Re: Allwinner D1

[Nick Hudson <nick.hudson%gmx.co.uk@localhost>]

> On 16 Mar 2026, at 02:37, Rui-Xiang Guo <rxg%lavabit.com@localhost> wrote:
> 
> On Mon, Oct 13, 2025 at 02:08:21PM -0700, Jason Thorpe wrote:
>> 
>> Well, it certainly looks implemented now (the clock parent for ehci and ohci is clock-controller@2001000, which is "allwinner,sun20i-d1-ccu?, which is sun20i_d1_ccu.c, it looks like it has everything needed, although I suppose it certainly possible that the clock/reset index constants used in the code doesn't match what the device tree is using; always a danger).
> 
> The sdmmc is working properly now, see boot.log -
> https://github.com/picohive/netbsd-mangopi-mq-pro/blob/main/boot.log
> 
> It requires
> cpufunc.c
> cpu_fdt.c.diff
> cpufunc.h.diff
> files.fdt.diff
> files.sunxi.diff
> sunxi_mmc.c.diff
> sunxi_wdt.c.diff
> GENERIC64.diff

Thanks for working on this.

Can you boot this kernel and let me know if it works - I’ve changed the cache stuff a little.

https://www.netbsd.org/~skrll/riscv64.netbsd

If it works I’ll commit.