[rkward-tracker] [ rkward-Bugs-1878594 ] FIXED IN SVN: X11 ONLY: crash/hang when using Gtk2 packages

SourceForge.net noreply at sourceforge.net
Fri Nov 19 13:17:45 UTC 2010


Bugs item #1878594, was opened at 2008-01-24 01:37
Message generated for change (Comment added) made by tfry
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=459007&aid=1878594&group_id=50231

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
>Resolution: Fixed
Priority: 5
Private: No
Submitted By: Nobody/Anonymous (nobody)
Assigned to: Nobody/Anonymous (nobody)
>Summary: FIXED IN SVN: X11 ONLY: crash/hang when using Gtk2 packages

Initial Comment:
library(RGtk2) # hangs

the message in konsole is 

Xlib: unexpected async reply (sequence 0x55)!

Can someone confirm? It might just be a Debian thing (using r-cran-rgtk2 here)

Ben Goodrich


----------------------------------------------------------------------

>Comment By: Thomas Friedrichsmeier (tfry)
Date: 2010-11-19 14:17

Message:
The experimental fix caused troubles, and I have reverted it.

Instead, in the SVN version, the R backend now runs in a separate process
(at least by default). This fixes the incompatiblities with rJava and RGtk2
on X11.

Please test!

----------------------------------------------------------------------

Comment By: Thomas Friedrichsmeier (tfry)
Date: 2010-10-21 13:36

Message:
An experimental fix is in SVN: http://p.sf.net/rkward/svn . If you are
affected by this problem, please give this a try.

----------------------------------------------------------------------

Comment By: Thomas Friedrichsmeier (tfry)
Date: 2009-09-07 12:24

Message:
Minor update:

After a recent fix, on Windows, RGtk2 and friends appear to work fine.

On Linux / X11 the situation is unchanged. Debugging indicates that the R
thread is stuck waiting in XSync() or _XReply() (typically from within
gdk_display_open()).

----------------------------------------------------------------------

Comment By: Thomas Friedrichsmeier (tfry)
Date: 2009-04-02 22:51

Message:
The issue has changed somewhat (without any changes on rkward's side).
   library (RGtk2)
does not hang, now. However, calling
   gdkDisplayOpen(":0")
after that, makes the R thread hang, while the main thread continues.

   library(gWidgetsRGtk2)
causes similar symptoms.

Any help is still much appreciated.

----------------------------------------------------------------------

Comment By: Thomas Friedrichsmeier (tfry)
Date: 2008-01-25 01:38

Message:
Logged In: YES 
user_id=300591
Originator: NO

I see the same issue (also on Debian, but downloaded RGtk2 from CRAN via
R).

The issue is not 100% reproducible, but perhaps 98% so.

If you run
> Sys.setenv (DISPLAY=":0.123412")
> library (RGtk2)
there is no hang, but RGtk2 complains about failing to initialize. Now do
> gdkDisplayOpen(":0.0")
and you get the hang, again, so this is where the problem is at.

Somehow the issue seems to be that RGtk2 (running in the "R thread" in
RKWard) tries to talk to the X server concurrently with Qt. However, I
don't really understand, why this should be a problem, since RGtk2 should
use its own XDisplay/connection, which should be fine, as far as I am
aware. Also in RKWard 0.4.9 there is no such problem.

Unfortunately, I have no good idea on this at all. Any help will be much
appreciated.

Starting rkward with
# rkward --sync
makes the problem go away most (but not all) of the time.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=459007&aid=1878594&group_id=50231




More information about the rkward-tracker mailing list