Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Clarify MIR semantics of checked binary operations #98738

Merged
merged 1 commit into from
Jul 5, 2022
Merged

Clarify MIR semantics of checked binary operations #98738

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
16 changes: 12 additions & 4 deletions compiler/rustc_middle/src/mir/syntax.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -990,11 +990,19 @@ pub enum Rvalue<'tcx> {
/// matching types and return a value of that type.
BinaryOp(BinOp, Box<(Operand<'tcx>, Operand<'tcx>)>),

/// Same as `BinaryOp`, but yields `(T, bool)` instead of `T`. In addition to performing the
/// same computation as the matching `BinaryOp`, checks if the infinite precison result would be
/// unequal to the actual result and sets the `bool` if this is the case.
/// Same as `BinaryOp`, but yields `(T, bool)` with a `bool` indicating an error condition.
///
/// This only supports addition, subtraction, multiplication, and shift operations on integers.
/// When overflow checking is disabled, the error condition is false. Otherwise, the error
tmiasko marked this conversation as resolved.
Show resolved Hide resolved
/// condition is determined as described below.
///
/// For addition, subtraction, and multiplication on integers the error condition is set when
/// the infinite precision result would be unequal to the actual result.
///
/// For shift operations on integers the error condition is set when the value of right-hand
/// side is greater than or equal to the number of bits in the type of the left-hand side, or
/// when the value of right-hand side is negative.
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Note that the "negative" thing is rather subtle currently: codegen doesn't actually check whether the type is signed or not. If we had u256, then the current codegen logic would break. See here for details.

///
/// Other combinations of types and operators are unsupported.
CheckedBinaryOp(BinOp, Box<(Operand<'tcx>, Operand<'tcx>)>),

/// Computes a value as described by the operation.
Expand Down