Bug 27681 – [Gtk] layout corruption

• Component: WebKitGTK
• Status: RESOLVED INVALID

[Luke Kenneth Casson Leighton]

once this is fixed:
https://bugs.webkit.org/show_bug.cgi?id=27679

or, if a copy of the subdirectory output/ is taken (e.g. with wget) and the web 2.0 (pure javascript) application is run locally, it can be seen that the layout is completely corrupted.

the entire content is displayed offset by approx 100 pixels right and 100 pixels down.  moving the mouse around, or clicking anywhere, shows the "true" position of the content, whereupon the content is redrawn in its correct location.

[Luke Kenneth Casson Leighton]

ok - an svn update and use of absolute latest git of libsoup (2.27.4 or something like that from march 2009) the segfault stops happening, and so the layout issue can be clearly seen.

[Jan Alonzo]

(In reply to comment #1)
> ok - an svn update and use of absolute latest git of libsoup (2.27.4 or
> something like that from march 2009) the segfault stops happening, and so the
> layout issue can be clearly seen.

Closing given that this works in libsoup 2.27.4 and recent WebKitGtk.

[Luke Kenneth Casson Leighton]

sorry, jan,

you closed the other two bugs which are segfaults but you did not read clearly what this bugreport says.

it says "layout corruption can only be seen if you don't get the segfault which, obviously, stops the page from being displayed".

that the segfault has been fixed does not mean that the layout corruption does not occur.

therefore, reopening this bugreport.

[Jan Alonzo]

(In reply to comment #3)
> sorry, jan,
> 
> you closed the other two bugs which are segfaults but you did not read clearly
> what this bugreport says.
> 
> it says "layout corruption can only be seen if you don't get the segfault
> which, obviously, stops the page from being displayed".
> 
> that the segfault has been fixed does not mean that the layout corruption does
> not occur.
> 
> therefore, reopening this bugreport.

Hi Luke. Which layout is corrupted? The page loads a bit slow, but I couldn't tell the difference once loaded.

[Luke Kenneth Casson Leighton]

(In reply to comment #4)
> (In reply to comment #3)
> > sorry, jan,
> > 
> > you closed the other two bugs which are segfaults but you did not read clearly
> > what this bugreport says.
> > 
> > it says "layout corruption can only be seen if you don't get the segfault
> > which, obviously, stops the page from being displayed".
> > 
> > that the segfault has been fixed does not mean that the layout corruption does
> > not occur.
> > 
> > therefore, reopening this bugreport.
> 
> Hi Luke. Which layout is corrupted? 

 http://pyjs.org/examples/kitchensink/output/KitchenSink.html

> The page loads a bit slow,

 yes.  pure javascript. rather a lot of it.

>  but I couldn't
> tell the difference once loaded.

 the fault i'm seeing is that .... ok, screenshot time, i'll add it as an attachment.

[Luke Kenneth Casson Leighton]

Created attachment 33508 [details]
showing layout corruption

you can see that the background colour is correct but that the text is offset.  things that are re-displayed at any time are displayed in their correct locations.  it's all very weird.

[Martin Robinson]

Using a recent build of WebKit GTK+ I cannot reproduce this issue. Feel free to re-open if you are still having issues.