KIO from inside a parsejob?

Andreas Pakulat apaku at gmx.de
Wed Jul 15 15:56:58 UTC 2009


On 15.07.09 16:52:04, Milian Wolff wrote:
> Andreas Pakulat wrote:
> >On 15.07.09 07:39:49, Niko Sams wrote:
> >>On Wed, Jul 15, 2009 at 01:48, Milian Wolff<mail at milianw.de> wrote:
> >>>Am Dienstag, 14. Juli 2009 22:43:19 schrieb Milian Wolff:
> >>>>Hey guys,
> >>>>
> >>>>could it be that KIO must not be accessed from a parsejob? Look at that, I
> >>>>can't explain it:
> >>>>
> >>>>http://mwolff.pastebin.com/f4e26ba49
> >>>>
> >>>>I try to fix the includebuilder for PHP and remote projects. My code now:
> >>>>
> >>>>http://mwolff.pastebin.com/f6b676fc0
> >>>>
> >>>>Here's the debug output on the console:
> >>>>http://mwolff.pastebin.com/f130f444b
> >>>>
> >>>>What's that about QPair's?! Is my build fubar? Help!
> >>>It's definitly the KIO::NetAccess::exists(). If I comment that out everything
> >>>seems to work fine.
> >>>
> >>>Niko: Have you ever tested that? Or did it work for you back then?
> >>I never tested with non-local files.
> >>
> >>I think we can't use KIO in a background thread (?).
> >
> >I suggest asking on kde-core-devel (unless you've examined kio's code ;)
> 
> Just asked in #kde-devel and pinotree said KIO is not threadsafe. So
> yeah, I think I'll have to remove that code...
> 
> Niko, I'd say we change the whole logic to the following:

Why not make your code using kio threadsafe then? Even if you cannot use
kio across threads, you can make sure that you're using it only from one
thread. Even if you can use kio just from the GUI thread you can still
do this.

Andreas

-- 
Today is the last day of your life so far.




More information about the KDevelop-devel mailing list