diff --git a/libm/src/math/copysignf.rs b/libm/src/math/copysignf.rs
deleted file mode 100644
index 8b9bed4c..00000000
--- a/libm/src/math/copysignf.rs
+++ /dev/null
@@ -1,8 +0,0 @@
-/// Sign of Y, magnitude of X (f32)
-///
-/// Constructs a number with the magnitude (absolute value) of its
-/// first argument, `x`, and the sign of its second argument, `y`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn copysignf(x: f32, y: f32) -> f32 {
-    super::generic::copysign(x, y)
-}
diff --git a/libm/src/math/copysignf128.rs b/libm/src/math/copysignf128.rs
deleted file mode 100644
index 7bd81d42..00000000
--- a/libm/src/math/copysignf128.rs
+++ /dev/null
@@ -1,8 +0,0 @@
-/// Sign of Y, magnitude of X (f128)
-///
-/// Constructs a number with the magnitude (absolute value) of its
-/// first argument, `x`, and the sign of its second argument, `y`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn copysignf128(x: f128, y: f128) -> f128 {
-    super::generic::copysign(x, y)
-}
diff --git a/libm/src/math/copysignf16.rs b/libm/src/math/copysignf16.rs
deleted file mode 100644
index 82065868..00000000
--- a/libm/src/math/copysignf16.rs
+++ /dev/null
@@ -1,8 +0,0 @@
-/// Sign of Y, magnitude of X (f16)
-///
-/// Constructs a number with the magnitude (absolute value) of its
-/// first argument, `x`, and the sign of its second argument, `y`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn copysignf16(x: f16, y: f16) -> f16 {
-    super::generic::copysign(x, y)
-}
diff --git a/libm/src/math/fabsf.rs b/libm/src/math/fabsf.rs
deleted file mode 100644
index e5820a26..00000000
--- a/libm/src/math/fabsf.rs
+++ /dev/null
@@ -1,39 +0,0 @@
-/// Absolute value (magnitude) (f32)
-///
-/// Calculates the absolute value (magnitude) of the argument `x`,
-/// by direct manipulation of the bit representation of `x`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn fabsf(x: f32) -> f32 {
-    select_implementation! {
-        name: fabsf,
-        use_arch: all(target_arch = "wasm32", intrinsics_enabled),
-        args: x,
-    }
-
-    super::generic::fabs(x)
-}
-
-// PowerPC tests are failing on LLVM 13: https://github.com/rust-lang/rust/issues/88520
-#[cfg(not(target_arch = "powerpc64"))]
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn sanity_check() {
-        assert_eq!(fabsf(-1.0), 1.0);
-        assert_eq!(fabsf(2.8), 2.8);
-    }
-
-    /// The spec: https://en.cppreference.com/w/cpp/numeric/math/fabs
-    #[test]
-    fn spec_tests() {
-        assert!(fabsf(f32::NAN).is_nan());
-        for f in [0.0, -0.0].iter().copied() {
-            assert_eq!(fabsf(f), 0.0);
-        }
-        for f in [f32::INFINITY, f32::NEG_INFINITY].iter().copied() {
-            assert_eq!(fabsf(f), f32::INFINITY);
-        }
-    }
-}
diff --git a/libm/src/math/fabsf128.rs b/libm/src/math/fabsf128.rs
deleted file mode 100644
index 46429ca4..00000000
--- a/libm/src/math/fabsf128.rs
+++ /dev/null
@@ -1,31 +0,0 @@
-/// Absolute value (magnitude) (f128)
-///
-/// Calculates the absolute value (magnitude) of the argument `x`,
-/// by direct manipulation of the bit representation of `x`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn fabsf128(x: f128) -> f128 {
-    super::generic::fabs(x)
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn sanity_check() {
-        assert_eq!(fabsf128(-1.0), 1.0);
-        assert_eq!(fabsf128(2.8), 2.8);
-    }
-
-    /// The spec: https://en.cppreference.com/w/cpp/numeric/math/fabs
-    #[test]
-    fn spec_tests() {
-        assert!(fabsf128(f128::NAN).is_nan());
-        for f in [0.0, -0.0].iter().copied() {
-            assert_eq!(fabsf128(f), 0.0);
-        }
-        for f in [f128::INFINITY, f128::NEG_INFINITY].iter().copied() {
-            assert_eq!(fabsf128(f), f128::INFINITY);
-        }
-    }
-}
diff --git a/libm/src/math/fabsf16.rs b/libm/src/math/fabsf16.rs
deleted file mode 100644
index eee42ac6..00000000
--- a/libm/src/math/fabsf16.rs
+++ /dev/null
@@ -1,31 +0,0 @@
-/// Absolute value (magnitude) (f16)
-///
-/// Calculates the absolute value (magnitude) of the argument `x`,
-/// by direct manipulation of the bit representation of `x`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn fabsf16(x: f16) -> f16 {
-    super::generic::fabs(x)
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn sanity_check() {
-        assert_eq!(fabsf16(-1.0), 1.0);
-        assert_eq!(fabsf16(2.8), 2.8);
-    }
-
-    /// The spec: https://en.cppreference.com/w/cpp/numeric/math/fabs
-    #[test]
-    fn spec_tests() {
-        assert!(fabsf16(f16::NAN).is_nan());
-        for f in [0.0, -0.0].iter().copied() {
-            assert_eq!(fabsf16(f), 0.0);
-        }
-        for f in [f16::INFINITY, f16::NEG_INFINITY].iter().copied() {
-            assert_eq!(fabsf16(f), f16::INFINITY);
-        }
-    }
-}
diff --git a/libm/src/math/fdimf.rs b/libm/src/math/fdimf.rs
deleted file mode 100644
index 367ef517..00000000
--- a/libm/src/math/fdimf.rs
+++ /dev/null
@@ -1,12 +0,0 @@
-/// Positive difference (f32)
-///
-/// Determines the positive difference between arguments, returning:
-/// * x - y if x > y, or
-/// * +0    if x <= y, or
-/// * NAN   if either argument is NAN.
-///
-/// A range error may occur.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn fdimf(x: f32, y: f32) -> f32 {
-    super::generic::fdim(x, y)
-}
diff --git a/libm/src/math/fdimf128.rs b/libm/src/math/fdimf128.rs
deleted file mode 100644
index 6f3d1d0f..00000000
--- a/libm/src/math/fdimf128.rs
+++ /dev/null
@@ -1,12 +0,0 @@
-/// Positive difference (f128)
-///
-/// Determines the positive difference between arguments, returning:
-/// * x - y if x > y, or
-/// * +0    if x <= y, or
-/// * NAN   if either argument is NAN.
-///
-/// A range error may occur.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn fdimf128(x: f128, y: f128) -> f128 {
-    super::generic::fdim(x, y)
-}
diff --git a/libm/src/math/fdimf16.rs b/libm/src/math/fdimf16.rs
deleted file mode 100644
index 37bd6885..00000000
--- a/libm/src/math/fdimf16.rs
+++ /dev/null
@@ -1,12 +0,0 @@
-/// Positive difference (f16)
-///
-/// Determines the positive difference between arguments, returning:
-/// * x - y if x > y, or
-/// * +0    if x <= y, or
-/// * NAN   if either argument is NAN.
-///
-/// A range error may occur.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn fdimf16(x: f16, y: f16) -> f16 {
-    super::generic::fdim(x, y)
-}
diff --git a/libm/src/math/floorf.rs b/libm/src/math/floorf.rs
deleted file mode 100644
index 16957b7f..00000000
--- a/libm/src/math/floorf.rs
+++ /dev/null
@@ -1,13 +0,0 @@
-/// Floor (f32)
-///
-/// Finds the nearest integer less than or equal to `x`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn floorf(x: f32) -> f32 {
-    select_implementation! {
-        name: floorf,
-        use_arch: all(target_arch = "wasm32", intrinsics_enabled),
-        args: x,
-    }
-
-    return super::generic::floor(x);
-}
diff --git a/libm/src/math/floorf128.rs b/libm/src/math/floorf128.rs
deleted file mode 100644
index 9a9fe415..00000000
--- a/libm/src/math/floorf128.rs
+++ /dev/null
@@ -1,7 +0,0 @@
-/// Floor (f128)
-///
-/// Finds the nearest integer less than or equal to `x`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn floorf128(x: f128) -> f128 {
-    return super::generic::floor(x);
-}
diff --git a/libm/src/math/floorf16.rs b/libm/src/math/floorf16.rs
deleted file mode 100644
index f9b868e0..00000000
--- a/libm/src/math/floorf16.rs
+++ /dev/null
@@ -1,7 +0,0 @@
-/// Floor (f16)
-///
-/// Finds the nearest integer less than or equal to `x`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn floorf16(x: f16) -> f16 {
-    return super::generic::floor(x);
-}
diff --git a/libm/src/math/fmodf.rs b/libm/src/math/fmodf.rs
deleted file mode 100644
index 4e95696e..00000000
--- a/libm/src/math/fmodf.rs
+++ /dev/null
@@ -1,5 +0,0 @@
-/// Calculate the remainder of `x / y`, the precise result of `x - trunc(x / y) * y`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn fmodf(x: f32, y: f32) -> f32 {
-    super::generic::fmod(x, y)
-}
diff --git a/libm/src/math/fmodf128.rs b/libm/src/math/fmodf128.rs
deleted file mode 100644
index ff0e0493..00000000
--- a/libm/src/math/fmodf128.rs
+++ /dev/null
@@ -1,5 +0,0 @@
-/// Calculate the remainder of `x / y`, the precise result of `x - trunc(x / y) * y`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn fmodf128(x: f128, y: f128) -> f128 {
-    super::generic::fmod(x, y)
-}
diff --git a/libm/src/math/fmodf16.rs b/libm/src/math/fmodf16.rs
deleted file mode 100644
index 11972a7d..00000000
--- a/libm/src/math/fmodf16.rs
+++ /dev/null
@@ -1,5 +0,0 @@
-/// Calculate the remainder of `x / y`, the precise result of `x - trunc(x / y) * y`.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn fmodf16(x: f16, y: f16) -> f16 {
-    super::generic::fmod(x, y)
-}
diff --git a/libm/src/math/ldexpf.rs b/libm/src/math/ldexpf.rs
deleted file mode 100644
index 95b27fc4..00000000
--- a/libm/src/math/ldexpf.rs
+++ /dev/null
@@ -1,4 +0,0 @@
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn ldexpf(x: f32, n: i32) -> f32 {
-    super::scalbnf(x, n)
-}
diff --git a/libm/src/math/ldexpf128.rs b/libm/src/math/ldexpf128.rs
deleted file mode 100644
index b35277d1..00000000
--- a/libm/src/math/ldexpf128.rs
+++ /dev/null
@@ -1,4 +0,0 @@
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn ldexpf128(x: f128, n: i32) -> f128 {
-    super::scalbnf128(x, n)
-}
diff --git a/libm/src/math/ldexpf16.rs b/libm/src/math/ldexpf16.rs
deleted file mode 100644
index 8de6cffd..00000000
--- a/libm/src/math/ldexpf16.rs
+++ /dev/null
@@ -1,4 +0,0 @@
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn ldexpf16(x: f16, n: i32) -> f16 {
-    super::scalbnf16(x, n)
-}
diff --git a/libm/src/math/roundf.rs b/libm/src/math/roundf.rs
deleted file mode 100644
index b5d7c9d6..00000000
--- a/libm/src/math/roundf.rs
+++ /dev/null
@@ -1,5 +0,0 @@
-/// Round `x` to the nearest integer, breaking ties away from zero.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn roundf(x: f32) -> f32 {
-    super::generic::round(x)
-}
diff --git a/libm/src/math/roundf128.rs b/libm/src/math/roundf128.rs
deleted file mode 100644
index fc316492..00000000
--- a/libm/src/math/roundf128.rs
+++ /dev/null
@@ -1,5 +0,0 @@
-/// Round `x` to the nearest integer, breaking ties away from zero.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn roundf128(x: f128) -> f128 {
-    super::generic::round(x)
-}
diff --git a/libm/src/math/roundf16.rs b/libm/src/math/roundf16.rs
deleted file mode 100644
index 8b356eaa..00000000
--- a/libm/src/math/roundf16.rs
+++ /dev/null
@@ -1,5 +0,0 @@
-/// Round `x` to the nearest integer, breaking ties away from zero.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn roundf16(x: f16) -> f16 {
-    super::generic::round(x)
-}
diff --git a/libm/src/math/scalbnf.rs b/libm/src/math/scalbnf.rs
deleted file mode 100644
index 57e7ba76..00000000
--- a/libm/src/math/scalbnf.rs
+++ /dev/null
@@ -1,4 +0,0 @@
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn scalbnf(x: f32, n: i32) -> f32 {
-    super::generic::scalbn(x, n)
-}
diff --git a/libm/src/math/scalbnf128.rs b/libm/src/math/scalbnf128.rs
deleted file mode 100644
index c1d2b485..00000000
--- a/libm/src/math/scalbnf128.rs
+++ /dev/null
@@ -1,4 +0,0 @@
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn scalbnf128(x: f128, n: i32) -> f128 {
-    super::generic::scalbn(x, n)
-}
diff --git a/libm/src/math/scalbnf16.rs b/libm/src/math/scalbnf16.rs
deleted file mode 100644
index 2209e1a1..00000000
--- a/libm/src/math/scalbnf16.rs
+++ /dev/null
@@ -1,4 +0,0 @@
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn scalbnf16(x: f16, n: i32) -> f16 {
-    super::generic::scalbn(x, n)
-}
diff --git a/libm/src/math/sqrtf.rs b/libm/src/math/sqrtf.rs
deleted file mode 100644
index c28a705e..00000000
--- a/libm/src/math/sqrtf.rs
+++ /dev/null
@@ -1,15 +0,0 @@
-/// The square root of `x` (f32).
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn sqrtf(x: f32) -> f32 {
-    select_implementation! {
-        name: sqrtf,
-        use_arch: any(
-            all(target_arch = "aarch64", target_feature = "neon"),
-            all(target_arch = "wasm32", intrinsics_enabled),
-            target_feature = "sse2"
-        ),
-        args: x,
-    }
-
-    super::generic::sqrt(x)
-}
diff --git a/libm/src/math/sqrtf128.rs b/libm/src/math/sqrtf128.rs
deleted file mode 100644
index eaef6ae0..00000000
--- a/libm/src/math/sqrtf128.rs
+++ /dev/null
@@ -1,5 +0,0 @@
-/// The square root of `x` (f128).
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn sqrtf128(x: f128) -> f128 {
-    return super::generic::sqrt(x);
-}
diff --git a/libm/src/math/sqrtf16.rs b/libm/src/math/sqrtf16.rs
deleted file mode 100644
index 7bedb7f8..00000000
--- a/libm/src/math/sqrtf16.rs
+++ /dev/null
@@ -1,11 +0,0 @@
-/// The square root of `x` (f16).
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn sqrtf16(x: f16) -> f16 {
-    select_implementation! {
-        name: sqrtf16,
-        use_arch: all(target_arch = "aarch64", target_feature = "fp16"),
-        args: x,
-    }
-
-    return super::generic::sqrt(x);
-}
diff --git a/libm/src/math/truncf.rs b/libm/src/math/truncf.rs
deleted file mode 100644
index 14533a26..00000000
--- a/libm/src/math/truncf.rs
+++ /dev/null
@@ -1,23 +0,0 @@
-/// Rounds the number toward 0 to the closest integral value (f32).
-///
-/// This effectively removes the decimal part of the number, leaving the integral part.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn truncf(x: f32) -> f32 {
-    select_implementation! {
-        name: truncf,
-        use_arch: all(target_arch = "wasm32", intrinsics_enabled),
-        args: x,
-    }
-
-    super::generic::trunc(x)
-}
-
-// PowerPC tests are failing on LLVM 13: https://github.com/rust-lang/rust/issues/88520
-#[cfg(not(target_arch = "powerpc64"))]
-#[cfg(test)]
-mod tests {
-    #[test]
-    fn sanity_check() {
-        assert_eq!(super::truncf(1.1), 1.0);
-    }
-}
diff --git a/libm/src/math/truncf128.rs b/libm/src/math/truncf128.rs
deleted file mode 100644
index 9dccc0d0..00000000
--- a/libm/src/math/truncf128.rs
+++ /dev/null
@@ -1,7 +0,0 @@
-/// Rounds the number toward 0 to the closest integral value (f128).
-///
-/// This effectively removes the decimal part of the number, leaving the integral part.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn truncf128(x: f128) -> f128 {
-    super::generic::trunc(x)
-}
diff --git a/libm/src/math/truncf16.rs b/libm/src/math/truncf16.rs
deleted file mode 100644
index d7c3d225..00000000
--- a/libm/src/math/truncf16.rs
+++ /dev/null
@@ -1,7 +0,0 @@
-/// Rounds the number toward 0 to the closest integral value (f16).
-///
-/// This effectively removes the decimal part of the number, leaving the integral part.
-#[cfg_attr(all(test, assert_no_panic), no_panic::no_panic)]
-pub fn truncf16(x: f16) -> f16 {
-    super::generic::trunc(x)
-}