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

Jump to bottom

Add more constants, functions, and tests for f16 and f128 #126608

Merged
merged 6 commits into from
Jun 26, 2024
Merged

Add more constants, functions, and tests for f16 and f128 #126608

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
fce07a8
Extract repeated constants from `f32` and `f64` source
tgross35 Jun 16, 2024
0314fe6
Reword docs for `f32` and `f64`
tgross35 Jun 21, 2024
99ed3d3
Add build.rs config for reliable `f16` and `f128`
tgross35 Jun 21, 2024
0eee055
Add doctests to existing `f16` and `f128` functions
tgross35 Jun 18, 2024
6e2d934
Add more `f16` and `f128` library functions and constants
tgross35 Jun 18, 2024
e181297
Add tests for `f16` and `f128`
tgross35 Jun 18, 2024
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
Prev Previous commit
Next Next commit
Reword docs for f32 and f64 
Better explain the reasoning for the `next_up`/`next_down` integer
implementation, as requested by Ralf.
  • Loading branch information
@tgross35
tgross35 committed Jun 24, 2024
commit 0314fe62bd178200a620b90b59becd735d6c2e66
10 changes: 6 additions & 4 deletions library/core/src/num/f32.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -796,8 +796,9 @@ impl f32 {
#[unstable(feature = "float_next_up_down", issue = "91399")]
#[rustc_const_unstable(feature = "float_next_up_down", issue = "91399")]
pub const fn next_up(self) -> Self {
// We must use strictly integer arithmetic to prevent denormals from
// flushing to zero after an arithmetic operation on some platforms.
// Some targets violate Rust's assumption of IEEE semantics, e.g. by flushing
// denormals to zero. This is in general unsound and unsupported, but here
// we do our best to still produce the correct result on such targets.
let bits = self.to_bits();
if self.is_nan() || bits == Self::INFINITY.to_bits() {
return self;
Expand Down Expand Up @@ -843,8 +844,9 @@ impl f32 {
#[unstable(feature = "float_next_up_down", issue = "91399")]
#[rustc_const_unstable(feature = "float_next_up_down", issue = "91399")]
pub const fn next_down(self) -> Self {
// We must use strictly integer arithmetic to prevent denormals from
// flushing to zero after an arithmetic operation on some platforms.
// Some targets violate Rust's assumption of IEEE semantics, e.g. by flushing
// denormals to zero. This is in general unsound and unsupported, but here
// we do our best to still produce the correct result on such targets.
let bits = self.to_bits();
if self.is_nan() || bits == Self::NEG_INFINITY.to_bits() {
return self;
Expand Down
10 changes: 6 additions & 4 deletions library/core/src/num/f64.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -804,8 +804,9 @@ impl f64 {
#[unstable(feature = "float_next_up_down", issue = "91399")]
#[rustc_const_unstable(feature = "float_next_up_down", issue = "91399")]
pub const fn next_up(self) -> Self {
// We must use strictly integer arithmetic to prevent denormals from
// flushing to zero after an arithmetic operation on some platforms.
// Some targets violate Rust's assumption of IEEE semantics, e.g. by flushing
// denormals to zero. This is in general unsound and unsupported, but here
// we do our best to still produce the correct result on such targets.
let bits = self.to_bits();
if self.is_nan() || bits == Self::INFINITY.to_bits() {
return self;
Expand Down Expand Up @@ -851,8 +852,9 @@ impl f64 {
#[unstable(feature = "float_next_up_down", issue = "91399")]
#[rustc_const_unstable(feature = "float_next_up_down", issue = "91399")]
pub const fn next_down(self) -> Self {
// We must use strictly integer arithmetic to prevent denormals from
// flushing to zero after an arithmetic operation on some platforms.
// Some targets violate Rust's assumption of IEEE semantics, e.g. by flushing
// denormals to zero. This is in general unsound and unsupported, but here
// we do our best to still produce the correct result on such targets.
let bits = self.to_bits();
if self.is_nan() || bits == Self::NEG_INFINITY.to_bits() {
return self;
Expand Down