65910 "VDD1 regulator" input supply > > is from vcc1 so we have to create a dummy/parent/some node to handle this. > > Example below has the details. > > Your device tree should accurately reflect the hardware. > > > Similarly we have to add 7 more fixed regulators entries to provide input > > supply to different PMIC regulators. > > If you have all these unrelated supplies in the system then your device > tree should accurately reflect that. However this would be a very In that case I have to add all these fixed regulators. > unusual hardware design. Usually there would be a very small set of > system supply rails, frequently only one. In case of AM335X-EVM one dedicated supply (5v) feeding to all the input supplies (vcc1-7 and vccio) of the regulators. tps65910 PMIC have vcc1-7 and vccio input supplies which are feeding to all the regulators. (like one - one and one - many). Tps65910 regulator drivers registration fails if we not providing the correct input supply with supply name. One option I can see is that fixed voltage regulators. If we do this then total 21 (8 input supply fixed regulators and others are out regulators) are present in the system. If this is the correct approach then I will submit next version of patch. Thanks AnilKumar
On Wed, Aug 15, 2012 at 04:47:02PM +0000, AnilKumar, Chimata wrote: > On Wed, Aug 15, 2012 at 18:23:39, Mark Brown wrote: > > If you have all these unrelated supplies in the system then your device > > tree should accurately reflect that. However this would be a very > In that case I have to add all these fixed regulators. Why? You then immediately go on to say... > > unusual hardware design. Usually there would be a very small set of > > system supply rails, frequently only one. > In case of AM335X-EVM one dedicated supply (5v) feeding to all the input > supplies (vcc1-7 and vccio) of the regulators. ...that they all map onto the same physical supply?
Mark, On Wed, Aug 15, 2012 at 22:24:32, Mark Brown wrote: > On Wed, Aug 15, 2012 at 04:47:02PM +0000, AnilKumar, Chimata wrote: > > On Wed, Aug 15, 2012 at 18:23:39, Mark Brown wrote: > > > > If you have all these unrelated supplies in the system then your device > > > tree should accurately reflect that. However this would be a very > > > In that case I have to add all these fixed regulators. > > Why? You then immediately go on to say... Because all the nodes/regulators should be provided/registered otherwise "regulator_dev_lookup" returns NULL if there are no regulator matches with the supply_name. I have gone through this thread http://patchwork.ozlabs.org/patch/166682/ But problem here is single input supply battery/main is not sufficient. pmic->desc[i].supply_name = info->vin_name; with this assignment regulator core searches for exact supply name vcc1 or vcc2 etc if there are no matches then it will not register any regulators. > > > > unusual hardware design. Usually there would be a very small set of > > > system supply rails, frequently only one. > > > In case of AM335X-EVM one dedicated supply (5v) feeding to all the input > > supplies (vcc1-7 and vccio) of the regulators. > > ...that they all map onto the same physical supply? > Yes single physical supply, if we want to add single fixed regulator node to feed all the regulators then driver have to modify to support this. Similar to revert of "regulator: tps65910: add support for input supply" Patch. Thanks AnilKumar
Mark, Small correction On Thu, Aug 16, 2012 at 18:52:25, AnilKumar, Chimata wrote: > Mark, > > On Wed, Aug 15, 2012 at 22:24:32, Mark Brown wrote: > > On Wed, Aug 15, 2012 at 04:47:02PM +0000, AnilKumar, Chimata wrote: > > > On Wed, Aug 15, 2012 at 18:23:39, Mark Brown wrote: > > > > > > If you have all these unrelated supplies in the system then your device > > > > tree should accurately reflect that. However this would be a very > > > > > In that case I have to add all these fixed regulators. > > > > Why? You then immediately go on to say... > > Because all the nodes/regulators should be provided/registered otherwise > "regulator_dev_lookup" returns NULL if there are no regulator matches with > the supply_name. > > I have gone through this thread > http://patchwork.ozlabs.org/patch/166682/ > > But problem here is single input supply battery/main is not sufficient. > > pmic->desc[i].supply_name = info->vin_name; > with this assignment regulator core searches for exact supply name vcc1 or > vcc2 etc if there are no matches then it will not register any regulators. > > > > > > > unusual hardware design. Usually there would be a very small set of > > > > system supply rails, frequently only one. > > > > > In case of AM335X-EVM one dedicated supply (5v) feeding to all the input > > > supplies (vcc1-7 and vccio) of the regulators. > > > > ...that they all map onto the same physical supply? > > > > Yes single physical supply, if we want to add single fixed regulator node > to feed all the regulators then driver have to modify to support this. > Similar to revert of "regulator: tps65910: add support for input supply" ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Copy/paste error Similar to revert of this patch "regulator: tps65910: set input_supply on desc unconditionally" Thanks AnilKumar
On Thu, Aug 16, 2012 at 01:22:27PM +0000, AnilKumar, Chimata wrote: > But problem here is single input supply battery/main is not sufficient. What makes you say this? > pmic->desc[i].supply_name = info->vin_name; > with this assignment regulator core searches for exact supply name vcc1 or > vcc2 etc if there are no matches then it will not register any regulators. Why does this mean you need multiple regulators? A single regulator can of course supply multiple devices.
Hi Mark, On Thu, Aug 16, 2012 at 19:30:56, Mark Brown wrote: > On Thu, Aug 16, 2012 at 01:22:27PM +0000, AnilKumar, Chimata wrote: > > > But problem here is single input supply battery/main is not sufficient. > > What makes you say this? > > > pmic->desc[i].supply_name = info->vin_name; > > with this assignment regulator core searches for exact supply name vcc1 or > > vcc2 etc if there are no matches then it will not register any regulators. > > Why does this mean you need multiple regulators? A single regulator can > of course supply multiple devices. I got your point but tps65910 driver is failing if we add single fixed regulator. Solution is driver needs to change or add multiple fixed regulators. Let's take this example ocp { vcc5v_supply: fixedregulator@0 { compatible = "regulator-fixed"; regulator-name = "vcc5v"; regulator-boot-on; }; }; /include/ "tps65910.dtsi" &tps { regulators { vcc1-supply = <&vcc5v_supply>; vcc2-supply = <&vcc5v_supply>; vcc3-supply = <&vcc5v_supply>; vcc4-supply = <&vcc5v_supply>; vcc5-supply = <&vcc5v_supply>; vcc6-supply = <&vcc5v_supply>; vcc7-supply = <&vcc5v_supply>; vccio-supply = <&vcc5v_supply>; vrtc_reg: regulator@0 { /* vrtc input is vcc7 */ regulator-always-on; }; ... ... }; }; These changes are added to am335x-evm.dts file after this change I am seeing below error. vrtc: tps65910 0-002d: Failed to find supply vcc7 tps65910 0-002d: failed to register tps65910-pmic regulator To get rid of this error we have to change the driver like this if (info->vin_name) { snprintf(in_supply, 32, "%s-supply", info->vin_name); if (of_find_property(np, in_supply, 0)) pmic_plat_data->input_supply[idx] = info->vin_name; } And pmic->desc[i].supply_name = pmic_plat_data->input_supply[i]; Regards AnilKumar
On Mon, Aug 20, 2012 at 06:27:01AM +0000, AnilKumar, Chimata wrote: > On Thu, Aug 16, 2012 at 19:30:56, Mark Brown wrote: > > Why does this mean you need multiple regulators? A single regulator can > > of course supply multiple devices. > I got your point but tps65910 driver is failing if we add single fixed > regulator. Solution is driver needs to change or add multiple fixed > regulators. No, really. One regulator can supply many things, not being a. > vrtc: > tps65910 0-002d: Failed to find supply vcc7 > tps65910 0-002d: failed to register tps65910-pmic regulator So you've specified a whole bunch of supplies and this one individual supply is not matching (and it's not the first...). You should investigate what's gone wrong there rather than bodging around the problem. I can't immediately say anything much based on the information you've given.
Hi Mark, On Tue, Aug 21, 2012 at 01:37:15, Mark Brown wrote: > On Mon, Aug 20, 2012 at 06:27:01AM +0000, AnilKumar, Chimata wrote: > > On Thu, Aug 16, 2012 at 19:30:56, Mark Brown wrote: > > > > Why does this mean you need multiple regulators? A single regulator can > > > of course supply multiple devices. > > > I got your point but tps65910 driver is failing if we add single fixed > > regulator. Solution is driver needs to change or add multiple fixed > > regulators. > > No, really. One regulator can supply many things, not being a. > > > vrtc: > > tps65910 0-002d: Failed to find supply vcc7 > > tps65910 0-002d: failed to register tps65910-pmic regulator > > So you've specified a whole bunch of supplies and this one individual > supply is not matching (and it's not the first...). You should > investigate what's gone wrong there rather than bodging around the > problem. I can't immediately say anything much based on the information > you've given. > Sorry for the noise in this issue, I found the root cause. I was defining the supply names like this which is wrong. &tps { regulators { vcc1-supply = <&vcc5v_supply>; vcc2-supply = <&vcc5v_supply>; vcc3-supply = <&vcc5v_supply>; vcc4-supply = <&vcc5v_supply>; vcc5-supply = <&vcc5v_supply>; vcc6-supply = <&vcc5v_supply>; vcc7-supply = <&vcc5v_supply>; vccio-supply = <&vcc5v_supply>; vrtc_reg: regulator@0 { /* vrtc input is vcc7 */ regulator-always-on; }; ... ... }; }; This was suppose to be &tps { vcc1-supply = <&vcc5v_supply>; vcc2-supply = <&vcc5v_supply>; vcc3-supply = <&vcc5v_supply>; vcc4-supply = <&vcc5v_supply>; vcc5-supply = <&vcc5v_supply>; vcc6-supply = <&vcc5v_supply>; vcc7-supply = <&vcc5v_supply>; vccio-supply = <&vcc5v_supply>; regulators { vrtc_reg: regulator@0 { /* vrtc input is vcc7 */ regulator-always-on; }; ... ... }; }; Thanks for your continuous feedback. Thanks AnilKumar
diff --git a/arch/arm/boot/dts/am335x-evm.dts b/arch/arm/boot/dts/am335x-evm.dts index d6a97d9..e4e1ccb 100644 --- a/arch/arm/boot/dts/am335x-evm.dts +++ b/arch/arm/boot/dts/am335x-evm.dts @@ -18,3 +18,31 @@ reg = <0x80000000 0x10000000>; /* 256 MB */ }; }; + +&i2c1 { + clock-frequency = <400000>; + + tps: tps@2D { + reg = <0x2D>; + }; +}; + +/include/ "tps65910.dtsi" + +&vdd1_reg { + /* VDD_MPU voltage limits 0.95V - 1.26V with +/-4% tolerance */ + regulator-name = "vdd_mpu"; + regulator-min-microvolt = <912500>; + regulator-max-microvolt = <1312500>; + regulator-boot-on; + regulator-always-on; +}; + +&vdd2_reg { + /* VDD_CORE voltage limits 0.95V - 1.1V with +/-4% tolerance */ + regulator-name = "vdd_core"; + regulator-min-microvolt = <912500>; + regulator-max-microvolt = <1150000>; + regulator-boot-on; + regulator-always-on; +};
Add tps65910 regulator device tree data to AM335x-EVM by adding regulator consumers with tightened constraints and regulator-name. TPS65910 regulator handle can be obtained by using this regulator name. This patch also add I2C node with I2C frequency and tps65910 PMIC I2C slave address. Signed-off-by: AnilKumar Ch <anilkumar@ti.com> --- arch/arm/boot/dts/am335x-evm.dts | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+)