Add wasm reference/pointers translation. (#1073)

cranelift-wasm/src/code_translator.rs

@@ -55,7 +55,7 @@ pub fn translate_operator<FE: FuncEnvironment + ?Sized>(
     environ: &mut FE,
 ) -> WasmResult<()> {
     if !state.reachable {
-        translate_unreachable_operator(module_translation_state, &op, builder, state)?;
+        translate_unreachable_operator(module_translation_state, &op, builder, state, environ)?;
         return Ok(());
     }
 
@@ -139,13 +139,13 @@ pub fn translate_operator<FE: FuncEnvironment + ?Sized>(
          ***********************************************************************************/
         Operator::Block { ty } => {
             let (params, results) = blocktype_params_results(module_translation_state, *ty)?;
-            let next = ebb_with_params(builder, results)?;
+            let next = ebb_with_params(builder, results, environ)?;
             state.push_block(next, params.len(), results.len());
         }
         Operator::Loop { ty } => {
             let (params, results) = blocktype_params_results(module_translation_state, *ty)?;
-            let loop_body = ebb_with_params(builder, params)?;
-            let next = ebb_with_params(builder, results)?;
+            let loop_body = ebb_with_params(builder, params, environ)?;
+            let next = ebb_with_params(builder, results, environ)?;
             builder.ins().jump(loop_body, state.peekn(params.len()));
             state.push_loop(loop_body, next, params.len(), results.len());
 
@@ -168,16 +168,16 @@ pub fn translate_operator<FE: FuncEnvironment + ?Sized>(
                 // destination ebb following the whole `if...end`. If we do end
                 // up discovering an `else`, then we will allocate an ebb for it
                 // and go back and patch the jump.
-                let destination = ebb_with_params(builder, results)?;
+                let destination = ebb_with_params(builder, results, environ)?;
                 let branch_inst = builder
                     .ins()
                     .brz(val, destination, state.peekn(params.len()));
                 (destination, ElseData::NoElse { branch_inst })
             } else {
                 // The `if` type signature is not valid without an `else` block,
                 // so we eagerly allocate the `else` block here.
-                let destination = ebb_with_params(builder, results)?;
-                let else_block = ebb_with_params(builder, params)?;
+                let destination = ebb_with_params(builder, results, environ)?;
+                let else_block = ebb_with_params(builder, params, environ)?;
                 builder
                     .ins()
                     .brz(val, else_block, state.peekn(params.len()));
@@ -229,7 +229,7 @@ pub fn translate_operator<FE: FuncEnvironment + ?Sized>(
                                 let (params, _results) =
                                     blocktype_params_results(module_translation_state, blocktype)?;
                                 debug_assert_eq!(params.len(), num_return_values);
-                                let else_ebb = ebb_with_params(builder, params)?;
+                                let else_ebb = ebb_with_params(builder, params, environ)?;
                                 builder.ins().jump(destination, state.peekn(params.len()));
                                 state.popn(params.len());
 
@@ -1352,11 +1352,12 @@ pub fn translate_operator<FE: FuncEnvironment + ?Sized>(
 /// Deals with a Wasm instruction located in an unreachable portion of the code. Most of them
 /// are dropped but special ones like `End` or `Else` signal the potential end of the unreachable
 /// portion so the translation state must be updated accordingly.
-fn translate_unreachable_operator(
+fn translate_unreachable_operator<FE: FuncEnvironment + ?Sized>(
     module_translation_state: &ModuleTranslationState,
     op: &Operator,
     builder: &mut FunctionBuilder,
     state: &mut FuncTranslationState,
+    environ: &mut FE,
 ) -> WasmResult<()> {
     debug_assert!(!state.reachable);
     match *op {
@@ -1397,7 +1398,7 @@ fn translate_unreachable_operator(
                             ElseData::NoElse { branch_inst } => {
                                 let (params, _results) =
                                     blocktype_params_results(module_translation_state, blocktype)?;
-                                let else_ebb = ebb_with_params(builder, params)?;
+                                let else_ebb = ebb_with_params(builder, params, environ)?;
 
                                 // We change the target of the branch instruction.
                                 builder.change_jump_destination(branch_inst, else_ebb);

cranelift-wasm/src/environ/dummy.rs

@@ -5,7 +5,9 @@
 //! [wasmtime-environ]: https://crates.io/crates/wasmtime-environ
 //! [Wasmtime]: https://github.com/bytecodealliance/wasmtime
 
-use crate::environ::{FuncEnvironment, GlobalVariable, ModuleEnvironment, ReturnMode, WasmResult};
+use crate::environ::{
+    FuncEnvironment, GlobalVariable, ModuleEnvironment, ReturnMode, TargetEnvironment, WasmResult,
+};
 use crate::func_translator::FuncTranslator;
 use crate::state::ModuleTranslationState;
 use crate::translation_utils::{
@@ -192,11 +194,13 @@ impl<'dummy_environment> DummyFuncEnvironment<'dummy_environment> {
     }
 }
 
-impl<'dummy_environment> FuncEnvironment for DummyFuncEnvironment<'dummy_environment> {
+impl<'dummy_environment> TargetEnvironment for DummyFuncEnvironment<'dummy_environment> {
     fn target_config(&self) -> TargetFrontendConfig {
         self.mod_info.config
     }
+}
 
+impl<'dummy_environment> FuncEnvironment for DummyFuncEnvironment<'dummy_environment> {
     fn return_mode(&self) -> ReturnMode {
         self.return_mode
     }
@@ -454,11 +458,13 @@ impl<'dummy_environment> FuncEnvironment for DummyFuncEnvironment<'dummy_environ
     }
 }
 
-impl<'data> ModuleEnvironment<'data> for DummyEnvironment {
+impl TargetEnvironment for DummyEnvironment {
     fn target_config(&self) -> TargetFrontendConfig {
         self.info.config
     }
+}
 
+impl<'data> ModuleEnvironment<'data> for DummyEnvironment {
     fn declare_signature(&mut self, sig: ir::Signature) -> WasmResult<()> {
         self.info.signatures.push(sig);
         Ok(())

cranelift-wasm/src/environ/mod.rs

@@ -6,5 +6,6 @@ mod spec;
 
 pub use crate::environ::dummy::DummyEnvironment;
 pub use crate::environ::spec::{
-    FuncEnvironment, GlobalVariable, ModuleEnvironment, ReturnMode, WasmError, WasmResult,
+    FuncEnvironment, GlobalVariable, ModuleEnvironment, ReturnMode, TargetEnvironment, WasmError,
+    WasmResult,
 };

cranelift-wasm/src/environ/spec.rs

@@ -103,12 +103,8 @@ pub enum ReturnMode {
     FallthroughReturn,
 }
 
-/// Environment affecting the translation of a single WebAssembly function.
-///
-/// A `FuncEnvironment` trait object is required to translate a WebAssembly function to Cranelift
-/// IR. The function environment provides information about the WebAssembly module as well as the
-/// runtime environment.
-pub trait FuncEnvironment {
+/// Environment affecting the translation of a WebAssembly.
+pub trait TargetEnvironment {
     /// Get the information needed to produce Cranelift IR for the given target.
     fn target_config(&self) -> TargetFrontendConfig;
 
@@ -124,13 +120,6 @@ pub trait FuncEnvironment {
         self.target_config().pointer_bytes()
     }
 
-    /// Should the code be structured to use a single `fallthrough_return` instruction at the end
-    /// of the function body, rather than `return` instructions as needed? This is used by VMs
-    /// to append custom epilogues.
-    fn return_mode(&self) -> ReturnMode {
-        ReturnMode::NormalReturns
-    }
-
     /// Get the Cranelift reference type to use for native references.
     ///
     /// This returns `R64` for 64-bit architectures and `R32` for 32-bit architectures.
@@ -141,6 +130,20 @@ pub trait FuncEnvironment {
             _ => panic!("unsupported pointer type"),
         }
     }
+}
+
+/// Environment affecting the translation of a single WebAssembly function.
+///
+/// A `FuncEnvironment` trait object is required to translate a WebAssembly function to Cranelift
+/// IR. The function environment provides information about the WebAssembly module as well as the
+/// runtime environment.
+pub trait FuncEnvironment: TargetEnvironment {
+    /// Should the code be structured to use a single `fallthrough_return` instruction at the end
+    /// of the function body, rather than `return` instructions as needed? This is used by VMs
+    /// to append custom epilogues.
+    fn return_mode(&self) -> ReturnMode {
+        ReturnMode::NormalReturns
+    }
 
     /// Set up the necessary preamble definitions in `func` to access the global variable
     /// identified by `index`.
@@ -384,10 +387,7 @@ pub trait FuncEnvironment {
 /// An object satisfying the `ModuleEnvironment` trait can be passed as argument to the
 /// [`translate_module`](fn.translate_module.html) function. These methods should not be called
 /// by the user, they are only for `cranelift-wasm` internal use.
-pub trait ModuleEnvironment<'data> {
-    /// Get the information needed to produce Cranelift IR for the current target.
-    fn target_config(&self) -> TargetFrontendConfig;
-
+pub trait ModuleEnvironment<'data>: TargetEnvironment {
     /// Provides the number of signatures up front. By default this does nothing, but
     /// implementations can use this to preallocate memory if desired.
     fn reserve_signatures(&mut self, _num: u32) -> WasmResult<()> {

cranelift-wasm/src/func_translator.rs

@@ -193,6 +193,7 @@ fn declare_locals<FE: FuncEnvironment + ?Sized>(
             builder.ins().vconst(ir::types::I8X16, constant_handle)
         }
         AnyRef => builder.ins().null(environ.reference_type()),
+        AnyFunc => builder.ins().null(environ.reference_type()),
         ty => return Err(wasm_unsupported!("unsupported local type {:?}", ty)),
     };
 

cranelift-wasm/src/lib.rs

@@ -58,8 +58,8 @@ mod state;
 mod translation_utils;
 
 pub use crate::environ::{
-    DummyEnvironment, FuncEnvironment, GlobalVariable, ModuleEnvironment, ReturnMode, WasmError,
-    WasmResult,
+    DummyEnvironment, FuncEnvironment, GlobalVariable, ModuleEnvironment, ReturnMode,
+    TargetEnvironment, WasmError, WasmResult,
 };
 pub use crate::func_translator::FuncTranslator;
 pub use crate::module_translator::translate_module;

cranelift-wasm/src/sections_translator.rs

@@ -44,14 +44,15 @@ pub fn parse_type_section(
                 params,
                 returns,
             } => {
-                let mut sig = Signature::new(environ.target_config().default_call_conv);
+                let mut sig =
+                    Signature::new(ModuleEnvironment::target_config(environ).default_call_conv);
                 sig.params.extend(params.iter().map(|ty| {
-                    let cret_arg: ir::Type = type_to_type(*ty)
+                    let cret_arg: ir::Type = type_to_type(*ty, environ)
                         .expect("only numeric types are supported in function signatures");
                     AbiParam::new(cret_arg)
                 }));
                 sig.returns.extend(returns.iter().map(|ty| {
-                    let cret_arg: ir::Type = type_to_type(*ty)
+                    let cret_arg: ir::Type = type_to_type(*ty, environ)
                         .expect("only numeric types are supported in function signatures");
                     AbiParam::new(cret_arg)
                 }));
@@ -106,7 +107,7 @@ pub fn parse_import_section<'data>(
             ImportSectionEntryType::Global(ref ty) => {
                 environ.declare_global_import(
                     Global {
-                        ty: type_to_type(ty.content_type).unwrap(),
+                        ty: type_to_type(ty.content_type, environ).unwrap(),
                         mutability: ty.mutable,
                         initializer: GlobalInit::Import,
                     },
@@ -117,7 +118,7 @@ pub fn parse_import_section<'data>(
             ImportSectionEntryType::Table(ref tab) => {
                 environ.declare_table_import(
                     Table {
-                        ty: match tabletype_to_type(tab.element_type)? {
+                        ty: match tabletype_to_type(tab.element_type, environ)? {
                             Some(t) => TableElementType::Val(t),
                             None => TableElementType::Func,
                         },
@@ -160,7 +161,7 @@ pub fn parse_table_section(
     for entry in tables {
         let table = entry?;
         environ.declare_table(Table {
-            ty: match tabletype_to_type(table.element_type)? {
+            ty: match tabletype_to_type(table.element_type, environ)? {
                 Some(t) => TableElementType::Val(t),
                 None => TableElementType::Func,
             },
@@ -215,6 +216,7 @@ pub fn parse_global_section(
             Operator::V128Const { value } => {
                 GlobalInit::V128Const(V128Imm::from(value.bytes().to_vec().as_slice()))
             }
+            Operator::RefNull => GlobalInit::RefNullConst,
             Operator::GetGlobal { global_index } => {
                 GlobalInit::GetGlobal(GlobalIndex::from_u32(global_index))
             }
@@ -226,7 +228,7 @@ pub fn parse_global_section(
             }
         };
         let global = Global {
-            ty: type_to_type(content_type).unwrap(),
+            ty: type_to_type(content_type, environ).unwrap(),
             mutability: mutable,
             initializer,
         };

cranelift-wasm/src/translation_utils.rs

@@ -1,5 +1,5 @@
 //! Helper functions and structures for the translation.
-use crate::environ::WasmResult;
+use crate::environ::{TargetEnvironment, WasmResult};
 use crate::state::ModuleTranslationState;
 use crate::wasm_unsupported;
 use core::u32;
@@ -83,6 +83,8 @@ pub enum GlobalInit {
     V128Const(V128Imm),
     /// A `get_global` of another global.
     GetGlobal(GlobalIndex),
+    /// A `ref.null`.
+    RefNullConst,
     ///< The global is imported from, and thus initialized by, a different module.
     Import,
 }
@@ -119,26 +121,34 @@ pub struct Memory {
 }
 
 /// Helper function translating wasmparser types to Cranelift types when possible.
-pub fn type_to_type(ty: wasmparser::Type) -> WasmResult<ir::Type> {
+pub fn type_to_type<PE: TargetEnvironment + ?Sized>(
+    ty: wasmparser::Type,
+    environ: &PE,
+) -> WasmResult<ir::Type> {
     match ty {
         wasmparser::Type::I32 => Ok(ir::types::I32),
         wasmparser::Type::I64 => Ok(ir::types::I64),
         wasmparser::Type::F32 => Ok(ir::types::F32),
         wasmparser::Type::F64 => Ok(ir::types::F64),
         wasmparser::Type::V128 => Ok(ir::types::I8X16),
+        wasmparser::Type::AnyRef | wasmparser::Type::AnyFunc => Ok(environ.reference_type()),
         ty => Err(wasm_unsupported!("type_to_type: wasm type {:?}", ty)),
     }
 }
 
 /// Helper function translating wasmparser possible table types to Cranelift types when possible,
 /// or None for Func tables.
-pub fn tabletype_to_type(ty: wasmparser::Type) -> WasmResult<Option<ir::Type>> {
+pub fn tabletype_to_type<PE: TargetEnvironment + ?Sized>(
+    ty: wasmparser::Type,
+    environ: &PE,
+) -> WasmResult<Option<ir::Type>> {
     match ty {
         wasmparser::Type::I32 => Ok(Some(ir::types::I32)),
         wasmparser::Type::I64 => Ok(Some(ir::types::I64)),
         wasmparser::Type::F32 => Ok(Some(ir::types::F32)),
         wasmparser::Type::F64 => Ok(Some(ir::types::F64)),
         wasmparser::Type::V128 => Ok(Some(ir::types::I8X16)),
+        wasmparser::Type::AnyRef => Ok(Some(environ.reference_type())),
         wasmparser::Type::AnyFunc => Ok(None),
         ty => Err(wasm_unsupported!(
             "tabletype_to_type: table wasm type {:?}",
@@ -159,6 +169,8 @@ pub fn blocktype_params_results(
             wasmparser::Type::F32 => (&[], &[wasmparser::Type::F32]),
             wasmparser::Type::F64 => (&[], &[wasmparser::Type::F64]),
             wasmparser::Type::V128 => (&[], &[wasmparser::Type::V128]),
+            wasmparser::Type::AnyRef => (&[], &[wasmparser::Type::AnyRef]),
+            wasmparser::Type::AnyFunc => (&[], &[wasmparser::Type::AnyFunc]),
             wasmparser::Type::EmptyBlockType => (&[], &[]),
             ty => return Err(wasm_unsupported!("blocktype_params_results: type {:?}", ty)),
         },
@@ -171,9 +183,10 @@ pub fn blocktype_params_results(
 }
 
 /// Create an `Ebb` with the given Wasm parameters.
-pub fn ebb_with_params(
+pub fn ebb_with_params<PE: TargetEnvironment + ?Sized>(
     builder: &mut FunctionBuilder,
     params: &[wasmparser::Type],
+    environ: &PE,
 ) -> WasmResult<ir::Ebb> {
     let ebb = builder.create_ebb();
     for ty in params.iter() {
@@ -190,6 +203,9 @@ pub fn ebb_with_params(
             wasmparser::Type::F64 => {
                 builder.append_ebb_param(ebb, ir::types::F64);
             }
+            wasmparser::Type::AnyRef | wasmparser::Type::AnyFunc => {
+                builder.append_ebb_param(ebb, environ.reference_type());
+            }
             wasmparser::Type::V128 => {
                 builder.append_ebb_param(ebb, ir::types::I8X16);
             }

src/clif-util.rs

@@ -119,6 +119,12 @@ fn add_enable_multi_value<'a>() -> clap::Arg<'a, 'a> {
         .help("Enable WASM's multi-value support")
 }
 
+fn add_enable_reference_types_flag<'a>() -> clap::Arg<'a, 'a> {
+    Arg::with_name("enable-reference-types")
+        .long("enable-reference-types")
+        .help("Enable WASM's reference types operations")
+}
+
 fn add_just_decode_flag<'a>() -> clap::Arg<'a, 'a> {
     Arg::with_name("just-decode")
         .short("t")
@@ -163,6 +169,7 @@ fn add_wasm_or_compile<'a>(cmd: &str) -> clap::App<'a, 'a> {
         .arg(add_debug_flag())
         .arg(add_enable_simd_flag())
         .arg(add_enable_multi_value())
+        .arg(add_enable_reference_types_flag())
         .arg(add_just_decode_flag())
         .arg(add_check_translation_flag())
 }
@@ -316,6 +323,7 @@ fn main() {
                     rest_cmd.is_present("value-ranges"),
                     rest_cmd.is_present("enable-simd"),
                     rest_cmd.is_present("enable-multi-value"),
+                    rest_cmd.is_present("enable-reference-types"),
                 )
             };