mbox series

[0/3] clk: meson: clk-pll driver update

Message ID 20180717095617.12240-1-jbrunet@baylibre.com
Headers show
Series clk: meson: clk-pll driver update | expand

Message

Jerome Brunet July 17, 2018, 9:56 a.m. UTC
This patchset is yet another round of update to the amlogic pll driver.

 1) Enable bit is added so we don't rely on the bootloader or the init
    value to enable to pll device.
 2) OD post dividers are removed from the pll driver. This simplify the
    driver and let us provide the clocks which exist between those
    dividera. Some device are actually using these clocks.
 3) The rates hard coded in parameter tables are remove. Instead, we
    only rely on the parent rate and the parameters to calculate the
    output rate, which is a lot better.

This series has been tested on the gxl libretech cc and axg s400.
I did not test it on meson8b yet.

Jerome Brunet (3):
  clk: meson: clk-pll: add enable bit
  clk: meson: clk-pll: remove od parameters
  clk: meson: clk-pll: drop hard-coded rates from pll tables

 drivers/clk/meson/axg.c     | 320 ++++++++++++++++------------
 drivers/clk/meson/axg.h     |   8 +-
 drivers/clk/meson/clk-pll.c | 156 +++++++++-----
 drivers/clk/meson/clkc.h    |  16 +-
 drivers/clk/meson/gxbb.c    | 504 ++++++++++++++++++++++----------------------
 drivers/clk/meson/gxbb.h    |  10 +-
 drivers/clk/meson/meson8b.c | 167 ++++++++-------
 drivers/clk/meson/meson8b.h |   5 +-
 8 files changed, 648 insertions(+), 538 deletions(-)

-- 
2.14.4

Comments

Martin Blumenstingl July 21, 2018, 8:17 p.m. UTC | #1
Hi Jerome,

On Tue, Jul 17, 2018 at 11:56 AM Jerome Brunet <jbrunet@baylibre.com> wrote:
>

> This patchset is yet another round of update to the amlogic pll driver.

>

>  1) Enable bit is added so we don't rely on the bootloader or the init

>     value to enable to pll device.

>  2) OD post dividers are removed from the pll driver. This simplify the

>     driver and let us provide the clocks which exist between those

>     dividera. Some device are actually using these clocks.

>  3) The rates hard coded in parameter tables are remove. Instead, we

>     only rely on the parent rate and the parameters to calculate the

>     output rate, which is a lot better.

>

> This series has been tested on the gxl libretech cc and axg s400.

> I did not test it on meson8b yet.

I had some comments on patch #2
once that is fixed I can help testing on Meson8b (if you give me a few days...)


Regards
Martin
Jerome Brunet July 21, 2018, 8:48 p.m. UTC | #2
On Sat, 2018-07-21 at 22:17 +0200, Martin Blumenstingl wrote:
> > This series has been tested on the gxl libretech cc and axg s400.

> > I did not test it on meson8b yet.

> 

> I had some comments on patch #2

> once that is fixed I can help testing on Meson8b (if you give me a few days...)


I don't intend to make another PR to clk in this cycle, unless something
critical comes up. No rush.