PLEASE TEST: IPI deadlock avoidance patch

[Craig Boston]

On Mon, Aug 30, 2004 at 08:11:58PM +0200, Daniel Eriksson wrote:
> Ok, I've run a few more tests now, and the results are not looking very good
> for gvinum in RAID-5 mode:

I'm afraid I have to confirm these results.  It took a while to get
everything back up and running (had problems building world for a while
because of a few key utilities that got bitten by the corruption).  I
have it sorted out now, and have successfully built world a few times
with SMP enabled, but using regular vinum instead of gvinum -- no sign
of any corrupt files.

So it seems that gvinum RAID-5 is a key factor in the corruption, and
SMP greatly increases the odds of it happening...

Is it worth creating a PR so this doesn't get lost amid the release
madness?

Craig