Hello again, Thanks a lot for replying, I really was near going postal ;-)
> You should replace the g_assert above with 'g_assert (GDK_IS_PIXMAP > (pixmap))' > > If it fails you need to track down where the supposedly valid pixmap > object gets turned into an invalid pointer. Is it properly initialized > in the first place? Does it get destroyed for one reason or another in > between? I now check with "g_assert (GDK_IS_PIXMAP(pixmap))" straight after creation of the pixmap and just before gdk_pixmap_ref(), and it never asserts but fails again at gdk_pixmap_ref. Hmmm... strange since I first thought that could be since the JNI-implementation of the JVM maybe does some weird things on ARM, but since it passes GDK_IS_PIXMAP am I right to assume that the pointer is ok? Thanks again, lg Clemens Creation-time: ----------------------------------------------------------------------------------------------- if (offScreen (env, obj) == JNI_FALSE) { GdkPixmap* pix = gdk_pixbuf_new (GDK_COLORSPACE_RGB, TRUE, 8, width, height); g_assert(GDK_IS_PIXMAP(pix)); createRawData (env, obj, pix); } else { GdkPixmap* pix = gdk_pixmap_new (NULL, width, height, gdk_rgb_get_visual ()->depth); g_assert(GDK_IS_PIXMAP(pix)); createRawData (env, obj, pix); } Use-Time: --------------------------------------------------------------------------------------------- pixmap = cp_gtk_image_get_pixmap (env, source); g_assert(pixmap != NULL); g_assert(GDK_IS_PIXMAP(pixmap)); gdk_pixmap_ref (pixmap); <------------------------> Dies here with message: "g_object_ref: assertion `G_IS_OBJECT (object)' failed" _______________________________________________ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers