Fwd: Problem report: Struct aliasing problem causes Thread_Ready_Chain corruption in 4.6.99.3

[Peer Stritzinger]

Joel,

> > But we'd need an actual piece of code exposing this problem. Anything
> > else is "poking around in the fog"
> >
> Peer's example is cpukit/score/src/threadresettimeslice.c.

Yes.

> Peer .. is it possible to construct a simple executable test case that
> breaks?  It
> sounds like it would be something like two tasks at the same priority
> and some
> code to schedule them like this.

I gave up on it while debugging.  I'm running here a big application
that doesn't show the
problem.  My customer adds some more app code to it and it triggers
only on heavily loaded systems after many hours.

But it could be easier when knowing what the code does.

> Hand examining PowerPC code isn't fun.

If I have a bit of quiet I actually like it ;-)  And the code GCC
produces for PowerPC is
pretty well optmized almost hand assembly quality.  On the other hand
this smartnes
together with some C programming tricks caused the problem and did me
in for three
weeks debugging ;-)

I'll try to get back with soemthing -- probably the analysis -- as
soon as possible.
Have to complete a revision for the customer first however.

Regards,
-- Peer