rustdoc: panic caused by a "reexport" of a type alias

[gustavla]

Code that compiles fine with rustc is causing an internal compiler error: unexpected panic with rustdoc.

I tried this code (bug.rs):

type Foo = isize;
pub use self::Foo::*;

The second line is essentially a typo, which can happen if Foo used to be an Enum for instance.

Again, the following works fine (although I'm not sure why, can you even add anything to the Foo namespace?):

rustc bug.rs --crate-type lib

However, if I run:

rustdoc bug.rs

I get the error:

error: internal compiler error: unexpected panic
note: the compiler unexpectedly panicked. this is a bug.
note: we would appreciate a bug report: https://github.com/rust-lang/rust/blob/master/CONTRIBUTING.md#bug-reports
thread 'rustc' panicked at 'glob not mapped to a module or enum', ../src/librustdoc/visit_ast.rs:246

This actually happened on a version of a package already pushed to crates.io (hdf5-sys 0.3.0), so anything that depends on the erroneous version and runs cargo doc will see this error.

Meta

Run on rustdoc 1.7.0-nightly (4ce1dafd1 2015-12-23).

Backtrace:

   1:        0x10767a558 - sys::backtrace::tracing::imp::write::h437867c953ecf350mZt
   2:        0x10768060f - panicking::log_panic::_<closure>::closure.41684
   3:        0x1076800b4 - panicking::log_panic::h11dc8b3c6952cc98VRx
   4:        0x107648006 - sys_common::unwind::begin_unwind_inner::haa584e089eabb124o2s
   5:        0x102f80b34 - sys_common::unwind::begin_unwind::begin_unwind::h18100502266587524809
   6:        0x103164cd3 - visit_ast::RustdocVisitor<'a, 'tcx>::resolve_id::h34d767a94ffda832qSt
   7:        0x1031629d6 - visit_ast::RustdocVisitor<'a, 'tcx>::visit_item::hec3b049092d8fa8cEVt
   8:        0x10315ba12 - visit_ast::RustdocVisitor<'a, 'tcx>::visit_mod_contents::h7551e6f2d22647cc1Ot
   9:        0x1030ef4ef - visit_ast::RustdocVisitor<'a, 'tcx>::visit::ha59cf730f9a0c027FHt
  10:        0x1030e8f12 - rustc_driver::driver::phase_3_run_analysis_passes::_<closure>::closure.42399
  11:        0x1030c5d27 - middle::ty::context::ctxt<'tcx>::create_and_enter::create_and_enter::h13379830180998392343
  12:        0x1030c2231 - driver::phase_3_run_analysis_passes::phase_3_run_analysis_passes::h13354236957594875836
  13:        0x102fbebb0 - core::run_core::h09e73f61b678853aZdk
  14:        0x102fba240 - sys_common::unwind::try::try_fn::try_fn::h13762566296504504003
  15:        0x107678088 - __rust_try
  16:        0x10766fd8e - sys_common::unwind::try::inner_try::hf68fe5ebaab44cceWYs
  17:        0x102fbbf01 - boxed::F.FnBox<A>::call_box::call_box::h3532894196117764030
  18:        0x10767f6ad - sys::thread::Thread::new::thread_start::h1257bf86ea35bf0aZbx
  19:     0x7fff8f30e267 - _pthread_body
  20:     0x7fff8f30e1e4 - _pthread_start

[mitaa]

This is easy to fix in rustdoc, but should this even compile?

[alexcrichton]

Whoa kinda crazy...

cc @jseyfried (resolve weridness, longstanding though!)