with Validation MIR around drop being fixed, we can remove the hack working around this
This commit is contained in:
parent
dd533a9ec6
commit
b3fa42fc0e
@ -11,7 +11,7 @@ use super::{
|
||||
EvalError, EvalResult, EvalErrorKind,
|
||||
EvalContext, DynamicLifetime,
|
||||
AccessKind, LockInfo,
|
||||
PrimVal, Value,
|
||||
Value,
|
||||
Lvalue, LvalueExtra,
|
||||
Machine,
|
||||
};
|
||||
@ -181,16 +181,11 @@ std::sync::atomic::AtomicBool::get_mut$|\
|
||||
// HACK: For now, bail out if we hit a dead local during recovery (can happen because sometimes we have
|
||||
// StorageDead before EndRegion).
|
||||
// TODO: We should rather fix the MIR.
|
||||
// HACK: Releasing on dead/undef local variables is a NOP. This can happen because of releases being added
|
||||
// before drop elaboration.
|
||||
// TODO: Fix the MIR so that these releases do not happen.
|
||||
match query.lval {
|
||||
Lvalue::Local { frame, local } => {
|
||||
let res = self.stack[frame].get_local(local);
|
||||
match (res, mode) {
|
||||
(Err(EvalError{ kind: EvalErrorKind::DeadLocal, ..}), ValidationMode::Recover(_)) |
|
||||
(Err(EvalError{ kind: EvalErrorKind::DeadLocal, ..}), ValidationMode::Release) |
|
||||
(Ok(Value::ByVal(PrimVal::Undef)), ValidationMode::Release) => {
|
||||
(Err(EvalError{ kind: EvalErrorKind::DeadLocal, ..}), ValidationMode::Recover(_)) => {
|
||||
return Ok(());
|
||||
}
|
||||
_ => {},
|
||||
|
Loading…
x
Reference in New Issue
Block a user