Add a common struct clk
PATCH 1/2 Add a common struct clk
We currently have ~21 definitions of struct clk in the ARM architecture, each defined on a per-platform basis. This makes it difficult to define platform- (or architecture-) independent clock sources without making assumptions about struct clk, and impossible to compile two platforms with different struct clks into a single image.
This change is an effort to unify struct clk where possible, by defining a common struct clk, containing a set of clock operations. Different clock implementations can set their own operations, and have a standard interface for generic code. The callback interface is exposed to the kernel proper, while the clock implementations only need to be seen by the platform internals.
This allows us to share clock code among platforms, and makes it possible to dynamically create clock devices in platform-independent code.
Platforms can enable the generic struct clock through CONFIG_USE_COMMON_STRUCT_CLK. In this case, the clock infrastructure consists of a common struct clk:
struct clk {
const struct clk_ops *ops;
unsigned int enable_count;
int flags;
union {
struct mutex mutex;
spinlock_t spinlock;
} lock;
};And a set of clock operations (defined per type of clock):
struct clk_ops {
int (*enable)(struct clk *);
void (*disable)(struct clk *);
unsigned long (*get_rate)(struct clk *);
[...]
};To define a hardware-specific clock, machine code can "subclass" the struct clock into a new struct (adding any device-specific data), and provide a set of operations:
struct clk_foo {
struct clk clk;
void __iomem *some_register;
};struct clk_ops clk_foo_ops = {
.get_rate = clk_foo_get_rate,
};
The common clock definitions are based on a development patch from Ben Herrenschmidt <[email protected]>.
--- arch/Kconfig | 3 include/linux/clk.h | 158 +++++++++++++++++++++++++++++++++++++++++--- kernel/Makefile | 1 kernel/clk.c | 100 +++++++++++++++++++++++++++ 4 files changed, 253 insertions(+), 9 deletions(-)
arch/Kconfig +3 -0
@@ -165,6 +165,9 @@ config HAVE_MIXED_BREAKPOINTS_REGS config HAVE_USER_RETURN_NOTIFIER bool +config USE_COMMON_STRUCT_CLK + bool + config HAVE_PERF_EVENTS_NMI bool help
include/linux/clk.h +149 -9
185 lines not on this page · raw message
kernel/Makefile +1 -0
@@ -106,6 +106,7 @@ obj-$(CONFIG_PERF_EVENTS) += perf_event.o obj-$(CONFIG_HAVE_HW_BREAKPOINT) += hw_breakpoint.o obj-$(CONFIG_USER_RETURN_NOTIFIER) += user-return-notifier.o obj-$(CONFIG_PADATA) += padata.o +obj-$(CONFIG_USE_COMMON_STRUCT_CLK) += clk.o ifneq ($(CONFIG_SCHED_OMIT_FRAME_POINTER),y) # According to Alan Modra <[email protected]>, the -fno-omit-frame-pointer is
kernel/clk.c new file +100 -0
100 lines not on this page · raw message
PATCH 2/2 clk: Generic support for fixed-rate clocks
Since most platforms will need a fixed-rate clock, add one. This will also serve as a basic example of an implementation of struct clk.
--- include/linux/clk.h | 13 +++++++++++++ kernel/clk.c | 14 ++++++++++++++ 2 files changed, 27 insertions(+)
include/linux/clk.h +13 -0
@@ -158,6 +158,19 @@ static inline void clk_common_init(struct clk *clk) mutex_init(&clk->lock.mutex); } +/* Simple fixed-rate clock */ +struct clk_fixed { +struct clk clk; +unsigned long rate; +}; + +extern struct clk_ops clk_fixed_ops; + +#define INIT_CLK_FIXED (name, r) { \ + .clk = INIT_CLK(name.clk, clk_fixed_ops), \ + .rate = (r) \ +} +#else /* !CONFIG_USE_COMMON_STRUCT_CLK */ /*
kernel/clk.c +14 -0
@@ -98,3 +98,17 @@ struct clk *clk_get_parent(struct clk *clk)return ERR_PTR (-ENOSYS ); }EXPORT_SYMBOL_GPL (clk_get_parent); + +/* clk_fixed support */ + +#define to_clk_fixed (clk) (container_of(clk, struct clk_fixed, clk)) + +static unsigned long clk_fixed_get_rate (struct clk *clk) +{ +return to_clk_fixed (clk)->rate; +} + +struct clk_ops clk_fixed_ops = { + .get_rate = clk_fixed_get_rate, +}; +EXPORT_SYMBOL_GPL (clk_fixed_ops);
Hi Jeremy,
Quoted 169 lines · Jeremy Kerr
We currently have ~21 definitions of struct clk in the ARM architecture, each defined on a per-platform basis. This makes it difficult to define platform- (or architecture-) independent clock sources without making assumptions about struct clk, and impossible to compile two platforms with different struct clks into a single image.
This change is an effort to unify struct clk where possible, by defining a common struct clk, containing a set of clock operations. Different clock implementations can set their own operations, and have a standard interface for generic code. The callback interface is exposed to the kernel proper, while the clock implementations only need to be seen by the platform internals.
This allows us to share clock code among platforms, and makes it possible to dynamically create clock devices in platform-independent code.
Platforms can enable the generic struct clock through
CONFIG_USE_COMMON_STRUCT_CLK. In this case, the clock infrastructure consists of a common struct clk:struct clk { const struct clk_ops *ops; unsigned int enable_count; int flags; union { struct mutex mutex; spinlock_t spinlock; } lock; };And a set of clock operations (defined per type of clock):
struct clk_ops { int (*enable)(struct clk *); void (*disable)(struct clk *); unsigned long (*get_rate)(struct clk *); [...] };To define a hardware-specific clock, machine code can "subclass" the struct clock into a new struct (adding any device-specific data), and provide a set of operations:
struct clk_foo { struct clk clk; void __iomem *some_register; };struct
clk_opsclk_foo_ops= {
.get_rate =clk_foo_get_rate,
};The common clock definitions are based on a development patch from Ben Herrenschmidt <[email protected]>.
Signed-off-by Jeremy Kerr <[email protected]>Acked-by Paulius Zaleckas <[email protected]>--- arch/Kconfig | 3 include/linux/clk.h | 158 +++++++++++++++++++++++++++++++++++++++++--- kernel/Makefile | 1 kernel/clk.c | 100 +++++++++++++++++++++++++++ 4 files changed, 253 insertions(+), 9 deletions(-)arch/Kconfig +3
-0@@ -165,6 +165,9 @@ config HAVE_MIXED_BREAKPOINTS_REGS config HAVE_USER_RETURN_NOTIFIER bool +config USE_COMMON_STRUCT_CLK + bool + config HAVE_PERF_EVENTS_NMI bool helpinclude/linux/clk.h +70
-2@@ -3,6 +3,7 @@* * Copyright (C) 2004 ARM Limited. * Written by Deep Blue Solutions Limited. +* Copyright (c) 2010 Jeremy Kerr <[email protected]> * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as @@ -11,18 +12,163 @@#ifndef __LINUX_CLK_H#define __LINUX_CLK_H +#include <linux/err.h> +#include <linux/mutex.h> +#include <linux/spinlock.h> +struct device ; -/* -* The base API. +#ifdef CONFIG_USE_COMMON_STRUCT_CLK + +#define CLK_ATOMIC 0x1 + +/* If we're using the common struct clk, we define the base clk object here */ + +/** +* struct clk - hardware independent clock structure +* @clk_ops: implementation-specific ops for this clock +* @enable_count: count of clk_enable() calls active on this clock +* @flags: platform-independent flags +* @lock: lock for enable/disable or other HW-specific ops +* +* The base clock object, used by drivers for hardware-independent manipulation +* of clock lines. This will be 'subclassed' by device-specific implementations, +* which add device-specific data to struct clk. For example: +* +* struct clk_foo { +* struct clk; +* [device specific fields] +* }; +* +* The clock driver code will manage the device-specific data, and pass +* clk_foo.clk to the common clock code. The clock driver will be called +* through the @ops callbacks. +* +* The @lock member provides either a spinlock or a mutex to protect (at least) +* @enable_count. The type of lock used will depend on @flags; if CLK_ATOMIC is +* set, then the core clock code will use a spinlock, otherwise a mutex. This +* lock will be acquired during clk_enable and clk_disable, so for atomic +* clocks, these ops callbacks must not sleep. +* +* The choice of atomic or non-atomic clock depends on how the clock is enabled. +* Typically, you'll want to use a non-atomic clock. For clocks that need to be +* enabled/disabled in interrupt context, use CLK_ATOMIC. Note that atomic +* clocks with parents will typically cascade enable/disable operations to +* their parent, so the parent of an atomic clock *must* be atomic too. +*/ +struct clk { +const struct clk_ops *ops; +unsigned int enable_count; +int flags; +union { +struct mutex mutex; +spinlock_t spinlock; + } lock; +}; + +/* static initialiser for non-atomic clocks */ +#define INIT_CLK (name, o) { \ + .ops = &o, \ + .enable_count = 0, \ + .flags = 0, \ + .lock.mutex = __MUTEX_INITIALIZER(name.lock.mutex), \ +} + +/* static initialiser for atomic clocks */ +#define INIT_CLK_ATOMIC (name, o) { \ + .ops = &o, \ + .enable_count = 0, \ + .flags = CLK_ATOMIC, \ + .lock.spinlock = __SPIN_LOCK_UNLOCKED(name.lock.spinlock), \ +} + +/**
include/linux/clk.h line 95
Quoted 10 lines
include/linux/clk.h +10
-0+* clk_ops: Callback operations for clocks; these are to be provided by the +* clock implementation, and will be called by drivers through the clk_* API. +* +* @enable: Enable the clock. This must not return until the clock is +* generating a valid clock signal, usable by consumer devices. +* Called with clk->lock held. +* +* @disable: Disable the clock. Called with clk->lock held. +* +* @get / @put: Called by the core clock code to notify the driver about
I wonder if this is valid kerneldoc. The tab before / looks (IMHO) ugly. Maybe specify "driver" a bit more to distinguish from "drivers" above. "clk_ops driver"?
include/linux/clk.h line 99
Quoted 4 lines
include/linux/clk.h +4
-0+* refounts as clk is passed to drivers. Optional. +* +* For other callbacks, see the corresponding clk_* functions. Parameters and +* return values are passed directly from/to these API functions directly, or
duplicated "directly"
Quoted 25 lines
include/linux/clk.h +23
-0+* -ENOSYS is returned if the callback is NULL, see kernel/clk.c for +* implementation details. All are optional. */ +struct clk_ops { +int (*enable)(struct clk *); +void (*disable)(struct clk *); +int (*get)(struct clk *); +void (*put)(struct clk *); +unsigned long (*get_rate)(struct clk *); +long (*round_rate)(struct clk *,unsigned long ); +int (*set_rate)(struct clk *,unsigned long ); +int (*set_parent)(struct clk *,struct clk *); +struct clk * (*get_parent)(struct clk *); +}; +static inline void __clk_lock (struct clk *clk) +{ +if (clk->flags &CLK_ATOMIC ) +spin_lock (&clk->lock.spinlock); +else +mutex_lock (&clk->lock.mutex); +} + +static inline void __clk_unlock (struct clk *clk) +{
include/linux/clk.h line 134
Quoted 10 lines
include/linux/clk.h +10
-0+if (clk->flags &CLK_ATOMIC ) +spin_unlock (&clk->lock.spinlock); +else +mutex_unlock (&clk->lock.mutex); +} + +/** +* __clk_get - update clock-specific refcounter +* +* @clk: The clock to refcount
"The clock to update the refcount for"?
Quoted 85 lines
include/linux/clk.h +32
-7+* +* Before a clock is returned from clk_get, this function should be called +* to update any clock-specific refcounting. +* +* Returns non-zero on success, zero on failure. +* +* Drivers should not need this function; it is only needed by the +* arch-specific clk_get() implementations. +*/ +int __clk_get (struct clk *clk); + +/** +* clk_common_init - initialise a clock for driver usage +* +* Used for runtime intialization of clocks; you don't need to call this +* if your clock has been (statically) initialized with INIT_CLK. +*/ +static inline void clk_common_init (struct clk *clk) +{ + clk->enable_count =0 ; +if (clk->flags &CLK_ATOMIC ) +spin_lock_init (&clk->lock.spinlock); +else +mutex_init (&clk->lock.mutex); +} + +#else /* !CONFIG_USE_COMMON_STRUCT_CLK */ /* -* struct clk - an machine class defined object / cookie. +* Global clock object, actual structure is declared per-machine */ struct clk ; +static inline void clk_common_init (struct clk *clk) { } + +#endif /* !CONFIG_USE_COMMON_STRUCT_CLK */ +/** * clk_get - lookup and obtain a reference to a clock producer. * @dev: device for clock "consumer" @@ -83,12 +229,6 @@ unsigned long clk_get_rate(struct clk *clk);*/ void clk_put (struct clk *clk); - -/* -* The remaining APIs are optional for machine class support. -*/ - -/** * clk_round_rate - adjust a rate to the exact rate a clock can provide * @clk: clock source kernel/Makefile +1
-0@@ -106,6 +106,7 @@ obj-$(CONFIG_PERF_EVENTS) += perf_event.o obj-$(CONFIG_HAVE_HW_BREAKPOINT) += hw_breakpoint.o obj-$(CONFIG_USER_RETURN_NOTIFIER) += user-return-notifier.o obj-$(CONFIG_PADATA) += padata.o +obj-$(CONFIG_USE_COMMON_STRUCT_CLK) += clk.o ifneq ($(CONFIG_SCHED_OMIT_FRAME_POINTER),y) # According to Alan Modra <[email protected]>, the -fno-omit-frame-pointer iskernel/clk.c new file +13
-0@@ -0,0 +1,100 @@ +/* +* Copyright (C) 2010 Canonical Ltd <[email protected]> +* +* This program is free software; you can redistribute it and/or modify +* it under the terms of the GNU General Public License version 2 as +* published by the Free Software Foundation. +* +* Standard functionality for the common clock API. +*/ + +#include <linux/clk.h> +#include <linux/module.h> +
kernel/clk.c line 23
Quoted 10 lines
kernel/clk.c new file +10
-0+int clk_enable (struct clk *clk) +{ +int ret =0 ; + +if (!clk->ops->enable) +return 0 ; + +__clk_lock (clk); +if (!clk->enable_count) + ret = clk->ops->enable (clk);
I wonder if it's worth to handle parents here, e.g.
if (!clk->enable_count) {
struct clk *parent = clk_get_parent(clk);
if (parent) {
ret = clk_enable(parent);
if (ret)
return ret;
}
ret = clk->ops->enable(clk);
if (likely(!ret))
clk->enable_count++;
else if (parent)
clk_disable(parent);
}as they are quite common.
And I wonder further if it makes the code a bit more efficient to use:
if (!clk->enable_count++) {
... enable clock (and maybe parent)
if (unlikely(ret))
clk->enable_count--
...kernel/clk.c line 36
Quoted 13 lines
kernel/clk.c new file +13
-0+ +if (!ret) + clk->enable_count++; +__clk_unlock (clk); + +return ret; +} +EXPORT_SYMBOL_GPL (clk_enable); + +void clk_disable (struct clk *clk) +{ +if (!clk->ops->disable) +return ;
WARN_ON(!clk->enable_count) ?
Quoted 70 lines
kernel/clk.c new file +64
-0+ +__clk_lock (clk); + +if (!--clk->enable_count) + clk->ops->disable (clk); + +__clk_unlock (clk); +} +EXPORT_SYMBOL_GPL (clk_disable); + +unsigned long clk_get_rate (struct clk *clk) +{ +if (clk->ops->get_rate) +return clk->ops->get_rate (clk); +return 0 ; +} +EXPORT_SYMBOL_GPL (clk_get_rate); + +int __clk_get (struct clk *clk) +{ +if (clk->ops->get) +return clk->ops->get (clk); +return 1 ; +} +EXPORT_SYMBOL_GPL (__clk_get); + +void clk_put (struct clk *clk) +{ +if (clk->ops->put) + clk->ops->put (clk); +} +EXPORT_SYMBOL_GPL (clk_put); + +long clk_round_rate (struct clk *clk,unsigned long rate) +{ +if (clk->ops->round_rate) +return clk->ops->round_rate (clk, rate); +return -ENOSYS ; +} +EXPORT_SYMBOL_GPL (clk_round_rate); + +int clk_set_rate (struct clk *clk,unsigned long rate) +{ +if (clk->ops->set_rate) +return clk->ops->set_rate (clk, rate); +return -ENOSYS ; +} +EXPORT_SYMBOL_GPL (clk_set_rate); + +int clk_set_parent (struct clk *clk,struct clk *parent) +{ +if (clk->ops->set_parent) +return clk->ops->set_parent (clk, parent); +return -ENOSYS ; +} +EXPORT_SYMBOL_GPL (clk_set_parent); + +struct clk *clk_get_parent (struct clk *clk) +{ +if (clk->ops->get_parent) +return clk->ops->get_parent (clk); +return ERR_PTR (-ENOSYS ); +} +EXPORT_SYMBOL_GPL (clk_get_parent);_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel at lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Signature · 2 lines
Pengutronix e.K. | Uwe Kleine-K?nig | Industrial Linux Solutions | http://www.pengutronix.de/ |
Hi Uwe
Quoted 11 lines
Quoted 11 lines
+/**
+ *clk_ops: Callback operations for clocks; these are to be provided by the + * clock implementation, and will be called by drivers through the clk_* API. + *
+ * @enable: Enable the clock. This must not return until the clock is
+ * generating a valid clock signal, usable by consumer devices.
+ * Called withclk->lockheld.
+ *
+ * @disable: Disable the clock. Called withclk->lockheld.
+ *
+ * @get / @put: Called by the core clock code to notify the driver
about
Quoted 3 lines
I wonder if this is valid kerneldoc. The tab before / looks (IMHO) ugly.
Not valid kernel doc, so I'll fix that up. The tab was unintentional.
Quoted 2 lines
Maybe specify "driver" a bit more to distinguish from "drivers" above. "
clk_opsdriver"?
This is actually for refcounting for uses by device drivers (ie, not the clock provider), I've updated the comment:
/**
* struct clk_ops - Callback operations for clocks; these are to be provided
* by the clock implementation, and will be called by drivers through the
* clk_* API.
*
* @enable: Enable the clock. This must not return until the clock is
* generating a valid clock signal, usable by consumer devices.
* Called with clk->lock held.
*
* @disable: Disable the clock. Called with clk->lock held.
*
* @get: Called by the core clock code to increment the clock's
* refount as clk is passed to device drivers. Optional.
*
* @put: Called by the core clock code to decrement the clocks's
* refounts as clk is released from device drivers. Optional.
*
* For other callbacks, see the corresponding clk_* functions. Parameters and
* return values are passed directly from/to these API functions, or
* -ENOSYS is returned if the callback is NULL, see kernel/clk.c for
* implementation details. All are optional.
*/
Quoted 6 lines
Quoted 4 lines
+/**
+ *__clk_get- update clock-specific refcounter
+ *
+ * @clk: The clock to refcount"The clock to update the refcount for"?
I'm using refcount as a verb here; if this isn't clear I can come up with something else. Your solution splits the 'clock' and the 'for' which may be difficult to parse too. Let me know if you have any other suggestions :)
Quoted 17 lines
I wonder if it's worth to handle parents here, e.g.
if (!clk->enable_count) { struct clk *parent = clk_get_parent(clk); if (parent) { ret = clk_enable(parent); if (ret) return ret; } ret = clk->ops->enable(clk); if (likely(!ret)) clk->enable_count++; else if (parent) clk_disable(parent); }
Quoted 1 line
as they are quite common.
I'm not convinced we should do the parent handling in the core clock code. It's fairly easy to do the parent enable/disable in platform code, which should have explicit knowledge about whether or not the clock has a parent, and the semantics of how the parent/child clocks interact.
However, happy to discuss this further if needs be.
Quoted 6 lines
kernel/clk.c line 36
Quoted 4 lines
kernel/clk.c new file +4
-0+void clk_disable (struct clk *clk) +{ +if (!clk->ops->disable) +return ;WARN_ON(!clk->enable_count) ?
Yep, good idea. I'll do this check with the lock acquired.
Thanks for the comments, I've updated my tree accordingly (along with some other kerneldoc fixups). I'll wait to see if there is any other feedback and re-post next week.
Cheers,
Jeremy
PATCH 1/2 Add a common struct clk
Quoted 26 lines · Jeremy Kerr
Hi Uwe
Quoted 11 lines
Quoted 11 lines
+/**
+ *clk_ops: Callback operations for clocks; these are to be provided by the + * clock implementation, and will be called by drivers through the clk_* API. + *
+ * @enable: Enable the clock. This must not return until the clock is
+ * generating a valid clock signal, usable by consumer devices.
+ * Called withclk->lockheld.
+ *
+ * @disable: Disable the clock. Called withclk->lockheld.
+ *
+ * @get / @put: Called by the core clock code to notify the driverabout
Quoted 3 lines
I wonder if this is valid kerneldoc. The tab before / looks (IMHO) ugly.
Not valid kernel doc, so I'll fix that up. The tab was unintentional.
Quoted 2 lines
Maybe specify "driver" a bit more to distinguish from "drivers" above. "
clk_opsdriver"?This is actually for refcounting for uses by device drivers (ie, not the clock provider), I've updated the comment:
Quoted 13 lines
/**
* structclk_ops- Callback operations for clocks; these are to be provided
* by the clock implementation, and will be called by drivers through the
* clk_* API.
*
* @enable: Enable the clock. This must not return until the clock is
* generating a valid clock signal, usable by consumer devices.
* Called withclk->lockheld.
*
* @disable: Disable the clock. Called withclk->lockheld.
*
* @get: Called by the core clock code to increment the clock's
* refount as clk is passed to device drivers. Optional.
s/refount/refcount/ (once more below)
Quoted 3 lines
*
* @put: Called by the core clock code to decrement the clocks's
* refounts as clk is released from device drivers. Optional.
again inconsistent tabbing.
IMHO the wording above is better. This makes me unsure if the callback has to fiddle with the refcount.
Quoted 4 lines
*
* For other callbacks, see the corresponding clk_* functions. Parameters and
* return values are passed directly from/to these API functions, or
*-ENOSYSis returned if the callback is NULL, seekernel/clk.cfor
This is not true for clk_get_rate. This returns 0 if the callback isn't set.
Quoted 8 lines
* implementation details. All are optional.
*/Quoted 5 lines
Quoted 4 lines
+/**
+ *__clk_get- update clock-specific refcounter
+ *
+ * @clk: The clock to refcount
Quoted 5 lines
Quoted 1 line
"The clock to update the refcount for"?
I'm using refcount as a verb here; if this isn't clear I can come up with something else. Your solution splits the 'clock' and the 'for' which may be difficult to parse too. Let me know if you have any other suggestions :)
hmm, don't know. Keeping it as is is OK for me, too, if you don't consider my suggestion to be better.
Quoted 17 lines
Quoted 17 lines
I wonder if it's worth to handle parents here, e.g.
if (!clk->enable_count) { struct clk *parent = clk_get_parent(clk); if (parent) { ret = clk_enable(parent); if (ret) return ret; } ret = clk->ops->enable(clk); if (likely(!ret)) clk->enable_count++; else if (parent) clk_disable(parent); }
Quoted 8 lines
Quoted 1 line
as they are quite common.
I'm not convinced we should do the parent handling in the core clock code. It's fairly easy to do the parent enable/disable in platform code, which should have explicit knowledge about whether or not the clock has a parent, and the semantics of how the parent/child clocks interact.
However, happy to discuss this further if needs be.
As already said in #armlinux, in the meantime I agree.
Quoted 8 lines
Quoted 6 lines
kernel/clk.c line 36
Quoted 4 lines
kernel/clk.c new file +4
-0+void clk_disable (struct clk *clk) +{ +if (!clk->ops->disable) +return ;WARN_ON(!clk->enable_count) ?Yep, good idea. I'll do this check with the lock acquired.
fine.
Quoted 3 lines
Thanks for the comments, I've updated my tree accordingly (along with some other kerneldoc fixups). I'll wait to see if there is any other feedback and re-post next week.
fine, too.
Thanks for your efforts
Uwe
PS: I don't know who looks at your git tree, but you might want to update the arch branches to base on your current work.
Signature · 2 lines
Pengutronix e.K. | Uwe Kleine-K?nig | Industrial Linux Solutions | http://www.pengutronix.de/ |