lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <d108104e-3ecd-43a9-b27f-4acd1c9e3c86@ti.com>
Date:   Fri, 6 Oct 2023 18:23:58 +0530
From:   "J, KEERTHY" <j-keerthy@...com>
To:     Nishanth Menon <nm@...com>
CC:     <robh+dt@...nel.org>, <vigneshr@...com>, <conor+dt@...nel.org>,
        <kristo@...nel.org>, <krzysztof.kozlowski+dt@...aro.org>,
        <u-kumar1@...com>, <linux-arm-kernel@...ts.infradead.org>,
        <devicetree@...r.kernel.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v7 5/7] arm64: dts: ti: k3-j784s4-mcu: Add the mcu domain
 watchdog instances



On 10/6/2023 5:31 PM, Nishanth Menon wrote:
> On 17:22-20231006, J, KEERTHY wrote:
>>
>>
>> On 10/6/2023 5:04 PM, Nishanth Menon wrote:
>>> On 09:58-20231006, Keerthy wrote:
>>>> There are totally 2 instances of watchdog module in MCU domain.
>>>> These instances are coupled with the MCU domain R5F instances.
>>>
>>>> Disabling them as they are not used by Linux.
>>> Device tree is hardware description - not tied to how Linux uses it.
>>>
>>> Reason these wdts are disabled by default is because they are tightly
>>> coupled with R5Fs.
>>>
>>>>
>>>> Signed-off-by: Keerthy <j-keerthy@...com>
>>>> ---
>>>>    .../boot/dts/ti/k3-j784s4-mcu-wakeup.dtsi     | 24 +++++++++++++++++++
>>>>    1 file changed, 24 insertions(+)
>>>>
>>>> diff --git a/arch/arm64/boot/dts/ti/k3-j784s4-mcu-wakeup.dtsi b/arch/arm64/boot/dts/ti/k3-j784s4-mcu-wakeup.dtsi
>>>> index a7b5c4cb7d3e..809a0b1cf038 100644
>>>> --- a/arch/arm64/boot/dts/ti/k3-j784s4-mcu-wakeup.dtsi
>>>> +++ b/arch/arm64/boot/dts/ti/k3-j784s4-mcu-wakeup.dtsi
>>>> @@ -714,4 +714,28 @@
>>>>    		ti,esm-pins = <63>;
>>>>    		bootph-pre-ram;
>>>>    	};
>>>> +
>>
>> Nishanth,
>>
>> Below i have addressed the coupling with R5Fs & MCU domains watcdogs.
>>
>>>> +	/*
>>>> +	 * The 2 RTI instances are couple with MCU R5Fs so keeping them
>>>> +	 * disabled as these will be used by their respective firmware
> 
> This description is best in the commit message
> 
>>>> +	 */
>>>> +	mcu_watchdog0: watchdog@...00000 {
>>>> +		compatible = "ti,j7-rti-wdt";
>>>> +		reg = <0x00 0x40600000 0x00 0x100>;
>>>> +		clocks = <&k3_clks 367 1>;
>>>> +		power-domains = <&k3_pds 367 TI_SCI_PD_EXCLUSIVE>;
>>>> +		assigned-clocks = <&k3_clks 367 0>;
>>>> +		assigned-clock-parents = <&k3_clks 367 4>;
>>>> +		status = "disabled";
>>>> +	};
>>>> +
>>>> +	mcu_watchdog1: watchdog@...10000 {
>>>> +		compatible = "ti,j7-rti-wdt";
>>>> +		reg = <0x00 0x40610000 0x00 0x100>;
>>>> +		clocks = <&k3_clks 368 1>;
>>>> +		power-domains = <&k3_pds 368 TI_SCI_PD_EXCLUSIVE>;
>>>> +		assigned-clocks = <&k3_clks 368 0>;
>>>> +		assigned-clock-parents = <&k3_clks 368 4>;
>>>
>>> Please DONOT ignore the review comments - I did ask the documentation in
>>> dts as well. reason being that this is what people will see rather than
>>> dig up the commit log. it should be intutive when reading the dts why
>>> nodes are disabled by default Vs the standard of leaving it enabled by
>>> default. Given esp that these peripherals do not have anything to do
>>> with board semantics (pinmux or something similar) to be complete.
>>
>> As mentioned above. I added single comment for addressing both the
>> watchdogs.
> 
> I missed it completely. Now that I think of it, I seem to have missed
> having seen it in previous rev reviews as well, and there is a reason
> for it: See [1] clarifying comment - nodes reserved for firmware usage
> have convention of "reserved" as status and documentation immediately
> above the status to help clarify the reason in-context. That is more
> readable than having to scroll up to find the rationale.
> 
> [1] https://lore.kernel.org/all/20231006114422.avymeap7h5ocs6zq@dreadlock/

Thanks Nishanth. I agree reserved is better and I will add a comment 
something like below:

/* Tightly coupled to R5F */
status = "reserved";

- Keerthy
> 

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ