Does ARC ever inject retain and release calls that you generally wouldn't see in a non-ARC environment?
For example, explicitly releasing an object from a getter:
- (NSArray *)dummyArray {
return [[NSArray alloc]init];
}
- (void)useDummyArray {
NSArray * arr = [self dummyArray];
//do something with arr
[arr release]; //unconventional injection of release.
}
Would ARC ever generate a release statement like the code above or would it autorelease the array returned by [self dummyArray];
The beauty of ARC is that you don't know, or need to know. However, you can give hints to the ARC static analyzer:
-(NSArray *) dummyArray NS_RETURNS_RETAINED { // this tells ARC that this function returns a retained value that should be released by the callee
return [[NSArray alloc] init];
}
-(NSArray *) otherDummyArray NS_RETURNS_NOT_RETAINED { // this tells ARC that the function returns a non-retained (autoreleased) value, which should NOT be released by the callee.
return [[NSArray alloc] init];
}
However, NS_RETURNS_NOT_RETAINED is the default, as long as your function name doesn't begin with init, in which NS_RETURNS_RETAINED becomes default.
So, in your specific scenario, it will almost always return an autorelease'd value. One major reason for this is support for interpolation with non-ARC code, which could result in leaks.
Related
I have a non-ARC project that is using an ARC library. I am confused if I should release the object returned by the library method or not. Here is some example:
- (void)test{
LibObject* obj1 = [[LibObject alloc] init];
LibObject* obj2 = [obj1 otherObj];
[obj1 release]; //should I release it?
[obj2 release]; //should I release it?
}
Best to my knowledge, if the objects are in the autorelease pool, I should leave it alone. Otherwise, I should release it.
However, the ARC document says that
When returning from such a function or method, ARC retains the value at the point of evaluation of the return statement, then leaves all local scopes, and then balances out the retain while ensuring that the value lives across the call boundary. In the worst case, this may involve an autorelease, but callers must not assume that the value is actually in the autorelease pool.
Does the document imply that I should always release the object since I cannot assume the object is autoreleased?
See Memory Management Policy in the "Advanced Memory Management Programming Guide":
You own any object you create
You create an object using a method
whose name begins with “alloc”, “new”, “copy”, or “mutableCopy” (for
example, alloc, newObject, or mutableCopy).
You must not relinquish ownership of an object you do not own
These rules are consistent with methods compiled with ARC.
In the ARC documentation, the behaviour of methods in the first category is described in "3.2.2. Retained return values":
When returning from such a function or method, ARC retains the value
at the point of evaluation of the return statement, ...
which means that the caller must release the object.
Methods in the second category are described in "3.2.3. Unretained return values":
When returning from such a function or method, ARC retains the value
at the point of evaluation of the return statement, then leaves all
local scopes, and then balances out the retain ...
which means the the caller must not release the object.
So in your example:
LibObject* obj1 = [[LibObject alloc] init];
LibObject* obj2 = [obj1 otherObj];
[obj1 release]; //should I release it? --> YES
[obj2 release]; //should I release it? --> NO
You own obj1, because it was created with alloc, so you have to release obj1.
You don't own obj2, therefore you must not release it.
Note: If you run the Xcode static analyzer on your code, it shows exactly where these rules are violated.
When you get an object from ARC enabled class to your non-ARC enabled class you have the responsibility to manage the memory.
ARC just simply put retain, release in your ARC enabled code during compile time. It won't manage the memory on other classes or objects that is in non-ARC mode.
You should release such objects after your need.
My understanding is that if you use the compiler flag -fno-objc-arc for this file you need to take care of retain/release yourself. So you should call [obj1 release]. If you weren't supposed to call it yourself, the compiler would warn you accordingly.
With my understanding is that in your non-ARC code, you just do all the retain / release with ARC-enabled library as always (just like interacting with other non-ARC code).
I have created a simple app to prove this behaviour.
Testing Code - compiled without ARC
- (void)testARC
{
ARCLib *al = [[ARCLib alloc] init];
NSLog(#"%p retainCount = %u", al, [al retainCount]);
[al release];
ARCLib *al2 = [ARCLib arcLib];
NSLog(#"%p retainCount = %u", al2, [al2 retainCount]);
}
ARCLib - compiled with ARC
#implementation ARCLib
+ (id)arcLib
{
return [[self alloc] init];
}
- (id)init
{
self = [super init];
if (self)
{
NSLog(#"%p init",self);
}
return self;
}
- (void)dealloc
{
NSLog(#"%p dealloc",self);
}
#end
Result
2012-12-13 20:15:21.879 ARCTest[15206:907] 0x1e821500 init
2012-12-13 20:15:21.883 ARCTest[15206:907] 0x1e821500 retainCount = 1
2012-12-13 20:15:21.884 ARCTest[15206:907] 0x1e821500 dealloc
2012-12-13 20:15:21.885 ARCTest[15206:907] 0x1dd26060 init
2012-12-13 20:15:21.886 ARCTest[15206:907] 0x1dd26060 retainCount = 1
2012-12-13 20:15:21.893 ARCTest[15206:907] 0x1dd26060 dealloc
Answer to your question
You should release your obj1, but don't need to release obj2.
I am doing my project in xcode 4.2 (Older Version). For my application, I just set the variables, arrays in dto class for using in entire app lifecycle. so I set with a property like this.
AppDTO(sub class of NSObject)
AppDTO.h
#property(nonatomic,retain)anotherAppDTO *aAppDTO;
#property(nonatomic,retain)NSMutableArray *array1;
#property(nonatomic,retain)NSMutableArray *array2;
#property(nonatomic,retain)NSString *string1,*string2,*string3;
AppDTO.m
- (id)init
{
self = [super init];
if (self) {
self.aAppDTO = [[anotherAppDTO alloc]init];
self.array1 = [[NSMutableArray alloc]init];
self.array2 = [[NSMutableArray alloc]init];
self.string1 = #"Hello";
self.string2= #"Hai";
}
}
-(void)dealloc
{
if(array1 != nil)
{
[array1 release];
array1 = nil;
}
if(array2 != nil)
{
[array2 release];
array2 = nil;
}
[aAppDTO release];
aAppDTO = nil;
[super dealloc];
}
when I analyze my app in Xcode 4.3.2, I get memory warning in self.array1 and self.array2 (Potential leak on object allocated on line….), but when I change self.array1 to array1, warning goes away.
What is the reason for using self. do I need to use self if I set #property(nonatomic,retain) to variables(like array1,array2,string1,string2).
Also in dealloc method, I heard we don't want to use [self.array1 release], instead we can use [array1 release]. Is it Correct?
Do I need to release my string in dealloc method.
Also I am releasing aAppDTO in dealloc method. if I allocate some objects in anotherAppDTO class, will it release automatically when I call [aAppDTO release] method.
Can anyone clarify me.
Many Thanks,
Anish
You get the warning because when you write :
self.array1 = [[NSMutableArray alloc]init];
is the same as :
[self setArray1: [[NSMutableArray alloc]init]];
As you can notice you are not allocating the underlying array1 private variable, but you are calling the setter of the property that since it is declared as retain it retains the object once assigned, this means that when you eventually will assign another object the second time with the setter the first object will remain with a retain count of one until the application will be closed (since you don't have any reference to that object anymore ...) .
Take a look at this great article to understand better Manual Reference Counting in Objective-C .
when i analyze my app in Xcode 4.3.2, i get memory warning in self.array1 and self.array2 (Potential leak on object allocated on line….), but when i change self.array1 to array1, warning goes away.
the analyzer's right. the parameter is retained when set. as well, you should favor direct access in initialization and dealloc. so, you should just write array1 = [[NSMutableArray alloc] init];, and be done.
What is the reason for using self. do i need to use self if i set #property(nonatomic,retain) to variables(like array1,array2,string1,string2).
those go through the accessor methods. if not in initialization or dealloc, you should favor going through the accessor methods because that is the common correct execution path for a fully constructed object.
Also in dealloc method, i heard we don't want to use [self.array1 release], instead we can use [array1 release]. Is it Correct?
correct.
Do i need to release my string in dealloc method.
yes.
Also I am releasing aAppDTO in dealloc method. if i allocate some objects in anotherAppDTO class, will it release automatically when i call [aAppDTO release] method.
when its reference count reaches 0, its dealloc will be called.
I think the others have answered your question.
I do want to draw your attention to Apple's excellent Advance Memory Management Programming Guide: Practical Memory Management, in which they walk through these sorts of scenarios. It's hard to take it all in on the first reading, but it really does cover this stuff. In answer to your question about the use of instance variables versus the accessor methods, I draw your attention to the section labeled to "Don't Use Accessor Methods in Initializer Methods and dealloc".
I'm new to Obj-C and I have a question concerning the autorelease. Is it ok to return an autoreleased variable for several methods? For example:
- (void) methodC {
Object anObj = [self methodB];
//Do something with anObj
}
- (Object *) methodB {
return [self methodA];
}
- (Object *) methodA {
Object anObj = [[anObj alloc] init];
release [anObj autorelease];
}
Will the variable remain valid even if it is returned up a method chain and used at the top? Or does it have to be retained somewhere along the way?
thank you
Yes, it will be valid in this case. You only have to worry about the variable being deallocated if somebody drains the autorelease pool. As long as you've written every function that returns along the way and you don't explicitly drain the autorelease pool, you don't have to worry about objects being deallocated from under you.
In the vast majority of cases, the code in the NSRunLoop takes care of draining the autorelease pool. When you return control from your application code to the API code (such as by returning from a touchesBegan handler etc.), you don't know if the autorelease pool will be drained, so you have to assume in the worst case that it will. In that case, you have to retain any objects you want to keep references to.
For example:
NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init];
Object *anObj = [self methodC]; // Your methodC as before -- anObj is valid
[pool drain]; // anObj will be deallocated here
The variable should remain valid. You only need to retain an object if it is actually "owned" by some other object and could be indirectly/unintentionally released along with it. For example, if you extracted an object from an array and then released the array, your object reference could become invalid unless you explicitly retain it.
For more details, see Object Ownership and Dismissal, particularly the sections on Autorelease and Validity of Shared Objects. The latter uses the following code to illustrate how you could "accidentally" make an object reference invalid.
heisenObject = [array objectAtIndex:n];
[array removeObjectAtIndex:n];
// heisenObject could now be invalid.
The following code shows how to mitigate this problem using retain.
heisenObject = [[array objectAtIndex:n] retain];
[array removeObjectAtIndex:n];
// use heisenObject.
[heisenObject release];
How do you release the memory in this situation? Do you have to?
- (NSString *) whatHappensHere {
NSMutableString * mutableString = [[NSMutableString alloc] initWithString:#"Hello"];
// ....
// more code ...
// ...
return mutableString;
}
With autorelease
- (NSString *) whatHappensHere {
NSMutableString * mutableString = [[NSMutableString alloc] initWithString:#"Hello"];
[mutableString autorelease];
return mutableString;
}
As willcodejavaforfood said, the convention is that any newly-allocated object returned from a method should be autorelease'd before being returned. This tells the Objective-C runtime that if nobody takes ownership of it with retain before (in most cases) the current iteration of the application event loop ends, it should be freed.
If it's just used locally in the calling function or returned up the stack, that works great and it gets freed sooner or later. If someone ends up wanting to keep it around, then they have to retain it and so will know they have to release it themselves later on.
(In fact, most of the non-init* utility constructors for ObjC base classes like strings, etc, do just that, or at least something functionally equivalent.)
Given the following property definition:
#property (nonatomic,retain) MyObject* foo;
does the following code cause a memory leak:
self.foo = [[MyObject alloc] init];
?
It looks like the alloc call increments the retain count on the object to 1, then the retain inside the property setter increases it to 1. But since the initial count is never decremented to 0, the object will stick around even when self is released. Is that analysis correct?
If so, it looks like I have two alternatives:
self.foo = [[[MyObject alloc] init] autorelease];
which is not recommended on the iPhone for performance reasons, or:
MyObject* x = [[MyObject alloc] init];
self.foo = x
[x release];
which is a bit cumbersome. Are there other alternatives?
Are there any alternatives?
No.
You are not going to be able write much of an iPhone application without using autorelease and the Cocoa Touch library uses them in many places. Understand what it's doing (adding the pointer to a list for removal on the next frame) and avoid using it in tight loops.
You can use class method on MyObject that does alloc/init/autorelease for you to clean it up.
+ (MyObject *)object {
return [[[MyObject alloc] init] autorelease];
}
self.foo = [MyObject object];
The easiest way to manage a retained property on the iPhone is the following (autorelease is not as bad as you think, at least for most uses):
-(id)init {
if (self = [super init]) {
self.someObject = [[[Object alloc] init] autorelease];
}
return self;
}
-(void)dealloc {
[someObject release];
[super dealloc];
}
The autorelease releases the reference to the floating instance which is assigned to self.object which retains its own reference, leaving you with the one reference you need (someObject). Then when the class is destroyed the only remaining reference is released, destroying the object.
As described in another answer, you can also create one or more "constructor" messages to create and autorelease the objects with optional parameters.
+(Object)object;
+(Object)objectWithCount:(int)count;
+(Object)objectFromFile:(NSString *)path;
One could define these as:
// No need to release o if fails because its already autoreleased
+(Object)objectFromFile:(NSString *)path {
Object *o = [[[Object alloc] init] autorelease];
if (![o loadFromFile:path]) {
return nil;
}
return o;
}
You are right, self.foo = [[MyObject alloc] init]; is leaking memory. Both alternatives are correct and can be used. Regarding the autorelease in such a statement: keep in mind that the object will released by the autorelease pool as soon as the current run loop ends, but it will most probably be retained a lot longer by self, so there is no issue with memory usage spikes here.