Incorrect code generation when inlining is off

[mbauman]

I don't know how to better describe (or search) for this:

$ julia-0.4 --inline=no
               _
   _       _ _(_)_     |  A fresh approach to technical computing
  (_)     | (_) (_)    |  Documentation: http://docs.julialang.org
   _ _   _| |_  __ _   |  Type "help()" for help.
  | | | | | | |/ _` |  |
  | | |_| | | | (_| |  |  Version 0.4.0-dev+3216 (2015-02-08 17:14 UTC)
 _/ |\__'_|_|_|\__'_|  |  Commit 470cacb* (0 days old master)
|__/                   |  x86_64-apple-darwin13.4.0

julia> type Foo{N}
         t::NTuple{N,Int}
       end

julia> f(x) = !isempty(x.t) && x.t[1]
f (generic function with 1 method)

julia> f(Foo(()))
ERROR: BoundsError: attempt to access ()
  at index [1]
 in getindex at /Users/mbauman/Code/julia-0.4/usr/lib/julia/sys.dylib
 in f at none:1
 in eval at no file

The call f(Foo(())) properly returns false when inlining is enabled.

[simonster]

maybe a dup of #9536

[mbauman]

Ah, yup, seems to be. Isaiah's speculative fixes-9536-but-breaks-other-things patch in #9536 (comment) fixes this. Thanks Simon.