| Message ID | 1442235619-4029-2-git-send-email-punit.agrawal@arm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Mon, Sep 14, 2015 at 02:00:14PM +0100, Punit Agrawal wrote: > The System Control Processor (SCP) provides access to SoC sensors via > the System Control and Power Interface (SCPI) Message Protocol. Add > bindings to allow probing of these sensors. Also support referencing > of the sensors for setting up thermal zones via the thermal DT > bindings. > > Signed-off-by: Punit Agrawal <punit.agrawal@arm.com> > Cc: Rob Herring <robh+dt@kernel.org> > Cc: Mark Rutland <mark.rutland@arm.com> > Cc: Sudeep Holla <sudeep.holla@arm.com> > --- > Documentation/devicetree/bindings/arm/arm,scpi.txt | 15 +++++++++++++++ > 1 file changed, 15 insertions(+) > > diff --git a/Documentation/devicetree/bindings/arm/arm,scpi.txt b/Documentation/devicetree/bindings/arm/arm,scpi.txt > index f002460..af66f7f 100644 > --- a/Documentation/devicetree/bindings/arm/arm,scpi.txt > +++ b/Documentation/devicetree/bindings/arm/arm,scpi.txt > @@ -72,8 +72,18 @@ Required sub-node properties: > - compatible : should be "arm,juno-scp-shmem" for Non-secure SRAM based > shared memory on Juno platforms > > +Sensor bindings for the sensors based on SCPI Message Protocol > +-------------------------------------------------------------- > +SCPI provides an API to access the various sensors on the SoC. > + > +Required properties: > +- compatible : should be "arm,scpi-sensors". > +- #thermal-sensor-cells: should be set to 1. This property follows the > + thermal device tree bindings[2]. You need to specify what the valid values for this cell are. Mark. > + > [0] http://infocenter.arm.com/help/topic/com.arm.doc.dui0922b/index.html > [1] Documentation/devicetree/bindings/clock/clock-bindings.txt > +[2] Documentation/devicetree/bindings/thermal/thermal.txt > > Example: > > @@ -122,6 +132,11 @@ scpi_protocol: scpi@2e000000 { > clock-output-names = "pxlclk0", "pxlclk1"; > }; > }; > + > + scpi_sensors0: sensors { > + compatible = "arm,scpi-sensors"; > + #thermal-sensor-cells = <1>; > + }; > }; > > cpu@0 { > -- > 2.5.0 >
Hi Mark, Thanks for taking a look. Mark Rutland <mark.rutland@arm.com> writes: > On Mon, Sep 14, 2015 at 02:00:14PM +0100, Punit Agrawal wrote: >> The System Control Processor (SCP) provides access to SoC sensors via >> the System Control and Power Interface (SCPI) Message Protocol. Add >> bindings to allow probing of these sensors. Also support referencing >> of the sensors for setting up thermal zones via the thermal DT >> bindings. >> >> Signed-off-by: Punit Agrawal <punit.agrawal@arm.com> >> Cc: Rob Herring <robh+dt@kernel.org> >> Cc: Mark Rutland <mark.rutland@arm.com> >> Cc: Sudeep Holla <sudeep.holla@arm.com> >> --- >> Documentation/devicetree/bindings/arm/arm,scpi.txt | 15 +++++++++++++++ >> 1 file changed, 15 insertions(+) >> >> diff --git a/Documentation/devicetree/bindings/arm/arm,scpi.txt b/Documentation/devicetree/bindings/arm/arm,scpi.txt >> index f002460..af66f7f 100644 >> --- a/Documentation/devicetree/bindings/arm/arm,scpi.txt >> +++ b/Documentation/devicetree/bindings/arm/arm,scpi.txt >> @@ -72,8 +72,18 @@ Required sub-node properties: >> - compatible : should be "arm,juno-scp-shmem" for Non-secure SRAM based >> shared memory on Juno platforms >> >> +Sensor bindings for the sensors based on SCPI Message Protocol >> +-------------------------------------------------------------- >> +SCPI provides an API to access the various sensors on the SoC. >> + >> +Required properties: >> +- compatible : should be "arm,scpi-sensors". >> +- #thermal-sensor-cells: should be set to 1. This property follows the >> + thermal device tree bindings[2]. > > You need to specify what the valid values for this cell are. The enumeration depends on the number of sensors exported by SCP firmware - which is platform dependent. I could add add something like if you think that is helpful - "Valid cell value is a number between 0..n-1, where n is the number of sensors exported by SCP firmware." Thanks, Punit > > Mark. > >> + >> [0] http://infocenter.arm.com/help/topic/com.arm.doc.dui0922b/index.html >> [1] Documentation/devicetree/bindings/clock/clock-bindings.txt >> +[2] Documentation/devicetree/bindings/thermal/thermal.txt >> >> Example: >> >> @@ -122,6 +132,11 @@ scpi_protocol: scpi@2e000000 { >> clock-output-names = "pxlclk0", "pxlclk1"; >> }; >> }; >> + >> + scpi_sensors0: sensors { >> + compatible = "arm,scpi-sensors"; >> + #thermal-sensor-cells = <1>; >> + }; >> }; >> >> cpu@0 { >> -- >> 2.5.0 >> > -- > To unsubscribe from this list: send the line "unsubscribe devicetree" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html
> >> +Sensor bindings for the sensors based on SCPI Message Protocol > >> +-------------------------------------------------------------- > >> +SCPI provides an API to access the various sensors on the SoC. > >> + > >> +Required properties: > >> +- compatible : should be "arm,scpi-sensors". > >> +- #thermal-sensor-cells: should be set to 1. This property follows the > >> + thermal device tree bindings[2]. > > > > You need to specify what the valid values for this cell are. > > The enumeration depends on the number of sensors exported by SCP > firmware - which is platform dependent. I could add add something like > if you think that is helpful - > > "Valid cell value is a number between 0..n-1, where n is the number > of sensors exported by SCP firmware." Can the FW identifer space have holes? Or are they always contiguous? If this is the same as the raw FW identifer value, specify that. Otherwise, you need to specify the mapping. There needs to be enough information for a dts author to figure out which values to place in the DT. Thanks, Mark.
Mark Rutland <mark.rutland@arm.com> writes: >> >> +Sensor bindings for the sensors based on SCPI Message Protocol >> >> +-------------------------------------------------------------- >> >> +SCPI provides an API to access the various sensors on the SoC. >> >> + >> >> +Required properties: >> >> +- compatible : should be "arm,scpi-sensors". >> >> +- #thermal-sensor-cells: should be set to 1. This property follows the >> >> + thermal device tree bindings[2]. >> > >> > You need to specify what the valid values for this cell are. >> >> The enumeration depends on the number of sensors exported by SCP >> firmware - which is platform dependent. I could add add something like >> if you think that is helpful - >> >> "Valid cell value is a number between 0..n-1, where n is the number >> of sensors exported by SCP firmware." > > Can the FW identifer space have holes? Or are they always contiguous? The way the SCP interface is defined, the sensor identifiers are contiguous, but not all are temperature sensors. > > If this is the same as the raw FW identifer value, specify that. > Otherwise, you need to specify the mapping. I'll update the patch to add mappings for Juno r0 (and r1 if I can get my hands on one). > > There needs to be enough information for a dts author to figure out > which values to place in the DT. I understand. Except sometimes it is hard to get the firmware to commit to not modify the ordering - discoverability and all that. :) > > Thanks, > Mark. > -- > To unsubscribe from this list: send the line "unsubscribe devicetree" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Sep 14, 2015 at 03:38:36PM +0100, Punit Agrawal wrote: > Mark Rutland <mark.rutland@arm.com> writes: > > >> >> +Sensor bindings for the sensors based on SCPI Message Protocol > >> >> +-------------------------------------------------------------- > >> >> +SCPI provides an API to access the various sensors on the SoC. > >> >> + > >> >> +Required properties: > >> >> +- compatible : should be "arm,scpi-sensors". > >> >> +- #thermal-sensor-cells: should be set to 1. This property follows the > >> >> + thermal device tree bindings[2]. > >> > > >> > You need to specify what the valid values for this cell are. > >> > >> The enumeration depends on the number of sensors exported by SCP > >> firmware - which is platform dependent. I could add add something like > >> if you think that is helpful - > >> > >> "Valid cell value is a number between 0..n-1, where n is the number > >> of sensors exported by SCP firmware." > > > > Can the FW identifer space have holes? Or are they always contiguous? > > The way the SCP interface is defined, the sensor identifiers are > contiguous, but not all are temperature sensors. Ok. So how exactly are they enumerated for this binding? > > If this is the same as the raw FW identifer value, specify that. > > Otherwise, you need to specify the mapping. > > I'll update the patch to add mappings for Juno r0 (and r1 if I can get > my hands on one). If there's identical logic mapping the two, we might just be able to describe that rather than having to add tables all the time. > > There needs to be enough information for a dts author to figure out > > which values to place in the DT. > > I understand. Except sometimes it is hard to get the firmware to commit to not > modify the ordering - discoverability and all that. :) If they do that, then things are broken regardless, no? I guess that'll be clear if/when I see how the mapping works. Thanks, Mark.
Mark Rutland <mark.rutland@arm.com> writes: > On Mon, Sep 14, 2015 at 03:38:36PM +0100, Punit Agrawal wrote: >> Mark Rutland <mark.rutland@arm.com> writes: >> >> >> >> +Sensor bindings for the sensors based on SCPI Message Protocol >> >> >> +-------------------------------------------------------------- >> >> >> +SCPI provides an API to access the various sensors on the SoC. >> >> >> + >> >> >> +Required properties: >> >> >> +- compatible : should be "arm,scpi-sensors". >> >> >> +- #thermal-sensor-cells: should be set to 1. This property follows the >> >> >> + thermal device tree bindings[2]. >> >> > >> >> > You need to specify what the valid values for this cell are. >> >> >> >> The enumeration depends on the number of sensors exported by SCP >> >> firmware - which is platform dependent. I could add add something like >> >> if you think that is helpful - >> >> >> >> "Valid cell value is a number between 0..n-1, where n is the number >> >> of sensors exported by SCP firmware." >> > >> > Can the FW identifer space have holes? Or are they always contiguous? >> >> The way the SCP interface is defined, the sensor identifiers are >> contiguous, but not all are temperature sensors. > > Ok. So how exactly are they enumerated for this binding? The sensor enumeration for r0 (which I've verified) and r1 can be found at [0]. The valid cell values are identifiers for the temperature sensors. > >> > If this is the same as the raw FW identifer value, specify that. >> > Otherwise, you need to specify the mapping. >> >> I'll update the patch to add mappings for Juno r0 (and r1 if I can get >> my hands on one). > > If there's identical logic mapping the two, we might just be able to > describe that rather than having to add tables all the time. > After seeing the mapping already published, I am wondering if there is any value in duplicating the information. If there are no objections, I'll update this patch to add pointers instead. >> > There needs to be enough information for a dts author to figure out >> > which values to place in the DT. >> >> I understand. Except sometimes it is hard to get the firmware to commit to not >> modify the ordering - discoverability and all that. :) > > If they do that, then things are broken regardless, no? I guess that'll > be clear if/when I see how the mapping works. [0] http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0922b/apas03s22.html > > Thanks, > Mark. > -- > To unsubscribe from this list: send the line "unsubscribe devicetree" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Sep 14, 2015 at 04:01:09PM +0100, Punit Agrawal wrote: > Mark Rutland <mark.rutland@arm.com> writes: > > > On Mon, Sep 14, 2015 at 03:38:36PM +0100, Punit Agrawal wrote: > >> Mark Rutland <mark.rutland@arm.com> writes: > >> > >> >> >> +Sensor bindings for the sensors based on SCPI Message Protocol > >> >> >> +-------------------------------------------------------------- > >> >> >> +SCPI provides an API to access the various sensors on the SoC. > >> >> >> + > >> >> >> +Required properties: > >> >> >> +- compatible : should be "arm,scpi-sensors". > >> >> >> +- #thermal-sensor-cells: should be set to 1. This property follows the > >> >> >> + thermal device tree bindings[2]. > >> >> > > >> >> > You need to specify what the valid values for this cell are. > >> >> > >> >> The enumeration depends on the number of sensors exported by SCP > >> >> firmware - which is platform dependent. I could add add something like > >> >> if you think that is helpful - > >> >> > >> >> "Valid cell value is a number between 0..n-1, where n is the number > >> >> of sensors exported by SCP firmware." > >> > > >> > Can the FW identifer space have holes? Or are they always contiguous? > >> > >> The way the SCP interface is defined, the sensor identifiers are > >> contiguous, but not all are temperature sensors. > > > > Ok. So how exactly are they enumerated for this binding? > > The sensor enumeration for r0 (which I've verified) and r1 can be found > at [0]. The valid cell values are identifiers for the temperature sensors. Ok. That's fine by me; I was confused and thought that there was some internal mapping. > >> > If this is the same as the raw FW identifer value, specify that. > >> > Otherwise, you need to specify the mapping. > >> > >> I'll update the patch to add mappings for Juno r0 (and r1 if I can get > >> my hands on one). > > > > If there's identical logic mapping the two, we might just be able to > > describe that rather than having to add tables all the time. > > > > After seeing the mapping already published, I am wondering if there is > any value in duplicating the information. If there are no objections, > I'll update this patch to add pointers instead. That's fine by me. The important part is that the value is a raw Sensor ID as the FW uses. So long as we state that, the IDs themselves can come from whatever documentation is valid for a particular instance. Thanks for the info! Thanks, Mark.
Mark Rutland <mark.rutland@arm.com> writes: > On Mon, Sep 14, 2015 at 04:01:09PM +0100, Punit Agrawal wrote: >> Mark Rutland <mark.rutland@arm.com> writes: >> [...] >> >> After seeing the mapping already published, I am wondering if there is >> any value in duplicating the information. If there are no objections, >> I'll update this patch to add pointers instead. > > That's fine by me. The important part is that the value is a raw Sensor > ID as the FW uses. So long as we state that, the IDs themselves can come > from whatever documentation is valid for a particular instance. > > Thanks for the info! Sure. I've updated the patch to explain what the valid cell values are along with a pointer to platform documentation for Juno R0 and R1. I've also added an example to show usage of the sensors to create a thermal zone. I'll post a new version shortly. Thanks for your having a look. > > Thanks, > Mark. > -- > To unsubscribe from this list: send the line "unsubscribe linux-pm" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, 2015-09-14 at 15:38 +0100, Punit Agrawal wrote: > Mark Rutland <mark.rutland@arm.com> writes: > > >> >> +Sensor bindings for the sensors based on SCPI Message Protocol > >> >> +-------------------------------------------------------------- > >> >> +SCPI provides an API to access the various sensors on the SoC. > >> >> + > >> >> +Required properties: > >> >> +- compatible : should be "arm,scpi-sensors". > >> >> +- #thermal-sensor-cells: should be set to 1. This property follows the > >> >> + thermal device tree bindings[2]. > >> > > >> > You need to specify what the valid values for this cell are. > >> > >> The enumeration depends on the number of sensors exported by SCP > >> firmware - which is platform dependent. I could add add something like > >> if you think that is helpful - > >> > >> "Valid cell value is a number between 0..n-1, where n is the number > >> of sensors exported by SCP firmware." > > > > Can the FW identifer space have holes? Or are they always contiguous? > > The way the SCP interface is defined, the sensor identifiers are > contiguous, Is there any documentation other than DUI0922A? [1] From what I can seen that just says it's a 16-bit value and doesn't put any particular constraints on its value. [1] http://community.arm.com/servlet/JiveServlet/download/8401-40-18262/DUI0922A_scp_message_interface.pdf
"Jon Medhurst (Tixy)" <tixy@linaro.org> writes: > On Mon, 2015-09-14 at 15:38 +0100, Punit Agrawal wrote: >> Mark Rutland <mark.rutland@arm.com> writes: >> >> >> >> +Sensor bindings for the sensors based on SCPI Message Protocol >> >> >> +-------------------------------------------------------------- >> >> >> +SCPI provides an API to access the various sensors on the SoC. >> >> >> + >> >> >> +Required properties: >> >> >> +- compatible : should be "arm,scpi-sensors". >> >> >> +- #thermal-sensor-cells: should be set to 1. This property follows the >> >> >> + thermal device tree bindings[2]. >> >> > >> >> > You need to specify what the valid values for this cell are. >> >> >> >> The enumeration depends on the number of sensors exported by SCP >> >> firmware - which is platform dependent. I could add add something like >> >> if you think that is helpful - >> >> >> >> "Valid cell value is a number between 0..n-1, where n is the number >> >> of sensors exported by SCP firmware." >> > >> > Can the FW identifer space have holes? Or are they always contiguous? >> >> The way the SCP interface is defined, the sensor identifiers are >> contiguous, > > Is there any documentation other than DUI0922A? [1] From what I can seen > that just says it's a 16-bit value and doesn't put any particular > constraints on its value. Although not explicitly stated, if you look at the Get Sensor Capability [2] and Get Sensor Info [3] commands you can indirectly infer that the Sensor IDs are contiguous. Not the strongest guarantee I know. All platforms currently using SCP (Juno R0 and R1) do indeed expose contiguous identifiers. Maybe we can convince the authors to make it explicit. > > [1] http://community.arm.com/servlet/JiveServlet/download/8401-40-18262/DUI0922A_scp_message_interface.pdf [2] http://arminfo.emea.arm.com/help/index.jsp?topic=/com.arm.doc.dui0922b/ch03s02s21.html [3] http://arminfo.emea.arm.com/help/index.jsp?topic=/com.arm.doc.dui0922b/BABCCCJJ.html
On Tue, 2015-09-15 at 10:37 +0100, Punit Agrawal wrote: > "Jon Medhurst (Tixy)" <tixy@linaro.org> writes: > > > On Mon, 2015-09-14 at 15:38 +0100, Punit Agrawal wrote: > >> Mark Rutland <mark.rutland@arm.com> writes: > >> [...] > >> The way the SCP interface is defined, the sensor identifiers are > >> contiguous, > > > > Is there any documentation other than DUI0922A? [1] From what I can seen > > that just says it's a 16-bit value and doesn't put any particular > > constraints on its value. > > Although not explicitly stated, if you look at the Get Sensor Capability > [2] and Get Sensor Info [3] commands you can indirectly infer that the > Sensor IDs are contiguous. I personally wouldn't even indirectly infer they are contiguous from what the document says. If I were implementing the firmware I would feel quite in my rights to, for example, use the top 8 bits of the ID for a sensor type and the bottom 8 for an index, if that made dispatching of requests more efficient. Or if some optional hardware was detected as missing, leaving some holes in ID space. As a specification of a 'standard' the document seems to be rather lacking. So, Sensor ID should be documented as being "an unsigned integer less than then number of sensors returned by the Get Sensor Capability command", or something like that. I guess clocks and other devices suffer from similar lack of specificity. > Not the strongest guarantee I know. > > All platforms currently using SCP (Juno R0 and R1) do indeed expose > contiguous identifiers. IMO, Linux drivers should be coded to the standard or written specification (where they are available) not the particular implementations available. > > > > [1] http://community.arm.com/servlet/JiveServlet/download/8401-40-18262/DUI0922A_scp_message_interface.pdf > [2] http://arminfo.emea.arm.com/help/index.jsp?topic=/com.arm.doc.dui0922b/ch03s02s21.html > [3] http://arminfo.emea.arm.com/help/index.jsp?topic=/com.arm.doc.dui0922b/BABCCCJJ.html I think those links are on ARM's intranet, they return NXDOMAIN for me.
On Tue, Sep 15, 2015 at 11:46:02AM +0100, Jon Medhurst (Tixy) wrote: > On Tue, 2015-09-15 at 10:37 +0100, Punit Agrawal wrote: > > "Jon Medhurst (Tixy)" <tixy@linaro.org> writes: > > > > > On Mon, 2015-09-14 at 15:38 +0100, Punit Agrawal wrote: > > >> Mark Rutland <mark.rutland@arm.com> writes: > > >> > [...] > > >> The way the SCP interface is defined, the sensor identifiers are > > >> contiguous, > > > > > > Is there any documentation other than DUI0922A? [1] From what I can seen > > > that just says it's a 16-bit value and doesn't put any particular > > > constraints on its value. > > > > Although not explicitly stated, if you look at the Get Sensor Capability > > [2] and Get Sensor Info [3] commands you can indirectly infer that the > > Sensor IDs are contiguous. > > I personally wouldn't even indirectly infer they are contiguous from > what the document says. If I were implementing the firmware I would feel > quite in my rights to, for example, use the top 8 bits of the ID for a > sensor type and the bottom 8 for an index, if that made dispatching of > requests more efficient. Or if some optional hardware was detected as > missing, leaving some holes in ID space. > > As a specification of a 'standard' the document seems to be rather > lacking. So, Sensor ID should be documented as being "an unsigned > integer less than then number of sensors returned by the Get Sensor > Capability command", or something like that. I guess clocks and other > devices suffer from similar lack of specificity. I think from the PoV of the binding, this doesn't matter. The value is just an arbitrary opaue token written down in a spec, that the FW understands how to interpret. I only asked about how the IDs were organised because I thought there was additional translation in the kernel, but this is not the case. The only potential problem is bit-width. Punit, I assume the value is 32-bit (or less) in the messages to the FW? If for the moment we assume the values can index a small array, then we can change that later (though ideally we'd handle the most geeneral case we're aware of from the outset). We just need to not infer any semantics from a value alone. Thanks, Mark.
On Tue, 2015-09-15 at 12:03 +0100, Mark Rutland wrote: > On Tue, Sep 15, 2015 at 11:46:02AM +0100, Jon Medhurst (Tixy) wrote: > > On Tue, 2015-09-15 at 10:37 +0100, Punit Agrawal wrote: > > > "Jon Medhurst (Tixy)" <tixy@linaro.org> writes: > > > > > > > On Mon, 2015-09-14 at 15:38 +0100, Punit Agrawal wrote: > > > >> Mark Rutland <mark.rutland@arm.com> writes: > > > >> > > [...] > > > >> The way the SCP interface is defined, the sensor identifiers are > > > >> contiguous, > > > > > > > > Is there any documentation other than DUI0922A? [1] From what I can seen > > > > that just says it's a 16-bit value and doesn't put any particular > > > > constraints on its value. > > > > > > Although not explicitly stated, if you look at the Get Sensor Capability > > > [2] and Get Sensor Info [3] commands you can indirectly infer that the > > > Sensor IDs are contiguous. > > > > I personally wouldn't even indirectly infer they are contiguous from > > what the document says. If I were implementing the firmware I would feel > > quite in my rights to, for example, use the top 8 bits of the ID for a > > sensor type and the bottom 8 for an index, if that made dispatching of > > requests more efficient. Or if some optional hardware was detected as > > missing, leaving some holes in ID space. > > > > As a specification of a 'standard' the document seems to be rather > > lacking. So, Sensor ID should be documented as being "an unsigned > > integer less than then number of sensors returned by the Get Sensor > > Capability command", or something like that. I guess clocks and other > > devices suffer from similar lack of specificity. > > I think from the PoV of the binding, this doesn't matter. The value is > just an arbitrary opaue token written down in a spec, that the FW > understands how to interpret. True, it's the Linux implementation in following patches that has assumptions, e.g. for loops iterating over id's 0..N-1 > > I only asked about how the IDs were organised because I thought there > was additional translation in the kernel, but this is not the case. > > The only potential problem is bit-width. Punit, I assume the value is > 32-bit (or less) in the messages to the FW? It's 16 bit.
"Jon Medhurst (Tixy)" <tixy@linaro.org> writes: > On Tue, 2015-09-15 at 10:37 +0100, Punit Agrawal wrote: >> "Jon Medhurst (Tixy)" <tixy@linaro.org> writes: >> >> > On Mon, 2015-09-14 at 15:38 +0100, Punit Agrawal wrote: >> >> Mark Rutland <mark.rutland@arm.com> writes: >> >> > [...] >> >> The way the SCP interface is defined, the sensor identifiers are >> >> contiguous, >> > >> > Is there any documentation other than DUI0922A? [1] From what I can seen >> > that just says it's a 16-bit value and doesn't put any particular >> > constraints on its value. >> >> Although not explicitly stated, if you look at the Get Sensor Capability >> [2] and Get Sensor Info [3] commands you can indirectly infer that the >> Sensor IDs are contiguous. > > I personally wouldn't even indirectly infer they are contiguous from > what the document says. If I were implementing the firmware I would feel > quite in my rights to, for example, use the top 8 bits of the ID for a > sensor type and the bottom 8 for an index, if that made dispatching of > requests more efficient. Or if some optional hardware was detected as > missing, leaving some holes in ID space. True. And without a command to convey the list of valid IDs, the consumer of the API would have to iterate over the entire 16bit space to locate valid IDs. > > As a specification of a 'standard' the document seems to be rather > lacking. So, Sensor ID should be documented as being "an unsigned > integer less than then number of sensors returned by the Get Sensor > Capability command", or something like that. I guess clocks and other > devices suffer from similar lack of specificity. I was thinking of suggesting something similar as an update. > >> Not the strongest guarantee I know. >> >> All platforms currently using SCP (Juno R0 and R1) do indeed expose >> contiguous identifiers. > > IMO, Linux drivers should be coded to the standard or written > specification (where they are available) not the particular > implementations available. > >> > >> > [1] http://community.arm.com/servlet/JiveServlet/download/8401-40-18262/DUI0922A_scp_message_interface.pdf >> [2] http://arminfo.emea.arm.com/help/index.jsp?topic=/com.arm.doc.dui0922b/ch03s02s21.html >> [3] http://arminfo.emea.arm.com/help/index.jsp?topic=/com.arm.doc.dui0922b/BABCCCJJ.html > > I think those links are on ARM's intranet, they return NXDOMAIN for me. Apologies, the below should work. [2] http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0922b/ch03s02s21.html [3] http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0922b/BABCCCJJ.html
On Tue, 2015-09-15 at 17:04 +0100, Punit Agrawal wrote: > "Jon Medhurst (Tixy)" <tixy@linaro.org> writes: > > > On Tue, 2015-09-15 at 10:37 +0100, Punit Agrawal wrote: > >> "Jon Medhurst (Tixy)" <tixy@linaro.org> writes: > >> > >> > On Mon, 2015-09-14 at 15:38 +0100, Punit Agrawal wrote: > >> >> Mark Rutland <mark.rutland@arm.com> writes: > >> >> > > [...] > >> >> The way the SCP interface is defined, the sensor identifiers are > >> >> contiguous, > >> > > >> > Is there any documentation other than DUI0922A? [1] From what I can seen > >> > that just says it's a 16-bit value and doesn't put any particular > >> > constraints on its value. > >> > >> Although not explicitly stated, if you look at the Get Sensor Capability > >> [2] and Get Sensor Info [3] commands you can indirectly infer that the > >> Sensor IDs are contiguous. > > > > I personally wouldn't even indirectly infer they are contiguous from > > what the document says. If I were implementing the firmware I would feel > > quite in my rights to, for example, use the top 8 bits of the ID for a > > sensor type and the bottom 8 for an index, if that made dispatching of > > requests more efficient. Or if some optional hardware was detected as > > missing, leaving some holes in ID space. > > True. And without a command to convey the list of valid IDs, the > consumer of the API would have to iterate over the entire 16bit space to > locate valid IDs. Or get IDs from device-tree :-) Anyway, I'm not arguing that the IDs shouldn't be 0..N-1, just that it should explicitly documented in the SCPI doc, which we're are in agreement on.
diff --git a/Documentation/devicetree/bindings/arm/arm,scpi.txt b/Documentation/devicetree/bindings/arm/arm,scpi.txt index f002460..af66f7f 100644 --- a/Documentation/devicetree/bindings/arm/arm,scpi.txt +++ b/Documentation/devicetree/bindings/arm/arm,scpi.txt @@ -72,8 +72,18 @@ Required sub-node properties: - compatible : should be "arm,juno-scp-shmem" for Non-secure SRAM based shared memory on Juno platforms +Sensor bindings for the sensors based on SCPI Message Protocol +-------------------------------------------------------------- +SCPI provides an API to access the various sensors on the SoC. + +Required properties: +- compatible : should be "arm,scpi-sensors". +- #thermal-sensor-cells: should be set to 1. This property follows the + thermal device tree bindings[2]. + [0] http://infocenter.arm.com/help/topic/com.arm.doc.dui0922b/index.html [1] Documentation/devicetree/bindings/clock/clock-bindings.txt +[2] Documentation/devicetree/bindings/thermal/thermal.txt Example: @@ -122,6 +132,11 @@ scpi_protocol: scpi@2e000000 { clock-output-names = "pxlclk0", "pxlclk1"; }; }; + + scpi_sensors0: sensors { + compatible = "arm,scpi-sensors"; + #thermal-sensor-cells = <1>; + }; }; cpu@0 {
The System Control Processor (SCP) provides access to SoC sensors via the System Control and Power Interface (SCPI) Message Protocol. Add bindings to allow probing of these sensors. Also support referencing of the sensors for setting up thermal zones via the thermal DT bindings. Signed-off-by: Punit Agrawal <punit.agrawal@arm.com> Cc: Rob Herring <robh+dt@kernel.org> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Sudeep Holla <sudeep.holla@arm.com> --- Documentation/devicetree/bindings/arm/arm,scpi.txt | 15 +++++++++++++++ 1 file changed, 15 insertions(+)