gEDA-dev: Another autosave crash....

[Ales Hvezda]

Hi Patrick,

[snip]
>I may be wrong but I do not think a mutex will help in any way.

	Yes, you are correct.  A mutex will not help here.  I didn't
realize that in gtk+ the whole GSource mechanism is implemented via the
main loop and not a thread.

[snip]
>Could people would have experienced this bug please try adding the
>following lines to s_toplevel_delete(), recompile and let us know if
>it solved the issue or not?
>
>  if (toplevel->auto_save_timeout != 0) {
>    gboolean ret;
>    ret = g_source_remove (toplevel->auto_save_timeout);
>    g_assert (ret);
>  }


	This hopefully fixes the bug, however, on my machine the
above g_assert was failing.  This is because I had the component insert
dialog box open as well on close, and the autosave callback was being
registered for the preview window (something we really do not want).
I have refactored/moved the init of the autosave callback to the only
place we want it (in x_window_setup().  I am still seeing a jump based
on a uninitalized variable from valgrind in the above if statement that
I do not understand yet.  However, I have commited these changes to CVS
now. Please verify.

	I hope all these changes greatly improve the stability of gschem.
Thanks for the bug fix Patrick!  

								-Ales