Auto merge of #128213 - matthiaskrgr:rollup-v40q1j3, r=matthiaskrgr

Rollup of 7 pull requests

Successful merges:

 - #126090 (Fix supertrait associated type unsoundness)
 - #127220 (Graciously handle `Drop` impls introducing more generic parameters than the ADT)
 - #127950 (Use `#[rustfmt::skip]` on some `use` groups to prevent reordering.)
 - #128085 (Various notes on match lowering)
 - #128150 (Stop using `unsized_const_parameters` in core/std)
 - #128194 (LLVM: LLVM-20.0 removes MMX types)
 - #128211 (fix: compilation issue w/ refactored type)

r? `@ghost`
`@rustbot` modify labels: rollup

Author: bors

tests/pass/intrinsics/portable-simd.rs

@@ -1,10 +1,23 @@
 //@compile-flags: -Zmiri-strict-provenance
-#![feature(portable_simd, adt_const_params, core_intrinsics, repr_simd)]
+#![feature(
+    portable_simd,
+    unsized_const_params,
+    adt_const_params,
+    rustc_attrs,
+    intrinsics,
+    core_intrinsics,
+    repr_simd
+)]
 #![allow(incomplete_features, internal_features)]
 use std::intrinsics::simd as intrinsics;
 use std::ptr;
 use std::simd::{prelude::*, StdFloat};
 
+extern "rust-intrinsic" {
+    #[rustc_nounwind]
+    pub fn simd_shuffle_generic<T, U, const IDX: &'static [u32]>(x: T, y: T) -> U;
+}
+
 fn simd_ops_f32() {
     let a = f32x4::splat(10.0);
     let b = f32x4::from_array([1.0, 2.0, 3.0, -4.0]);