[Webkit-unassigned] [Bug 100639] While Using WebGL, MiniBrowser segfaults on Refreshing the page.

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=100639


Kenneth Rohde Christiansen <kenneth at webkit.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #171191|review?                     |review+
               Flag|                            |




--- Comment #5 from Kenneth Rohde Christiansen <kenneth at webkit.org>  2012-10-30 02:53:45 PST ---
(From update of attachment 171191)
View in context: https://bugs.webkit.org/attachment.cgi?id=171191&action=review

> Source/WebCore/ChangeLog:12
> +        TextureMapperSurfaceBackingStore can import texture's from a GraphicSurface.
> +        In such cases GraphicsSurfaceGLX creates an XPixmap to read texture content
> +        from a given WindowId but doesn't create any new window.However, OffScreenRootWindow
> +        always tries to unmap window(in it's destruction) resulting in segfault.Now,We check
> +        if the window is valid before trying to unmap it.

Could you add some spaces after . and start of new sentences

> Source/WebCore/platform/graphics/surfaces/qt/GraphicsSurfaceGLX.cpp:96
>      ~OffScreenRootWindow()
>      {
>          if (!--m_refCount) {
> -            XUnmapWindow(m_display, m_window);
> -            XDestroyWindow(m_display, m_window);
> -            if (m_display)
> +            if (m_display) {
> +                if (m_window) {
> +                    XUnmapWindow(m_display, m_window);
> +                    XDestroyWindow(m_display, m_window);
> +                    m_window = 0;
> +                }
>                  XCloseDisplay(m_display);
> -            m_display = 0;
> +                m_display = 0;
> +            }
>          }
>      }

I would try avoiding indentating tha much

like

if (--m_refCount)
    return;

...

That is like the webkit way

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