[PATCH 0/4] clk: sunxi-ng: a523: Fix single-divider clocks

Chen-Yu Tsai wens at kernel.org
Sun Sep 13 09:19:15 PDT 2026


On Mon, Sep 14, 2026 at 12:16 AM Chen-Yu Tsai <wens at kernel.org> wrote:
>
> On Sun, 06 Sep 2026 15:06:37 +0800, Junhui Liu wrote:
> > Sashiko pointed out two issues in my series adding clock driver support
> > for the Allwinner A733 SoC [1][2].
> >
> > First, ccu_mp_set_rate() generates an invalid mask when an MP clock is
> > defined with a zero-width P divider, clearing unrelated register fields.
> > Second, ccu_mp_set_rate() does not handle CCU_FEATURE_UPDATE_BIT, so
> > divider changes are not latched by the hardware.
> >
> > [...]
>
> Applied to sunxi/clk-for-7.4 in sunxi, thanks!
>
> [1/4] clk: sunxi-ng: div: Add feature support to M clock macro
>       https://git.kernel.org/sunxi/linux/c/da543a093f65
> [2/4] clk: sunxi-ng: sun55i-a523-ccu: Use M-only clocks for MBUS, IOMMU and DRAM
>       https://git.kernel.org/sunxi/linux/c/88735a2ffbea
> [3/4] clk: sunxi-ng: sun55i-a523-ccu: Use P-only clocks for high-speed timers
>       https://git.kernel.org/sunxi/linux/c/4b41d0ffa9d9
> [4/4] clk: sunxi-ng: sun55i-a523-r-ccu: Use P-only clocks for timers
>       https://git.kernel.org/sunxi/linux/c/b260db91603a

After pushing things out, I realized that I based the topic branch on the
wrong base commit. I rebased things and forced pushed, so the commit hashes
here won't be correct.

Rest assured that things have indeed been merged.

> Best regards,
> --
> Chen-Yu Tsai <wens at kernel.org>
>
>



More information about the linux-arm-kernel mailing list