i-g-t test for gem flink race

[Dave Airlie]

As we talked on irc, I decided to write a test case and it actually
seems to trigger the race.

Really its insane userspace behaviour and I'm not really sure we
should care to fix, I can't see it causing an oops, more userspace
does something insane, it gets a bad result, but I'll contemplate it a
bit more.

http://cgit.freedesktop.org/~airlied/intel-gpu-tools/commit/?h=prime&id=2e7433f87b337116ab848c12b05a1a05891640df

is the current test, it fails for me here usually once in
100,000->200,000 cycles.

Dave.