sync

package
v1.2.78 Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Jul 23, 2023 License: MIT Imports: 15 Imported by: 1

Documentation

Index

Examples

Constants

View Source 
const DefaultMaxIdleResourcesPerBucket = 2

DefaultMaxIdleResourcesPerBucket is the default value of LruPool's MaxIdleResourcesPerBucket.

Variables

View Source 
var DefaultLruPool = &LruPool{
	MaxIdleResources:    100,
	IdleResourceTimeout: 90 * time.Second,
}

DefaultLruPool is new resources as needed and caches them for reuse by subsequent calls.

View Source 
var ErrThreadClosed = errors.New("sync: Thread closed")

ErrThreadClosed is returned by the Thread's Do methods after a call to `Shutdown`.

View Source 
var ErrUntilClosed = errors.New("sync: Until closed")

ErrUntilClosed is returned by the Until's Do method after a call to Close.

Functions

func WithLock 

func WithLock(lk sync.Locker, fn func())

WithLock runs while holding lk.

Types

type ConditionVariable 

type ConditionVariable struct {
	// contains filtered or unexported fields
}

ConditionVariable is an object able to block the calling thread until notified to resume. see http://www.cplusplus.com/reference/condition_variable/condition_variable/

func (*ConditionVariable) Broadcast 

func (c *ConditionVariable) Broadcast()

Broadcast wakes all goroutines waiting on c.

It is allowed but not required for the caller to hold c.L during the call.

func (*ConditionVariable) Signal 

func (c *ConditionVariable) Signal()

Signal wakes one goroutine waiting on c, if there is any.

It is allowed but not required for the caller to hold c.L during the call.

func (*ConditionVariable) Wait 

func (c *ConditionVariable) Wait(lck sync.Locker)

Wait atomically unlocks c.L and suspends execution of the calling goroutine. After later resuming execution, Wait locks c.L before returning. Unlike in other systems, Wait cannot return unless awoken by Broadcast or Signal.

Because c.L is not locked when Wait first resumes, the caller typically cannot assume that the condition is true when Wait returns. Instead, the caller should Wait in a loop: 

c.L.Lock()
for !condition() {
    c.Wait()
}
... make use of condition ...
c.L.Unlock()

Wait wait until notified

func (*ConditionVariable) WaitContext 

func (c *ConditionVariable) WaitContext(ctx context.Context, lck sync.Locker) error

WaitContext wait until notified or time point The execution of the current thread (which shall have locked lck's mutex) is blocked either until notified or until ctx done, whichever happens first.

func (*ConditionVariable) WaitFor 

func (c *ConditionVariable) WaitFor(lck sync.Locker, timeout time.Duration) error

WaitFor The execution of the current goroutine (which shall have locked lck's mutex) is blocked during rel_time, or until notified (if the latter happens first). WaitFor wait for timeout or until notified It behaves as if implemented as: return wait_until (lck, chrono::steady_clock::now() + rel_time);

func (*ConditionVariable) WaitForPred 

func (c *ConditionVariable) WaitForPred(lck sync.Locker, timeout time.Duration, pred func() bool) bool

WaitForPred wait for timeout or until notified If pred is nil, do as pred returns false always, If pred is specified, the function only blocks if pred returns false, and notifications can only unblock the thread when it becomes true (which is especially useful to check against spurious wake-up calls). It behaves as if implemented as: return wait_until (lck, chrono::steady_clock::now() + rel_time, std::move(pred));

func (*ConditionVariable) WaitPred 

func (c *ConditionVariable) WaitPred(lck sync.Locker, pred func() bool)

WaitPred wait until notified If pred is specified (2), the function only blocks if pred returns false, and notifications can only unblock the thread when it becomes true (which is specially useful to check against spurious wake-up calls). This version (2) behaves as if implemented as: while (!pred()) wait(lck);

func (*ConditionVariable) WaitPredContext 

func (c *ConditionVariable) WaitPredContext(ctx context.Context, lck sync.Locker, pred func() bool) bool

WaitPredContext wait until notified or ctx done The execution of the current thread (which shall have locked lck's mutex) is blocked either until notified or until ctx, whichever happens first. If pred is nil, do as pred returns false always, If pred is specified, the function only blocks if pred returns false, and notifications can only unblock the thread when it becomes true (which is especially useful to check against spurious wake-up calls). It behaves as if implemented as: while (!pred()) 

if ( wait_until(ctx,lck) == cv_status::timeout)
  return pred();

return true;

func (*ConditionVariable) WaitUntil 

func (c *ConditionVariable) WaitUntil(lck sync.Locker, d time.Time) error

WaitUntil wait until notified or time point The execution of the current thread (which shall have locked lck's mutex) is blocked either until notified or until abs_time, whichever happens first.

func (*ConditionVariable) WaitUntilPred 

func (c *ConditionVariable) WaitUntilPred(lck sync.Locker, d time.Time, pred func() bool) bool

WaitUntilPred wait until notified or ctx done The execution of the current thread (which shall have locked lck's mutex) is blocked either until notified or until ctx, whichever happens first. If pred is nil, do as pred returns false always, If pred is specified, the function only blocks if pred returns false, and notifications can only unblock the thread when it becomes true (which is especially useful to check against spurious wake-up calls). It behaves as if implemented as: while (!pred()) 

if ( wait_until(lck,abs_time) == cv_status::timeout)
  return pred();

return true;

type EmptyThreadDoOption added in v1.2.33 

type EmptyThreadDoOption struct{}

EmptyThreadDoOption does not alter the configuration. It can be embedded in another structure to build custom options.

This API is EXPERIMENTAL.

type Event 

type Event struct {
	// contains filtered or unexported fields
}

Event represents a one-time event that may occur in the future. we use a fast path to avoid the more-expensive "select" below after the Event has returned once.

func NewEvent 

func NewEvent() *Event

NewEvent returns a new, ready-to-use Event.

func (*Event) Done 

func (e *Event) Done() <-chan struct{}

Done returns a channel that will be closed when Fire is called.

func (*Event) Fire 

func (e *Event) Fire() bool

Fire causes e to complete. It is safe to call multiple times, and concurrently. It returns true iff this call to Fire caused the signaling channel returned by Done to close.

func (*Event) HasFired 

func (e *Event) HasFired() bool

HasFired returns true if Fire has been called. we use a fast path to avoid the more-expensive "select" below after the Event has returned once.

type LruPool 

type LruPool struct {
	// New optionally specifies a function to generate
	// a value when Get would otherwise return nil.
	// It may not be changed concurrently with calls to Get.
	New func(ctx context.Context, req interface{}) (resp interface{}, err error)

	// DisableKeepAlives, if true, disables keep-alives and
	// will only use the resource to the server for a single request.
	DisableKeepAlives bool

	// MaxIdleResources controls the maximum number of idle (keep-alive)
	// resources across all buckets. Zero means no limit.
	MaxIdleResources int

	// MaxIdleResourcesPerBucket, if non-zero, controls the maximum idle
	// (keep-alive) resources to keep per-bucket. If zero,
	// DefaultMaxIdleResourcesPerBucket is used.
	MaxIdleResourcesPerBucket int

	// MaxResourcesPerBucket optionally limits the total number of
	// resources per bucket, including resources in the newResource,
	// active, and idle states. On limit violation, news will block.
	//
	// Zero means no limit.
	MaxResourcesPerBucket int

	// IdleResourceTimeout is the maximum amount of time an idle
	// (keep-alive) resource will remain idle before closing
	// itself.
	// Zero means no limit.
	IdleResourceTimeout time.Duration
	// contains filtered or unexported fields
}

LruPool is an implementation of sync.Pool with LRU.

By default, LruPool caches resources for future re-use. This may leave many open resources when accessing many buckets. This behavior can be managed using LruPool's CloseIdleResources method and the MaxIdleResourcesPerBucket and DisableKeepAlives fields.

LruPools should be reused instead of created as needed. LruPools are safe for concurrent use by multiple goroutines.

A LruPool is a low-level primitive for making resources.

func (*LruPool) CloseIdleResources 

func (t *LruPool) CloseIdleResources()

CloseIdleResources closes any connections which were previously connected from previous requests but are now sitting idle in a "keep-alive" state. It does not interrupt any connections currently in use.

func (*LruPool) Get 

func (t *LruPool) Get(ctx context.Context, req interface{}) (v interface{}, put context.CancelFunc)

Get creates a new PersistResource to the target as specified in the key. If this doesn't return an error, the PersistResource is ready to write requests to.

func (*LruPool) GetByKey 

func (t *LruPool) GetByKey(ctx context.Context, key interface{}, req interface{}) (v interface{}, put context.CancelFunc)

GetByKey creates a new PersistResource to the target as specified in the key. If this doesn't return an error, the PersistResource is ready to write requests to.

func (*LruPool) GetByKeyOrError 

func (t *LruPool) GetByKeyOrError(ctx context.Context, key interface{}, req interface{}) (pc *PersistResource, err error)

GetByKeyOrError creates a new PersistResource to the target as specified in the key. If this doesn't return an error, the PersistResource is ready to write requests to.

func (*LruPool) GetOrError 

func (t *LruPool) GetOrError(ctx context.Context, req interface{}) (v interface{}, put context.CancelFunc, err error)

GetOrError creates a new PersistResource to the target as specified in the key. If this doesn't return an error, the PersistResource is ready to write requests to.

func (*LruPool) Put 

func (t *LruPool) Put(presource *PersistResource)

func (*LruPool) RemoveIdleResource 

func (t *LruPool) RemoveIdleResource(presource *PersistResource) bool

RemoveIdleResource marks presource as dead.

type PersistResource 

type PersistResource struct {
	// contains filtered or unexported fields

} // isBroken reports whether this resource is in a known broken state.

PersistResource wraps a resource, usually a persistent one (but may be used for non-keep-alive requests as well)

func (*PersistResource) Get 

func (pc *PersistResource) Get() interface{}

func (*PersistResource) Put 

func (pc *PersistResource) Put()

type Subject 

type Subject struct {
	// contains filtered or unexported fields
}

Subject implements a condition variable like with channel, a rendezvous point for goroutines waiting for or announcing the occurrence of an event.

The caller typically cannot assume that the condition is true when Subscribe chan returns. Instead, the caller should Wait in a loop: 

time.After(timeout, c.PublishBroadcast()) // for timeout or periodic event
c.PublishBroadcast() // for async notify event directly
eventC, cancel := c.Subscribe()
for !condition() {
    select{
    case event, closed := <- eventC:
        ... make use of event ...
    }
}
... make use of condition ...

func (*Subject) PublishBroadcast 

func (s *Subject) PublishBroadcast(ctx context.Context, event interface{}) error

PublishBroadcast wakes all listeners waiting on c. PublishBroadcast blocks until event is received or dropped. event will be dropped if ctx is Done before event is received.

func (*Subject) PublishSignal 

func (s *Subject) PublishSignal(ctx context.Context, event interface{}) error

PublishSignal wakes one listener waiting on c, if there is any. PublishSignal blocks until event is received or dropped.

func (*Subject) Subscribe 

func (s *Subject) Subscribe() (<-chan interface{}, context.CancelFunc)

Subscribe returns a channel that's closed when awoken by PublishSignal or PublishBroadcast. never be canceled. Successive calls to Subscribe return different values. The close of the Subscribe channel may happen asynchronously, after the cancel function returns.

type Thread 

type Thread struct {
	GoRoutine bool // Use thread as goroutine, that is without runtime.LockOSThread()

	// The Leak is published as a variable directly.
	GoroutineLeak *expvar_.Leak // represents whether goroutine is leaked, take effects if not nil
	OSThreadLeak  *expvar_.Leak // represents whether runtime.LockOSThread is leaked, take effects  if not nil
	HandlerLeak   *expvar_.Leak // represents whether handler in Do is blocked is leaked, take effects  if not nil
	// contains filtered or unexported fields
}

Thread should be used for such as calling OS services or non-Go library functions that depend on per-thread state, as runtime.LockOSThread().

func (*Thread) Do 

func (t *Thread) Do(ctx context.Context, f func(), opts ...ThreadDoOption) error

Do will call the function f in the same thread or escape thread. f is enqueued only if ctx is not canceled and Thread is not Shutdown and Not escape

func (*Thread) Shutdown 

func (t *Thread) Shutdown()

func (*Thread) WatchStats added in v1.2.50 

func (t *Thread) WatchStats()

WatchStats bind Leak var to "sync.Thread"

type ThreadDoOption added in v1.2.33 

type ThreadDoOption interface {
	// contains filtered or unexported methods
}

A ThreadDoOption sets options.

func WiththreadDoEscapeThread added in v1.2.33 

func WiththreadDoEscapeThread(v bool) ThreadDoOption

WiththreadDoEscapeThread sets EscapeThread in threadDo. call the function f in the same thread or escape thread

type ThreadDoOptionFunc added in v1.2.33 

type ThreadDoOptionFunc func(*threadDo)

ThreadDoOptionFunc wraps a function that modifies threadDo into an implementation of the ThreadDoOption interface.

type TimeoutCond 

type TimeoutCond struct {

	// L is held while observing or changing the condition
	L sync.Locker
	// contains filtered or unexported fields
}

func NewTimeoutCond 

func NewTimeoutCond(l sync.Locker) *TimeoutCond

NewTimeoutCond returns a new TimeoutCond with Locker l.

func (*TimeoutCond) Broadcast 

func (c *TimeoutCond) Broadcast()

Broadcast wakes all goroutines waiting on c.

It is allowed but not required for the caller to hold c.L during the call.

func (*TimeoutCond) Signal 

func (c *TimeoutCond) Signal()

Signal wakes one goroutine waiting on c, if there is any.

It is allowed but not required for the caller to hold c.L during the call.

func (*TimeoutCond) Wait 

func (c *TimeoutCond) Wait()

Wait atomically unlocks c.L and suspends execution of the calling goroutine. After later resuming execution, Wait locks c.L before returning. Unlike in other systems, Wait cannot return unless awoken by Broadcast or Signal.

Because c.L is not locked when Wait first resumes, the caller typically cannot assume that the condition is true when Wait returns. Instead, the caller should Wait in a loop: 

c.L.Lock()
for !condition() {
    c.Wait()
}
... make use of condition ...
c.L.Unlock()

Wait wait until notified

func (*TimeoutCond) WaitContext 

func (c *TimeoutCond) WaitContext(ctx context.Context) error

WaitContext wait until notified or time point The execution of the current thread (which shall have locked lck's mutex) is blocked either until notified or until ctx done, whichever happens first.

func (*TimeoutCond) WaitFor 

func (c *TimeoutCond) WaitFor(timeout time.Duration) error

WaitFor The execution of the current goroutine (which shall have locked lck's mutex) is blocked during rel_time, or until notified (if the latter happens first). WaitFor wait for timeout or until notified It behaves as if implemented as: return wait_until (lck, chrono::steady_clock::now() + rel_time);

func (*TimeoutCond) WaitForPred 

func (c *TimeoutCond) WaitForPred(timeout time.Duration, pred func() bool) bool

WaitForPred wait for timeout or until notified If pred is nil, do as pred returns false always, If pred is specified, the function only blocks if pred returns false, and notifications can only unblock the thread when it becomes true (which is especially useful to check against spurious wake-up calls). It behaves as if implemented as: return wait_until (lck, chrono::steady_clock::now() + rel_time, std::move(pred));

func (*TimeoutCond) WaitPred 

func (c *TimeoutCond) WaitPred(pred func() bool)

WaitPred wait until notified If pred is specified (2), the function only blocks if pred returns false, and notifications can only unblock the thread when it becomes true (which is specially useful to check against spurious wake-up calls). This version (2) behaves as if implemented as: while (!pred()) wait(lck);

func (*TimeoutCond) WaitPredContext 

func (c *TimeoutCond) WaitPredContext(ctx context.Context, pred func() bool) bool

WaitPredContext wait until notified or ctx done The execution of the current thread (which shall have locked lck's mutex) is blocked either until notified or until ctx, whichever happens first. If pred is nil, do as pred returns false always, If pred is specified, the function only blocks if pred returns false, and notifications can only unblock the thread when it becomes true (which is especially useful to check against spurious wake-up calls). It behaves as if implemented as: while (!pred()) 

if ( wait_until(ctx,lck) == cv_status::timeout)
  return pred();

return true;

func (*TimeoutCond) WaitUntil 

func (c *TimeoutCond) WaitUntil(d time.Time) error

WaitUntil wait until notified or time point The execution of the current thread (which shall have locked lck's mutex) is blocked either until notified or until abs_time, whichever happens first.

func (*TimeoutCond) WaitUntilPred 

func (c *TimeoutCond) WaitUntilPred(d time.Time, pred func() bool) bool

WaitUntilPred wait until notified or time point The execution of the current thread (which shall have locked c.L's mutex) is blocked either until notified or until abs_time, whichever happens first. If pred is nil, do as pred returns false always, If pred is specified, the function only blocks if pred returns false, and notifications can only unblock the thread when it becomes true (which is especially useful to check against spurious wake-up calls). It behaves as if implemented as: while (!pred()) 

if ( wait_until(lck,abs_time) == cv_status::timeout)
  return pred();

return true;

type Until 

type Until struct {
	// contains filtered or unexported fields
}

Until represents a class of work and forms a namespace in which units of work can be executed with duplicate suppression. It blocks on certain Do actions and unblock when Retry is called.

func (*Until) Close 

func (u *Until) Close()

func (*Until) Do 

func (u *Until) Do(ctx context.Context, fn func() (interface{}, error)) (val interface{}, err error)

Do executes and returns the results of the given function. It may block in the following cases: - the current fn is nil - the err returned by the current fn is not nil When one of these situations happens, Do blocks until the Retry is called.

func (*Until) Retry 

func (u *Until) Retry()

Retry unblocks all blocked pick.

type Walk 

type Walk struct {
	Burst int // Burst will be set to 1 if less than 1
	// contains filtered or unexported fields
}
Example 
package main

import (
	"context"
	"fmt"

	"github.com/searKing/golang/go/sync"
)

func main() {

	// chan WalkInfo
	walkChan := make(chan interface{}, 0)

	p := sync.Walk{
		Burst: 1,
	}
	defer p.Wait()

	p.Walk(context.Background(), walkChan, func(name interface{}) error {
		fmt.Printf("%s\n", name)
		return nil
	})

	for i := 0; i < 5; i++ {
		walkChan <- fmt.Sprintf("%d", i)
	}
	close(walkChan)
}

Output:

0
1
2
3
4

func (*Walk) Error 

func (p *Walk) Error() error

func (*Walk) TrySetError 

func (p *Walk) TrySetError(err error)

func (*Walk) Wait 

func (p *Walk) Wait() error

Wait blocks until the WaitGroup counter is zero.

func (*Walk) Walk 

func (p *Walk) Walk(ctx context.Context, taskChan <-chan interface{}, procFn WalkFunc) (doneC <-chan struct{})

Walk will consume all tasks parallel and block until ctx.Done() or taskChan is closed. Walk returns a channel that's closed when work done on behalf of this walk should be canceled. Done may return nil if this walk can never be canceled. Successive calls to Done return the same value. The close of the Done channel may happen asynchronously, after the cancel function returns.

type WalkFunc 

type WalkFunc func(task interface{}) error

WalkFunc is the type of the function called for each task processed by Walk. The path argument contains the argument to Walk as a task.

In the case of an error, the info argument will be nil. If an error is returned, processing stops.

Source Files

View all Source files

Directories

Path Synopsis
Package leaderelection implements leader election of a set of endpoints.
 Package leaderelection implements leader election of a set of endpoints.

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.