2015-04-24 18:56:37 +00:00
|
|
|
#ifndef _GEN_PV_LOCK_SLOWPATH
|
|
|
|
#error "do not include this file"
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#include <linux/hash.h>
|
|
|
|
#include <linux/bootmem.h>
|
2015-07-12 01:19:19 +00:00
|
|
|
#include <linux/debug_locks.h>
|
2015-04-24 18:56:37 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Implement paravirt qspinlocks; the general idea is to halt the vcpus instead
|
|
|
|
* of spinning them.
|
|
|
|
*
|
|
|
|
* This relies on the architecture to provide two paravirt hypercalls:
|
|
|
|
*
|
|
|
|
* pv_wait(u8 *ptr, u8 val) -- suspends the vcpu if *ptr == val
|
|
|
|
* pv_kick(cpu) -- wakes a suspended vcpu
|
|
|
|
*
|
|
|
|
* Using these we implement __pv_queued_spin_lock_slowpath() and
|
|
|
|
* __pv_queued_spin_unlock() to replace native_queued_spin_lock_slowpath() and
|
|
|
|
* native_queued_spin_unlock().
|
|
|
|
*/
|
|
|
|
|
|
|
|
#define _Q_SLOW_VAL (3U << _Q_LOCKED_OFFSET)
|
|
|
|
|
2015-11-10 00:09:27 +00:00
|
|
|
/*
|
|
|
|
* Queue Node Adaptive Spinning
|
|
|
|
*
|
|
|
|
* A queue node vCPU will stop spinning if the vCPU in the previous node is
|
|
|
|
* not running. The one lock stealing attempt allowed at slowpath entry
|
|
|
|
* mitigates the slight slowdown for non-overcommitted guest with this
|
|
|
|
* aggressive wait-early mechanism.
|
|
|
|
*
|
|
|
|
* The status of the previous node will be checked at fixed interval
|
|
|
|
* controlled by PV_PREV_CHECK_MASK. This is to ensure that we won't
|
|
|
|
* pound on the cacheline of the previous node too heavily.
|
|
|
|
*/
|
|
|
|
#define PV_PREV_CHECK_MASK 0xff
|
|
|
|
|
2015-07-11 20:36:52 +00:00
|
|
|
/*
|
|
|
|
* Queue node uses: vcpu_running & vcpu_halted.
|
|
|
|
* Queue head uses: vcpu_running & vcpu_hashed.
|
|
|
|
*/
|
2015-04-24 18:56:37 +00:00
|
|
|
enum vcpu_state {
|
|
|
|
vcpu_running = 0,
|
2015-07-11 20:36:52 +00:00
|
|
|
vcpu_halted, /* Used only in pv_wait_node */
|
|
|
|
vcpu_hashed, /* = pv_hash'ed + vcpu_halted */
|
2015-04-24 18:56:37 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
struct pv_node {
|
|
|
|
struct mcs_spinlock mcs;
|
|
|
|
struct mcs_spinlock __res[3];
|
|
|
|
|
|
|
|
int cpu;
|
|
|
|
u8 state;
|
|
|
|
};
|
|
|
|
|
2015-12-10 20:17:46 +00:00
|
|
|
/*
|
|
|
|
* Include queued spinlock statistics code
|
|
|
|
*/
|
|
|
|
#include "qspinlock_stat.h"
|
|
|
|
|
2015-11-10 21:18:56 +00:00
|
|
|
/*
|
|
|
|
* By replacing the regular queued_spin_trylock() with the function below,
|
|
|
|
* it will be called once when a lock waiter enter the PV slowpath before
|
|
|
|
* being queued. By allowing one lock stealing attempt here when the pending
|
|
|
|
* bit is off, it helps to reduce the performance impact of lock waiter
|
|
|
|
* preemption without the drawback of lock starvation.
|
|
|
|
*/
|
|
|
|
#define queued_spin_trylock(l) pv_queued_spin_steal_lock(l)
|
|
|
|
static inline bool pv_queued_spin_steal_lock(struct qspinlock *lock)
|
|
|
|
{
|
|
|
|
struct __qspinlock *l = (void *)lock;
|
|
|
|
|
2016-07-14 12:26:11 +00:00
|
|
|
if (!(atomic_read(&lock->val) & _Q_LOCKED_PENDING_MASK) &&
|
|
|
|
(cmpxchg(&l->locked, 0, _Q_LOCKED_VAL) == 0)) {
|
|
|
|
qstat_inc(qstat_pv_lock_stealing, true);
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
return false;
|
2015-11-10 21:18:56 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The pending bit is used by the queue head vCPU to indicate that it
|
|
|
|
* is actively spinning on the lock and no lock stealing is allowed.
|
|
|
|
*/
|
|
|
|
#if _Q_PENDING_BITS == 8
|
|
|
|
static __always_inline void set_pending(struct qspinlock *lock)
|
|
|
|
{
|
|
|
|
struct __qspinlock *l = (void *)lock;
|
|
|
|
|
|
|
|
WRITE_ONCE(l->pending, 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
static __always_inline void clear_pending(struct qspinlock *lock)
|
|
|
|
{
|
|
|
|
struct __qspinlock *l = (void *)lock;
|
|
|
|
|
|
|
|
WRITE_ONCE(l->pending, 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The pending bit check in pv_queued_spin_steal_lock() isn't a memory
|
|
|
|
* barrier. Therefore, an atomic cmpxchg() is used to acquire the lock
|
|
|
|
* just to be sure that it will get it.
|
|
|
|
*/
|
|
|
|
static __always_inline int trylock_clear_pending(struct qspinlock *lock)
|
|
|
|
{
|
|
|
|
struct __qspinlock *l = (void *)lock;
|
|
|
|
|
|
|
|
return !READ_ONCE(l->locked) &&
|
|
|
|
(cmpxchg(&l->locked_pending, _Q_PENDING_VAL, _Q_LOCKED_VAL)
|
|
|
|
== _Q_PENDING_VAL);
|
|
|
|
}
|
|
|
|
#else /* _Q_PENDING_BITS == 8 */
|
|
|
|
static __always_inline void set_pending(struct qspinlock *lock)
|
|
|
|
{
|
2016-04-17 23:01:27 +00:00
|
|
|
atomic_or(_Q_PENDING_VAL, &lock->val);
|
2015-11-10 21:18:56 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static __always_inline void clear_pending(struct qspinlock *lock)
|
|
|
|
{
|
2016-04-17 23:01:27 +00:00
|
|
|
atomic_andnot(_Q_PENDING_VAL, &lock->val);
|
2015-11-10 21:18:56 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static __always_inline int trylock_clear_pending(struct qspinlock *lock)
|
|
|
|
{
|
|
|
|
int val = atomic_read(&lock->val);
|
|
|
|
|
|
|
|
for (;;) {
|
|
|
|
int old, new;
|
|
|
|
|
|
|
|
if (val & _Q_LOCKED_MASK)
|
|
|
|
break;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Try to clear pending bit & set locked bit
|
|
|
|
*/
|
|
|
|
old = val;
|
|
|
|
new = (val & ~_Q_PENDING_MASK) | _Q_LOCKED_VAL;
|
|
|
|
val = atomic_cmpxchg(&lock->val, old, new);
|
|
|
|
|
|
|
|
if (val == old)
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
#endif /* _Q_PENDING_BITS == 8 */
|
|
|
|
|
2015-04-24 18:56:37 +00:00
|
|
|
/*
|
|
|
|
* Lock and MCS node addresses hash table for fast lookup
|
|
|
|
*
|
|
|
|
* Hashing is done on a per-cacheline basis to minimize the need to access
|
|
|
|
* more than one cacheline.
|
|
|
|
*
|
|
|
|
* Dynamically allocate a hash table big enough to hold at least 4X the
|
|
|
|
* number of possible cpus in the system. Allocation is done on page
|
|
|
|
* granularity. So the minimum number of hash buckets should be at least
|
|
|
|
* 256 (64-bit) or 512 (32-bit) to fully utilize a 4k page.
|
|
|
|
*
|
|
|
|
* Since we should not be holding locks from NMI context (very rare indeed) the
|
|
|
|
* max load factor is 0.75, which is around the point where open addressing
|
|
|
|
* breaks down.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
struct pv_hash_entry {
|
|
|
|
struct qspinlock *lock;
|
|
|
|
struct pv_node *node;
|
|
|
|
};
|
|
|
|
|
|
|
|
#define PV_HE_PER_LINE (SMP_CACHE_BYTES / sizeof(struct pv_hash_entry))
|
|
|
|
#define PV_HE_MIN (PAGE_SIZE / sizeof(struct pv_hash_entry))
|
|
|
|
|
|
|
|
static struct pv_hash_entry *pv_lock_hash;
|
|
|
|
static unsigned int pv_lock_hash_bits __read_mostly;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Allocate memory for the PV qspinlock hash buckets
|
|
|
|
*
|
|
|
|
* This function should be called from the paravirt spinlock initialization
|
|
|
|
* routine.
|
|
|
|
*/
|
|
|
|
void __init __pv_init_lock_hash(void)
|
|
|
|
{
|
|
|
|
int pv_hash_size = ALIGN(4 * num_possible_cpus(), PV_HE_PER_LINE);
|
|
|
|
|
|
|
|
if (pv_hash_size < PV_HE_MIN)
|
|
|
|
pv_hash_size = PV_HE_MIN;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Allocate space from bootmem which should be page-size aligned
|
|
|
|
* and hence cacheline aligned.
|
|
|
|
*/
|
|
|
|
pv_lock_hash = alloc_large_system_hash("PV qspinlock",
|
|
|
|
sizeof(struct pv_hash_entry),
|
|
|
|
pv_hash_size, 0, HASH_EARLY,
|
|
|
|
&pv_lock_hash_bits, NULL,
|
|
|
|
pv_hash_size, pv_hash_size);
|
|
|
|
}
|
|
|
|
|
|
|
|
#define for_each_hash_entry(he, offset, hash) \
|
|
|
|
for (hash &= ~(PV_HE_PER_LINE - 1), he = &pv_lock_hash[hash], offset = 0; \
|
|
|
|
offset < (1 << pv_lock_hash_bits); \
|
|
|
|
offset++, he = &pv_lock_hash[(hash + offset) & ((1 << pv_lock_hash_bits) - 1)])
|
|
|
|
|
|
|
|
static struct qspinlock **pv_hash(struct qspinlock *lock, struct pv_node *node)
|
|
|
|
{
|
|
|
|
unsigned long offset, hash = hash_ptr(lock, pv_lock_hash_bits);
|
|
|
|
struct pv_hash_entry *he;
|
2015-11-10 00:09:25 +00:00
|
|
|
int hopcnt = 0;
|
2015-04-24 18:56:37 +00:00
|
|
|
|
|
|
|
for_each_hash_entry(he, offset, hash) {
|
2015-11-10 00:09:25 +00:00
|
|
|
hopcnt++;
|
2015-04-24 18:56:37 +00:00
|
|
|
if (!cmpxchg(&he->lock, NULL, lock)) {
|
|
|
|
WRITE_ONCE(he->node, node);
|
2015-11-10 00:09:25 +00:00
|
|
|
qstat_hop(hopcnt);
|
2015-04-24 18:56:37 +00:00
|
|
|
return &he->lock;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Hard assume there is a free entry for us.
|
|
|
|
*
|
|
|
|
* This is guaranteed by ensuring every blocked lock only ever consumes
|
|
|
|
* a single entry, and since we only have 4 nesting levels per CPU
|
|
|
|
* and allocated 4*nr_possible_cpus(), this must be so.
|
|
|
|
*
|
|
|
|
* The single entry is guaranteed by having the lock owner unhash
|
|
|
|
* before it releases.
|
|
|
|
*/
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct pv_node *pv_unhash(struct qspinlock *lock)
|
|
|
|
{
|
|
|
|
unsigned long offset, hash = hash_ptr(lock, pv_lock_hash_bits);
|
|
|
|
struct pv_hash_entry *he;
|
|
|
|
struct pv_node *node;
|
|
|
|
|
|
|
|
for_each_hash_entry(he, offset, hash) {
|
|
|
|
if (READ_ONCE(he->lock) == lock) {
|
|
|
|
node = READ_ONCE(he->node);
|
|
|
|
WRITE_ONCE(he->lock, NULL);
|
|
|
|
return node;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Hard assume we'll find an entry.
|
|
|
|
*
|
|
|
|
* This guarantees a limited lookup time and is itself guaranteed by
|
|
|
|
* having the lock owner do the unhash -- IFF the unlock sees the
|
|
|
|
* SLOW flag, there MUST be a hash entry.
|
|
|
|
*/
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
|
2015-11-10 00:09:27 +00:00
|
|
|
/*
|
|
|
|
* Return true if when it is time to check the previous node which is not
|
|
|
|
* in a running state.
|
|
|
|
*/
|
|
|
|
static inline bool
|
|
|
|
pv_wait_early(struct pv_node *prev, int loop)
|
|
|
|
{
|
|
|
|
if ((loop & PV_PREV_CHECK_MASK) != 0)
|
|
|
|
return false;
|
|
|
|
|
2017-01-10 07:56:46 +00:00
|
|
|
return READ_ONCE(prev->state) != vcpu_running || vcpu_is_preempted(prev->cpu);
|
2015-11-10 00:09:27 +00:00
|
|
|
}
|
|
|
|
|
2015-04-24 18:56:37 +00:00
|
|
|
/*
|
|
|
|
* Initialize the PV part of the mcs_spinlock node.
|
|
|
|
*/
|
|
|
|
static void pv_init_node(struct mcs_spinlock *node)
|
|
|
|
{
|
|
|
|
struct pv_node *pn = (struct pv_node *)node;
|
|
|
|
|
|
|
|
BUILD_BUG_ON(sizeof(struct pv_node) > 5*sizeof(struct mcs_spinlock));
|
|
|
|
|
|
|
|
pn->cpu = smp_processor_id();
|
|
|
|
pn->state = vcpu_running;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Wait for node->locked to become true, halt the vcpu after a short spin.
|
2015-07-11 20:36:52 +00:00
|
|
|
* pv_kick_node() is used to set _Q_SLOW_VAL and fill in hash table on its
|
|
|
|
* behalf.
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
2015-11-10 00:09:27 +00:00
|
|
|
static void pv_wait_node(struct mcs_spinlock *node, struct mcs_spinlock *prev)
|
2015-04-24 18:56:37 +00:00
|
|
|
{
|
|
|
|
struct pv_node *pn = (struct pv_node *)node;
|
2015-11-10 00:09:27 +00:00
|
|
|
struct pv_node *pp = (struct pv_node *)prev;
|
2015-04-24 18:56:37 +00:00
|
|
|
int loop;
|
2015-11-10 00:09:27 +00:00
|
|
|
bool wait_early;
|
2015-04-24 18:56:37 +00:00
|
|
|
|
2016-05-31 16:53:47 +00:00
|
|
|
for (;;) {
|
2015-11-10 00:09:27 +00:00
|
|
|
for (wait_early = false, loop = SPIN_THRESHOLD; loop; loop--) {
|
2015-04-24 18:56:37 +00:00
|
|
|
if (READ_ONCE(node->locked))
|
|
|
|
return;
|
2015-11-10 00:09:27 +00:00
|
|
|
if (pv_wait_early(pp, loop)) {
|
|
|
|
wait_early = true;
|
|
|
|
break;
|
|
|
|
}
|
2015-04-24 18:56:37 +00:00
|
|
|
cpu_relax();
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Order pn->state vs pn->locked thusly:
|
|
|
|
*
|
|
|
|
* [S] pn->state = vcpu_halted [S] next->locked = 1
|
|
|
|
* MB MB
|
2015-07-11 20:36:52 +00:00
|
|
|
* [L] pn->locked [RmW] pn->state = vcpu_hashed
|
2015-04-24 18:56:37 +00:00
|
|
|
*
|
2015-07-11 20:36:52 +00:00
|
|
|
* Matches the cmpxchg() from pv_kick_node().
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
2015-05-12 08:51:55 +00:00
|
|
|
smp_store_mb(pn->state, vcpu_halted);
|
2015-04-24 18:56:37 +00:00
|
|
|
|
2015-11-10 00:09:25 +00:00
|
|
|
if (!READ_ONCE(node->locked)) {
|
|
|
|
qstat_inc(qstat_pv_wait_node, true);
|
2015-11-10 00:09:27 +00:00
|
|
|
qstat_inc(qstat_pv_wait_early, wait_early);
|
2015-04-24 18:56:37 +00:00
|
|
|
pv_wait(&pn->state, vcpu_halted);
|
2015-11-10 00:09:25 +00:00
|
|
|
}
|
2015-04-24 18:56:37 +00:00
|
|
|
|
|
|
|
/*
|
2015-11-10 00:09:25 +00:00
|
|
|
* If pv_kick_node() changed us to vcpu_hashed, retain that
|
2015-11-10 21:18:56 +00:00
|
|
|
* value so that pv_wait_head_or_lock() knows to not also try
|
|
|
|
* to hash this lock.
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
2015-07-11 20:36:52 +00:00
|
|
|
cmpxchg(&pn->state, vcpu_halted, vcpu_running);
|
2015-04-24 18:56:37 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If the locked flag is still not set after wakeup, it is a
|
|
|
|
* spurious wakeup and the vCPU should wait again. However,
|
|
|
|
* there is a pretty high overhead for CPU halting and kicking.
|
|
|
|
* So it is better to spin for a while in the hope that the
|
|
|
|
* MCS lock will be released soon.
|
|
|
|
*/
|
2015-11-10 00:09:25 +00:00
|
|
|
qstat_inc(qstat_pv_spurious_wakeup, !READ_ONCE(node->locked));
|
2015-04-24 18:56:37 +00:00
|
|
|
}
|
2015-07-11 20:36:52 +00:00
|
|
|
|
2015-04-24 18:56:37 +00:00
|
|
|
/*
|
|
|
|
* By now our node->locked should be 1 and our caller will not actually
|
|
|
|
* spin-wait for it. We do however rely on our caller to do a
|
|
|
|
* load-acquire for us.
|
|
|
|
*/
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2015-07-11 20:36:52 +00:00
|
|
|
* Called after setting next->locked = 1 when we're the lock owner.
|
|
|
|
*
|
2015-11-10 21:18:56 +00:00
|
|
|
* Instead of waking the waiters stuck in pv_wait_node() advance their state
|
|
|
|
* such that they're waiting in pv_wait_head_or_lock(), this avoids a
|
|
|
|
* wake/sleep cycle.
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
2015-07-11 20:36:52 +00:00
|
|
|
static void pv_kick_node(struct qspinlock *lock, struct mcs_spinlock *node)
|
2015-04-24 18:56:37 +00:00
|
|
|
{
|
|
|
|
struct pv_node *pn = (struct pv_node *)node;
|
2015-07-11 20:36:52 +00:00
|
|
|
struct __qspinlock *l = (void *)lock;
|
2015-04-24 18:56:37 +00:00
|
|
|
|
|
|
|
/*
|
2015-07-11 20:36:52 +00:00
|
|
|
* If the vCPU is indeed halted, advance its state to match that of
|
|
|
|
* pv_wait_node(). If OTOH this fails, the vCPU was running and will
|
|
|
|
* observe its next->locked value and advance itself.
|
2015-04-24 18:56:37 +00:00
|
|
|
*
|
2015-07-11 20:36:52 +00:00
|
|
|
* Matches with smp_store_mb() and cmpxchg() in pv_wait_node()
|
|
|
|
*/
|
|
|
|
if (cmpxchg(&pn->state, vcpu_halted, vcpu_hashed) != vcpu_halted)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Put the lock into the hash table and set the _Q_SLOW_VAL.
|
2015-04-24 18:56:37 +00:00
|
|
|
*
|
2015-07-11 20:36:52 +00:00
|
|
|
* As this is the same vCPU that will check the _Q_SLOW_VAL value and
|
|
|
|
* the hash table later on at unlock time, no atomic instruction is
|
|
|
|
* needed.
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
2015-07-11 20:36:52 +00:00
|
|
|
WRITE_ONCE(l->locked, _Q_SLOW_VAL);
|
|
|
|
(void)pv_hash(lock, pn);
|
2015-04-24 18:56:37 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2015-11-10 21:18:56 +00:00
|
|
|
* Wait for l->locked to become clear and acquire the lock;
|
|
|
|
* halt the vcpu after a short spin.
|
2015-04-24 18:56:37 +00:00
|
|
|
* __pv_queued_spin_unlock() will wake us.
|
2015-11-10 21:18:56 +00:00
|
|
|
*
|
|
|
|
* The current value of the lock will be returned for additional processing.
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
2015-11-10 21:18:56 +00:00
|
|
|
static u32
|
|
|
|
pv_wait_head_or_lock(struct qspinlock *lock, struct mcs_spinlock *node)
|
2015-04-24 18:56:37 +00:00
|
|
|
{
|
|
|
|
struct pv_node *pn = (struct pv_node *)node;
|
|
|
|
struct __qspinlock *l = (void *)lock;
|
|
|
|
struct qspinlock **lp = NULL;
|
2015-11-10 00:09:25 +00:00
|
|
|
int waitcnt = 0;
|
2015-04-24 18:56:37 +00:00
|
|
|
int loop;
|
|
|
|
|
2015-07-11 20:36:52 +00:00
|
|
|
/*
|
|
|
|
* If pv_kick_node() already advanced our state, we don't need to
|
|
|
|
* insert ourselves into the hash table anymore.
|
|
|
|
*/
|
|
|
|
if (READ_ONCE(pn->state) == vcpu_hashed)
|
|
|
|
lp = (struct qspinlock **)1;
|
|
|
|
|
2015-12-10 20:17:45 +00:00
|
|
|
/*
|
|
|
|
* Tracking # of slowpath locking operations
|
|
|
|
*/
|
|
|
|
qstat_inc(qstat_pv_lock_slowpath, true);
|
|
|
|
|
2015-11-10 00:09:25 +00:00
|
|
|
for (;; waitcnt++) {
|
2015-11-10 00:09:27 +00:00
|
|
|
/*
|
|
|
|
* Set correct vCPU state to be used by queue node wait-early
|
|
|
|
* mechanism.
|
|
|
|
*/
|
|
|
|
WRITE_ONCE(pn->state, vcpu_running);
|
|
|
|
|
2015-11-10 21:18:56 +00:00
|
|
|
/*
|
|
|
|
* Set the pending bit in the active lock spinning loop to
|
|
|
|
* disable lock stealing before attempting to acquire the lock.
|
|
|
|
*/
|
|
|
|
set_pending(lock);
|
2015-04-24 18:56:37 +00:00
|
|
|
for (loop = SPIN_THRESHOLD; loop; loop--) {
|
2015-11-10 21:18:56 +00:00
|
|
|
if (trylock_clear_pending(lock))
|
|
|
|
goto gotlock;
|
2015-04-24 18:56:37 +00:00
|
|
|
cpu_relax();
|
|
|
|
}
|
2015-11-10 21:18:56 +00:00
|
|
|
clear_pending(lock);
|
|
|
|
|
2015-04-24 18:56:37 +00:00
|
|
|
|
|
|
|
if (!lp) { /* ONCE */
|
|
|
|
lp = pv_hash(lock, pn);
|
2015-07-11 20:36:52 +00:00
|
|
|
|
2015-04-24 18:56:37 +00:00
|
|
|
/*
|
2015-07-13 15:58:30 +00:00
|
|
|
* We must hash before setting _Q_SLOW_VAL, such that
|
|
|
|
* when we observe _Q_SLOW_VAL in __pv_queued_spin_unlock()
|
|
|
|
* we'll be sure to be able to observe our hash entry.
|
2015-04-24 18:56:37 +00:00
|
|
|
*
|
2015-07-13 15:58:30 +00:00
|
|
|
* [S] <hash> [Rmw] l->locked == _Q_SLOW_VAL
|
|
|
|
* MB RMB
|
|
|
|
* [RmW] l->locked = _Q_SLOW_VAL [L] <unhash>
|
2015-04-24 18:56:37 +00:00
|
|
|
*
|
2015-07-13 15:58:30 +00:00
|
|
|
* Matches the smp_rmb() in __pv_queued_spin_unlock().
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
2015-11-10 21:18:56 +00:00
|
|
|
if (xchg(&l->locked, _Q_SLOW_VAL) == 0) {
|
2015-04-24 18:56:37 +00:00
|
|
|
/*
|
2015-11-10 21:18:56 +00:00
|
|
|
* The lock was free and now we own the lock.
|
|
|
|
* Change the lock value back to _Q_LOCKED_VAL
|
|
|
|
* and unhash the table.
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
2015-11-10 21:18:56 +00:00
|
|
|
WRITE_ONCE(l->locked, _Q_LOCKED_VAL);
|
2015-04-24 18:56:37 +00:00
|
|
|
WRITE_ONCE(*lp, NULL);
|
2015-11-10 21:18:56 +00:00
|
|
|
goto gotlock;
|
2015-04-24 18:56:37 +00:00
|
|
|
}
|
|
|
|
}
|
2016-07-14 08:15:56 +00:00
|
|
|
WRITE_ONCE(pn->state, vcpu_hashed);
|
2015-11-10 00:09:25 +00:00
|
|
|
qstat_inc(qstat_pv_wait_head, true);
|
|
|
|
qstat_inc(qstat_pv_wait_again, waitcnt);
|
2015-04-24 18:56:37 +00:00
|
|
|
pv_wait(&l->locked, _Q_SLOW_VAL);
|
|
|
|
|
|
|
|
/*
|
2016-05-31 16:53:47 +00:00
|
|
|
* Because of lock stealing, the queue head vCPU may not be
|
|
|
|
* able to acquire the lock before it has to wait again.
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2015-11-10 21:18:56 +00:00
|
|
|
* The cmpxchg() or xchg() call before coming here provides the
|
|
|
|
* acquire semantics for locking. The dummy ORing of _Q_LOCKED_VAL
|
|
|
|
* here is to indicate to the compiler that the value will always
|
|
|
|
* be nozero to enable better code optimization.
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
2015-11-10 21:18:56 +00:00
|
|
|
gotlock:
|
|
|
|
return (u32)(atomic_read(&lock->val) | _Q_LOCKED_VAL);
|
2015-04-24 18:56:37 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2015-11-10 00:09:24 +00:00
|
|
|
* PV versions of the unlock fastpath and slowpath functions to be used
|
|
|
|
* instead of queued_spin_unlock().
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
2015-11-10 00:09:24 +00:00
|
|
|
__visible void
|
|
|
|
__pv_queued_spin_unlock_slowpath(struct qspinlock *lock, u8 locked)
|
2015-04-24 18:56:37 +00:00
|
|
|
{
|
|
|
|
struct __qspinlock *l = (void *)lock;
|
|
|
|
struct pv_node *node;
|
|
|
|
|
2015-07-21 10:13:43 +00:00
|
|
|
if (unlikely(locked != _Q_SLOW_VAL)) {
|
|
|
|
WARN(!debug_locks_silent,
|
|
|
|
"pvqspinlock: lock 0x%lx has corrupted value 0x%x!\n",
|
|
|
|
(unsigned long)lock, atomic_read(&lock->val));
|
2015-07-12 01:19:19 +00:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2015-07-13 15:58:30 +00:00
|
|
|
/*
|
|
|
|
* A failed cmpxchg doesn't provide any memory-ordering guarantees,
|
|
|
|
* so we need a barrier to order the read of the node data in
|
|
|
|
* pv_unhash *after* we've read the lock being _Q_SLOW_VAL.
|
|
|
|
*
|
2015-11-10 21:18:56 +00:00
|
|
|
* Matches the cmpxchg() in pv_wait_head_or_lock() setting _Q_SLOW_VAL.
|
2015-07-13 15:58:30 +00:00
|
|
|
*/
|
|
|
|
smp_rmb();
|
|
|
|
|
2015-04-24 18:56:37 +00:00
|
|
|
/*
|
|
|
|
* Since the above failed to release, this must be the SLOW path.
|
|
|
|
* Therefore start by looking up the blocked node and unhashing it.
|
|
|
|
*/
|
|
|
|
node = pv_unhash(lock);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Now that we have a reference to the (likely) blocked pv_node,
|
|
|
|
* release the lock.
|
|
|
|
*/
|
|
|
|
smp_store_release(&l->locked, 0);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* At this point the memory pointed at by lock can be freed/reused,
|
|
|
|
* however we can still use the pv_node to kick the CPU.
|
2015-07-11 20:36:52 +00:00
|
|
|
* The other vCPU may not really be halted, but kicking an active
|
|
|
|
* vCPU is harmless other than the additional latency in completing
|
|
|
|
* the unlock.
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
2015-11-10 00:09:25 +00:00
|
|
|
qstat_inc(qstat_pv_kick_unlock, true);
|
2015-09-11 18:37:34 +00:00
|
|
|
pv_kick(node->cpu);
|
2015-04-24 18:56:37 +00:00
|
|
|
}
|
2015-11-10 00:09:24 +00:00
|
|
|
|
2015-04-24 18:56:37 +00:00
|
|
|
/*
|
|
|
|
* Include the architecture specific callee-save thunk of the
|
|
|
|
* __pv_queued_spin_unlock(). This thunk is put together with
|
2015-11-10 00:09:24 +00:00
|
|
|
* __pv_queued_spin_unlock() to make the callee-save thunk and the real unlock
|
|
|
|
* function close to each other sharing consecutive instruction cachelines.
|
|
|
|
* Alternatively, architecture specific version of __pv_queued_spin_unlock()
|
|
|
|
* can be defined.
|
2015-04-24 18:56:37 +00:00
|
|
|
*/
|
|
|
|
#include <asm/qspinlock_paravirt.h>
|
|
|
|
|
2015-11-10 00:09:24 +00:00
|
|
|
#ifndef __pv_queued_spin_unlock
|
|
|
|
__visible void __pv_queued_spin_unlock(struct qspinlock *lock)
|
|
|
|
{
|
|
|
|
struct __qspinlock *l = (void *)lock;
|
|
|
|
u8 locked;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We must not unlock if SLOW, because in that case we must first
|
|
|
|
* unhash. Otherwise it would be possible to have multiple @lock
|
|
|
|
* entries, which would be BAD.
|
|
|
|
*/
|
2016-09-19 09:23:52 +00:00
|
|
|
locked = cmpxchg_release(&l->locked, _Q_LOCKED_VAL, 0);
|
2015-11-10 00:09:24 +00:00
|
|
|
if (likely(locked == _Q_LOCKED_VAL))
|
|
|
|
return;
|
|
|
|
|
|
|
|
__pv_queued_spin_unlock_slowpath(lock, locked);
|
|
|
|
}
|
|
|
|
#endif /* __pv_queued_spin_unlock */
|