Re: [PATCH V3 4/5] clk: imx: support fracn gppll

From: Abel Vesa
Date: Wed Feb 23 2022 - 05:47:58 EST


On 22-02-23 11:43:08, Sascha Hauer wrote:
> Hi,
>
> Sorry for the late review, but this PLL seems to have some copy-pasted
> code I just cleaned up in another PLL, so I thought I have a closer
> look.
>

Thanks for helping with the review, Sascha.

Peng, I'll wait for these comments to be addressed then.

>
> On Wed, Feb 23, 2022 at 02:43:57PM +0800, Peng Fan (OSS) wrote:
> > From: Peng Fan <peng.fan@xxxxxxx>
> >
> > This PLL module is a Fractional-N synthesizer,
> > supporting 30-bit numerator and denominator. Numerator is a signed
> > number. It has feature to adjust fractional portion of feedback
> > divider dynamically. This fracn gppll is used in i.MX93.
> >
> > Signed-off-by: Peng Fan <peng.fan@xxxxxxx>
> > ---
> > drivers/clk/imx/Makefile | 1 +
> > drivers/clk/imx/clk-fracn-gppll.c | 328 ++++++++++++++++++++++++++++++
> > drivers/clk/imx/clk.h | 21 ++
> > 3 files changed, 350 insertions(+)
> > create mode 100644 drivers/clk/imx/clk-fracn-gppll.c
> >
> > diff --git a/drivers/clk/imx/Makefile b/drivers/clk/imx/Makefile
> > index 36c04922d789..60c8a4bb7574 100644
> > --- a/drivers/clk/imx/Makefile
> > +++ b/drivers/clk/imx/Makefile
> > @@ -5,6 +5,7 @@ mxc-clk-objs += clk-busy.o
> > mxc-clk-objs += clk-composite-7ulp.o
> > mxc-clk-objs += clk-composite-8m.o
> > mxc-clk-objs += clk-composite-93.o
> > +mxc-clk-objs += clk-fracn-gppll.o
> > mxc-clk-objs += clk-cpu.o
> > mxc-clk-objs += clk-divider-gate.o
> > mxc-clk-objs += clk-fixup-div.o
> > diff --git a/drivers/clk/imx/clk-fracn-gppll.c b/drivers/clk/imx/clk-fracn-gppll.c
> > new file mode 100644
> > index 000000000000..6c9946a4bdb7
> > --- /dev/null
> > +++ b/drivers/clk/imx/clk-fracn-gppll.c
> > @@ -0,0 +1,328 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +/*
> > + * Copyright 2021 NXP
> > + */
> > +
> > +#include <linux/bits.h>
> > +#include <linux/clk-provider.h>
> > +#include <linux/err.h>
> > +#include <linux/export.h>
> > +#include <linux/io.h>
> > +#include <linux/iopoll.h>
> > +#include <linux/slab.h>
> > +#include <linux/jiffies.h>
> > +
> > +#include "clk.h"
> > +
> > +#define PLL_CTRL 0x0
> > +#define CLKMUX_BYPASS BIT(2)
> > +#define CLKMUX_EN BIT(1)
> > +#define POWERUP_MASK BIT(0)
> > +
> > +#define PLL_ANA_PRG 0x10
> > +#define PLL_SPREAD_SPECTRUM 0x30
> > +
> > +#define PLL_NUMERATOR 0x40
> > +#define PLL_MFN_MASK GENMASK(31, 2)
> > +#define PLL_MFN_SHIFT 2
> > +
> > +#define PLL_DENOMINATOR 0x50
> > +#define PLL_MFD_MASK GENMASK(29, 0)
> > +
> > +#define PLL_DIV 0x60
> > +#define PLL_MFI_MASK GENMASK(24, 16)
> > +#define PLL_MFI_SHIFT 16
> > +#define PLL_RDIV_MASK GENMASK(15, 13)
> > +#define PLL_RDIV_SHIFT 13
> > +#define PLL_ODIV_MASK GENMASK(7, 0)
> > +
> > +#define PLL_DFS_CTRL(x) (0x70 + (x) * 0x10)
> > +
> > +#define PLL_STATUS 0xF0
> > +#define LOCK_STATUS BIT(0)
> > +
> > +#define DFS_STATUS 0xF4
> > +
> > +#define LOCK_TIMEOUT_US 200
> > +
> > +#define PLL_FRACN_GP(_rate, _mfi, _mfn, _mfd, _rdiv, _odiv) \
> > + { \
> > + .rate = (_rate), \
> > + .mfi = (_mfi), \
> > + .mfn = (_mfn), \
> > + .mfd = (_mfd), \
> > + .rdiv = (_rdiv), \
> > + .odiv = (_odiv), \
> > + }
> > +
> > +struct clk_fracn_gppll {
> > + struct clk_hw hw;
> > + void __iomem *base;
> > + const struct imx_fracn_gppll_rate_table *rate_table;
> > + int rate_count;
> > +};
> > +
> > +#define to_clk_fracn_gppll(_hw) container_of(_hw, struct clk_fracn_gppll, hw)
>
> Consider using a static inline function instead.
>
> > +
> > +/*
> > + * Fvco = 𝐹𝑟𝑒𝑓∙(𝑀𝐹𝐼+𝑀𝐹𝑁/𝑀𝐹𝐷)
> > + * Fout = Fvco / (rdiv * odiv)
> > + */
> > +static const struct imx_fracn_gppll_rate_table fracn_tbl[] = {
> > + PLL_FRACN_GP(650000000U, 81, 0, 0, 0, 3),
> > + PLL_FRACN_GP(594000000U, 198, 0, 0, 0, 8),
> > + PLL_FRACN_GP(560000000U, 70, 0, 0, 0, 3),
> > + PLL_FRACN_GP(400000000U, 50, 0, 0, 0, 3),
> > + PLL_FRACN_GP(393216000U, 81, 92, 100, 0, 5)
> > +};
> > +
> > +struct imx_fracn_gppll_clk imx_fracn_gppll = {
> > + .rate_table = fracn_tbl,
> > + .rate_count = ARRAY_SIZE(fracn_tbl),
> > +};
> > +EXPORT_SYMBOL_GPL(imx_fracn_gppll);
> > +
> > +static const struct imx_fracn_gppll_rate_table *
> > +imx_get_pll_settings(struct clk_fracn_gppll *pll, unsigned long rate)
> > +{
> > + const struct imx_fracn_gppll_rate_table *rate_table = pll->rate_table;
> > + int i;
> > +
> > + for (i = 0; i < pll->rate_count; i++)
> > + if (rate == rate_table[i].rate)
> > + return &rate_table[i];
> > +
> > + return NULL;
> > +}
> > +
> > +static long clk_fracn_gppll_round_rate(struct clk_hw *hw, unsigned long rate,
> > + unsigned long *prate)
> > +{
> > + struct clk_fracn_gppll *pll = to_clk_fracn_gppll(hw);
> > + const struct imx_fracn_gppll_rate_table *rate_table = pll->rate_table;
> > + int i;
> > +
> > + /* Assumming rate_table is in descending order */
>
> s/Assumming/Assuming/
>
> > + for (i = 0; i < pll->rate_count; i++)
> > + if (rate >= rate_table[i].rate)
> > + return rate_table[i].rate;
> > +
> > + if (i == pll->rate_count)
>
> This is always true when you're here,
>
> > + pr_err("Not able to round rate for %s: %lu\n", clk_hw_get_name(hw), rate);
>
> but this message shouldn't be needed at all. When a rate is passed in
> here that is too low then the driver should silently round up to the
> lowest supported rate.
>
> > +
> > + /* return minimum supported value */
> > + return rate_table[i - 1].rate;
>
> IMO rate_table[pll->rate_count - 1].rate makes it clearer what is meant
> here.
>
> > +}
> > +
> > +static unsigned long clk_fracn_gppll_recalc_rate(struct clk_hw *hw, unsigned long parent_rate)
> > +{
> > + struct clk_fracn_gppll *pll = to_clk_fracn_gppll(hw);
> > + const struct imx_fracn_gppll_rate_table *rate_table = pll->rate_table;
> > + u32 pll_numerator, pll_denominator, pll_div;
> > + u32 mfi, mfn, mfd, rdiv, odiv;
> > + u64 fvco = parent_rate;
> > + long rate = 0;
> > + int i;
> > +
> > + pll_numerator = readl_relaxed(pll->base + PLL_NUMERATOR);
> > + mfn = (pll_numerator & PLL_MFN_MASK) >> PLL_MFN_SHIFT;
>
> Have a look at FIELD_GET/FIELD_PREP, it really makes setting and reading
> fields easier and nicer to look at.
>
> > +
> > + pll_denominator = readl_relaxed(pll->base + PLL_DENOMINATOR);
> > + mfd = pll_denominator & PLL_MFD_MASK;
> > +
> > + pll_div = readl_relaxed(pll->base + PLL_DIV);
> > + mfi = (pll_div & PLL_MFI_MASK) >> PLL_MFI_SHIFT;
> > +
> > + rdiv = (pll_div & PLL_RDIV_MASK) >> PLL_RDIV_SHIFT;
> > + rdiv = rdiv + 1;
> > + odiv = pll_div & PLL_ODIV_MASK;
> > + switch (odiv) {
> > + case 0:
> > + odiv = 2;
> > + break;
> > + case 1:
> > + odiv = 3;
> > + break;
> > + default:
> > + break;
> > + }
> > +
> > + /*
> > + * Sometimes, the recalculated rate has deviation due to
> > + * the frac part. So find the accurate pll rate from the table
> > + * first, if no match rate in the table, use the rate calculated
> > + * from the equation below.
> > + */
> > + for (i = 0; i < pll->rate_count; i++) {
> > + if (rate_table[i].mfn == mfn && rate_table[i].mfi == mfi &&
> > + rate_table[i].mfd == mfd && rate_table[i].rdiv == rdiv &&
> > + rate_table[i].odiv == odiv)
> > + rate = rate_table[i].rate;
> > + }
> > +
> > + /* Fvco = 𝐹𝑟𝑒𝑓∙(𝑀𝐹𝐼+𝑀𝐹𝑁/𝑀𝐹𝐷) */
> > + fvco = fvco * mfi + fvco * mfn / mfd;
> > +
> > + do_div(fvco, rdiv * odiv);
> > +
> > + return rate ? (unsigned long) rate : (unsigned long)fvco;
>
> You could bail out early here instead of calculating fvco which you then
> don't use.
>
> > +}
> > +
> > +static int clk_fracn_gppll_wait_lock(struct clk_fracn_gppll *pll)
> > +{
> > + u32 val;
> > +
> > + return readl_poll_timeout(pll->base + PLL_STATUS, val,
> > + val & LOCK_STATUS, 0, LOCK_TIMEOUT_US);
> > +}
> > +
> > +static int clk_fracn_gppll_set_rate(struct clk_hw *hw, unsigned long drate,
> > + unsigned long prate)
> > +{
> > + struct clk_fracn_gppll *pll = to_clk_fracn_gppll(hw);
> > + const struct imx_fracn_gppll_rate_table *rate;
> > + u32 tmp, pll_div, ana_mfn;
> > + int ret;
> > +
> > + rate = imx_get_pll_settings(pll, drate);
> > + if (!rate) {
> > + pr_err("%s: Invalid rate : %lu for pll clk %s\n", __func__,
> > + drate, clk_hw_get_name(hw));
> > + return -EINVAL;
> > + }
>
> The core does a round_rate() before doing a set_rate(), so this should
> never happen. I think this can be dropped, in doubt the resulting NULL
> pointer deref provides enough developer information to debug this
> internal error.
>
> > +
> > + /* Disable output */
> > + tmp = readl_relaxed(pll->base + PLL_CTRL);
> > + tmp &= ~CLKMUX_EN;
> > + writel_relaxed(tmp, pll->base + PLL_CTRL);
> > +
> > + /* Power Down */
> > + tmp &= ~POWERUP_MASK;
> > + writel_relaxed(tmp, pll->base + PLL_CTRL);
> > +
> > + /* Disable BYPASS */
> > + tmp &= ~CLKMUX_BYPASS;
> > + writel_relaxed(tmp, pll->base + PLL_CTRL);
> > +
> > + pll_div = (rate->rdiv << PLL_RDIV_SHIFT) | rate->odiv | (rate->mfi << PLL_MFI_SHIFT);
> > + writel_relaxed(pll_div, pll->base + PLL_DIV);
> > + writel_relaxed(rate->mfd, pll->base + PLL_DENOMINATOR);
> > + writel_relaxed(rate->mfn << PLL_MFN_SHIFT, pll->base + PLL_NUMERATOR);
> > +
> > + /* Wait for 5us according to fracn mode pll doc */
> > + udelay(5);
> > +
> > + /* Enable Powerup */
> > + tmp |= POWERUP_MASK;
> > + writel_relaxed(tmp, pll->base + PLL_CTRL);
> > +
> > + /* Wait Lock*/
>
> Nitpick: Space missing at end of comment.
>
> > + ret = clk_fracn_gppll_wait_lock(pll);
> > + if (ret)
> > + return ret;
> > +
> > + /* Enable output */
> > + tmp |= CLKMUX_EN;
> > + writel_relaxed(tmp, pll->base + PLL_CTRL);
> > +
> > + ana_mfn = (readl_relaxed(pll->base + PLL_STATUS) & PLL_MFN_MASK) >> PLL_MFN_SHIFT;
> > +
> > + WARN(ana_mfn != rate->mfn, "ana_mfn != rate->mfn\n");
> > +
> > + return 0;
> > +}
> > +
> > +static int clk_fracn_gppll_prepare(struct clk_hw *hw)
> > +{
> > + struct clk_fracn_gppll *pll = to_clk_fracn_gppll(hw);
> > + u32 val;
> > + int ret;
> > +
> > + val = readl_relaxed(pll->base + PLL_CTRL);
> > + if (val & POWERUP_MASK)
> > + return 0;
> > +
> > + val |= CLKMUX_BYPASS;
> > + writel_relaxed(val, pll->base + PLL_CTRL);
> > +
> > + val |= POWERUP_MASK;
> > + writel_relaxed(val, pll->base + PLL_CTRL);
> > +
> > + val |= CLKMUX_EN;
> > + writel_relaxed(val, pll->base + PLL_CTRL);
> > +
> > + ret = clk_fracn_gppll_wait_lock(pll);
> > + if (ret)
> > + return ret;
> > +
> > + val &= ~CLKMUX_BYPASS;
> > + writel_relaxed(val, pll->base + PLL_CTRL);
> > +
> > + return 0;
> > +}
> > +
> > +static int clk_fracn_gppll_is_prepared(struct clk_hw *hw)
> > +{
> > + struct clk_fracn_gppll *pll = to_clk_fracn_gppll(hw);
> > + u32 val;
> > +
> > + val = readl_relaxed(pll->base + PLL_CTRL);
> > +
> > + return (val & POWERUP_MASK) ? 1 : 0;
> > +}
> > +
> > +static void clk_fracn_gppll_unprepare(struct clk_hw *hw)
> > +{
> > + struct clk_fracn_gppll *pll = to_clk_fracn_gppll(hw);
> > + u32 val;
> > +
> > + val = readl_relaxed(pll->base + PLL_CTRL);
> > + val &= ~POWERUP_MASK;
> > + writel_relaxed(val, pll->base + PLL_CTRL);
> > +}
> > +
> > +static const struct clk_ops clk_fracn_gppll_ops = {
> > + .prepare = clk_fracn_gppll_prepare,
> > + .unprepare = clk_fracn_gppll_unprepare,
> > + .is_prepared = clk_fracn_gppll_is_prepared,
> > + .recalc_rate = clk_fracn_gppll_recalc_rate,
> > + .round_rate = clk_fracn_gppll_round_rate,
> > + .set_rate = clk_fracn_gppll_set_rate,
> > +};
> > +
> > +struct clk_hw *imx_clk_fracn_gppll(const char *name, const char *parent_name, void __iomem *base,
> > + const struct imx_fracn_gppll_clk *pll_clk)
> > +{
> > + struct clk_fracn_gppll *pll;
> > + struct clk_hw *hw;
> > + struct clk_init_data init;
> > + int ret;
> > + u32 val;
> > +
> > + pll = kzalloc(sizeof(*pll), GFP_KERNEL);
> > + if (!pll)
> > + return ERR_PTR(-ENOMEM);
> > +
> > + init.name = name;
> > + init.flags = pll_clk->flags;
> > + init.parent_names = &parent_name;
> > + init.num_parents = 1;
> > + init.ops = &clk_fracn_gppll_ops;
> > +
> > + pll->base = base;
> > + pll->hw.init = &init;
> > + pll->rate_table = pll_clk->rate_table;
> > + pll->rate_count = pll_clk->rate_count;
> > +
> > + hw = &pll->hw;
> > +
> > + ret = clk_hw_register(NULL, hw);
> > + if (ret) {
> > + pr_err("%s: failed to register pll %s %d\n", __func__, name, ret);
> > + kfree(pll);
> > + return ERR_PTR(ret);
> > + }
> > +
> > + return hw;
> > +}
> > +EXPORT_SYMBOL_GPL(imx_clk_fracn_gppll);
> > diff --git a/drivers/clk/imx/clk.h b/drivers/clk/imx/clk.h
> > index 63eb7c53b123..a7cbbcd1a3f4 100644
> > --- a/drivers/clk/imx/clk.h
> > +++ b/drivers/clk/imx/clk.h
> > @@ -72,6 +72,27 @@ extern struct imx_pll14xx_clk imx_1416x_pll;
> > extern struct imx_pll14xx_clk imx_1443x_pll;
> > extern struct imx_pll14xx_clk imx_1443x_dram_pll;
> >
> > +/* NOTE: Rate table should be kept sorted in descending order. */
> > +struct imx_fracn_gppll_rate_table {
> > + unsigned int rate;
> > + unsigned int mfi;
> > + unsigned int mfn;
> > + unsigned int mfd;
> > + unsigned int rdiv;
> > + unsigned int odiv;
> > +};
> > +
> > +struct imx_fracn_gppll_clk {
> > + const struct imx_fracn_gppll_rate_table *rate_table;
> > + int rate_count;
> > + int flags;
> > +};
> > +
> > +struct clk_hw *imx_clk_fracn_gppll(const char *name, const char *parent_name, void __iomem *base,
> > + const struct imx_fracn_gppll_clk *pll_clk);
> > +
> > +extern struct imx_fracn_gppll_clk imx_fracn_gppll;
> > +
> > #define imx_clk_cpu(name, parent_name, div, mux, pll, step) \
> > to_clk(imx_clk_hw_cpu(name, parent_name, div, mux, pll, step))
> >
> > --
> > 2.25.1
> >
> >
>
> --
> Pengutronix e.K. | |
> Steuerwalder Str. 21 | http://www.pengutronix.de/ |
> 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
> Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |