I've got class:
ClassX.m
#property (assign) BOOL wasProcessed;
-(void) methodA { //<- this can be called many times in short period of time
dispatch_async(dispatch_get_main_queue(), ^{
[self methodB];
});
}
- (void) methodB {
if (!self.wasProcessed) {
self.wasProcessed = YES;
//... some code
}
}
Since dispatch_async is used so a few calls to methodB can be processed concurrently at the same time and following code needs to be atomic:
if (!self.wasProcessed) {
self.wasProcessed = YES; //e.g two calls can enter here before setting YES and it would be bad because I want to process it only one time
How can those 2 lines be made atomic (checking and setting variable)? I dont want to make atomic code that is after "self.wasProcessed = YES;" so moving whole if to #synchronize(self) won't be good solution. If there is anything wrong with my thinking please point it out as I'm not very experienced in those topics, Thank you.
Try #synchronized. While the enclosed code is being executed on a thread, it will block other threads from executing it.
- (void) methodB {
#synchronized(self) {
if (!self.wasProcessed) {
self.wasProcessed = YES;
//... some code
}
}
}
-(void) methodA {
dispatch_async(dispatch_get_main_queue(), ^{
[[NSOperationQueue mainQueue] addOperationWithBlock:^(){
[self methodB];
}];
});
}
Your's methodB will be only called in main thread, so it will be never performed simultaneously.
Related
I have three methods, two of them run at the same time. And the third method should be started only when the first and second method together complete their work. Either the first or second method, competitors, can finish their work first.
- (void)method1 {
//DO Long Work
isMethod1Complete = YES;
[self method3];
}
- (void)method2 {
//DO Long Work
isMethod2Complete = YES;
[self method3];
}
- (void)method3 {
if (isMethod1Complete && isMethod2Complete) {
//DO Work once
}
}
Method 3 should always be called once. But the problem is that there is a situation that method1 and method2 have finished working at the same time, and method3 is called twice. Tell me how to solve this problem in objective c for iOS?
Update:A concrete example, I have two services that call delegates when they finish their work.
- (void)method1Handler {
isMethod1Complete = YES;
[self method3];
}
- (void)method2Handler {
isMethod1Complete = YES;
[self method3];
}
How can this be solved without blocks?
For blocks, Rob's example is the best.
You say:
I have three methods, two of them run at the same time.
That means that they must be asynchronous or running on background queues (otherwise there's no way for them to run at the same time).
So, the idea is that you should give them both completion handlers (which will be called when they're done):
- (void)method1WithCompletion:(void(^ _Nonnull)(void))completion {
dispatch_async(dispatch_get_global_queue(QOS_CLASS_DEFAULT, 0), ^{
//DO Long Work asynchronously
completion();
});
}
- (void)method2WithCompletion:(void(^ _Nonnull)(void))completion {
dispatch_async(dispatch_get_global_queue(QOS_CLASS_DEFAULT, 0), ^{
//DO Long Work asynchronously
completion();
});
}
- (void)method3 {
// final task
}
In the above example, I added explicit dispatch_async calls to a background queue to ensure that the two long tasks run asynchronously. But if the code is already doing something asynchronous (e.g. a network request), then you will likely not need these dispatch_async calls, but just put the completion() call inside the completion handler provided by whatever API you are already using. But without more information about what method1 and method2 are doing, I cannot be more specific.
But, setting that aside, once your method1 and method2 have their own completion handlers, you can use dispatch_group_notify to identify what should be done when all of the dispatch_group_enter calls are balanced by their corresponding dispatch_group_leave calls:
dispatch_group_t group = dispatch_group_create();
dispatch_group_enter(group);
[self method1WithCompletion:^{
dispatch_group_leave(group);
}];
dispatch_group_enter(group);
[self method2WithCompletion:^{
dispatch_group_leave(group);
}];
dispatch_group_notify(group, dispatch_get_main_queue(), ^{
[self method3];
});
In subsequent comments, you mentioned that you are not using a completion block-based API, but rather a delegate-protocol-based API. You have a few options, for example:
You can use the same above closure pattern, but just save the completion handlers as block properties, e.g.:
For example, define block properties:
#property (nonatomic, copy, nullable) void (^completionOne)(void);
#property (nonatomic, copy, nullable) void (^completionTwo)(void);
Then, your method1 and method2 would save these blocks:
- (void)method1WithCompletion:(void(^ _Nonnull)(void))completion {
self.completionOne = completion;
// start your time consuming asynchronous process
}
// and your completion delegate method can then call the saved closure
// and then remove it
- (void)method1DidComplete {
self.completionOne();
self.completionOne = nil;
}
- (void)method2WithCompletion:(void(^ _Nonnull)(void))completion {
self.completionTwo = completion;
// start second asynchronous process
}
// same as above
- (void)method2DidComplete {
self.completionTwo();
self.completionTwo = nil;
}
The delegate-protocol completion API would then just call the saved block properties (and probably reset them to nil to free the memory associated with those blocks).
Then you can use the dispatch group notify process as shown in my original answer, above.
Alternatively, rather than using blocks, you can just use dispatch group by itself. For example, define dispatch group property:
#property (nonatomic, strong, nullable) dispatch_group_t group;
Then, you create your group and start your two tasks:
self.group = dispatch_group_create();
[self method1];
[self method2];
dispatch_group_notify(self.group, dispatch_get_main_queue(), ^{
[self method3];
});
And, the two methods then dispatch_group_enter when you start the tasks and dispatch_group_leave in their respective completion handler delegate methods:
- (void)method1 {
dispatch_group_enter(self.group);
// start first asynchronous process
}
// in your delegate completion method, you "leave" the group
- (void)method1DidComplete {
dispatch_group_leave(self.group);
}
- (void)method2 {
dispatch_group_enter(self.group);
// start second asynchronous process
}
- (void)method2DidComplete {
dispatch_group_leave(self.group);
}
- (void)method3 {
// you might as well remove the group now that you're done with it
self.group = nil;
// final task
NSLog(#"doing three");
}
Personally, I would generally lean towards the first option (that way, the dispatch group stuff is contained in a single method), but either approach works.
Why not dispatching the "call" to method3 in a serial queue?
dispatch_queue_t notSimQ;
notSimQ = dispatch_queue_create("notSimQ", NULL);
- (void)method1 {
//DO Long Work
isMethod1Complete = YES;
dispatch_async( notSimQ, // or sync
^{
[self method3];
});
}
- (void)method2 { … } // similiasr
- (void)method3 { … } // unchanged
The calls to method3 are never in competition.
- (void)method1 {
//DO Long Work
isMethod2Complete = YES;
dispatch_async(dispatch_get_main_queue(), ^(void){
[self method3];
}
}
- (void)method2 { … }
- (void)method3 { … }
- (void)viewDidLoad {
[super viewDidLoad];
[self method1];
[self method2];
}
- (void)method1 {
//DO Long Work
isMethod1Complete = YES;
}
- (void)method2 {
//DO Long Work
isMethod2Complete = YES;
dispatch_async(dispatch_get_main_queue(), ^(void){
[self method3];
}
}
- (void)method3 {
if (isMethod1Complete && isMethod2Complete) {
//DO Work once
}
}
How can I do things when two asynchronous requests are finished ?
perform the asynchronous initRequest and asynchronous refreshHeader, they finished after I perform the requestEnd method.
#implementation ViewController
- (void)viewDidLoad {
[super viewDidLoad];
[self initRequest];
[self requestEnd];
}
- (void)initRequest{
[GBCompanyRequest requestWithSuccess:^(EvaluationAverage *evaluationAverage) {
} failure:^(NSString *message) {
}];
[self refreshHeader];
}
- (void)refreshHeader{
[GBCompanyEvaluationRequest requestWithSuccess:^(EvaluationAverage *evaluationAverage) {
} failure:^(NSString *message) {
}];
}
- (void)requestEnd{
NSLog(#"How can I do things when two asynchronous requests are finished");
}
#end
My suggestion is use dispatch_group_t
dispatch_group_t : Generally use for call/execute group of synchronously or asynchronously method/block and it will notify you after all the method/block execution are done.
For more information about how to use, you can read this Q/A.
Waiting until two async blocks are executed before starting another block
I'am now using dispatch_group to manage multi-threads in my project. something like this:
- (void)functionA{
self.taskGroup = dispatch_group_create();
// Call functionB more than one time
[self functionB];
...
[self functionB];
dispatch_group_notify(..., ^{
// Do something if all [self functionB] complete
});
- (void)functionB{
dispatch_group_enter(self.taskGroup);
if (condition) {
dispatch_group_leave(self.taskGroup);
return;
}
[self doSomethingInBackground:^{
NSLog(#"completed!");
dispatch_group_leave(self.taskGroup);
}];
}
My question is how to automatically call dispatch_group_leave when functionB is complete, so that I don't need to call it before every return statement in functionB or any background task called by functionB.
As #Sk0prion mentioned in comments, I can simply involve macro here.
#define m_return dispatch_group_leave(self.taskGroup);return
I've gotten in a few cases when something receives multiple refresh calls in quick succession, eg:
- ViewController receives multiple KVO notifications.
- Datamanger class that is called from setters to refresh when multiple settings change.
Ideally I would like to execute only the last refresh call from a series (drop all the intermediate ones).
Right now I'm using an isRefreshing property and a needRefresh to block excessive refreshes, eg:
- (id)init {
...
[self observeValueForKeyPath:#"isRefreshing" ....];
}
- (void)setParameter:(NSInteger)parameter {
....
[self refresh];
}
/* and many more kinds of updates require a refresh */
- (void)setAnotherProperty:(NSArray*)array {
....
[self refresh];
}
- (void)refresh {
if (self.isRefreshing) {
self.needRefresh = YES;
return;
}
self.isRefreshing = YES;
...
self.isRefreshing = NO;
}
- observeValueForKeyPath..... {
if (!self.isRefreshing && self.needsRefresh) {
self.needsRefresh = NO;
[self refresh];
}
}
Is there a better solution for this kind of problem?
You can create a NSOperationQueue with concurrency set to one and only submit a new operation to it when its operation count is zero. (Or use cancellation logic to remove pending jobs so that only one new one is queued if there's a job in progress.)
What you're doing is reasonable for a single-threaded system but would become fairly complicated for multiple threads.
Looks like you should delay refreshing for a while.
You can use different techniques to do so. It is enough only one flag.
For example you may use async block to make a delay for a one main run-loop cycle
- (void)setParameter:(NSInteger)parameter {
....
[self requestRefrhesh];
}
- (void)setAnotherProperty:(NSArray*)array {
....
[self requestRefrhesh];
}
...
-(void) requestRefrhesh {
if (self.refreshRequested) {
return;
} else {
self.refreshRequested = YES;
dispatch_async(dispatch_get_main_queue(), ^(void){
//Run in main UI thread
//make your UI changes here
self.refreshRequested = NO;
});
}
}
I have been thinking about a problem that seemingly would be simple to implement, yet an efficient and threadsafe solution is stymying me. What I want to do is create some sort of worker object. Several callers may ask it to work from different threads. A requirement is that requests must not queue up. In other words if somebody asks the worker to do work but sees it is already doing work, it should just return early.
A simple first pass is this:
#interface Worker : NSObject
#property (nonatomic, assign, getter = isWorking) BOOL working;
- (void)doWork;
#end
#implementation Worker
{
dispatch_queue_t _workerQueue; //... a private serial queue
}
- (void)doWork
{
if ( self.isWorking )
{
return;
}
self.working = YES;
dispatch_async(_workerQueue, ^{
// Do time consuming work here ... Done!
self.working = NO;
});
}
#end
The problem with this is that the isWorking property is not threadsafe. Marking it atomic won't help here, as accesses to it need to be synchronized across a few statements.
To make it threadsafe I would need to protect the isWorking with a lock:
#interface Worker : NSObject
#property (nonatomic, assign, getter = isWorking) BOOL working;
- (void)doWork;
#end
#implementation Worker
{
dispatch_queue_t _workerQueue; //... a private serial queue
NSLock *_lock; // assume this is created
}
- (void)doWork
{
[_lock lock];
if ( self.isWorking )
{
[_lock unlock];
return;
}
self.working = YES;
[_lock unlock];
dispatch_async(_workerQueue, ^{
// Do time consuming work here ... Done!
[_lock lock];
self.working = NO;
[_lock unlock];
});
}
#end
While I do believe this would be threadsafe, I think it's pretty crummy to have to take and give up a lock (an expensive operation) so frequently.
So, is there a more elegant solution?
dispatch_semaphore is the idiomatic way to limit access to a finite resource, if you're already using GCD.
// Add an ivar:
dispatch_semaphore_t _semaphore;
// To initialize:
_semaphore = dispatch_semaphore_create(1);
// To "do work" from any thread:
- (void)doWork
{
if (dispatch_semaphore_wait(_semaphore, DISPATCH_TIME_NOW) == 0) {
// We got the semaphore without waiting, so we're first in line.
dispatch_async(_workerQueue, ^{
// do time consuming work here, then when done:
dispatch_semaphore_signal(_semaphore);
});
} else {
// We would have had to wait for the semaphore, so somebody must have
// been doing work already, and we should do nothing.
}
}
Here's a blog post explaining in more detail.
You may be able to use an atomic test-and-set operation here. GCC provides __atomic_test_and_set for this purpose. Here's how you might use it in C (untested):
static volatile bool working = FALSE;
if(__atomic_test_and_set(&working, __ATOMIC_ACQUIRE)) {
// Already was working.
}else{
// Do work, possibly in another thread.
// When done:
__atomic_clear(&working, __ATOMIC_RELEASE);
}
Easy, huh?
For making a property thread-safe you could simply use #synchronize.