mirror of https://lore.kernel.org/linux-amlogic/
 help / color / mirror / Atom feed
From: mturquette@baylibre.com (Michael Turquette)
To: linus-amlogic@lists.infradead.org
Subject: [PATCH 0/7] Add support for AmLogic GXBB clock controller
Date: Tue, 14 Jun 2016 18:41:03 -0700	[thread overview]
Message-ID: <20160615014103.5467.86678@quark.deferred.io> (raw)
In-Reply-To: <m2d1njsikh.fsf@baylibre.com>

Hi Kevin,

Quoting Kevin Hilman (2016-06-14 11:59:10)
> Hi Mike,
> 
> Michael Turquette <mturquette@baylibre.com> writes:
> 
> > This series is based on the AmLogic Meson8b rewrite/cleanup[0]. The
> > AmLogic GXBB is an ARMv8-based SoC, fed by a 24MHZ xtal, and it provides
> > several PLLs, muxes dividers and gates to drive CPUs and peripherals.
> >
> > While based on the Meson8b clock controller driver, this series adds
> > supports for Multi-phase Locked Loops (mpll) and support for PLLs with
> > fractional rates.
> >
> > This series introduces the clock controller driver for the gxbb,
> > including the DT binding description and accompanying change in the gxbb
> > dtsi. Only three clocks are exposed via the headers in the dt-bindings
> > include-chroot. More can be added later, but since these values
> > represent an ABI I wanted to start small. I'm not sure about the names
> > for the gate clocks, so I especially tried to avoid putting those in the
> > binding (with the exception of the Ethernet gate, which Kevin has needed
> > for a while).
> 
> I tested this on top of my integration branch for v4.7-rc[1] along with
> the ethernet driver.
> 
> First, I modified the ethernet node to use the new clock[2], and then I
> added the ethernet driver to the build as modules[3] since
> unloading/reloading the module is testing the clock API[3].
> 
> Then, I tested unloading/reloading the dwmac-meson module on
> meson-gxbb-p200 and meson-gxbb-odroidc2 and found that the system hung
> on module reload.
> 
> As you suggested off-list, this might be because it's walking up the
> tree and disabling clk81, which would cause such a problem, so I added
> CLK_IS_CRITICAL to clk81, and that got things working just fine.

Great! I'm glad that fixed the problem.

> 
> With that change, I can now unload/reload the ethernet module (which is
> prepare/enable'ing and unprepare/disable'ing the clocks.
> 
> Tested-by: Kevin Hilman <khilman@baylibre.com>

Cool, thanks for testing. I'll fold your change to set the
CLK_IS_CRITICAL flag in below on v2, which I'll push to the list early
next week after I return from travel.

Regards,
Mike

> 
> Kevin
> 
> [1] https://git.kernel.org/cgit/linux/kernel/git/khilman/linux-amlogic.git/log/?h=v4.7/integ
> [2]
> diff --git a/arch/arm64/boot/dts/amlogic/meson-gxbb.dtsi b/arch/arm64/boot/dts/amlogic/meson-gxbb.dtsi
> index e3210eb89bc8..834258a36cfb 100644
> --- a/arch/arm64/boot/dts/amlogic/meson-gxbb.dtsi
> +++ b/arch/arm64/boot/dts/amlogic/meson-gxbb.dtsi
> @@ -45,6 +45,7 @@
>  #include <dt-bindings/interrupt-controller/arm-gic.h>
>  #include <dt-bindings/gpio/meson-gxbb-gpio.h>
>  #include <dt-bindings/reset/amlogic,meson-gxbb-reset.h>
> +#include <dt-bindings/clock/gxbb-clkc.h>
>  
>  / {
>         compatible = "amlogic,meson-gxbb";
> @@ -332,7 +333,7 @@
>                                0x0 0xc8834540 0x0 0x4>;
>                         interrupts = <0 8 1>;
>                         interrupt-names = "macirq";
> -                       clocks = <&xtal>;
> +                       clocks = <&clkc CLKID_ETH>;
>                         clock-names = "stmmaceth";
>                         phy-mode = "rgmii";
>                         status = "disabled";
> 
> [3]
> CONFIG_STMMAC_ETH=m
> CONFIG_PHYLIB=m
> CONFIG_MICREL_PHY=m
> CONFIG_REALTEK_PHY=m
> 
> [4]
> diff --git a/drivers/clk/meson/gxbb.c b/drivers/clk/meson/gxbb.c
> index 26c78a2a614d..007b7157cf4b 100644
> --- a/drivers/clk/meson/gxbb.c
> +++ b/drivers/clk/meson/gxbb.c
> @@ -558,7 +558,7 @@ static struct clk_gate gxbb_clk81 = {
>                 .ops = &clk_gate_ops,
>                 .parent_names = (const char *[]){ "mpeg_clk_div" },
>                 .num_parents = 1,
> -               .flags = (CLK_SET_RATE_PARENT | CLK_IGNORE_UNUSED),
> +               .flags = (CLK_SET_RATE_PARENT | CLK_IGNORE_UNUSED | CLK_IS_CRITICAL),
>         },
>  };

      reply	other threads:[~2016-06-15  1:41 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-06-10  0:32 Michael Turquette
2016-06-10  0:32 ` [PATCH 1/7] clk: meson: add peripheral gate macro Michael Turquette
2016-06-10  0:32 ` [PATCH 2/7] clk: meson: add mpll support Michael Turquette
2016-06-10  0:32 ` [PATCH 3/7] clk: meson: fractional pll support Michael Turquette
2016-06-10  0:32 ` [PATCH 4/7] clk: gxbb: Document bindings for the GXBB clock controller Michael Turquette
2016-06-10  0:32 ` [PATCH 5/7] clk: gxbb: add AmLogic GXBB clk controller driver Michael Turquette
2016-06-10  0:32 ` [PATCH 6/7] arm64: amlogic: select gxbb clk driver Michael Turquette
2016-06-10  0:32 ` [PATCH 7/7] arm64: dts: gxbb clock controller Michael Turquette
2016-06-14 19:01   ` Kevin Hilman
2016-06-15  1:41     ` Michael Turquette
2016-06-10  9:12 ` [PATCH 0/7] Add support for AmLogic GXBB " Neil Armstrong
2016-06-13 17:55   ` Kevin Hilman
2016-06-14  5:02     ` Neil Armstrong
2016-06-14 18:59 ` Kevin Hilman
2016-06-15  1:41   ` Michael Turquette [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20160615014103.5467.86678@quark.deferred.io \
    --to=mturquette@baylibre.com \
    --cc=linus-amlogic@lists.infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®