Bug 34416 - [Qt] Build without SSL support is broken
Summary: [Qt] Build without SSL support is broken
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: New Bugs (show other bugs)
Version: 528+ (Nightly build)
Hardware: PC Windows XP
: P3 Normal
Assignee: Nobody
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-02-01 04:45 PST by Henry Haverinen
Modified: 2010-02-01 07:09 PST (History)
4 users (show)

See Also:

Attachments
build fix (2.96 KB, patch)
2010-02-01 06:59 PST, Henry Haverinen 		no flags Details | Formatted Diff | Diff
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Henry Haverinen 2010-02-01 04:45:03 PST 
If Qt is configured without SSL support, then the build breaks in SocketStreamHandler.
Comment 1 Yael 2010-02-01 06:42:50 PST 
Sorry, I forgot about that.
Comment 2 Simon Hausmann 2010-02-01 06:48:31 PST 
Henry has a patch coming up :)
Comment 3 Kenneth Rohde Christiansen 2010-02-01 06:52:57 PST 
Our product manager is programming now? :-) rocks!
Comment 4 Henry Haverinen 2010-02-01 06:59:24 PST 
Created attachment 47839 [details]
build fix
Comment 5 Kenneth Rohde Christiansen 2010-02-01 07:02:18 PST 
Comment on attachment 47839 [details]
build fix

You need to put the patch up for review, I have done so now.


> +    } else
>          m_socket = new QTcpSocket(this);
> +
> +    if (!m_socket)
> +        return;

Why did you add this early return. The new is supposed to return something valid, unless we are out of memory.
Comment 6 Simon Hausmann 2010-02-01 07:04:02 PST 
Committed r54134: <http://trac.webkit.org/changeset/54134>
Comment 7 Henry Haverinen 2010-02-01 07:07:42 PST 
(In reply to comment #5)
> (From update of attachment 47839 [details])
> You need to put the patch up for review, I have done so now.
> > +    } else
> >          m_socket = new QTcpSocket(this);
> > +
> > +    if (!m_socket)
> > +        return;
> Why did you add this early return. The new is supposed to return something
> valid, unless we are out of memory.

The return is not related to the new just above it, but to an #ifndef'd new above it. Without SSL support, there's no way to create a valid socket for this case.
Comment 8 Kenneth Rohde Christiansen 2010-02-01 07:09:16 PST 
(In reply to comment #7)
> The return is not related to the new just above it, but to an #ifndef'd new
> above it. Without SSL support, there's no way to create a valid socket for this
> case.

OK, great! :-)