2005-04-17 05:20:36 +07:00
|
|
|
#ifndef __LINUX_PERCPU_H
|
|
|
|
#define __LINUX_PERCPU_H
|
2006-09-26 13:31:21 +07:00
|
|
|
|
2007-07-16 13:39:57 +07:00
|
|
|
#include <linux/preempt.h>
|
2005-04-17 05:20:36 +07:00
|
|
|
#include <linux/slab.h> /* For kmalloc() */
|
|
|
|
#include <linux/smp.h>
|
2006-09-26 13:31:21 +07:00
|
|
|
#include <linux/cpumask.h>
|
2009-03-06 12:33:58 +07:00
|
|
|
#include <linux/pfn.h>
|
2006-09-26 13:31:21 +07:00
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
#include <asm/percpu.h>
|
|
|
|
|
2009-03-06 12:33:58 +07:00
|
|
|
/* enough to cover all DEFINE_PER_CPUs in modules */
|
2007-05-03 00:27:11 +07:00
|
|
|
#ifdef CONFIG_MODULES
|
2009-03-06 12:33:58 +07:00
|
|
|
#define PERCPU_MODULE_RESERVE (8 << 10)
|
2007-05-03 00:27:11 +07:00
|
|
|
#else
|
2009-03-06 12:33:58 +07:00
|
|
|
#define PERCPU_MODULE_RESERVE 0
|
2005-04-17 05:20:36 +07:00
|
|
|
#endif
|
|
|
|
|
2009-03-06 12:33:58 +07:00
|
|
|
#ifndef PERCPU_ENOUGH_ROOM
|
2007-05-03 00:27:11 +07:00
|
|
|
#define PERCPU_ENOUGH_ROOM \
|
2009-03-06 12:33:58 +07:00
|
|
|
(ALIGN(__per_cpu_end - __per_cpu_start, SMP_CACHE_BYTES) + \
|
|
|
|
PERCPU_MODULE_RESERVE)
|
|
|
|
#endif
|
2007-05-03 00:27:11 +07:00
|
|
|
|
2006-09-26 13:30:53 +07:00
|
|
|
/*
|
|
|
|
* Must be an lvalue. Since @var must be a simple identifier,
|
|
|
|
* we force a syntax error here if it isn't.
|
|
|
|
*/
|
|
|
|
#define get_cpu_var(var) (*({ \
|
2006-10-06 14:43:58 +07:00
|
|
|
extern int simple_identifier_##var(void); \
|
2006-09-26 13:30:53 +07:00
|
|
|
preempt_disable(); \
|
|
|
|
&__get_cpu_var(var); }))
|
2005-04-17 05:20:36 +07:00
|
|
|
#define put_cpu_var(var) preempt_enable()
|
|
|
|
|
|
|
|
#ifdef CONFIG_SMP
|
|
|
|
|
2009-02-24 09:57:21 +07:00
|
|
|
/* minimum unit size, also is the maximum supported allocation size */
|
2009-03-06 12:33:58 +07:00
|
|
|
#define PCPU_MIN_UNIT_SIZE PFN_ALIGN(64 << 10)
|
2009-02-24 09:57:21 +07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* PERCPU_DYNAMIC_RESERVE indicates the amount of free area to piggy
|
2009-03-06 12:33:59 +07:00
|
|
|
* back on the first chunk for dynamic percpu allocation if arch is
|
|
|
|
* manually allocating and mapping it for faster access (as a part of
|
|
|
|
* large page mapping for example).
|
2009-02-24 09:57:21 +07:00
|
|
|
*
|
2009-03-06 12:33:59 +07:00
|
|
|
* The following values give between one and two pages of free space
|
|
|
|
* after typical minimal boot (2-way SMP, single disk and NIC) with
|
|
|
|
* both defconfig and a distro config on x86_64 and 32. More
|
|
|
|
* intelligent way to determine this would be nice.
|
2009-02-24 09:57:21 +07:00
|
|
|
*/
|
2009-03-06 12:33:59 +07:00
|
|
|
#if BITS_PER_LONG > 32
|
|
|
|
#define PERCPU_DYNAMIC_RESERVE (20 << 10)
|
|
|
|
#else
|
|
|
|
#define PERCPU_DYNAMIC_RESERVE (12 << 10)
|
|
|
|
#endif
|
2009-02-24 09:57:21 +07:00
|
|
|
|
2009-02-20 14:29:08 +07:00
|
|
|
extern void *pcpu_base_addr;
|
2009-08-14 13:00:51 +07:00
|
|
|
extern const unsigned long *pcpu_unit_offsets;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
percpu: introduce pcpu_alloc_info and pcpu_group_info
Till now, non-linear cpu->unit map was expressed using an integer
array which maps each cpu to a unit and used only by lpage allocator.
Although how many units have been placed in a single contiguos area
(group) is known while building unit_map, the information is lost when
the result is recorded into the unit_map array. For lpage allocator,
as all allocations are done by lpages and whether two adjacent lpages
are in the same group or not is irrelevant, this didn't cause any
problem. Non-linear cpu->unit mapping will be used for sparse
embedding and this grouping information is necessary for that.
This patch introduces pcpu_alloc_info which contains all the
information necessary for initializing percpu allocator.
pcpu_alloc_info contains array of pcpu_group_info which describes how
units are grouped and mapped to cpus. pcpu_group_info also has
base_offset field to specify its offset from the chunk's base address.
pcpu_build_alloc_info() initializes this field as if all groups are
allocated back-to-back as is currently done but this will be used to
sparsely place groups.
pcpu_alloc_info is a rather complex data structure which contains a
flexible array which in turn points to nested cpu_map arrays.
* pcpu_alloc_alloc_info() and pcpu_free_alloc_info() are provided to
help dealing with pcpu_alloc_info.
* pcpu_lpage_build_unit_map() is updated to build pcpu_alloc_info,
generalized and renamed to pcpu_build_alloc_info().
@cpu_distance_fn may be NULL indicating that all cpus are of
LOCAL_DISTANCE.
* pcpul_lpage_dump_cfg() is updated to process pcpu_alloc_info,
generalized and renamed to pcpu_dump_alloc_info(). It now also
prints which group each alloc unit belongs to.
* pcpu_setup_first_chunk() now takes pcpu_alloc_info instead of the
separate parameters. All first chunk allocators are updated to use
pcpu_build_alloc_info() to build alloc_info and call
pcpu_setup_first_chunk() with it. This has the side effect of
packing units for sparse possible cpus. ie. if cpus 0, 2 and 4 are
possible, they'll be assigned unit 0, 1 and 2 instead of 0, 2 and 4.
* x86 setup_pcpu_lpage() is updated to deal with alloc_info.
* sparc64 setup_per_cpu_areas() is updated to build alloc_info.
Although the changes made by this patch are pretty pervasive, it
doesn't cause any behavior difference other than packing of sparse
cpus. It mostly changes how information is passed among
initialization functions and makes room for more flexibility.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: David Miller <davem@davemloft.net>
2009-08-14 13:00:51 +07:00
|
|
|
struct pcpu_group_info {
|
|
|
|
int nr_units; /* aligned # of units */
|
|
|
|
unsigned long base_offset; /* base address offset */
|
|
|
|
unsigned int *cpu_map; /* unit->cpu map, empty
|
|
|
|
* entries contain NR_CPUS */
|
|
|
|
};
|
|
|
|
|
|
|
|
struct pcpu_alloc_info {
|
|
|
|
size_t static_size;
|
|
|
|
size_t reserved_size;
|
|
|
|
size_t dyn_size;
|
|
|
|
size_t unit_size;
|
|
|
|
size_t atom_size;
|
|
|
|
size_t alloc_size;
|
|
|
|
size_t __ai_size; /* internal, don't use */
|
|
|
|
int nr_groups; /* 0 if grouping unnecessary */
|
|
|
|
struct pcpu_group_info groups[];
|
|
|
|
};
|
|
|
|
|
2009-08-14 13:00:50 +07:00
|
|
|
enum pcpu_fc {
|
|
|
|
PCPU_FC_AUTO,
|
|
|
|
PCPU_FC_EMBED,
|
|
|
|
PCPU_FC_PAGE,
|
|
|
|
|
|
|
|
PCPU_FC_NR,
|
|
|
|
};
|
|
|
|
extern const char *pcpu_fc_names[PCPU_FC_NR];
|
|
|
|
|
|
|
|
extern enum pcpu_fc pcpu_chosen_fc;
|
|
|
|
|
2009-08-14 13:00:50 +07:00
|
|
|
typedef void * (*pcpu_fc_alloc_fn_t)(unsigned int cpu, size_t size,
|
|
|
|
size_t align);
|
2009-07-04 06:10:59 +07:00
|
|
|
typedef void (*pcpu_fc_free_fn_t)(void *ptr, size_t size);
|
|
|
|
typedef void (*pcpu_fc_populate_pte_fn_t)(unsigned long addr);
|
2009-07-04 06:11:00 +07:00
|
|
|
typedef int (pcpu_fc_cpu_distance_fn_t)(unsigned int from, unsigned int to);
|
2009-02-20 14:29:08 +07:00
|
|
|
|
percpu: introduce pcpu_alloc_info and pcpu_group_info
Till now, non-linear cpu->unit map was expressed using an integer
array which maps each cpu to a unit and used only by lpage allocator.
Although how many units have been placed in a single contiguos area
(group) is known while building unit_map, the information is lost when
the result is recorded into the unit_map array. For lpage allocator,
as all allocations are done by lpages and whether two adjacent lpages
are in the same group or not is irrelevant, this didn't cause any
problem. Non-linear cpu->unit mapping will be used for sparse
embedding and this grouping information is necessary for that.
This patch introduces pcpu_alloc_info which contains all the
information necessary for initializing percpu allocator.
pcpu_alloc_info contains array of pcpu_group_info which describes how
units are grouped and mapped to cpus. pcpu_group_info also has
base_offset field to specify its offset from the chunk's base address.
pcpu_build_alloc_info() initializes this field as if all groups are
allocated back-to-back as is currently done but this will be used to
sparsely place groups.
pcpu_alloc_info is a rather complex data structure which contains a
flexible array which in turn points to nested cpu_map arrays.
* pcpu_alloc_alloc_info() and pcpu_free_alloc_info() are provided to
help dealing with pcpu_alloc_info.
* pcpu_lpage_build_unit_map() is updated to build pcpu_alloc_info,
generalized and renamed to pcpu_build_alloc_info().
@cpu_distance_fn may be NULL indicating that all cpus are of
LOCAL_DISTANCE.
* pcpul_lpage_dump_cfg() is updated to process pcpu_alloc_info,
generalized and renamed to pcpu_dump_alloc_info(). It now also
prints which group each alloc unit belongs to.
* pcpu_setup_first_chunk() now takes pcpu_alloc_info instead of the
separate parameters. All first chunk allocators are updated to use
pcpu_build_alloc_info() to build alloc_info and call
pcpu_setup_first_chunk() with it. This has the side effect of
packing units for sparse possible cpus. ie. if cpus 0, 2 and 4 are
possible, they'll be assigned unit 0, 1 and 2 instead of 0, 2 and 4.
* x86 setup_pcpu_lpage() is updated to deal with alloc_info.
* sparc64 setup_per_cpu_areas() is updated to build alloc_info.
Although the changes made by this patch are pretty pervasive, it
doesn't cause any behavior difference other than packing of sparse
cpus. It mostly changes how information is passed among
initialization functions and makes room for more flexibility.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: David Miller <davem@davemloft.net>
2009-08-14 13:00:51 +07:00
|
|
|
extern struct pcpu_alloc_info * __init pcpu_alloc_alloc_info(int nr_groups,
|
|
|
|
int nr_units);
|
|
|
|
extern void __init pcpu_free_alloc_info(struct pcpu_alloc_info *ai);
|
|
|
|
|
|
|
|
extern struct pcpu_alloc_info * __init pcpu_build_alloc_info(
|
|
|
|
size_t reserved_size, ssize_t dyn_size,
|
|
|
|
size_t atom_size,
|
2009-08-14 13:00:51 +07:00
|
|
|
pcpu_fc_cpu_distance_fn_t cpu_distance_fn);
|
|
|
|
|
2009-08-14 13:00:51 +07:00
|
|
|
extern int __init pcpu_setup_first_chunk(const struct pcpu_alloc_info *ai,
|
|
|
|
void *base_addr);
|
2009-02-24 09:57:21 +07:00
|
|
|
|
2009-08-14 13:00:49 +07:00
|
|
|
#ifdef CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK
|
2009-08-14 13:00:52 +07:00
|
|
|
extern int __init pcpu_embed_first_chunk(size_t reserved_size, ssize_t dyn_size,
|
|
|
|
size_t atom_size,
|
|
|
|
pcpu_fc_cpu_distance_fn_t cpu_distance_fn,
|
|
|
|
pcpu_fc_alloc_fn_t alloc_fn,
|
|
|
|
pcpu_fc_free_fn_t free_fn);
|
2009-08-14 13:00:49 +07:00
|
|
|
#endif
|
2009-03-10 14:27:48 +07:00
|
|
|
|
2009-08-14 13:00:49 +07:00
|
|
|
#ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK
|
2009-08-14 13:00:51 +07:00
|
|
|
extern int __init pcpu_page_first_chunk(size_t reserved_size,
|
2009-07-04 06:10:59 +07:00
|
|
|
pcpu_fc_alloc_fn_t alloc_fn,
|
|
|
|
pcpu_fc_free_fn_t free_fn,
|
|
|
|
pcpu_fc_populate_pte_fn_t populate_pte_fn);
|
2009-08-14 13:00:49 +07:00
|
|
|
#endif
|
2009-07-04 06:10:59 +07:00
|
|
|
|
2009-02-20 14:29:08 +07:00
|
|
|
/*
|
|
|
|
* Use this to get to a cpu's version of the per-cpu object
|
|
|
|
* dynamically allocated. Non-atomic access to the current CPU's
|
|
|
|
* version should probably be combined with get_cpu()/put_cpu().
|
|
|
|
*/
|
2009-02-20 14:29:08 +07:00
|
|
|
#define per_cpu_ptr(ptr, cpu) SHIFT_PERCPU_PTR((ptr), per_cpu_offset((cpu)))
|
|
|
|
|
2009-03-06 12:33:59 +07:00
|
|
|
extern void *__alloc_reserved_percpu(size_t size, size_t align);
|
2009-02-20 14:29:08 +07:00
|
|
|
extern void *__alloc_percpu(size_t size, size_t align);
|
|
|
|
extern void free_percpu(void *__pdata);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2009-03-30 17:07:44 +07:00
|
|
|
#ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA
|
|
|
|
extern void __init setup_per_cpu_areas(void);
|
|
|
|
#endif
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
#else /* CONFIG_SMP */
|
|
|
|
|
2009-02-20 14:29:08 +07:00
|
|
|
#define per_cpu_ptr(ptr, cpu) ({ (void)(cpu); (ptr); })
|
2006-09-26 13:31:21 +07:00
|
|
|
|
2009-02-20 14:29:08 +07:00
|
|
|
static inline void *__alloc_percpu(size_t size, size_t align)
|
2006-09-26 13:31:21 +07:00
|
|
|
{
|
2009-02-20 14:29:08 +07:00
|
|
|
/*
|
|
|
|
* Can't easily make larger alignment work with kmalloc. WARN
|
|
|
|
* on it. Larger alignment should only be used for module
|
|
|
|
* percpu sections on SMP for which this path isn't used.
|
|
|
|
*/
|
2009-02-26 08:54:17 +07:00
|
|
|
WARN_ON_ONCE(align > SMP_CACHE_BYTES);
|
2009-02-25 20:36:45 +07:00
|
|
|
return kzalloc(size, GFP_KERNEL);
|
2006-09-26 13:31:21 +07:00
|
|
|
}
|
|
|
|
|
2009-02-20 14:29:08 +07:00
|
|
|
static inline void free_percpu(void *p)
|
2006-09-26 13:31:21 +07:00
|
|
|
{
|
2009-02-20 14:29:08 +07:00
|
|
|
kfree(p);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
2009-03-30 17:07:44 +07:00
|
|
|
static inline void __init setup_per_cpu_areas(void) { }
|
|
|
|
|
2009-07-15 21:35:14 +07:00
|
|
|
static inline void *pcpu_lpage_remapped(void *kaddr)
|
|
|
|
{
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
#endif /* CONFIG_SMP */
|
|
|
|
|
2009-02-20 14:29:08 +07:00
|
|
|
#define alloc_percpu(type) (type *)__alloc_percpu(sizeof(type), \
|
|
|
|
__alignof__(type))
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2009-04-11 02:02:40 +07:00
|
|
|
/*
|
|
|
|
* Optional methods for optimized non-lvalue per-cpu variable access.
|
|
|
|
*
|
|
|
|
* @var can be a percpu variable or a field of it and its size should
|
|
|
|
* equal char, int or long. percpu_read() evaluates to a lvalue and
|
|
|
|
* all others to void.
|
|
|
|
*
|
|
|
|
* These operations are guaranteed to be atomic w.r.t. preemption.
|
|
|
|
* The generic versions use plain get/put_cpu_var(). Archs are
|
|
|
|
* encouraged to implement single-instruction alternatives which don't
|
|
|
|
* require preemption protection.
|
|
|
|
*/
|
|
|
|
#ifndef percpu_read
|
|
|
|
# define percpu_read(var) \
|
|
|
|
({ \
|
|
|
|
typeof(per_cpu_var(var)) __tmp_var__; \
|
|
|
|
__tmp_var__ = get_cpu_var(var); \
|
|
|
|
put_cpu_var(var); \
|
|
|
|
__tmp_var__; \
|
|
|
|
})
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#define __percpu_generic_to_op(var, val, op) \
|
|
|
|
do { \
|
|
|
|
get_cpu_var(var) op val; \
|
|
|
|
put_cpu_var(var); \
|
|
|
|
} while (0)
|
|
|
|
|
|
|
|
#ifndef percpu_write
|
|
|
|
# define percpu_write(var, val) __percpu_generic_to_op(var, (val), =)
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifndef percpu_add
|
|
|
|
# define percpu_add(var, val) __percpu_generic_to_op(var, (val), +=)
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifndef percpu_sub
|
|
|
|
# define percpu_sub(var, val) __percpu_generic_to_op(var, (val), -=)
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifndef percpu_and
|
|
|
|
# define percpu_and(var, val) __percpu_generic_to_op(var, (val), &=)
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifndef percpu_or
|
|
|
|
# define percpu_or(var, val) __percpu_generic_to_op(var, (val), |=)
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifndef percpu_xor
|
|
|
|
# define percpu_xor(var, val) __percpu_generic_to_op(var, (val), ^=)
|
|
|
|
#endif
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
#endif /* __LINUX_PERCPU_H */
|