http://llvm.org/bugs/show_bug.cgi?id=14851

Duncan Sands <[email protected]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
                 CC|                            |[email protected]
         Resolution|INVALID                     |

--- Comment #5 from Duncan Sands <[email protected]> 2013-01-08 22:26:06 CST ---
I agree it is OK from the C/C++ point of view, but it is wrong per the current
IR semantics for inbounds getelementptr.  Inbounds is about staying inside
allocated objects (the struct S in this case), not about staying inside fields
of an allocated object (the array foo).  Personally I think this is kind of
silly, but that's just the way it is right now.  The same issue came up for
allocas in Chandler's SROA rewrite.

-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
_______________________________________________
LLVMbugs mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs

Reply via email to