Re: [PATCH 05/15] arm64: dts: imx8mp-evk: enable uart1/3 ports

From: Peng Fan <peng.fan@oss.nxp.com>
To: Marco Felsch <m.felsch@pengutronix.de>
Cc: robh+dt@kernel.org, krzysztof.kozlowski+dt@linaro.org,
	shawnguo@kernel.org, s.hauer@pengutronix.de,
	devicetree@vger.kernel.org, Peng Fan <peng.fan@nxp.com>,
	linux-kernel@vger.kernel.org, linux-imx@nxp.com,
	kernel@pengutronix.de, festevam@gmail.com,
	linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH 05/15] arm64: dts: imx8mp-evk: enable uart1/3 ports
Date: Fri, 21 Oct 2022 17:31:50 +0800	[thread overview]
Message-ID: <5ddc1dac-b6ee-34c4-63eb-63c9a254d984@oss.nxp.com> (raw)
In-Reply-To: <20221021090910.zkijqqt6mpukzqdp@pengutronix.de>

Hi Marco,

On 10/21/2022 5:09 PM, Marco Felsch wrote:
> On 22-10-21, Peng Fan wrote:
>> Hi Marco,
>>
>> On 10/20/2022 7:07 PM, Marco Felsch wrote:
>>> Hi Peng,
>>>
>>> On 22-10-20, Peng Fan (OSS) wrote:
>>>> From: Peng Fan <peng.fan@nxp.com>
>>>>
>>>> Enable uart1/3 ports for evk board.
>>>>
>>>> Signed-off-by: Peng Fan <peng.fan@nxp.com>
>>>> ---
>>>>    arch/arm64/boot/dts/freescale/imx8mp-evk.dts | 36 ++++++++++++++++++++
>>>>    1 file changed, 36 insertions(+)
>>>>
>>>> diff --git a/arch/arm64/boot/dts/freescale/imx8mp-evk.dts b/arch/arm64/boot/dts/freescale/imx8mp-evk.dts
>>>> index 2e29bb3c041c..366f709f8790 100644
>>>> --- a/arch/arm64/boot/dts/freescale/imx8mp-evk.dts
>>>> +++ b/arch/arm64/boot/dts/freescale/imx8mp-evk.dts
>>>> @@ -428,6 +428,15 @@ &snvs_pwrkey {
>>>>    	status = "okay";
>>>>    };
>>>> +&uart1 { /* BT */
>>>> +	pinctrl-names = "default";
>>>> +	pinctrl-0 = <&pinctrl_uart1>;
>>>> +	assigned-clocks = <&clk IMX8MP_CLK_UART1>;
>>>> +	assigned-clock-parents = <&clk IMX8MP_SYS_PLL1_80M>;
>>>
>>> I'm curious, what is the default parent and why is this wrong? For the
>>> already exisiting uart2 we don't do that. Same applies for uart3.
>>
>> The default parent is OSC_24M. The uart2 is for console, so 24M is ok.
>> As I recall, we met issue 24M not able to get higher baudrate.
> 
> What did you mean by higher baudrate, is it everything > 115200? When
> the console baudrates can be fullfilled with the PLL1_80M as well
> wouldn't it be worth to fix the imx8mp.dtsi instead?

To console, we use 115200, 24M could fullfill it.

BaudRate = (clk / ref_clk_div) / (16 * (ubmr + 1) / (ubir + 1))

If you have 24M ref_clk, the max baudrate is 1.5M, with setting
ubmr, ubir to 0, and ref clk divider to 1.

So more higher baudrate, 24M could not fulfill.

THanks,
Peng.

> 
> Regards,
>    Marco

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel