Re: g_idle_add() clarification

[Ian Chapman <ichapman videotron ca>]

Thanks Chris,
    I think that not only is my code broken my idea on how to do what I 
want is broken.  I must fix that before I start on the code.

    My starting point was two button, scan off and scan on in glade.  
Scan on sets a flag and it scans til the off button is pressed.  So the 
off button could not be pressed til end-of-scan, a situation commonly 
known as a deadly embrace in hardware design.  After that g_idle_add() 
was pointed out to me and I've been innovating rather blindly ever since.

    I interpreted g_idle_add() as inserting Scanning() into the main 
loop and if Scanning returned FALSE it was removed from the main loop.  
So I guess I'm not understanding the following explanation of
guint g_idle_add(GSourceFunc function, gpointer data); :-

Adds a function to be called whenever there are no higher priority 
events pending to the default main loop.  The function is given the 
default idle priority, G_PRIORITY_DEFAULT_IDLE. If the fRegards 
Ian.unction returns FALSE it is automatically removed from the list of 
event sources and will not be called again.
This internally creates a main loop source using g_idle_source_new() and 
attaches it to the main loop context using g_source_attach(). You can do 
these steps manually if you need greater control.
Parameters
function,  function to call
data, data to pass to function .

Also Chris you indicated that maybe g_idle_add() should not be in a 
callback.  Maybe it should be in main() after gtk_widget_show (window);  
I'm stuck for the moment not understanding.

Richard's suggestion improved things but my code and idea is still 
broken on exiting Scanning()

        Regards Ian.

On 01/20/2015 04:57 PM, Chris Vine wrote:
> On Tue, 20 Jan 2015 16:47:27 -0500
> Ian Chapman <ichapman videotron ca> wrote:
> > Thanks for your reply Richard.
> > Yes, I know that.  I thought Scanning() was a lower level than the
> > gtk window and thus the StartScan_active() would exit before the
> > mainloop would add Scanning() and then only execute Scanning() when
> > nothing else (signals) was active.
> >
> > I was also asking about the GLib-CRITICAL stuff on exiting
> > Scanning().  Ian.
> >
> > On 01/20/2015 03:34 PM, richard boaz wrote:
> > > hi,
> > >
> > > you are calling the function Scanning() in the g_idle_add() call.g_idle_add()
> > >
> > > richard
> > >
> > > On Tue, Jan 20, 2015 at 3:23 PM, Ian Chapman <ichapman videotron ca
> > > <mailto:ichapman videotron ca>> wrote:
> > >
> > >      Hi all,
> > >
> > >      Sorry to bother but I'm lost with the workings of
> > >
> > >      guint g_idle_add ()
> > >
> > >      My code is as
> > >      follows
> > >
> > >      extern "C"
> > >
> > >      void on_StartScan_activate() //from gtk menu item
> > >
> > >      { printf("Start Scan Activated. \n");
> > >
> > >      scan_on_fg =1;
> > >
> > >      Scanning_lvl = g_idle_add((GSourceFunc) Scanning(), (gpointer)
> > > 0);g_idle_add()
> > > g_idle_add()
> > >      printf("Exiting StartScan.\n");  //main.cpp, Line 129
> > >
> > >      return; }
> > >
> > >      I get to Scanning() okay but I don't exit
> > > on_StartScan_activeat() until I exit Scanning() (not what I
> > > expected).
> > >
> > >      Worse still when I step through Scanning as I exit I get
> > >
> > >      (gdb) next
> > >
> > >      (main:24011): GLib-CRITICAL **: g_idle_add_full: assertion
> > >      `function != NULL' failed
> > >
> > >      on_StartScan_activate () at main.cpp:129
> > >
> > >      (gdb) next
> > >
> > >      Exiting StartScan.
> > >
> > >
> > >      guint Scanning(void)
> > > g_idle_add()
> > >      {
> > >
> > >      while(1 == scan_on_fg)
> > >
> > >      { //Bean counting with rest of GUI in control, I hoped. }
> > >
> > >      return(FALSE); //Remove this thread.
> > >
> > >      }
> > >
> > >
> > >      Regards Ian.
> What Richard is telling you is that your code is broken.  Fix that
> before dealing with warnings.  The point is that your broken code is
> doing what the warning is telling you about - passing a FALSE value
> (which equates here to NULL here), as returned by Scanning(), to the
> function pointer argument of g_idle_add().
>
> You have other problems.  You are passing a callback to g_idle_add()
> which does not have the correct signature and surpressing this failure
> by a cast.  (Why?.)  Fix that as well and see what warnings are left
> over.
>
> Chris
> _______________________________________________
> gtk-list mailing list
> gtk-list gnome org
> https://mail.gnome.org/mailman/listinfo/gtk-list