| Message ID | 20240119100639.84029-2-changhuang.liang@starfivetech.com |
|---|---|
| State | New |
| Headers | show |
| Series | Add camera subsystem for StarFive | expand |
On Fri, Jan 19, 2024 at 02:06:38AM -0800, Changhuang Liang wrote: > Drop some unused properties for clocks, resets and interrupts for > StarFive JH7110 camera subsystem. What do you mean "unused"? Do these clocks etc exist but are not used by the driver? Or do they not exist at all? The two are very different! Thanks, Conor. > > Reviewed-by: Hal Feng <hal.feng@starfivetech.com> > Reviewed-by: Jack Zhu <jack.zhu@starfivetech.com> > Signed-off-by: Changhuang Liang <changhuang.liang@starfivetech.com> > --- > .../bindings/media/starfive,jh7110-camss.yaml | 31 +++++-------------- > 1 file changed, 8 insertions(+), 23 deletions(-) > > diff --git a/Documentation/devicetree/bindings/media/starfive,jh7110-camss.yaml b/Documentation/devicetree/bindings/media/starfive,jh7110-camss.yaml > index c66586d90fa2..2504381058b3 100644 > --- a/Documentation/devicetree/bindings/media/starfive,jh7110-camss.yaml > +++ b/Documentation/devicetree/bindings/media/starfive,jh7110-camss.yaml > @@ -4,14 +4,14 @@ > $id: http://devicetree.org/schemas/media/starfive,jh7110-camss.yaml# > $schema: http://devicetree.org/meta-schemas/core.yaml# > > -title: Starfive SoC CAMSS ISP > +title: StarFive SoC CAMSS ISP > > maintainers: > - Jack Zhu <jack.zhu@starfivetech.com> > - Changhuang Liang <changhuang.liang@starfivetech.com> > > description: > - The Starfive CAMSS ISP is a Camera interface for Starfive JH7110 SoC. It > + The StarFive CAMSS ISP is a Camera interface for StarFive JH7110 SoC. It > consists of a VIN controller (Video In Controller, a top-level control unit) > and an ISP. > > @@ -28,26 +28,21 @@ properties: > - const: isp > > clocks: > - maxItems: 7 > + maxItems: 3 > > clock-names: > items: > - - const: apb_func > - const: wrapper_clk_c > - - const: dvp_inv > - - const: axiwr > - - const: mipi_rx0_pxl > - const: ispcore_2x > - const: isp_axi > > resets: > - maxItems: 6 > + maxItems: 5 > > reset-names: > items: > - const: wrapper_p > - const: wrapper_c > - - const: axird > - const: axiwr > - const: isp_top_n > - const: isp_top_axi > @@ -57,7 +52,7 @@ properties: > - description: JH7110 ISP Power Domain Switch Controller. > > interrupts: > - maxItems: 4 > + maxItems: 3 > > ports: > $ref: /schemas/graph.yaml#/properties/ports > @@ -125,34 +120,24 @@ examples: > reg = <0x19840000 0x10000>, > <0x19870000 0x30000>; > reg-names = "syscon", "isp"; > - clocks = <&ispcrg 0>, > - <&ispcrg 13>, > - <&ispcrg 2>, > - <&ispcrg 12>, > - <&ispcrg 1>, > + clocks = <&ispcrg 13>, > <&syscrg 51>, > <&syscrg 52>; > - clock-names = "apb_func", > - "wrapper_clk_c", > - "dvp_inv", > - "axiwr", > - "mipi_rx0_pxl", > + clock-names = "wrapper_clk_c", > "ispcore_2x", > "isp_axi"; > resets = <&ispcrg 0>, > <&ispcrg 1>, > - <&ispcrg 10>, > <&ispcrg 11>, > <&syscrg 41>, > <&syscrg 42>; > reset-names = "wrapper_p", > "wrapper_c", > - "axird", > "axiwr", > "isp_top_n", > "isp_top_axi"; > power-domains = <&pwrc 5>; > - interrupts = <92>, <87>, <88>, <90>; > + interrupts = <92>, <87>, <90>; > > ports { > #address-cells = <1>; > -- > 2.25.1 > > > _______________________________________________ > linux-riscv mailing list > linux-riscv@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-riscv
On Fri, Jan 19, 2024 at 12:57:22PM +0000, Changhuang Liang wrote: > EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe > > Hi , Conor > > > 主题: Re: [PATCH v1 1/2] dt-bindings: media: starfive: Match driver and yaml > > property names > > > > On Fri, Jan 19, 2024 at 02:06:38AM -0800, Changhuang Liang wrote: > > > Drop some unused properties for clocks, resets and interrupts for > > > StarFive JH7110 camera subsystem. > > > > What do you mean "unused"? > > > > Do these clocks etc exist but are not used by the driver? > > > > Or do they not exist at all? > > > > The two are very different! > These clocks etc exist but are not used by the driver. That's not an acceptable reason for removing them from the binding. If they exist, they should be documented, regardless of whether the driver makes use of them. NAK. Also, your mail didn't end up on LKML, so there's something wrong with your setup. Thanks, Conor.
> On Fri, Jan 19, 2024 at 12:57:22PM +0000, Changhuang Liang wrote: > > EXTERNAL EMAIL: Do not click links or open attachments unless you know > > the content is safe > > > > Hi , Conor > > > > > 主题: Re: [PATCH v1 1/2] dt-bindings: media: starfive: Match driver > > > and yaml property names > > > > > > On Fri, Jan 19, 2024 at 02:06:38AM -0800, Changhuang Liang wrote: > > > > Drop some unused properties for clocks, resets and interrupts for > > > > StarFive JH7110 camera subsystem. > > > > > > What do you mean "unused"? > > > > > > Do these clocks etc exist but are not used by the driver? > > > > > > Or do they not exist at all? > > > > > > The two are very different! > > > These clocks etc exist but are not used by the driver. > > That's not an acceptable reason for removing them from the binding. If they > exist, they should be documented, regardless of whether the driver makes use > of them. NAK. If so, how to avoid the warning of dtbs_check. Thanks, Changhuang
On Mon, Jan 22, 2024 at 01:33:56AM +0000, Changhuang Liang wrote: > > > On Fri, Jan 19, 2024 at 12:57:22PM +0000, Changhuang Liang wrote: > > > EXTERNAL EMAIL: Do not click links or open attachments unless you know > > > the content is safe > > > > > > Hi , Conor > > > > > > > 主题: Re: [PATCH v1 1/2] dt-bindings: media: starfive: Match driver > > > > and yaml property names > > > > > > > > On Fri, Jan 19, 2024 at 02:06:38AM -0800, Changhuang Liang wrote: > > > > > Drop some unused properties for clocks, resets and interrupts for > > > > > StarFive JH7110 camera subsystem. > > > > > > > > What do you mean "unused"? > > > > > > > > Do these clocks etc exist but are not used by the driver? > > > > > > > > Or do they not exist at all? > > > > > > > > The two are very different! > > > > > These clocks etc exist but are not used by the driver. > > > > That's not an acceptable reason for removing them from the binding. If they > > exist, they should be documented, regardless of whether the driver makes use > > of them. NAK. > > If so, how to avoid the warning of dtbs_check. By also adding the clocks, resets and interrupts to the dts.
On Mon, Jan 22, 2024 at 08:24:51AM +0000, Conor Dooley wrote: > On Mon, Jan 22, 2024 at 01:33:56AM +0000, Changhuang Liang wrote: > > > > > On Fri, Jan 19, 2024 at 12:57:22PM +0000, Changhuang Liang wrote: > > > > EXTERNAL EMAIL: Do not click links or open attachments unless you know > > > > the content is safe > > > > > > > > Hi , Conor > > > > > > > > > 主题: Re: [PATCH v1 1/2] dt-bindings: media: starfive: Match driver > > > > > and yaml property names > > > > > > > > > > On Fri, Jan 19, 2024 at 02:06:38AM -0800, Changhuang Liang wrote: > > > > > > Drop some unused properties for clocks, resets and interrupts for > > > > > > StarFive JH7110 camera subsystem. > > > > > > > > > > What do you mean "unused"? > > > > > > > > > > Do these clocks etc exist but are not used by the driver? > > > > > > > > > > Or do they not exist at all? > > > > > > > > > > The two are very different! > > > > > > > These clocks etc exist but are not used by the driver. > > > > > > That's not an acceptable reason for removing them from the binding. If they > > > exist, they should be documented, regardless of whether the driver makes use > > > of them. NAK. > > > > If so, how to avoid the warning of dtbs_check. > > By also adding the clocks, resets and interrupts to the dts. Going through patchwork stuff now that the merge window is done. I'm gonna mark the dts patch as changes requeted. The binding (and dts) should describe all of the clocks the hardware has, whether or not you choose to use them all in software does not matter. Can you please resend the dts patch, with all of the clocks, resets and interrupts present? Thanks, Conor.
Hi, Conor, > On Mon, Jan 22, 2024 at 08:24:51AM +0000, Conor Dooley wrote: [...] > > > > > > Do these clocks etc exist but are not used by the driver? > > > > > > > > > > > > Or do they not exist at all? > > > > > > > > > > > > The two are very different! > > > > > > > > > These clocks etc exist but are not used by the driver. > > > > > > > > That's not an acceptable reason for removing them from the > > > > binding. If they exist, they should be documented, regardless of > > > > whether the driver makes use of them. NAK. > > > > > > If so, how to avoid the warning of dtbs_check. > > > > By also adding the clocks, resets and interrupts to the dts. > > Going through patchwork stuff now that the merge window is done. I'm gonna > mark the dts patch as changes requeted. The binding (and dts) should > describe all of the clocks the hardware has, whether or not you choose to use > them all in software does not matter. Can you please resend the dts patch, > with all of the clocks, resets and interrupts present? > You have applied the dts patch. https://lore.kernel.org/all/20240219032741.18387-1-changhuang.liang@starfivetech.com/ Regards, Changhuang
On 28 March 2024 01:06:42 GMT, Changhuang Liang <changhuang.liang@starfivetech.com> wrote: >Hi, Conor, > >> On Mon, Jan 22, 2024 at 08:24:51AM +0000, Conor Dooley wrote: >[...] >> > > > > > Do these clocks etc exist but are not used by the driver? >> > > > > > >> > > > > > Or do they not exist at all? >> > > > > > >> > > > > > The two are very different! >> > > > >> > > > > These clocks etc exist but are not used by the driver. >> > > > >> > > > That's not an acceptable reason for removing them from the >> > > > binding. If they exist, they should be documented, regardless of >> > > > whether the driver makes use of them. NAK. >> > > >> > > If so, how to avoid the warning of dtbs_check. >> > >> > By also adding the clocks, resets and interrupts to the dts. >> >> Going through patchwork stuff now that the merge window is done. I'm gonna >> mark the dts patch as changes requeted. The binding (and dts) should >> describe all of the clocks the hardware has, whether or not you choose to use >> them all in software does not matter. Can you please resend the dts patch, >> with all of the clocks, resets and interrupts present? >> > >You have applied the dts patch. >https://lore.kernel.org/all/20240219032741.18387-1-changhuang.liang@starfivetech.com/ Oh, sorry for the noise then. I guess I was mislead by patchwork marking this as new :)
diff --git a/Documentation/devicetree/bindings/media/starfive,jh7110-camss.yaml b/Documentation/devicetree/bindings/media/starfive,jh7110-camss.yaml index c66586d90fa2..2504381058b3 100644 --- a/Documentation/devicetree/bindings/media/starfive,jh7110-camss.yaml +++ b/Documentation/devicetree/bindings/media/starfive,jh7110-camss.yaml @@ -4,14 +4,14 @@ $id: http://devicetree.org/schemas/media/starfive,jh7110-camss.yaml# $schema: http://devicetree.org/meta-schemas/core.yaml# -title: Starfive SoC CAMSS ISP +title: StarFive SoC CAMSS ISP maintainers: - Jack Zhu <jack.zhu@starfivetech.com> - Changhuang Liang <changhuang.liang@starfivetech.com> description: - The Starfive CAMSS ISP is a Camera interface for Starfive JH7110 SoC. It + The StarFive CAMSS ISP is a Camera interface for StarFive JH7110 SoC. It consists of a VIN controller (Video In Controller, a top-level control unit) and an ISP. @@ -28,26 +28,21 @@ properties: - const: isp clocks: - maxItems: 7 + maxItems: 3 clock-names: items: - - const: apb_func - const: wrapper_clk_c - - const: dvp_inv - - const: axiwr - - const: mipi_rx0_pxl - const: ispcore_2x - const: isp_axi resets: - maxItems: 6 + maxItems: 5 reset-names: items: - const: wrapper_p - const: wrapper_c - - const: axird - const: axiwr - const: isp_top_n - const: isp_top_axi @@ -57,7 +52,7 @@ properties: - description: JH7110 ISP Power Domain Switch Controller. interrupts: - maxItems: 4 + maxItems: 3 ports: $ref: /schemas/graph.yaml#/properties/ports @@ -125,34 +120,24 @@ examples: reg = <0x19840000 0x10000>, <0x19870000 0x30000>; reg-names = "syscon", "isp"; - clocks = <&ispcrg 0>, - <&ispcrg 13>, - <&ispcrg 2>, - <&ispcrg 12>, - <&ispcrg 1>, + clocks = <&ispcrg 13>, <&syscrg 51>, <&syscrg 52>; - clock-names = "apb_func", - "wrapper_clk_c", - "dvp_inv", - "axiwr", - "mipi_rx0_pxl", + clock-names = "wrapper_clk_c", "ispcore_2x", "isp_axi"; resets = <&ispcrg 0>, <&ispcrg 1>, - <&ispcrg 10>, <&ispcrg 11>, <&syscrg 41>, <&syscrg 42>; reset-names = "wrapper_p", "wrapper_c", - "axird", "axiwr", "isp_top_n", "isp_top_axi"; power-domains = <&pwrc 5>; - interrupts = <92>, <87>, <88>, <90>; + interrupts = <92>, <87>, <90>; ports { #address-cells = <1>;