Re: [PATCH v5 RESEND 0/5] Add initial Milk-V Duo S board support

From: Joshua Milas <josh.milas@gmail.com>
To: Inochi Amaoto <inochiama@gmail.com>
Cc: tglx@kernel.org, robh@kernel.org, krzk+dt@kernel.org,
	conor+dt@kernel.org, pjw@kernel.org, samuel.holland@sifive.com,
	unicorn_wang@outlook.com, daniel.lezcano@linaro.org,
	palmer@dabbelt.com, aou@eecs.berkeley.edu, alex@ghiti.fr,
	liujingqi@lanxincomputing.com, alexander.sverdlin@gmail.com,
	rabenda.cn@gmail.com, dlan@kernel.org, chao.wei@sophgo.com,
	anup@brainfault.org, linux-kernel@vger.kernel.org,
	devicetree@vger.kernel.org, linux-riscv@lists.infradead.org,
	sophgo@lists.linux.dev, hanguidong02@gmail.com,
	michael.opdenacker@rootcommit.com
Subject: Re: [PATCH v5 RESEND 0/5] Add initial Milk-V Duo S board support
Date: Mon, 8 Jun 2026 09:28:10 -0400	[thread overview]
Message-ID: <aibDam5mwkaIJhHw@sleek> (raw)
In-Reply-To: <ahteTfYC5QMlpjF2@inochi.infowork>

Hi Inochi,

Thanks for the feedback. The next patch will have a large change with the
pinctrl, I believe I will have to drop the tags for that. For my reference
would the tags be applied to the commits or the top level patch?

I have the pinctrl configs staged in my tree but I am having some issues.

1. spi3 conflicts with sdhci1 (wifi) when enabled since they share pins,
   I am not sure how the original DT had it working. Im assuming they had it
   enabled and the user controlled it through the pinmux program? We originally
   tried to enable only what the vendor had in their DT. Taking another look
   at the vendor DT spi3 is the only SPI enabled, so we probably should not
   choose a different spi node. Should we just disable spi3?
2. I am having trouble getting the card to enumerate with just the DT. I was
   able to get it enumerated with some kernel changes and I am trying
   to translate those changes into the DT.
3. I added at least one pinctrl config for each device. For the devices that
   have multiple selections, should I add all the configs?

Have a great day,
- Joshua Milas