stage2 ARM: fix register allocation in genArmBinOp

Previously, this would reuse an operand even if reuseOperand returned
false for both operands.

genArmBinOpCode was also changed to be more Three-address code oriented
in the process.
This commit is contained in:
joachimschmidt557 2021-02-07 20:01:11 +01:00 committed by Jakub Konka
parent c2beaba85a
commit 6a5a6386c6

View File

@ -1295,38 +1295,31 @@ fn Function(comptime arch: std.Target.Cpu.Arch) type {
const rhs = try self.resolveInst(op_rhs);
// Destination must be a register
// Source may be register, memory or an immediate
//
// So there are two options: (lhs is src and rhs is dest)
// or (rhs is src and lhs is dest)
const lhs_is_dest = blk: {
if (self.reuseOperand(inst, 0, lhs)) {
break :blk true;
} else if (self.reuseOperand(inst, 1, rhs)) {
break :blk false;
} else {
break :blk lhs == .register;
}
};
var dst_mcv: MCValue = undefined;
var src_mcv: MCValue = undefined;
var src_inst: *ir.Inst = undefined;
if (lhs_is_dest) {
var lhs_mcv: MCValue = undefined;
var rhs_mcv: MCValue = undefined;
if (self.reuseOperand(inst, 0, lhs)) {
// LHS is the destination
// RHS is the source
src_inst = op_rhs;
src_mcv = rhs;
dst_mcv = if (lhs != .register) try self.copyToNewRegister(inst, lhs) else lhs;
} else {
lhs_mcv = if (lhs != .register) try self.copyToNewRegister(inst, lhs) else lhs;
rhs_mcv = rhs;
dst_mcv = lhs_mcv;
} else if (self.reuseOperand(inst, 1, rhs)) {
// RHS is the destination
// LHS is the source
src_inst = op_lhs;
src_mcv = lhs;
dst_mcv = if (rhs != .register) try self.copyToNewRegister(inst, rhs) else rhs;
lhs_mcv = lhs;
rhs_mcv = if (rhs != .register) try self.copyToNewRegister(inst, rhs) else rhs;
dst_mcv = rhs_mcv;
} else {
// TODO save 1 copy instruction by directly allocating the destination register
// LHS is the destination
// RHS is the source
lhs_mcv = try self.copyToNewRegister(inst, lhs);
rhs_mcv = rhs;
dst_mcv = lhs_mcv;
}
try self.genArmBinOpCode(inst.src, dst_mcv.register, src_mcv, lhs_is_dest, op);
try self.genArmBinOpCode(inst.src, dst_mcv.register, lhs_mcv, rhs_mcv, op);
return dst_mcv;
}
@ -1334,11 +1327,17 @@ fn Function(comptime arch: std.Target.Cpu.Arch) type {
self: *Self,
src: usize,
dst_reg: Register,
src_mcv: MCValue,
lhs_is_dest: bool,
lhs_mcv: MCValue,
rhs_mcv: MCValue,
op: ir.Inst.Tag,
) !void {
const operand = switch (src_mcv) {
assert(lhs_mcv == .register or lhs_mcv == .register);
const swap_lhs_and_rhs = rhs_mcv == .register and lhs_mcv != .register;
const op1 = if (swap_lhs_and_rhs) rhs_mcv.register else lhs_mcv.register;
const op2 = if (swap_lhs_and_rhs) lhs_mcv else rhs_mcv;
const operand = switch (op2) {
.none => unreachable,
.undef => unreachable,
.dead, .unreach => unreachable,
@ -1352,37 +1351,37 @@ fn Function(comptime arch: std.Target.Cpu.Arch) type {
// Load immediate into register if it doesn't fit
// as an operand
break :blk Instruction.Operand.fromU32(@intCast(u32, imm)) orelse
Instruction.Operand.reg(try self.copyToTmpRegister(src, src_mcv), Instruction.Operand.Shift.none);
Instruction.Operand.reg(try self.copyToTmpRegister(src, op2), Instruction.Operand.Shift.none);
},
.register => |src_reg| Instruction.Operand.reg(src_reg, Instruction.Operand.Shift.none),
.register => |reg| Instruction.Operand.reg(reg, Instruction.Operand.Shift.none),
.stack_offset,
.embedded_in_code,
.memory,
=> Instruction.Operand.reg(try self.copyToTmpRegister(src, src_mcv), Instruction.Operand.Shift.none),
=> Instruction.Operand.reg(try self.copyToTmpRegister(src, op2), Instruction.Operand.Shift.none),
};
switch (op) {
.add => {
writeInt(u32, try self.code.addManyAsArray(4), Instruction.add(.al, dst_reg, dst_reg, operand).toU32());
writeInt(u32, try self.code.addManyAsArray(4), Instruction.add(.al, dst_reg, op1, operand).toU32());
},
.sub => {
if (lhs_is_dest) {
writeInt(u32, try self.code.addManyAsArray(4), Instruction.sub(.al, dst_reg, dst_reg, operand).toU32());
if (swap_lhs_and_rhs) {
writeInt(u32, try self.code.addManyAsArray(4), Instruction.rsb(.al, dst_reg, op1, operand).toU32());
} else {
writeInt(u32, try self.code.addManyAsArray(4), Instruction.rsb(.al, dst_reg, dst_reg, operand).toU32());
writeInt(u32, try self.code.addManyAsArray(4), Instruction.sub(.al, dst_reg, op1, operand).toU32());
}
},
.bool_and, .bit_and => {
writeInt(u32, try self.code.addManyAsArray(4), Instruction.@"and"(.al, dst_reg, dst_reg, operand).toU32());
writeInt(u32, try self.code.addManyAsArray(4), Instruction.@"and"(.al, dst_reg, op1, operand).toU32());
},
.bool_or, .bit_or => {
writeInt(u32, try self.code.addManyAsArray(4), Instruction.orr(.al, dst_reg, dst_reg, operand).toU32());
writeInt(u32, try self.code.addManyAsArray(4), Instruction.orr(.al, dst_reg, op1, operand).toU32());
},
.not, .xor => {
writeInt(u32, try self.code.addManyAsArray(4), Instruction.eor(.al, dst_reg, dst_reg, operand).toU32());
writeInt(u32, try self.code.addManyAsArray(4), Instruction.eor(.al, dst_reg, op1, operand).toU32());
},
.cmp_eq => {
writeInt(u32, try self.code.addManyAsArray(4), Instruction.cmp(.al, dst_reg, operand).toU32());
writeInt(u32, try self.code.addManyAsArray(4), Instruction.cmp(.al, op1, operand).toU32());
},
else => unreachable, // not a binary instruction
}
@ -2135,7 +2134,7 @@ fn Function(comptime arch: std.Target.Cpu.Arch) type {
const src_mcv = rhs;
const dst_mcv = if (lhs != .register) try self.copyToNewRegister(&inst.base, lhs) else lhs;
try self.genArmBinOpCode(inst.base.src, dst_mcv.register, src_mcv, true, .cmp_eq);
try self.genArmBinOpCode(inst.base.src, dst_mcv.register, dst_mcv, src_mcv, .cmp_eq);
const info = inst.lhs.ty.intInfo(self.target.*);
return switch (info.signedness) {
.signed => MCValue{ .compare_flags_signed = op },