Rollup merge of #74888 - infinity0:ignore-endian-big, r=nikomatsakis

compiletest: ignore-endian-big, fixes #74829, fixes #74885

See discussion on #74829

I tested it on a Debian s390x machine, works well.

src/test/mir-opt/const-promotion-extern-static.rs

@@ -1,7 +1,7 @@
+// ignore-endian-big
 extern "C" {
     static X: i32;
 }
-
 static Y: i32 = 42;
 
 // EMIT_MIR const_promotion_extern_static.BAR.PromoteTemps.diff

src/test/mir-opt/const_allocation.rs

@@ -1,5 +1,5 @@
+// ignore-endian-big
 // EMIT_MIR_FOR_EACH_BIT_WIDTH
-
 static FOO: &[(Option<i32>, &[&str])] =
     &[(None, &[]), (None, &["foo", "bar"]), (Some(42), &["meh", "mop", "möp"])];
 

src/test/mir-opt/const_allocation2.rs

@@ -1,5 +1,5 @@
+// ignore-endian-big
 // EMIT_MIR_FOR_EACH_BIT_WIDTH
-
 // EMIT_MIR const_allocation2.main.ConstProp.after.mir
 fn main() {
     FOO;

src/test/mir-opt/const_allocation3.rs

@@ -1,5 +1,5 @@
+// ignore-endian-big
 // EMIT_MIR_FOR_EACH_BIT_WIDTH
-
 // EMIT_MIR const_allocation3.main.ConstProp.after.mir
 fn main() {
     FOO;

src/test/mir-opt/inline/inline-into-box-place.rs

@@ -1,8 +1,8 @@
+// ignore-endian-big
 // ignore-wasm32-bare compiled with panic=abort by default
 // compile-flags: -Z mir-opt-level=3
 // EMIT_MIR_FOR_EACH_BIT_WIDTH
 #![feature(box_syntax)]
-
 // EMIT_MIR inline_into_box_place.main.Inline.diff
 fn main() {
     let _x: Box<Vec<u32>> = box Vec::new();

src/test/ui/simd/simd-intrinsic-generic-bitmask.rs

@@ -2,6 +2,7 @@
 #![allow(non_camel_case_types)]
 
 // ignore-emscripten
+// ignore-endian-big behavior of simd_bitmask is endian-specific
 
 // Test that the simd_bitmask intrinsic produces correct results.
 

src/test/ui/simd/simd-intrinsic-generic-select.rs

@@ -2,10 +2,7 @@
 #![allow(non_camel_case_types)]
 
 // ignore-emscripten
-// ignore-mips       behavior of simd_select_bitmask is endian-specific
-// ignore-mips64     behavior of simd_select_bitmask is endian-specific
-// ignore-powerpc    behavior of simd_select_bitmask is endian-specific
-// ignore-powerpc64  behavior of simd_select_bitmask is endian-specific
+// ignore-endian-big behavior of simd_select_bitmask is endian-specific
 
 // Test that the simd_select intrinsics produces correct results.