On Wed, May 29, 2013 at 02:33:07PM +0300, Mika Kuoppala wrote:
> Chris Wilson writes:
>
> > It appears that a beneficial side-effect of Mika's more accurate hangman
> > work is to speed up hang detection and execution. This exposes a bug in
> > the reset code that then treats repeated simulated h
Chris Wilson writes:
> It appears that a beneficial side-effect of Mika's more accurate hangman
> work is to speed up hang detection and execution. This exposes a bug in
> the reset code that then treats repeated simulated hangs as an
> indication that the machine is wedged. Jiggle the code aroun
It appears that a beneficial side-effect of Mika's more accurate hangman
work is to speed up hang detection and execution. This exposes a bug in
the reset code that then treats repeated simulated hangs as an
indication that the machine is wedged. Jiggle the code around so that we
only do the simula
