stage2: improve LLVM backend for enums

* support lowering enum types and constants to LLVM IR
 * fix cmp instruction to support enum operands
This commit is contained in:
Andrew Kelley 2021-09-16 21:57:46 -07:00
parent 091a98f524
commit b58d8aa05f
3 changed files with 59 additions and 32 deletions

View File

@ -588,6 +588,12 @@ pub const DeclGen = struct {
const info = t.intInfo(self.module.getTarget());
return self.context.intType(info.bits);
},
.Enum => {
var buffer: Type.Payload.Bits = undefined;
const int_ty = t.enumTagType(&buffer);
const bit_count = int_ty.intInfo(self.module.getTarget()).bits;
return self.context.intType(bit_count);
},
.Float => switch (t.floatBits(self.module.getTarget())) {
16 => return self.context.halfType(),
32 => return self.context.floatType(),
@ -686,7 +692,6 @@ pub const DeclGen = struct {
.BoundFn => @panic("TODO remove BoundFn from the language"),
.Enum,
.Union,
.Opaque,
.Frame,
@ -723,6 +728,17 @@ pub const DeclGen = struct {
}
return llvm_int;
},
.Enum => {
const llvm_type = try self.llvmType(tv.ty);
const uint: u64 = uint: {
if (tv.val.castTag(.enum_field_index)) |payload| {
break :uint payload.data;
}
break :uint tv.val.toUnsignedInt();
};
const llvm_int = llvm_type.constInt(uint, .False);
return llvm_int;
},
.Float => {
if (tv.ty.floatBits(self.module.getTarget()) <= 64) {
const llvm_ty = try self.llvmType(tv.ty);
@ -907,7 +923,18 @@ pub const DeclGen = struct {
.ComptimeFloat => unreachable,
.Type => unreachable,
.EnumLiteral => unreachable,
else => return self.todo("implement const of type '{}'", .{tv.ty}),
.Void => unreachable,
.NoReturn => unreachable,
.Undefined => unreachable,
.Null => unreachable,
.BoundFn => unreachable,
.Opaque => unreachable,
.Union,
.Frame,
.AnyFrame,
.Vector,
=> return self.todo("implement const of type '{}'", .{tv.ty}),
}
}
@ -1195,21 +1222,15 @@ pub const FuncGen = struct {
const bin_op = self.air.instructions.items(.data)[inst].bin_op;
const lhs = try self.resolveInst(bin_op.lhs);
const rhs = try self.resolveInst(bin_op.rhs);
const inst_ty = self.air.typeOfIndex(inst);
const operand_ty = self.air.typeOf(bin_op.lhs);
switch (self.air.typeOf(bin_op.lhs).zigTypeTag()) {
.Int, .Bool, .Pointer, .ErrorSet => {
const is_signed = inst_ty.isSignedInt();
const operation = switch (op) {
.eq => .EQ,
.neq => .NE,
.lt => @as(llvm.IntPredicate, if (is_signed) .SLT else .ULT),
.lte => @as(llvm.IntPredicate, if (is_signed) .SLE else .ULE),
.gt => @as(llvm.IntPredicate, if (is_signed) .SGT else .UGT),
.gte => @as(llvm.IntPredicate, if (is_signed) .SGE else .UGE),
};
return self.builder.buildICmp(operation, lhs, rhs, "");
const int_ty = switch (operand_ty.zigTypeTag()) {
.Enum => blk: {
var buffer: Type.Payload.Bits = undefined;
const int_ty = operand_ty.enumTagType(&buffer);
break :blk int_ty;
},
.Int, .Bool, .Pointer, .ErrorSet => operand_ty,
.Float => {
const operation: llvm.RealPredicate = switch (op) {
.eq => .OEQ,
@ -1222,7 +1243,17 @@ pub const FuncGen = struct {
return self.builder.buildFCmp(operation, lhs, rhs, "");
},
else => unreachable,
}
};
const is_signed = int_ty.isSignedInt();
const operation = switch (op) {
.eq => .EQ,
.neq => .NE,
.lt => @as(llvm.IntPredicate, if (is_signed) .SLT else .ULT),
.lte => @as(llvm.IntPredicate, if (is_signed) .SLE else .ULE),
.gt => @as(llvm.IntPredicate, if (is_signed) .SGT else .UGT),
.gte => @as(llvm.IntPredicate, if (is_signed) .SGE else .UGE),
};
return self.builder.buildICmp(operation, lhs, rhs, "");
}
fn airBlock(self: *FuncGen, inst: Air.Inst.Index) !?*const llvm.Value {

View File

@ -118,3 +118,15 @@ test "cmpxchg on a global variable" {
_ = @cmpxchgWeak(u32, &a_global_variable, 1234, 42, .Acquire, .Monotonic);
try expect(a_global_variable == 42);
}
test "atomic load and rmw with enum" {
const Value = enum(u8) { a, b, c };
var x = Value.a;
try expect(@atomicLoad(Value, &x, .SeqCst) != .b);
_ = @atomicRmw(Value, &x, .Xchg, .c, .SeqCst);
try expect(@atomicLoad(Value, &x, .SeqCst) == .c);
try expect(@atomicLoad(Value, &x, .SeqCst) != .a);
try expect(@atomicLoad(Value, &x, .SeqCst) != .b);
}

View File

@ -3,22 +3,6 @@ const expect = std.testing.expect;
const expectEqual = std.testing.expectEqual;
const builtin = @import("builtin");
test "atomic load and rmw with enum" {
const Value = enum(u8) {
a,
b,
c,
};
var x = Value.a;
try expect(@atomicLoad(Value, &x, .SeqCst) != .b);
_ = @atomicRmw(Value, &x, .Xchg, .c, .SeqCst);
try expect(@atomicLoad(Value, &x, .SeqCst) == .c);
try expect(@atomicLoad(Value, &x, .SeqCst) != .a);
try expect(@atomicLoad(Value, &x, .SeqCst) != .b);
}
test "atomic store" {
var x: u32 = 0;
@atomicStore(u32, &x, 1, .SeqCst);