Rfork is not splitting the stack when creating a new thread,
so the parent and child are executing on the same stack.
However, if the parent returns and keeps executing before
the child can read the arguments from the parent stack,
the child will not see the right arguments. The solution
is to load the needed pieces from the parent stack into
register before INT $64.

Thanks to Russ Cox for the explanation.

LGTM=rsc
R=rsc
CC=ality, golang-codereviews
https://golang.org/cl/64140043