Rollup merge of #129835 - RalfJung:float-tests, r=workingjubilee

enable const-float-classify test, and test_next_up/down on 32bit x86

The  test_next_up/down tests have been disabled on all 32bit x86 targets, which goes too far -- they should definitely work on our (tier 1) i686 target, it is only without SSE that we might run into trouble due to #114479. However, I cannot reproduce that trouble any more -- maybe that got fixed by #123351?

The  const-float-classify test relied on const traits "because we can", and got disabled when const traits got removed. That's an unfortunate reduction in test coverage of our float functionality, so let's restore the test in a way that does not rely on const traits.

The const-float tests are actually testing runtime behavior as well, and I don't think that runtime behavior is covered anywhere else. Probably they shouldn't be called "const-float", but we don't have a `tests/ui/float` folder... should I create one and move them there? Are there any other ui tests that should be moved there?

I also removed some FIXME referring to not use x87 for Rust-to-Rust-calls -- that has happened in #123351 so this got fixed indeed. Does that mean we can simplify all that float code again? I am not sure how to test it. Is running the test suite with an i586 target enough?

Cc ```@tgross35``` ```@workingjubilee```

library/core/src/num/f16.rs

@@ -435,6 +435,7 @@ impl f16 {
         // WASM, see llvm/llvm-project#96437). These are platforms bugs, and Rust will misbehave on
         // such platforms, but we can at least try to make things seem as sane as possible by being
         // careful here.
+        // see also https://github.com/rust-lang/rust/issues/114479
         if self.is_infinite() {
             // Thus, a value may compare unequal to infinity, despite having a "full" exponent mask.
             FpCategory::Infinite

library/core/src/num/f32.rs

@@ -662,10 +662,7 @@ impl f32 {
         // hardware flushes subnormals to zero. These are platforms bugs, and Rust will misbehave on
         // such hardware, but we can at least try to make things seem as sane as possible by being
         // careful here.
-        //
-        // FIXME(jubilee): Using x87 operations is never necessary in order to function
-        // on x86 processors for Rust-to-Rust calls, so this issue should not happen.
-        // Code generation should be adjusted to use non-C calling conventions, avoiding this.
+        // see also https://github.com/rust-lang/rust/issues/114479
         if self.is_infinite() {
             // A value may compare unequal to infinity, despite having a "full" exponent mask.
             FpCategory::Infinite

library/core/src/num/f64.rs

@@ -660,10 +660,7 @@ impl f64 {
         // float semantics Rust relies on: x87 uses a too-large exponent, and some hardware flushes
         // subnormals to zero. These are platforms bugs, and Rust will misbehave on such hardware,
         // but we can at least try to make things seem as sane as possible by being careful here.
-        //
-        // FIXME(jubilee): Using x87 operations is never necessary in order to function
-        // on x86 processors for Rust-to-Rust calls, so this issue should not happen.
-        // Code generation should be adjusted to use non-C calling conventions, avoiding this.
+        // see also https://github.com/rust-lang/rust/issues/114479
         //
         // Thus, a value may compare unequal to infinity, despite having a "full" exponent mask.
         // And it may not be NaN, as it can simply be an "overextended" finite value.

library/std/src/f32/tests.rs

@@ -2,31 +2,24 @@ use crate::f32::consts;
 use crate::num::{FpCategory as Fp, *};
 
 /// Smallest number
-#[allow(dead_code)] // unused on x86
 const TINY_BITS: u32 = 0x1;
 
 /// Next smallest number
-#[allow(dead_code)] // unused on x86
 const TINY_UP_BITS: u32 = 0x2;
 
 /// Exponent = 0b11...10, Sifnificand 0b1111..10. Min val > 0
-#[allow(dead_code)] // unused on x86
 const MAX_DOWN_BITS: u32 = 0x7f7f_fffe;
 
 /// Zeroed exponent, full significant
-#[allow(dead_code)] // unused on x86
 const LARGEST_SUBNORMAL_BITS: u32 = 0x007f_ffff;
 
 /// Exponent = 0b1, zeroed significand
-#[allow(dead_code)] // unused on x86
 const SMALLEST_NORMAL_BITS: u32 = 0x0080_0000;
 
 /// First pattern over the mantissa
-#[allow(dead_code)] // unused on x86
 const NAN_MASK1: u32 = 0x002a_aaaa;
 
 /// Second pattern over the mantissa
-#[allow(dead_code)] // unused on x86
 const NAN_MASK2: u32 = 0x0055_5555;
 
 #[allow(unused_macros)]
@@ -353,9 +346,6 @@ fn test_is_sign_negative() {
     assert!((-f32::NAN).is_sign_negative());
 }
 
-// Ignore test on x87 floating point, these platforms do not guarantee NaN
-// payloads are preserved and flush denormals to zero, failing the tests.
-#[cfg(not(target_arch = "x86"))]
 #[test]
 fn test_next_up() {
     let tiny = f32::from_bits(TINY_BITS);
@@ -386,9 +376,6 @@ fn test_next_up() {
     assert_f32_biteq!(nan2.next_up(), nan2);
 }
 
-// Ignore test on x87 floating point, these platforms do not guarantee NaN
-// payloads are preserved and flush denormals to zero, failing the tests.
-#[cfg(not(target_arch = "x86"))]
 #[test]
 fn test_next_down() {
     let tiny = f32::from_bits(TINY_BITS);

library/std/src/f64/tests.rs

@@ -2,31 +2,24 @@ use crate::f64::consts;
 use crate::num::{FpCategory as Fp, *};
 
 /// Smallest number
-#[allow(dead_code)] // unused on x86
 const TINY_BITS: u64 = 0x1;
 
 /// Next smallest number
-#[allow(dead_code)] // unused on x86
 const TINY_UP_BITS: u64 = 0x2;
 
 /// Exponent = 0b11...10, Sifnificand 0b1111..10. Min val > 0
-#[allow(dead_code)] // unused on x86
 const MAX_DOWN_BITS: u64 = 0x7fef_ffff_ffff_fffe;
 
 /// Zeroed exponent, full significant
-#[allow(dead_code)] // unused on x86
 const LARGEST_SUBNORMAL_BITS: u64 = 0x000f_ffff_ffff_ffff;
 
 /// Exponent = 0b1, zeroed significand
-#[allow(dead_code)] // unused on x86
 const SMALLEST_NORMAL_BITS: u64 = 0x0010_0000_0000_0000;
 
 /// First pattern over the mantissa
-#[allow(dead_code)] // unused on x86
 const NAN_MASK1: u64 = 0x000a_aaaa_aaaa_aaaa;
 
 /// Second pattern over the mantissa
-#[allow(dead_code)] // unused on x86
 const NAN_MASK2: u64 = 0x0005_5555_5555_5555;
 
 #[allow(unused_macros)]
@@ -343,9 +336,6 @@ fn test_is_sign_negative() {
     assert!((-f64::NAN).is_sign_negative());
 }
 
-// Ignore test on x87 floating point, these platforms do not guarantee NaN
-// payloads are preserved and flush denormals to zero, failing the tests.
-#[cfg(not(target_arch = "x86"))]
 #[test]
 fn test_next_up() {
     let tiny = f64::from_bits(TINY_BITS);
@@ -375,9 +365,6 @@ fn test_next_up() {
     assert_f64_biteq!(nan2.next_up(), nan2);
 }
 
-// Ignore test on x87 floating point, these platforms do not guarantee NaN
-// payloads are preserved and flush denormals to zero, failing the tests.
-#[cfg(not(target_arch = "x86"))]
 #[test]
 fn test_next_down() {
     let tiny = f64::from_bits(TINY_BITS);