Re: Martian Rover Spirit's mulitasking software reset.

["jimmaureenrogers@worldnet.att.net" <jimmaureenrogers@worldnet.att.net>]

Martin Dowie wrote:
> My understanding is that there was a "priority inverstion", so had they
> used "pragma Locking_Policy (Ceiling_Locking);" (RM  D.3 Priority
> Ceiling Locking) then the problem could have been avioded.
>
> Of course, they might not have used it even if they were using Ada. :-)
> Cheers

Problems such as priority inversions, deadlocks, livelocks, and race
conditions are far too common when writing Pthreads. While Pthreads
provides tools which, when used in the proper combination, allow the
programmer to avoid such problems, the programmer is expected to
know all the necessary combinations.

This follows the C programming philosophy. Assume that the programmer
is an expert and makes no mistakes.

Another possible contributor to the problem when using C is that
concurrency issues are not common considerations for C programmers.
C provides no syntax or for concurrency. C programmers are not as
familiar with the design of concurrent systems as are Ada programmers.

None of these reasons ensures that Ada programmers could not have
produced an equivalent mess. They do provide a higher likelihood of
producing a successful program.

Jim Rogers