Adjust canonical ABI option syntax to be less verbose

[alexcrichton]

Currently the syntax for specifying a memory with a canonical option is:

(memory (core memory $memory))

and optionally instance alias sugar can also be used:

(memory (core memory $libc "memory"))

This PR proposes changing these two syntaxes to:

(memory $memory)
(memory $libc "memory")

with the theory that the "core" part is already implied by the canonical
option itself and otherwise saying "memory" twice is redundant.

[lukewagner]

Totally agreed the current syntax feels overly verbose in these scenarios so thanks for bringing this up!

So, first, I was already on the edge about whether we needed the core-prefix here (given the static context of "this is definitely a core index"), so agreed we should drop that part.

If we wanted to drop the (sort ...) that wraps the <core:instanceidx> <name>, then I think we'd need to define some custom syntactic sugar just for the instantiate instruction because, as is, if the AST has, e.g., (realloc <core:funcidx>), then the inline alias sugar only lets you substitute in a (func $inst "exp") for the <core:funcidx>, producing (realloc (func $inst "exp")). memory is a more ambiguous case since memory is a case of sort, but in this context, the memory token is serving as a label of an immediate, which could have just as well been named not-memory, e.g. into. Because the sugar-substitution is purely lexical (it doesn't care about sort vs. immediate), you could say that the inline alias sugar already allows you to write what you're proposing in this PR. (FWIW, I tried to think of a better immediate label than memory to this wasn't a special case, but I couldn't find one: into only makes sense for lifting; mem feels contrived.)

So: drop the core-prefix and but keep the (func ...) for realloc/post-return?

[alexcrichton]

👍 sounds reasonable to me!

[lukewagner]

Thanks! Could you also update examples/SharedEverythingDynamicLinking.md (where, it appears, I wasn't even already fully following the explainer...).

[lukewagner]

Suggested change

	| (realloc (func <core:funcidx>))
	| (post-return (func <core:funcidx>))
	| (realloc <core:funcidx>)
	| (post-return <core:funcidx>)

The idea being that you can either write (canon lift $f (realloc $r) (func $l)) or, via inline-alias-sugar, (canon lift $f (realloc (func $i "r")) (func $l)).

[alexcrichton]

Sure thing, should be updated now I believe.

[lukewagner]

Thanks!