Re: [PATCH v6 6/8] clk: sunxi-ng: a733: Add mod clocks support

[Andre Przywara <andre.przywara@arm.com>]

Hi Vinicius,

On 10/28/26 19:09, Vinicius Pedrosa wrote:
> Hi Junhui,
> 
> In patch 6/8, usb2-u2-ref is declared as a plain mux, but the manual

can you please reply inline? Copy the original email, then reply after 
the code that you have a comment on? See the other replies on this 
series for examples.
After all this is the reason we do this patch-inline-in-email approach 
for review, so that the context is clear, and makes it easier for the 
author, other reviewers and maintainers to follow the review comments.

Subscribing to the linux-sunxi list is probably the easiest to get the 
emails, or use lore to fetch the respective emails. For one-offs you can 
also download the mbox file from lore, then import them in your email 
application.

Cheers,
Andre

P.S. This applies in general to your email replies, I see them as 
completely from scratch text, just with the right In-reply-to header, 
but without any context.

> has a gate at bit 31 of 0x1348, and its reset value is 0. So the CCU
> driver can't enable it.
> 
> -static SUNXI_CCU_MUX_DATA(usb2_u2_ref_clk, "usb2-u2-ref", usb01_ref_parents, 0x1348, 24, 3, 0);
> +static SUNXI_CCU_MUX_DATA_WITH_GATE(usb2_u2_ref_clk, "usb2-u2-ref", usb01_ref_parents,
> +                                    0x1348, 24, 3, BIT(31), 0);
> 
> I hit this on my Cubie A7Z with a USB stick on the DWC3 port (the DT
> and the USB2 PHY driver for that port are local, not posted):
> 
> - v6 as posted: 0x02003348 stays 0 and the DWC3 soft reset times out
>    (-110), 4/4 boots, one of them cold.
> - same kernel, only bit 31 set from U-Boot: the host works, 3/3.
> - with the change above: works, 3/3.
> 
> Same DTB and initramfs in every run. The SerDes gates were also the
> same, so I haven't checked whether they matter on their own.
> 
> My boot chain (U-Boot SPL, TF-A, U-Boot) leaves bit 31 clear. I don't
> know what the vendor one does.
> 
> Thanks,
> Vinicius