Re: About: timeout handler of epicsTimer

[Andrew Johnson <anj@aps.anl.gov>]

Jun-ichi Odagiri wrote:
> I'm wondering if the following statement about epicsTimer is correct or not.
>
> #Start of the statement
> Any handler of epicsTimer runs, when the timer expired, not in an interrupt
> context but in a normal task's (thread's) context.
> This goes for the case where the iocCore program is running on vxWorks or
> RTEMS, or whatever OS supported in Base.
> #End of the statement

Your statement is true; unlike the old vxWorks watchdog timers that we 
used to use in iocCore before the OSI layer was added, an epicsTimer's 
notify routine or callback is now always executed in thread context.

> What I'd like to know is, in other words, if a handler of epicsTimer is
> allowed to take semaphores in the handler or not.
>
> I guess the statement is TRUE because I have found an example of a handler
> to call printf, in "IOC Application Developer's Guide (Chp.19, Section
> 10.4)".

Yes it is allowed, but you should be aware of the implications of this: 
If the semaphore blocks, then none of the other epicsTimers that use the 
same epicsTimerQueue will be executed until the semaphore is released 
and the callback/notify routine returns.  It is not advisable for a 
timer on the system shared timer queue to block for any signficant 
period of time in its callback/notify routine.  The use of printf() and 
similar functions is fine though because it should not block for any 
significant period of time.

- Andrew
--
* * Matt Santos / / For a Brighter America * *