Add reference output tests for JS operations (#1894)

* Add reference output tests for JS operations

This commit starts adding a test suite which checks in, to the
repository, test assertions for both the JS and wasm file outputs of a
Rust crate compiled with `#[wasm_bindgen]`. These aren't intended to be
exhaustive or large scale tests, but rather micro-tests to help observe
the changes in `wasm-bindgen`'s output over time.

The motivation for this commit is basically overhauling how all the GC
passes work in `wasm-bindgen` today. The reorganization is also included
in this commit as well.

Previously `wasm-bindgen` would, in an ad-hoc fashion, run the GC passes
of `walrus` in a bunch of places to ensure that less "garbage" was seen
by future passes. This not only was a source of slowdown but it also was
pretty brittle since `wasm-bindgen` kept breaking if extra iteams leaked
through.

The strategy taken in this commit is to have one precise location for a
GC pass, and everything goes through there. This is achieved by:

* All internal exports are removed immediately when generating the
  nonstandard wasm interface types section. Internal exports,
  intrinsics, and runtime support are all referenced by the various
  instructions and/or sections that use them. This means that we now
  have precise tracking of what an adapter uses.

* This in turn enables us to implement the `add_gc_roots` function for
  `walrus` custom sections, which in turn allows walrus GC passes to do
  what `unexport_unused_intrinsics` did before. That function is now no
  longer necessary, but effectively works the same way. All intrinsics
  are unexported at the beginning and then they're selectively
  re-imported and re-exported through the JS glue generation pass as
  necessary and defined by the bindings.

* Passes like the `anyref` pass are now much more precise about the
  intrinsics that they work with. The `anyref` pass also deletes any
  internal intrinsics found and also does some rewriting of the adapters
  aftewards now to hook up calls to the heap count import to the heap
  count intrinsic in the wasm module.

* Fix handling of __wbindgen_realloc

The final user of the `require_internal_export` function was
`__wbindgen_realloc`. This usage has now been removed by updating how we
handle usage of the `realloc` function.

The wasm interface types standard doesn't have a `realloc` function
slot, nor do I think it ever will. This means that as a polyfill for
wasm interface types we'll always have to support the lack of `realloc`.
For direct Rust to JS, however, we can still optionally handle
`realloc`. This is all handled with a few internal changes.

* Custom `StringToMemory` instructions now exist. These have an extra
  `realloc` slot to store an intrinsic, if found.
* Our custom instructions are lowered to the standard instructions when
  generating an interface types section.
* The `realloc` function, if present, is passed as an argument like the
  malloc function when passing strings to wasm. If it's not present we
  use a slower fallback, but if it's present we use the faster
  implementation.

This should mean that there's little-to-no impact on existing users of
`wasm-bindgen`, but this should continue to still work for wasm
interface types polyfills and such. Additionally the GC passes now work
in that they don't delete `__wbindgen_realloc` which we later try to
reference.

* Add an empty test for the anyref pass

* Precisely track I32FromOptionAnyref's dependencies

This depends on the anyref table and a function to allocate an index if
the anyref pass is running, so be sure to track that in the instruction
itself for GC rooting.

* Trim extraneous exports from nop anyref module

Or if you're otherwise not using anyref slices, don't force some
intrinsics to exist.

* Remove globals from reference tests

Looks like these values adjust in slight but insignificant ways over
time

* Update the anyref xform tests

crates/cli-support/src/anyref.rs

@@ -1,3 +1,6 @@
+use crate::descriptor::VectorKind;
+use crate::intrinsic::Intrinsic;
+use crate::wit::AuxImport;
 use crate::wit::{AdapterKind, Instruction, NonstandardWitSection};
 use crate::wit::{AdapterType, InstructionData, StackChange, WasmBindgenAux};
 use anyhow::Error;
@@ -17,7 +20,7 @@ pub fn process(module: &mut Module) -> Result<(), Error> {
         .implements
         .iter()
         .cloned()
-        .map(|(core, adapter)| (adapter, core))
+        .map(|(core, _, adapter)| (adapter, core))
         .collect::<HashMap<_, _>>();
 
     // Transform all exported functions in the module, using the bindings listed
@@ -45,13 +48,73 @@ pub fn process(module: &mut Module) -> Result<(), Error> {
 
     let meta = cfg.run(module)?;
 
+    let mut aux = module
+        .customs
+        .delete_typed::<WasmBindgenAux>()
+        .expect("wit custom section should exist");
     let section = module
         .customs
-        .get_typed_mut::<WasmBindgenAux>()
+        .get_typed_mut::<NonstandardWitSection>()
         .expect("wit custom section should exist");
-    section.anyref_table = Some(meta.table);
-    section.anyref_alloc = meta.alloc;
-    section.anyref_drop_slice = meta.drop_slice;
+
+    // If the module looks like it's going to use some of these exports, store
+    // them in the aux section to get used.
+    //
+    // FIXME: this is not great, we should ideally have precise tracking of what
+    // requires what. These are used by catch clauses and anyref slices going
+    // in/out of wasm. The catch clauses are a bit weird but anyref slices
+    // should ideally track in their own instructions what table/functions
+    // they're referencing. This doesn't fit well in today's model of
+    // slice-related instructions, though, so let's just cop out and only enable
+    // these coarsely.
+    aux.anyref_table = Some(meta.table);
+    if module_needs_anyref_metadata(&aux, section) {
+        aux.anyref_alloc = meta.alloc;
+        aux.anyref_drop_slice = meta.drop_slice;
+    }
+
+    // Additonally we may need to update some adapter instructions other than
+    // those found for the anyref pass. These are some general "fringe support"
+    // things necessary to get absolutely everything working.
+    for (_, adapter) in section.adapters.iter_mut() {
+        let instrs = match &mut adapter.kind {
+            AdapterKind::Local { instructions } => instructions,
+            AdapterKind::Import { .. } => continue,
+        };
+        for instr in instrs {
+            match instr.instr {
+                // Calls to the heap live count intrinsic are now routed to the
+                // actual wasm function which keeps track of this.
+                Instruction::CallAdapter(adapter) => {
+                    let id = match meta.live_count {
+                        Some(id) => id,
+                        None => continue,
+                    };
+                    let import = match aux.import_map.get(&adapter) {
+                        Some(import) => import,
+                        None => continue,
+                    };
+                    match import {
+                        AuxImport::Intrinsic(Intrinsic::AnyrefHeapLiveCount) => {}
+                        _ => continue,
+                    }
+                    instr.instr = Instruction::Standard(wit_walrus::Instruction::CallCore(id));
+                }
+
+                // Optional anyref values are now managed in the wasm module, so
+                // we need to store where they're managed.
+                Instruction::I32FromOptionAnyref {
+                    ref mut table_and_alloc,
+                } => {
+                    *table_and_alloc = meta.alloc.map(|id| (meta.table, id));
+                }
+                _ => continue,
+            };
+        }
+    }
+
+    module.customs.add(*aux);
+
     Ok(())
 }
 
@@ -267,3 +330,55 @@ fn export_xform(cx: &mut Context, export: Export, instrs: &mut Vec<InstructionDa
         instrs.remove(idx);
     }
 }
+
+/// This function shouldn't need to exist, see the fixme at the call-site.
+fn module_needs_anyref_metadata(aux: &WasmBindgenAux, section: &NonstandardWitSection) -> bool {
+    use Instruction::*;
+
+    // our `handleError` intrinsic uses a few pieces of metadata to store
+    // indices directly into the wasm module.
+    if aux.imports_with_catch.len() > 0 {
+        return true;
+    }
+
+    // Look for any instructions which may use `VectorKind::Anyref`. If there
+    // are any then we'll need our intrinsics/tables/etc, otherwise we shouldn't
+    // ever need them.
+    section.adapters.iter().any(|(_, adapter)| {
+        let instructions = match &adapter.kind {
+            AdapterKind::Local { instructions } => instructions,
+            AdapterKind::Import { .. } => return false,
+        };
+        instructions.iter().any(|instr| match instr.instr {
+            VectorToMemory {
+                kind: VectorKind::Anyref,
+                ..
+            }
+            | MutableSliceToMemory {
+                kind: VectorKind::Anyref,
+                ..
+            }
+            | OptionVector {
+                kind: VectorKind::Anyref,
+                ..
+            }
+            | VectorLoad {
+                kind: VectorKind::Anyref,
+                ..
+            }
+            | OptionVectorLoad {
+                kind: VectorKind::Anyref,
+                ..
+            }
+            | View {
+                kind: VectorKind::Anyref,
+                ..
+            }
+            | OptionView {
+                kind: VectorKind::Anyref,
+                ..
+            } => true,
+            _ => false,
+        })
+    })
+}