On Dec 6, 2011, at 4:59 PM, Ken Thomases wrote:
> On Dec 6, 2011, at 2:47 PM, Larry Campbell wrote:
>
>> void allocAndRaise()
>> {
>> [[[NSString alloc] initWithString:@"foo bar and zot"] autorelease];
>> [NSException raise:@"foo" format:@"bar"];
>> }
>>
>> //#define LEAK
>>
>> void foo()
>> {
>> NSAutoreleasePool *pool = [NSAutoreleasePool new];
>> NS_DURING {
>> allocAndRaise();
>> } NS_HANDLER {
>> #ifdef LEAK
>> [localException raise];
>> #else
>> [localException retain]; // I don't think this workaround should
>> be necessary, but it is
>> [pool release];
>> [[localException autorelease] raise];
>> #endif
>> } NS_ENDHANDLER
>> [pool release];
>> }
>
> The workaround is necessary because the exception object raised in
> allocAndRaise() has been autoreleased. It is no longer owned by anything and
> is only still around because the pool containing the pending release(s) has
> not been drained. When you drain it, that executes the pending release(s)
> and ultimately deallocates it.By "workaround" I didn't mean just the "[localException retain]", I meant the "[pool release]", without which the retain/autorelease of localException would be unnecessary. > That is one of the reasons you are discouraged from releasing/draining > autorelease pools in exception handlers or @finally blocks. But if you don't do that, it leaks the pool. Which, though smaller, is still a leak. For what it's worth, gnustep doesn't leak in this situation. (Well, it has a different leak, for which I just submitted a fix.) - lc
smime.p7s
Description: S/MIME cryptographic signature
_______________________________________________ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
