Auto merge of #122459 - Nadrieril:sort-eq, r=oli-obk

match lowering: sort `Eq` candidates in the failure case too

This is a slight tweak to MIR gen of matches. Take a match like:
```rust
match (s, flag) {
    ("a", _) if foo() => 1,
    ("b", true) => 2,
    ("a", false) => 3,
    (_, true) => 4,
    _ => 5,
}
```
If we switch on `s == "a"`, the first candidate matches, and we learn almost nothing about the second candidate. So there's a choice:
1. (what we do today) stop sorting candidates, keep the "b" case grouped with everything below. This could allow us to be clever here and test on `flag == true` next.
2. (what this PR does) sort "b" into the failure case. The "b" will be alone (fewer opportunities for picking a good test), but that means the two "a" cases require a single test.

Today, we aren't clever in which tests we pick, so this is an unambiguous win. In a future where we pick tests better, idk. Grouping tests as much as possible feels like a generally good strategy.

This was proposed in #29623 (9 years ago :D)

Author: bors

compiler/rustc_mir_build/src/build/matches/test.rs

@@ -650,12 +650,14 @@ impl<'a, 'tcx> Builder<'a, 'tcx> {
                 }
             }
 
-            // FIXME(#29623): return `Some(1)` when the values are different.
-            (TestKind::Eq { value: test_val, .. }, TestCase::Constant { value: case_val })
-                if test_val == case_val =>
-            {
-                fully_matched = true;
-                Some(TestBranch::Success)
+            (TestKind::Eq { value: test_val, .. }, TestCase::Constant { value: case_val }) => {
+                if test_val == case_val {
+                    fully_matched = true;
+                    Some(TestBranch::Success)
+                } else {
+                    fully_matched = false;
+                    Some(TestBranch::Failure)
+                }
             }
 
             (