[Webkit-unassigned] [Bug 136038] Resource leak in object - memory allocated in constructor but not freed in destructor.

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=136038


Darin Adler <darin at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #236811|review?                     |review-
               Flag|                            |




--- Comment #6 from Darin Adler <darin at apple.com>  2014-08-19 09:00:48 PST ---
(From update of attachment 236811)
View in context: https://bugs.webkit.org/attachment.cgi?id=236811&action=review

It’s kind of weak to use unique_ptr for m_nonHTTPProtocolHost but not do the same thing for the hosts in m_hosts.

> Source/WebCore/ChangeLog:3
> +        Fixed resource leak in object 

This is not a good name for a bug. It’s much too vague. Almost like “fixed bug in project”.

And also, it’s inaccurate. This patch doesn’t fix a leak. It makes some inconsequential changes in a class, which I suppose could be helpful if someone didn’t realize the object was an immortal singleton.

> Source/WebCore/ChangeLog:8
> +        Changed m_nonHTTPProtocolHost from HostInformation* to std::unique_ptr<HostInformation> and added ASSERT_NOT_REACHED() in the destructor.

It’s a little strange to do both of these things. It’s correct to assert that the destructor is never called. That’s what makes it clear we don’t really need a unique_ptr. But I suppose it’s OK to use unique_ptr just to be a little clearer about object ownership.

> Source/WebCore/loader/ResourceLoadScheduler.cpp:107
> +    m_nonHTTPProtocolHost = std::make_unique<ResourceLoadScheduler::HostInformation>(String(), maxRequestsInFlightForNonHTTPProtocols);

Why change from construction syntax to assignment syntax? For unique_ptr that’s less efficient for no good reason. Just seems like gratuitous “make coding style a little worse” change.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.