Re: [PATCH 3/10] clk: sunxi-ng: Implement multiplier offsets

From: Chen-Yu Tsai
Date: Sat Nov 12 2016 - 23:18:36 EST

Next message: Chen-Yu Tsai: "Re: [PATCH 5/10] clk: sunxi-ng: Implement multiplier maximum"
Previous message: Baoquan He: "[PATCH v2 2/2] bnx2: Wait for in-flight DMA to complete at probe stage"
In reply to: Maxime Ripard: "[PATCH 3/10] clk: sunxi-ng: Implement multiplier offsets"
Next in thread: Maxime Ripard: "[PATCH 1/10] clk: sunxi-ng: multiplier: Add fractionnal support"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Wed, Nov 9, 2016 at 1:23 AM, Maxime Ripard
<maxime.ripard@xxxxxxxxxxxxxxxxxx> wrote:
> The multipliers we've seen so far all had an offset of one. However, on the

Explaining that the offset refers to the difference between the register value
and the actual multiplier/divider applied to the clock rate would be nice.

> earlier Allwinner SoCs, the multipliers could have no offset at all.
>
> Implement an additional field for the multipliers to specify that offset.

You are also doing this for dividers. Please mention that.

And you should mention that you're only doing this for "linear" factors,
not power-of-two ones.

>
> Signed-off-by: Maxime Ripard <maxime.ripard@xxxxxxxxxxxxxxxxxx>
> ---
> drivers/clk/sunxi-ng/ccu_div.h | 10 +++++++++-
> drivers/clk/sunxi-ng/ccu_mp.c | 10 +++++++---
> drivers/clk/sunxi-ng/ccu_mult.c | 4 ++--
> drivers/clk/sunxi-ng/ccu_mult.h | 20 ++++++++++++++------
> drivers/clk/sunxi-ng/ccu_nk.c | 14 ++++++++++----
> drivers/clk/sunxi-ng/ccu_nkm.c | 18 +++++++++++++-----
> drivers/clk/sunxi-ng/ccu_nkmp.c | 17 +++++++++++++----
> drivers/clk/sunxi-ng/ccu_nm.c | 13 ++++++++++---
> 8 files changed, 78 insertions(+), 28 deletions(-)
>
> diff --git a/drivers/clk/sunxi-ng/ccu_div.h b/drivers/clk/sunxi-ng/ccu_div.h
> index 06540f7cf41c..08d074451204 100644
> --- a/drivers/clk/sunxi-ng/ccu_div.h
> +++ b/drivers/clk/sunxi-ng/ccu_div.h
> @@ -41,6 +41,7 @@ struct ccu_div_internal {
> u8 width;
>
> u32 max;
> + u32 offset;
>
> u32 flags;
>
> @@ -58,20 +59,27 @@ struct ccu_div_internal {
> #define _SUNXI_CCU_DIV_TABLE(_shift, _width, _table) \
> _SUNXI_CCU_DIV_TABLE_FLAGS(_shift, _width, _table, 0)
>
> -#define _SUNXI_CCU_DIV_MAX_FLAGS(_shift, _width, _max, _flags) \
> +#define _SUNXI_CCU_DIV_OFFSET_MAX_FLAGS(_shift, _width, _off, _max, _flags) \
> { \
> .shift = _shift, \
> .width = _width, \
> .flags = _flags, \
> .max = _max, \
> + .offset = _off, \
> }
>
> +#define _SUNXI_CCU_DIV_MAX_FLAGS(_shift, _width, _max, _flags) \
> + _SUNXI_CCU_DIV_OFFSET_MAX_FLAGS(_shift, _width, 1, _max, _flags)
> +
> #define _SUNXI_CCU_DIV_FLAGS(_shift, _width, _flags) \
> _SUNXI_CCU_DIV_MAX_FLAGS(_shift, _width, 0, _flags)
>
> #define _SUNXI_CCU_DIV_MAX(_shift, _width, _max) \
> _SUNXI_CCU_DIV_MAX_FLAGS(_shift, _width, _max, 0)
>
> +#define _SUNXI_CCU_DIV_OFFSET(_shift, _width, _offset) \
> + _SUNXI_CCU_DIV_OFFSET_MAX_FLAGS(_shift, _width, _offset, 0, 0)
> +

With this macro, you can have a divider offset of anything but 1, but
no specified maximum. You should handle this in the callbacks somehow.
Same goes for the multiplier. Otherwise with max = (1 << width), you'll
overflow the register field when offset = 0.

Also, if specified, does the maximum apply before or after the offset
is applied? Some clarification is required.

> #define _SUNXI_CCU_DIV(_shift, _width) \
> _SUNXI_CCU_DIV_FLAGS(_shift, _width, 0)
>
> diff --git a/drivers/clk/sunxi-ng/ccu_mp.c b/drivers/clk/sunxi-ng/ccu_mp.c
> index ebb1b31568a5..22c2ca7a2a22 100644
> --- a/drivers/clk/sunxi-ng/ccu_mp.c
> +++ b/drivers/clk/sunxi-ng/ccu_mp.c
> @@ -89,11 +89,14 @@ static unsigned long ccu_mp_recalc_rate(struct clk_hw *hw,
>
> m = reg >> cmp->m.shift;
> m &= (1 << cmp->m.width) - 1;
> + m += cmp->m.offset;
> + if (!m)
> + m++;
>
> p = reg >> cmp->p.shift;
> p &= (1 << cmp->p.width) - 1;
>
> - return (parent_rate >> p) / (m + 1);
> + return (parent_rate >> p) / m;
> }
>
> static int ccu_mp_determine_rate(struct clk_hw *hw,
> @@ -124,9 +127,10 @@ static int ccu_mp_set_rate(struct clk_hw *hw, unsigned long rate,
> reg = readl(cmp->common.base + cmp->common.reg);
> reg &= ~GENMASK(cmp->m.width + cmp->m.shift - 1, cmp->m.shift);
> reg &= ~GENMASK(cmp->p.width + cmp->p.shift - 1, cmp->p.shift);
> + reg |= (m - cmp->m.offset) << cmp->m.shift;
> + reg |= ilog2(p) << cmp->p.shift;
>
> - writel(reg | (ilog2(p) << cmp->p.shift) | ((m - 1) << cmp->m.shift),
> - cmp->common.base + cmp->common.reg);
> + writel(reg, cmp->common.base + cmp->common.reg);
>
> spin_unlock_irqrestore(cmp->common.lock, flags);
>
> diff --git a/drivers/clk/sunxi-ng/ccu_mult.c b/drivers/clk/sunxi-ng/ccu_mult.c
> index 826302464650..bf5e11c803f9 100644
> --- a/drivers/clk/sunxi-ng/ccu_mult.c
> +++ b/drivers/clk/sunxi-ng/ccu_mult.c
> @@ -85,7 +85,7 @@ static unsigned long ccu_mult_recalc_rate(struct clk_hw *hw,
> ccu_mux_helper_adjust_parent_for_prediv(&cm->common, &cm->mux, -1,
> &parent_rate);
>
> - return parent_rate * (val + 1);
> + return parent_rate * (val + cm->mult.offset);
> }
>
> static int ccu_mult_determine_rate(struct clk_hw *hw,
> @@ -122,7 +122,7 @@ static int ccu_mult_set_rate(struct clk_hw *hw, unsigned long rate,
> reg = readl(cm->common.base + cm->common.reg);
> reg &= ~GENMASK(cm->mult.width + cm->mult.shift - 1, cm->mult.shift);
>
> - writel(reg | ((_cm.mult - 1) << cm->mult.shift),
> + writel(reg | ((_cm.mult - cm->mult.offset) << cm->mult.shift),
> cm->common.base + cm->common.reg);
>
> spin_unlock_irqrestore(cm->common.lock, flags);
> diff --git a/drivers/clk/sunxi-ng/ccu_mult.h b/drivers/clk/sunxi-ng/ccu_mult.h
> index bd2e38b5a32a..84839641dfdf 100644
> --- a/drivers/clk/sunxi-ng/ccu_mult.h
> +++ b/drivers/clk/sunxi-ng/ccu_mult.h
> @@ -6,20 +6,28 @@
> #include "ccu_mux.h"
>
> struct ccu_mult_internal {
> + u8 offset;
> u8 shift;
> u8 width;
> u8 min;
> };
>
> -#define _SUNXI_CCU_MULT_MIN(_shift, _width, _min) \
> - { \
> - .shift = _shift, \
> - .width = _width, \
> - .min = _min, \
> +#define _SUNXI_CCU_MULT_OFFSET_MIN(_shift, _width, _offset, _min) \
> + { \
> + .min = _min, \
> + .offset = _offset, \
> + .shift = _shift, \
> + .width = _width, \
> }
>
> +#define _SUNXI_CCU_MULT_MIN(_shift, _width, _min) \
> + _SUNXI_CCU_MULT_OFFSET_MIN(_shift, _width, 1, _min)
> +
> +#define _SUNXI_CCU_MULT_OFFSET(_shift, _width, _offset) \
> + _SUNXI_CCU_MULT_OFFSET_MIN(_shift, _width, _offset, 1)
> +
> #define _SUNXI_CCU_MULT(_shift, _width) \
> - _SUNXI_CCU_MULT_MIN(_shift, _width, 1)
> + _SUNXI_CCU_MULT_OFFSET_MIN(_shift, _width, 1, 1)
>
> struct ccu_mult {
> u32 enable;
> diff --git a/drivers/clk/sunxi-ng/ccu_nk.c b/drivers/clk/sunxi-ng/ccu_nk.c
> index eaf0fdf78d2b..90117d3ead8c 100644
> --- a/drivers/clk/sunxi-ng/ccu_nk.c
> +++ b/drivers/clk/sunxi-ng/ccu_nk.c
> @@ -76,12 +76,17 @@ static unsigned long ccu_nk_recalc_rate(struct clk_hw *hw,
>
> n = reg >> nk->n.shift;
> n &= (1 << nk->n.width) - 1;
> + n += nk->n.offset;
> + if (!n)
> + n++;
>
> k = reg >> nk->k.shift;
> k &= (1 << nk->k.width) - 1;
> + k += nk->k.offset;
> + if (!k)
> + k++;
>
> - rate = parent_rate * (n + 1) * (k + 1);
> -
> + rate = parent_rate * n * k;
> if (nk->common.features & CCU_FEATURE_FIXED_POSTDIV)
> rate /= nk->fixed_post_div;
>
> @@ -135,8 +140,9 @@ static int ccu_nk_set_rate(struct clk_hw *hw, unsigned long rate,
> reg &= ~GENMASK(nk->n.width + nk->n.shift - 1, nk->n.shift);
> reg &= ~GENMASK(nk->k.width + nk->k.shift - 1, nk->k.shift);
>
> - writel(reg | ((_nk.k - 1) << nk->k.shift) | ((_nk.n - 1) << nk->n.shift),
> - nk->common.base + nk->common.reg);
> + reg |= (_nk.k - nk->k.offset) << nk->k.shift;
> + reg |= (_nk.n - nk->n.offset) << nk->n.shift;
> + writel(reg, nk->common.base + nk->common.reg);
>
> spin_unlock_irqrestore(nk->common.lock, flags);
>
> diff --git a/drivers/clk/sunxi-ng/ccu_nkm.c b/drivers/clk/sunxi-ng/ccu_nkm.c
> index fd3c6a9d987c..e0b4c914d7ac 100644
> --- a/drivers/clk/sunxi-ng/ccu_nkm.c
> +++ b/drivers/clk/sunxi-ng/ccu_nkm.c
> @@ -82,14 +82,23 @@ static unsigned long ccu_nkm_recalc_rate(struct clk_hw *hw,
>
> n = reg >> nkm->n.shift;
> n &= (1 << nkm->n.width) - 1;
> + n += nkm->n.offset;
> + if (!n)
> + n++;
>
> k = reg >> nkm->k.shift;
> k &= (1 << nkm->k.width) - 1;
> + k += nkm->k.offset;
> + if (!k)
> + k++;
>
> m = reg >> nkm->m.shift;
> m &= (1 << nkm->m.width) - 1;
> + m += nkm->m.offset;
> + if (!m)
> + m++;
>
> - rate = parent_rate * (n + 1) * (k + 1) / (m + 1);
> + rate = parent_rate * n * k / m;
> if (nkm->common.features & CCU_FEATURE_FIXED_POSTDIV)
> rate /= nkm->fixed_post_div;
>
> @@ -156,10 +165,9 @@ static int ccu_nkm_set_rate(struct clk_hw *hw, unsigned long rate,
> reg &= ~GENMASK(nkm->k.width + nkm->k.shift - 1, nkm->k.shift);
> reg &= ~GENMASK(nkm->m.width + nkm->m.shift - 1, nkm->m.shift);
>
> - reg |= (_nkm.n - 1) << nkm->n.shift;
> - reg |= (_nkm.k - 1) << nkm->k.shift;
> - reg |= (_nkm.m - 1) << nkm->m.shift;
> -
> + reg |= (_nkm.n - nkm->n.offset) << nkm->n.shift;
> + reg |= (_nkm.k - nkm->k.offset) << nkm->k.shift;
> + reg |= (_nkm.m - nkm->m.offset) << nkm->m.shift;
> writel(reg, nkm->common.base + nkm->common.reg);
>
> spin_unlock_irqrestore(nkm->common.lock, flags);
> diff --git a/drivers/clk/sunxi-ng/ccu_nkmp.c b/drivers/clk/sunxi-ng/ccu_nkmp.c
> index 684c42da3ebb..da2bba02b845 100644
> --- a/drivers/clk/sunxi-ng/ccu_nkmp.c
> +++ b/drivers/clk/sunxi-ng/ccu_nkmp.c
> @@ -88,17 +88,26 @@ static unsigned long ccu_nkmp_recalc_rate(struct clk_hw *hw,
>
> n = reg >> nkmp->n.shift;
> n &= (1 << nkmp->n.width) - 1;
> + n += nkmp->n.offset;
> + if (!n)
> + n++;
>
> k = reg >> nkmp->k.shift;
> k &= (1 << nkmp->k.width) - 1;
> + k += nkmp->k.offset;
> + if (!k)
> + k++;
>
> m = reg >> nkmp->m.shift;
> m &= (1 << nkmp->m.width) - 1;
> + m += nkmp->m.offset;
> + if (!m)
> + m++;
>
> p = reg >> nkmp->p.shift;
> p &= (1 << nkmp->p.width) - 1;
>
> - return (parent_rate * (n + 1) * (k + 1) >> p) / (m + 1);
> + return parent_rate * n * k >> p / m;
> }
>
> static long ccu_nkmp_round_rate(struct clk_hw *hw, unsigned long rate,
> @@ -148,9 +157,9 @@ static int ccu_nkmp_set_rate(struct clk_hw *hw, unsigned long rate,
> reg &= ~GENMASK(nkmp->m.width + nkmp->m.shift - 1, nkmp->m.shift);
> reg &= ~GENMASK(nkmp->p.width + nkmp->p.shift - 1, nkmp->p.shift);
>
> - reg |= (_nkmp.n - 1) << nkmp->n.shift;
> - reg |= (_nkmp.k - 1) << nkmp->k.shift;
> - reg |= (_nkmp.m - 1) << nkmp->m.shift;
> + reg |= (_nkmp.n - nkmp->n.offset) << nkmp->n.shift;
> + reg |= (_nkmp.k - nkmp->k.offset) << nkmp->k.shift;
> + reg |= (_nkmp.m - nkmp->m.offset) << nkmp->m.shift;
> reg |= ilog2(_nkmp.p) << nkmp->p.shift;
>
> writel(reg, nkmp->common.base + nkmp->common.reg);
> diff --git a/drivers/clk/sunxi-ng/ccu_nm.c b/drivers/clk/sunxi-ng/ccu_nm.c
> index c9f3b6c982f0..158d74e0215f 100644
> --- a/drivers/clk/sunxi-ng/ccu_nm.c
> +++ b/drivers/clk/sunxi-ng/ccu_nm.c
> @@ -80,11 +80,17 @@ static unsigned long ccu_nm_recalc_rate(struct clk_hw *hw,
>
> n = reg >> nm->n.shift;
> n &= (1 << nm->n.width) - 1;
> + n += nm->n.offset;
> + if (!n)
> + n++;
>
> m = reg >> nm->m.shift;
> m &= (1 << nm->m.width) - 1;
> + m += nm->m.offset;
> + if (!m)
> + m++;
>
> - return parent_rate * (n + 1) / (m + 1);
> + return parent_rate * n / m;
> }
>
> static long ccu_nm_round_rate(struct clk_hw *hw, unsigned long rate,
> @@ -129,8 +135,9 @@ static int ccu_nm_set_rate(struct clk_hw *hw, unsigned long rate,
> reg &= ~GENMASK(nm->n.width + nm->n.shift - 1, nm->n.shift);
> reg &= ~GENMASK(nm->m.width + nm->m.shift - 1, nm->m.shift);
>
> - writel(reg | ((_nm.m - 1) << nm->m.shift) | ((_nm.n - 1) << nm->n.shift),
> - nm->common.base + nm->common.reg);
> + reg |= (_nm.n - nm->n.offset) << nm->n.shift;
> + reg |= (_nm.m - nm->m.offset) << nm->m.shift;
> + writel(reg, nm->common.base + nm->common.reg);
>
> spin_unlock_irqrestore(nm->common.lock, flags);
>
> --
> git-series 0.8.11

The existing bits in this patch look good.

Regards
ChenYu

Next message: Chen-Yu Tsai: "Re: [PATCH 5/10] clk: sunxi-ng: Implement multiplier maximum"
Previous message: Baoquan He: "[PATCH v2 2/2] bnx2: Wait for in-flight DMA to complete at probe stage"
In reply to: Maxime Ripard: "[PATCH 3/10] clk: sunxi-ng: Implement multiplier offsets"
Next in thread: Maxime Ripard: "[PATCH 1/10] clk: sunxi-ng: multiplier: Add fractionnal support"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]