Experimental Physics and
Industrial Control System

Andrew Johnson <[email protected]> · Mon, 11 Jun 2012 09:59:13 -0500

Hi again,

Earlier I wrote:
> Your 'casr 2' is being run at a much higher priority than the other Channel
> Access activities.  I suspect your CPU is too busy handling higher priority
> operations to spend any time actually running the low-level CA threads that
> communicate with client.

To finish my theory of what's happening here...

When 'casr 2' is run from the shell it goes through each of the client 
connections in turn and takes a mutex for the channel.  If the mutex is 
already owned by a thread that is otherwise CPU-starved, the act of taking the 
mutex triggers the vxWorks thread priority inversion protection and raises the 
priority of the thread that owns the mutex up to the shell's priority.  This 
gives it a chance to actually run and complete what it was trying to do.  
That's why I think your 'casr 2' causes the duplicate circuits to get cleaned 
up properly.

- Andrew
-- 
Never interrupt your enemy when he is making a mistake.
-- Napoleon Bonaparte

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System