[Openmcl-devel] Semaphore troubles

[James M. Lawrence]

On Wed, May 9, 2012 at 7:43 PM, Erik Pearson <erik at defunweb.com> wrote:
> HI James,
>
> Interesting question.
>
> What about the case where the two tasks pushed onto the task stack complete
> their subsequent pushes to the receive stack before the first call to
> popping the receive stack?

Hi, thanks for investigating.

As I mentioned, using a sleep of 0.5 also leads to failure. I just
tried some higher times:

1 second -- failure at 103 iterations
2 seconds -- failure at 49 iterations
3 seconds -- failure at 7 iterations
10 seconds -- failure at 13 iterations

Doesn't that deflate your theory? Also, can you explain why the
problem goes away without the loop?