[llvm-commits] Control reaches the end of non-void function

[Chris Lattner]

On Mar 27, 2008, at 4:35 PM, David Chisnall wrote:

> Some versions of GCC can't tell that an assert that always fails  
> counts as a return path, so emits warnings. I've enclosed a diff  
> that fixes this in all of the places it currently appears.

Applied, thanks!
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20080324/060415.html

-Chris