Experimental Physics and
Industrial Control System

Andrew Johnson <[email protected]> · Thu, 25 Feb 2010 12:02:22 -0600

On Thursday 25 February 2010 07:43:51 Benjamin Franksen wrote:
>
> It remains to draw practical conclusions for how to avoid such race
> conditions in SNL code. Should we generally be wary of pvPut and monitor to
> the same state variable?

That is certainly my recommendation, since doing so avoids having to think 
through the possible consequences for the state program.  Thanks for posting 
your detailed analysis, I think it shows that the issue is a very subtle one.

When I've suggested not doing that in the past some people have said that they 
use the technique without problems, but I suspect that they could still suffer 
from it under some circumstances, they just haven't seen it /yet/ (or maybe 
their application doesn't fail spectacularly when it happens).

> Could the compiler warn us about that? Or is this too restrictive?

If someone wants to develop the code to add that as a warning I would be in 
favour of merging it.  I suspect that it wouldn't be too hard to do, you'd do 
the check whenever you process a pvPut() call (but I'm not a compiler writer 
so I could be wrong about the difficulty).  As long as it's a warning I don't 
think it would be too restrictive (C and C++ compilers have a long history of 
adding warnings as the developers discover more common anti-patterns).

- Andrew
-- 
The best FOSS code is written to be read by other humans -- Harald Welte

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System