2017-03-17 06:18:50 +00:00
|
|
|
// SPDX-License-Identifier: GPL-2.0
|
|
|
|
|
|
|
|
#include <linux/export.h>
|
|
|
|
#include <linux/log2.h>
|
|
|
|
#include <linux/percpu.h>
|
|
|
|
#include <linux/preempt.h>
|
|
|
|
#include <linux/rcupdate.h>
|
|
|
|
#include <linux/sched.h>
|
2022-09-24 05:33:13 +00:00
|
|
|
#include <linux/sched/clock.h>
|
2017-03-17 06:18:50 +00:00
|
|
|
#include <linux/sched/rt.h>
|
|
|
|
#include <linux/slab.h>
|
|
|
|
|
|
|
|
#include "six.h"
|
|
|
|
|
|
|
|
#ifdef DEBUG
|
|
|
|
#define EBUG_ON(cond) BUG_ON(cond)
|
|
|
|
#else
|
|
|
|
#define EBUG_ON(cond) do {} while (0)
|
|
|
|
#endif
|
|
|
|
|
2023-02-05 00:38:43 +00:00
|
|
|
#define six_acquire(l, t, r, ip) lock_acquire(l, 0, t, r, 1, NULL, ip)
|
|
|
|
#define six_release(l, ip) lock_release(l, ip)
|
2017-03-17 06:18:50 +00:00
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
static void do_six_unlock_type(struct six_lock *lock, enum six_lock_type type);
|
|
|
|
|
2017-03-17 06:18:50 +00:00
|
|
|
struct six_lock_vals {
|
|
|
|
/* Value we add to the lock in order to take the lock: */
|
|
|
|
u64 lock_val;
|
|
|
|
|
|
|
|
/* If the lock has this value (used as a mask), taking the lock fails: */
|
|
|
|
u64 lock_fail;
|
|
|
|
|
|
|
|
/* Value we add to the lock in order to release the lock: */
|
|
|
|
u64 unlock_val;
|
|
|
|
|
|
|
|
/* Mask that indicates lock is held for this type: */
|
|
|
|
u64 held_mask;
|
|
|
|
|
|
|
|
/* Waitlist we wakeup when releasing the lock: */
|
|
|
|
enum six_lock_type unlock_wakeup;
|
|
|
|
};
|
|
|
|
|
|
|
|
#define __SIX_LOCK_HELD_read __SIX_VAL(read_lock, ~0)
|
|
|
|
#define __SIX_LOCK_HELD_intent __SIX_VAL(intent_lock, ~0)
|
|
|
|
#define __SIX_LOCK_HELD_write __SIX_VAL(seq, 1)
|
|
|
|
|
|
|
|
#define LOCK_VALS { \
|
|
|
|
[SIX_LOCK_read] = { \
|
|
|
|
.lock_val = __SIX_VAL(read_lock, 1), \
|
|
|
|
.lock_fail = __SIX_LOCK_HELD_write + __SIX_VAL(write_locking, 1),\
|
|
|
|
.unlock_val = -__SIX_VAL(read_lock, 1), \
|
|
|
|
.held_mask = __SIX_LOCK_HELD_read, \
|
|
|
|
.unlock_wakeup = SIX_LOCK_write, \
|
|
|
|
}, \
|
|
|
|
[SIX_LOCK_intent] = { \
|
|
|
|
.lock_val = __SIX_VAL(intent_lock, 1), \
|
|
|
|
.lock_fail = __SIX_LOCK_HELD_intent, \
|
|
|
|
.unlock_val = -__SIX_VAL(intent_lock, 1), \
|
|
|
|
.held_mask = __SIX_LOCK_HELD_intent, \
|
|
|
|
.unlock_wakeup = SIX_LOCK_intent, \
|
|
|
|
}, \
|
|
|
|
[SIX_LOCK_write] = { \
|
|
|
|
.lock_val = __SIX_VAL(seq, 1), \
|
|
|
|
.lock_fail = __SIX_LOCK_HELD_read, \
|
|
|
|
.unlock_val = __SIX_VAL(seq, 1), \
|
|
|
|
.held_mask = __SIX_LOCK_HELD_write, \
|
|
|
|
.unlock_wakeup = SIX_LOCK_read, \
|
|
|
|
}, \
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void six_set_owner(struct six_lock *lock, enum six_lock_type type,
|
2022-08-26 23:22:24 +00:00
|
|
|
union six_lock_state old,
|
|
|
|
struct task_struct *owner)
|
2017-03-17 06:18:50 +00:00
|
|
|
{
|
|
|
|
if (type != SIX_LOCK_intent)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (!old.intent_lock) {
|
|
|
|
EBUG_ON(lock->owner);
|
2022-08-26 23:22:24 +00:00
|
|
|
lock->owner = owner;
|
2017-03-17 06:18:50 +00:00
|
|
|
} else {
|
|
|
|
EBUG_ON(lock->owner != current);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline unsigned pcpu_read_count(struct six_lock *lock)
|
|
|
|
{
|
|
|
|
unsigned read_count = 0;
|
|
|
|
int cpu;
|
|
|
|
|
|
|
|
for_each_possible_cpu(cpu)
|
|
|
|
read_count += *per_cpu_ptr(lock->readers, cpu);
|
|
|
|
return read_count;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This is probably up there with the more evil things I've done */
|
|
|
|
#define waitlist_bitnr(id) ilog2((((union six_lock_state) { .waiters = 1 << (id) }).l))
|
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
static int __do_six_trylock_type(struct six_lock *lock,
|
|
|
|
enum six_lock_type type,
|
|
|
|
struct task_struct *task,
|
|
|
|
bool try)
|
2017-03-17 06:18:50 +00:00
|
|
|
{
|
|
|
|
const struct six_lock_vals l[] = LOCK_VALS;
|
|
|
|
union six_lock_state old, new;
|
2022-08-26 23:22:24 +00:00
|
|
|
int ret;
|
2017-03-17 06:18:50 +00:00
|
|
|
u64 v;
|
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
EBUG_ON(type == SIX_LOCK_write && lock->owner != task);
|
2017-03-17 06:18:50 +00:00
|
|
|
EBUG_ON(type == SIX_LOCK_write && (lock->state.seq & 1));
|
|
|
|
EBUG_ON(type == SIX_LOCK_write && (try != !(lock->state.write_locking)));
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Percpu reader mode:
|
|
|
|
*
|
|
|
|
* The basic idea behind this algorithm is that you can implement a lock
|
|
|
|
* between two threads without any atomics, just memory barriers:
|
|
|
|
*
|
|
|
|
* For two threads you'll need two variables, one variable for "thread a
|
|
|
|
* has the lock" and another for "thread b has the lock".
|
|
|
|
*
|
|
|
|
* To take the lock, a thread sets its variable indicating that it holds
|
|
|
|
* the lock, then issues a full memory barrier, then reads from the
|
|
|
|
* other thread's variable to check if the other thread thinks it has
|
|
|
|
* the lock. If we raced, we backoff and retry/sleep.
|
|
|
|
*/
|
|
|
|
|
|
|
|
if (type == SIX_LOCK_read && lock->readers) {
|
|
|
|
preempt_disable();
|
|
|
|
this_cpu_inc(*lock->readers); /* signal that we own lock */
|
|
|
|
|
|
|
|
smp_mb();
|
|
|
|
|
|
|
|
old.v = READ_ONCE(lock->state.v);
|
|
|
|
ret = !(old.v & l[type].lock_fail);
|
|
|
|
|
|
|
|
this_cpu_sub(*lock->readers, !ret);
|
|
|
|
preempt_enable();
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If we failed because a writer was trying to take the
|
|
|
|
* lock, issue a wakeup because we might have caused a
|
|
|
|
* spurious trylock failure:
|
|
|
|
*/
|
2022-08-25 14:49:52 +00:00
|
|
|
if (old.write_locking)
|
2022-08-26 23:22:24 +00:00
|
|
|
ret = -1 - SIX_LOCK_write;
|
2017-03-17 06:18:50 +00:00
|
|
|
} else if (type == SIX_LOCK_write && lock->readers) {
|
|
|
|
if (try) {
|
|
|
|
atomic64_add(__SIX_VAL(write_locking, 1),
|
|
|
|
&lock->state.counter);
|
|
|
|
smp_mb__after_atomic();
|
2022-10-15 04:34:38 +00:00
|
|
|
} else if (!(lock->state.waiters & (1 << SIX_LOCK_write))) {
|
|
|
|
atomic64_add(__SIX_VAL(waiters, 1 << SIX_LOCK_write),
|
|
|
|
&lock->state.counter);
|
|
|
|
/*
|
|
|
|
* pairs with barrier after unlock and before checking
|
|
|
|
* for readers in unlock path
|
|
|
|
*/
|
|
|
|
smp_mb__after_atomic();
|
2017-03-17 06:18:50 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
ret = !pcpu_read_count(lock);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* On success, we increment lock->seq; also we clear
|
|
|
|
* write_locking unless we failed from the lock path:
|
|
|
|
*/
|
|
|
|
v = 0;
|
|
|
|
if (ret)
|
|
|
|
v += __SIX_VAL(seq, 1);
|
|
|
|
if (ret || try)
|
|
|
|
v -= __SIX_VAL(write_locking, 1);
|
|
|
|
|
|
|
|
if (try && !ret) {
|
|
|
|
old.v = atomic64_add_return(v, &lock->state.counter);
|
2022-08-26 23:22:24 +00:00
|
|
|
if (old.waiters & (1 << SIX_LOCK_read))
|
|
|
|
ret = -1 - SIX_LOCK_read;
|
2017-03-17 06:18:50 +00:00
|
|
|
} else {
|
|
|
|
atomic64_add(v, &lock->state.counter);
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
v = READ_ONCE(lock->state.v);
|
|
|
|
do {
|
|
|
|
new.v = old.v = v;
|
|
|
|
|
|
|
|
if (!(old.v & l[type].lock_fail)) {
|
|
|
|
new.v += l[type].lock_val;
|
|
|
|
|
|
|
|
if (type == SIX_LOCK_write)
|
|
|
|
new.write_locking = 0;
|
2022-08-25 14:49:52 +00:00
|
|
|
} else if (!try && !(new.waiters & (1 << type)))
|
2017-03-17 06:18:50 +00:00
|
|
|
new.waiters |= 1 << type;
|
|
|
|
else
|
|
|
|
break; /* waiting bit already set */
|
|
|
|
} while ((v = atomic64_cmpxchg_acquire(&lock->state.counter,
|
|
|
|
old.v, new.v)) != old.v);
|
|
|
|
|
|
|
|
ret = !(old.v & l[type].lock_fail);
|
|
|
|
|
|
|
|
EBUG_ON(ret && !(lock->state.v & l[type].held_mask));
|
|
|
|
}
|
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
if (ret > 0)
|
|
|
|
six_set_owner(lock, type, old, task);
|
2017-03-17 06:18:50 +00:00
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
EBUG_ON(type == SIX_LOCK_write && (try || ret > 0) && (lock->state.write_locking));
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
static inline void __six_lock_wakeup(struct six_lock *lock, enum six_lock_type lock_type)
|
|
|
|
{
|
|
|
|
struct six_lock_waiter *w, *next;
|
|
|
|
struct task_struct *task;
|
|
|
|
bool saw_one;
|
|
|
|
int ret;
|
|
|
|
again:
|
|
|
|
ret = 0;
|
|
|
|
saw_one = false;
|
|
|
|
raw_spin_lock(&lock->wait_lock);
|
|
|
|
|
|
|
|
list_for_each_entry_safe(w, next, &lock->wait_list, list) {
|
|
|
|
if (w->lock_want != lock_type)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
if (saw_one && lock_type != SIX_LOCK_read)
|
|
|
|
goto unlock;
|
|
|
|
saw_one = true;
|
|
|
|
|
|
|
|
ret = __do_six_trylock_type(lock, lock_type, w->task, false);
|
|
|
|
if (ret <= 0)
|
|
|
|
goto unlock;
|
|
|
|
|
|
|
|
__list_del(w->list.prev, w->list.next);
|
|
|
|
task = w->task;
|
|
|
|
/*
|
|
|
|
* Do no writes to @w besides setting lock_acquired - otherwise
|
|
|
|
* we would need a memory barrier:
|
|
|
|
*/
|
|
|
|
barrier();
|
|
|
|
w->lock_acquired = true;
|
|
|
|
wake_up_process(task);
|
|
|
|
}
|
|
|
|
|
|
|
|
clear_bit(waitlist_bitnr(lock_type), (unsigned long *) &lock->state.v);
|
|
|
|
unlock:
|
|
|
|
raw_spin_unlock(&lock->wait_lock);
|
|
|
|
|
|
|
|
if (ret < 0) {
|
|
|
|
lock_type = -ret - 1;
|
|
|
|
goto again;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void six_lock_wakeup(struct six_lock *lock,
|
|
|
|
union six_lock_state state,
|
|
|
|
enum six_lock_type lock_type)
|
|
|
|
{
|
|
|
|
if (lock_type == SIX_LOCK_write && state.read_lock)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (!(state.waiters & (1 << lock_type)))
|
|
|
|
return;
|
|
|
|
|
|
|
|
__six_lock_wakeup(lock, lock_type);
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool do_six_trylock_type(struct six_lock *lock,
|
|
|
|
enum six_lock_type type,
|
|
|
|
bool try)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = __do_six_trylock_type(lock, type, current, try);
|
|
|
|
if (ret < 0)
|
|
|
|
__six_lock_wakeup(lock, -ret - 1);
|
|
|
|
|
|
|
|
return ret > 0;
|
|
|
|
}
|
|
|
|
|
2017-03-17 06:18:50 +00:00
|
|
|
__always_inline __flatten
|
2023-02-05 00:38:43 +00:00
|
|
|
static bool __six_trylock_type(struct six_lock *lock, enum six_lock_type type,
|
|
|
|
unsigned long ip)
|
2017-03-17 06:18:50 +00:00
|
|
|
{
|
|
|
|
if (!do_six_trylock_type(lock, type, true))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
if (type != SIX_LOCK_write)
|
2023-02-05 00:38:43 +00:00
|
|
|
six_acquire(&lock->dep_map, 1, type == SIX_LOCK_read, ip);
|
2017-03-17 06:18:50 +00:00
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
__always_inline __flatten
|
|
|
|
static bool __six_relock_type(struct six_lock *lock, enum six_lock_type type,
|
2023-02-05 00:38:43 +00:00
|
|
|
unsigned seq, unsigned long ip)
|
2017-03-17 06:18:50 +00:00
|
|
|
{
|
|
|
|
const struct six_lock_vals l[] = LOCK_VALS;
|
|
|
|
union six_lock_state old;
|
|
|
|
u64 v;
|
|
|
|
|
|
|
|
EBUG_ON(type == SIX_LOCK_write);
|
|
|
|
|
|
|
|
if (type == SIX_LOCK_read &&
|
|
|
|
lock->readers) {
|
|
|
|
bool ret;
|
|
|
|
|
|
|
|
preempt_disable();
|
|
|
|
this_cpu_inc(*lock->readers);
|
|
|
|
|
|
|
|
smp_mb();
|
|
|
|
|
|
|
|
old.v = READ_ONCE(lock->state.v);
|
|
|
|
ret = !(old.v & l[type].lock_fail) && old.seq == seq;
|
|
|
|
|
|
|
|
this_cpu_sub(*lock->readers, !ret);
|
|
|
|
preempt_enable();
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Similar to the lock path, we may have caused a spurious write
|
|
|
|
* lock fail and need to issue a wakeup:
|
|
|
|
*/
|
2022-08-25 14:49:52 +00:00
|
|
|
if (old.write_locking)
|
|
|
|
six_lock_wakeup(lock, old, SIX_LOCK_write);
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
if (ret)
|
2023-02-05 00:38:43 +00:00
|
|
|
six_acquire(&lock->dep_map, 1, type == SIX_LOCK_read, ip);
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
v = READ_ONCE(lock->state.v);
|
|
|
|
do {
|
|
|
|
old.v = v;
|
|
|
|
|
|
|
|
if (old.seq != seq || old.v & l[type].lock_fail)
|
|
|
|
return false;
|
|
|
|
} while ((v = atomic64_cmpxchg_acquire(&lock->state.counter,
|
|
|
|
old.v,
|
|
|
|
old.v + l[type].lock_val)) != old.v);
|
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
six_set_owner(lock, type, old, current);
|
2017-03-17 06:18:50 +00:00
|
|
|
if (type != SIX_LOCK_write)
|
2023-02-05 00:38:43 +00:00
|
|
|
six_acquire(&lock->dep_map, 1, type == SIX_LOCK_read, ip);
|
2017-03-17 06:18:50 +00:00
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef CONFIG_SIX_LOCK_SPIN_ON_OWNER
|
|
|
|
|
2023-02-05 19:09:30 +00:00
|
|
|
static inline bool six_can_spin_on_owner(struct six_lock *lock)
|
2017-03-17 06:18:50 +00:00
|
|
|
{
|
|
|
|
struct task_struct *owner;
|
2023-02-05 19:09:30 +00:00
|
|
|
bool ret;
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
if (need_resched())
|
2023-02-05 19:09:30 +00:00
|
|
|
return false;
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
owner = READ_ONCE(lock->owner);
|
2023-02-05 19:09:30 +00:00
|
|
|
ret = !owner || owner_on_cpu(owner);
|
2017-03-17 06:18:50 +00:00
|
|
|
rcu_read_unlock();
|
2023-02-05 19:09:30 +00:00
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void six_set_nospin(struct six_lock *lock)
|
|
|
|
{
|
|
|
|
union six_lock_state old, new;
|
|
|
|
u64 v = READ_ONCE(lock->state.v);
|
|
|
|
|
|
|
|
do {
|
|
|
|
new.v = old.v = v;
|
|
|
|
new.nospin = true;
|
|
|
|
} while ((v = atomic64_cmpxchg(&lock->state.counter, old.v, new.v)) != old.v);
|
2017-03-17 06:18:50 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline bool six_spin_on_owner(struct six_lock *lock,
|
2023-02-05 19:09:30 +00:00
|
|
|
struct task_struct *owner,
|
|
|
|
u64 end_time)
|
2017-03-17 06:18:50 +00:00
|
|
|
{
|
|
|
|
bool ret = true;
|
2023-02-05 19:09:30 +00:00
|
|
|
unsigned loop = 0;
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
while (lock->owner == owner) {
|
|
|
|
/*
|
|
|
|
* Ensure we emit the owner->on_cpu, dereference _after_
|
|
|
|
* checking lock->owner still matches owner. If that fails,
|
|
|
|
* owner might point to freed memory. If it still matches,
|
|
|
|
* the rcu_read_lock() ensures the memory stays valid.
|
|
|
|
*/
|
|
|
|
barrier();
|
|
|
|
|
2023-02-05 19:09:30 +00:00
|
|
|
if (!owner_on_cpu(owner) || need_resched()) {
|
|
|
|
ret = false;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!(++loop & 0xf) && (time_after64(sched_clock(), end_time))) {
|
|
|
|
six_set_nospin(lock);
|
2017-03-17 06:18:50 +00:00
|
|
|
ret = false;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
cpu_relax();
|
|
|
|
}
|
|
|
|
rcu_read_unlock();
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline bool six_optimistic_spin(struct six_lock *lock, enum six_lock_type type)
|
|
|
|
{
|
|
|
|
struct task_struct *task = current;
|
2023-02-05 19:09:30 +00:00
|
|
|
u64 end_time;
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
if (type == SIX_LOCK_write)
|
|
|
|
return false;
|
|
|
|
|
|
|
|
preempt_disable();
|
|
|
|
if (!six_can_spin_on_owner(lock))
|
|
|
|
goto fail;
|
|
|
|
|
|
|
|
if (!osq_lock(&lock->osq))
|
|
|
|
goto fail;
|
|
|
|
|
2023-02-05 19:09:30 +00:00
|
|
|
end_time = sched_clock() + 10 * NSEC_PER_USEC;
|
|
|
|
|
2017-03-17 06:18:50 +00:00
|
|
|
while (1) {
|
|
|
|
struct task_struct *owner;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If there's an owner, wait for it to either
|
|
|
|
* release the lock or go to sleep.
|
|
|
|
*/
|
|
|
|
owner = READ_ONCE(lock->owner);
|
2023-02-05 19:09:30 +00:00
|
|
|
if (owner && !six_spin_on_owner(lock, owner, end_time))
|
2017-03-17 06:18:50 +00:00
|
|
|
break;
|
|
|
|
|
|
|
|
if (do_six_trylock_type(lock, type, false)) {
|
|
|
|
osq_unlock(&lock->osq);
|
|
|
|
preempt_enable();
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* When there's no owner, we might have preempted between the
|
|
|
|
* owner acquiring the lock and setting the owner field. If
|
|
|
|
* we're an RT task that will live-lock because we won't let
|
|
|
|
* the owner complete.
|
|
|
|
*/
|
|
|
|
if (!owner && (need_resched() || rt_task(task)))
|
|
|
|
break;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The cpu_relax() call is a compiler barrier which forces
|
|
|
|
* everything in this loop to be re-loaded. We don't need
|
|
|
|
* memory barriers as we'll eventually observe the right
|
|
|
|
* values at the cost of a few extra spins.
|
|
|
|
*/
|
|
|
|
cpu_relax();
|
|
|
|
}
|
|
|
|
|
|
|
|
osq_unlock(&lock->osq);
|
|
|
|
fail:
|
|
|
|
preempt_enable();
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If we fell out of the spin path because of need_resched(),
|
|
|
|
* reschedule now, before we try-lock again. This avoids getting
|
|
|
|
* scheduled out right after we obtained the lock.
|
|
|
|
*/
|
|
|
|
if (need_resched())
|
|
|
|
schedule();
|
|
|
|
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
#else /* CONFIG_SIX_LOCK_SPIN_ON_OWNER */
|
|
|
|
|
|
|
|
static inline bool six_optimistic_spin(struct six_lock *lock, enum six_lock_type type)
|
|
|
|
{
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
noinline
|
|
|
|
static int __six_lock_type_slowpath(struct six_lock *lock, enum six_lock_type type,
|
2022-08-27 20:22:51 +00:00
|
|
|
struct six_lock_waiter *wait,
|
2023-02-05 00:38:43 +00:00
|
|
|
six_lock_should_sleep_fn should_sleep_fn, void *p,
|
|
|
|
unsigned long ip)
|
2017-03-17 06:18:50 +00:00
|
|
|
{
|
|
|
|
union six_lock_state old;
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
if (type == SIX_LOCK_write) {
|
|
|
|
EBUG_ON(lock->state.write_locking);
|
|
|
|
atomic64_add(__SIX_VAL(write_locking, 1), &lock->state.counter);
|
|
|
|
smp_mb__after_atomic();
|
|
|
|
}
|
|
|
|
|
|
|
|
if (six_optimistic_spin(lock, type))
|
2022-08-26 23:22:24 +00:00
|
|
|
goto out;
|
2017-03-17 06:18:50 +00:00
|
|
|
|
2023-02-05 00:38:43 +00:00
|
|
|
lock_contended(&lock->dep_map, ip);
|
2017-03-17 06:18:50 +00:00
|
|
|
|
2022-08-27 20:22:51 +00:00
|
|
|
wait->task = current;
|
|
|
|
wait->lock_want = type;
|
2022-08-26 23:22:24 +00:00
|
|
|
wait->lock_acquired = false;
|
2022-08-25 14:49:52 +00:00
|
|
|
|
|
|
|
raw_spin_lock(&lock->wait_lock);
|
|
|
|
if (!(lock->state.waiters & (1 << type)))
|
|
|
|
set_bit(waitlist_bitnr(type), (unsigned long *) &lock->state.v);
|
2022-08-26 23:22:24 +00:00
|
|
|
/*
|
|
|
|
* Retry taking the lock after taking waitlist lock, have raced with an
|
|
|
|
* unlock:
|
|
|
|
*/
|
|
|
|
ret = __do_six_trylock_type(lock, type, current, false);
|
|
|
|
if (ret <= 0) {
|
|
|
|
wait->start_time = local_clock();
|
2022-09-24 05:33:13 +00:00
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
if (!list_empty(&lock->wait_list)) {
|
|
|
|
struct six_lock_waiter *last =
|
|
|
|
list_last_entry(&lock->wait_list,
|
|
|
|
struct six_lock_waiter, list);
|
2022-09-24 05:33:13 +00:00
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
if (time_before_eq64(wait->start_time, last->start_time))
|
|
|
|
wait->start_time = last->start_time + 1;
|
|
|
|
}
|
2022-09-24 05:33:13 +00:00
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
list_add_tail(&wait->list, &lock->wait_list);
|
|
|
|
}
|
2022-08-25 14:49:52 +00:00
|
|
|
raw_spin_unlock(&lock->wait_lock);
|
2017-03-17 06:18:50 +00:00
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
if (unlikely(ret > 0)) {
|
|
|
|
ret = 0;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (unlikely(ret < 0)) {
|
|
|
|
__six_lock_wakeup(lock, -ret - 1);
|
|
|
|
ret = 0;
|
|
|
|
}
|
|
|
|
|
2017-03-17 06:18:50 +00:00
|
|
|
while (1) {
|
|
|
|
set_current_state(TASK_UNINTERRUPTIBLE);
|
2022-08-26 23:22:24 +00:00
|
|
|
|
|
|
|
if (wait->lock_acquired)
|
2017-03-17 06:18:50 +00:00
|
|
|
break;
|
|
|
|
|
|
|
|
ret = should_sleep_fn ? should_sleep_fn(lock, p) : 0;
|
2022-08-26 23:22:24 +00:00
|
|
|
if (unlikely(ret)) {
|
|
|
|
raw_spin_lock(&lock->wait_lock);
|
|
|
|
if (!wait->lock_acquired)
|
|
|
|
list_del(&wait->list);
|
|
|
|
raw_spin_unlock(&lock->wait_lock);
|
|
|
|
|
|
|
|
if (wait->lock_acquired)
|
|
|
|
do_six_unlock_type(lock, type);
|
2017-03-17 06:18:50 +00:00
|
|
|
break;
|
2022-08-26 23:22:24 +00:00
|
|
|
}
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
schedule();
|
|
|
|
}
|
|
|
|
|
|
|
|
__set_current_state(TASK_RUNNING);
|
2022-08-26 23:22:24 +00:00
|
|
|
out:
|
|
|
|
if (ret && type == SIX_LOCK_write && lock->state.write_locking) {
|
2017-03-17 06:18:50 +00:00
|
|
|
old.v = atomic64_sub_return(__SIX_VAL(write_locking, 1),
|
|
|
|
&lock->state.counter);
|
|
|
|
six_lock_wakeup(lock, old, SIX_LOCK_read);
|
|
|
|
}
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2022-08-27 20:22:51 +00:00
|
|
|
__always_inline __flatten
|
|
|
|
static int __six_lock_type_waiter(struct six_lock *lock, enum six_lock_type type,
|
|
|
|
struct six_lock_waiter *wait,
|
2023-02-05 00:38:43 +00:00
|
|
|
six_lock_should_sleep_fn should_sleep_fn, void *p,
|
|
|
|
unsigned long ip)
|
2017-03-17 06:18:50 +00:00
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
2022-09-24 05:33:13 +00:00
|
|
|
wait->start_time = 0;
|
|
|
|
|
2017-03-17 06:18:50 +00:00
|
|
|
if (type != SIX_LOCK_write)
|
2023-02-05 00:38:43 +00:00
|
|
|
six_acquire(&lock->dep_map, 0, type == SIX_LOCK_read, ip);
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
ret = do_six_trylock_type(lock, type, true) ? 0
|
2023-02-05 00:38:43 +00:00
|
|
|
: __six_lock_type_slowpath(lock, type, wait, should_sleep_fn, p, ip);
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
if (ret && type != SIX_LOCK_write)
|
2023-02-05 00:38:43 +00:00
|
|
|
six_release(&lock->dep_map, ip);
|
2017-03-17 06:18:50 +00:00
|
|
|
if (!ret)
|
2023-02-05 00:38:43 +00:00
|
|
|
lock_acquired(&lock->dep_map, ip);
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2022-08-27 20:22:51 +00:00
|
|
|
__always_inline
|
|
|
|
static int __six_lock_type(struct six_lock *lock, enum six_lock_type type,
|
2023-02-05 00:38:43 +00:00
|
|
|
six_lock_should_sleep_fn should_sleep_fn, void *p,
|
|
|
|
unsigned long ip)
|
2022-08-27 20:22:51 +00:00
|
|
|
{
|
|
|
|
struct six_lock_waiter wait;
|
|
|
|
|
2023-02-05 00:38:43 +00:00
|
|
|
return __six_lock_type_waiter(lock, type, &wait, should_sleep_fn, p, ip);
|
2022-08-27 20:22:51 +00:00
|
|
|
}
|
|
|
|
|
2017-03-17 06:18:50 +00:00
|
|
|
__always_inline __flatten
|
2022-08-26 23:22:24 +00:00
|
|
|
static void do_six_unlock_type(struct six_lock *lock, enum six_lock_type type)
|
2017-03-17 06:18:50 +00:00
|
|
|
{
|
|
|
|
const struct six_lock_vals l[] = LOCK_VALS;
|
|
|
|
union six_lock_state state;
|
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
if (type == SIX_LOCK_intent)
|
2017-03-17 06:18:50 +00:00
|
|
|
lock->owner = NULL;
|
|
|
|
|
|
|
|
if (type == SIX_LOCK_read &&
|
|
|
|
lock->readers) {
|
|
|
|
smp_mb(); /* unlock barrier */
|
|
|
|
this_cpu_dec(*lock->readers);
|
|
|
|
smp_mb(); /* between unlocking and checking for waiters */
|
|
|
|
state.v = READ_ONCE(lock->state.v);
|
|
|
|
} else {
|
2023-02-05 19:09:30 +00:00
|
|
|
u64 v = l[type].unlock_val;
|
|
|
|
|
|
|
|
if (type != SIX_LOCK_read)
|
|
|
|
v -= lock->state.v & __SIX_VAL(nospin, 1);
|
|
|
|
|
2017-03-17 06:18:50 +00:00
|
|
|
EBUG_ON(!(lock->state.v & l[type].held_mask));
|
2023-02-05 19:09:30 +00:00
|
|
|
state.v = atomic64_add_return_release(v, &lock->state.counter);
|
2017-03-17 06:18:50 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
six_lock_wakeup(lock, state, l[type].unlock_wakeup);
|
|
|
|
}
|
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
__always_inline __flatten
|
2023-02-05 00:38:43 +00:00
|
|
|
static void __six_unlock_type(struct six_lock *lock, enum six_lock_type type,
|
|
|
|
unsigned long ip)
|
2022-08-26 23:22:24 +00:00
|
|
|
{
|
|
|
|
EBUG_ON(type == SIX_LOCK_write &&
|
|
|
|
!(lock->state.v & __SIX_LOCK_HELD_intent));
|
|
|
|
EBUG_ON((type == SIX_LOCK_write ||
|
|
|
|
type == SIX_LOCK_intent) &&
|
|
|
|
lock->owner != current);
|
|
|
|
|
|
|
|
if (type != SIX_LOCK_write)
|
2023-02-05 00:38:43 +00:00
|
|
|
six_release(&lock->dep_map, ip);
|
2022-08-26 23:22:24 +00:00
|
|
|
|
|
|
|
if (type == SIX_LOCK_intent &&
|
|
|
|
lock->intent_lock_recurse) {
|
|
|
|
--lock->intent_lock_recurse;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
do_six_unlock_type(lock, type);
|
|
|
|
}
|
|
|
|
|
2017-03-17 06:18:50 +00:00
|
|
|
#define __SIX_LOCK(type) \
|
2023-02-05 00:38:43 +00:00
|
|
|
bool six_trylock_ip_##type(struct six_lock *lock, unsigned long ip) \
|
2017-03-17 06:18:50 +00:00
|
|
|
{ \
|
2023-02-05 00:38:43 +00:00
|
|
|
return __six_trylock_type(lock, SIX_LOCK_##type, ip); \
|
2017-03-17 06:18:50 +00:00
|
|
|
} \
|
2023-02-05 00:38:43 +00:00
|
|
|
EXPORT_SYMBOL_GPL(six_trylock_ip_##type); \
|
2017-03-17 06:18:50 +00:00
|
|
|
\
|
2023-02-05 00:38:43 +00:00
|
|
|
bool six_relock_ip_##type(struct six_lock *lock, u32 seq, unsigned long ip)\
|
2017-03-17 06:18:50 +00:00
|
|
|
{ \
|
2023-02-05 00:38:43 +00:00
|
|
|
return __six_relock_type(lock, SIX_LOCK_##type, seq, ip); \
|
2017-03-17 06:18:50 +00:00
|
|
|
} \
|
2023-02-05 00:38:43 +00:00
|
|
|
EXPORT_SYMBOL_GPL(six_relock_ip_##type); \
|
2017-03-17 06:18:50 +00:00
|
|
|
\
|
2023-02-05 00:38:43 +00:00
|
|
|
int six_lock_ip_##type(struct six_lock *lock, \
|
|
|
|
six_lock_should_sleep_fn should_sleep_fn, void *p, \
|
|
|
|
unsigned long ip) \
|
2017-03-17 06:18:50 +00:00
|
|
|
{ \
|
2023-02-05 00:38:43 +00:00
|
|
|
return __six_lock_type(lock, SIX_LOCK_##type, should_sleep_fn, p, ip);\
|
2017-03-17 06:18:50 +00:00
|
|
|
} \
|
2023-02-05 00:38:43 +00:00
|
|
|
EXPORT_SYMBOL_GPL(six_lock_ip_##type); \
|
2017-03-17 06:18:50 +00:00
|
|
|
\
|
2023-02-05 00:38:43 +00:00
|
|
|
int six_lock_ip_waiter_##type(struct six_lock *lock, \
|
2022-08-27 20:22:51 +00:00
|
|
|
struct six_lock_waiter *wait, \
|
2023-02-05 00:38:43 +00:00
|
|
|
six_lock_should_sleep_fn should_sleep_fn, void *p,\
|
|
|
|
unsigned long ip) \
|
2022-08-27 20:22:51 +00:00
|
|
|
{ \
|
2023-02-05 00:38:43 +00:00
|
|
|
return __six_lock_type_waiter(lock, SIX_LOCK_##type, wait, should_sleep_fn, p, ip);\
|
2022-08-27 20:22:51 +00:00
|
|
|
} \
|
2023-02-05 00:38:43 +00:00
|
|
|
EXPORT_SYMBOL_GPL(six_lock_ip_waiter_##type); \
|
2022-08-27 20:22:51 +00:00
|
|
|
\
|
2023-02-05 00:38:43 +00:00
|
|
|
void six_unlock_ip_##type(struct six_lock *lock, unsigned long ip) \
|
2017-03-17 06:18:50 +00:00
|
|
|
{ \
|
2023-02-05 00:38:43 +00:00
|
|
|
__six_unlock_type(lock, SIX_LOCK_##type, ip); \
|
2017-03-17 06:18:50 +00:00
|
|
|
} \
|
2023-02-05 00:38:43 +00:00
|
|
|
EXPORT_SYMBOL_GPL(six_unlock_ip_##type);
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
__SIX_LOCK(read)
|
|
|
|
__SIX_LOCK(intent)
|
|
|
|
__SIX_LOCK(write)
|
|
|
|
|
|
|
|
#undef __SIX_LOCK
|
|
|
|
|
|
|
|
/* Convert from intent to read: */
|
|
|
|
void six_lock_downgrade(struct six_lock *lock)
|
|
|
|
{
|
|
|
|
six_lock_increment(lock, SIX_LOCK_read);
|
|
|
|
six_unlock_intent(lock);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(six_lock_downgrade);
|
|
|
|
|
|
|
|
bool six_lock_tryupgrade(struct six_lock *lock)
|
|
|
|
{
|
|
|
|
union six_lock_state old, new;
|
|
|
|
u64 v = READ_ONCE(lock->state.v);
|
|
|
|
|
|
|
|
do {
|
|
|
|
new.v = old.v = v;
|
|
|
|
|
|
|
|
if (new.intent_lock)
|
|
|
|
return false;
|
|
|
|
|
|
|
|
if (!lock->readers) {
|
|
|
|
EBUG_ON(!new.read_lock);
|
|
|
|
new.read_lock--;
|
|
|
|
}
|
|
|
|
|
|
|
|
new.intent_lock = 1;
|
|
|
|
} while ((v = atomic64_cmpxchg_acquire(&lock->state.counter,
|
|
|
|
old.v, new.v)) != old.v);
|
|
|
|
|
|
|
|
if (lock->readers)
|
|
|
|
this_cpu_dec(*lock->readers);
|
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
six_set_owner(lock, SIX_LOCK_intent, old, current);
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(six_lock_tryupgrade);
|
|
|
|
|
|
|
|
bool six_trylock_convert(struct six_lock *lock,
|
|
|
|
enum six_lock_type from,
|
|
|
|
enum six_lock_type to)
|
|
|
|
{
|
|
|
|
EBUG_ON(to == SIX_LOCK_write || from == SIX_LOCK_write);
|
|
|
|
|
|
|
|
if (to == from)
|
|
|
|
return true;
|
|
|
|
|
|
|
|
if (to == SIX_LOCK_read) {
|
|
|
|
six_lock_downgrade(lock);
|
|
|
|
return true;
|
|
|
|
} else {
|
|
|
|
return six_lock_tryupgrade(lock);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(six_trylock_convert);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Increment read/intent lock count, assuming we already have it read or intent
|
|
|
|
* locked:
|
|
|
|
*/
|
|
|
|
void six_lock_increment(struct six_lock *lock, enum six_lock_type type)
|
|
|
|
{
|
|
|
|
const struct six_lock_vals l[] = LOCK_VALS;
|
|
|
|
|
2023-02-05 00:38:43 +00:00
|
|
|
six_acquire(&lock->dep_map, 0, type == SIX_LOCK_read, _RET_IP_);
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
/* XXX: assert already locked, and that we don't overflow: */
|
|
|
|
|
|
|
|
switch (type) {
|
|
|
|
case SIX_LOCK_read:
|
|
|
|
if (lock->readers) {
|
|
|
|
this_cpu_inc(*lock->readers);
|
|
|
|
} else {
|
|
|
|
EBUG_ON(!lock->state.read_lock &&
|
|
|
|
!lock->state.intent_lock);
|
|
|
|
atomic64_add(l[type].lock_val, &lock->state.counter);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case SIX_LOCK_intent:
|
|
|
|
EBUG_ON(!lock->state.intent_lock);
|
|
|
|
lock->intent_lock_recurse++;
|
|
|
|
break;
|
|
|
|
case SIX_LOCK_write:
|
|
|
|
BUG();
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(six_lock_increment);
|
|
|
|
|
|
|
|
void six_lock_wakeup_all(struct six_lock *lock)
|
|
|
|
{
|
2022-08-26 23:22:24 +00:00
|
|
|
union six_lock_state state = lock->state;
|
2017-03-17 06:18:50 +00:00
|
|
|
struct six_lock_waiter *w;
|
|
|
|
|
2022-08-26 23:22:24 +00:00
|
|
|
six_lock_wakeup(lock, state, SIX_LOCK_read);
|
|
|
|
six_lock_wakeup(lock, state, SIX_LOCK_intent);
|
|
|
|
six_lock_wakeup(lock, state, SIX_LOCK_write);
|
|
|
|
|
2017-03-17 06:18:50 +00:00
|
|
|
raw_spin_lock(&lock->wait_lock);
|
2022-08-25 14:49:52 +00:00
|
|
|
list_for_each_entry(w, &lock->wait_list, list)
|
2017-03-17 06:18:50 +00:00
|
|
|
wake_up_process(w->task);
|
|
|
|
raw_spin_unlock(&lock->wait_lock);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(six_lock_wakeup_all);
|
|
|
|
|
|
|
|
void six_lock_pcpu_free(struct six_lock *lock)
|
|
|
|
{
|
|
|
|
BUG_ON(lock->readers && pcpu_read_count(lock));
|
|
|
|
BUG_ON(lock->state.read_lock);
|
|
|
|
|
|
|
|
free_percpu(lock->readers);
|
|
|
|
lock->readers = NULL;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(six_lock_pcpu_free);
|
|
|
|
|
|
|
|
void six_lock_pcpu_alloc(struct six_lock *lock)
|
|
|
|
{
|
|
|
|
#ifdef __KERNEL__
|
|
|
|
if (!lock->readers)
|
|
|
|
lock->readers = alloc_percpu(unsigned);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(six_lock_pcpu_alloc);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Returns lock held counts, for both read and intent
|
|
|
|
*/
|
|
|
|
struct six_lock_count six_lock_counts(struct six_lock *lock)
|
|
|
|
{
|
2022-08-22 03:08:53 +00:00
|
|
|
struct six_lock_count ret;
|
|
|
|
|
|
|
|
ret.n[SIX_LOCK_read] = 0;
|
|
|
|
ret.n[SIX_LOCK_intent] = lock->state.intent_lock + lock->intent_lock_recurse;
|
|
|
|
ret.n[SIX_LOCK_write] = lock->state.seq & 1;
|
2017-03-17 06:18:50 +00:00
|
|
|
|
|
|
|
if (!lock->readers)
|
2022-08-22 03:08:53 +00:00
|
|
|
ret.n[SIX_LOCK_read] += lock->state.read_lock;
|
2017-03-17 06:18:50 +00:00
|
|
|
else {
|
|
|
|
int cpu;
|
|
|
|
|
|
|
|
for_each_possible_cpu(cpu)
|
2022-08-22 03:08:53 +00:00
|
|
|
ret.n[SIX_LOCK_read] += *per_cpu_ptr(lock->readers, cpu);
|
2017-03-17 06:18:50 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(six_lock_counts);
|