WIP: Generate an IR.

The compiler has come a long way, but we still can't do int64, because it requires a non-cell storage size. There's no (sane) way to express conversions between int32 and int64 within the AST, because we have no uniform way of inserting conversion nodes. This is already a deep problem that has been hacked around for operator overloads and property accessors, and it doesn't scale. The solution is obvious: transform the AST into an IR. That's what we should have done from the beginning but didn't. Unfortunately it requires a *lot* of refactoring and a ton of boilerplate. So far, I have most of the boilerplate done, but the refactoring is only halfway there. CodeGenerator has not been ported to the IR yet. Once this gigantic patch is done, we'll have the following changes: - `struct value` will be eliminated, and good riddance. - The AST will be immutable after parsing. - The semantic analysis phase will output a new IR tree. - CodeGenerator will generate off the IR instead. Since the IR is a transformation of the AST, I'm expecting minimal changes to the end result. - functag_t will be replaced by FunctionType. V2: CG-IR can now assemble trivial programs. V3: CG-IR supports basic calls; 341 test failures. V4: CG-IR supports binary ops; 333 test failures. V5: CG-IR supports do-while and if; 329 test failures. V6: CG-IR supports args, local incdec, switch; 319 test failures.
alliedmodders · Aug 9, 2024 · 3894d96 · 3894d96
1 parent fd82555
commit 3894d96
Show file tree

Hide file tree

Showing 34 changed files with 2,754 additions and 2,013 deletions.
diff --git a/compiler/AMBuilder b/compiler/AMBuilder
@@ -12,6 +12,7 @@ module.sources += [
   'data-queue.cpp',
   'errors.cpp',
   'expressions.cpp',
+  'ir.cpp',
   'lexer.cpp',
   'main.cpp',
   'name-resolution.cpp',

diff --git a/compiler/array-helpers.cpp b/compiler/array-helpers.cpp
@@ -822,7 +822,7 @@ bool Semantics::CheckArrayDeclaration(VarDeclBase* decl) {
 class CompoundEmitter final
 {
   public:
-    CompoundEmitter(Type* type, Expr* init)
+    CompoundEmitter(QualType type, Expr* init)
       : type_(type),
         init_(init),
         pending_zeroes_(0)
@@ -852,11 +852,11 @@ class CompoundEmitter final
     size_t AddString(StringExpr* expr);
     void AddInlineArray(LayoutFieldDecl* field, ArrayExpr* expr);
     void AddInlineEnumStruct(EnumStructDecl* es, ArrayExpr* expr);
-    void EmitPadding(size_t rank_size, Type* type, size_t emitted, bool ellipses,
+    void EmitPadding(size_t rank_size, QualType type, size_t emitted, bool ellipses,
                      const ke::Maybe<cell> prev1, const ke::Maybe<cell> prev2);
 
   private:
-    Type* type_;
+    QualType type_;
     Expr* init_;
     tr::vector<cell> iv_;
     tr::vector<cell> data_;
@@ -948,7 +948,7 @@ cell CompoundEmitter::Emit(ArrayType* rank, Expr* init) {
     // This only works because enum structs are flattened and don't support
     // internal IVs. No plans to change this as it would greatly increase
     // complexity unless we radically changed arrays.
-    EmitPadding(rank->size(), rank->inner(), emitted, ellipses, prev1, prev2);
+    EmitPadding(rank->size(), QualType(rank->inner()), emitted, ellipses, prev1, prev2);
 
     return (start * sizeof(cell)) | kDataFlag;
 }
@@ -967,7 +967,7 @@ void CompoundEmitter::AddInlineEnumStruct(EnumStructDecl* es, ArrayExpr* array)
             assert(field);
 
             auto rank_type = field->type()->to<ArrayType>();
-            EmitPadding(rank_type->size(), rank_type->inner(), emitted, false, {}, {});
+            EmitPadding(rank_type->size(), QualType(rank_type->inner()), emitted, false, {}, {});
         } else if (ArrayExpr* expr = item->as<ArrayExpr>()) {
             // Subarrays can only appear in an enum struct. Normal 2D cases
             // would flow through the check at the start of this function.
@@ -994,12 +994,12 @@ void CompoundEmitter::AddInlineArray(LayoutFieldDecl* field, ArrayExpr* array) {
     }
 
     auto rank_size = field->type()->to<ArrayType>()->size();
-    EmitPadding(rank_size, field->type(), array->exprs().size(),
+    EmitPadding(rank_size, QualType(field->type()), array->exprs().size(),
                 array->ellipses(), prev1, prev2);
 }
 
 void
-CompoundEmitter::EmitPadding(size_t rank_size, Type* type, size_t emitted, bool ellipses,
+CompoundEmitter::EmitPadding(size_t rank_size, QualType type, size_t emitted, bool ellipses,
                              const ke::Maybe<cell> prev1, const ke::Maybe<cell> prev2)
 {
     // Pad remainder to zeroes if the array was explicitly sized.
@@ -1053,20 +1053,20 @@ CompoundEmitter::add_data(cell value)
     data_.emplace_back(value);
 }
 
-void BuildCompoundInitializer(Type* type, Expr* init, ArrayData* array) {
+void BuildCompoundInitializer(QualType type, Expr* init, ArrayData* array,
+                              std::optional<cell_t> base_address)
+{
     CompoundEmitter emitter(type, init);
     emitter.Emit();
 
     array->iv = std::move(emitter.iv());
     array->data = std::move(emitter.data());
     array->zeroes = emitter.pending_zeroes();
-}
 
-void BuildCompoundInitializer(VarDeclBase* decl, ArrayData* array, cell base_address) {
-    BuildCompoundInitializer(decl->type(), decl->init_rhs(), array);
-
-    for (auto& v : array->iv)
-        v += base_address;
+    if (base_address) {
+        for (auto& v : array->iv)
+            v += *base_address;
+    }
 }
 
 } // namespace cc

diff --git a/compiler/array-helpers.h b/compiler/array-helpers.h
@@ -38,8 +38,8 @@ bool ResolveArrayType(Semantics* sema, const token_pos_t& pos, typeinfo_t* type,
 // Perform type and size checks of an array and its initializer if present.
 bool CheckArrayInitialization(Semantics* sema, const typeinfo_t& type, Expr* init);
 
-void BuildCompoundInitializer(VarDeclBase* decl, ArrayData* array, cell_t base_addr);
-void BuildCompoundInitializer(Type* type, Expr* init, ArrayData* array);
+void BuildCompoundInitializer(QualType type, Expr* init, ArrayData* array,
+                              std::optional<cell_t> base_address = {});
 
 cell_t CalcArraySize(Type* type);
 

diff --git a/compiler/assembler.cpp b/compiler/assembler.cpp
@@ -65,7 +65,7 @@ struct function_entry {
     function_entry(const function_entry& other) = delete;
     function_entry& operator =(const function_entry& other) = delete;
 
-    FunctionDecl* decl = nullptr;
+    ir::Function* fun;
     std::string name;
 };
 
@@ -123,7 +123,7 @@ class RttiBuilder
     RttiBuilder(CompileContext& cc, CodeGenerator& cg, SmxNameTable* names);
 
     void finish(SmxBuilder& builder);
-    void add_method(FunctionDecl* fun);
+    void add_method(ir::Function* fun);
     void add_native(FunctionDecl* sym);
 
   private:
@@ -360,17 +360,19 @@ RttiBuilder::add_debug_var(SmxRttiTable<smx_rtti_debug_var>* table, DebugString&
     var.type_id = type_id;
 }
 
-void RttiBuilder::add_method(FunctionDecl* fun) {
+void RttiBuilder::add_method(ir::Function* fun) {
     assert(fun->is_live());
 
     uint32_t index = methods_->count();
     smx_rtti_method& method = methods_->add();
-    method.name = names_->add(fun->name());
-    method.pcode_start = fun->cg()->label.offset();
-    method.pcode_end = fun->cg()->pcode_end;
-    method.signature = encode_signature(fun->canonical());
-
-    if (!fun->cg()->dbgstrs)
+    method.name = names_->add(fun->decl()->name());
+    method.pcode_start = fun->label().offset();
+    method.pcode_end = fun->pcode_end();
+    method.signature = encode_signature(fun->decl()->canonical());
+
+    (void)index;
+#if 0
+    if (!fun->dbgstrs)
         return;
 
     smx_rtti_debug_method debug;
@@ -392,6 +394,7 @@ void RttiBuilder::add_method(FunctionDecl* fun) {
     // Only add a method table entry if we actually had locals.
     if (debug.first_local != dbg_locals_->count())
         dbg_methods_->add(debug);
+#endif
 }
 
 void RttiBuilder::add_native(FunctionDecl* fun) {
@@ -721,59 +724,44 @@ Assembler::Assemble(SmxByteBuffer* buffer)
     std::vector<function_entry> functions;
     std::unordered_set<Decl*> symbols;
 
-    // Sort globals.
-    std::vector<Decl*> global_symbols;
-    cc_.globals()->ForEachSymbol([&](Decl* decl) -> void {
-        global_symbols.push_back(decl);
+    auto mod = cg_.mod();
 
-        // This is only to assert that we embedded pointers properly in the assembly buffer.
-        symbols.emplace(decl);
+    // Sort globals.
+    std::sort(mod->functions().begin(), mod->functions().end(),
+              [](const ir::Function* a, const ir::Function* b) {
+        return a->decl()->name()->str() < b->decl()->name()->str();
     });
-    for (const auto& decl : cc_.functions()) {
-        if (symbols.count(decl))
-            continue;
-        if (decl->canonical() != decl)
+
+    for (const auto& fun : mod->functions()) {
+        auto decl = fun->decl();
+
+        if (decl->is_native() || !fun->body())
             continue;
-        global_symbols.push_back(decl);
-        symbols.emplace(decl);
-    }
 
-    std::sort(global_symbols.begin(), global_symbols.end(),
-              [](const Decl* a, const Decl *b) -> bool {
-        return a->name()->str() < b->name()->str();
-    });
+        function_entry entry;
+        entry.fun = fun;
+        if (decl->is_public()) {
+            entry.name = decl->name()->str();
+        } else {
+            // Create a private name.
+            entry.name = ke::StringPrintf(".%d.%s", fun->label().offset(), decl->name()->chars());
+        }
+
+        functions.emplace_back(std::move(entry));
+    }
 
+#if 0
     // Build the easy symbol tables.
     for (const auto& decl : global_symbols) {
-        if (auto fun = decl->as<FunctionDecl>()) {
-            if (fun->is_native())
-                continue;
-
-            if (!fun->body())
-                continue;
-            if (!fun->is_live())
-                continue;
-            if (fun->canonical() != fun)
-                continue;
-
-            function_entry entry;
-            entry.decl = fun;
-            if (fun->is_public()) {
-                entry.name = fun->name()->str();
-            } else {
-                // Create a private name.
-                entry.name = ke::StringPrintf(".%d.%s", fun->cg()->label.offset(), fun->name()->chars());
-            }
-
-            functions.emplace_back(std::move(entry));
-        } else if (auto var = decl->as<VarDecl>()) {
+        if (auto var = decl->as<VarDecl>()) {
             if (var->is_public() || (var->is_used() && !var->as<ConstDecl>())) {
                 sp_file_pubvars_t& pubvar = pubvars->add();
                 pubvar.address = var->addr();
                 pubvar.name = names->add(var->name());
             }
         }
     }
+#endif
 
     // The public list must be sorted.
     std::sort(functions.begin(), functions.end(),
@@ -783,31 +771,27 @@ Assembler::Assemble(SmxByteBuffer* buffer)
     for (size_t i = 0; i < functions.size(); i++) {
         function_entry& f = functions[i];
 
-        assert(f.decl->cg()->label.offset() > 0);
-        assert(f.decl->impl());
-        assert(f.decl->cg()->pcode_end > f.decl->cg()->label.offset());
-
         sp_file_publics_t& pubfunc = publics->add();
-        pubfunc.address = f.decl->cg()->label.offset();
+        pubfunc.address = f.fun->label().offset();
         pubfunc.name = names->add(*cc_.atoms(), f.name.c_str());
 
         auto id = (uint32_t(i) << 1) | 1;
         if (!Label::ValueFits(id))
             report(421);
-        cg_.LinkPublicFunction(f.decl, id);
+        cg_.LinkPublicFunction(f.fun, id);
 
-        rtti.add_method(f.decl);
+        rtti.add_method(f.fun);
     }
 
     // Populate the native table.
     for (size_t i = 0; i < cg_.native_list().size(); i++) {
-        FunctionDecl* sym = cg_.native_list()[i];
-        assert(size_t(sym->cg()->label.offset()) == i);
+        ir::Function* sym = cg_.native_list()[i];
+        assert(size_t(sym->label().offset()) == i);
 
         sp_file_natives_t& entry = natives->add();
-        entry.name = names->add(sym->name());
+        entry.name = names->add(sym->decl()->name());
 
-        rtti.add_native(sym);
+        rtti.add_native(sym->decl());
     }
 
     // Set up the code section.

diff --git a/compiler/assembler.h b/compiler/assembler.h
@@ -26,6 +26,7 @@
 #include "libsmx/data-pool.h"
 #include "libsmx/smx-builder.h"
 #include "libsmx/smx-encoding.h"
+#include "ir.h"
 #include "sc.h"
 #include "shared/byte-buffer.h"
 #include "shared/string-pool.h"

diff --git a/compiler/ast-types.h b/compiler/ast-types.h
@@ -19,6 +19,8 @@
 //  3.  This notice may not be removed or altered from any source distribution.
 #pragma once
 
+#include <stdint.h>
+
 #define AST_STMT_TYPE_LIST(FOR_EACH) \
     FOR_EACH(StmtList) \
     FOR_EACH(BlockStmt) \
@@ -79,6 +81,48 @@
     FOR_EACH(StructExpr) \
     FOR_EACH(StructInitFieldExpr)
 
+#define IR_NODE_TYPE_LIST(FOR_EACH) \
+    /* Decls */ \
+    FOR_EACH(Function) \
+    FOR_EACH(Variable) \
+    /* Statements */ \
+    FOR_EACH(Return) \
+    FOR_EACH(ValueInsn) \
+    FOR_EACH(Exit) \
+    FOR_EACH(Break) \
+    FOR_EACH(Continue) \
+    FOR_EACH(Assert) \
+    FOR_EACH(If) \
+    FOR_EACH(DoWhile) \
+    FOR_EACH(Delete) \
+    FOR_EACH(ForLoop) \
+    FOR_EACH(Switch) \
+    FOR_EACH(FunctionDef) \
+    /* Values */ \
+    FOR_EACH(ConstVal) \
+    FOR_EACH(CharArrayLiteral) \
+    FOR_EACH(VariableRef) \
+    FOR_EACH(TypeRef) \
+    FOR_EACH(FunctionRef) \
+    FOR_EACH(IndexOp) \
+    FOR_EACH(Load) \
+    FOR_EACH(TernaryOp) \
+    FOR_EACH(BinaryOp) \
+    FOR_EACH(Array) \
+    FOR_EACH(CommaOp) \
+    FOR_EACH(CallOp) \
+    FOR_EACH(TempRef) \
+    FOR_EACH(PropertyRef) \
+    FOR_EACH(FieldRef) \
+    FOR_EACH(UnaryOp) \
+    FOR_EACH(CallUserOp) \
+    FOR_EACH(IncDecOp) \
+    FOR_EACH(Store) \
+    FOR_EACH(ThisRef)
+
+namespace sp {
+namespace cc {
+
 enum class ExprKind : uint8_t
 {
 #define _(Name) Name,
@@ -92,3 +136,13 @@ enum class StmtKind : uint8_t
     AST_STMT_TYPE_LIST(_)
 #undef _
 };
+
+enum class IrKind : uint8_t
+{
+#define _(Name) Name,
+    IR_NODE_TYPE_LIST(_)
+#undef _
+};
+
+} // namespace cc
+} // namespace sp