struct field reordering (minimizing padding)

2026-08-09 14:12:25 +02:00
parent 729488e702
commit 6688822de2
4 changed files with 111 additions and 15 deletions
@@ -40,7 +40,7 @@ roadmap and milestone history.
 - narrow immutable zero-terminated byte pointer/slice conversion to `*c_char` / `?*c_char` without general `u8`/`c_char` interchange
 - optionals with `null`, `orelse`, postfix `?`, conditional unwraps, guarded unwraps, and left-to-right short-circuiting multi-unwraps
 - nominal distinct types with explicit scalar backing conversion during construction and explicit scalar backing extraction, native enums with optional explicit integer backing and explicit backing-to-scalar casts, contextual enum literals, and imported C enums as target-backed integer aliases
- source-order native structs, opaque nominal records with `Name :: opaque`, complete `c_struct { ... }`, keyed record literals, native untagged unions, and native tagged unions `union(Enum)` / `union(enum)`
+- compiler-reordered native structs with fields laid out by decreasing alignment (declaration order breaks ties and remains the reflection/diagnostic order), opaque nominal records with `Name :: opaque`, complete source-order `c_struct { ... }`, keyed record literals, native untagged unions, and native tagged unions `union(Enum)` / `union(enum)`
 - named native struct fields may declare defaults with `field T = expression`; keyed literals use defaults for omitted fields and explicit initializers override them
 - void-payload tagged-union variants, anonymous struct payloads, contextual `.variant`, `.variant{payload}`, and `.variant{field = value}` construction
 - native sum composition with `A | B` for unbacked enums and tagged unions, optionally grouped as `(A | B)`, using program-global `u16` variant ids
@@ -1117,6 +1117,15 @@ emit_instruction_stream :: proc(
 				} else if item.kind == .Range && arg_index == 2 {
 					element_type = types.BOOL
 				}
+				aggregate_index := arg_index
+				if item.kind == .Struct {
+					aggregate_index = types.physical_field_index(
+						&emitter.module.types,
+						instruction.type,
+						arg_index,
+						emitter.module.target,
+					)
+				}
 				final := arg_index == total-1
 				if final {
 					fmt.sbprintf(&emitter.builder, "  %%v%d = insertvalue %s ", instruction_index, type_name)
@@ -1134,7 +1143,7 @@ emit_instruction_stream :: proc(
 				} else {
 					write_constant(&emitter.builder, i64(item.sentinel), element_type, &emitter.module.types)
 				}
-				fmt.sbprintf(&emitter.builder, ", %d\n", arg_index)
+				fmt.sbprintf(&emitter.builder, ", %d\n", aggregate_index)
 			}
 		case .Null:
 			item, ok := types.node(&emitter.module.types, instruction.type)
@@ -1334,13 +1343,19 @@ emit_instruction_stream :: proc(
 				emit_recovery_value(emitter, instruction_index, instruction, "invalid field reference")
 				continue
 			}
+			physical_index := types.physical_field_index(
+				&emitter.module.types,
+				base_type,
+				field_index,
+				emitter.module.target,
+			)
 			if types.is_union(base_type, &emitter.module.types) {
 				fmt.sbprintf(&emitter.builder, "  %%v%d = getelementptr i8, ptr %%v%d, i64 %d\n", instruction_index, instruction.a, types.union_payload_offset(base_type, &emitter.module.types, emitter.module.target))
 			} else {
 				fmt.sbprintf(
 					&emitter.builder,
 					"  %%v%d = getelementptr %s, ptr %%v%d, i32 0, i32 %d\n",
-					instruction_index, llvm_type(base_type, &emitter.module.types), instruction.a, field_index,
+					instruction_index, llvm_type(base_type, &emitter.module.types), instruction.a, physical_index,
 				)
 			}
 		case .Load:
@@ -2593,12 +2608,30 @@ emit_types :: proc(emitter: ^Emitter) {
 			strings.write_string(&emitter.builder, " }\n")
 			continue
 		}
+		fields := types.fields_for(&emitter.module.types, id)
 		strings.write_string(&emitter.builder, "{ ")
-		for field, field_index in types.fields_for(&emitter.module.types, id) {
-			if field_index > 0 {
+		previous_index := -1
+		for physical_index in 0..<len(fields) {
+			logical_index := -1
+			for _, candidate_index in fields {
+				if previous_index >= 0 &&
+				   !types.field_layout_precedes(
+						&emitter.module.types, id, previous_index, candidate_index, emitter.module.target,
+				   ) {
+					continue
+				}
+				if logical_index < 0 ||
+				   types.field_layout_precedes(
+						&emitter.module.types, id, candidate_index, logical_index, emitter.module.target,
+				   ) {
+					logical_index = candidate_index
+				}
+			}
+			if physical_index > 0 {
 				strings.write_string(&emitter.builder, ", ")
 			}
-			strings.write_string(&emitter.builder, llvm_type(field.type, &emitter.module.types))
+			strings.write_string(&emitter.builder, llvm_type(fields[logical_index].type, &emitter.module.types))
+			previous_index = logical_index
 		}
 		strings.write_string(&emitter.builder, " }\n")
 	}
@@ -636,6 +636,43 @@ fields_for :: proc(store: ^Store, value: Type) -> []Field {
 	return store.fields[start:end]
 }

+field_layout_precedes :: proc(
+	store: ^Store,
+	value: Type,
+	left, right: int,
+	selected := target.DEFAULT,
+) -> bool {
+	fields := fields_for(store, value)
+	item, ok := node(store, value)
+	if !ok || item.kind != .Struct || item.c_layout ||
+	   left < 0 || left >= len(fields) || right < 0 || right >= len(fields) {
+		return left < right
+	}
+	left_alignment := alignment_of(fields[left].type, store, selected)
+	right_alignment := alignment_of(fields[right].type, store, selected)
+	return left_alignment > right_alignment ||
+	       (left_alignment == right_alignment && left < right)
+}
+
+physical_field_index :: proc(
+	store: ^Store,
+	value: Type,
+	logical_index: int,
+	selected := target.DEFAULT,
+) -> int {
+	fields := fields_for(store, value)
+	if logical_index < 0 || logical_index >= len(fields) {
+		return logical_index
+	}
+	result := 0
+	for _, other_index in fields {
+		if field_layout_precedes(store, value, other_index, logical_index, selected) {
+			result += 1
+		}
+	}
+	return result
+}
+
 params_for :: proc(store: ^Store, value: Type) -> []Field {
 	item, ok := node(store, value)
 	if !ok || item.kind != .Function {
@@ -1709,12 +1746,19 @@ size :: proc(value: Type, store: ^Store, selected := target.DEFAULT) -> u64 {
 		}
 		offset: u64
 		max_align: u64 = 1
+		if !item.c_layout {
+			for field in fields_for(store, value) {
+				offset += size(field.type, store, selected)
+				max_align = max(max_align, u64(alignment_of(field.type, store, selected)))
+			}
+		} else {
 			for field in fields_for(store, value) {
 				field_align := u64(alignment_of(field.type, store, selected))
 				offset = (offset+field_align-1)/field_align*field_align
 				offset += size(field.type, store, selected)
 				max_align = max(max_align, field_align)
 			}
+		}
 		return (offset+max_align-1)/max_align*max_align
 	case .Union:
 		item, _ := node(store, value)
@@ -1887,6 +1887,18 @@ Bool_Last :: struct {
 	flag bool
 }

+Optimized :: struct {
+	prefix u8
+	wide u64
+	suffix u8
+}
+
+Source_Order_C :: c_struct {
+	prefix c_uchar
+	wide c_ulong
+	suffix c_uchar
+}
+
 Opaque :: opaque
 Color :: enum {
 	red
@@ -1937,12 +1949,19 @@ main func() i32 {
 	if (sizeof!(bool) != 1) return 15
 	if (sizeof!(Bool_First) != 2) return 16
 	if (sizeof!(Bool_Last) != 2) return 17
+	optimized Optimized := Optimized{prefix = 40, wide = 72623859790382856, suffix = 41}
+	if (sizeof!(Optimized) != 16) return 18
+	if (alignof!(Optimized) != 8) return 19
+	if (sizeof!(Source_Order_C) != 24) return 20
+	if (optimized.prefix != 40) return 21
+	if (optimized.wide != 72623859790382856) return 22
+	if (optimized.suffix != 41) return 23
 	first :: make_bool_first()
-	if (!first.flag) return 18
-	if (first.byte != 41) return 19
+	if (!first.flag) return 24
+	if (first.byte != 41) return 25
 	last :: make_bool_last()
-	if (!last.flag) return 20
-	if (last.byte != 42) return 21
+	if (!last.flag) return 26
+	if (last.byte != 42) return 27
 	return 0
 }
 `