Skip to content

GitLab

To protect your data, the CISO officer has suggested users to enable GitLab 2FA as soon as possible.

Commit d9758e2c authored by qinsoon's avatar qinsoon
Browse files

floatingpoint addition

parent abefe290
Hide whitespace changes
Inline Side-by-side
run_test.sh
View file @ d9758e2c
#!/bin/sh
RUST_BACKTRACE=1 RUST_TEST_THREADS=1 cargo test "$@"
RUSTFLAGS=-Zincremental=target/incr-cache RUST_BACKTRACE=1 RUST_TEST_THREADS=1 cargo test "$@"
src/compiler/backend/arch/x86_64/asm_backend.rs
View file @ d9758e2c
......@@ -1375,36 +1375,79 @@ impl CodeGenerator for ASMCodeGen {
false
)
}
fn emit_add_r64_r64(&mut self, dest: &P<Value>, src: &P<Value>) {
trace!("emit: add {}, {} -> {}", dest, src, dest);
let (reg1, id1, loc1) = self.prepare_reg(src, 4 + 1);
let (reg2, id2, loc2) = self.prepare_reg(dest, 4 + 1 + reg1.len() + 1);
let asm = format!("addq {},{}", reg1, reg2);
fn emit_movsd_f64_f64 (&mut self, dest: &P<Value>, src: &P<Value>) {
trace!("emit: movsd {} -> {}", src, dest);
let (reg1, id1, loc1) = self.prepare_reg(src, 5 + 1);
let (reg2, id2, loc2) = self.prepare_reg(dest, 5 + 1 + reg1.len() + 1);
let asm = format!("movsd {},{}", reg1, reg2);
self.add_asm_inst(
asm,
hashmap!{
id2 => vec![loc2.clone()]
id2 => vec![loc2]
},
hashmap!{
id1 => vec![loc1],
id2 => vec![loc2]
id1 => vec![loc1]
},
false
)
}
// load
fn emit_movsd_f64_mem64(&mut self, dest: &P<Value>, src: &P<Value>) {
trace!("emit: movsd {} -> {}", src, dest);
let (mem, uses) = self.prepare_mem(src, 5 + 1);
let (reg, id2, loc2) = self.prepare_reg(dest, 5 + 1 + mem.len() + 1);
let asm = format!("movsd {},{}", mem, reg);
self.add_asm_inst(
asm,
hashmap!{
id2 => vec![loc2]
},
uses,
true
)
}
// store
fn emit_movsd_mem64_f64(&mut self, dest: &P<Value>, src: &P<Value>) {
trace!("emit: movsd {} -> {}", src, dest);
let (reg, id1, loc1) = self.prepare_reg(src, 5 + 1);
let (mem, mut uses) = self.prepare_mem(dest, 5 + 1 + reg.len() + 1);
// the register we used for the memory location is counted as 'use'
// use the vec from mem as 'use' (push use reg from src to it)
if uses.contains_key(&id1) {
uses.get_mut(&id1).unwrap().push(loc1);
} else {
uses.insert(id1, vec![loc1]);
}
let asm = format!("movsd {},{}", reg, mem);
self.add_asm_inst(
asm,
hashmap!{},
uses,
true
)
}
fn emit_lea_r64(&mut self, dest: &P<Value>, src: &P<Value>) {
trace!("emit: lea {} -> {}", src, dest);
let (mem, uses) = self.prepare_mem(src, 4 + 1);
let (reg, id2, loc2) = self.prepare_reg(dest, 4 + 1 + mem.len() + 1);
let asm = format!("leaq {},{}", mem, reg);
self.add_asm_inst(
asm,
hashmap!{
......@@ -1412,16 +1455,16 @@ impl CodeGenerator for ASMCodeGen {
},
uses,
true
)
)
}
fn emit_and_r64_imm32(&mut self, dest: &P<Value>, src: i32) {
trace!("emit: and {}, {} -> {}", src, dest, dest);
let (reg1, id1, loc1) = self.prepare_reg(dest, 4 + 1 + 1 + src.to_string().len() + 1);
let asm = format!("andq ${},{}", src, reg1);
self.add_asm_inst(
asm,
hashmap!{
......@@ -1433,14 +1476,35 @@ impl CodeGenerator for ASMCodeGen {
false
)
}
fn emit_and_r64_r64(&mut self, dest: &P<Value>, src: &P<Value>) {
trace!("emit: and {}, {} -> {}", src, dest, dest);
let (reg1, id1, loc1) = self.prepare_reg(src, 4 + 1);
let (reg2, id2, loc2) = self.prepare_reg(dest, 4 + 1 + reg1.len() + 1);
let (reg2, id2, loc2) = self.prepare_reg(dest, 4 + 1 + reg1.len() + 1);
let asm = format!("andq {},{}", reg1, reg2);
self.add_asm_inst(
asm,
hashmap!{
id2 => vec![loc2.clone()]
},
hashmap!{
id1 => vec![loc1],
id2 => vec![loc2]
},
false
)
}
fn emit_add_r64_r64(&mut self, dest: &P<Value>, src: &P<Value>) {
trace!("emit: add {}, {} -> {}", dest, src, dest);
let (reg1, id1, loc1) = self.prepare_reg(src, 4 + 1);
let (reg2, id2, loc2) = self.prepare_reg(dest, 4 + 1 + reg1.len() + 1);
let asm = format!("addq {},{}", reg1, reg2);
self.add_asm_inst(
asm,
......@@ -1453,7 +1517,7 @@ impl CodeGenerator for ASMCodeGen {
},
false
)
}
}
fn emit_add_r64_mem64(&mut self, dest: &P<Value>, src: &P<Value>) {
trace!("emit: add {}, {} -> {}", dest, src, dest);
......@@ -1478,6 +1542,32 @@ impl CodeGenerator for ASMCodeGen {
false
)
}
fn emit_addsd_f64_f64 (&mut self, dest: &P<Value>, src: &P<Value>) {
trace!("emit: addsd {}, {} -> {}", dest, src, dest);
let (reg1, id1, loc1) = self.prepare_reg(src, 5 + 1);
let (reg2, id2, loc2) = self.prepare_reg(dest, 5 + 1 + reg1.len() + 1);
let asm = format!("addsd {},{}", reg1, reg2);
self.add_asm_inst(
asm,
hashmap!{
id2 => vec![loc2.clone()]
},
hashmap!{
id1 => vec![loc1],
id2 => vec![loc2]
},
false
)
}
fn emit_addsd_f64_mem64(&mut self, dest: &P<Value>, src: &P<Value>) {
trace!("emit: addsd {}, {} -> {}", dest, src, dest);
unimplemented!()
}
fn emit_sub_r64_r64(&mut self, dest: &P<Value>, src: &P<Value>) {
trace!("emit: sub {}, {} -> {}", dest, src, dest);
......
src/compiler/backend/arch/x86_64/codegen.rs
View file @ d9758e2c
......@@ -26,26 +26,33 @@ pub trait CodeGenerator {
fn emit_cmp_r64_imm32(&mut self, op1: &P<Value>, op2: i32);
fn emit_cmp_r64_mem64(&mut self, op1: &P<Value>, op2: &P<Value>);
fn emit_mov_r64_imm32(&mut self, dest: &P<Value>, src: i32);
fn emit_mov_r64_mem64(&mut self, dest: &P<Value>, src: &P<Value>); // load
fn emit_mov_r64_r64(&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_mov_mem64_r64(&mut self, dest: &P<Value>, src: &P<Value>); // store
fn emit_mov_r64_imm32 (&mut self, dest: &P<Value>, src: i32);
fn emit_mov_r64_mem64 (&mut self, dest: &P<Value>, src: &P<Value>); // load
fn emit_mov_r64_r64 (&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_mov_mem64_r64 (&mut self, dest: &P<Value>, src: &P<Value>); // store
fn emit_mov_mem64_imm32(&mut self, dest: &P<Value>, src: i32);
fn emit_movsd_f64_f64 (&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_movsd_f64_mem64(&mut self, dest: &P<Value>, src: &P<Value>); // load
fn emit_movsd_mem64_f64(&mut self, dest: &P<Value>, src: &P<Value>); // store
fn emit_lea_r64(&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_and_r64_imm32(&mut self, dest: &P<Value>, src: i32);
fn emit_and_r64_r64(&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_and_r64_r64 (&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_add_r64_r64(&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_add_r64_r64 (&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_add_r64_mem64(&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_add_r64_imm32(&mut self, dest: &P<Value>, src: i32);
fn emit_addsd_f64_f64 (&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_addsd_f64_mem64(&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_sub_r64_r64(&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_sub_r64_mem64(&mut self, dest: &P<Value>, src: &P<Value>);
fn emit_sub_r64_imm32(&mut self, dest: &P<Value>, src: i32);
fn emit_mul_r64(&mut self, src: &P<Value>);
fn emit_mul_r64 (&mut self, src: &P<Value>);
fn emit_mul_mem64(&mut self, src: &P<Value>);
fn emit_jmp(&mut self, dest: MuName);
......
src/compiler/backend/arch/x86_64/inst_sel.rs
View file @ d9758e2c
......@@ -171,7 +171,7 @@ impl <'a> InstructionSelection {
let reg_op1 = self.emit_ireg(&ops[op1], f_content, f_context, vm);
let reg_op2 = self.emit_ireg(&ops[op2], f_content, f_context, vm);
let res_tmp = self.emit_get_result(node);
let res_tmp = self.get_result_value(node);
// mov op1, res
self.backend.emit_mov_r64_r64(&res_tmp, &reg_op1);
......@@ -182,7 +182,7 @@ impl <'a> InstructionSelection {
let reg_op1 = self.emit_ireg(&ops[op1], f_content, f_context, vm);
let reg_op2 = self.node_iimm_to_i32(&ops[op2]);
let res_tmp = self.emit_get_result(node);
let res_tmp = self.get_result_value(node);
// mov op1, res
self.backend.emit_mov_r64_r64(&res_tmp, &reg_op1);
......@@ -196,7 +196,7 @@ impl <'a> InstructionSelection {
let reg_op1 = self.emit_ireg(&ops[op1], f_content, f_context, vm);
let reg_op2 = self.emit_mem(&ops[op2]);
let res_tmp = self.emit_get_result(node);
let res_tmp = self.get_result_value(node);
// mov op1, res
self.backend.emit_mov_r64_r64(&res_tmp, &reg_op1);
......@@ -215,7 +215,7 @@ impl <'a> InstructionSelection {
let reg_op1 = self.emit_ireg(&ops[op1], f_content, f_context, vm);
let reg_op2 = self.emit_ireg(&ops[op2], f_content, f_context, vm);
let res_tmp = self.emit_get_result(node);
let res_tmp = self.get_result_value(node);
// mov op1, res
self.backend.emit_mov_r64_r64(&res_tmp, &reg_op1);
......@@ -226,7 +226,7 @@ impl <'a> InstructionSelection {
let reg_op1 = self.emit_ireg(&ops[op1], f_content, f_context, vm);
let imm_op2 = self.node_iimm_to_i32(&ops[op2]);
let res_tmp = self.emit_get_result(node);
let res_tmp = self.get_result_value(node);
// mov op1, res
self.backend.emit_mov_r64_r64(&res_tmp, &reg_op1);
......@@ -240,7 +240,7 @@ impl <'a> InstructionSelection {
let reg_op1 = self.emit_ireg(&ops[op1], f_content, f_context, vm);
let mem_op2 = self.emit_mem(&ops[op2]);
let res_tmp = self.emit_get_result(node);
let res_tmp = self.get_result_value(node);
// mov op1, res
self.backend.emit_mov_r64_r64(&res_tmp, &reg_op1);
......@@ -284,7 +284,7 @@ impl <'a> InstructionSelection {
// put imm in a temporary
// here we use result reg as temporary
let res_tmp = self.emit_get_result(node);
let res_tmp = self.get_result_value(node);
self.backend.emit_mov_r64_imm32(&res_tmp, imm_op2);
self.backend.emit_mul_r64(&res_tmp);
......@@ -297,9 +297,41 @@ impl <'a> InstructionSelection {
}
// mov rax -> result
let res_tmp = self.emit_get_result(node);
let res_tmp = self.get_result_value(node);
self.backend.emit_mov_r64_r64(&res_tmp, &rax);
},
// floating point
op::BinOp::FAdd => {
if self.match_fpreg(&ops[op1]) && self.match_fpreg(&ops[op2]) {
trace!("emit add-fpreg-fpreg");
let reg_op1 = self.emit_fpreg(&ops[op1], f_content, f_context, vm);
let reg_op2 = self.emit_fpreg(&ops[op2], f_content, f_context, vm);
let res_tmp = self.get_result_value(node);
// movsd op1, res
self.backend.emit_movsd_f64_f64(&res_tmp, &reg_op1);
// add op2 res
self.backend.emit_addsd_f64_f64(&res_tmp, &reg_op2);
} else if self.match_fpreg(&ops[op1]) && self.match_mem(&ops[op2]) {
trace!("emit add-fpreg-mem");
let reg_op1 = self.emit_fpreg(&ops[op1], f_content, f_context, vm);
let mem_op2 = self.emit_mem(&ops[op2]);
let res_tmp = self.get_result_value(node);
// mov op1, res
self.backend.emit_movsd_f64_f64(&res_tmp, &reg_op1);
// sub op2 res
self.backend.emit_addsd_f64_mem64(&res_tmp, &mem_op2);
} else if self.match_mem(&ops[op1]) && self.match_fpreg(&ops[op2]) {
trace!("emit add-mem-fpreg");
unimplemented!();
} else {
unimplemented!()
}
}
_ => unimplemented!()
}
......@@ -321,7 +353,7 @@ impl <'a> InstructionSelection {
}
let resolved_loc = self.node_mem_to_value(loc_op, vm);
let res_temp = self.emit_get_result(node);
let res_temp = self.get_result_value(node);
if self.match_ireg(node) {
// emit mov(GPR)
......@@ -371,7 +403,7 @@ impl <'a> InstructionSelection {
let ops = inst.ops.read().unwrap();
let ref op = ops[op_index];
let res_tmp = self.emit_get_result(node);
let res_tmp = self.get_result_value(node);
let hdr_size = mm::objectmodel::OBJECT_HEADER_SIZE;
if hdr_size == 0 {
......@@ -440,7 +472,7 @@ impl <'a> InstructionSelection {
// put start as result
// ASM: mov %start -> %result
let tmp_res = self.emit_get_result(node);
let tmp_res = self.get_result_value(node);
self.backend.emit_mov_r64_r64(&tmp_res, &tmp_start);
// ASM jmp alloc_end
......@@ -533,8 +565,14 @@ impl <'a> InstructionSelection {
fn emit_load_base_offset (&mut self, dest: &P<Value>, base: &P<Value>, offset: i32, vm: &VM) {
let mem = self.make_memory_op_base_offset(base, offset, dest.ty.clone(), vm);
self.backend.emit_mov_r64_mem64(dest, &mem);
if dest.is_int_reg() {
self.backend.emit_mov_r64_mem64(dest, &mem);
} else if dest.is_fp_reg() {
self.backend.emit_movsd_f64_mem64(dest, &mem);
} else {
unimplemented!();
}
}
fn emit_store_base_offset (&mut self, base: &P<Value>, offset: i32, src: &P<Value>, vm: &VM) {
......@@ -929,7 +967,7 @@ impl <'a> InstructionSelection {
// unload arguments
let mut gpr_arg_count = 0;
// TODO: let mut fpr_arg_count = 0;
let mut fpr_arg_count = 0;
// initial stack arg is at RBP+16
// arg <- RBP + 16
// return addr
......@@ -949,7 +987,17 @@ impl <'a> InstructionSelection {
stack_arg_offset += arg_size as i32;
}
} else if arg.is_fp_reg() {
unimplemented!();
if fpr_arg_count < x86_64::ARGUMENT_FPRs.len() {
self.backend.emit_movsd_f64_f64(&arg, &x86_64::ARGUMENT_FPRs[fpr_arg_count]);
fpr_arg_count += 1;
} else {
// unload from stack
self.emit_load_base_offset(&arg, &x86_64::RBP.clone(), stack_arg_offset, vm);
// move stack_arg_offset by the size of 'arg'
let arg_size = vm.get_backend_type_info(arg.ty.id()).size;
stack_arg_offset += arg_size as i32;
}
} else {
// args that are not fp or int (possibly struct/array/etc)
unimplemented!();
......@@ -971,7 +1019,7 @@ impl <'a> InstructionSelection {
};
let mut gpr_ret_count = 0;
// TODO: let mut fpr_ret_count = 0;
let mut fpr_ret_count = 0;
for i in ret_val_indices {
let ref ret_val = ops[*i];
if self.match_ireg(ret_val) {
......@@ -984,6 +1032,11 @@ impl <'a> InstructionSelection {
self.backend.emit_mov_r64_imm32(&x86_64::RETURN_GPRs[gpr_ret_count], imm_ret_val);
gpr_ret_count += 1;
} else if self.match_fpreg(ret_val) {
let reg_ret_val = self.emit_fpreg(ret_val, f_content, f_context, vm);
self.backend.emit_movsd_f64_f64(&x86_64::RETURN_FPRs[fpr_ret_count], &reg_ret_val);
fpr_ret_count += 1;
} else {
unimplemented!();
}
......@@ -1061,7 +1114,7 @@ impl <'a> InstructionSelection {
let ref value = inst.value.as_ref().unwrap()[0];
if types::is_scalar(&value.ty) {
if value.is_int_reg() {
true
} else {
false
......@@ -1076,30 +1129,63 @@ impl <'a> InstructionSelection {
}
}
}
fn match_fpreg(&mut self, op: &TreeNode) -> bool {
match op.v {
TreeNode_::Instruction(ref inst) => {
if inst.value.is_some() {
if inst.value.as_ref().unwrap().len() > 1 {
return false;
}
let ref value = inst.value.as_ref().unwrap()[0];
if value.is_fp_reg() {
true
} else {
false
}
} else {
false
}
}
TreeNode_::Value(ref pv) => {
pv.is_fp_reg()
}
}
}
fn emit_ireg(&mut self, op: &P<TreeNode>, f_content: &FunctionContent, f_context: &mut FunctionContext, vm: &VM) -> P<Value> {
match op.v {
TreeNode_::Instruction(_) => {
self.instruction_select(op, f_content, f_context, vm);
self.emit_get_result(op)
self.get_result_value(op)
},
TreeNode_::Value(ref pv) => {
match pv.v {
Value_::Constant(_)
| Value_::Global(_)
| Value_::Memory(_) => panic!("expected ireg"),
Value_::SSAVar(_) => {
pv.clone()
},
Value_::SSAVar(_) => pv.clone(),
_ => panic!("expected ireg")
}
}
}
}
#[allow(unused_variables)]
fn match_fpreg(&mut self, op: &P<TreeNode>) -> bool {
unimplemented!()
fn emit_fpreg(&mut self, op: &P<TreeNode>, f_content: &FunctionContent, f_context: &mut FunctionContext, vm: &VM) -> P<Value> {
match op.v {
TreeNode_::Instruction(_) => {
self.instruction_select(op, f_content, f_context, vm);
self.get_result_value(op)
},
TreeNode_::Value(ref pv) => {
match pv.v {
Value_::SSAVar(_) => pv.clone(),
_ => panic!("expected fpreg")
}
}
}
}
fn match_iimm(&mut self, op: &P<TreeNode>) -> bool {
......@@ -1224,7 +1310,7 @@ impl <'a> InstructionSelection {
unimplemented!()
}
fn emit_get_result(&mut self, node: &TreeNode) -> P<Value> {
fn get_result_value(&mut self, node: &TreeNode) -> P<Value> {
match node.v {
TreeNode_::Instruction(ref inst) => {
if inst.value.is_some() {
......
src/compiler/backend/arch/x86_64/mod.rs
View file @ d9758e2c
......@@ -143,7 +143,9 @@ lazy_static!{
XMM1.clone()
];
pub static ref ARGUMENT_FPRs : [P<Value>; 6] = [
pub static ref ARGUMENT_FPRs : [P<Value>; 8] = [
XMM0.clone(),
XMM1.clone(),
XMM2.clone(),
XMM3.clone(),
XMM4.clone(),
......
tests/lib.rs
View file @ d9758e2c
......@@ -2,6 +2,8 @@ extern crate mu;
#[macro_use]
extern crate log;
extern crate simple_logger;
#[macro_use]
extern crate maplit;
mod test_ir;
mod test_compiler;
......
tests/test_compiler/mod.rs
View file @ d9758e2c
......@@ -5,4 +5,5 @@ mod test_global;
mod test_compiler;
mod test_alloc;
mod test_exception;
mod test_thread;
\ No newline at end of file
mod test_thread;
mod test_floatingpoint;
\ No newline at end of file
tests/test_compiler/test_floatingpoint.rs 0 → 100644
View file @ d9758e2c
extern crate mu;
extern crate log;
extern crate simple_logger;
extern crate libloading;
use self::mu::ast::types::*;
use self::mu::ast::ir::*;
use self::mu::ast::inst::*;
use self::mu::ast::op::*;
use self::mu::vm::*;
use self::mu::compiler::*;
use std::sync::RwLock;
use std::sync::Arc;
use aot;
#[test]
fn test_fp_add() {
simple_logger::init_with_level(log::LogLevel::Trace).ok();
let vm = Arc::new(fp_add());
let compiler = Compiler::new(CompilerPolicy::default(), vm.clone());
let func_id = vm.id_of("fp_add");
{