ITStudio_OUC/OxideJS
Archived
0
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: BigInt SUB/MUL with 0n no longer raise Division by zero

Browse Source 
binary_arith! applied an unconditional is_zero() guard, so x - 0n and
x * 0n (and wrapped-object variants) incorrectly threw RangeError.
Only DIV/MOD legitimately reject BigInt zero divisors; SUB/MUL must
compute normally. test262 language/expressions/subtraction:
29/6 -> 30/5 (bigint-arithmetic now passes).
This commit is contained in:
爱特每羊队
2026-08-12 03:30:03 +08:00
parent f3c37911f7
commit b844505ec7
1 changed files with 9 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files
crates/oxide_vm/src/vm.rs
+9 -8
View File
@@ -80,7 +80,7 @@ macro_rules! throw_err {
}
macro_rules! binary_arith {
($self:ident, $a:expr, $b:expr, $rd:expr, $op:tt) => {{
($self:ident, $a:expr, $b:expr, $rd:expr, $op:tt, $check_zero:expr) => {{
let lv = $self.regs[$a];
let rv = $self.regs[$b];
if lv.is_int() && rv.is_int() {
@@ -88,8 +88,9 @@ macro_rules! binary_arith {
} else if lv.is_bigint() && rv.is_bigint() {
let l = $self.bigint_value(lv).clone();
let r = $self.bigint_value(rv).clone();
if r.is_zero() {
// BigInt 除/模零：规范要求 RangeError（Number 路径走 f64 inf/NaN）。
if r.is_zero() && $check_zero {
// 仅除/模对 BigInt 零除数抛 RangeError；SUB/MUL 等其余二元运算
// 遇到 0n 必须正常计算（Number 路径走 f64 inf/NaN）。
$self.raise_error_kind("RangeError", "Division by zero")?;
$self.regs[$rd] = JsValue::undefined();
} else {
@@ -104,7 +105,7 @@ macro_rules! binary_arith {
// 包装对象 coerce 后暴露双 BigInt（如 Object(2n) / 2n）。
let lv = $self.bigint_value(l).clone();
let rv = $self.bigint_value(r).clone();
if rv.is_zero() {
if rv.is_zero() && $check_zero {
$self.raise_error_kind("RangeError", "Division by zero")?;
$self.regs[$rd] = JsValue::undefined();
} else {
@@ -1254,19 +1255,19 @@ impl Vm {
}
OpCode::SUB => {
binary_arith!(self, a, b, rd, -);
binary_arith!(self, a, b, rd, -, false);
}
OpCode::MUL => {
binary_arith!(self, a, b, rd, *);
binary_arith!(self, a, b, rd, *, false);
}
OpCode::DIV => {
binary_arith!(self, a, b, rd, /);
binary_arith!(self, a, b, rd, /, true);
}
OpCode::MOD => {
binary_arith!(self, a, b, rd, %);
binary_arith!(self, a, b, rd, %, true);
}
OpCode::NEG => {