Re: [PATCH] clk: qcom: clk-alpha-pll: Fix the L value of Zonda PLLs with a negative alpha

[Konrad Dybcio]

On 11/3/26 10:48 AM, Abel Vesa wrote:
> The alpha value of the Zonda PLLs is signed, so for a rate with a
> fractional part of one half or more, clk_zonda_pll_set_rate() programs
> L as the integer part rounded up, through zonda_pll_adjust_l_val().
> 
> Since commit a4d89b11aca3 ("clk: qcom: clk-alpha-pll: Simplify the
> zonda_pll_adjust_l_val()"), that function adds the rounding to the
> requested rate instead of to its quotient by the parent rate, so such a
> rate change writes the rate in Hz, truncated to 32 bits, into the L
> register.
> 
> Use the quotient again.
> 
> Fixes: a4d89b11aca3 ("clk: qcom: clk-alpha-pll: Simplify the zonda_pll_adjust_l_val()")
> Cc: stable@xxxxxxxxxxxxxxx
> Signed-off-by: Abel Vesa <abel.vesa@xxxxxxxxxxxxxxxx>
> ---
> Prepared with help from an LLM-based assistant.
> ---

https://lore.kernel.org/linux-arm-msm/20261027-nords_dprx-v3-1-5d98b2c6c263@xxxxxxxxxxxxxxxx/

Konrad