Rather than incrementing a spilled pointer

I guess these became obsolete when the communication stuff moved into
the stdlib.

This seems to confuse LLVM in some very specific situations.

Closes #883

Closes #885

update docs to reflect the current state of syntax extension and macros

rustc: When revoking a cleanup of a unique pointer, zero it out so that the GC won't try to visit it

I don't know exactly what's going on but this optimization is giving
me problems with the check-fast runner. I also don't see how it is
correct for external tags.

When a task fails, we will throw an exception, then catch it at the bottom of
the stack.

On Windows we don't do this yet because the exception doesn't propagate
correctly.

No cleanups yet.

Issue #236

This is so that when a child dies after the parent, it still holds a valid
pointer and can call supervisor->kill() safely.

The motivation here is that the bottom of each stack needs to contain a C++
try/catch block so that we can unwind. This is already the case for main, but
not spawned tasks.

Issue #236

Issue #236

Issue #236

Closes #820

Move expressions where the RHS is an upvar are now forbidden within
block expressions.

trans::type_of now has a constraint saying that its type argument
is statically sized. This eliminates the "impossible happened" case
in type_of. Yay!

I note that this change decreased translation time for stage2/rustc
from 16.1 s to 14.0 s. I also think many of the remaining checks
could be eliminated with some mildly clever use of constrained types
and further preconditions. Future work!

Propagate the if-check constraint into the consequent even when
there's no else branch. (Oops!)