1 The Common Clk Framework
2 Mike Turquette <mturquette@ti.com>
4 This document endeavours to explain the common clk framework details,
5 and how to port a platform over to this framework. It is not yet a
6 detailed explanation of the clock api in include/linux/clk.h, but
7 perhaps someday it will include that information.
9 Part 1 - introduction and interface split
11 The common clk framework is an interface to control the clock nodes
12 available on various devices today. This may come in the form of clock
13 gating, rate adjustment, muxing or other operations. This framework is
14 enabled with the CONFIG_COMMON_CLK option.
16 The interface itself is divided into two halves, each shielded from the
17 details of its counterpart. First is the common definition of struct
18 clk which unifies the framework-level accounting and infrastructure that
19 has traditionally been duplicated across a variety of platforms. Second
20 is a common implementation of the clk.h api, defined in
21 drivers/clk/clk.c. Finally there is struct clk_ops, whose operations
22 are invoked by the clk api implementation.
24 The second half of the interface is comprised of the hardware-specific
25 callbacks registered with struct clk_ops and the corresponding
26 hardware-specific structures needed to model a particular clock. For
27 the remainder of this document any reference to a callback in struct
28 clk_ops, such as .enable or .set_rate, implies the hardware-specific
29 implementation of that code. Likewise, references to struct clk_foo
30 serve as a convenient shorthand for the implementation of the
31 hardware-specific bits for the hypothetical "foo" hardware.
33 Tying the two halves of this interface together is struct clk_hw, which
34 is defined in struct clk_foo and pointed to within struct clk. This
35 allows for easy navigation between the two discrete halves of the common
38 Part 2 - common data structures and api
40 Below is the common struct clk definition from
41 include/linux/clk-private.h, modified for brevity:
45 const struct clk_ops *ops;
50 struct hlist_head children;
51 struct hlist_node child_node;
55 The members above make up the core of the clk tree topology. The clk
56 api itself defines several driver-facing functions which operate on
57 struct clk. That api is documented in include/linux/clk.h.
59 Platforms and devices utilizing the common struct clk use the struct
60 clk_ops pointer in struct clk to perform the hardware-specific parts of
61 the operations defined in clk.h:
64 int (*prepare)(struct clk_hw *hw);
65 void (*unprepare)(struct clk_hw *hw);
66 int (*enable)(struct clk_hw *hw);
67 void (*disable)(struct clk_hw *hw);
68 int (*is_enabled)(struct clk_hw *hw);
69 unsigned long (*recalc_rate)(struct clk_hw *hw,
70 unsigned long parent_rate);
71 long (*round_rate)(struct clk_hw *hw,
73 unsigned long *parent_rate);
74 long (*determine_rate)(struct clk_hw *hw,
76 unsigned long *best_parent_rate,
77 struct clk **best_parent_clk);
78 int (*set_parent)(struct clk_hw *hw, u8 index);
79 u8 (*get_parent)(struct clk_hw *hw);
80 int (*set_rate)(struct clk_hw *hw,
82 unsigned long parent_rate);
83 int (*set_rate_and_parent)(struct clk_hw *hw,
85 unsigned long parent_rate,
87 unsigned long (*recalc_accuracy)(struct clk_hw *hw,
88 unsigned long parent_accuracy);
89 void (*init)(struct clk_hw *hw);
90 int (*debug_init)(struct clk_hw *hw,
91 struct dentry *dentry);
94 Part 3 - hardware clk implementations
96 The strength of the common struct clk comes from its .ops and .hw pointers
97 which abstract the details of struct clk from the hardware-specific bits, and
98 vice versa. To illustrate consider the simple gateable clk implementation in
99 drivers/clk/clk-gate.c:
108 struct clk_gate contains struct clk_hw hw as well as hardware-specific
109 knowledge about which register and bit controls this clk's gating.
110 Nothing about clock topology or accounting, such as enable_count or
111 notifier_count, is needed here. That is all handled by the common
112 framework code and struct clk.
114 Let's walk through enabling this clk from driver code:
117 clk = clk_get(NULL, "my_gateable_clk");
122 The call graph for clk_enable is very simple:
125 clk->ops->enable(clk->hw);
128 [resolves struct clk gate with to_clk_gate(hw)]
129 clk_gate_set_bit(gate);
131 And the definition of clk_gate_set_bit:
133 static void clk_gate_set_bit(struct clk_gate *gate)
137 reg = __raw_readl(gate->reg);
138 reg |= BIT(gate->bit_idx);
139 writel(reg, gate->reg);
142 Note that to_clk_gate is defined as:
144 #define to_clk_gate(_hw) container_of(_hw, struct clk_gate, clk)
146 This pattern of abstraction is used for every clock hardware
149 Part 4 - supporting your own clk hardware
151 When implementing support for a new type of clock it only necessary to
152 include the following header:
154 #include <linux/clk-provider.h>
156 include/linux/clk.h is included within that header and clk-private.h
157 must never be included from the code which implements the operations for
158 a clock. More on that below in Part 5.
160 To construct a clk hardware structure for your platform you must define
165 ... hardware specific data goes here ...
168 To take advantage of your data you'll need to support valid operations
171 struct clk_ops clk_foo_ops {
172 .enable = &clk_foo_enable;
173 .disable = &clk_foo_disable;
176 Implement the above functions using container_of:
178 #define to_clk_foo(_hw) container_of(_hw, struct clk_foo, hw)
180 int clk_foo_enable(struct clk_hw *hw)
184 foo = to_clk_foo(hw);
186 ... perform magic on foo ...
191 Below is a matrix detailing which clk_ops are mandatory based upon the
192 hardware capabilities of that clock. A cell marked as "y" means
193 mandatory, a cell marked as "n" implies that either including that
194 callback is invalid or otherwise unnecessary. Empty cells are either
195 optional or must be evaluated on a case-by-case basis.
197 clock hardware characteristics
198 -----------------------------------------------------------
199 | gate | change rate | single parent | multiplexer | root |
200 |------|-------------|---------------|-------------|------|
202 .unprepare | | | | | |
204 .enable | y | | | | |
205 .disable | y | | | | |
206 .is_enabled | y | | | | |
208 .recalc_rate | | y | | | |
209 .round_rate | | y [1] | | | |
210 .determine_rate | | y [1] | | | |
211 .set_rate | | y | | | |
213 .set_parent | | | n | y | n |
214 .get_parent | | | n | y | n |
216 .recalc_accuracy| | | | | |
219 -----------------------------------------------------------
220 [1] either one of round_rate or determine_rate is required.
222 Finally, register your clock at run-time with a hardware-specific
223 registration function. This function simply populates struct clk_foo's
224 data and then passes the common struct clk parameters to the framework
229 See the basic clock types in drivers/clk/clk-*.c for examples.
231 Part 5 - static initialization of clock data
233 For platforms with many clocks (often numbering into the hundreds) it
234 may be desirable to statically initialize some clock data. This
235 presents a problem since the definition of struct clk should be hidden
236 from everyone except for the clock core in drivers/clk/clk.c.
238 To get around this problem struct clk's definition is exposed in
239 include/linux/clk-private.h along with some macros for more easily
240 initializing instances of the basic clock types. These clocks must
241 still be initialized with the common clock framework via a call to
244 clk-private.h must NEVER be included by code which implements struct
245 clk_ops callbacks, nor must it be included by any logic which pokes
246 around inside of struct clk at run-time. To do so is a layering
249 To better enforce this policy, always follow this simple rule: any
250 statically initialized clock data MUST be defined in a separate file
251 from the logic that implements its ops. Basically separate the logic
252 from the data and all is well.
254 Part 6 - Disabling clock gating of unused clocks
256 Sometimes during development it can be useful to be able to bypass the
257 default disabling of unused clocks. For example, if drivers aren't enabling
258 clocks properly but rely on them being on from the bootloader, bypassing
259 the disabling means that the driver will remain functional while the issues
262 To bypass this disabling, include "clk_ignore_unused" in the bootargs to the
267 The common clock framework uses two global locks, the prepare lock and the
270 The enable lock is a spinlock and is held across calls to the .enable,
271 .disable and .is_enabled operations. Those operations are thus not allowed to
272 sleep, and calls to the clk_enable(), clk_disable() and clk_is_enabled() API
273 functions are allowed in atomic context.
275 The prepare lock is a mutex and is held across calls to all other operations.
276 All those operations are allowed to sleep, and calls to the corresponding API
277 functions are not allowed in atomic context.
279 This effectively divides operations in two groups from a locking perspective.
281 Drivers don't need to manually protect resources shared between the operations
282 of one group, regardless of whether those resources are shared by multiple
283 clocks or not. However, access to resources that are shared between operations
284 of the two groups needs to be protected by the drivers. An example of such a
285 resource would be a register that controls both the clock rate and the clock
286 enable/disable state.
288 The clock framework is reentrant, in that a driver is allowed to call clock
289 framework functions from within its implementation of clock operations. This
290 can for instance cause a .set_rate operation of one clock being called from
291 within the .set_rate operation of another clock. This case must be considered
292 in the driver implementations, but the code flow is usually controlled by the
295 Note that locking must also be considered when code outside of the common
296 clock framework needs to access resources used by the clock operations. This
297 is considered out of scope of this document.