Fix FloatingPoint.frexp returning unsigned exponent

C's frexp writes a signed int for the exponent. The Pony wrapper
declared it as U32, silently reinterpreting negative exponents as
large positive values.

Closes #5103

Author: SeanTAllen

.release-notes/fix-frexp-signed-exponent.md

@@ -0,0 +1,13 @@
+## Fix `FloatingPoint.frexp` returning unsigned exponent
+
+`FloatingPoint.frexp` (and its implementations on `F32` and `F64`) returned the exponent as `U32` when C's `frexp` writes a signed `int`. Negative exponents were silently reinterpreted as large positive values.
+
+The return type is now `(A, I32)` instead of `(A, U32)`. If you destructure the result and type the exponent, update it:
+
+```pony
+// Before
+(let mantissa, let exp: U32) = my_float.frexp()
+
+// After
+(let mantissa, let exp: I32) = my_float.frexp()
+```

packages/builtin/float.pony

@@ -56,7 +56,7 @@ use @asinh[F64](x: F64)
 use @atanh[F64](x: F64)
 use @"llvm.copysign.f32"[F32](x: F32, sign: F32)
 use @"llvm.copysign.f64"[F64](x: F64, sign: F64)
-use @frexp[F64](value: F64, exponent: Pointer[U32])
+use @frexp[F64](value: F64, exponent: Pointer[I32])
 use @ldexpf[F32](value: F32, exponent: I32)
 use @ldexp[F64](value: F64, exponent: I32)
 
@@ -202,8 +202,8 @@ primitive F32 is FloatingPoint[F32]
   fun ldexp(x: F32, exponent: I32): F32 =>
     @ldexpf(x, exponent)
 
-  fun frexp(): (F32, U32) =>
-    var exponent: U32 = 0
+  fun frexp(): (F32, I32) =>
+    var exponent: I32 = 0
     var mantissa = @frexp(f64(), addressof exponent)
     (mantissa.f32(), exponent)
 
@@ -419,8 +419,8 @@ primitive F64 is FloatingPoint[F64]
   fun ldexp(x: F64, exponent: I32): F64 =>
     @ldexp(x, exponent)
 
-  fun frexp(): (F64, U32) =>
-    var exponent: U32 = 0
+  fun frexp(): (F64, I32) =>
+    var exponent: I32 = 0
     var mantissa = @frexp(this, addressof exponent)
     (mantissa, exponent)
 

packages/builtin/real.pony

@@ -668,7 +668,7 @@ trait val FloatingPoint[A: FloatingPoint[A] val] is Real[A]
   fun nan(): Bool
 
   fun ldexp(x: A, exponent: I32): A
-  fun frexp(): (A, U32)
+  fun frexp(): (A, I32)
   fun log(): A
   fun log2(): A
   fun log10(): A