@@ -4,6 +4,7 @@ obj-y += \
clk.o \
clk-busy.o \
clk-cpu.o \
+ clk-divider-gate.o \
clk-fixup-div.o \
clk-fixup-mux.o \
clk-gate-exclusive.o \
new file mode 100644
@@ -0,0 +1,219 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * Copyright 2018 NXP.
+ * Dong Aisheng <aisheng.dong@nxp.com>
+ */
+
+#include <linux/clk-provider.h>
+#include <linux/err.h>
+#include <linux/io.h>
+#include <linux/slab.h>
+
+struct clk_divider_gate {
+ struct clk_divider divider;
+ u32 cached_val;
+};
+
+static inline struct clk_divider_gate *to_clk_divider_gate(struct clk_hw *hw)
+{
+ struct clk_divider *div = to_clk_divider(hw);
+
+ return container_of(div, struct clk_divider_gate, divider);
+}
+
+static unsigned long clk_divider_gate_recalc_rate_ro(struct clk_hw *hw,
+ unsigned long parent_rate)
+{
+ struct clk_divider *div = to_clk_divider(hw);
+ unsigned int val;
+
+ val = clk_readl(div->reg) >> div->shift;
+ val &= clk_div_mask(div->width);
+ if (!val)
+ return 0;
+
+ return divider_recalc_rate(hw, parent_rate, val, div->table,
+ div->flags, div->width);
+}
+
+static unsigned long clk_divider_gate_recalc_rate(struct clk_hw *hw,
+ unsigned long parent_rate)
+{
+ struct clk_divider_gate *div_gate = to_clk_divider_gate(hw);
+ struct clk_divider *div = to_clk_divider(hw);
+ unsigned long flags = 0;
+ unsigned int val;
+
+ spin_lock_irqsave(div->lock, flags);
+
+ if (!clk_hw_is_enabled(hw)) {
+ val = div_gate->cached_val;
+ } else {
+ val = clk_readl(div->reg) >> div->shift;
+ val &= clk_div_mask(div->width);
+ }
+
+ spin_unlock_irqrestore(div->lock, flags);
+
+ if (!val)
+ return 0;
+
+ return divider_recalc_rate(hw, parent_rate, val, div->table,
+ div->flags, div->width);
+}
+
+static long clk_divider_round_rate(struct clk_hw *hw, unsigned long rate,
+ unsigned long *prate)
+{
+ return clk_divider_ops.round_rate(hw, rate, prate);
+}
+
+static int clk_divider_gate_set_rate(struct clk_hw *hw, unsigned long rate,
+ unsigned long parent_rate)
+{
+ struct clk_divider_gate *div_gate = to_clk_divider_gate(hw);
+ struct clk_divider *div = to_clk_divider(hw);
+ unsigned long flags = 0;
+ int value;
+ u32 val;
+
+ value = divider_get_val(rate, parent_rate, div->table,
+ div->width, div->flags);
+ if (value < 0)
+ return value;
+
+ spin_lock_irqsave(div->lock, flags);
+
+ if (clk_hw_is_enabled(hw)) {
+ val = clk_readl(div->reg);
+ val &= ~(clk_div_mask(div->width) << div->shift);
+ val |= (u32)value << div->shift;
+ clk_writel(val, div->reg);
+ } else {
+ div_gate->cached_val = value;
+ }
+
+ spin_unlock_irqrestore(div->lock, flags);
+
+ return 0;
+}
+
+static int clk_divider_enable(struct clk_hw *hw)
+{
+ struct clk_divider_gate *div_gate = to_clk_divider_gate(hw);
+ struct clk_divider *div = to_clk_divider(hw);
+ unsigned long flags = 0;
+ u32 val;
+
+ if (!div_gate->cached_val) {
+ pr_err("%s: no valid preset rate\n", clk_hw_get_name(hw));
+ return -EINVAL;
+ }
+
+ spin_lock_irqsave(div->lock, flags);
+ /* restore div val */
+ val = clk_readl(div->reg);
+ val |= div_gate->cached_val << div->shift;
+ clk_writel(val, div->reg);
+
+ spin_unlock_irqrestore(div->lock, flags);
+
+ return 0;
+}
+
+static void clk_divider_disable(struct clk_hw *hw)
+{
+ struct clk_divider_gate *div_gate = to_clk_divider_gate(hw);
+ struct clk_divider *div = to_clk_divider(hw);
+ unsigned long flags = 0;
+ u32 val;
+
+ spin_lock_irqsave(div->lock, flags);
+
+ /* store the current div val */
+ val = clk_readl(div->reg) >> div->shift;
+ val &= clk_div_mask(div->width);
+ div_gate->cached_val = val;
+ clk_writel(0, div->reg);
+
+ spin_unlock_irqrestore(div->lock, flags);
+}
+
+static int clk_divider_is_enabled(struct clk_hw *hw)
+{
+ struct clk_divider *div = to_clk_divider(hw);
+ u32 val;
+
+ val = clk_readl(div->reg) >> div->shift;
+ val &= clk_div_mask(div->width);
+
+ return val ? 1 : 0;
+}
+
+static const struct clk_ops clk_divider_gate_ro_ops = {
+ .recalc_rate = clk_divider_gate_recalc_rate_ro,
+ .round_rate = clk_divider_round_rate,
+};
+
+static const struct clk_ops clk_divider_gate_ops = {
+ .recalc_rate = clk_divider_gate_recalc_rate,
+ .round_rate = clk_divider_round_rate,
+ .set_rate = clk_divider_gate_set_rate,
+ .enable = clk_divider_enable,
+ .disable = clk_divider_disable,
+ .is_enabled = clk_divider_is_enabled,
+};
+
+/*
+ * NOTE: In order to resue the most code from the common divider,
+ * we also design our divider following the way that provids an extra
+ * clk_divider_flags, however it's fixed to CLK_DIVIDER_ONE_BASED by
+ * default as our HW is. Besides that it supports only CLK_DIVIDER_READ_ONLY
+ * flag which can be specified by user flexibly.
+ */
+struct clk_hw *imx_clk_divider_gate(const char *name, const char *parent_name,
+ unsigned long flags, void __iomem *reg,
+ u8 shift, u8 width, u8 clk_divider_flags,
+ const struct clk_div_table *table,
+ spinlock_t *lock)
+{
+ struct clk_init_data init;
+ struct clk_divider_gate *div_gate;
+ struct clk_hw *hw;
+ u32 val;
+ int ret;
+
+ div_gate = kzalloc(sizeof(*div_gate), GFP_KERNEL);
+ if (!div_gate)
+ return ERR_PTR(-ENOMEM);
+
+ init.name = name;
+ if (clk_divider_flags & CLK_DIVIDER_READ_ONLY)
+ init.ops = &clk_divider_gate_ro_ops;
+ else
+ init.ops = &clk_divider_gate_ops;
+ init.flags = flags;
+ init.parent_names = parent_name ? &parent_name : NULL;
+ init.num_parents = parent_name ? 1 : 0;
+
+ div_gate->divider.reg = reg;
+ div_gate->divider.shift = shift;
+ div_gate->divider.width = width;
+ div_gate->divider.lock = lock;
+ div_gate->divider.table = table;
+ div_gate->divider.hw.init = &init;
+ div_gate->divider.flags = CLK_DIVIDER_ONE_BASED | clk_divider_flags;
+ /* cache gate status */
+ val = clk_readl(reg) >> shift;
+ val &= clk_div_mask(width);
+ div_gate->cached_val = val;
+
+ hw = &div_gate->divider.hw;
+ ret = clk_hw_register(NULL, hw);
+ if (ret) {
+ kfree(div_gate);
+ hw = ERR_PTR(ret);
+ }
+
+ return hw;
+}
@@ -220,4 +220,8 @@ struct clk *imx_clk_cpu(const char *name, const char *parent_name,
struct clk *div, struct clk *mux, struct clk *pll,
struct clk *step);
+struct clk_hw *imx_clk_divider_gate(const char *name, const char *parent_name,
+ unsigned long flags, void __iomem *reg, u8 shift, u8 width,
+ u8 clk_divider_flags, const struct clk_div_table *table,
+ spinlock_t *lock);
#endif
For dividers with zero indicating clock is disabled, instead of giving a warning each time like "clkx: Zero divisor and CLK_DIVIDER_ALLOW_ZERO not set" in exist code, we'd like to introduce enable/disable function for it. e.g. 000b - Clock disabled 001b - Divide by 1 010b - Divide by 2 ... Set rate when the clk is disabled will cache the rate request and only when the clk is enabled will the driver actually program the hardware to have the requested divider value. Similarly, when the clk is disabled we'll write a 0 there, but when the clk is enabled we'll restore whatever rate (divider) was chosen last. It does mean that recalc rate will be sort of odd, because when the clk is off it will return 0, and when the clk is on it will return the right rate. So to make things work, we'll need to return the cached rate in recalc rate when the clk is off and read the hardware when the clk is on. NOTE for the default off divider, the recalc rate will still return 0 as there's still no proper preset rate. Enable such divider will give user a reminder error message. Cc: Stephen Boyd <sboyd@codeaurora.org> Cc: Michael Turquette <mturquette@baylibre.com> Cc: Shawn Guo <shawnguo@kernel.org> Signed-off-by: Dong Aisheng <aisheng.dong@nxp.com> --- ChangeLog: v5->v6: * move gatable divider code into imx specific clock folder as suggested by Michael. Patch title also updated accordingly. v4->v5: * no changes v3->v4: * no changes v2->v3: * split normal and gate ops * fix the possible racy v1->v2: * add enable/disable for the type of CLK_DIVIDER_ZERO_GATE dividers --- drivers/clk/imx/Makefile | 1 + drivers/clk/imx/clk-divider-gate.c | 219 +++++++++++++++++++++++++++++++++++++ drivers/clk/imx/clk.h | 4 + 3 files changed, 224 insertions(+) create mode 100644 drivers/clk/imx/clk-divider-gate.c