From: Michael Turquette <mturquette@baylibre.com>
To: Tony Lindgren <tony@atomide.com>,
"Matthijs van Duin" <matthijsvanduin@gmail.com>
Cc: "Stephen Boyd" <sboyd@codeaurora.org>,
"Tero Kristo" <t-kristo@ti.com>,
linux-clk@vger.kernel.org,
"linux-omap@vger.kernel.org" <linux-omap@vger.kernel.org>,
"linux-arm-kernel@lists.infradead.org"
<linux-arm-kernel@lists.infradead.org>,
"Brian Hutchinson" <b.hutchman@gmail.com>,
"Delio Brignoli" <dbrignoli@audioscience.com>,
"Neil Armstrong" <narmstrong@baylibre.com>,
"Philipp Rosenberger" <ilu@linutronix.de>,
"Russell King - ARM Linux" <linux@arm.linux.org.uk>
Subject: Re: [PATCH 2/2] clk: ti: Add support for dm814x ADPLL
Date: Mon, 29 Feb 2016 14:19:58 -0800 [thread overview]
Message-ID: <20160229221958.3140.29381@quark.deferred.io> (raw)
In-Reply-To: <20160226173505.GA13417@atomide.com>
Quoting Tony Lindgren (2016-02-26 09:35:05)
> * Matthijs van Duin <matthijsvanduin@gmail.com> [160223 05:48]:
> > On 17 February 2016 at 22:20, Tony Lindgren <tony@atomide.com> wrote:
> > >* Michael Turquette <mturquette@baylibre.com> [160217 12:53]:
> > >> Do you think you will ever have a driver that wants to gate these
> > >> clocks? Probably not for MPU/DDR, but the HAND_OFF flag is a better =
fit
> > >> if you do. It's the same as CRITICAL but transfers the prepare/enable
> > >> reference counting to the consumer driver after that driver calls
> > >> clk_get() and clk_prepare_enable() on the HAND_OFF clk.
> > >
> > > OK I'll use CLK_ENABLE_HAND_OFF then. Who knows maybe somebody
> > > somewhere has code running on one of the coprocessors in SRAM that
> > > actually allows gating these :)
> =
> Mike, as I mentioned, looks like the CLK_ENABLE_HAND_OFF series needs
> some changes for the warnings and to avoid hang on rmmod. So below
> is the adpll driver with the binding as a single patch and COMPILE_TEST
> enabled so we avoid the dependency to CLK_ENABLE_HAND_OFF series.
> =
> I'll submit a minimal follow-up patch for CLK_ENABLE_HAND_OFF
> later as we're getting quite close to the merge window.
> =
> I will also send a separate patch to allow COMPILE_TEST to properly
> build for drivers/clk/ti.
> =
> > An arrangement somewhat similar to the one the am335x has should
> > indeed be doable, but note that it would be the coprocessor that would
> > gate and ungate the clocks, not linux. The L3 clock is actually even
> > more essential and can never be gated at all.
> > =
> > In the absence of help from a coprocessor linux could still put the L3
> > and MPU PLLs into bypass for power saving, and DDR can be gated if
> > it's first put into self-refresh and helper code in on-chip SRAM is
> > used to manage the transitions.
> > =
> > On 18 February 2016 at 00:02, Michael Turquette <mturquette@baylibre.co=
m> wrote:
> > > Low power mp3 decoder executing out of ARM icache with DDR clock cut =
;-)
> > =
> > L1 cache is rather tricky since you can't lock it. You can however
> > lock 448 KB of the 512 KB L2 cache, which should be plenty of space.
> > There's also quite a bit of SRAM on chip usable for buffering data.
> > =
> > The C674x DSP would probably also be a good candidate for such an appli=
cation.
> =
> Sounds doable, the CLK_ENABLE_HAND_OFF won't matter in that case
> though :)
> =
> Regards,
> =
> Tony
> =
> 8< ----------------------
> From: Tony Lindgren <tony@atomide.com>
> Date: Mon, 22 Feb 2016 14:15:37 -0800
> Subject: [PATCH] clk: ti: Add support for dm814x ADPLL
> =
> On dm814x we have 13 ADPLLs with 3 to 4 outputs on each. The
> ADPLLs have several dividers and muxes controlled by a shared
> control register for each PLL.
> =
> Note that for the clocks to work as device drivers for booting on
> dm814x, this patch depends on "ARM: OMAP2+: Change core_initcall
> levels to postcore_initcall" that has already been merged.
> =
> Also note that this patch does not implement clk_set_rate for the
> PLL, that will be posted later on when available.
> =
> Cc: Michael Turquette <mturquette@baylibre.com>
> Cc: Stephen Boyd <sboyd@codeaurora.org>
> Acked-by: Tero Kristo <t-kristo@ti.com>
> Signed-off-by: Tony Lindgren <tony@atomide.com>
Applied.
Thanks,
Mike
> =
> diff --git a/Documentation/devicetree/bindings/clock/ti/adpll.txt b/Docum=
entation/devicetree/bindings/clock/ti/adpll.txt
> new file mode 100644
> index 0000000..4c8a2ce
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/clock/ti/adpll.txt
> @@ -0,0 +1,41 @@
> +Binding for Texas Instruments ADPLL clock.
> +
> +Binding status: Unstable - ABI compatibility may be broken in the future
> +
> +This binding uses the common clock binding[1]. It assumes a
> +register-mapped ADPLL with two to three selectable input clocks
> +and three to four children.
> +
> +[1] Documentation/devicetree/bindings/clock/clock-bindings.txt
> +
> +Required properties:
> +- compatible : shall be one of "ti,dm814-adpll-s-clock" or
> + "ti,dm814-adpll-lj-clock" depending on the type of the ADPLL
> +- #clock-cells : from common clock binding; shall be set to 1.
> +- clocks : link phandles of parent clocks clkinp and clkinpulow, note
> + that the adpll-s-clock also has an optional clkinphif
> +- reg : address and length of the register set for controlling the ADPLL.
> +
> +Examples:
> + adpll_mpu_ck: adpll@40 {
> + #clock-cells =3D <1>;
> + compatible =3D "ti,dm814-adpll-s-clock";
> + reg =3D <0x40 0x40>;
> + clocks =3D <&devosc_ck &devosc_ck &devosc_ck>;
> + clock-names =3D "clkinp", "clkinpulow", "clkinphif";
> + clock-output-names =3D "481c5040.adpll.dcoclkldo",
> + "481c5040.adpll.clkout",
> + "481c5040.adpll.clkoutx2",
> + "481c5040.adpll.clkouthif";
> + };
> +
> + adpll_dsp_ck: adpll@80 {
> + #clock-cells =3D <1>;
> + compatible =3D "ti,dm814-adpll-lj-clock";
> + reg =3D <0x80 0x30>;
> + clocks =3D <&devosc_ck &devosc_ck>;
> + clock-names =3D "clkinp", "clkinpulow";
> + clock-output-names =3D "481c5080.adpll.dcoclkldo",
> + "481c5080.adpll.clkout",
> + "481c5080.adpll.clkoutldo";
> + };
> diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig
> index eca8e01..0b7364a 100644
> --- a/drivers/clk/Kconfig
> +++ b/drivers/clk/Kconfig
> @@ -202,6 +202,7 @@ config COMMON_CLK_CDCE706
> =
> source "drivers/clk/bcm/Kconfig"
> source "drivers/clk/hisilicon/Kconfig"
> +source "drivers/clk/ti/Kconfig"
> source "drivers/clk/qcom/Kconfig"
> =
> endmenu
> diff --git a/drivers/clk/ti/Kconfig b/drivers/clk/ti/Kconfig
> new file mode 100644
> index 0000000..2713417
> --- /dev/null
> +++ b/drivers/clk/ti/Kconfig
> @@ -0,0 +1,6 @@
> +config COMMON_CLK_TI_ADPLL
> + tristate "Clock driver for dm814x ADPLL"
> + depends on ARCH_OMAP2PLUS || COMPILE_TEST
> + default y if SOC_TI81XX
> + ---help---
> + ADPLL clock driver for the dm814x SoC using common clock framew=
ork.
> diff --git a/drivers/clk/ti/Makefile b/drivers/clk/ti/Makefile
> index d4ac960..dfe91d7 100644
> --- a/drivers/clk/ti/Makefile
> +++ b/drivers/clk/ti/Makefile
> @@ -18,3 +18,5 @@ obj-$(CONFIG_SOC_AM43XX) +=3D $(clk-common=
) dpll3xxx.o clk-43xx.o
> ifdef CONFIG_ATAGS
> obj-$(CONFIG_ARCH_OMAP3) +=3D clk-3xxx-legacy.o
> endif
> +
> +obj-$(CONFIG_COMMON_CLK_TI_ADPLL) +=3D adpll.o
> diff --git a/drivers/clk/ti/adpll.c b/drivers/clk/ti/adpll.c
> new file mode 100644
> index 0000000..f741d79
> --- /dev/null
> +++ b/drivers/clk/ti/adpll.c
> @@ -0,0 +1,983 @@
> +/*
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License as
> + * published by the Free Software Foundation version 2.
> + *
> + * This program is distributed "as is" WITHOUT ANY WARRANTY of any
> + * kind, whether express or implied; without even the implied warranty
> + * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + * GNU General Public License for more details.
> + */
> +
> +#include <linux/clk.h>
> +#include <linux/clkdev.h>
> +#include <linux/clk-provider.h>
> +#include <linux/delay.h>
> +#include <linux/err.h>
> +#include <linux/math64.h>
> +#include <linux/module.h>
> +#include <linux/of_device.h>
> +#include <linux/string.h>
> +
> +#define ADPLL_PLLSS_MMR_LOCK_OFFSET 0x00 /* Managed by MPPULL */
> +#define ADPLL_PLLSS_MMR_LOCK_ENABLED 0x1f125B64
> +#define ADPLL_PLLSS_MMR_UNLOCK_MAGIC 0x1eda4c3d
> +
> +#define ADPLL_PWRCTRL_OFFSET 0x00
> +#define ADPLL_PWRCTRL_PONIN 5
> +#define ADPLL_PWRCTRL_PGOODIN 4
> +#define ADPLL_PWRCTRL_RET 3
> +#define ADPLL_PWRCTRL_ISORET 2
> +#define ADPLL_PWRCTRL_ISOSCAN 1
> +#define ADPLL_PWRCTRL_OFFMODE 0
> +
> +#define ADPLL_CLKCTRL_OFFSET 0x04
> +#define ADPLL_CLKCTRL_CLKDCOLDOEN 29
> +#define ADPLL_CLKCTRL_IDLE 23
> +#define ADPLL_CLKCTRL_CLKOUTEN 20
> +#define ADPLL_CLKINPHIFSEL_ADPLL_S 19 /* REVISIT: which bit? */
> +#define ADPLL_CLKCTRL_CLKOUTLDOEN_ADPLL_LJ 19
> +#define ADPLL_CLKCTRL_ULOWCLKEN 18
> +#define ADPLL_CLKCTRL_CLKDCOLDOPWDNZ 17
> +#define ADPLL_CLKCTRL_M2PWDNZ 16
> +#define ADPLL_CLKCTRL_M3PWDNZ_ADPLL_S 15
> +#define ADPLL_CLKCTRL_LOWCURRSTDBY_ADPLL_S 13
> +#define ADPLL_CLKCTRL_LPMODE_ADPLL_S 12
> +#define ADPLL_CLKCTRL_REGM4XEN_ADPLL_S 10
> +#define ADPLL_CLKCTRL_SELFREQDCO_ADPLL_LJ 10
> +#define ADPLL_CLKCTRL_TINITZ 0
> +
> +#define ADPLL_TENABLE_OFFSET 0x08
> +#define ADPLL_TENABLEDIV_OFFSET 0x8c
> +
> +#define ADPLL_M2NDIV_OFFSET 0x10
> +#define ADPLL_M2NDIV_M2 16
> +#define ADPLL_M2NDIV_M2_ADPLL_S_WIDTH 5
> +#define ADPLL_M2NDIV_M2_ADPLL_LJ_WIDTH 7
> +
> +#define ADPLL_MN2DIV_OFFSET 0x14
> +#define ADPLL_MN2DIV_N2 16
> +
> +#define ADPLL_FRACDIV_OFFSET 0x18
> +#define ADPLL_FRACDIV_REGSD 24
> +#define ADPLL_FRACDIV_FRACTIONALM 0
> +#define ADPLL_FRACDIV_FRACTIONALM_MASK 0x3ffff
> +
> +#define ADPLL_BWCTRL_OFFSET 0x1c
> +#define ADPLL_BWCTRL_BWCONTROL 1
> +#define ADPLL_BWCTRL_BW_INCR_DECRZ 0
> +
> +#define ADPLL_RESERVED_OFFSET 0x20
> +
> +#define ADPLL_STATUS_OFFSET 0x24
> +#define ADPLL_STATUS_PONOUT 31
> +#define ADPLL_STATUS_PGOODOUT 30
> +#define ADPLL_STATUS_LDOPWDN 29
> +#define ADPLL_STATUS_RECAL_BSTATUS3 28
> +#define ADPLL_STATUS_RECAL_OPPIN 27
> +#define ADPLL_STATUS_PHASELOCK 10
> +#define ADPLL_STATUS_FREQLOCK 9
> +#define ADPLL_STATUS_BYPASSACK 8
> +#define ADPLL_STATUS_LOSSREF 6
> +#define ADPLL_STATUS_CLKOUTENACK 5
> +#define ADPLL_STATUS_LOCK2 4
> +#define ADPLL_STATUS_M2CHANGEACK 3
> +#define ADPLL_STATUS_HIGHJITTER 1
> +#define ADPLL_STATUS_BYPASS 0
> +#define ADPLL_STATUS_PREPARED_MASK (BIT(ADPLL_STATUS_PHASELOCK) | \
> + BIT(ADPLL_STATUS_FREQLOCK))
> +
> +#define ADPLL_M3DIV_OFFSET 0x28 /* Only on MPUPLL */
> +#define ADPLL_M3DIV_M3 0
> +#define ADPLL_M3DIV_M3_WIDTH 5
> +#define ADPLL_M3DIV_M3_MASK 0x1f
> +
> +#define ADPLL_RAMPCTRL_OFFSET 0x2c /* Only on MPUPLL */
> +#define ADPLL_RAMPCTRL_CLKRAMPLEVEL 19
> +#define ADPLL_RAMPCTRL_CLKRAMPRATE 16
> +#define ADPLL_RAMPCTRL_RELOCK_RAMP_EN 0
> +
> +#define MAX_ADPLL_INPUTS 3
> +#define MAX_ADPLL_OUTPUTS 4
> +#define ADPLL_MAX_RETRIES 5
> +
> +#define to_dco(_hw) container_of(_hw, struct ti_adpll_dco_data, hw)
> +#define to_adpll(_hw) container_of(_hw, struct ti_adpll_data, dco)
> +#define to_clkout(_hw) container_of(_hw, struct ti_adpll_clkout_data, hw)
> +
> +enum ti_adpll_clocks {
> + TI_ADPLL_DCO,
> + TI_ADPLL_DCO_GATE,
> + TI_ADPLL_N2,
> + TI_ADPLL_M2,
> + TI_ADPLL_M2_GATE,
> + TI_ADPLL_BYPASS,
> + TI_ADPLL_HIF,
> + TI_ADPLL_DIV2,
> + TI_ADPLL_CLKOUT,
> + TI_ADPLL_CLKOUT2,
> + TI_ADPLL_M3,
> +};
> +
> +#define TI_ADPLL_NR_CLOCKS (TI_ADPLL_M3 + 1)
> +
> +enum ti_adpll_inputs {
> + TI_ADPLL_CLKINP,
> + TI_ADPLL_CLKINPULOW,
> + TI_ADPLL_CLKINPHIF,
> +};
> +
> +enum ti_adpll_s_outputs {
> + TI_ADPLL_S_DCOCLKLDO,
> + TI_ADPLL_S_CLKOUT,
> + TI_ADPLL_S_CLKOUTX2,
> + TI_ADPLL_S_CLKOUTHIF,
> +};
> +
> +enum ti_adpll_lj_outputs {
> + TI_ADPLL_LJ_CLKDCOLDO,
> + TI_ADPLL_LJ_CLKOUT,
> + TI_ADPLL_LJ_CLKOUTLDO,
> +};
> +
> +struct ti_adpll_platform_data {
> + const bool is_type_s;
> + const int nr_max_inputs;
> + const int nr_max_outputs;
> + const int output_index;
> +};
> +
> +struct ti_adpll_clock {
> + struct clk *clk;
> + struct clk_lookup *cl;
> + void (*unregister)(struct clk *clk);
> +};
> +
> +struct ti_adpll_dco_data {
> + struct clk_hw hw;
> +};
> +
> +struct ti_adpll_clkout_data {
> + struct ti_adpll_data *adpll;
> + struct clk_gate gate;
> + struct clk_hw hw;
> +};
> +
> +struct ti_adpll_data {
> + struct device *dev;
> + const struct ti_adpll_platform_data *c;
> + struct device_node *np;
> + unsigned long pa;
> + void __iomem *iobase;
> + void __iomem *regs;
> + spinlock_t lock; /* For ADPLL shared register access */
> + const char *parent_names[MAX_ADPLL_INPUTS];
> + struct clk *parent_clocks[MAX_ADPLL_INPUTS];
> + struct ti_adpll_clock *clocks;
> + struct clk_onecell_data outputs;
> + struct ti_adpll_dco_data dco;
> +};
> +
> +static const char *ti_adpll_clk_get_name(struct ti_adpll_data *d,
> + int output_index,
> + const char *postfix)
> +{
> + const char *name;
> + int err;
> +
> + if (output_index >=3D 0) {
> + err =3D of_property_read_string_index(d->np,
> + "clock-output-names",
> + output_index,
> + &name);
> + if (err)
> + return NULL;
> + } else {
> + const char *base_name =3D "adpll";
> + char *buf;
> +
> + buf =3D devm_kzalloc(d->dev, 8 + 1 + strlen(base_name) + =
1 +
> + strlen(postfix), GFP_KERNEL);
> + if (!buf)
> + return NULL;
> + sprintf(buf, "%08lx.%s.%s", d->pa, base_name, postfix);
> + name =3D buf;
> + }
> +
> + return name;
> +}
> +
> +#define ADPLL_MAX_CON_ID 16 /* See MAX_CON_ID */
> +
> +static int ti_adpll_setup_clock(struct ti_adpll_data *d, struct clk *clo=
ck,
> + int index, int output_index, const char *=
name,
> + void (*unregister)(struct clk *clk))
> +{
> + struct clk_lookup *cl;
> + const char *postfix =3D NULL;
> + char con_id[ADPLL_MAX_CON_ID];
> +
> + d->clocks[index].clk =3D clock;
> + d->clocks[index].unregister =3D unregister;
> +
> + /* Separate con_id in format "pll040dcoclkldo" to fit MAX_CON_ID =
*/
> + postfix =3D strrchr(name, '.');
> + if (strlen(postfix) > 1) {
> + if (strlen(postfix) > ADPLL_MAX_CON_ID)
> + dev_warn(d->dev, "clock %s con_id lookup may fail=
\n",
> + name);
> + snprintf(con_id, 16, "pll%03lx%s", d->pa & 0xfff, postfix=
+ 1);
> + cl =3D clkdev_create(clock, con_id, NULL);
> + if (!cl)
> + return -ENOMEM;
> + d->clocks[index].cl =3D cl;
> + } else {
> + dev_warn(d->dev, "no con_id for clock %s\n", name);
> + }
> +
> + if (output_index < 0)
> + return 0;
> +
> + d->outputs.clks[output_index] =3D clock;
> + d->outputs.clk_num++;
> +
> + return 0;
> +}
> +
> +static int ti_adpll_init_divider(struct ti_adpll_data *d,
> + enum ti_adpll_clocks index,
> + int output_index, char *name,
> + struct clk *parent_clock,
> + void __iomem *reg,
> + u8 shift, u8 width,
> + u8 clk_divider_flags)
> +{
> + const char *child_name;
> + const char *parent_name;
> + struct clk *clock;
> +
> + child_name =3D ti_adpll_clk_get_name(d, output_index, name);
> + if (!child_name)
> + return -EINVAL;
> +
> + parent_name =3D __clk_get_name(parent_clock);
> + clock =3D clk_register_divider(d->dev, child_name, parent_name, 0,
> + reg, shift, width, clk_divider_flags,
> + &d->lock);
> + if (IS_ERR(clock)) {
> + dev_err(d->dev, "failed to register divider %s: %li\n",
> + name, PTR_ERR(clock));
> + return PTR_ERR(clock);
> + }
> +
> + return ti_adpll_setup_clock(d, clock, index, output_index, child_=
name,
> + clk_unregister_divider);
> +}
> +
> +static int ti_adpll_init_mux(struct ti_adpll_data *d,
> + enum ti_adpll_clocks index,
> + char *name, struct clk *clk0,
> + struct clk *clk1,
> + void __iomem *reg,
> + u8 shift)
> +{
> + const char *child_name;
> + const char *parents[2];
> + struct clk *clock;
> +
> + child_name =3D ti_adpll_clk_get_name(d, -ENODEV, name);
> + if (!child_name)
> + return -ENOMEM;
> + parents[0] =3D __clk_get_name(clk0);
> + parents[1] =3D __clk_get_name(clk1);
> + clock =3D clk_register_mux(d->dev, child_name, parents, 2, 0,
> + reg, shift, 1, 0, &d->lock);
> + if (IS_ERR(clock)) {
> + dev_err(d->dev, "failed to register mux %s: %li\n",
> + name, PTR_ERR(clock));
> + return PTR_ERR(clock);
> + }
> +
> + return ti_adpll_setup_clock(d, clock, index, -ENODEV, child_name,
> + clk_unregister_mux);
> +}
> +
> +static int ti_adpll_init_gate(struct ti_adpll_data *d,
> + enum ti_adpll_clocks index,
> + int output_index, char *name,
> + struct clk *parent_clock,
> + void __iomem *reg,
> + u8 bit_idx,
> + u8 clk_gate_flags)
> +{
> + const char *child_name;
> + const char *parent_name;
> + struct clk *clock;
> +
> + child_name =3D ti_adpll_clk_get_name(d, output_index, name);
> + if (!child_name)
> + return -EINVAL;
> +
> + parent_name =3D __clk_get_name(parent_clock);
> + clock =3D clk_register_gate(d->dev, child_name, parent_name, 0,
> + reg, bit_idx, clk_gate_flags,
> + &d->lock);
> + if (IS_ERR(clock)) {
> + dev_err(d->dev, "failed to register gate %s: %li\n",
> + name, PTR_ERR(clock));
> + return PTR_ERR(clock);
> + }
> +
> + return ti_adpll_setup_clock(d, clock, index, output_index, child_=
name,
> + clk_unregister_gate);
> +}
> +
> +static int ti_adpll_init_fixed_factor(struct ti_adpll_data *d,
> + enum ti_adpll_clocks index,
> + char *name,
> + struct clk *parent_clock,
> + unsigned int mult,
> + unsigned int div)
> +{
> + const char *child_name;
> + const char *parent_name;
> + struct clk *clock;
> +
> + child_name =3D ti_adpll_clk_get_name(d, -ENODEV, name);
> + if (!child_name)
> + return -ENOMEM;
> +
> + parent_name =3D __clk_get_name(parent_clock);
> + clock =3D clk_register_fixed_factor(d->dev, child_name, parent_na=
me,
> + 0, mult, div);
> + if (IS_ERR(clock))
> + return PTR_ERR(clock);
> +
> + return ti_adpll_setup_clock(d, clock, index, -ENODEV, child_name,
> + clk_unregister);
> +}
> +
> +static void ti_adpll_set_idle_bypass(struct ti_adpll_data *d)
> +{
> + unsigned long flags;
> + u32 v;
> +
> + spin_lock_irqsave(&d->lock, flags);
> + v =3D readl_relaxed(d->regs + ADPLL_CLKCTRL_OFFSET);
> + v |=3D BIT(ADPLL_CLKCTRL_IDLE);
> + writel_relaxed(v, d->regs + ADPLL_CLKCTRL_OFFSET);
> + spin_unlock_irqrestore(&d->lock, flags);
> +}
> +
> +static void ti_adpll_clear_idle_bypass(struct ti_adpll_data *d)
> +{
> + unsigned long flags;
> + u32 v;
> +
> + spin_lock_irqsave(&d->lock, flags);
> + v =3D readl_relaxed(d->regs + ADPLL_CLKCTRL_OFFSET);
> + v &=3D ~BIT(ADPLL_CLKCTRL_IDLE);
> + writel_relaxed(v, d->regs + ADPLL_CLKCTRL_OFFSET);
> + spin_unlock_irqrestore(&d->lock, flags);
> +}
> +
> +static bool ti_adpll_clock_is_bypass(struct ti_adpll_data *d)
> +{
> + u32 v;
> +
> + v =3D readl_relaxed(d->regs + ADPLL_STATUS_OFFSET);
> +
> + return v & BIT(ADPLL_STATUS_BYPASS);
> +}
> +
> +/*
> + * Locked and bypass are not actually mutually exclusive: if you only c=
are
> + * about the DCO clock and not CLKOUT you can clear M2PWDNZ before enabl=
ing
> + * the PLL, resulting in status (FREQLOCK | PHASELOCK | BYPASS) after lo=
ck.
> + */
> +static bool ti_adpll_is_locked(struct ti_adpll_data *d)
> +{
> + u32 v =3D readl_relaxed(d->regs + ADPLL_STATUS_OFFSET);
> +
> + return (v & ADPLL_STATUS_PREPARED_MASK) =3D=3D ADPLL_STATUS_PREPA=
RED_MASK;
> +}
> +
> +static int ti_adpll_wait_lock(struct ti_adpll_data *d)
> +{
> + int retries =3D ADPLL_MAX_RETRIES;
> +
> + do {
> + if (ti_adpll_is_locked(d))
> + return 0;
> + usleep_range(200, 300);
> + } while (retries--);
> +
> + dev_err(d->dev, "pll failed to lock\n");
> + return -ETIMEDOUT;
> +}
> +
> +static int ti_adpll_prepare(struct clk_hw *hw)
> +{
> + struct ti_adpll_dco_data *dco =3D to_dco(hw);
> + struct ti_adpll_data *d =3D to_adpll(dco);
> +
> + ti_adpll_clear_idle_bypass(d);
> + ti_adpll_wait_lock(d);
> +
> + return 0;
> +}
> +
> +static void ti_adpll_unprepare(struct clk_hw *hw)
> +{
> + struct ti_adpll_dco_data *dco =3D to_dco(hw);
> + struct ti_adpll_data *d =3D to_adpll(dco);
> +
> + ti_adpll_set_idle_bypass(d);
> +}
> +
> +static int ti_adpll_is_prepared(struct clk_hw *hw)
> +{
> + struct ti_adpll_dco_data *dco =3D to_dco(hw);
> + struct ti_adpll_data *d =3D to_adpll(dco);
> +
> + return ti_adpll_is_locked(d);
> +}
> +
> +/*
> + * Note that the DCO clock is never subject to bypass: if the PLL is off,
> + * dcoclk is low.
> + */
> +static unsigned long ti_adpll_recalc_rate(struct clk_hw *hw,
> + unsigned long parent_rate)
> +{
> + struct ti_adpll_dco_data *dco =3D to_dco(hw);
> + struct ti_adpll_data *d =3D to_adpll(dco);
> + u32 frac_m, divider, v;
> + u64 rate;
> + unsigned long flags;
> +
> + if (ti_adpll_clock_is_bypass(d))
> + return 0;
> +
> + spin_lock_irqsave(&d->lock, flags);
> + frac_m =3D readl_relaxed(d->regs + ADPLL_FRACDIV_OFFSET);
> + frac_m &=3D ADPLL_FRACDIV_FRACTIONALM_MASK;
> + rate =3D readw_relaxed(d->regs + ADPLL_MN2DIV_OFFSET) << 18;
> + rate +=3D frac_m;
> + rate *=3D parent_rate;
> + divider =3D (readw_relaxed(d->regs + ADPLL_M2NDIV_OFFSET) + 1) <<=
18;
> + spin_unlock_irqrestore(&d->lock, flags);
> +
> + do_div(rate, divider);
> +
> + if (d->c->is_type_s) {
> + v =3D readl_relaxed(d->regs + ADPLL_CLKCTRL_OFFSET);
> + if (v & BIT(ADPLL_CLKCTRL_REGM4XEN_ADPLL_S))
> + rate *=3D 4;
> + rate *=3D 2;
> + }
> +
> + return rate;
> +}
> +
> +/* PLL parent is always clkinp, bypass only affects the children */
> +static u8 ti_adpll_get_parent(struct clk_hw *hw)
> +{
> + return 0;
> +}
> +
> +static struct clk_ops ti_adpll_ops =3D {
> + .prepare =3D ti_adpll_prepare,
> + .unprepare =3D ti_adpll_unprepare,
> + .is_prepared =3D ti_adpll_is_prepared,
> + .recalc_rate =3D ti_adpll_recalc_rate,
> + .get_parent =3D ti_adpll_get_parent,
> +};
> +
> +static int ti_adpll_init_dco(struct ti_adpll_data *d)
> +{
> + struct clk_init_data init;
> + struct clk *clock;
> + const char *postfix;
> + int width, err;
> +
> + d->outputs.clks =3D devm_kzalloc(d->dev, sizeof(struct clk *) *
> + MAX_ADPLL_OUTPUTS,
> + GFP_KERNEL);
> + if (!d->outputs.clks)
> + return -ENOMEM;
> +
> + if (d->c->output_index < 0)
> + postfix =3D "dco";
> + else
> + postfix =3D NULL;
> +
> + init.name =3D ti_adpll_clk_get_name(d, d->c->output_index, postfi=
x);
> + if (!init.name)
> + return -EINVAL;
> +
> + init.parent_names =3D d->parent_names;
> + init.num_parents =3D d->c->nr_max_inputs;
> + init.ops =3D &ti_adpll_ops;
> + init.flags =3D CLK_GET_RATE_NOCACHE;
> + d->dco.hw.init =3D &init;
> +
> + if (d->c->is_type_s)
> + width =3D 5;
> + else
> + width =3D 4;
> +
> + /* Internal input clock divider N2 */
> + err =3D ti_adpll_init_divider(d, TI_ADPLL_N2, -ENODEV, "n2",
> + d->parent_clocks[TI_ADPLL_CLKINP],
> + d->regs + ADPLL_MN2DIV_OFFSET,
> + ADPLL_MN2DIV_N2, width, 0);
> + if (err)
> + return err;
> +
> + clock =3D devm_clk_register(d->dev, &d->dco.hw);
> + if (IS_ERR(clock))
> + return PTR_ERR(clock);
> +
> + return ti_adpll_setup_clock(d, clock, TI_ADPLL_DCO, d->c->output_=
index,
> + init.name, NULL);
> +}
> +
> +static int ti_adpll_clkout_enable(struct clk_hw *hw)
> +{
> + struct ti_adpll_clkout_data *co =3D to_clkout(hw);
> + struct clk_hw *gate_hw =3D &co->gate.hw;
> +
> + __clk_hw_set_clk(gate_hw, hw);
> +
> + return clk_gate_ops.enable(gate_hw);
> +}
> +
> +static void ti_adpll_clkout_disable(struct clk_hw *hw)
> +{
> + struct ti_adpll_clkout_data *co =3D to_clkout(hw);
> + struct clk_hw *gate_hw =3D &co->gate.hw;
> +
> + __clk_hw_set_clk(gate_hw, hw);
> + clk_gate_ops.disable(gate_hw);
> +}
> +
> +static int ti_adpll_clkout_is_enabled(struct clk_hw *hw)
> +{
> + struct ti_adpll_clkout_data *co =3D to_clkout(hw);
> + struct clk_hw *gate_hw =3D &co->gate.hw;
> +
> + __clk_hw_set_clk(gate_hw, hw);
> +
> + return clk_gate_ops.is_enabled(gate_hw);
> +}
> +
> +/* Setting PLL bypass puts clkout and clkoutx2 into bypass */
> +static u8 ti_adpll_clkout_get_parent(struct clk_hw *hw)
> +{
> + struct ti_adpll_clkout_data *co =3D to_clkout(hw);
> + struct ti_adpll_data *d =3D co->adpll;
> +
> + return ti_adpll_clock_is_bypass(d);
> +}
> +
> +static int ti_adpll_init_clkout(struct ti_adpll_data *d,
> + enum ti_adpll_clocks index,
> + int output_index, int gate_bit,
> + char *name, struct clk *clk0,
> + struct clk *clk1)
> +{
> + struct ti_adpll_clkout_data *co;
> + struct clk_init_data init;
> + struct clk_ops *ops;
> + const char *parent_names[2];
> + const char *child_name;
> + struct clk *clock;
> + int err;
> +
> + co =3D devm_kzalloc(d->dev, sizeof(*co), GFP_KERNEL);
> + if (!co)
> + return -ENOMEM;
> + co->adpll =3D d;
> +
> + err =3D of_property_read_string_index(d->np,
> + "clock-output-names",
> + output_index,
> + &child_name);
> + if (err)
> + return err;
> +
> + ops =3D devm_kzalloc(d->dev, sizeof(*ops), GFP_KERNEL);
> + if (!ops)
> + return -ENOMEM;
> +
> + init.name =3D child_name;
> + init.ops =3D ops;
> + init.flags =3D CLK_IS_BASIC;
> + co->hw.init =3D &init;
> + parent_names[0] =3D __clk_get_name(clk0);
> + parent_names[1] =3D __clk_get_name(clk1);
> + init.parent_names =3D parent_names;
> + init.num_parents =3D 2;
> +
> + ops->get_parent =3D ti_adpll_clkout_get_parent;
> + ops->determine_rate =3D __clk_mux_determine_rate;
> + if (gate_bit) {
> + co->gate.lock =3D &d->lock;
> + co->gate.reg =3D d->regs + ADPLL_CLKCTRL_OFFSET;
> + co->gate.bit_idx =3D gate_bit;
> + ops->enable =3D ti_adpll_clkout_enable;
> + ops->disable =3D ti_adpll_clkout_disable;
> + ops->is_enabled =3D ti_adpll_clkout_is_enabled;
> + }
> +
> + clock =3D devm_clk_register(d->dev, &co->hw);
> + if (IS_ERR(clock)) {
> + dev_err(d->dev, "failed to register output %s: %li\n",
> + name, PTR_ERR(clock));
> + return PTR_ERR(clock);
> + }
> +
> + return ti_adpll_setup_clock(d, clock, index, output_index, child_=
name,
> + NULL);
> +}
> +
> +static int ti_adpll_init_children_adpll_s(struct ti_adpll_data *d)
> +{
> + int err;
> +
> + if (!d->c->is_type_s)
> + return 0;
> +
> + /* Internal mux, sources from divider N2 or clkinpulow */
> + err =3D ti_adpll_init_mux(d, TI_ADPLL_BYPASS, "bypass",
> + d->clocks[TI_ADPLL_N2].clk,
> + d->parent_clocks[TI_ADPLL_CLKINPULOW],
> + d->regs + ADPLL_CLKCTRL_OFFSET,
> + ADPLL_CLKCTRL_ULOWCLKEN);
> + if (err)
> + return err;
> +
> + /* Internal divider M2, sources DCO */
> + err =3D ti_adpll_init_divider(d, TI_ADPLL_M2, -ENODEV, "m2",
> + d->clocks[TI_ADPLL_DCO].clk,
> + d->regs + ADPLL_M2NDIV_OFFSET,
> + ADPLL_M2NDIV_M2,
> + ADPLL_M2NDIV_M2_ADPLL_S_WIDTH,
> + CLK_DIVIDER_ONE_BASED);
> + if (err)
> + return err;
> +
> + /* Internal fixed divider, after M2 before clkout */
> + err =3D ti_adpll_init_fixed_factor(d, TI_ADPLL_DIV2, "div2",
> + d->clocks[TI_ADPLL_M2].clk,
> + 1, 2);
> + if (err)
> + return err;
> +
> + /* Output clkout with a mux and gate, sources from div2 or bypass=
*/
> + err =3D ti_adpll_init_clkout(d, TI_ADPLL_CLKOUT, TI_ADPLL_S_CLKOU=
T,
> + ADPLL_CLKCTRL_CLKOUTEN, "clkout",
> + d->clocks[TI_ADPLL_DIV2].clk,
> + d->clocks[TI_ADPLL_BYPASS].clk);
> + if (err)
> + return err;
> +
> + /* Output clkoutx2 with a mux and gate, sources from M2 or bypass=
*/
> + err =3D ti_adpll_init_clkout(d, TI_ADPLL_CLKOUT2, TI_ADPLL_S_CLKO=
UTX2, 0,
> + "clkout2", d->clocks[TI_ADPLL_M2].clk,
> + d->clocks[TI_ADPLL_BYPASS].clk);
> + if (err)
> + return err;
> +
> + /* Internal mux, sources from DCO and clkinphif */
> + if (d->parent_clocks[TI_ADPLL_CLKINPHIF]) {
> + err =3D ti_adpll_init_mux(d, TI_ADPLL_HIF, "hif",
> + d->clocks[TI_ADPLL_DCO].clk,
> + d->parent_clocks[TI_ADPLL_CLKINPH=
IF],
> + d->regs + ADPLL_CLKCTRL_OFFSET,
> + ADPLL_CLKINPHIFSEL_ADPLL_S);
> + if (err)
> + return err;
> + }
> +
> + /* Output clkouthif with a divider M3, sources from hif */
> + err =3D ti_adpll_init_divider(d, TI_ADPLL_M3, TI_ADPLL_S_CLKOUTHI=
F, "m3",
> + d->clocks[TI_ADPLL_HIF].clk,
> + d->regs + ADPLL_M3DIV_OFFSET,
> + ADPLL_M3DIV_M3,
> + ADPLL_M3DIV_M3_WIDTH,
> + CLK_DIVIDER_ONE_BASED);
> + if (err)
> + return err;
> +
> + /* Output clock dcoclkldo is the DCO */
> +
> + return 0;
> +}
> +
> +static int ti_adpll_init_children_adpll_lj(struct ti_adpll_data *d)
> +{
> + int err;
> +
> + if (d->c->is_type_s)
> + return 0;
> +
> + /* Output clkdcoldo, gated output of DCO */
> + err =3D ti_adpll_init_gate(d, TI_ADPLL_DCO_GATE, TI_ADPLL_LJ_CLKD=
COLDO,
> + "clkdcoldo", d->clocks[TI_ADPLL_DCO].clk,
> + d->regs + ADPLL_CLKCTRL_OFFSET,
> + ADPLL_CLKCTRL_CLKDCOLDOEN, 0);
> + if (err)
> + return err;
> +
> + /* Internal divider M2, sources from DCO */
> + err =3D ti_adpll_init_divider(d, TI_ADPLL_M2, -ENODEV,
> + "m2", d->clocks[TI_ADPLL_DCO].clk,
> + d->regs + ADPLL_M2NDIV_OFFSET,
> + ADPLL_M2NDIV_M2,
> + ADPLL_M2NDIV_M2_ADPLL_LJ_WIDTH,
> + CLK_DIVIDER_ONE_BASED);
> + if (err)
> + return err;
> +
> + /* Output clkoutldo, gated output of M2 */
> + err =3D ti_adpll_init_gate(d, TI_ADPLL_M2_GATE, TI_ADPLL_LJ_CLKOU=
TLDO,
> + "clkoutldo", d->clocks[TI_ADPLL_M2].clk,
> + d->regs + ADPLL_CLKCTRL_OFFSET,
> + ADPLL_CLKCTRL_CLKOUTLDOEN_ADPLL_LJ,
> + 0);
> + if (err)
> + return err;
> +
> + /* Internal mux, sources from divider N2 or clkinpulow */
> + err =3D ti_adpll_init_mux(d, TI_ADPLL_BYPASS, "bypass",
> + d->clocks[TI_ADPLL_N2].clk,
> + d->parent_clocks[TI_ADPLL_CLKINPULOW],
> + d->regs + ADPLL_CLKCTRL_OFFSET,
> + ADPLL_CLKCTRL_ULOWCLKEN);
> + if (err)
> + return err;
> +
> + /* Output clkout, sources M2 or bypass */
> + err =3D ti_adpll_init_clkout(d, TI_ADPLL_CLKOUT, TI_ADPLL_S_CLKOU=
T,
> + ADPLL_CLKCTRL_CLKOUTEN, "clkout",
> + d->clocks[TI_ADPLL_M2].clk,
> + d->clocks[TI_ADPLL_BYPASS].clk);
> + if (err)
> + return err;
> +
> + return 0;
> +}
> +
> +static void ti_adpll_free_resources(struct ti_adpll_data *d)
> +{
> + int i;
> +
> + for (i =3D TI_ADPLL_M3; i >=3D 0; i--) {
> + struct ti_adpll_clock *ac =3D &d->clocks[i];
> +
> + if (!ac || IS_ERR_OR_NULL(ac->clk))
> + continue;
> + if (ac->cl)
> + clkdev_drop(ac->cl);
> + if (ac->unregister)
> + ac->unregister(ac->clk);
> + }
> +}
> +
> +/* MPU PLL manages the lock register for all PLLs */
> +static void ti_adpll_unlock_all(void __iomem *reg)
> +{
> + u32 v;
> +
> + v =3D readl_relaxed(reg);
> + if (v =3D=3D ADPLL_PLLSS_MMR_LOCK_ENABLED)
> + writel_relaxed(ADPLL_PLLSS_MMR_UNLOCK_MAGIC, reg);
> +}
> +
> +static int ti_adpll_init_registers(struct ti_adpll_data *d)
> +{
> + int register_offset =3D 0;
> +
> + if (d->c->is_type_s) {
> + register_offset =3D 8;
> + ti_adpll_unlock_all(d->iobase + ADPLL_PLLSS_MMR_LOCK_OFFS=
ET);
> + }
> +
> + d->regs =3D d->iobase + register_offset + ADPLL_PWRCTRL_OFFSET;
> +
> + return 0;
> +}
> +
> +static int ti_adpll_init_inputs(struct ti_adpll_data *d)
> +{
> + const char *error =3D "need at least %i inputs";
> + struct clk *clock;
> + int nr_inputs;
> +
> + nr_inputs =3D of_clk_get_parent_count(d->np);
> + if (nr_inputs < d->c->nr_max_inputs) {
> + dev_err(d->dev, error, nr_inputs);
> + return -EINVAL;
> + }
> + of_clk_parent_fill(d->np, d->parent_names, nr_inputs);
> +
> + clock =3D devm_clk_get(d->dev, d->parent_names[0]);
> + if (IS_ERR(clock)) {
> + dev_err(d->dev, "could not get clkinp\n");
> + return PTR_ERR(clock);
> + }
> + d->parent_clocks[TI_ADPLL_CLKINP] =3D clock;
> +
> + clock =3D devm_clk_get(d->dev, d->parent_names[1]);
> + if (IS_ERR(clock)) {
> + dev_err(d->dev, "could not get clkinpulow clock\n");
> + return PTR_ERR(clock);
> + }
> + d->parent_clocks[TI_ADPLL_CLKINPULOW] =3D clock;
> +
> + if (d->c->is_type_s) {
> + clock =3D devm_clk_get(d->dev, d->parent_names[2]);
> + if (IS_ERR(clock)) {
> + dev_err(d->dev, "could not get clkinphif clock\n"=
);
> + return PTR_ERR(clock);
> + }
> + d->parent_clocks[TI_ADPLL_CLKINPHIF] =3D clock;
> + }
> +
> + return 0;
> +}
> +
> +static const struct ti_adpll_platform_data ti_adpll_type_s =3D {
> + .is_type_s =3D true,
> + .nr_max_inputs =3D MAX_ADPLL_INPUTS,
> + .nr_max_outputs =3D MAX_ADPLL_OUTPUTS,
> + .output_index =3D TI_ADPLL_S_DCOCLKLDO,
> +};
> +
> +static const struct ti_adpll_platform_data ti_adpll_type_lj =3D {
> + .is_type_s =3D false,
> + .nr_max_inputs =3D MAX_ADPLL_INPUTS - 1,
> + .nr_max_outputs =3D MAX_ADPLL_OUTPUTS - 1,
> + .output_index =3D -EINVAL,
> +};
> +
> +static const struct of_device_id ti_adpll_match[] =3D {
> + { .compatible =3D "ti,dm814-adpll-s-clock", &ti_adpll_type_s },
> + { .compatible =3D "ti,dm814-adpll-lj-clock", &ti_adpll_type_lj },
> + {},
> +};
> +MODULE_DEVICE_TABLE(of, ti_adpll_match);
> +
> +static int ti_adpll_probe(struct platform_device *pdev)
> +{
> + struct device_node *node =3D pdev->dev.of_node;
> + struct device *dev =3D &pdev->dev;
> + const struct of_device_id *match;
> + const struct ti_adpll_platform_data *pdata;
> + struct ti_adpll_data *d;
> + struct resource *res;
> + int err;
> +
> + match =3D of_match_device(ti_adpll_match, dev);
> + if (match)
> + pdata =3D match->data;
> + else
> + return -ENODEV;
> +
> + d =3D devm_kzalloc(dev, sizeof(*d), GFP_KERNEL);
> + if (!d)
> + return -ENOMEM;
> + d->dev =3D dev;
> + d->np =3D node;
> + d->c =3D pdata;
> + dev_set_drvdata(d->dev, d);
> + spin_lock_init(&d->lock);
> +
> + res =3D platform_get_resource(pdev, IORESOURCE_MEM, 0);
> + if (!res)
> + return -ENODEV;
> + d->pa =3D res->start;
> +
> + d->iobase =3D devm_ioremap_resource(dev, res);
> + if (IS_ERR(d->iobase)) {
> + dev_err(dev, "could not get IO base: %li\n",
> + PTR_ERR(d->iobase));
> + return PTR_ERR(d->iobase);
> + }
> +
> + err =3D ti_adpll_init_registers(d);
> + if (err)
> + return err;
> +
> + err =3D ti_adpll_init_inputs(d);
> + if (err)
> + return err;
> +
> + d->clocks =3D devm_kzalloc(d->dev, sizeof(struct ti_adpll_clock) *
> + TI_ADPLL_NR_CLOCKS,
> + GFP_KERNEL);
> + if (!d->clocks)
> + goto free;
> +
> + err =3D ti_adpll_init_dco(d);
> + if (err) {
> + dev_err(dev, "could not register dco: %i\n", err);
> + goto free;
> + }
> +
> + err =3D ti_adpll_init_children_adpll_s(d);
> + if (err)
> + goto free;
> + err =3D ti_adpll_init_children_adpll_lj(d);
> + if (err)
> + goto free;
> +
> + err =3D of_clk_add_provider(d->np, of_clk_src_onecell_get, &d->ou=
tputs);
> + if (err)
> + goto free;
> +
> + return 0;
> +
> +free:
> + WARN_ON(1);
> + ti_adpll_free_resources(d);
> +
> + return err;
> +}
> +
> +static int ti_adpll_remove(struct platform_device *pdev)
> +{
> + struct ti_adpll_data *d =3D dev_get_drvdata(&pdev->dev);
> +
> + ti_adpll_free_resources(d);
> +
> + return 0;
> +}
> +
> +static struct platform_driver ti_adpll_driver =3D {
> + .driver =3D {
> + .name =3D "ti-adpll",
> + .of_match_table =3D ti_adpll_match,
> + },
> + .probe =3D ti_adpll_probe,
> + .remove =3D ti_adpll_remove,
> +};
> +
> +static int __init ti_adpll_init(void)
> +{
> + return platform_driver_register(&ti_adpll_driver);
> +}
> +core_initcall(ti_adpll_init);
> +
> +static void __exit ti_adpll_exit(void)
> +{
> + platform_driver_unregister(&ti_adpll_driver);
> +}
> +module_exit(ti_adpll_exit);
> +
> +MODULE_DESCRIPTION("Clock driver for dm814x ADPLL");
> +MODULE_ALIAS("platform:dm814-adpll-clock");
> +MODULE_AUTHOR("Tony LIndgren <tony@atomide.com>");
> +MODULE_LICENSE("GPL v2");
> diff --git a/drivers/clk/ti/clk-814x.c b/drivers/clk/ti/clk-814x.c
> index 9e85fcc..2323643 100644
> --- a/drivers/clk/ti/clk-814x.c
> +++ b/drivers/clk/ti/clk-814x.c
> @@ -5,8 +5,10 @@
> */
> =
> #include <linux/kernel.h>
> +#include <linux/clk.h>
> #include <linux/clk-provider.h>
> #include <linux/clk/ti.h>
> +#include <linux/of_platform.h>
> =
> #include "clock.h"
> =
> @@ -27,11 +29,62 @@ static struct ti_dt_clk dm814_clks[] =3D {
> { .node_name =3D NULL },
> };
> =
> +static bool timer_clocks_initialized;
> +
> +int __init dm814x_adpll_early_init(void)
> +{
> + struct device_node *np;
> +
> + if (!timer_clocks_initialized)
> + return -ENODEV;
> +
> + np =3D of_find_node_by_name(NULL, "pllss");
> + if (!np) {
> + pr_err("Could not find node for plls\n");
> + return -ENODEV;
> + }
> +
> + of_platform_populate(np, NULL, NULL, NULL);
> +
> + return 0;
> +}
> +core_initcall(dm814x_adpll_early_init);
> +
> +static const char * const init_clocks[] =3D {
> + "pll040clkout", /* MPU 481c5040.adpll.clkout */
> + "pll290clkout", /* DDR 481c5290.adpll.clkout */
> +};
> +
> +int __init dm814x_adpll_enable_init_clocks(void)
> +{
> + int i, err;
> +
> + if (!timer_clocks_initialized)
> + return -ENODEV;
> +
> + for (i =3D 0; i < ARRAY_SIZE(init_clocks); i++) {
> + struct clk *clock;
> +
> + clock =3D clk_get(NULL, init_clocks[i]);
> + if (WARN(IS_ERR(clock), "could not find init clock %s\n",
> + init_clocks[i]))
> + continue;
> + err =3D clk_prepare_enable(clock);
> + if (WARN(err, "could not enable init clock %s\n",
> + init_clocks[i]))
> + continue;
> + }
> +
> + return 0;
> +}
> +postcore_initcall(dm814x_adpll_enable_init_clocks);
> +
> int __init dm814x_dt_clk_init(void)
> {
> ti_dt_clocks_register(dm814_clks);
> omap2_clk_disable_autoidle_all();
> omap2_clk_enable_init_clocks(NULL, 0);
> + timer_clocks_initialized =3D true;
> =
> return 0;
> }
prev parent reply other threads:[~2016-02-29 22:19 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-02-12 21:20 [PATCHv5 0/2] Clock driver for dm814x and dra62x ADPLL Tony Lindgren
2016-02-12 21:20 ` [PATCH 1/2] dt-bindings: clock: adpll: Add binding documentation for TI adpll Tony Lindgren
2016-02-17 1:04 ` Michael Turquette
2016-02-17 17:28 ` Tony Lindgren
2016-02-12 21:20 ` [PATCH 2/2] clk: ti: Add support for dm814x ADPLL Tony Lindgren
2016-02-17 1:19 ` Michael Turquette
2016-02-17 17:39 ` Tony Lindgren
2016-02-17 20:52 ` Michael Turquette
2016-02-17 21:20 ` Tony Lindgren
2016-02-17 23:02 ` Michael Turquette
2016-02-23 13:47 ` Matthijs van Duin
2016-02-26 17:35 ` Tony Lindgren
2016-02-29 22:19 ` 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=20160229221958.3140.29381@quark.deferred.io \
--to=mturquette@baylibre.com \
--cc=b.hutchman@gmail.com \
--cc=dbrignoli@audioscience.com \
--cc=ilu@linutronix.de \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-clk@vger.kernel.org \
--cc=linux-omap@vger.kernel.org \
--cc=linux@arm.linux.org.uk \
--cc=matthijsvanduin@gmail.com \
--cc=narmstrong@baylibre.com \
--cc=sboyd@codeaurora.org \
--cc=t-kristo@ti.com \
--cc=tony@atomide.com \
/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;
as well as URLs for NNTP newsgroup(s).