| //@ compile-flags:-g | |
| //@ disable-gdb-pretty-printers | |
| //@ ignore-backends: gcc | |
| //@ gdb-command:set print union on | |
| //@ gdb-command:run | |
| //@ gdb-command:print case1 | |
| //@ gdb-check:$1 = generic_struct_style_enum::Regular<u16, u32, i64>::Case1{a: 0, b: 31868, c: 31868, d: 31868, e: 31868} | |
| //@ gdb-command:print case2 | |
| //@ gdb-check:$2 = generic_struct_style_enum::Regular<i16, u32, i64>::Case2{a: 0, b: 286331153, c: 286331153} | |
| //@ gdb-command:print case3 | |
| //@ gdb-check:$3 = generic_struct_style_enum::Regular<u16, i32, u64>::Case3{a: 0, b: 6438275382588823897} | |
| //@ gdb-command:print univariant | |
| //@ gdb-check:$4 = generic_struct_style_enum::Univariant<i32>::TheOnlyCase{a: -1} | |
| use self::Regular::{Case1, Case2, Case3}; | |
| use self::Univariant::TheOnlyCase; | |
| // NOTE: This is a copy of the non-generic test case. The `Txx` type parameters have to be | |
| // substituted with something of size `xx` bits and the same alignment as an integer type of the | |
| // same size. | |
| // The first element is to ensure proper alignment, irrespective of the machines word size. Since | |
| // the size of the discriminant value is machine dependent, this has be taken into account when | |
| // datatype layout should be predictable as in this case. | |
| enum Regular<T16, T32, T64> { | |
| Case1 { a: T64, b: T16, c: T16, d: T16, e: T16}, | |
| Case2 { a: T64, b: T32, c: T32}, | |
| Case3 { a: T64, b: T64 } | |
| } | |
| enum Univariant<T> { | |
| TheOnlyCase { a: T } | |
| } | |
| fn main() { | |
| // In order to avoid endianness trouble all of the following test values consist of a single | |
| // repeated byte. This way each interpretation of the union should look the same, no matter if | |
| // this is a big or little endian machine. | |
| // 0b0111110001111100011111000111110001111100011111000111110001111100 = 8970181431921507452 | |
| // 0b01111100011111000111110001111100 = 2088533116 | |
| // 0b0111110001111100 = 31868 | |
| // 0b01111100 = 124 | |
| let case1: Regular<u16, u32, i64> = Case1 { a: 0, b: 31868, c: 31868, d: 31868, e: 31868 }; | |
| // 0b0001000100010001000100010001000100010001000100010001000100010001 = 1229782938247303441 | |
| // 0b00010001000100010001000100010001 = 286331153 | |
| // 0b0001000100010001 = 4369 | |
| // 0b00010001 = 17 | |
| let case2: Regular<i16, u32, i64> = Case2 { a: 0, b: 286331153, c: 286331153 }; | |
| // 0b0101100101011001010110010101100101011001010110010101100101011001 = 6438275382588823897 | |
| // 0b01011001010110010101100101011001 = 1499027801 | |
| // 0b0101100101011001 = 22873 | |
| // 0b01011001 = 89 | |
| let case3: Regular<u16, i32, u64> = Case3 { a: 0, b: 6438275382588823897 }; | |
| let univariant = TheOnlyCase { a: -1 }; | |
| zzz(); // #break | |
| } | |
| fn zzz() {()} |