diff options
Diffstat (limited to 'src/libstd/f64.rs')
| -rw-r--r-- | src/libstd/f64.rs | 125 |
1 files changed, 48 insertions, 77 deletions
diff --git a/src/libstd/f64.rs b/src/libstd/f64.rs index d73d7cd2c7b..f4d804fd508 100644 --- a/src/libstd/f64.rs +++ b/src/libstd/f64.rs @@ -9,8 +9,9 @@ // except according to those terms. //! This module provides constants which are specific to the implementation -//! of the `f64` floating point data type. Mathematically significant -//! numbers are provided in the `consts` sub-module. +//! of the `f64` floating point data type. +//! +//! Mathematically significant numbers are provided in the `consts` sub-module. //! //! *[See also the `f64` primitive type](../primitive.f64.html).* @@ -23,6 +24,8 @@ use core::num; use intrinsics; #[cfg(not(test))] use num::FpCategory; +#[cfg(not(test))] +use sys::cmath; #[stable(feature = "rust1", since = "1.0.0")] pub use core::f64::{RADIX, MANTISSA_DIGITS, DIGITS, EPSILON}; @@ -35,53 +38,6 @@ pub use core::f64::{MIN, MIN_POSITIVE, MAX}; #[stable(feature = "rust1", since = "1.0.0")] pub use core::f64::consts; -#[allow(dead_code)] -mod cmath { - use libc::{c_double, c_int}; - - #[link_name = "m"] - extern { - pub fn acos(n: c_double) -> c_double; - pub fn asin(n: c_double) -> c_double; - pub fn atan(n: c_double) -> c_double; - pub fn atan2(a: c_double, b: c_double) -> c_double; - pub fn cbrt(n: c_double) -> c_double; - pub fn cosh(n: c_double) -> c_double; - pub fn erf(n: c_double) -> c_double; - pub fn erfc(n: c_double) -> c_double; - pub fn expm1(n: c_double) -> c_double; - pub fn fdim(a: c_double, b: c_double) -> c_double; - pub fn fmod(a: c_double, b: c_double) -> c_double; - pub fn frexp(n: c_double, value: &mut c_int) -> c_double; - pub fn ilogb(n: c_double) -> c_int; - pub fn ldexp(x: c_double, n: c_int) -> c_double; - pub fn logb(n: c_double) -> c_double; - pub fn log1p(n: c_double) -> c_double; - pub fn nextafter(x: c_double, y: c_double) -> c_double; - pub fn modf(n: c_double, iptr: &mut c_double) -> c_double; - pub fn sinh(n: c_double) -> c_double; - pub fn tan(n: c_double) -> c_double; - pub fn tanh(n: c_double) -> c_double; - pub fn tgamma(n: c_double) -> c_double; - - // These are commonly only available for doubles - - pub fn j0(n: c_double) -> c_double; - pub fn j1(n: c_double) -> c_double; - pub fn jn(i: c_int, n: c_double) -> c_double; - - pub fn y0(n: c_double) -> c_double; - pub fn y1(n: c_double) -> c_double; - pub fn yn(i: c_int, n: c_double) -> c_double; - - #[cfg_attr(all(windows, target_env = "msvc"), link_name = "__lgamma_r")] - pub fn lgamma_r(n: c_double, sign: &mut c_int) -> c_double; - - #[cfg_attr(all(windows, target_env = "msvc"), link_name = "_hypot")] - pub fn hypot(x: c_double, y: c_double) -> c_double; - } -} - #[cfg(not(test))] #[lang = "f64"] impl f64 { @@ -997,10 +953,13 @@ impl f64 { /// Raw transmutation to `u64`. /// - /// Converts the `f64` into its raw memory representation, - /// similar to the `transmute` function. + /// This is currently identical to `transmute::<f64, u64>(self)` on all platforms. + /// + /// See `from_bits` for some discussion of the portability of this operation + /// (there are almost no issues). /// - /// Note that this function is distinct from casting. + /// Note that this function is distinct from `as` casting, which attempts to + /// preserve the *numeric* value, and not the bitwise value. /// /// # Examples /// @@ -1017,17 +976,33 @@ impl f64 { /// Raw transmutation from `u64`. /// - /// Converts the given `u64` containing the float's raw memory - /// representation into the `f64` type, similar to the - /// `transmute` function. + /// This is currently identical to `transmute::<u64, f64>(v)` on all platforms. + /// It turns out this is incredibly portable, for two reasons: + /// + /// * Floats and Ints have the same endianess on all supported platforms. + /// * IEEE-754 very precisely specifies the bit layout of floats. /// - /// There is only one difference to a bare `transmute`: - /// Due to the implications onto Rust's safety promises being - /// uncertain, if the representation of a signaling NaN "sNaN" float - /// is passed to the function, the implementation is allowed to - /// return a quiet NaN instead. + /// However there is one caveat: prior to the 2008 version of IEEE-754, how + /// to interpret the NaN signaling bit wasn't actually specified. Most platforms + /// (notably x86 and ARM) picked the interpretation that was ultimately + /// standardized in 2008, but some didn't (notably MIPS). As a result, all + /// signaling NaNs on MIPS are quiet NaNs on x86, and vice-versa. /// - /// Note that this function is distinct from casting. + /// Rather than trying to preserve signaling-ness cross-platform, this + /// implementation favours preserving the exact bits. This means that + /// any payloads encoded in NaNs will be preserved even if the result of + /// this method is sent over the network from an x86 machine to a MIPS one. + /// + /// If the results of this method are only manipulated by the same + /// architecture that produced them, then there is no portability concern. + /// + /// If the input isn't NaN, then there is no portability concern. + /// + /// If you don't care about signalingness (very likely), then there is no + /// portability concern. + /// + /// Note that this function is distinct from `as` casting, which attempts to + /// preserve the *numeric* value, and not the bitwise value. /// /// # Examples /// @@ -1036,25 +1011,11 @@ impl f64 { /// let v = f64::from_bits(0x4029000000000000); /// let difference = (v - 12.5).abs(); /// assert!(difference <= 1e-5); - /// // Example for a signaling NaN value: - /// let snan = 0x7FF0000000000001; - /// assert_ne!(f64::from_bits(snan).to_bits(), snan); /// ``` #[stable(feature = "float_bits_conv", since = "1.20.0")] #[inline] - pub fn from_bits(mut v: u64) -> Self { - const EXP_MASK: u64 = 0x7FF0000000000000; - const FRACT_MASK: u64 = 0x000FFFFFFFFFFFFF; - if v & EXP_MASK == EXP_MASK && v & FRACT_MASK != 0 { - // While IEEE 754-2008 specifies encodings for quiet NaNs - // and signaling ones, certain MIPS and PA-RISC - // CPUs treat signaling NaNs differently. - // Therefore to be safe, we pass a known quiet NaN - // if v is any kind of NaN. - // The check above only assumes IEEE 754-1985 to be - // valid. - v = unsafe { ::mem::transmute(NAN) }; - } + pub fn from_bits(v: u64) -> Self { + // It turns out the safety issues with sNaN were overblown! Hooray! unsafe { ::mem::transmute(v) } } } @@ -1641,5 +1602,15 @@ mod tests { assert_approx_eq!(f64::from_bits(0x4029000000000000), 12.5); assert_approx_eq!(f64::from_bits(0x4094e40000000000), 1337.0); assert_approx_eq!(f64::from_bits(0xc02c800000000000), -14.25); + + // Check that NaNs roundtrip their bits regardless of signalingness + // 0xA is 0b1010; 0x5 is 0b0101 -- so these two together clobbers all the mantissa bits + let masked_nan1 = f64::NAN.to_bits() ^ 0x000A_AAAA_AAAA_AAAA; + let masked_nan2 = f64::NAN.to_bits() ^ 0x0005_5555_5555_5555; + assert!(f64::from_bits(masked_nan1).is_nan()); + assert!(f64::from_bits(masked_nan2).is_nan()); + + assert_eq!(f64::from_bits(masked_nan1).to_bits(), masked_nan1); + assert_eq!(f64::from_bits(masked_nan2).to_bits(), masked_nan2); } } |