Experimental Physics and
Industrial Control System

Till Straumann <[email protected]> · Tue, 26 Nov 2002 15:24:47 -0800

Andrew Johnson wrote:

IMO, clamping everyghing to < 91 with the shell running at 91 is safe

- no runaway (base) task can lock the console. However, the user

still has the option of doing really critical work at a higher

priority, if necessary - at the risk of having a locked-up system

should such a task

run away, of course.

... or some other task runs away that owns a resource that the high

priority task needs and has its priority raised through inheritance.

This is still an issue that the application designer should know about

(but beware if that semaphore is hidden behind some OS API...).

Yeah - (although theoretically possible), I consider this rather

unlikely. A RT critical task should only share very 'fast' resources

(i.e. objects that are mutexed for very few lines of code [low runaway

error probability]) with a low priority task. A high priority, RT

critical task who operates on/shares complex resources (doing things

e.g. like printf, malloc or networking) is probably ill-designed.

It might seem prudent to have a control character in the console port's

ISR that can raise the shell's priority in an emergency - I remember we

discussed this idea round the table at JLAB anyway.

Sure - that would be nifty and doable under RTEMS - but what to do under

vxWorks? Another problem is that such a solution is probably highly BSP

specific.

OTOH - rather than building a generic solution into the console ISR [for

multiple OSes/BSPs] that burden could be off-loaded to the user who

really needs the high priority task. For the particular board/OS he/she

uses it's easy to use any IRQ (e.g. a serial line, an onboard switch or

a watchdog) to suspend the critical task on request.

Now another question. RTEMS, and I expect vxWorks as well, have

system tasks handling things like network I/O. What priority should

these run at? At the moment I've got the RTEMS network tasks at

higher than the highest EPICS task priority. From the above

discussions it would seem that running the network tasks at an RTEMS

priority equivalent to EPICS task priority 91 would be the right thing

to do.

 From past experience with vxWorks, you probably do not want your

console to be below the network tasks in case nasty things happen that

can lock you out and prevent any chance of debugging.

That makes sense to me.

IIRC, in vxWorks

the only tasks higher than tShell handle message logging and system

callbacks from exceptions (things that can't be done in ISR context).

- Andrew

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System