[PATCH, ppc] Fix hw *points for embedded ppc in a threaded environment

[Pedro Alves]

On 04/22/2013 04:32 PM, Luis Machado wrote:

> For the intermitent x86 failure, raising the usleep time a bit (1 to 5) seems to solve it.

>  I blame this on something other than GDB then.

Blaming the test itself is plausible.  However, the extra usleep time just
means the race window is smaller, not fixed, so still subject to the
occasional racy failure.

-- 
Pedro Alves