Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <652841a9-a97e-0c87-8742-701be6633b82@ti.com>
Date:   Fri, 2 Sep 2022 20:04:57 +0530
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.13.0
Subject: Re: [PATCH RESEND 2/6] arm64: dts: ti: k3-j721s2-main: Add SERDES and
 WIZ device tree node
Content-Language: en-US
To:     Matt Ranostay <mranostay@ti.com>
CC:     <devicetree@vger.kernel.org>,
        <linux-arm-kernel@lists.infradead.org>,
        <linux-kernel@vger.kernel.org>, Nishanth Menon <nm@ti.com>
References: <20220810094000.248487-1-mranostay@ti.com>
 <20220810094000.248487-2-mranostay@ti.com>
 <8c130e92-9970-54d3-d263-37ae2f722b34@ti.com> <YxFEk5zU4l9Yraek@ubuntu>
From:   Vignesh Raghavendra <vigneshr@ti.com>
In-Reply-To: <YxFEk5zU4l9Yraek@ubuntu>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Precedence: bulk


On 02/09/22 05:17, Matt Ranostay wrote:
> On Thu, Sep 01, 2022 at 02:15:28PM +0530, Vignesh Raghavendra wrote:
>>
>>
>> On 10/08/22 15:09, Matt Ranostay wrote:
>>> From: Aswath Govindraju <a-govindraju@ti.com>
>>>
>>> Add dt node for the single instance of WIZ (SERDES wrapper) and
>>> SERDES module shared by PCIe, eDP and USB.
>>>
>>> Cc: Vignesh Raghavendra <vigneshr@ti.com>
>>> Cc: Nishanth Menon <nm@ti.com>
>>> Acked-by: Matt Ranostay <mranostay@ti.com>
>>> Signed-off-by: Aswath Govindraju <a-govindraju@ti.com>
>>> ---
>>>  arch/arm64/boot/dts/ti/k3-j721s2-main.dtsi | 70 ++++++++++++++++++++++
>>>  1 file changed, 70 insertions(+)
>>>
>>> diff --git a/arch/arm64/boot/dts/ti/k3-j721s2-main.dtsi b/arch/arm64/boot/dts/ti/k3-j721s2-main.dtsi
>>> index f7e359da8690..f1e02d896168 100644
>>> --- a/arch/arm64/boot/dts/ti/k3-j721s2-main.dtsi
>>> +++ b/arch/arm64/boot/dts/ti/k3-j721s2-main.dtsi
>>> @@ -5,6 +5,13 @@
>>>   * Copyright (C) 2021 Texas Instruments Incorporated - https://www.ti.com/
>>>   */
>>>  
>>> +/ {
>>> +	serdes_refclk: serdes-refclk {
>>
>> clk-X
>>
> 

Sorry for the confusion, I meant clock-<abc>. DT specification says all
clock nodes should follow above naming convention

> So something like 'serdes_refclk: clock-cmnrefclk' would seem to match the current
> naming style of other K3 platforms (albeit except j7200).
> 

Make sense...

>>> +		#clock-cells = <0>;
>>> +		compatible = "fixed-clock";
>>> +	};
>>> +};
>>> +
>>>  &cbass_main {
>>>  	msmc_ram: sram@70000000 {
>>>  		compatible = "mmio-sram";
>>> @@ -33,6 +40,13 @@ scm_conf: scm-conf@104000 {
>>>  		#size-cells = <1>;
>>>  		ranges = <0x00 0x00 0x00104000 0x18000>;
>>>  
>>> +		serdes_ln_ctrl: mux-controller@80 {
>>
>> mux-controller-X
>>
> 
> 80 is the index into the mmio region. So should it still have @ for the node
> name?
> 

@addr can only be used with nodes that have reg property. And since
there is no reg here -> would follow mux-controller-<abc> convention.

> Also this is how the rest of the K3 platforms reference the mux controller.


We need to fix other dtsi files to drop @addr part as they are now
falling make dtbs_check

> 
>>> +			compatible = "mmio-mux";
>>> +			#mux-control-cells = <1>;
>>> +			mux-reg-masks = <0x80 0x3>, <0x84 0x3>, /* SERDES0 lane0/1 select */
>>> +					<0x88 0x3>, <0x8c 0x3>; /* SERDES0 lane2/3 select */
>>> +		};
>>> +
>>>  		usb_serdes_mux: mux-controller@0 {
>>>  			compatible = "mmio-mux";
>>>  			#mux-control-cells = <1>;
>>> @@ -728,6 +742,62 @@ usb0: usb@6000000 {
>>>  		};
>>>  	};
>>>  
>>> +	serdes_wiz0: wiz@5060000 {
>>> +		compatible = "ti,j721e-wiz-10g";
>>> +		#address-cells = <1>;
>>> +		#size-cells = <1>;
>>> +		power-domains = <&k3_pds 365 TI_SCI_PD_EXCLUSIVE>;
>>> +		clocks = <&k3_clks 365 0>, <&k3_clks 365 3>, <&serdes_refclk>;
>>> +		clock-names = "fck", "core_ref_clk", "ext_ref_clk";
>>> +		num-lanes = <4>;
>>> +		#reset-cells = <1>;
>>> +		ranges = <0x5060000 0x0 0x5060000 0x10000>;
>>> +
>>> +		assigned-clocks = <&k3_clks 365 3>;
>>> +		assigned-clock-parents = <&k3_clks 365 7>;
>>> +
>>> +		wiz0_pll0_refclk: pll0-refclk {
>>
>> please use generic node name clk-X here and elsewhere in the patch
> 
> So something like 'wiz0_pll0_refclk: clk-refpll0'?


Sorry, clock-refpll0  perhaps

[...]

-- 
Regards
Vignesh