[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <8c801213-2ca7-4336-a8a2-05d43a4a9251@ti.com>
Date: Tue, 22 Apr 2025 22:51:52 +0530
From: "Kumar, Udit" <u-kumar1@...com>
To: Judith Mendez <jm@...com>, Nishanth Menon <nm@...com>
CC: Vignesh Raghavendra <vigneshr@...com>, Tero Kristo <kristo@...nel.org>,
Rob Herring <robh@...nel.org>,
Krzysztof Kozlowski <krzk+dt@...nel.org>,
Conor Dooley <conor+dt@...nel.org>,
<linux-arm-kernel@...ts.infradead.org>, <devicetree@...r.kernel.org>,
<linux-kernel@...r.kernel.org>, Moteen Shah
<m-shah@...com>,
<u-kumar1@...com>
Subject: Re: [PATCH v2 1/5] arm64: dts: ti: k3-am6*: Set eMMC clock parents to
default
On 4/22/2025 10:47 PM, Judith Mendez wrote:
> Hi Udit,
>
> On 4/22/25 9:08 AM, Kumar, Udit wrote:
>>
>> On 4/22/2025 6:07 PM, Nishanth Menon wrote:
>>> On 09:36-20250421, Judith Mendez wrote:
>>>> Hi Udit,
>>>>
>>>> On 4/19/25 10:00 AM, Kumar, Udit wrote:
>>>>> On 4/18/2025 5:00 AM, Judith Mendez wrote:
>>>>>> Set eMMC clock parents to the defaults which is
>>>>>> MAIN_PLL0_HSDIV5_CLKOUT
>>>>>> for eMMC. This change is necessary since DM is not implementing the
>>>>>> correct procedure to switch PLL clock source for eMMC and we have a
>>>>>> non-glich-free mux. To remove any potential issues, lets switch
>>>>>> back to
>>>>>> the defaults.
>>>>> IMO, we need to fix DM if not then documentation [0] .
>>>> DM cannot be fixed for only one IP and documentation says what clock
>>>> parents are supported, it does not have to say what are the issues
>>>> that come with using a specific clock parent.
>>> As I understand the arasan IP requirement is that the IP must be
>>> held in
>>> reset while the clock is switched, which is not reasonable to implement
>>> given decoupled systems like DM and Linux OS.
>>
>>
>> Thanks , then fix should be extended to sdhci1 as well
>
> sdhci1 and sdhci2 should already be using the defaults:
> https://gist.github.com/jmenti/e280232ae3d5bb5097df7b0f04c85e8a
> This change is only needed for eMMC.
>
> ~ Judith
Ok, thanks for confirmation
Acked-by: Udit Kumar <u-kumar1@...com>
>
>>
>>
>>>
>>>>> Then only this patch is ok because as per document [0]
>>>>>
>>>>> removed clock by this patch is valid parent for eMMC.
>>>> The clock parent currently set is a valid parent, but we have non-
>>>> glitch-free muxes and to avoid any potential issues with these, we
>>>> should switch back to the defaults. It seems like we randomly switched
>>>> from the default for no good reason and it has been copy paste per
>>>> platforms since then, so we are switching back to the defaults now.
>>>>
>>>> ~ Judith
>>>>
>>>>> [0]
>>>>> https://software-dl.ti.com/tisci/esd/latest/5_soc_doc/j722s/clocks.html
>>>>>
>>>>>
>>>>> Thanks
>>>>>
>>>>> Udit
>>>>>
>>>>>> Fixes: c37c58fdeb8a ("arm64: dts: ti: k3-am62: Add more peripheral
>>>>>> nodes")
>>>>>> Fixes: d3ae4e8d8b6a ("arm64: dts: ti: k3-am62a-main: Add sdhci0
>>>>>> instance")
>>>>>> Fixes: b5080c7c1f7e ("arm64: dts: ti: k3-am62p: Add nodes for
>>>>>> more IPs")
>>> Please follow ./Documentation/process/stable-kernel-rules.rst - this is
>>> easier to backport if these were to be split into 3 different patches.
>>> Please do not forget to add Cc: stable@...r.kernel.org as per the rules
>>> as well.
>>>
>>>>>> Signed-off-by: Judith Mendez <jm@...com>
>>>>>> ---
>>>>>> arch/arm64/boot/dts/ti/k3-am62-main.dtsi | 2 --
>>>>>> arch/arm64/boot/dts/ti/k3-am62a-main.dtsi | 2 --
>>>>>> arch/arm64/boot/dts/ti/k3-am62p-j722s-common-main.dtsi | 2 --
>>>>>> 3 files changed, 6 deletions(-)
>>>>>>
>>>>>> [..]
>
Powered by blists - more mailing lists