RFR(XS): 8167300: Local scheduling failures during gcm should be fatal

[Roland Westrelin]

Hi Vladimir,

>>> Second, is not better for debugging to add assert in a place where we
>>> record failure instead of waiting when we exit compilation?
>>
>> Are you suggesting this?
>
> Yes, it is in addition to your changes.

Why in addition?
Shouldn't this be all we need:

http://cr.openjdk.java.net/~roland/8167300/webrev.01/

Roland.