Commit c5905afb authored by Ingo Molnar's avatar Ingo Molnar

static keys: Introduce 'struct static_key', static_key_true()/false() and...

static keys: Introduce 'struct static_key', static_key_true()/false() and static_key_slow_[inc|dec]()

So here's a boot tested patch on top of Jason's series that does
all the cleanups I talked about and turns jump labels into a
more intuitive to use facility. It should also address the
various misconceptions and confusions that surround jump labels.

Typical usage scenarios:

        #include <linux/static_key.h>

        struct static_key key = STATIC_KEY_INIT_TRUE;

        if (static_key_false(&key))
                do unlikely code
        else
                do likely code

Or:

        if (static_key_true(&key))
                do likely code
        else
                do unlikely code

The static key is modified via:

        static_key_slow_inc(&key);
        ...
        static_key_slow_dec(&key);

The 'slow' prefix makes it abundantly clear that this is an
expensive operation.

I've updated all in-kernel code to use this everywhere. Note
that I (intentionally) have not pushed through the rename
blindly through to the lowest levels: the actual jump-label
patching arch facility should be named like that, so we want to
decouple jump labels from the static-key facility a bit.

On non-jump-label enabled architectures static keys default to
likely()/unlikely() branches.
Signed-off-by: default avatarIngo Molnar <mingo@elte.hu>
Acked-by: default avatarJason Baron <jbaron@redhat.com>
Acked-by: default avatarSteven Rostedt <rostedt@goodmis.org>
Cc: a.p.zijlstra@chello.nl
Cc: mathieu.desnoyers@efficios.com
Cc: davem@davemloft.net
Cc: ddaney.cavm@gmail.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20120222085809.GA26397@elte.huSigned-off-by: default avatarIngo Molnar <mingo@elte.hu>
parent 1cfa60dc
......@@ -47,18 +47,29 @@ config KPROBES
If in doubt, say "N".
config JUMP_LABEL
bool "Optimize trace point call sites"
bool "Optimize very unlikely/likely branches"
depends on HAVE_ARCH_JUMP_LABEL
help
This option enables a transparent branch optimization that
makes certain almost-always-true or almost-always-false branch
conditions even cheaper to execute within the kernel.
Certain performance-sensitive kernel code, such as trace points,
scheduler functionality, networking code and KVM have such
branches and include support for this optimization technique.
If it is detected that the compiler has support for "asm goto",
the kernel will compile trace point locations with just a
nop instruction. When trace points are enabled, the nop will
be converted to a jump to the trace function. This technique
lowers overhead and stress on the branch prediction of the
processor.
On i386, options added to the compiler flags may increase
the size of the kernel slightly.
the kernel will compile such branches with just a nop
instruction. When the condition flag is toggled to true, the
nop will be converted to a jump instruction to execute the
conditional block of instructions.
This technique lowers overhead and stress on the branch prediction
of the processor and generally makes the kernel faster. The update
of the condition is slower, but those are always very rare.
( On 32-bit x86, the necessary options added to the compiler
flags may increase the size of the kernel slightly. )
config OPTPROBES
def_bool y
......
......@@ -281,9 +281,9 @@ paravirt_init_missing_ticks_accounting(int cpu)
pv_time_ops.init_missing_ticks_accounting(cpu);
}
struct jump_label_key;
extern struct jump_label_key paravirt_steal_enabled;
extern struct jump_label_key paravirt_steal_rq_enabled;
struct static_key;
extern struct static_key paravirt_steal_enabled;
extern struct static_key paravirt_steal_rq_enabled;
static inline int
paravirt_do_steal_accounting(unsigned long *new_itm)
......
......@@ -634,8 +634,8 @@ struct pv_irq_ops pv_irq_ops = {
* pv_time_ops
* time operations
*/
struct jump_label_key paravirt_steal_enabled;
struct jump_label_key paravirt_steal_rq_enabled;
struct static_key paravirt_steal_enabled;
struct static_key paravirt_steal_rq_enabled;
static int
ia64_native_do_steal_accounting(unsigned long *new_itm)
......
......@@ -20,7 +20,7 @@
#define WORD_INSN ".word"
#endif
static __always_inline bool arch_static_branch(struct jump_label_key *key)
static __always_inline bool arch_static_branch(struct static_key *key)
{
asm goto("1:\tnop\n\t"
"nop\n\t"
......
......@@ -17,7 +17,7 @@
#define JUMP_ENTRY_TYPE stringify_in_c(FTR_ENTRY_LONG)
#define JUMP_LABEL_NOP_SIZE 4
static __always_inline bool arch_static_branch(struct jump_label_key *key)
static __always_inline bool arch_static_branch(struct static_key *key)
{
asm goto("1:\n\t"
"nop\n\t"
......
......@@ -13,7 +13,7 @@
#define ASM_ALIGN ".balign 4"
#endif
static __always_inline bool arch_static_branch(struct jump_label_key *key)
static __always_inline bool arch_static_branch(struct static_key *key)
{
asm goto("0: brcl 0,0\n"
".pushsection __jump_table, \"aw\"\n"
......
......@@ -7,7 +7,7 @@
#define JUMP_LABEL_NOP_SIZE 4
static __always_inline bool arch_static_branch(struct jump_label_key *key)
static __always_inline bool arch_static_branch(struct static_key *key)
{
asm goto("1:\n\t"
"nop\n\t"
......
......@@ -9,12 +9,12 @@
#define JUMP_LABEL_NOP_SIZE 5
#define JUMP_LABEL_INITIAL_NOP ".byte 0xe9 \n\t .long 0\n\t"
#define STATIC_KEY_INITIAL_NOP ".byte 0xe9 \n\t .long 0\n\t"
static __always_inline bool arch_static_branch(struct jump_label_key *key)
static __always_inline bool arch_static_branch(struct static_key *key)
{
asm goto("1:"
JUMP_LABEL_INITIAL_NOP
STATIC_KEY_INITIAL_NOP
".pushsection __jump_table, \"aw\" \n\t"
_ASM_ALIGN "\n\t"
_ASM_PTR "1b, %l[l_yes], %c0 \n\t"
......
......@@ -230,9 +230,9 @@ static inline unsigned long long paravirt_sched_clock(void)
return PVOP_CALL0(unsigned long long, pv_time_ops.sched_clock);
}
struct jump_label_key;
extern struct jump_label_key paravirt_steal_enabled;
extern struct jump_label_key paravirt_steal_rq_enabled;
struct static_key;
extern struct static_key paravirt_steal_enabled;
extern struct static_key paravirt_steal_rq_enabled;
static inline u64 paravirt_steal_clock(int cpu)
{
......
......@@ -438,9 +438,9 @@ void __init kvm_guest_init(void)
static __init int activate_jump_labels(void)
{
if (has_steal_clock) {
jump_label_inc(&paravirt_steal_enabled);
static_key_slow_inc(&paravirt_steal_enabled);
if (steal_acc)
jump_label_inc(&paravirt_steal_rq_enabled);
static_key_slow_inc(&paravirt_steal_rq_enabled);
}
return 0;
......
......@@ -202,8 +202,8 @@ static void native_flush_tlb_single(unsigned long addr)
__native_flush_tlb_single(addr);
}
struct jump_label_key paravirt_steal_enabled;
struct jump_label_key paravirt_steal_rq_enabled;
struct static_key paravirt_steal_enabled;
struct static_key paravirt_steal_rq_enabled;
static u64 native_steal_clock(int cpu)
{
......
......@@ -234,7 +234,7 @@ static void audit_vcpu_spte(struct kvm_vcpu *vcpu)
}
static bool mmu_audit;
static struct jump_label_key mmu_audit_key;
static struct static_key mmu_audit_key;
static void __kvm_mmu_audit(struct kvm_vcpu *vcpu, int point)
{
......@@ -250,7 +250,7 @@ static void __kvm_mmu_audit(struct kvm_vcpu *vcpu, int point)
static inline void kvm_mmu_audit(struct kvm_vcpu *vcpu, int point)
{
if (static_branch((&mmu_audit_key)))
if (static_key_false((&mmu_audit_key)))
__kvm_mmu_audit(vcpu, point);
}
......@@ -259,7 +259,7 @@ static void mmu_audit_enable(void)
if (mmu_audit)
return;
jump_label_inc(&mmu_audit_key);
static_key_slow_inc(&mmu_audit_key);
mmu_audit = true;
}
......@@ -268,7 +268,7 @@ static void mmu_audit_disable(void)
if (!mmu_audit)
return;
jump_label_dec(&mmu_audit_key);
static_key_slow_dec(&mmu_audit_key);
mmu_audit = false;
}
......
......@@ -9,15 +9,15 @@
*
* Jump labels provide an interface to generate dynamic branches using
* self-modifying code. Assuming toolchain and architecture support the result
* of a "if (static_branch(&key))" statement is a unconditional branch (which
* of a "if (static_key_false(&key))" statement is a unconditional branch (which
* defaults to false - and the true block is placed out of line).
*
* However at runtime we can change the 'static' branch target using
* jump_label_{inc,dec}(). These function as a 'reference' count on the key
* However at runtime we can change the branch target using
* static_key_slow_{inc,dec}(). These function as a 'reference' count on the key
* object and for as long as there are references all branches referring to
* that particular key will point to the (out of line) true block.
*
* Since this relies on modifying code the jump_label_{inc,dec}() functions
* Since this relies on modifying code the static_key_slow_{inc,dec}() functions
* must be considered absolute slow paths (machine wide synchronization etc.).
* OTOH, since the affected branches are unconditional their runtime overhead
* will be absolutely minimal, esp. in the default (off) case where the total
......@@ -26,12 +26,26 @@
*
* When the control is directly exposed to userspace it is prudent to delay the
* decrement to avoid high frequency code modifications which can (and do)
* cause significant performance degradation. Struct jump_label_key_deferred and
* jump_label_dec_deferred() provide for this.
* cause significant performance degradation. Struct static_key_deferred and
* static_key_slow_dec_deferred() provide for this.
*
* Lacking toolchain and or architecture support, it falls back to a simple
* conditional branch.
*/
*
* struct static_key my_key = STATIC_KEY_INIT_TRUE;
*
* if (static_key_true(&my_key)) {
* }
*
* will result in the true case being in-line and starts the key with a single
* reference. Mixing static_key_true() and static_key_false() on the same key is not
* allowed.
*
* Not initializing the key (static data is initialized to 0s anyway) is the
* same as using STATIC_KEY_INIT_FALSE and static_key_false() is
* equivalent with static_branch().
*
*/
#include <linux/types.h>
#include <linux/compiler.h>
......@@ -39,16 +53,17 @@
#if defined(CC_HAVE_ASM_GOTO) && defined(CONFIG_JUMP_LABEL)
struct jump_label_key {
struct static_key {
atomic_t enabled;
/* Set lsb bit to 1 if branch is default true, 0 ot */
struct jump_entry *entries;
#ifdef CONFIG_MODULES
struct jump_label_mod *next;
struct static_key_mod *next;
#endif
};
struct jump_label_key_deferred {
struct jump_label_key key;
struct static_key_deferred {
struct static_key key;
unsigned long timeout;
struct delayed_work work;
};
......@@ -66,13 +81,34 @@ struct module;
#ifdef HAVE_JUMP_LABEL
#ifdef CONFIG_MODULES
#define JUMP_LABEL_INIT {ATOMIC_INIT(0), NULL, NULL}
#else
#define JUMP_LABEL_INIT {ATOMIC_INIT(0), NULL}
#endif
#define JUMP_LABEL_TRUE_BRANCH 1UL
static
inline struct jump_entry *jump_label_get_entries(struct static_key *key)
{
return (struct jump_entry *)((unsigned long)key->entries
& ~JUMP_LABEL_TRUE_BRANCH);
}
static inline bool jump_label_get_branch_default(struct static_key *key)
{
if ((unsigned long)key->entries & JUMP_LABEL_TRUE_BRANCH)
return true;
return false;
}
static __always_inline bool static_key_false(struct static_key *key)
{
return arch_static_branch(key);
}
static __always_inline bool static_branch(struct jump_label_key *key)
static __always_inline bool static_key_true(struct static_key *key)
{
return !static_key_false(key);
}
/* Deprecated. Please use 'static_key_false() instead. */
static __always_inline bool static_branch(struct static_key *key)
{
return arch_static_branch(key);
}
......@@ -88,21 +124,24 @@ extern void arch_jump_label_transform(struct jump_entry *entry,
extern void arch_jump_label_transform_static(struct jump_entry *entry,
enum jump_label_type type);
extern int jump_label_text_reserved(void *start, void *end);
extern void jump_label_inc(struct jump_label_key *key);
extern void jump_label_dec(struct jump_label_key *key);
extern void jump_label_dec_deferred(struct jump_label_key_deferred *key);
extern bool jump_label_enabled(struct jump_label_key *key);
extern void static_key_slow_inc(struct static_key *key);
extern void static_key_slow_dec(struct static_key *key);
extern void static_key_slow_dec_deferred(struct static_key_deferred *key);
extern bool static_key_enabled(struct static_key *key);
extern void jump_label_apply_nops(struct module *mod);
extern void jump_label_rate_limit(struct jump_label_key_deferred *key,
unsigned long rl);
extern void
jump_label_rate_limit(struct static_key_deferred *key, unsigned long rl);
#define STATIC_KEY_INIT_TRUE ((struct static_key) \
{ .enabled = ATOMIC_INIT(1), .entries = (void *)1 })
#define STATIC_KEY_INIT_FALSE ((struct static_key) \
{ .enabled = ATOMIC_INIT(0), .entries = (void *)0 })
#else /* !HAVE_JUMP_LABEL */
#include <linux/atomic.h>
#define JUMP_LABEL_INIT {ATOMIC_INIT(0)}
struct jump_label_key {
struct static_key {
atomic_t enabled;
};
......@@ -110,30 +149,45 @@ static __always_inline void jump_label_init(void)
{
}
struct jump_label_key_deferred {
struct jump_label_key key;
struct static_key_deferred {
struct static_key key;
};
static __always_inline bool static_branch(struct jump_label_key *key)
static __always_inline bool static_key_false(struct static_key *key)
{
if (unlikely(atomic_read(&key->enabled)) > 0)
return true;
return false;
}
static __always_inline bool static_key_true(struct static_key *key)
{
if (unlikely(atomic_read(&key->enabled)))
if (likely(atomic_read(&key->enabled)) > 0)
return true;
return false;
}
static inline void jump_label_inc(struct jump_label_key *key)
/* Deprecated. Please use 'static_key_false() instead. */
static __always_inline bool static_branch(struct static_key *key)
{
if (unlikely(atomic_read(&key->enabled)) > 0)
return true;
return false;
}
static inline void static_key_slow_inc(struct static_key *key)
{
atomic_inc(&key->enabled);
}
static inline void jump_label_dec(struct jump_label_key *key)
static inline void static_key_slow_dec(struct static_key *key)
{
atomic_dec(&key->enabled);
}
static inline void jump_label_dec_deferred(struct jump_label_key_deferred *key)
static inline void static_key_slow_dec_deferred(struct static_key_deferred *key)
{
jump_label_dec(&key->key);
static_key_slow_dec(&key->key);
}
static inline int jump_label_text_reserved(void *start, void *end)
......@@ -144,9 +198,9 @@ static inline int jump_label_text_reserved(void *start, void *end)
static inline void jump_label_lock(void) {}
static inline void jump_label_unlock(void) {}
static inline bool jump_label_enabled(struct jump_label_key *key)
static inline bool static_key_enabled(struct static_key *key)
{
return !!atomic_read(&key->enabled);
return (atomic_read(&key->enabled) > 0);
}
static inline int jump_label_apply_nops(struct module *mod)
......@@ -154,13 +208,20 @@ static inline int jump_label_apply_nops(struct module *mod)
return 0;
}
static inline void jump_label_rate_limit(struct jump_label_key_deferred *key,
static inline void
jump_label_rate_limit(struct static_key_deferred *key,
unsigned long rl)
{
}
#define STATIC_KEY_INIT_TRUE ((struct static_key) \
{ .enabled = ATOMIC_INIT(1) })
#define STATIC_KEY_INIT_FALSE ((struct static_key) \
{ .enabled = ATOMIC_INIT(0) })
#endif /* HAVE_JUMP_LABEL */
#define jump_label_key_enabled ((struct jump_label_key){ .enabled = ATOMIC_INIT(1), })
#define jump_label_key_disabled ((struct jump_label_key){ .enabled = ATOMIC_INIT(0), })
#define STATIC_KEY_INIT STATIC_KEY_INIT_FALSE
#define jump_label_enabled static_key_enabled
#endif /* _LINUX_JUMP_LABEL_H */
......@@ -214,8 +214,8 @@ enum {
#include <linux/skbuff.h>
#ifdef CONFIG_RPS
#include <linux/jump_label.h>
extern struct jump_label_key rps_needed;
#include <linux/static_key.h>
extern struct static_key rps_needed;
#endif
struct neighbour;
......
......@@ -163,13 +163,13 @@ extern struct ctl_path nf_net_ipv4_netfilter_sysctl_path[];
extern struct list_head nf_hooks[NFPROTO_NUMPROTO][NF_MAX_HOOKS];
#if defined(CONFIG_JUMP_LABEL)
#include <linux/jump_label.h>
extern struct jump_label_key nf_hooks_needed[NFPROTO_NUMPROTO][NF_MAX_HOOKS];
#include <linux/static_key.h>
extern struct static_key nf_hooks_needed[NFPROTO_NUMPROTO][NF_MAX_HOOKS];
static inline bool nf_hooks_active(u_int8_t pf, unsigned int hook)
{
if (__builtin_constant_p(pf) &&
__builtin_constant_p(hook))
return static_branch(&nf_hooks_needed[pf][hook]);
return static_key_false(&nf_hooks_needed[pf][hook]);
return !list_empty(&nf_hooks[pf][hook]);
}
......
......@@ -514,7 +514,7 @@ struct perf_guest_info_callbacks {
#include <linux/ftrace.h>
#include <linux/cpu.h>
#include <linux/irq_work.h>
#include <linux/jump_label.h>
#include <linux/static_key.h>
#include <linux/atomic.h>
#include <asm/local.h>
......@@ -1038,7 +1038,7 @@ static inline int is_software_event(struct perf_event *event)
return event->pmu->task_ctx_nr == perf_sw_context;
}
extern struct jump_label_key perf_swevent_enabled[PERF_COUNT_SW_MAX];
extern struct static_key perf_swevent_enabled[PERF_COUNT_SW_MAX];
extern void __perf_sw_event(u32, u64, struct pt_regs *, u64);
......@@ -1066,7 +1066,7 @@ perf_sw_event(u32 event_id, u64 nr, struct pt_regs *regs, u64 addr)
{
struct pt_regs hot_regs;
if (static_branch(&perf_swevent_enabled[event_id])) {
if (static_key_false(&perf_swevent_enabled[event_id])) {
if (!regs) {
perf_fetch_caller_regs(&hot_regs);
regs = &hot_regs;
......@@ -1075,12 +1075,12 @@ perf_sw_event(u32 event_id, u64 nr, struct pt_regs *regs, u64 addr)
}
}
extern struct jump_label_key_deferred perf_sched_events;
extern struct static_key_deferred perf_sched_events;
static inline void perf_event_task_sched_in(struct task_struct *prev,
struct task_struct *task)
{
if (static_branch(&perf_sched_events.key))
if (static_key_false(&perf_sched_events.key))
__perf_event_task_sched_in(prev, task);
}
......@@ -1089,7 +1089,7 @@ static inline void perf_event_task_sched_out(struct task_struct *prev,
{
perf_sw_event(PERF_COUNT_SW_CONTEXT_SWITCHES, 1, NULL, 0);
if (static_branch(&perf_sched_events.key))
if (static_key_false(&perf_sched_events.key))
__perf_event_task_sched_out(prev, next);
}
......
#include <linux/jump_label.h>
......@@ -17,7 +17,7 @@
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/rcupdate.h>
#include <linux/jump_label.h>
#include <linux/static_key.h>
struct module;
struct tracepoint;
......@@ -29,7 +29,7 @@ struct tracepoint_func {
struct tracepoint {
const char *name; /* Tracepoint name */
struct jump_label_key key;
struct static_key key;
void (*regfunc)(void);
void (*unregfunc)(void);
struct tracepoint_func __rcu *funcs;
......@@ -145,7 +145,7 @@ static inline void tracepoint_synchronize_unregister(void)
extern struct tracepoint __tracepoint_##name; \
static inline void trace_##name(proto) \
{ \
if (static_branch(&__tracepoint_##name.key)) \
if (static_key_false(&__tracepoint_##name.key)) \
__DO_TRACE(&__tracepoint_##name, \
TP_PROTO(data_proto), \
TP_ARGS(data_args), \
......@@ -188,7 +188,7 @@ static inline void tracepoint_synchronize_unregister(void)
__attribute__((section("__tracepoints_strings"))) = #name; \
struct tracepoint __tracepoint_##name \
__attribute__((section("__tracepoints"))) = \
{ __tpstrtab_##name, JUMP_LABEL_INIT, reg, unreg, NULL };\
{ __tpstrtab_##name, STATIC_KEY_INIT_FALSE, reg, unreg, NULL };\
static struct tracepoint * const __tracepoint_ptr_##name __used \
__attribute__((section("__tracepoints_ptrs"))) = \
&__tracepoint_##name;
......
......@@ -55,7 +55,7 @@
#include <linux/uaccess.h>
#include <linux/memcontrol.h>
#include <linux/res_counter.h>
#include <linux/jump_label.h>
#include <linux/static_key.h>
#include <linux/filter.h>
#include <linux/rculist_nulls.h>
......@@ -924,13 +924,13 @@ inline void sk_refcnt_debug_release(const struct sock *sk)
#endif /* SOCK_REFCNT_DEBUG */
#if defined(CONFIG_CGROUP_MEM_RES_CTLR_KMEM) && defined(CONFIG_NET)
extern struct jump_label_key memcg_socket_limit_enabled;
extern struct static_key memcg_socket_limit_enabled;
static inline struct cg_proto *parent_cg_proto(struct proto *proto,
struct cg_proto *cg_proto)
{
return proto->proto_cgroup(parent_mem_cgroup(cg_proto->memcg));
}
#define mem_cgroup_sockets_enabled static_branch(&memcg_socket_limit_enabled)
#define mem_cgroup_sockets_enabled static_key_false(&memcg_socket_limit_enabled)
#else
#define mem_cgroup_sockets_enabled 0
static inline struct cg_proto *parent_cg_proto(struct proto *proto,
......
......@@ -128,7 +128,7 @@ enum event_type_t {
* perf_sched_events : >0 events exist
* perf_cgroup_events: >0 per-cpu cgroup events exist on this cpu
*/
struct jump_label_key_deferred perf_sched_events __read_mostly;
struct static_key_deferred perf_sched_events __read_mostly;
static DEFINE_PER_CPU(atomic_t, perf_cgroup_events);
static atomic_t nr_mmap_events __read_mostly;
......@@ -2769,7 +2769,7 @@ static void free_event(struct perf_event *event)
if (!event->parent) {
if (event->attach_state & PERF_ATTACH_TASK)
jump_label_dec_deferred(&perf_sched_events);
static_key_slow_dec_deferred(&perf_sched_events);
if (event->attr.mmap || event->attr.mmap_data)
atomic_dec(&nr_mmap_events);
if (event->attr.comm)
......@@ -2780,7 +2780,7 @@ static void free_event(struct perf_event *event)
put_callchain_buffers();
if (is_cgroup_event(event)) {
atomic_dec(&per_cpu(perf_cgroup_events, event->cpu));
jump_label_dec_deferred(&perf_sched_events);
static_key_slow_dec_deferred(&perf_sched_events);
}
}
......@@ -4982,7 +4982,7 @@ fail:
return err;
}
struct jump_label_key perf_swevent_enabled[PERF_COUNT_SW_MAX];
struct static_key perf_swevent_enabled[PERF_COUNT_SW_MAX];
static void sw_perf_event_destroy(struct perf_event *event)
{
......@@ -4990,7 +4990,7 @@ static void sw_perf_event_destroy(struct perf_event *event)
WARN_ON(event->parent);
jump_label_dec(&perf_swevent_enabled[event_id]);
static_key_slow_dec(&perf_swevent_enabled[event_id]);
swevent_hlist_put(event);
}
......@@ -5020,7 +5020,7 @@ static int perf_swevent_init(struct perf_event *event)
if (err)
return err;
jump_label_inc(&perf_swevent_enabled[event_id]);