Experimental Physics and
Industrial Control System

Pete Jemian <[email protected]> · Fri, 24 Oct 2014 12:29:37 -0500

That's the real test.  ... stop MEDM and see if problem continues.

Another test is, with MEDM and Winteracter application both running,

stop the IOC and see if the focus stops switching.

Seems that an experienced Motif programmer might understand how this

interaction is even possible.

You asked:
> What kind of events are created by the EPICS system which
> could be picked up by an application which is not EPICS-aware?

MEDM receives CA monitor updates and converts some of those updates into

updates to its Motif interface. These updates in MEDM are the path,

however remote such a possibility exists, that Motif events from MEDM

are being received by your Winteracter application.

On 10/24/2014 11:46 AM, Mark Rivers wrote:

Hi Eliot,

If you stop medm from processing events by putting it in edit mode does the problem immediately stop?

If you kill medm and run another EPICS Motif CA application (probe, stripTool, etc.) do you have the problem?

Mark

________________________________________
From: [email protected] [[email protected]] on behalf of Jennings, Guy [[email protected]]
Sent: Friday, October 24, 2014 11:10 AM
To: [email protected]
Subject: Re: Focus deadlock

Doesn’t the common factor seem to be motif?

On Oct 24, 2014, at 9:50 AM, Specht, Eliot D. <[email protected]> wrote:

Pete,

Can you help me to make a well-posed question for the Winteracter developer? What kind of events are created by the EPICS system which could be picked up by an application which is not EPICS-aware? That is, how would you describe a CA monitor event to a developer who is unfamiliar with EPICS?

Eliot Specht

Subject:               Re: Focus deadlock
From:    Pete Jemian <[email protected]>
To:          <[email protected]>
Date:     Fri, 24 Oct 2014 07:50:49 -0500
Can this be a an issue where the GUI is receiving CA monitor events directly and focusing the associated window with each event? If so, can you throttle widget updates in Winteracter to a fixed interval rather than updating the GUI for each CA monitor event received?

On 10/24/2014 7:36 AM, Specht, Eliot D. wrote:
I’m seeing a conflict between:

“medm pilatusDetector.adl” and a Winteracter program.  Winteracter
creates GUIs for Fortran programs using Motif. The Winteracter program
is simply waiting for input while the problem occurs. The result is that
input focus shifts rapidly between two Winteracter windows. The medm
windows do not get input focus, so it’s hard to understand how medm
contributes to the problem.

Subject:               Re: Focus deadlock

From:    Pete Jemian <[email protected]>

To:          <[email protected]>

Date:     Thu, 23 Oct 2014 13:26:17 -0500

What are the GUI applications involved?

On 10/23/2014 1:15 PM, Specht, Eliot D. wrote:

I’m running an application on a Linux x86_64 machine which opens several

windows for user input and data display. It runs fine unless I have an

EPICS areadetector IOC running in the background, in which case I get a

“focus deadlock”, where focus shifts rapidly between two application

windows and the screen becomes unresponsive to keyboard or mouse input.

The problem occurs under Gnome but not KDE. Running KDE is an effective

fix, but it would be nice to get to the bottom of the problem. Can

anyone suggest what’s going on?

Eliot Specht                        Voice: (865)574-7682

Oak Ridge National Laboratory       FAX:   (865)574-4913

P.O. Box 2008, MS 6064

Oak Ridge, TN 37831-6064

Eliot Specht                        Voice: (865)574-7682

Oak Ridge National Laboratory       FAX:   (865)574-4913

P.O. Box 2008, MS 6064

Oak Ridge, TN 37831-6064

--
----------------------------------------------------------
Pete R. Jemian, Ph.D.                <[email protected]>
Beam line Controls and Data Acquisition, Group Leader
Advanced Photon Source,   Argonne National Laboratory
Argonne, IL  60439                   630 - 252 - 3189
-----------------------------------------------------------
    Education is the one thing for which people
       are willing to pay yet not receive.
-----------------------------------------------------------

Eliot Specht                        Voice: (865)574-7682
Oak Ridge National Laboratory       FAX:   (865)574-4913
P.O. Box 2008, MS 6064
Oak Ridge, TN 37831-6064

--
----------------------------------------------------------
Pete R. Jemian, Ph.D.                <[email protected]>
Beam line Controls and Data Acquisition, Group Leader
Advanced Photon Source,   Argonne National Laboratory
Argonne, IL  60439                   630 - 252 - 3189
-----------------------------------------------------------
   Education is the one thing for which people
      are willing to pay yet not receive.
-----------------------------------------------------------

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System