On Sat, 2018-07-21 at 21:48 +0200, Martin Blumenstingl wrote:
> > @@ -250,11 +264,15 @@ static const struct reg_sequence axg_hifi_init_regs[]
> > = {
> > { .reg = HHI_HIFI_PLL_CNTL3, .def = 0x0a6a3a88 },
> > { .reg = HHI_HIFI_PLL_CNTL4, .def = 0xc000004d },
> > { .reg = HHI_HIFI_PLL_CNTL5, .def = 0x00058000 },
> > - { .reg = HHI_HIFI_PLL_CNTL, .def = 0x40010250 },
>
> is this change on purpose? this line set en, m, n, l and od before
> maybe you can document it in the commit message
Yes the change is on purpose, but as you pointed out it is worth a comment
Actually, when taking od out of the pll driver, I remembered this 'initial
setting' and it kinda bothered me
If the od divider registers after the DCO, the value could have changed with it,
which is why I wanted to remove the write on this register
As you pointed out, in this register, we find m, n, od ... and enable. In a way,
removing this register setting was the reason why I wanted to add the enable bit
to begin with :)
