mbox series

[0/3] clk: meson: three small clk-pll fixes

Message ID 20201226121556.975418-1-martin.blumenstingl@googlemail.com (mailing list archive)
Headers show
Series clk: meson: three small clk-pll fixes | expand

Message

Martin Blumenstingl Dec. 26, 2020, 12:15 p.m. UTC
Hi Jerome,

while working on some changes for the 32-bit SoCs I hit a corner-case
in the HDMI PLL: there's some rate doubling going. The PLL only locks
in a specific rate range but the M/N table is not aware of that. This
means for now (I am planning to fix that) that we can end up in a 
ituation where the PLL doesn't lock and meson_clk_pll_set_rate() is
supposed to still behave in this case. So here's three small patches
for that.

For me it's fine to queue these patches for -next. I am not aware of
any breakage upstream, only some of my pending patches prefer to have
these fixes.

Slightly unrelated: if you know anything about that clock doubling then
please let me know!


Best regards,
Martin


Martin Blumenstingl (3):
  clk: meson: clk-pll: fix initializing the old rate (fallback) for a
    PLL
  clk: meson: clk-pll: make "ret" a signed integer
  clk: meson: clk-pll: propagate the error from meson_clk_pll_set_rate()

 drivers/clk/meson/clk-pll.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

Comments

Jerome Brunet Jan. 4, 2021, 11:43 a.m. UTC | #1
On Sat 26 Dec 2020 at 13:15, Martin Blumenstingl <martin.blumenstingl@googlemail.com> wrote:

> Hi Jerome,
>
> while working on some changes for the 32-bit SoCs I hit a corner-case
> in the HDMI PLL: there's some rate doubling going. The PLL only locks
> in a specific rate range but the M/N table is not aware of that. This
> means for now (I am planning to fix that) that we can end up in a 
> ituation where the PLL doesn't lock and meson_clk_pll_set_rate() is
> supposed to still behave in this case. So here's three small patches
> for that.
>
> For me it's fine to queue these patches for -next. I am not aware of
> any breakage upstream, only some of my pending patches prefer to have
> these fixes.
>
> Slightly unrelated: if you know anything about that clock doubling then
> please let me know!
>
>
> Best regards,
> Martin
>
>
> Martin Blumenstingl (3):
>   clk: meson: clk-pll: fix initializing the old rate (fallback) for a
>     PLL
>   clk: meson: clk-pll: make "ret" a signed integer
>   clk: meson: clk-pll: propagate the error from meson_clk_pll_set_rate()
>
>  drivers/clk/meson/clk-pll.c | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)

Applied, Thx