| Message ID | 20240621112303.1607621-9-claudiu.beznea.uj@bp.renesas.com (mailing list archive) |
|---|---|
| State | Superseded |
| Delegated to: | Geert Uytterhoeven |
| Headers | show |
| Series | i2c: riic: Add support for Renesas RZ/G3S | expand |
Hi Claudiu, > -----Original Message----- > From: Claudiu <claudiu.beznea@tuxon.dev> > Sent: Friday, June 21, 2024 12:23 PM > Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support > > From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > > Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is compatible with the version available on > Renesas RZ/V2H (R9A09G075). Most of the IP variants that the RIIC driver is working with supports > fast mode plus. > However, it happens that on the same SoC to have IP instatiations that support fast mode plus as > well as IP instantiation that doesn't support it. For this, introduced the renesas,riic-no-fast- > mode-plus property. > > Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > --- > Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 ++++++++ > 1 file changed, 8 insertions(+) > > diff --git a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > index 91ecf17b7a81..c0964edbca69 100644 > --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > @@ -25,6 +25,10 @@ properties: > - renesas,riic-r9a07g054 # RZ/V2L > - const: renesas,riic-rz # RZ/A or RZ/G2L > > + - items: > + - const: renesas,riic-r9a08g045 # RZ/G3S > + - const: renesas,riic-r9a09g057 > + > - const: renesas,riic-r9a09g057 # RZ/V2H(P) > > reg: > @@ -66,6 +70,10 @@ properties: > resets: > maxItems: 1 > > + renesas,riic-no-fast-mode-plus: > + description: specifies if fast mode plus is not supported > + type: Boolean Can't this info, as part of device data?? Based on frequency and device data is enough to derive this info?? Cheers, Biju > + > required: > - compatible > - reg > -- > 2.39.2 >
On 21.06.2024 15:34, Biju Das wrote: > Hi Claudiu, > >> -----Original Message----- >> From: Claudiu <claudiu.beznea@tuxon.dev> >> Sent: Friday, June 21, 2024 12:23 PM >> Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support >> >> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> >> >> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is compatible with the version available on >> Renesas RZ/V2H (R9A09G075). Most of the IP variants that the RIIC driver is working with supports >> fast mode plus. >> However, it happens that on the same SoC to have IP instatiations that support fast mode plus as >> well as IP instantiation that doesn't support it. For this, introduced the renesas,riic-no-fast- >> mode-plus property. >> >> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> >> --- >> Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 ++++++++ >> 1 file changed, 8 insertions(+) >> >> diff --git a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >> b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >> index 91ecf17b7a81..c0964edbca69 100644 >> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >> @@ -25,6 +25,10 @@ properties: >> - renesas,riic-r9a07g054 # RZ/V2L >> - const: renesas,riic-rz # RZ/A or RZ/G2L >> >> + - items: >> + - const: renesas,riic-r9a08g045 # RZ/G3S >> + - const: renesas,riic-r9a09g057 >> + >> - const: renesas,riic-r9a09g057 # RZ/V2H(P) >> >> reg: >> @@ -66,6 +70,10 @@ properties: >> resets: >> maxItems: 1 >> >> + renesas,riic-no-fast-mode-plus: >> + description: specifies if fast mode plus is not supported >> + type: Boolean > > Can't this info, as part of device data?? Based on frequency and device data > is enough to derive this info?? We can't rely completely on device data because on RZ/G3S we have 2 RIIC channels that support fast mode plus and 2 that doesn't support it. Thank you, Claudiu Beznea > > Cheers, > Biju > >> + >> required: >> - compatible >> - reg >> -- >> 2.39.2 >> >
Hi claudiu, > -----Original Message----- > From: claudiu beznea <claudiu.beznea@tuxon.dev> > Sent: Friday, June 21, 2024 1:55 PM > Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support > > > > On 21.06.2024 15:34, Biju Das wrote: > > Hi Claudiu, > > > >> -----Original Message----- > >> From: Claudiu <claudiu.beznea@tuxon.dev> > >> Sent: Friday, June 21, 2024 12:23 PM > >> Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the > >> R9A08G045 support > >> > >> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > >> > >> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is compatible > >> with the version available on Renesas RZ/V2H (R9A09G075). Most of the > >> IP variants that the RIIC driver is working with supports fast mode plus. > >> However, it happens that on the same SoC to have IP instatiations > >> that support fast mode plus as well as IP instantiation that doesn't > >> support it. For this, introduced the renesas,riic-no-fast- mode-plus property. > >> > >> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > >> --- > >> Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 ++++++++ > >> 1 file changed, 8 insertions(+) > >> > >> diff --git a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >> b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >> index 91ecf17b7a81..c0964edbca69 100644 > >> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >> @@ -25,6 +25,10 @@ properties: > >> - renesas,riic-r9a07g054 # RZ/V2L > >> - const: renesas,riic-rz # RZ/A or RZ/G2L > >> > >> + - items: > >> + - const: renesas,riic-r9a08g045 # RZ/G3S > >> + - const: renesas,riic-r9a09g057 > >> + > >> - const: renesas,riic-r9a09g057 # RZ/V2H(P) > >> > >> reg: > >> @@ -66,6 +70,10 @@ properties: > >> resets: > >> maxItems: 1 > >> > >> + renesas,riic-no-fast-mode-plus: > >> + description: specifies if fast mode plus is not supported > >> + type: Boolean > > > > Can't this info, as part of device data?? Based on frequency and > > device data is enough to derive this info?? > > We can't rely completely on device data because on RZ/G3S we have 2 RIIC channels that support fast > mode plus and 2 that doesn't support it. Can't array of bits for this channels won't help?? Based on the bit info, you have that details and check against frequency defined in DT. Cheers, Biju
On 21.06.2024 15:56, Biju Das wrote: > > Hi claudiu, > >> -----Original Message----- >> From: claudiu beznea <claudiu.beznea@tuxon.dev> >> Sent: Friday, June 21, 2024 1:55 PM >> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support >> >> >> >> On 21.06.2024 15:34, Biju Das wrote: >>> Hi Claudiu, >>> >>>> -----Original Message----- >>>> From: Claudiu <claudiu.beznea@tuxon.dev> >>>> Sent: Friday, June 21, 2024 12:23 PM >>>> Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the >>>> R9A08G045 support >>>> >>>> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> >>>> >>>> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is compatible >>>> with the version available on Renesas RZ/V2H (R9A09G075). Most of the >>>> IP variants that the RIIC driver is working with supports fast mode plus. >>>> However, it happens that on the same SoC to have IP instatiations >>>> that support fast mode plus as well as IP instantiation that doesn't >>>> support it. For this, introduced the renesas,riic-no-fast- mode-plus property. >>>> >>>> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> >>>> --- >>>> Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 ++++++++ >>>> 1 file changed, 8 insertions(+) >>>> >>>> diff --git a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>> b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>> index 91ecf17b7a81..c0964edbca69 100644 >>>> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>> @@ -25,6 +25,10 @@ properties: >>>> - renesas,riic-r9a07g054 # RZ/V2L >>>> - const: renesas,riic-rz # RZ/A or RZ/G2L >>>> >>>> + - items: >>>> + - const: renesas,riic-r9a08g045 # RZ/G3S >>>> + - const: renesas,riic-r9a09g057 >>>> + >>>> - const: renesas,riic-r9a09g057 # RZ/V2H(P) >>>> >>>> reg: >>>> @@ -66,6 +70,10 @@ properties: >>>> resets: >>>> maxItems: 1 >>>> >>>> + renesas,riic-no-fast-mode-plus: >>>> + description: specifies if fast mode plus is not supported >>>> + type: Boolean >>> >>> Can't this info, as part of device data?? Based on frequency and >>> device data is enough to derive this info?? >> >> We can't rely completely on device data because on RZ/G3S we have 2 RIIC channels that support fast >> mode plus and 2 that doesn't support it. > > Can't array of bits for this channels won't help?? Can you give an example? I'm not sure I understand how you would prefer me to use the array of bits. Thank you, Claudiu Beznea > Based on the bit info, you have that details and check against frequency defined in DT. > > Cheers, > Biju
> -----Original Message----- > From: claudiu beznea <claudiu.beznea@tuxon.dev> > Sent: Friday, June 21, 2024 2:06 PM Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support > > > > On 21.06.2024 15:56, Biju Das wrote: > > > > Hi claudiu, > > > >> -----Original Message----- > >> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >> Sent: Friday, June 21, 2024 1:55 PM > >> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document > >> the R9A08G045 support > >> > >> > >> > >> On 21.06.2024 15:34, Biju Das wrote: > >>> Hi Claudiu, > >>> > >>>> -----Original Message----- > >>>> From: Claudiu <claudiu.beznea@tuxon.dev> > >>>> Sent: Friday, June 21, 2024 12:23 PM > >>>> Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the > >>>> R9A08G045 support > >>>> > >>>> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > >>>> > >>>> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is compatible > >>>> with the version available on Renesas RZ/V2H (R9A09G075). Most of > >>>> the IP variants that the RIIC driver is working with supports fast mode plus. > >>>> However, it happens that on the same SoC to have IP instatiations > >>>> that support fast mode plus as well as IP instantiation that > >>>> doesn't support it. For this, introduced the renesas,riic-no-fast- mode-plus property. > >>>> > >>>> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > >>>> --- > >>>> Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 > >>>> ++++++++ > >>>> 1 file changed, 8 insertions(+) > >>>> > >>>> diff --git > >>>> a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>> b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>> index 91ecf17b7a81..c0964edbca69 100644 > >>>> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>> @@ -25,6 +25,10 @@ properties: > >>>> - renesas,riic-r9a07g054 # RZ/V2L > >>>> - const: renesas,riic-rz # RZ/A or RZ/G2L > >>>> > >>>> + - items: > >>>> + - const: renesas,riic-r9a08g045 # RZ/G3S > >>>> + - const: renesas,riic-r9a09g057 > >>>> + > >>>> - const: renesas,riic-r9a09g057 # RZ/V2H(P) > >>>> > >>>> reg: > >>>> @@ -66,6 +70,10 @@ properties: > >>>> resets: > >>>> maxItems: 1 > >>>> > >>>> + renesas,riic-no-fast-mode-plus: > >>>> + description: specifies if fast mode plus is not supported > >>>> + type: Boolean > >>> > >>> Can't this info, as part of device data?? Based on frequency and > >>> device data is enough to derive this info?? > >> > >> We can't rely completely on device data because on RZ/G3S we have 2 > >> RIIC channels that support fast mode plus and 2 that doesn't support it. > > > > Can't array of bits for this channels won't help?? > > Can you give an example? I'm not sure I understand how you would prefer me to use the array of > bits. struct riic_of_data { u8 regs[RIIC_REG_END]; u16 fast_mode_info info; /* 1 means fast mode plus supported, starting with channel 0*/ }; .info = 0x3, means channel 0 and 1 has fast mode plus supported .info = 0x0, none of the channel supported fast mode plus. Cheers, Biju
On 21.06.2024 16:10, Biju Das wrote: > > >> -----Original Message----- >> From: claudiu beznea <claudiu.beznea@tuxon.dev> >> Sent: Friday, June 21, 2024 2:06 PM > Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support >> >> >> >> On 21.06.2024 15:56, Biju Das wrote: >>> >>> Hi claudiu, >>> >>>> -----Original Message----- >>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> >>>> Sent: Friday, June 21, 2024 1:55 PM >>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document >>>> the R9A08G045 support >>>> >>>> >>>> >>>> On 21.06.2024 15:34, Biju Das wrote: >>>>> Hi Claudiu, >>>>> >>>>>> -----Original Message----- >>>>>> From: Claudiu <claudiu.beznea@tuxon.dev> >>>>>> Sent: Friday, June 21, 2024 12:23 PM >>>>>> Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the >>>>>> R9A08G045 support >>>>>> >>>>>> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> >>>>>> >>>>>> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is compatible >>>>>> with the version available on Renesas RZ/V2H (R9A09G075). Most of >>>>>> the IP variants that the RIIC driver is working with supports fast mode plus. >>>>>> However, it happens that on the same SoC to have IP instatiations >>>>>> that support fast mode plus as well as IP instantiation that >>>>>> doesn't support it. For this, introduced the renesas,riic-no-fast- mode-plus property. >>>>>> >>>>>> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> >>>>>> --- >>>>>> Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 >>>>>> ++++++++ >>>>>> 1 file changed, 8 insertions(+) >>>>>> >>>>>> diff --git >>>>>> a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>> b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>> index 91ecf17b7a81..c0964edbca69 100644 >>>>>> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>> @@ -25,6 +25,10 @@ properties: >>>>>> - renesas,riic-r9a07g054 # RZ/V2L >>>>>> - const: renesas,riic-rz # RZ/A or RZ/G2L >>>>>> >>>>>> + - items: >>>>>> + - const: renesas,riic-r9a08g045 # RZ/G3S >>>>>> + - const: renesas,riic-r9a09g057 >>>>>> + >>>>>> - const: renesas,riic-r9a09g057 # RZ/V2H(P) >>>>>> >>>>>> reg: >>>>>> @@ -66,6 +70,10 @@ properties: >>>>>> resets: >>>>>> maxItems: 1 >>>>>> >>>>>> + renesas,riic-no-fast-mode-plus: >>>>>> + description: specifies if fast mode plus is not supported >>>>>> + type: Boolean >>>>> >>>>> Can't this info, as part of device data?? Based on frequency and >>>>> device data is enough to derive this info?? >>>> >>>> We can't rely completely on device data because on RZ/G3S we have 2 >>>> RIIC channels that support fast mode plus and 2 that doesn't support it. >>> >>> Can't array of bits for this channels won't help?? >> >> Can you give an example? I'm not sure I understand how you would prefer me to use the array of >> bits. > > struct riic_of_data { > u8 regs[RIIC_REG_END]; > u16 fast_mode_info info; /* 1 means fast mode plus supported, starting with channel 0*/ > }; > > .info = 0x3, means channel 0 and 1 has fast mode plus supported > .info = 0x0, none of the channel supported fast mode plus. If I understand the proposal correctly, a match b/w struct riic_of_data::info bit + frequency and the nodes in device tree is still needed, right? As the RZ/G3S RIIC channels are using the same compatible. W/o a match how I cannot detect in the driver who is, e.g., channel 1 that supports FMP w/o hardcoding some RIIC channel data in the driver (e.g. RIIC channel address)? Also, for future SoCs that will suffer the same symptom but for different channels (and channels with different addresses) the driver will have to be adapted to match b/w the channel bit in struct riic_of_data::info and channel node from DT. > > Cheers, > Biju
Hi Claudiu, > -----Original Message----- > From: claudiu beznea <claudiu.beznea@tuxon.dev> > Sent: Friday, June 21, 2024 2:30 PM > Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support > > > > On 21.06.2024 16:10, Biju Das wrote: > > > > > >> -----Original Message----- > >> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >> Sent: Friday, June 21, 2024 2:06 PM > > Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document > > the R9A08G045 support > >> > >> > >> > >> On 21.06.2024 15:56, Biju Das wrote: > >>> > >>> Hi claudiu, > >>> > >>>> -----Original Message----- > >>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >>>> Sent: Friday, June 21, 2024 1:55 PM > >>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document > >>>> the R9A08G045 support > >>>> > >>>> > >>>> > >>>> On 21.06.2024 15:34, Biju Das wrote: > >>>>> Hi Claudiu, > >>>>> > >>>>>> -----Original Message----- > >>>>>> From: Claudiu <claudiu.beznea@tuxon.dev> > >>>>>> Sent: Friday, June 21, 2024 12:23 PM > >>>>>> Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document > >>>>>> the > >>>>>> R9A08G045 support > >>>>>> > >>>>>> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > >>>>>> > >>>>>> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is > >>>>>> compatible with the version available on Renesas RZ/V2H > >>>>>> (R9A09G075). Most of the IP variants that the RIIC driver is working with supports fast mode > plus. > >>>>>> However, it happens that on the same SoC to have IP instatiations > >>>>>> that support fast mode plus as well as IP instantiation that > >>>>>> doesn't support it. For this, introduced the renesas,riic-no-fast- mode-plus property. > >>>>>> > >>>>>> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > >>>>>> --- > >>>>>> Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 > >>>>>> ++++++++ > >>>>>> 1 file changed, 8 insertions(+) > >>>>>> > >>>>>> diff --git > >>>>>> a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>> b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>> index 91ecf17b7a81..c0964edbca69 100644 > >>>>>> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>> @@ -25,6 +25,10 @@ properties: > >>>>>> - renesas,riic-r9a07g054 # RZ/V2L > >>>>>> - const: renesas,riic-rz # RZ/A or RZ/G2L > >>>>>> > >>>>>> + - items: > >>>>>> + - const: renesas,riic-r9a08g045 # RZ/G3S > >>>>>> + - const: renesas,riic-r9a09g057 > >>>>>> + > >>>>>> - const: renesas,riic-r9a09g057 # RZ/V2H(P) > >>>>>> > >>>>>> reg: > >>>>>> @@ -66,6 +70,10 @@ properties: > >>>>>> resets: > >>>>>> maxItems: 1 > >>>>>> > >>>>>> + renesas,riic-no-fast-mode-plus: > >>>>>> + description: specifies if fast mode plus is not supported > >>>>>> + type: Boolean > >>>>> > >>>>> Can't this info, as part of device data?? Based on frequency and > >>>>> device data is enough to derive this info?? > >>>> > >>>> We can't rely completely on device data because on RZ/G3S we have 2 > >>>> RIIC channels that support fast mode plus and 2 that doesn't support it. > >>> > >>> Can't array of bits for this channels won't help?? > >> > >> Can you give an example? I'm not sure I understand how you would > >> prefer me to use the array of bits. > > > > struct riic_of_data { > > u8 regs[RIIC_REG_END]; > > u16 fast_mode_info info; /* 1 means fast mode plus supported, > > starting with channel 0*/ }; > > > > .info = 0x3, means channel 0 and 1 has fast mode plus supported .info > > = 0x0, none of the channel supported fast mode plus. > > If I understand the proposal correctly, a match b/w struct riic_of_data::info bit + frequency and > the nodes in device tree is still needed, right? As the RZ/G3S RIIC channels are using the same > compatible. > W/o a match how I cannot detect in the driver who is, e.g., channel 1 that supports FMP w/o > hardcoding some RIIC channel data in the driver (e.g. RIIC channel address)? bit array gives the capability info on various channels. If someone define fast_mode_plus frequency in DT node and channel is not fast_mode_plus(from the capability info) then you should return error. Here you need to use SoC specific compatible as each SoC has different capabilities. Cheers, Biju > > Also, for future SoCs that will suffer the same symptom but for different channels (and channels > with different addresses) the driver will have to be adapted to match b/w the channel bit in struct > riic_of_data::info and channel node from DT. > > > > > Cheers, > > Biju
On Fri, Jun 21, 2024 at 02:22:59PM +0300, Claudiu wrote: > From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > > Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is compatible with > the version available on Renesas RZ/V2H (R9A09G075). Most of the IP > variants that the RIIC driver is working with supports fast mode plus. > However, it happens that on the same SoC to have IP instatiations that > support fast mode plus as well as IP instantiation that doesn't support > it. For this, introduced the renesas,riic-no-fast-mode-plus property. > > Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > --- > Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 ++++++++ > 1 file changed, 8 insertions(+) > > diff --git a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > index 91ecf17b7a81..c0964edbca69 100644 > --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > @@ -25,6 +25,10 @@ properties: > - renesas,riic-r9a07g054 # RZ/V2L > - const: renesas,riic-rz # RZ/A or RZ/G2L > > + - items: > + - const: renesas,riic-r9a08g045 # RZ/G3S > + - const: renesas,riic-r9a09g057 > + > - const: renesas,riic-r9a09g057 # RZ/V2H(P) > > reg: > @@ -66,6 +70,10 @@ properties: > resets: > maxItems: 1 > > + renesas,riic-no-fast-mode-plus: > + description: specifies if fast mode plus is not supported > + type: boolean The correct type here is "flag", not boolean.
On 21.06.2024 17:06, Biju Das wrote: > Hi Claudiu, > >> -----Original Message----- >> From: claudiu beznea <claudiu.beznea@tuxon.dev> >> Sent: Friday, June 21, 2024 2:30 PM >> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support >> >> >> >> On 21.06.2024 16:10, Biju Das wrote: >>> >>> >>>> -----Original Message----- >>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> >>>> Sent: Friday, June 21, 2024 2:06 PM >>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document >>> the R9A08G045 support >>>> >>>> >>>> >>>> On 21.06.2024 15:56, Biju Das wrote: >>>>> >>>>> Hi claudiu, >>>>> >>>>>> -----Original Message----- >>>>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> >>>>>> Sent: Friday, June 21, 2024 1:55 PM >>>>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document >>>>>> the R9A08G045 support >>>>>> >>>>>> >>>>>> >>>>>> On 21.06.2024 15:34, Biju Das wrote: >>>>>>> Hi Claudiu, >>>>>>> >>>>>>>> -----Original Message----- >>>>>>>> From: Claudiu <claudiu.beznea@tuxon.dev> >>>>>>>> Sent: Friday, June 21, 2024 12:23 PM >>>>>>>> Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document >>>>>>>> the >>>>>>>> R9A08G045 support >>>>>>>> >>>>>>>> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> >>>>>>>> >>>>>>>> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is >>>>>>>> compatible with the version available on Renesas RZ/V2H >>>>>>>> (R9A09G075). Most of the IP variants that the RIIC driver is working with supports fast mode >> plus. >>>>>>>> However, it happens that on the same SoC to have IP instatiations >>>>>>>> that support fast mode plus as well as IP instantiation that >>>>>>>> doesn't support it. For this, introduced the renesas,riic-no-fast- mode-plus property. >>>>>>>> >>>>>>>> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> >>>>>>>> --- >>>>>>>> Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 >>>>>>>> ++++++++ >>>>>>>> 1 file changed, 8 insertions(+) >>>>>>>> >>>>>>>> diff --git >>>>>>>> a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>>>> b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>>>> index 91ecf17b7a81..c0964edbca69 100644 >>>>>>>> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>>>> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>>>> @@ -25,6 +25,10 @@ properties: >>>>>>>> - renesas,riic-r9a07g054 # RZ/V2L >>>>>>>> - const: renesas,riic-rz # RZ/A or RZ/G2L >>>>>>>> >>>>>>>> + - items: >>>>>>>> + - const: renesas,riic-r9a08g045 # RZ/G3S >>>>>>>> + - const: renesas,riic-r9a09g057 >>>>>>>> + >>>>>>>> - const: renesas,riic-r9a09g057 # RZ/V2H(P) >>>>>>>> >>>>>>>> reg: >>>>>>>> @@ -66,6 +70,10 @@ properties: >>>>>>>> resets: >>>>>>>> maxItems: 1 >>>>>>>> >>>>>>>> + renesas,riic-no-fast-mode-plus: >>>>>>>> + description: specifies if fast mode plus is not supported >>>>>>>> + type: Boolean >>>>>>> >>>>>>> Can't this info, as part of device data?? Based on frequency and >>>>>>> device data is enough to derive this info?? >>>>>> >>>>>> We can't rely completely on device data because on RZ/G3S we have 2 >>>>>> RIIC channels that support fast mode plus and 2 that doesn't support it. >>>>> >>>>> Can't array of bits for this channels won't help?? >>>> >>>> Can you give an example? I'm not sure I understand how you would >>>> prefer me to use the array of bits. >>> >>> struct riic_of_data { >>> u8 regs[RIIC_REG_END]; >>> u16 fast_mode_info info; /* 1 means fast mode plus supported, >>> starting with channel 0*/ }; >>> >>> .info = 0x3, means channel 0 and 1 has fast mode plus supported .info >>> = 0x0, none of the channel supported fast mode plus. >> >> If I understand the proposal correctly, a match b/w struct riic_of_data::info bit + frequency and >> the nodes in device tree is still needed, right? As the RZ/G3S RIIC channels are using the same >> compatible. >> W/o a match how I cannot detect in the driver who is, e.g., channel 1 that supports FMP w/o >> hardcoding some RIIC channel data in the driver (e.g. RIIC channel address)? > > bit array gives the capability info on various channels. > > If someone define fast_mode_plus frequency in DT node and channel is not fast_mode_plus(from the capability info) > then you should return error. > > Here you need to use SoC specific compatible as each SoC has different capabilities. And I would add, as it is in this case: there are multiple instantiation of the RIIC in RZ/G3S SoC. RIIC 0 and 1 supports FMP, RIIC 2 and 3 does not. For all RIICs (0, 1, 2, 3) we use the same compatible (as all are part of the same SoC). How to do the match b/w DT RIIC channel and driver with the solution you propose w/o hardcoding some RIIC channel data in the driver? > > Cheers, > Biju > > >> >> Also, for future SoCs that will suffer the same symptom but for different channels (and channels >> with different addresses) the driver will have to be adapted to match b/w the channel bit in struct >> riic_of_data::info and channel node from DT. >> >>> >>> Cheers, >>> Biju
> -----Original Message----- > From: claudiu beznea <claudiu.beznea@tuxon.dev> > Sent: Friday, June 21, 2024 3:31 PM > To: Biju Das <biju.das.jz@bp.renesas.com>; Chris Brandt <Chris.Brandt@renesas.com>; > andi.shyti@kernel.org; robh@kernel.org; krzk+dt@kernel.org; conor+dt@kernel.org; > geert+renesas@glider.be; magnus.damm@gmail.com; mturquette@baylibre.com; sboyd@kernel.org; > p.zabel@pengutronix.de; wsa+renesas@sang-engineering.com > Cc: linux-renesas-soc@vger.kernel.org; linux-i2c@vger.kernel.org; devicetree@vger.kernel.org; > linux-kernel@vger.kernel.org; linux-clk@vger.kernel.org; Claudiu Beznea > <claudiu.beznea.uj@bp.renesas.com> > Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support > > > > On 21.06.2024 17:06, Biju Das wrote: > > Hi Claudiu, > > > >> -----Original Message----- > >> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >> Sent: Friday, June 21, 2024 2:30 PM > >> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document > >> the R9A08G045 support > >> > >> > >> > >> On 21.06.2024 16:10, Biju Das wrote: > >>> > >>> > >>>> -----Original Message----- > >>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >>>> Sent: Friday, June 21, 2024 2:06 PM > >>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document > >>> the R9A08G045 support > >>>> > >>>> > >>>> > >>>> On 21.06.2024 15:56, Biju Das wrote: > >>>>> > >>>>> Hi claudiu, > >>>>> > >>>>>> -----Original Message----- > >>>>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >>>>>> Sent: Friday, June 21, 2024 1:55 PM > >>>>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: > >>>>>> Document the R9A08G045 support > >>>>>> > >>>>>> > >>>>>> > >>>>>> On 21.06.2024 15:34, Biju Das wrote: > >>>>>>> Hi Claudiu, > >>>>>>> > >>>>>>>> -----Original Message----- > >>>>>>>> From: Claudiu <claudiu.beznea@tuxon.dev> > >>>>>>>> Sent: Friday, June 21, 2024 12:23 PM > >>>>>>>> Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document > >>>>>>>> the > >>>>>>>> R9A08G045 support > >>>>>>>> > >>>>>>>> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > >>>>>>>> > >>>>>>>> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is > >>>>>>>> compatible with the version available on Renesas RZ/V2H > >>>>>>>> (R9A09G075). Most of the IP variants that the RIIC driver is > >>>>>>>> working with supports fast mode > >> plus. > >>>>>>>> However, it happens that on the same SoC to have IP > >>>>>>>> instatiations that support fast mode plus as well as IP > >>>>>>>> instantiation that doesn't support it. For this, introduced the renesas,riic-no-fast- > mode-plus property. > >>>>>>>> > >>>>>>>> Signed-off-by: Claudiu Beznea > >>>>>>>> <claudiu.beznea.uj@bp.renesas.com> > >>>>>>>> --- > >>>>>>>> Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 > >>>>>>>> ++++++++ > >>>>>>>> 1 file changed, 8 insertions(+) > >>>>>>>> > >>>>>>>> diff --git > >>>>>>>> a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>> b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>> index 91ecf17b7a81..c0964edbca69 100644 > >>>>>>>> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>> @@ -25,6 +25,10 @@ properties: > >>>>>>>> - renesas,riic-r9a07g054 # RZ/V2L > >>>>>>>> - const: renesas,riic-rz # RZ/A or RZ/G2L > >>>>>>>> > >>>>>>>> + - items: > >>>>>>>> + - const: renesas,riic-r9a08g045 # RZ/G3S > >>>>>>>> + - const: renesas,riic-r9a09g057 > >>>>>>>> + > >>>>>>>> - const: renesas,riic-r9a09g057 # RZ/V2H(P) > >>>>>>>> > >>>>>>>> reg: > >>>>>>>> @@ -66,6 +70,10 @@ properties: > >>>>>>>> resets: > >>>>>>>> maxItems: 1 > >>>>>>>> > >>>>>>>> + renesas,riic-no-fast-mode-plus: > >>>>>>>> + description: specifies if fast mode plus is not supported > >>>>>>>> + type: Boolean > >>>>>>> > >>>>>>> Can't this info, as part of device data?? Based on frequency and > >>>>>>> device data is enough to derive this info?? > >>>>>> > >>>>>> We can't rely completely on device data because on RZ/G3S we have > >>>>>> 2 RIIC channels that support fast mode plus and 2 that doesn't support it. > >>>>> > >>>>> Can't array of bits for this channels won't help?? > >>>> > >>>> Can you give an example? I'm not sure I understand how you would > >>>> prefer me to use the array of bits. > >>> > >>> struct riic_of_data { > >>> u8 regs[RIIC_REG_END]; > >>> u16 fast_mode_info info; /* 1 means fast mode plus supported, > >>> starting with channel 0*/ }; > >>> > >>> .info = 0x3, means channel 0 and 1 has fast mode plus supported > >>> .info = 0x0, none of the channel supported fast mode plus. > >> > >> If I understand the proposal correctly, a match b/w struct > >> riic_of_data::info bit + frequency and the nodes in device tree is > >> still needed, right? As the RZ/G3S RIIC channels are using the same compatible. > >> W/o a match how I cannot detect in the driver who is, e.g., channel 1 > >> that supports FMP w/o hardcoding some RIIC channel data in the driver (e.g. RIIC channel > address)? > > > > bit array gives the capability info on various channels. > > > > If someone define fast_mode_plus frequency in DT node and channel is > > not fast_mode_plus(from the capability info) then you should return error. > > > > Here you need to use SoC specific compatible as each SoC has different capabilities. > > And I would add, as it is in this case: there are multiple instantiation of the RIIC in RZ/G3S SoC. > RIIC 0 and 1 supports FMP, RIIC 2 and 3 does not. > > For all RIICs (0, 1, 2, 3) we use the same compatible (as all are part of the same SoC). How to do > the match b/w DT RIIC channel and driver with the solution you propose w/o hardcoding some RIIC > channel data in the driver? .info =0x3, so you know from the capability, for this soc, bus 0 and 1 supports FMP. Cheers, Biju > > > > > Cheers, > > Biju > > > > > >> > >> Also, for future SoCs that will suffer the same symptom but for > >> different channels (and channels with different addresses) the driver > >> will have to be adapted to match b/w the channel bit in struct riic_of_data::info and channel > node from DT. > >> > >>> > >>> Cheers, > >>> Biju
On 21.06.2024 17:37, Biju Das wrote: > > >> -----Original Message----- >> From: claudiu beznea <claudiu.beznea@tuxon.dev> >> Sent: Friday, June 21, 2024 3:31 PM >> To: Biju Das <biju.das.jz@bp.renesas.com>; Chris Brandt <Chris.Brandt@renesas.com>; >> andi.shyti@kernel.org; robh@kernel.org; krzk+dt@kernel.org; conor+dt@kernel.org; >> geert+renesas@glider.be; magnus.damm@gmail.com; mturquette@baylibre.com; sboyd@kernel.org; >> p.zabel@pengutronix.de; wsa+renesas@sang-engineering.com >> Cc: linux-renesas-soc@vger.kernel.org; linux-i2c@vger.kernel.org; devicetree@vger.kernel.org; >> linux-kernel@vger.kernel.org; linux-clk@vger.kernel.org; Claudiu Beznea >> <claudiu.beznea.uj@bp.renesas.com> >> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support >> >> >> >> On 21.06.2024 17:06, Biju Das wrote: >>> Hi Claudiu, >>> >>>> -----Original Message----- >>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> >>>> Sent: Friday, June 21, 2024 2:30 PM >>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document >>>> the R9A08G045 support >>>> >>>> >>>> >>>> On 21.06.2024 16:10, Biju Das wrote: >>>>> >>>>> >>>>>> -----Original Message----- >>>>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> >>>>>> Sent: Friday, June 21, 2024 2:06 PM >>>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document >>>>> the R9A08G045 support >>>>>> >>>>>> >>>>>> >>>>>> On 21.06.2024 15:56, Biju Das wrote: >>>>>>> >>>>>>> Hi claudiu, >>>>>>> >>>>>>>> -----Original Message----- >>>>>>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> >>>>>>>> Sent: Friday, June 21, 2024 1:55 PM >>>>>>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: >>>>>>>> Document the R9A08G045 support >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> On 21.06.2024 15:34, Biju Das wrote: >>>>>>>>> Hi Claudiu, >>>>>>>>> >>>>>>>>>> -----Original Message----- >>>>>>>>>> From: Claudiu <claudiu.beznea@tuxon.dev> >>>>>>>>>> Sent: Friday, June 21, 2024 12:23 PM >>>>>>>>>> Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document >>>>>>>>>> the >>>>>>>>>> R9A08G045 support >>>>>>>>>> >>>>>>>>>> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> >>>>>>>>>> >>>>>>>>>> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is >>>>>>>>>> compatible with the version available on Renesas RZ/V2H >>>>>>>>>> (R9A09G075). Most of the IP variants that the RIIC driver is >>>>>>>>>> working with supports fast mode >>>> plus. >>>>>>>>>> However, it happens that on the same SoC to have IP >>>>>>>>>> instatiations that support fast mode plus as well as IP >>>>>>>>>> instantiation that doesn't support it. For this, introduced the renesas,riic-no-fast- >> mode-plus property. >>>>>>>>>> >>>>>>>>>> Signed-off-by: Claudiu Beznea >>>>>>>>>> <claudiu.beznea.uj@bp.renesas.com> >>>>>>>>>> --- >>>>>>>>>> Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 >>>>>>>>>> ++++++++ >>>>>>>>>> 1 file changed, 8 insertions(+) >>>>>>>>>> >>>>>>>>>> diff --git >>>>>>>>>> a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>>>>>> b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>>>>>> index 91ecf17b7a81..c0964edbca69 100644 >>>>>>>>>> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>>>>>> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >>>>>>>>>> @@ -25,6 +25,10 @@ properties: >>>>>>>>>> - renesas,riic-r9a07g054 # RZ/V2L >>>>>>>>>> - const: renesas,riic-rz # RZ/A or RZ/G2L >>>>>>>>>> >>>>>>>>>> + - items: >>>>>>>>>> + - const: renesas,riic-r9a08g045 # RZ/G3S >>>>>>>>>> + - const: renesas,riic-r9a09g057 >>>>>>>>>> + >>>>>>>>>> - const: renesas,riic-r9a09g057 # RZ/V2H(P) >>>>>>>>>> >>>>>>>>>> reg: >>>>>>>>>> @@ -66,6 +70,10 @@ properties: >>>>>>>>>> resets: >>>>>>>>>> maxItems: 1 >>>>>>>>>> >>>>>>>>>> + renesas,riic-no-fast-mode-plus: >>>>>>>>>> + description: specifies if fast mode plus is not supported >>>>>>>>>> + type: Boolean >>>>>>>>> >>>>>>>>> Can't this info, as part of device data?? Based on frequency and >>>>>>>>> device data is enough to derive this info?? >>>>>>>> >>>>>>>> We can't rely completely on device data because on RZ/G3S we have >>>>>>>> 2 RIIC channels that support fast mode plus and 2 that doesn't support it. >>>>>>> >>>>>>> Can't array of bits for this channels won't help?? >>>>>> >>>>>> Can you give an example? I'm not sure I understand how you would >>>>>> prefer me to use the array of bits. >>>>> >>>>> struct riic_of_data { >>>>> u8 regs[RIIC_REG_END]; >>>>> u16 fast_mode_info info; /* 1 means fast mode plus supported, >>>>> starting with channel 0*/ }; >>>>> >>>>> .info = 0x3, means channel 0 and 1 has fast mode plus supported >>>>> .info = 0x0, none of the channel supported fast mode plus. >>>> >>>> If I understand the proposal correctly, a match b/w struct >>>> riic_of_data::info bit + frequency and the nodes in device tree is >>>> still needed, right? As the RZ/G3S RIIC channels are using the same compatible. >>>> W/o a match how I cannot detect in the driver who is, e.g., channel 1 >>>> that supports FMP w/o hardcoding some RIIC channel data in the driver (e.g. RIIC channel >> address)? >>> >>> bit array gives the capability info on various channels. >>> >>> If someone define fast_mode_plus frequency in DT node and channel is >>> not fast_mode_plus(from the capability info) then you should return error. >>> >>> Here you need to use SoC specific compatible as each SoC has different capabilities. >> >> And I would add, as it is in this case: there are multiple instantiation of the RIIC in RZ/G3S SoC. >> RIIC 0 and 1 supports FMP, RIIC 2 and 3 does not. >> >> For all RIICs (0, 1, 2, 3) we use the same compatible (as all are part of the same SoC). How to do >> the match b/w DT RIIC channel and driver with the solution you propose w/o hardcoding some RIIC >> channel data in the driver? > > .info =0x3, so you know from the capability, for this soc, bus 0 and 1 supports FMP. I understand this part. What I don't understand is: when probing the driver for, e.g., bus 0, how do I know I probe the driver for bus 0? compatible is the same for all buses. > > Cheers, > Biju > > >> >>> >>> Cheers, >>> Biju >>> >>> >>>> >>>> Also, for future SoCs that will suffer the same symptom but for >>>> different channels (and channels with different addresses) the driver >>>> will have to be adapted to match b/w the channel bit in struct riic_of_data::info and channel >> node from DT. >>>> >>>>> >>>>> Cheers, >>>>> Biju
> -----Original Message----- > From: claudiu beznea <claudiu.beznea@tuxon.dev> > Sent: Friday, June 21, 2024 3:59 PM > Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support > > > > On 21.06.2024 17:37, Biju Das wrote: > > > > > >> -----Original Message----- > >> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >> Sent: Friday, June 21, 2024 3:31 PM > >> To: Biju Das <biju.das.jz@bp.renesas.com>; Chris Brandt > >> <Chris.Brandt@renesas.com>; andi.shyti@kernel.org; robh@kernel.org; > >> krzk+dt@kernel.org; conor+dt@kernel.org; > >> geert+renesas@glider.be; magnus.damm@gmail.com; > >> geert+mturquette@baylibre.com; sboyd@kernel.org; > >> p.zabel@pengutronix.de; wsa+renesas@sang-engineering.com > >> Cc: linux-renesas-soc@vger.kernel.org; linux-i2c@vger.kernel.org; > >> devicetree@vger.kernel.org; linux-kernel@vger.kernel.org; > >> linux-clk@vger.kernel.org; Claudiu Beznea > >> <claudiu.beznea.uj@bp.renesas.com> > >> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document > >> the R9A08G045 support > >> > >> > >> > >> On 21.06.2024 17:06, Biju Das wrote: > >>> Hi Claudiu, > >>> > >>>> -----Original Message----- > >>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >>>> Sent: Friday, June 21, 2024 2:30 PM > >>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document > >>>> the R9A08G045 support > >>>> > >>>> > >>>> > >>>> On 21.06.2024 16:10, Biju Das wrote: > >>>>> > >>>>> > >>>>>> -----Original Message----- > >>>>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >>>>>> Sent: Friday, June 21, 2024 2:06 PM > >>>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: > >>>>> Document the R9A08G045 support > >>>>>> > >>>>>> > >>>>>> > >>>>>> On 21.06.2024 15:56, Biju Das wrote: > >>>>>>> > >>>>>>> Hi claudiu, > >>>>>>> > >>>>>>>> -----Original Message----- > >>>>>>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >>>>>>>> Sent: Friday, June 21, 2024 1:55 PM > >>>>>>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: > >>>>>>>> Document the R9A08G045 support > >>>>>>>> > >>>>>>>> > >>>>>>>> > >>>>>>>> On 21.06.2024 15:34, Biju Das wrote: > >>>>>>>>> Hi Claudiu, > >>>>>>>>> > >>>>>>>>>> -----Original Message----- > >>>>>>>>>> From: Claudiu <claudiu.beznea@tuxon.dev> > >>>>>>>>>> Sent: Friday, June 21, 2024 12:23 PM > >>>>>>>>>> Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: > >>>>>>>>>> Document the > >>>>>>>>>> R9A08G045 support > >>>>>>>>>> > >>>>>>>>>> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > >>>>>>>>>> > >>>>>>>>>> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is > >>>>>>>>>> compatible with the version available on Renesas RZ/V2H > >>>>>>>>>> (R9A09G075). Most of the IP variants that the RIIC driver is > >>>>>>>>>> working with supports fast mode > >>>> plus. > >>>>>>>>>> However, it happens that on the same SoC to have IP > >>>>>>>>>> instatiations that support fast mode plus as well as IP > >>>>>>>>>> instantiation that doesn't support it. For this, introduced > >>>>>>>>>> the renesas,riic-no-fast- > >> mode-plus property. > >>>>>>>>>> > >>>>>>>>>> Signed-off-by: Claudiu Beznea > >>>>>>>>>> <claudiu.beznea.uj@bp.renesas.com> > >>>>>>>>>> --- > >>>>>>>>>> Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 > >>>>>>>>>> ++++++++ > >>>>>>>>>> 1 file changed, 8 insertions(+) > >>>>>>>>>> > >>>>>>>>>> diff --git > >>>>>>>>>> a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>>>> b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>>>> index 91ecf17b7a81..c0964edbca69 100644 > >>>>>>>>>> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>>>> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>>>> @@ -25,6 +25,10 @@ properties: > >>>>>>>>>> - renesas,riic-r9a07g054 # RZ/V2L > >>>>>>>>>> - const: renesas,riic-rz # RZ/A or RZ/G2L > >>>>>>>>>> > >>>>>>>>>> + - items: > >>>>>>>>>> + - const: renesas,riic-r9a08g045 # RZ/G3S > >>>>>>>>>> + - const: renesas,riic-r9a09g057 > >>>>>>>>>> + > >>>>>>>>>> - const: renesas,riic-r9a09g057 # RZ/V2H(P) > >>>>>>>>>> > >>>>>>>>>> reg: > >>>>>>>>>> @@ -66,6 +70,10 @@ properties: > >>>>>>>>>> resets: > >>>>>>>>>> maxItems: 1 > >>>>>>>>>> > >>>>>>>>>> + renesas,riic-no-fast-mode-plus: > >>>>>>>>>> + description: specifies if fast mode plus is not supported > >>>>>>>>>> + type: Boolean > >>>>>>>>> > >>>>>>>>> Can't this info, as part of device data?? Based on frequency > >>>>>>>>> and device data is enough to derive this info?? > >>>>>>>> > >>>>>>>> We can't rely completely on device data because on RZ/G3S we > >>>>>>>> have > >>>>>>>> 2 RIIC channels that support fast mode plus and 2 that doesn't support it. > >>>>>>> > >>>>>>> Can't array of bits for this channels won't help?? > >>>>>> > >>>>>> Can you give an example? I'm not sure I understand how you would > >>>>>> prefer me to use the array of bits. > >>>>> > >>>>> struct riic_of_data { > >>>>> u8 regs[RIIC_REG_END]; > >>>>> u16 fast_mode_info info; /* 1 means fast mode plus supported, > >>>>> starting with channel 0*/ }; > >>>>> > >>>>> .info = 0x3, means channel 0 and 1 has fast mode plus supported > >>>>> .info = 0x0, none of the channel supported fast mode plus. > >>>> > >>>> If I understand the proposal correctly, a match b/w struct > >>>> riic_of_data::info bit + frequency and the nodes in device tree is > >>>> still needed, right? As the RZ/G3S RIIC channels are using the same compatible. > >>>> W/o a match how I cannot detect in the driver who is, e.g., channel > >>>> 1 that supports FMP w/o hardcoding some RIIC channel data in the > >>>> driver (e.g. RIIC channel > >> address)? > >>> > >>> bit array gives the capability info on various channels. > >>> > >>> If someone define fast_mode_plus frequency in DT node and channel is > >>> not fast_mode_plus(from the capability info) then you should return error. > >>> > >>> Here you need to use SoC specific compatible as each SoC has different capabilities. > >> > >> And I would add, as it is in this case: there are multiple instantiation of the RIIC in RZ/G3S > SoC. > >> RIIC 0 and 1 supports FMP, RIIC 2 and 3 does not. > >> > >> For all RIICs (0, 1, 2, 3) we use the same compatible (as all are > >> part of the same SoC). How to do the match b/w DT RIIC channel and > >> driver with the solution you propose w/o hardcoding some RIIC channel data in the driver? > > > > .info =0x3, so you know from the capability, for this soc, bus 0 and 1 supports FMP. > > I understand this part. What I don't understand is: when probing the driver for, e.g., bus 0, how > do I know I probe the driver for bus 0? compatible is the same for all buses. You are right. I missed this point. Cheers, biju
Hi Claudiu, > -----Original Message----- > From: claudiu beznea <claudiu.beznea@tuxon.dev> > Sent: Friday, June 21, 2024 3:59 PM > Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document the R9A08G045 support > > > > On 21.06.2024 17:37, Biju Das wrote: > > > > > >> -----Original Message----- > >> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >> Sent: Friday, June 21, 2024 3:31 PM > >> To: Biju Das <biju.das.jz@bp.renesas.com>; Chris Brandt > >> <Chris.Brandt@renesas.com>; andi.shyti@kernel.org; robh@kernel.org; > >> krzk+dt@kernel.org; conor+dt@kernel.org; > >> geert+renesas@glider.be; magnus.damm@gmail.com; > >> geert+mturquette@baylibre.com; sboyd@kernel.org; > >> p.zabel@pengutronix.de; wsa+renesas@sang-engineering.com > >> Cc: linux-renesas-soc@vger.kernel.org; linux-i2c@vger.kernel.org; > >> devicetree@vger.kernel.org; linux-kernel@vger.kernel.org; > >> linux-clk@vger.kernel.org; Claudiu Beznea > >> <claudiu.beznea.uj@bp.renesas.com> > >> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document > >> the R9A08G045 support > >> > >> > >> > >> On 21.06.2024 17:06, Biju Das wrote: > >>> Hi Claudiu, > >>> > >>>> -----Original Message----- > >>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >>>> Sent: Friday, June 21, 2024 2:30 PM > >>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: Document > >>>> the R9A08G045 support > >>>> > >>>> > >>>> > >>>> On 21.06.2024 16:10, Biju Das wrote: > >>>>> > >>>>> > >>>>>> -----Original Message----- > >>>>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >>>>>> Sent: Friday, June 21, 2024 2:06 PM > >>>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: > >>>>> Document the R9A08G045 support > >>>>>> > >>>>>> > >>>>>> > >>>>>> On 21.06.2024 15:56, Biju Das wrote: > >>>>>>> > >>>>>>> Hi claudiu, > >>>>>>> > >>>>>>>> -----Original Message----- > >>>>>>>> From: claudiu beznea <claudiu.beznea@tuxon.dev> > >>>>>>>> Sent: Friday, June 21, 2024 1:55 PM > >>>>>>>> Subject: Re: [PATCH 08/12] dt-bindings: i2c: renesas,riic: > >>>>>>>> Document the R9A08G045 support > >>>>>>>> > >>>>>>>> > >>>>>>>> > >>>>>>>> On 21.06.2024 15:34, Biju Das wrote: > >>>>>>>>> Hi Claudiu, > >>>>>>>>> > >>>>>>>>>> -----Original Message----- > >>>>>>>>>> From: Claudiu <claudiu.beznea@tuxon.dev> > >>>>>>>>>> Sent: Friday, June 21, 2024 12:23 PM > >>>>>>>>>> Subject: [PATCH 08/12] dt-bindings: i2c: renesas,riic: > >>>>>>>>>> Document the > >>>>>>>>>> R9A08G045 support > >>>>>>>>>> > >>>>>>>>>> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > >>>>>>>>>> > >>>>>>>>>> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is > >>>>>>>>>> compatible with the version available on Renesas RZ/V2H > >>>>>>>>>> (R9A09G075). Most of the IP variants that the RIIC driver is > >>>>>>>>>> working with supports fast mode > >>>> plus. > >>>>>>>>>> However, it happens that on the same SoC to have IP > >>>>>>>>>> instatiations that support fast mode plus as well as IP > >>>>>>>>>> instantiation that doesn't support it. For this, introduced > >>>>>>>>>> the renesas,riic-no-fast- > >> mode-plus property. > >>>>>>>>>> > >>>>>>>>>> Signed-off-by: Claudiu Beznea > >>>>>>>>>> <claudiu.beznea.uj@bp.renesas.com> > >>>>>>>>>> --- > >>>>>>>>>> Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 8 > >>>>>>>>>> ++++++++ > >>>>>>>>>> 1 file changed, 8 insertions(+) > >>>>>>>>>> > >>>>>>>>>> diff --git > >>>>>>>>>> a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>>>> b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>>>> index 91ecf17b7a81..c0964edbca69 100644 > >>>>>>>>>> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>>>> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > >>>>>>>>>> @@ -25,6 +25,10 @@ properties: > >>>>>>>>>> - renesas,riic-r9a07g054 # RZ/V2L > >>>>>>>>>> - const: renesas,riic-rz # RZ/A or RZ/G2L > >>>>>>>>>> > >>>>>>>>>> + - items: > >>>>>>>>>> + - const: renesas,riic-r9a08g045 # RZ/G3S > >>>>>>>>>> + - const: renesas,riic-r9a09g057 > >>>>>>>>>> + > >>>>>>>>>> - const: renesas,riic-r9a09g057 # RZ/V2H(P) > >>>>>>>>>> > >>>>>>>>>> reg: > >>>>>>>>>> @@ -66,6 +70,10 @@ properties: > >>>>>>>>>> resets: > >>>>>>>>>> maxItems: 1 > >>>>>>>>>> > >>>>>>>>>> + renesas,riic-no-fast-mode-plus: > >>>>>>>>>> + description: specifies if fast mode plus is not supported > >>>>>>>>>> + type: Boolean > >>>>>>>>> > >>>>>>>>> Can't this info, as part of device data?? Based on frequency > >>>>>>>>> and device data is enough to derive this info?? > >>>>>>>> > >>>>>>>> We can't rely completely on device data because on RZ/G3S we > >>>>>>>> have > >>>>>>>> 2 RIIC channels that support fast mode plus and 2 that doesn't support it. > >>>>>>> > >>>>>>> Can't array of bits for this channels won't help?? > >>>>>> > >>>>>> Can you give an example? I'm not sure I understand how you would > >>>>>> prefer me to use the array of bits. > >>>>> > >>>>> struct riic_of_data { > >>>>> u8 regs[RIIC_REG_END]; > >>>>> u16 fast_mode_info info; /* 1 means fast mode plus supported, > >>>>> starting with channel 0*/ }; > >>>>> > >>>>> .info = 0x3, means channel 0 and 1 has fast mode plus supported > >>>>> .info = 0x0, none of the channel supported fast mode plus. > >>>> > >>>> If I understand the proposal correctly, a match b/w struct > >>>> riic_of_data::info bit + frequency and the nodes in device tree is > >>>> still needed, right? As the RZ/G3S RIIC channels are using the same compatible. > >>>> W/o a match how I cannot detect in the driver who is, e.g., channel > >>>> 1 that supports FMP w/o hardcoding some RIIC channel data in the > >>>> driver (e.g. RIIC channel > >> address)? > >>> > >>> bit array gives the capability info on various channels. > >>> > >>> If someone define fast_mode_plus frequency in DT node and channel is > >>> not fast_mode_plus(from the capability info) then you should return error. > >>> > >>> Here you need to use SoC specific compatible as each SoC has different capabilities. > >> > >> And I would add, as it is in this case: there are multiple instantiation of the RIIC in RZ/G3S > SoC. > >> RIIC 0 and 1 supports FMP, RIIC 2 and 3 does not. > >> > >> For all RIICs (0, 1, 2, 3) we use the same compatible (as all are > >> part of the same SoC). How to do the match b/w DT RIIC channel and > >> driver with the solution you propose w/o hardcoding some RIIC channel data in the driver? > > > > .info =0x3, so you know from the capability, for this soc, bus 0 and 1 supports FMP. > > I understand this part. What I don't understand is: when probing the driver for, e.g., bus 0, how > do I know I probe the driver for bus 0? compatible is the same for all buses. I am not a DT expert. Another option is maybe a different compatible for "renesas,riic- r9a08g045-fmp" to describe buses with fmp and "renesas,riic- r9a08g045" for buses that don't support fmp and maybe you can validate at dt-binding level to check against frequencies?? If non-fmp bus, dt-binding check should check the frequency and if it is greater the fast mode frequency Then the check should fail, not sure the dt-binding check allows (x > y) comparison ?? Cheers, Biju
Hi Claudiu, On Fri, Jun 21, 2024 at 1:23 PM Claudiu <claudiu.beznea@tuxon.dev> wrote: > From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > > Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is compatible with > the version available on Renesas RZ/V2H (R9A09G075). Most of the IP > variants that the RIIC driver is working with supports fast mode plus. > However, it happens that on the same SoC to have IP instatiations that > support fast mode plus as well as IP instantiation that doesn't support > it. For this, introduced the renesas,riic-no-fast-mode-plus property. > > Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> Thanks for your patch! > --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml > @@ -25,6 +25,10 @@ properties: > - renesas,riic-r9a07g054 # RZ/V2L > - const: renesas,riic-rz # RZ/A or RZ/G2L > > + - items: > + - const: renesas,riic-r9a08g045 # RZ/G3S > + - const: renesas,riic-r9a09g057 > + LGTM. > - const: renesas,riic-r9a09g057 # RZ/V2H(P) > > reg: > @@ -66,6 +70,10 @@ properties: > resets: > maxItems: 1 > > + renesas,riic-no-fast-mode-plus: > + description: specifies if fast mode plus is not supported > + type: boolean > + Do you really need this? The bus' clock-frequency property should take into account the combined capabilities of all of controller, target, and wiring. It is up to the DTS writer to validate that all timing conditions are met. > required: > - compatible > - reg Gr{oetje,eeting}s, Geert
Hi, Geert, On 24.06.2024 18:40, Geert Uytterhoeven wrote: > Hi Claudiu, > > On Fri, Jun 21, 2024 at 1:23 PM Claudiu <claudiu.beznea@tuxon.dev> wrote: >> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> >> >> Document the Renesas RZ/G3S (R9A08G045) RIIC IP. This is compatible with >> the version available on Renesas RZ/V2H (R9A09G075). Most of the IP >> variants that the RIIC driver is working with supports fast mode plus. >> However, it happens that on the same SoC to have IP instatiations that >> support fast mode plus as well as IP instantiation that doesn't support >> it. For this, introduced the renesas,riic-no-fast-mode-plus property. >> >> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com> > > Thanks for your patch! > >> --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >> +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml >> @@ -25,6 +25,10 @@ properties: >> - renesas,riic-r9a07g054 # RZ/V2L >> - const: renesas,riic-rz # RZ/A or RZ/G2L >> >> + - items: >> + - const: renesas,riic-r9a08g045 # RZ/G3S >> + - const: renesas,riic-r9a09g057 >> + > > LGTM. > >> - const: renesas,riic-r9a09g057 # RZ/V2H(P) >> >> reg: >> @@ -66,6 +70,10 @@ properties: >> resets: >> maxItems: 1 >> >> + renesas,riic-no-fast-mode-plus: >> + description: specifies if fast mode plus is not supported >> + type: boolean >> + > > Do you really need this? > The bus' clock-frequency property should take into account the combined > capabilities of all of controller, target, and wiring. It is up to the > DTS writer to validate that all timing conditions are met. On a second thought, I tend to agree with you on this. I added it to comply with "chapter 47.5.15 I2C Bus Interface Access Timing, note 7, Tfmin cannot meet the specification in fast-mode plus for the RIIC ch2 and ch3" statement from hw manual. Thank you, Claudiu Beznea > >> required: >> - compatible >> - reg > > Gr{oetje,eeting}s, > > Geert >
diff --git a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml index 91ecf17b7a81..c0964edbca69 100644 --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml @@ -25,6 +25,10 @@ properties: - renesas,riic-r9a07g054 # RZ/V2L - const: renesas,riic-rz # RZ/A or RZ/G2L + - items: + - const: renesas,riic-r9a08g045 # RZ/G3S + - const: renesas,riic-r9a09g057 + - const: renesas,riic-r9a09g057 # RZ/V2H(P) reg: @@ -66,6 +70,10 @@ properties: resets: maxItems: 1 + renesas,riic-no-fast-mode-plus: + description: specifies if fast mode plus is not supported + type: boolean + required: - compatible - reg