Skip to content

Commit ff0d44e

Browse files
krishna-veerareddykrishna-veerareddy
committed
Add imprecise_flops lint
Add lint to detect floating point operations that can be computed more accurately at the cost of performance. `cbrt`, `ln_1p` and `exp_m1` library functions call their equivalent cmath implementations which is slower but more accurate so moving checks for these under this new lint.
1 parent e94a167 commit ff0d44e

14 files changed

+67
-20
lines changed

‎CHANGELOG.md

+1
Original file line numberDiff line numberDiff line change
@@ -1169,6 +1169,7 @@ Released 2018-09-13
11691169
[`ifs_same_cond`]: https://rust-lang.github.io/rust-clippy/master/index.html#ifs_same_cond
11701170
[`implicit_hasher`]: https://rust-lang.github.io/rust-clippy/master/index.html#implicit_hasher
11711171
[`implicit_return`]: https://rust-lang.github.io/rust-clippy/master/index.html#implicit_return
1172+
[`imprecise_flops`]: https://rust-lang.github.io/rust-clippy/master/index.html#imprecise_flops
11721173
[`inconsistent_digit_grouping`]: https://rust-lang.github.io/rust-clippy/master/index.html#inconsistent_digit_grouping
11731174
[`indexing_slicing`]: https://rust-lang.github.io/rust-clippy/master/index.html#indexing_slicing
11741175
[`ineffective_bit_mask`]: https://rust-lang.github.io/rust-clippy/master/index.html#ineffective_bit_mask

‎README.md

+1-1
Original file line numberDiff line numberDiff line change
@@ -5,7 +5,7 @@
55

66
A collection of lints to catch common mistakes and improve your [Rust](https://github.com/rust-lang/rust) code.
77

8-
[There are 357 lints included in this crate!](https://rust-lang.github.io/rust-clippy/master/index.html)
8+
[There are 358 lints included in this crate!](https://rust-lang.github.io/rust-clippy/master/index.html)
99

1010
We have a bunch of lint categories to allow you to choose how much Clippy is supposed to ~~annoy~~ help you:
1111

‎clippy_lints/src/floating_point_arithmetic.rs

+44-11
Original file line numberDiff line numberDiff line change
@@ -16,6 +16,39 @@ use std::f64::consts as f64_consts;
1616
use sugg::{format_numeric_literal, Sugg};
1717
use syntax::ast;
1818

19+
declare_clippy_lint! {
20+
/// **What it does:** Looks for floating-point expressions that
21+
/// can be expressed using built-in methods to improve accuracy
22+
/// at the cost of performance.
23+
///
24+
/// **Why is this bad?** Negatively impacts accuracy.
25+
///
26+
/// **Known problems:** None
27+
///
28+
/// **Example:**
29+
///
30+
/// ```rust
31+
///
32+
/// let a = 3f32;
33+
/// let _ = a.powf(1.0 / 3.0);
34+
/// let _ = (1.0 + a).ln();
35+
/// let _ = a.exp() - 1.0;
36+
/// ```
37+
///
38+
/// is better expressed as
39+
///
40+
/// ```rust
41+
///
42+
/// let a = 3f32;
43+
/// let _ = a.cbrt();
44+
/// let _ = a.ln_1p();
45+
/// let _ = a.exp_m1();
46+
/// ```
47+
pub IMPRECISE_FLOPS,
48+
nursery,
49+
"usage of imprecise floating point operations"
50+
}
51+
1952
declare_clippy_lint! {
2053
/// **What it does:** Looks for floating-point expressions that
2154
/// can be expressed using built-in methods to improve both
@@ -34,12 +67,9 @@ declare_clippy_lint! {
3467
/// let _ = (2f32).powf(a);
3568
/// let _ = E.powf(a);
3669
/// let _ = a.powf(1.0 / 2.0);
37-
/// let _ = a.powf(1.0 / 3.0);
3870
/// let _ = a.log(2.0);
3971
/// let _ = a.log(10.0);
4072
/// let _ = a.log(E);
41-
/// let _ = (1.0 + a).ln();
42-
/// let _ = a.exp() - 1.0;
4373
/// let _ = a.powf(2.0);
4474
/// let _ = a * 2.0 + 4.0;
4575
/// ```
@@ -53,12 +83,9 @@ declare_clippy_lint! {
5383
/// let _ = a.exp2();
5484
/// let _ = a.exp();
5585
/// let _ = a.sqrt();
56-
/// let _ = a.cbrt();
5786
/// let _ = a.log2();
5887
/// let _ = a.log10();
5988
/// let _ = a.ln();
60-
/// let _ = a.ln_1p();
61-
/// let _ = a.exp_m1();
6289
/// let _ = a.powi(2);
6390
/// let _ = a.mul_add(2.0, 4.0);
6491
/// ```
@@ -67,7 +94,10 @@ declare_clippy_lint! {
6794
"usage of sub-optimal floating point operations"
6895
}
6996

70-
declare_lint_pass!(FloatingPointArithmetic => [SUBOPTIMAL_FLOPS]);
97+
declare_lint_pass!(FloatingPointArithmetic => [
98+
IMPRECISE_FLOPS,
99+
SUBOPTIMAL_FLOPS
100+
]);
71101

72102
// Returns the specialized log method for a given base if base is constant
73103
// and is one of 2, 10 and e
@@ -156,7 +186,7 @@ fn check_ln1p(cx: &LateContext<'_, '_>, expr: &Expr<'_>, args: &[Expr<'_>]) {
156186

157187
span_lint_and_sugg(
158188
cx,
159-
SUBOPTIMAL_FLOPS,
189+
IMPRECISE_FLOPS,
160190
expr.span,
161191
"ln(1 + x) can be computed more accurately",
162192
"consider using",
@@ -215,18 +245,21 @@ fn check_powf(cx: &LateContext<'_, '_>, expr: &Expr<'_>, args: &[Expr<'_>]) {
215245

216246
// Check argument
217247
if let Some((value, _)) = constant(cx, cx.tables, &args[1]) {
218-
let (help, suggestion) = if F32(1.0 / 2.0) == value || F64(1.0 / 2.0) == value {
248+
let (lint, help, suggestion) = if F32(1.0 / 2.0) == value || F64(1.0 / 2.0) == value {
219249
(
250+
SUBOPTIMAL_FLOPS,
220251
"square-root of a number can be computed more efficiently and accurately",
221252
format!("{}.sqrt()", Sugg::hir(cx, &args[0], "..")),
222253
)
223254
} else if F32(1.0 / 3.0) == value || F64(1.0 / 3.0) == value {
224255
(
256+
IMPRECISE_FLOPS,
225257
"cube-root of a number can be computed more accurately",
226258
format!("{}.cbrt()", Sugg::hir(cx, &args[0], "..")),
227259
)
228260
} else if let Some(exponent) = get_integer_from_float_constant(&value) {
229261
(
262+
SUBOPTIMAL_FLOPS,
230263
"exponentiation with integer powers can be computed more efficiently",
231264
format!(
232265
"{}.powi({})",
@@ -240,7 +273,7 @@ fn check_powf(cx: &LateContext<'_, '_>, expr: &Expr<'_>, args: &[Expr<'_>]) {
240273

241274
span_lint_and_sugg(
242275
cx,
243-
SUBOPTIMAL_FLOPS,
276+
lint,
244277
expr.span,
245278
help,
246279
"consider using",
@@ -264,7 +297,7 @@ fn check_expm1(cx: &LateContext<'_, '_>, expr: &Expr<'_>) {
264297
then {
265298
span_lint_and_sugg(
266299
cx,
267-
SUBOPTIMAL_FLOPS,
300+
IMPRECISE_FLOPS,
268301
expr.span,
269302
"(e.pow(x) - 1) can be computed more accurately",
270303
"consider using",

‎clippy_lints/src/lib.rs

+2
Original file line numberDiff line numberDiff line change
@@ -538,6 +538,7 @@ pub fn register_plugins(store: &mut rustc_lint::LintStore, sess: &Session, conf:
538538
&fallible_impl_from::FALLIBLE_IMPL_FROM,
539539
&float_literal::EXCESSIVE_PRECISION,
540540
&float_literal::LOSSY_FLOAT_LITERAL,
541+
&floating_point_arithmetic::IMPRECISE_FLOPS,
541542
&floating_point_arithmetic::SUBOPTIMAL_FLOPS,
542543
&format::USELESS_FORMAT,
543544
&formatting::POSSIBLE_MISSING_COMMA,
@@ -1648,6 +1649,7 @@ pub fn register_plugins(store: &mut rustc_lint::LintStore, sess: &Session, conf:
16481649
store.register_group(true, "clippy::nursery", Some("clippy_nursery"), vec![
16491650
LintId::of(&attrs::EMPTY_LINE_AFTER_OUTER_ATTR),
16501651
LintId::of(&fallible_impl_from::FALLIBLE_IMPL_FROM),
1652+
LintId::of(&floating_point_arithmetic::IMPRECISE_FLOPS),
16511653
LintId::of(&floating_point_arithmetic::SUBOPTIMAL_FLOPS),
16521654
LintId::of(&missing_const_for_fn::MISSING_CONST_FOR_FN),
16531655
LintId::of(&mutable_debug_assertion::DEBUG_ASSERT_WITH_MUT_CALL),

‎src/lintlist/mod.rs

+8-1
Original file line numberDiff line numberDiff line change
@@ -6,7 +6,7 @@ pub use lint::Lint;
66
pub use lint::LINT_LEVELS;
77

88
// begin lint list, do not remove this comment, it’s used in `update_lints`
9-
pub const ALL_LINTS: [Lint; 357] = [
9+
pub const ALL_LINTS: [Lint; 358] = [
1010
Lint {
1111
name: "absurd_extreme_comparisons",
1212
group: "correctness",
@@ -749,6 +749,13 @@ pub const ALL_LINTS: [Lint; 357] = [
749749
deprecation: None,
750750
module: "implicit_return",
751751
},
752+
Lint {
753+
name: "imprecise_flops",
754+
group: "nursery",
755+
desc: "usage of imprecise floating point operations",
756+
deprecation: None,
757+
module: "floating_point_arithmetic",
758+
},
752759
Lint {
753760
name: "inconsistent_digit_grouping",
754761
group: "style",

‎tests/ui/floating_point_exp.fixed

+1-1
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,5 @@
11
// run-rustfix
2-
#![warn(clippy::suboptimal_flops)]
2+
#![warn(clippy::imprecise_flops)]
33

44
fn main() {
55
let x = 2f32;

‎tests/ui/floating_point_exp.rs

+1-1
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,5 @@
11
// run-rustfix
2-
#![warn(clippy::suboptimal_flops)]
2+
#![warn(clippy::imprecise_flops)]
33

44
fn main() {
55
let x = 2f32;

‎tests/ui/floating_point_exp.stderr

+1-1
Original file line numberDiff line numberDiff line change
@@ -4,7 +4,7 @@ error: (e.pow(x) - 1) can be computed more accurately
44
LL | let _ = x.exp() - 1.0;
55
| ^^^^^^^^^^^^^ help: consider using: `x.exp_m1()`
66
|
7-
= note: `-D clippy::suboptimal-flops` implied by `-D warnings`
7+
= note: `-D clippy::imprecise-flops` implied by `-D warnings`
88

99
error: (e.pow(x) - 1) can be computed more accurately
1010
--> $DIR/floating_point_exp.rs:7:13

‎tests/ui/floating_point_log.fixed

+1-1
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
// run-rustfix
22
#![allow(dead_code, clippy::double_parens)]
3-
#![warn(clippy::suboptimal_flops)]
3+
#![warn(clippy::suboptimal_flops, clippy::imprecise_flops)]
44

55
const TWO: f32 = 2.0;
66
const E: f32 = std::f32::consts::E;

‎tests/ui/floating_point_log.rs

+1-1
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
// run-rustfix
22
#![allow(dead_code, clippy::double_parens)]
3-
#![warn(clippy::suboptimal_flops)]
3+
#![warn(clippy::suboptimal_flops, clippy::imprecise_flops)]
44

55
const TWO: f32 = 2.0;
66
const E: f32 = std::f32::consts::E;

‎tests/ui/floating_point_log.stderr

+2
Original file line numberDiff line numberDiff line change
@@ -53,6 +53,8 @@ error: ln(1 + x) can be computed more accurately
5353
|
5454
LL | let _ = (1f32 + 2.).ln();
5555
| ^^^^^^^^^^^^^^^^ help: consider using: `2.0f32.ln_1p()`
56+
|
57+
= note: `-D clippy::imprecise-flops` implied by `-D warnings`
5658

5759
error: ln(1 + x) can be computed more accurately
5860
--> $DIR/floating_point_log.rs:25:13

‎tests/ui/floating_point_powf.fixed

+1-1
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,5 @@
11
// run-rustfix
2-
#![warn(clippy::suboptimal_flops)]
2+
#![warn(clippy::suboptimal_flops, clippy::imprecise_flops)]
33

44
fn main() {
55
let x = 3f32;

‎tests/ui/floating_point_powf.rs

+1-1
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,5 @@
11
// run-rustfix
2-
#![warn(clippy::suboptimal_flops)]
2+
#![warn(clippy::suboptimal_flops, clippy::imprecise_flops)]
33

44
fn main() {
55
let x = 3f32;

‎tests/ui/floating_point_powf.stderr

+2
Original file line numberDiff line numberDiff line change
@@ -47,6 +47,8 @@ error: cube-root of a number can be computed more accurately
4747
|
4848
LL | let _ = x.powf(1.0 / 3.0);
4949
| ^^^^^^^^^^^^^^^^^ help: consider using: `x.cbrt()`
50+
|
51+
= note: `-D clippy::imprecise-flops` implied by `-D warnings`
5052

5153
error: exponentiation with integer powers can be computed more efficiently
5254
--> $DIR/floating_point_powf.rs:14:13

0 commit comments

Comments
 (0)