From: Geert Uytterhoeven <geert@linux-m68k.org> To: Alex Riesen <alexander.riesen@cetitec.com>, Kieran Bingham <kieran.bingham@ideasonboard.com>, Mauro Carvalho Chehab <mchehab@kernel.org>, Hans Verkuil <hverkuil-cisco@xs4all.nl>, Laurent Pinchart <laurent.pinchart@ideasonboard.com>, Rob Herring <robh+dt@kernel.org>, Mark Rutland <mark.rutland@arm.com>, driverdevel <devel@driverdev.osuosl.org>, Linux Media Mailing List <linux-media@vger.kernel.org>, Linux Kernel Mailing List <linux-kernel@vger.kernel.org>, "open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" <devicetree@vger.kernel.org>, Linux-Renesas <linux-renesas-soc@vger.kernel.org> Subject: Re: [PATCH 8/8] arm64: dts: renesas: salvator: add a connection from adv748x codec (HDMI input) to the R-Car SoC Date: Mon, 2 Mar 2020 16:32:32 +0100 [thread overview] Message-ID: <CAMuHMdW21rYXoOSE8azHNqYjng_j41rsL=Fo2bZc=1ULi9+pLw@mail.gmail.com> (raw) In-Reply-To: <20200302150706.GB3717@pflmari> Hi Alex, On Mon, Mar 2, 2020 at 4:07 PM Alex Riesen <alexander.riesen@cetitec.com> wrote: > Geert Uytterhoeven, Mon, Mar 02, 2020 14:47:46 +0100: > > On Mon, Mar 2, 2020 at 2:40 PM Alex Riesen <alexander.riesen@cetitec.com> wrote: > > > > > --- a/arch/arm64/boot/dts/renesas/salvator-common.dtsi > > > > > +++ b/arch/arm64/boot/dts/renesas/salvator-common.dtsi > > > > > @@ -322,6 +322,10 @@ > > > > > clock-frequency = <22579200>; > > > > > }; > > > > > > > > > > +&audio_clk_c { > > > > > + clock-frequency = <12288000>; > > > > > +}; > > > > > > > > Does the ADV7482 always generate a 12.288 MHz clock signal? > > > > Or is this programmable? > > > > > > Oops. It looks like it is and the value is derived from the sampling rate > > > (48kHz) and the master clock multiplier. Both hard-coded in the board file. > > > > Where are these hardcoded in the board file? > > In the endpoint definition, arch/arm64/boot/dts/renesas/r8a7795-es1-salvator-x.dts > > So the frequency can be set at the run-time, perhaps even derived from > endpoint connected to the output. In this case, rsnd_endpoint3, > which has the "mclk-fs" setting. Not sure if the sampling rate > can be set to something else for the HDMI, though. > > > Even if they are, technically this is a clock output of the ADV7482. > > ... which I hope to correct as soon as I steal the hardware from whoever stole > it from me... > > > > > > video-receiver@70 { > > > > > compatible = "adi,adv7482"; > > > > > ... > > > > > + clocks = <&rcar_sound 3>, <&audio_clk_c>; > > > > > + clock-names = "clk-hdmi-video", "clk-hdmi-i2s-mclk"; > > > > > > > > The above declares the Audio CLK C to be a clock input of the ADV7482, while > > > > it is an output. > > > > > > I would gladly give it right direction if I *really* understood what I was > > > doing... > > > > :-) > > > > > > Furthermore, the DT bindings do not document that clocks can be specified. > > > > > > Should the DT bindings document that the clock cannot be specified than? > > > > It currently does say so, as it doesn't list "clocks" in its properties section. > > The bindings documentation file, which we're talking about here and which does > not list the specifiable input clocks in its properties, is it the > > Documentation/devicetree/bindings/media/i2c/adv748x.txt > > ? Yes. > > And this absence of documentation also means that whatever clocks (both input > in "clocks=" and output in "#clock-cells") listed in a specific .dts are just > an integration detail? No, the absence probably means that any clock-related properties in a .dts file will just be ignored. Looking at the driver source, it indeed has no support related to clocks at all. > Does this below makes more sense, than? > > video-receiver@70 { > compatible = "adi,adv7482"; > clocks = <&rcar_sound 3>; > clock-names = "clk-hdmi-video"; > adv748x_mclk: mclk { > compatible = "fixed-clock"; > #clock-cells = <0>; > /* frequency hard-coded for illustration */ > clock-frequency = <12288000>; > clock-output-names = "clk-hdmi-i2s-mclk"; > }; > }; The #clock-cells should be in the main video-receiver node. Probably there is more than one clock output, so #clock-cells may be 1? There is no need for a fixed-clock compatible, nor for clock-frequency and clock-output-names. But most important: this should be documented in the adv748x DT bindings, and implemented in the adv748x driver. > Now I'm a bit hazy on how to declare that the MCLK output of the > video-receiver@70 is connected to the Audio Clock C of the SoC... > Probably remove use of "audio_clk_c" completely? Yes, the current audio_clk_c definition in the DTS assumes a fixed crystal. > > > > > @@ -686,7 +700,8 @@ > > > > > }; > > > > > > > > > > sound_pins: sound { > > > > > - groups = "ssi01239_ctrl", "ssi0_data", "ssi1_data_a"; > > > > > + groups = "ssi01239_ctrl", "ssi0_data", "ssi1_data_a", > > > > > + "ssi4_data"; > > > > > > > > Missing "ss4_ctrl", for the SCK4 and WS4 pins. > > > > > > I'll add them. > > > As the device seems to function even without thoes, does this mean the > > > pins in the group are used "on demand" by whatever needs them? > > > > Probably the SCK4/WS4 functions are the reset-state defaults. > > That ... might require some trial and testing: when I add them to the group, > the reset defaults will be overridden by the platform initialization, which is > not necessarily the reset default. Will see. Or by the boot loader. Anyway, you need to specify these in the DTS. > > > Does a "clocks = ..." statement always mean input clocks? > > > > Yes it does. > > If a device has clock outputs and is thus a clock provider, it should > > have a #clock-cells property, and this should be documented in the bindings. > > > > A clock consumer will refer to clocks of a provider using the "clocks" > > property, specifying a clock specifier (phandle and zero or more indices) > > for each clock referenced. > > Something like this? > > &rcar_sound { > clocks = ..., > <&adv748x_mclk>, > <&cpg CPG_CORE CPG_AUDIO_CLK_I>; > clock-names = ..., > "clk_c", > "clk_i"; > }; More or less. Might become find_a_better_label_choice: video-receiver@70 { ... }; &rcar_sound { clock = ..., <&find_a_better_label_choice 0>, ... }; as there may be multiple clock outputs on the ADV7482. Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds
WARNING: multiple messages have this Message-ID (diff)
From: Geert Uytterhoeven <geert@linux-m68k.org> To: Alex Riesen <alexander.riesen@cetitec.com>, Kieran Bingham <kieran.bingham@ideasonboard.com>, Mauro Carvalho Chehab <mchehab@kernel.org>, Hans Verkuil <hverkuil-cisco@xs4all.nl>, Laurent Pinchart <laurent.pinchart@ideasonboard.com>, Rob Herring <robh+dt@kernel.org>, Mark Rutland <mark.rutland@arm.com>, driverdevel <devel@driverdev.osuosl.org>, Linux Media Mailing List <linux-media@vger.kernel.org>, Linux Kernel Mailing List <linux-kernel@vger.kernel.org>, "open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" <devicetree@vger.kernel.org>, Linux-Renesas <linux-renesas-soc@vger.kernel.org> Subject: Re: [PATCH 8/8] arm64: dts: renesas: salvator: add a connection from adv748x codec (HDMI input) to the R-Car SoC Date: Mon, 2 Mar 2020 16:32:32 +0100 [thread overview] Message-ID: <CAMuHMdW21rYXoOSE8azHNqYjng_j41rsL=Fo2bZc=1ULi9+pLw@mail.gmail.com> (raw) In-Reply-To: <20200302150706.GB3717@pflmari> Hi Alex, On Mon, Mar 2, 2020 at 4:07 PM Alex Riesen <alexander.riesen@cetitec.com> wrote: > Geert Uytterhoeven, Mon, Mar 02, 2020 14:47:46 +0100: > > On Mon, Mar 2, 2020 at 2:40 PM Alex Riesen <alexander.riesen@cetitec.com> wrote: > > > > > --- a/arch/arm64/boot/dts/renesas/salvator-common.dtsi > > > > > +++ b/arch/arm64/boot/dts/renesas/salvator-common.dtsi > > > > > @@ -322,6 +322,10 @@ > > > > > clock-frequency = <22579200>; > > > > > }; > > > > > > > > > > +&audio_clk_c { > > > > > + clock-frequency = <12288000>; > > > > > +}; > > > > > > > > Does the ADV7482 always generate a 12.288 MHz clock signal? > > > > Or is this programmable? > > > > > > Oops. It looks like it is and the value is derived from the sampling rate > > > (48kHz) and the master clock multiplier. Both hard-coded in the board file. > > > > Where are these hardcoded in the board file? > > In the endpoint definition, arch/arm64/boot/dts/renesas/r8a7795-es1-salvator-x.dts > > So the frequency can be set at the run-time, perhaps even derived from > endpoint connected to the output. In this case, rsnd_endpoint3, > which has the "mclk-fs" setting. Not sure if the sampling rate > can be set to something else for the HDMI, though. > > > Even if they are, technically this is a clock output of the ADV7482. > > ... which I hope to correct as soon as I steal the hardware from whoever stole > it from me... > > > > > > video-receiver@70 { > > > > > compatible = "adi,adv7482"; > > > > > ... > > > > > + clocks = <&rcar_sound 3>, <&audio_clk_c>; > > > > > + clock-names = "clk-hdmi-video", "clk-hdmi-i2s-mclk"; > > > > > > > > The above declares the Audio CLK C to be a clock input of the ADV7482, while > > > > it is an output. > > > > > > I would gladly give it right direction if I *really* understood what I was > > > doing... > > > > :-) > > > > > > Furthermore, the DT bindings do not document that clocks can be specified. > > > > > > Should the DT bindings document that the clock cannot be specified than? > > > > It currently does say so, as it doesn't list "clocks" in its properties section. > > The bindings documentation file, which we're talking about here and which does > not list the specifiable input clocks in its properties, is it the > > Documentation/devicetree/bindings/media/i2c/adv748x.txt > > ? Yes. > > And this absence of documentation also means that whatever clocks (both input > in "clocks=" and output in "#clock-cells") listed in a specific .dts are just > an integration detail? No, the absence probably means that any clock-related properties in a .dts file will just be ignored. Looking at the driver source, it indeed has no support related to clocks at all. > Does this below makes more sense, than? > > video-receiver@70 { > compatible = "adi,adv7482"; > clocks = <&rcar_sound 3>; > clock-names = "clk-hdmi-video"; > adv748x_mclk: mclk { > compatible = "fixed-clock"; > #clock-cells = <0>; > /* frequency hard-coded for illustration */ > clock-frequency = <12288000>; > clock-output-names = "clk-hdmi-i2s-mclk"; > }; > }; The #clock-cells should be in the main video-receiver node. Probably there is more than one clock output, so #clock-cells may be 1? There is no need for a fixed-clock compatible, nor for clock-frequency and clock-output-names. But most important: this should be documented in the adv748x DT bindings, and implemented in the adv748x driver. > Now I'm a bit hazy on how to declare that the MCLK output of the > video-receiver@70 is connected to the Audio Clock C of the SoC... > Probably remove use of "audio_clk_c" completely? Yes, the current audio_clk_c definition in the DTS assumes a fixed crystal. > > > > > @@ -686,7 +700,8 @@ > > > > > }; > > > > > > > > > > sound_pins: sound { > > > > > - groups = "ssi01239_ctrl", "ssi0_data", "ssi1_data_a"; > > > > > + groups = "ssi01239_ctrl", "ssi0_data", "ssi1_data_a", > > > > > + "ssi4_data"; > > > > > > > > Missing "ss4_ctrl", for the SCK4 and WS4 pins. > > > > > > I'll add them. > > > As the device seems to function even without thoes, does this mean the > > > pins in the group are used "on demand" by whatever needs them? > > > > Probably the SCK4/WS4 functions are the reset-state defaults. > > That ... might require some trial and testing: when I add them to the group, > the reset defaults will be overridden by the platform initialization, which is > not necessarily the reset default. Will see. Or by the boot loader. Anyway, you need to specify these in the DTS. > > > Does a "clocks = ..." statement always mean input clocks? > > > > Yes it does. > > If a device has clock outputs and is thus a clock provider, it should > > have a #clock-cells property, and this should be documented in the bindings. > > > > A clock consumer will refer to clocks of a provider using the "clocks" > > property, specifying a clock specifier (phandle and zero or more indices) > > for each clock referenced. > > Something like this? > > &rcar_sound { > clocks = ..., > <&adv748x_mclk>, > <&cpg CPG_CORE CPG_AUDIO_CLK_I>; > clock-names = ..., > "clk_c", > "clk_i"; > }; More or less. Might become find_a_better_label_choice: video-receiver@70 { ... }; &rcar_sound { clock = ..., <&find_a_better_label_choice 0>, ... }; as there may be multiple clock outputs on the ADV7482. Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds _______________________________________________ devel mailing list devel@linuxdriverproject.org http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
next prev parent reply other threads:[~2020-03-02 15:32 UTC|newest] Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top 2020-01-13 14:19 [PATCH 0/8] media: i2c: adv748x: add support for HDMI audio Alex Riesen 2020-01-13 14:19 ` Alex Riesen 2020-01-13 14:15 ` [PATCH 1/8] media: adv748x: add a device-specific wrapper for register block read Alex Riesen 2020-01-13 14:15 ` Alex Riesen 2020-01-13 14:15 ` [PATCH 2/8] media: adv748x: add audio mute control and output selection ioctls Alex Riesen 2020-01-13 14:15 ` Alex Riesen 2020-03-13 8:16 ` Hans Verkuil 2020-03-13 8:16 ` Hans Verkuil 2020-03-13 10:26 ` Alex Riesen 2020-03-13 10:26 ` Alex Riesen 2020-03-13 10:52 ` Hans Verkuil 2020-03-13 10:52 ` Hans Verkuil 2020-03-13 11:00 ` Alex Riesen 2020-03-13 11:00 ` Alex Riesen 2020-01-13 14:15 ` [PATCH 3/8] media: adv748x: add log_status ioctl Alex Riesen 2020-01-13 14:15 ` Alex Riesen 2020-01-13 14:15 ` [PATCH 4/8] media: adv748x: reserve space for the audio (I2S) port in the driver structures Alex Riesen 2020-01-13 14:15 ` Alex Riesen 2020-01-13 14:15 ` [PATCH 5/8] media: adv748x: add an ASoC DAI definition to the driver Alex Riesen 2020-01-13 14:15 ` Alex Riesen 2020-01-13 14:15 ` [PATCH 6/8] media: adv748x: reduce amount of code for bitwise modification of device registers Alex Riesen 2020-01-13 14:15 ` Alex Riesen 2020-01-13 14:15 ` [PATCH 7/8] dt-bindings: adv748x: add information about serial audio interface (I2S/TDM) Alex Riesen 2020-01-13 14:15 ` Alex Riesen 2020-01-13 22:32 ` Rob Herring 2020-01-13 22:32 ` Rob Herring 2020-01-13 14:15 ` [PATCH 8/8] arm64: dts: renesas: salvator: add a connection from adv748x codec (HDMI input) to the R-Car SoC Alex Riesen 2020-01-13 14:15 ` Alex Riesen 2020-03-02 12:28 ` Geert Uytterhoeven 2020-03-02 12:28 ` Geert Uytterhoeven 2020-03-02 13:40 ` Alex Riesen 2020-03-02 13:40 ` Alex Riesen 2020-03-02 13:47 ` Geert Uytterhoeven 2020-03-02 13:47 ` Geert Uytterhoeven 2020-03-02 15:07 ` Alex Riesen 2020-03-02 15:07 ` Alex Riesen 2020-03-02 15:32 ` Geert Uytterhoeven [this message] 2020-03-02 15:32 ` Geert Uytterhoeven 2020-03-02 16:09 ` Alex Riesen 2020-03-02 16:09 ` Alex Riesen 2020-03-02 16:13 ` Geert Uytterhoeven 2020-03-02 16:13 ` Geert Uytterhoeven 2020-03-05 14:36 ` Alex Riesen 2020-03-05 14:36 ` Alex Riesen 2020-03-06 12:21 ` Geert Uytterhoeven 2020-03-06 12:21 ` Geert Uytterhoeven 2020-03-06 13:16 ` Laurent Pinchart 2020-03-06 13:16 ` Laurent Pinchart 2020-03-06 13:41 ` Alex Riesen 2020-03-06 13:41 ` Alex Riesen 2020-03-06 13:45 ` Laurent Pinchart 2020-03-06 13:45 ` Laurent Pinchart 2020-03-09 1:31 ` Kuninori Morimoto 2020-03-09 1:31 ` Kuninori Morimoto 2020-03-09 11:09 ` Alex Riesen 2020-03-09 11:09 ` Alex Riesen 2020-03-10 1:07 ` Kuninori Morimoto 2020-03-10 1:07 ` Kuninori Morimoto 2020-03-10 8:17 ` Alex Riesen 2020-03-10 8:17 ` Alex Riesen 2020-03-10 10:39 ` Laurent Pinchart 2020-03-10 10:39 ` Laurent Pinchart
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to='CAMuHMdW21rYXoOSE8azHNqYjng_j41rsL=Fo2bZc=1ULi9+pLw@mail.gmail.com' \ --to=geert@linux-m68k.org \ --cc=alexander.riesen@cetitec.com \ --cc=devel@driverdev.osuosl.org \ --cc=devicetree@vger.kernel.org \ --cc=hverkuil-cisco@xs4all.nl \ --cc=kieran.bingham@ideasonboard.com \ --cc=laurent.pinchart@ideasonboard.com \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-media@vger.kernel.org \ --cc=linux-renesas-soc@vger.kernel.org \ --cc=mark.rutland@arm.com \ --cc=mchehab@kernel.org \ --cc=robh+dt@kernel.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.