internal compiler error: incorrect size with sdl::video::PixelFormat

[robn]

Possible (probable) duplicate of #13186.

Compiling the following program against with rust@30fe550 against brson/rust-sdl@804adf6:

extern crate sdl;

static my_pixfmt: sdl::video::PixelFormat = sdl::video::PixelFormat {
    palette: None,
    bpp: 24,
    r_loss: 0,
    g_loss: 0,
    b_loss: 0,
    a_loss: 0,
    r_shift: 16,
    g_shift: 8,
    b_shift: 0,
    a_shift: 0,
    r_mask: 0xff,
    g_mask: 0xff00,
    b_mask: 0xff0000,
    a_mask: 0,
    color_key: 0,
    alpha: 0,
};

fn main() {
    print!("{:?}", my_pixfmt);
}

gives:

$ RUST_BACKTRACE=1 rustc -L rust-sdl -o test test.rs
{ { i8, [15 x i8] }, i8, i8, i8, i8, i8, i8, i8, i8, i8, i32, i32, i32, i32, i32, i8 } { { i8, [15 x i8] } { i8 0, [15 x i8] undef }, i8 24, i8 0, i8 0, i8 0, i8 0, i8 16, i8 8, i8 0, i8 0, i32 255, i32 65280, i32 16711680, i32 0, i32 0, i8 0 }
{ { i8, [7 x i8], [1 x i64] }, i8, i8, i8, i8, i8, i8, i8, i8, i8, i32, i32, i32, i32, i32, i8 } undef
error: internal compiler error: const expr(8: sdl::video::PixelFormat{palette: None,
                        bpp: 24,
                        r_loss: 0,
                        g_loss: 0,
                        b_loss: 0,
                        a_loss: 0,
                        r_shift: 16,
                        g_shift: 8,
                        b_shift: 0,
                        a_shift: 0,
                        r_mask: 255,
                        g_mask: 65280,
                        b_mask: 16711680,
                        a_mask: 0,
                        color_key: 0,
                        alpha: 0,}) of type sdl::video::PixelFormat has size 52 instead of 56
note: the compiler hit an unexpected failure path. this is a bug.
note: we would appreciate a bug report: http://static.rust-lang.org/doc/master/complement-bugreport.html
note: run with `RUST_BACKTRACE=1` for a backtrace
task 'rustc' failed at '~Any', /home/robn/code/rust/rust/src/libsyntax/diagnostic.rs:155
stack backtrace:
   1:     0x7f46a1812930 - rt::backtrace::imp::write::h726c0ee05b0a1251rta::v0.11.pre
   2:     0x7f46a1772370 - rt::unwind::begin_unwind_inner::h9b7e455dab1796daw39::v0.11.pre
   3:     0x7f46a0c39430 - rt::unwind::begin_unwind::h5910056841063238677::v0.11.pre
   4:     0x7f46a0c3a490 - diagnostic::Handler::bug::h0f9bbab1cbb7e822AYb::v0.11.pre
   5:     0x7f46a2429510 - driver::session::Session::bug::h6e55d6e6cbf0d844bhh::v0.11.pre
   6:     0x7f46a24db640 - middle::trans::consts::const_expr::ha281951f96861546GFj::v0.11.pre
   7:     0x7f46a2426850 - middle::trans::base::get_item_val::h4b8a8b57d5346ddbBNp::v0.11.pre
   8:     0x7f46a24dd3c0 - middle::trans::consts::trans_const::h6abb7670a66f1aecstk::v0.11.pre
   9:     0x7f46a2425440 - middle::trans::base::trans_item::h65d6980b59f31439Btp::v0.11.pre
  10:     0x7f46a250d2c0 - middle::trans::base::trans_mod::hdd24677d9163c30cHyp::v0.11.pre
  11:     0x7f46a25151a0 - middle::trans::base::trans_crate::h0920ca7f9115adc25dq::v0.11.pre
  12:     0x7f46a2c404a0 - driver::driver::phase_4_translate_to_llvm::hb81995af2fcfe7361Cf::v0.11.pre
  13:     0x7f46a2c42a90 - driver::driver::compile_input::hfaa2edba3ed249a4QSf::v0.11.pre
  14:     0x7f46a2c67490 - run_compiler::h0c58ff7ffab7f82efpn::v0.11.pre
  15:     0x7f46a2c7f210 - main_args::closure.91455
  16:     0x7f46a2c7d710 - monitor::closure.91330
  17:     0x7f46a2c79000 - task::TaskBuilder::try::closure.91096
  18:     0x7f46a1e9d650 - task::spawn_opts::closure.7103
  19:     0x7f46a180d970 - rt::task::Task::run::closure.39989
  20:     0x7f46a18196b0 - rust_try
  21:     0x7f46a180d7b0 - rt::task::Task::run::hc79a6e8202de0a541T7::v0.11.pre
  22:     0x7f46a1e9d420 - task::spawn_opts::closure.7076
  23:     0x7f46a1811470 - rt::thread::thread_start::h915d8ac90bd84ba1My8::v0.11.pre
  24:     0x7f469f52ffa0 - start_thread
  25:     0x7f46a1445a09 - __clone
  26:                0x0 - <unknown>

Unfortunately I was unable to reduce this to a case that doesn't require rust-sdl. The PixelFormat structure is a bit too complex for me to know which bits are important for this case.

Happy to help reduce it further if I can. Let me know.

[ghost]

Duplicate of #13186. Closing.