Some MIR dumps differ between little and big endian

[oli-obk]

in #77413 it was noted that MIR dumps in our test suite somtimes differ between little and big endian, making our test suite fail on big endian. This is just about MIR pretty printing.

All that needs to be done is to pull out

rust/compiler/rustc_mir/src/util/pretty.rs

Lines 432 to 436 in 5ded394

	match ty.kind() {
	ty::Int(_) | ty::Uint(_) | ty::Bool | ty::Char | ty::Float(_) => {}
	// Unit type
	ty::Tuple(tys) if tys.is_empty() => {}
	ty::FnDef(..) => {}

into a separate function that returns true if the verbose printing should be used and then make it recurse on ty::Tuple and ty::Array elements.

You can then run rerun tests separately for 64bit and 32bit (the latter via e.g. --target i686-unknown-linux-gnu)

[hencrice]

@rustbot claim

I would like to give this a try as my first Rust contribution :)

[infinity0]

This is still not fixed, I see the test failures still in 1.51 https://buildd.debian.org/status/fetch.php?pkg=rustc&arch=s390x&ver=1.51.0%2Bdfsg1-1%7Eexp2&stamp=1619448544&raw=0

[infinity0]

e.g. on s390x:

-	+                                          // + literal: Const { ty: (u32, bool), val: Value(ByRef { alloc: Allocation { bytes: [2, 0, 0, 0, 0, 0, 0, 0], relocations: Relocations(SortedMap { data: [] }), init_mask: InitMask { blocks: [31], len: Size { raw: 8 } }, size: Size { raw: 8 }, align: Align { pow2: 2 }, mutability: Not, extra: () }, offset: Size { raw: 0 } }) }
+	+                                          // + literal: Const { ty: (u32, bool), val: Value(ByRef { alloc: Allocation { bytes: [0, 0, 0, 2, 0, 0, 0, 0], relocations: Relocations(SortedMap { data: [] }), init_mask: InitMask { blocks: [31], len: Size { raw: 8 } }, size: Size { raw: 8 }, align: Align { pow2: 2 }, mutability: Not, extra: () }, offset: Size { raw: 0 } }) }

[oli-obk]

we need the same fix (as the one in #79999 ) for mir::Constant

[fee1-dead]

@rustbot claim