In a table-driven test, a developer runs subtests in a loop using `t.Run`. They add `t.Parallel()` inside each subtest to speed things up. With the loop written as below (Go 1.21), what is the classic bug?
func TestThings(t *testing.T) {
cases := []struct{ name string; in, want int }{
{"a", 1, 2}, {"b", 2, 3},
}
for _, tc := range cases {
t.Run(tc.name, func(t *testing.T) {
t.Parallel()
if got := tc.in + 1; got != tc.want {
t.Errorf("%s: got %d", tc.name, got)
}
})
}
}