[llvm-commits] [llvm] r40128 - /llvm/trunk/test/CFrontend/2004-02-13-Memset.c

[Chris Lattner]

On Jul 22, 2007, at 3:40 AM, Duncan Sands wrote:

>>>> If so, we should fix TreeToLLVM::EmitBuiltinMemCopy to treat a null
>>>> return as alignment = 1.
>>>
>>> A 0 return could also mean that the argument wasn't a pointer - I
>>> will check if this is problematic or not.
>>
>> Okay, we could just do an explicit check before that to decide if the
>> argument is a pointer.  If not, bail out before calling
>> get_pointer_alignment.  Then, after the get_pointer_alignment call is
>> made, if it returns zero, just set the alignment to 1.
>
> Sure!  But in fact the argument was already known to be a pointer
> in all cases.  I've committed a fix.

Nice!  Does it make sense to change the dejagnu test back to -O0?

-Chris