user0/rust
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Auto merge of #45205 - rkruppe:saturating-casts, r=eddyb

Browse source 
Saturating casts between integers and floats

Introduces a new flag, `-Z saturating-float-casts`, which makes code generation for int->float and float->int casts safe (`undef`-free), implementing [the saturating semantics laid out by](https://github.com/rust-lang/rust/issues/10184#issuecomment-299229143) @jorendorff for float->int casts and overflowing to infinity for `u128::MAX` -> `f32`.
Constant evaluation in trans was changed to behave like HIR const eval already did, i.e., saturate for u128->f32 and report an error for problematic float->int casts.

Many thanks to @eddyb, whose APFloat port simplified many parts of this patch, and made HIR constant evaluation recognize dangerous float casts as mentioned above.
Also thanks to @ActuallyaDeviloper whose branchless implementation served as inspiration for this implementation.

cc #10184 #41799
fixes #45134
This commit is contained in:
bors 2017-11-08 17:27:56 +00:00
commit 7ca430df71
11 changed files with 535 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

65
src/test/codegen/unchecked-float-casts.rs Normal file
View file

@ -0,0 +1,65 @@
// Copyright 2017 The Rust Project Developers. See the COPYRIGHT
// file at the top-level directory of this distribution and at
// http://rust-lang.org/COPYRIGHT.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.
// compile-flags: -C no-prepopulate-passes
// This file tests that we don't generate any code for saturation if
// -Z saturating-float-casts is not enabled.
#![crate_type = "lib"]
#![feature(i128_type)]
// CHECK-LABEL: @f32_to_u32
#[no_mangle]
pub fn f32_to_u32(x: f32) -> u32 {
// CHECK: fptoui
// CHECK-NOT: fcmp
// CHECK-NOT: icmp
// CHECK-NOT: select
x as u32
}
// CHECK-LABEL: @f32_to_i32
#[no_mangle]
pub fn f32_to_i32(x: f32) -> i32 {
// CHECK: fptosi
// CHECK-NOT: fcmp
// CHECK-NOT: icmp
// CHECK-NOT: select
x as i32
}
#[no_mangle]
pub fn f64_to_u8(x: f32) -> u16 {
// CHECK-NOT: fcmp
// CHECK-NOT: icmp
// CHECK-NOT: select
x as u16
}
// CHECK-LABEL: @i32_to_f64
#[no_mangle]
pub fn i32_to_f64(x: i32) -> f64 {
// CHECK: sitofp
// CHECK-NOT: fcmp
// CHECK-NOT: icmp
// CHECK-NOT: select
x as f64
}
// CHECK-LABEL: @u128_to_f32
#[no_mangle]
pub fn u128_to_f32(x: u128) -> f32 {
// CHECK: uitofp
// CHECK-NOT: fcmp
// CHECK-NOT: icmp
// CHECK-NOT: select
x as f32
}